Scientific Papers 16 2010 University ědeckých prací ... · solid and high sugar content food have been reported [16-20]. Comfrey (Symphytum officinale L.) belongs to the borage

University of Pardubice

Sborník vědeckých pracíUniverzity Pardubice

Faculty of Chemical Technology

Fakulta chemicko-technologická

Scientific Papers 16 2010

series A

Contents

Červenka L., Kubínová J., Juszczak L., Witzak T.:Study of Moisture Adsorption Process in Comfrey(Symphytum officinale L.) Roots at 25 °C .................................... 5

Guzsvány V., Nakajima H., Soh N., Nakano K., Švancara I.,Imato T.: Applicability of Antimony Film Electrodesin Anodic Stripping Voltammetry Combined withSequential Injection Analysis ..................................................... 19

Wassel A.A., Abdullatif S.: New Type of PVC-Membrane Ion-Selective Electrodes and Their Applicability toDetermine Some Antidepressant Drugs ..................................... 33

El-Kabbany F., Hassan S.T., Hafez M.: Infrared SpectroscopicAnalysis of Polymorphism in Diphenyl Carbazide .................... 57

Janíèek P., Drašar È., Krejèová A., Loš�ák P., Navrátil J.:Optical Properties of Tl-Doped Bi2Se3 Single Crystals ............. 75

Kanïár R., Žáková P., Fröhlichová M., Štramová X., Bencová S.:Determination of Coproporphyrin I and III in HumanUrine Using HPLC with Fluorescence Detection ....................... 87

Kanïár R., Žáková P.: Simultaneous Determination ofVanillylmandelic and Homovanillic Acids inUrine Using LC with Coulometric ElectrochemicalDetection Based on SPE Sample Preparation ............................ 99

Lišková V., Pinto T.A.D., Burgert L., Hrdina R.: Use of NewReactive Dyes for Dying of Polyamide 6 and Wool ................ 111

Matìjková E., Kalendová A., Obrdlík Š.: Preparation ofSurface Modified Kaolines by Silicone Hydro-phobization ................................................................................ 133

Obrdlík Š., Kalendová A., Matìjková E.: AnticorrosionProperties of Ion-Exchangable Pigments inEpoxide Coatings ...................................................................... 145

Chýlková J., Šelešovská R., Stuchlík M., Machalíková J.:Contamination of Atmosphere with VolatileHydrocarbons During Handling Oil Productsof Petrol and Diesel Fuel Types ................................................ 153

Bednaøíková M., Linhartová M., Hyršlová J.: AspectsConcerning Attractiveness of Company as Employer .............. 167

Branská L., Šilhavá K.: Increasing of Equipment Reliabilityby RCM Application ................................................................. 179

Franc P., Tetøevová L.: Cost Benefit Analysis and the Social

Time Preference Rate to Estimate Social DiscountRate in the Czech Republic ....................................................... 191

Hyršlová J., Bednaøíková M., Tesnerová P.: Quality CostsMonitoring — Pilot Project in Lasselsberger’sProduction Plant ........................................................................ 203

Paták M., Vlèková V.: Time Series Forecasting as a Toolof Operative Management ........................................................ 219

Sabolová V.: Triple Helix Model in the Czech Tertiary Education ............. 229

1 To whom correspondence should be addressed.

5

SCIENTIFIC PAPERSOF THE UNIVERSITY OF PARDUBICE

Series AFaculty of Chemical Technology

16 (2010)

STUDY OF MOISTURE ADSORPTION PROCESSIN COMFREY (SYMPHYTUM

OFFICINALE L.) ROOTS AT 25 °C

Libor ČERVENKAa1, Jana KUBÍNOVÁa, Lesław JUSZCZAKb and Teresa WITZAKc

aDepartment of Analytical Chemistry,The University of Pardubice, CZ–532 10 Pardubice,

bDepartment of Analysis and Evaluation of Food Quality,University of Agriculture, PL–30 149 Krakow

Received September 13, 2010

Sorption isotherms of comfrey (Symphytum officinale L.) root samples withdifferent particle size were obtained at 25 °C. The shape of isotherm was similarto those of high-sugar-content foods and the particle size did not affect adsorptionprocess in the aw range used in this study. Blahovec–Yanniotis model wasconsidered to give the best fit over the whole range of aw tested. Variousparameters describing the properties of sorbed water derived from GAB andBlahovec–Yanniotis models have been discussed. DSC method was used tomeasure the glass transition temperatures (Tg) of root samples in relation to wateractivity. The safe moisture content was determined in 13.39 g/100g in dried basisat 25 °C. Combining of the Tg line with sorption isotherm in one plot showed that

6 Červenka L. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 5–18

the glass transition temperature concept overestimated the temperature stabilityfor both root samples.

Introduction

The knowledge of the moisture content and water activity (aw) relationships offood products is very important. These can be estimated by determining ofmoisture sorption isotherm that can be used for defining storage conditions and inmaking shelf life determination [1]. Unfavorable storage conditions may lead todeterioration of the product caused both by native enzymes and bymicroorganisms’ activity [2]. A large amount of data has been publishedconsidering the sorption behavior of foods and food products including meat [3],cereal products [4] or fruit and vegetables [5,6]. Medicinal herbs and spices usedin the folk medicine are also in the scope of the interest [7-9].

Many empirical relations describing the sorption characteristics of foods andfood ingredients have been proposed in the literature [10]. The GAB(Guggnheim–Anderson–de Boer) equation is a general model for predicting thesorption isotherm for most products including herbs [8,11]. The monolayermoisture content, a key parameter corresponding to the physical and chemicalstability of foods, can be derived from this model. For the purpose of optimizingthe storage condition of various products, the sorption model with parameterswhich can be easily applied to the practice is a better choice. This situationrepresents the GAB equation with its monolayer moisture content. Similarequations have been proposed including Blahovec–Yanniotis model [12] andCaurie models [13]. Blahovec–Yanniotis equation allows distinguishing betweenthe amount of water bonded to the sorption site of non-soluble solids and theamount of water presents in aqueous solution. Caurie [13] proposed a model whichresulted in various parameters such as monolayer value, density of sorbed water,surface area of sorption and the number of adsorbed monolayers. All theseparameters may be helpful in understanding the moisture behavior and thus settingappropriate storage conditions.

In the recent years, the concepts related to the aw have been enriched bythose of glass transition temperature (Tg). It is defined as the temperature at whichthe material changes from the glassy to the rubbery state for a given heating rate.The food biopolymers are stable at glassy state where rotational mobility ofmolecules is restricted while changes may occur at or above Tg [14]. Despite thefact that Tg and sorption isotherms represent two different criteria, Sablani [15]proposed both concepts to be used for determination of food stability. Tg is productspecific and is a function of moisture content (or water activity) of material. Inaddition, they concluded the both concepts need a substantial revision since thereis a discrepancy in setting the so-told safe temperature for storage of various food

Červenka L. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 5–18 7

products. Glass transition temperature as a function of moisture content for somesolid and high sugar content food have been reported [16-20].

Comfrey (Symphytum officinale L.) belongs to the borage family, andhistorically it was used to treat gastrointestinal disorders. In 2001, however, theU.S. Food and Drug Administration banned the sale of oral products containingcomfrey because of the content of pyrrolizidine alkaloids [21]. Comfrey has beenalso studied for its anti-inflammatory [22] and antifungal activity [23]. Recentmedical studies revealed that comfrey root extracts have pronounced analgesic andanti-exudative properties and may serve as a therapeutic agent in the treatment offacture, ankle distortion or osteoarthritis [24,25]. Comfrey also contains polymericsubstances exhibiting antioxidant activity against active oxygen species withoutcytotoxic side effects [26,27].

In today’s market, the herbs can be found in many types of preparations, forinstance syrups, infusions or herbal tea. Dry pieces of herb are usually stored atambient temperature without regard to relative humidity. Since dried herbs havehigh prevalence of moulds, yeasts [28] or coliforms [29], it is necessary to assessappropriate storage conditions to prevent growing of, or toxin productionaccompanying microflora.

The purpose of this study was to determine the relationship betweenmoisture content, water activity and glass transition temperature of comfrey rootsamples at the temperature of 25 °C and thus obtain the critical moisture contentor water activity level for safe storage.

Materials and Methods

Sample Preparation

Chopped and pre-dried pieces of roots (approximate size 1.0×0.5 cm) ofSymphytum officinale L. were purchased in local supply. The approximatecomposition was determined according to AOAC procedures [30]: moisturecontent (method 920.36), ash content (method 942.05), ether extract (method932.02) and crude protein (method 2001.11). The total carbohydrate content wascalculated from the difference. All the measurements were done in triplicate in twoconsecutive trials. The results are summarized in Table I.

For the sorption study, the samples were pre-dried in forced-air oven at thetemperature of 50 °C for 24 hours, and then the roots were manually ground in amortar and sieved to give a fine particle size of # 1.0 mm. The particles greaterthan 1.0 mm (in the range of 1.0-5.0 mm) were also used in this study. Both typesof samples were subsequently dried over silica gel to constant weight.


Table I Approximate chemical composition of comfrey (Symphytum officinalis L.) (in g/100g dry basis, average ±SD, six replicates)

Comfrey

Ash 10.48 ± 0.22

Crude protein (nitrogen content ×6.25) 5.22 ± 0.95

Ether extract 0.57 ± 0.16

Total carbohydratea 70.83aCalculated from differences

Determination of Moisture Adsorption Isotherm

For the adsorption studies, following salt slurries were used according to Stolloff[31]: LiCl, CH3COOK, MgCl2, K2CO3, Mg(NO3)2, CoCl2, NaCl, KCl and KNO3.The moisture adsorption isotherms were determined gravimetrically by exposingroot samples (approximately 1.0 g) in aluminum pans to different relative humiditygenerated by salt slurries in the approximate range of 0.12 to 0.93 aw. A fewcrystals of thymol were placed in desiccators with relative humidity higher than75 % to prevent mould growth. After 2-3 weeks, the samples were graduallywithdrawn and the true equilibrated water activity at 25 °C was measured inNovasina instrument. After measurement, the sample was immediately placed inforced-air oven and the appropriate moisture content was determined accordingto AOAC procedure (method 920.36).Each sample was weighed by means ofanalytical balance (sensitivity ±0.001 g). The results were expressed in g per 100gof dry basis (d. b.). The isotherms were obtained by plotting the moisture contentversus aw. Each isotherm was constructed using data of three replicates.

Isotherm Models

The experimental sorption data of two samples at 25 °C were fitted to threesorption equations, namely Guggenheim–Anderson–deBoer (GAB),Blahovec–Yanniotis (B–Y) and Caurie’s equation. The GAB equation is the mostversatile for various food products and biomaterials [6] and was used in thefollowing form

(1)


where aw is the water activity, M is the experimental moisture contents, M0 is themonolayer moisture contents on dry basis (g/100g), C is the constant related to theheat of sorption.

Blahovec–Yanniotis [12] introduced a 4-parameter model based on theassumption that a part of the water is sorbed on non-soluble solids forming amonolayer while the rest of the water is available as solvent for soluble solids. Theequation was expressed in the form

(2)

where aw is the water activity, M is the experimental moisture content, a1, a2, b1,b2 are constants. The first term in Eq. (2) represents the amount of water bondedto the sorption sites of the non-soluble solids and the second term represents theamount of the water in an aqueous solution.

The parameters of the equations (Eqs. (1) and (2)) were estimated usingnon-linear regression technique (Gauss–Newton) in QCExpertTM v 3.0 (TriloBytes.r.o., Pardubice, the Czech Republic). The goodness of fit was evaluated with themean relative percentage deviation (%E) using the formula

(3)

where Me are experimental and Mp are predicted moisture contents. A model isconsidered acceptable if the %E values are below 10 % [32]. The sign test ofresidual values was used as an additional indicator of the model’s quality. Thefollowing testing model was applied

(4)

where ns and nt are the numbers of sequences of the same sign for experimentaland theoretical residuals, Dt is error of theoretical sign sequences, C is correctionfor continuity. If the sign test reveals a trend in residual values, the model shouldnot be accepted [33].

Caurie’s equation was applied in its linear form previously published byCaurie [13]

(5)


where aw is the water activity, M and Mc are the experimental and monolayermoisture contents on dry basis (g/100g), respectively. D is the constant related tothe density of sorbed water (g ml–1). Monolayer values (Mc) were calculated usingCaurie’s plot of ln(1–aw)/aw versus ln(1/M) in the aw range from 0.12 to 0.93. Thesurface area (A, m2 g–1) and the number of adsorbed monolayers (N) werecalculated according to Caurie [13] using the formulae

(6)

(7)

where S is the slope of Caurie’s plot. The constants of Caurie’s equations were computed by the least squares

method. The statistical differences were calculated using analysis of variance(ANOVA) at the probability level of p = 0.05.

Glass Transition Temperature

The samples were adjusted to various water activities by adsorption process asdescribed above. After the measurement of water activity, the glass transitiontemperature was immediately measured using differential scanning calorimeter(DSC 204 F1 Phoenix, Netzsch, Germany) equipped with an intracoolerprecalibrated with indium. An amount of about 8.5 mg was hermetically sealed inaluminum pans. Each sample was cooled to the temperature of –100 °C at the rateof 10 °C min–1 and the scanning was performed by heating at the same rate from–100 °C to 100 °C with the empty aluminum pan as a reference. The glasstransition temperature was determined from the midpoint of the heat capacitychange observed on the second run.

The glass transition temperature and moisture content relationship wasmodeled using Gordon–Taylor equation [34]

(8)

where Xs and Xw are the mass fractions of solid and water, respectively. Tg, Tgs, Tgw

are the glass transition temperature of mixture, solids and water, respectively. Tgw

= 138 K, h is the Gordon–Taylor parameter.


Results and Discussion

Moisture Adsorption Isotherms

The results of the experimental measurements of the equilibrium moisture contentsof root samples at 25 °C during adsorption are given in Fig. 1.

Fig. 1 Adsorption isotherm of comfrey (Symphytum officinale L.) at 25 °C

The shape of the isotherms is characteristic of high-sugar-containing foodsadsorbing relatively small amount of water at low water activities and largeamounts at high water activities, particularly above 0.60aw. The similar shape wasdetermined in other plant parts such as eucalyptus leaves [8], miscanthus stemsand leaves [35] or sufflower and tarragon [36]. The effect of particle size ofsamples did not affect the adsorption process in the aw range used in this study (p> 0.05). As Strange and Onwulata [37] stated the effect of particle size on themoisture adsorption should not be generalized. Whereas oat fiber adsorbed morewater with decreasing the particle size, wheat bran adsorbed less water and cornbran did not correlate with particle size in their study. On the other hand,Mathlouthi and Roge [38] found that moisture adsorption increased withdecreasing of sucrose crystal size distribution.

Sorption Isotherm Models

Two isotherm equations (Eqs (1) and (2)) were used for establishing the degree offit to the experimental data. Estimated parameters for these equations are presentedin Table II.


Table II Estimated parameters for selected models of adsorption isotherm equations forcomfrey (Symphytum officinale L.) at 25 °C

Parameters of isotherm Comfrey

root particles

#1.0 mm 1.0-5.0 mm

GAB (Eq. (1))

M0 5.861 6.563

C 264.8 17.25

K 0.974 0.946

%E 5.211 3.876

Sg 3.287 1.172

p 0.001 0.241

pattern random

Blahovec–Yanniotis (Eq. (2))

a1 0.015 0.016

b1 0.091 0.164

a2 0.272 0.142

b2 0.278 0.131

%E 3.751 2.674

Sg 1.401 0.788

p 0.161 0.431

pattern random

Caurie (Eq. (4))

Mc 7.223 6.949

D 1.805 1.795

A 108.9 105.4

N 4.000 3.871

R2 0.974 0.974

M0 – monolayer moisture content (g/100 g d. b.); Mc – bonded water (g/100 g d. b.); D – densityof sorbed water (g ml–1); A – surface of sorption (m2 g–1); N – number of adsorbed monolayers;%E – mean relative percentage deviation; Sg – sign test of residuals; p – probability level; C, K,a1, a2, b1, b2 – parameters of adsorption models, R2 – coefficient of determination


Both GAB and B–Y models fitted well the experimental data with %E <4.00. However, the sign test showed that the GAB equation was not acceptable forfitting the experimental data of comfrey root samples with particle size # 1.0 mm.Blahovec–Yanniotis equation well predicted the experimental values in both rootsamples (Table II). This equation (Eq. (2)) well described the sorption data ofmany food products including starchy and high protein foods, fruits andvegetables, nuts, legumes, and seeds [39]. The advantage of this model over theGAB equation in fitting the experimental data is mainly in the region of the highwater activity. Therefore, we concluded that this model better described thesorption data of comfrey and soapwort root samples in this study.

Properties of Adsorbed Water

The physical state of water in foods determines their stability during storage [2].Therefore it is suitable to generate the information related to various aspects ofwater in food. The monolayer moisture contents of both root samples estimatedfrom the GAB equation ranged from 5.8-6.6 g/100g d. b. for comfrey. Themonolayer values for root samples with particle size # 1.0 mm were significantlylower (p < 0.05) in comparison with the particle size greater than 1.0 mm. Blaho-vec and Yanniotis presented a different approach to moisture sorption behavior infood systems [8,39]. Using their mathematical model, one can analyze thecontribution of tightly adsorbed water and solution water in the sample. It is clearfrom Fig. 2 that the contribution of solution water is small at low water activitiesbut at higher aw values (> 0.75) the contribution is significant for all the samples.

Fig. 2 Contribution of surface adsorption term and solution term in adsorption isothermof comfrey (Symphytum officinale L.) Root samples at 25 °C according toBlahovec–Yanniotis equation (Eq. (2))


Although the overall adsorption isotherms (Fig. 1) did not show anydifference between fine particles and those of size of 1.0-5.0 mm, some trend wasobserved by applying two terms of Blahovec–Yanniotis equation (Eq. (2)). Theroot samples ground into fine particles (# 1.0 mm) tightly adsorbed higher amountof water in comparison with root particle of 1.0-5.0 mm size (Fig. 2).Consequently, free or solution water became dominate above 0.70 and 0.60aw forroot particles # 1.0 mm and 1.0-5.0 mm, respectively.

It was described that in the range of 0.2-0.75aw the water interacts primarilywith polar surface groups [40]. The grinding of sample into fine particles mayexpose more hydrophilic functional groups, therefore the root sample withparticles # 1.0 mm adsorbed a higher amount of water.

The parameters derived from linear regression of Caurie’s equation (Eq. (5))and computed parameters according to Eqs (6) and (7) are presented in Table II.Bonded or non-freezable water is similar to those obtained in GAB models,density of sorbed water ranged from 1.79 to 1.81 g ml–1 for all the samples, surfacearea of sorption was greater for fine particles (# 1.00 mm) than for particles > 1.00mm. This parameter supports our previous finding that root samples with fineparticles adsorbed more water as compared to particles > 1.00 mm.

Glass Transition Temperature

The effect of moisture content on glass transition temperature is shown in Fig. 3.

Fig. 3 Gordon–Taylor model (Eq. (8)) to predict Tg as function of water activity forcomfrey (Symphytum officinale L.)


Glass transition temperature was found to vary with moisture content for both rootsamples. The decrease with an increase in moisture content was reported in manysugar rich foods such as jaggery granules [17], spray dried açai juice [20],raspberry [19], cassava starch [16] or rice [18]. The Gordon–Taylor equation hasbeen successfully used for prediction of Tg for both root samples. The parametersestimated by nonlinear regression are shown in Table III. This table also presentsmean relative percentage deviation (%E) showing that Gordon–Taylor modelreasonably predicts Tgs.

Table III Values used in prediction of Tg for Gordon–Taylor equation (Eq. (7))

Parameter Comfrey

(Symphytum officinale L.)

Root particles #1.0 mm

Tgs, K 341.0

k 0.210

%E 40237

Fig. 4 Water sorption isotherms and glass transition temperatures of comfrey(Symphytum officinale L.) as functions of water activity

Combining the glass transition temperature concept with sorption isothermis a useful tool for estimation of critical values for aw and moisture content.According to Roos [41], such critical values are defined as those decreasing theTg to ambient temperature. For estimation of critical moisture content for comfreyand soapwort root samples, the adsorption isotherm using Blahovec–Yanniotisequation and the plot of Tg vs. aw was constructed (Fig. 4). The critical moisturecontents at the temperature of 25 °C are evaluated to be 13.39g/100g d. b. for com-


Table IV Comparison of water sorption isotherm and glass transition concepts for comfrey(Symphytum officinale L.) root samples (particles #1.0 mm) at the temperature of25 °C

Sorption isotherm model Glass transition model

M0 (Eq. (1))g/100 g d.b.

aw

correspondingto monolayerwater content

Tg from glasstransition

model°C

Tg°C

moisturecontent

g/100 g d.b.

aw atcorrespondingwater activity

5.86 0.144 60.93 40202 13.39 0.561

frey root samples. The corresponding critical aw value was 0.561 (Table IV). This means that

when the comfrey root sample is stored at the temperature of25 °C, the maximumrelative humidity which it can be exposed to is 56.1 % and its moisture content is13.39 g/100g d. b. It was previously published that there is a surprisingdiscrepancy between the predictions of stable temperature using sorption isothermand glass transition concepts [15]. For example, Saymaladevi et al. [19] found thatglass transition concept underestimated the stability temperatures for freeze-driedraspberries. The opposite effect was found in our work, where the sorptionisotherm model increased the critical temperature (Tg) to 60.93 for comfrey. On theother hand, the study of Zimeri and Kokini [42] supported the results obtained inthis study although they did not construct the combined plot of sorption isothermand glass transition line. They found that one type of native inulin (Raftiline) hadmonolayer moisture content 7.22 % at 25 °C and the corresponding Tg can beestimated in the range of 60-70 °C. Moreover, they also concluded that glasstransition temperature of inulin depended on the crystallinity, which is the functionof moisture content. The detailed study on physicochemical changes ofpolysaccharides or phenolic compounds at a range of water contents may betterelucidate the suitability of moisture sorption isotherm or glass transitiontemperature concepts to predict stability of root samples.

Conclusion

Blahovec–Yanniotis fits well to the experimental data for comfrey root samples.The particle size did not affect the overall adsorption properties of root samples;however, using adsorption and solution term of Blahovec–Yanniotis equation, theparticle size affected the sorption behavior. Due to higher surface area of sorptionand probably the higher content of polar groups, the fine particles of root samplesadsorbed more water. The glass transition was determined in 13.39 gmoisture/100g d. b. for comfrey. The combination of Gordon–Taylor and GABequations into one plot revealed that there is a discrepancy between predicting the


Tg values. The glass transition temperature concept overestimated the temperaturestability for both root samples.

Acknowledgements

Financial help for this project was provided by Ministry of Education, Youth andSports of the Czech Republic (project No. 0021627502) and the Czech ScienceFoundation (GA203/08/1536).

References

[1] Mathlouthi M.: Food Control 12, 409 (2001).[2] Ross Y.H.: Water Acitivity and Plasticization in: Food Shelf Life Stability

(N.A.M. Eskin, D.S. Robinson, Eds), pp 4-36, CRC Press, Boca Raton, 2001.[3] Singh R.R.B., Rao K.H., Anjaneyulu A.S.R., Patil G.R.: Food. Res. Int. 34,

143 (2001).[4] Turhan M., Oymael B., Ekiz H.I.: J. Food Eng. 57, 269 (2003).[5] Falade K.O., Adetunji A.I., Aworh O.C.: Eur. Food Res. Technol. 217, 230

(2003).[6] Samaniego-Esguerra C.M., Boag I.F., Robertson G.L.: J. Food Eng. 33, 115

(1991).[7] Kouhila M., Belghit A., Daguenet M., Boutaleb B.C.: J. Food Eng. 47, 281

(2001).[8] Kouhila M., Kechaou N., Otmani M., Fliyou M., Lahsasni S.: Dry Technol.

20, 2027 (2002).[9] Soysal Y., Oztekin S.: J. Agr. Eng. Res. 78, 57 (2001).[10] Al-Muhtaseb A.H., McMinn W.A.M., Magee T.R.A.: Food Bioprod.

Process. 80, 118 (2002).[11] Cervenka L.: J. Food Nutr. Res. 47, 131 (2008).[12] Blahovec J., Yanniotis S.: J. Food Eng. 91, 72 (2009).[13] Caurie M.: Derivation of Full Range Moisture Isotherms in: Water Activity:

Influences on Food Quality (L.B. Rockland, G.F. Stewert, Eds), pp 63-87,Academic Press, New York, 1981.

[15] Sablani S.S., Kasapis S., Rahman M.S.: J. Food Eng. 78, 266 (2007).[16] Perdomo J., Cova A., Sandoval A.J., Garcia L., Laredo E., Muller A.J.:

Carbohyd. Polym. 76, 305 (2009).[17] Rao P.V.K.J., Das M., Das S.K.: Int. J. Food Sci. Tech. 45, 94 (2010).[18] Sablani S.S., Bruno L., Kasapis S., Symaladevi R.M.: J. Food Eng. 90, 110

(2009).[19] Syamaladevi R.M., Sablani S.S., Tang J.M., Powers J., Swanson B. G.: J.


Food Eng. 91, 460 (2009).[20] Tonon R.V., Baroni A.F., Brabet C., Gibert O., Pallet D., Hubinger M.D.: J.

Food Eng. 94, 215 (2009).[21] Schaneberg B.T., Molyneux R.J., Khan I.A.: Phytochem. Analysis 15, 36

(2004).[22] Tunon H., Olavsdotter C., Bohlin L.: J. Ethnopharmacol. 48, 61 (1995).[23] Kartal M., Kurucu S., Choudary M.I.: Turk. J. Med. Sci. 31, 487 (2001).[24] Koll R., Buhr M., Dieter R., Pabst H., Predel H.G., Petrowicz O., Giannetti

B., Klingenburg S., Staiger C.: Phytomedicine 11, 470 (2004).[25] Grube B., Grunwald J., Krug L., Staiger C.: Phytomedicine 14, 2 (2007).[26] Barbakadze V., van den Berg A.J.J., Beukelman C.J., Kemmink J., van

Ufford H.C.Q.: Chem. Nat. Compd. 45, 6 (2009).[27] Barthomeuf C.M., Debiton E., Barbakadze V.V., Kemertelidze E.P.: J. Agr.

Food Chem. 49, 3942 (2001).[28] Tournas V., Katsoudas E., Miracco E.J.: Int. J. Food Microbiol. 108, 178

(2006).[29] Mpuchane S., Gashe B.A.: J. Food Protect. 59, 28 (1996).[30] AOAC. Official Methods of Analysis of AOAC International, 18th edition, (W.

Horwitz, G.W. Latimer Eds), AOAC International, Gaithersburg 2005.[31] Stoloff L.: J. Assoc. Off. Anal. Chem. 61, 1167 (1978).[32] Palou E., Lopez-Malo A., Argaiz A.: J. Food Eng. 31, 85 (1997).[33] Meloun M., Militky J.: Statistical Treatment of Experimental Data (in

Czech), Plus, Ltd., Prague 1997.[34] Gordon M., Taylor M.S.: J. Appl. Chem. 2, 493 (1952).[35] Arabhosseini A., Huisman W., Muller J.: Biomass Bioenerg. 34, 411 (2010).[36] Kaya S., Kahyaoglu T.: J. Food Eng. 78, 413 (2007).[37] Strange E.D., Onwulata C.I.: J. Food Quality 25, 63 (2002).[38] Mathlouthi M., Roge B.: Food Chem. 82, 61 (2003).[39] Yanniotis S., Blahovec J.: Lwt-Food Sci. Technol. 42, 1688 (2009).[40] Reid D.S., Fennema O.R.: Water and Ice in: Fennema’s Food Chemistry (S.

Damodaran, K.L. Parkin, O.R. Fennema Eds), pp 17-82, CRC Press, BocaRaton, 2008.

[41] Roos Y.H.: J. Food Process. Pres. 16, 433 (1993).[42] Zimeri J.E., Kokini J.L.: Carbohyd. Polym. 48, 299 (2002).


19



16 (2010)

APPLICABILITY OF ANTIMONY FILMELECTRODES IN ANODIC STRIPPING

VOLTAMMETRY COMBINED WITH SEQUENTIALINJECTION ANALYSIS

Valéria GUZSVÁNYa1, Hizuru NAKAJIMAb, Nobuaki SOHb,Koji NAKANOb, Ivan ŠVANCARAc and Toshimiko IMATOb

aDepartment of Chemistry, Biochemistry and Environmental Protection,Faculty of Sciences, RS–21 000 Novi Sad,

bDepartment of Applied Chemistry, Kyushu University, JP–819 0395 Fukuoka,cDepartment of Analytical Chemistry,

The University of Pardubice, CZ–532 10 Pardubice


The possibilities to employ antimony-film plated glassy carbon electrode (SbF-GCE) in anodic stripping voltammetry combined with sequential injection analysis(ASV-SIA) were examined and tested for the determination of Pb(II) and Cd(II)selected as the model ions. It has been found that SbF-GCE can be operated atthe low :g l–1 concentration level via the in-situ preparation — deposition froma plating solution of 0.5 M HCl, mixed with 0.75 mg l–1 Sb(III), and at a potentialof –1.5 V vs. Ag/AgCl. The same solution could then be used for pre-concentrationand stripping, when the reproducibility of the respective signals was characterised

20 Guzsvány V. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 19–31

by means of the RSD < 2.8 %, and the detection limit (3F) of about 1.2 µg l–1

Pb(II) and 1.4 µg l-–1 Cd(II). Furthermore, as found, the presence of weaklycomplexing halide salts, KCl or KBr, could further improve the overall signal-to-noise ratio, whereas the addition of KSCN into the solution(s) offered thepossibility to detect some additional ions; namely, of highly electronegative metalelements such as Zn(II), Mn(II), and Cr(III).

Introduction

In modern instrumental measurements, electrochemical stripping analysis (ESA)with pre-concentration step is still widely used for the determination of heavymetal ions at the trace or even ultratrace level [1]. Besides a number of well knownadvantages, the individual techniques of ESA can also be eventually coupled withflow injection analysis (FIA) or sequential injection analysis (SIA), thus offeringfurther improvement of selectivity and, mainly, the automation of the wholeanalytical procedure [2-4]. Although the latter is also possible in the commonstationary mode [5], both ESA-FIA and ESA-SIA combinations are preferable dueto better availability of the respective instrumentation, as well as more effectiveelectroanalytical performance [1,2].

According to the newest trends such as the “green chemistry concept” [6],the choice of the respective detection units for measurements in flowing streamsusually comprises various types of non-mercury electrodes from which especiallybismuth-based electrodes (BiEs [7-9]) have undergone a remarkable progressduring the last decade [10], including successful applications in both FIA (see,e.g., Refs [11,12]) and SIA [13,14].

Another attractive group of non-mercury electrodes propagated recently isrepresented by antimony electrodes (SbEs), comprising (i) antimony film-platedcarbonaceous substrate electrodes (SbFEs [15,16]), (ii) antimony(III) oxide, (iii)antimony(III) precipitate, or (iv) antimony powder bulk-modified carbon pasteelectrodes (i.e., Sb2O3-CPEs [17], SbOL-CPEs [18], and Sb-CPEs [19]), togetherwith related configurations prepared via screen-printing technologies (e.g., Sb2O3-SPEs [20]). To date, the characterisation and selected applications of antimony-based electrodes were the subject of interest in a dozen of contributions [15-17,19-29]; none of which has dealt so far with their employment in flowing streams.

The first attempts for such a combination have been made by our grouprecently [30-32]; firstly, by employing a bismuth film-plated electrode [30],followed by experiments with the antimony analogue [31] in a motivation to unifyimportant experimental parameters for using both BiEs and SbEs under the sameconditions [32]. Since the previous reports have emphasised some principaldifferences in the behaviour of these electrodes (see, e.g., Refs [15] and [19]), suchuniversal procedure would be a new finding, which is the key topic of the latter

Guzsvány V. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 19–31 21

article. Herein, the attention is paid again to the antimony film electrodeconfiguration and its continuing examination for the determination of typicalheavy metal ions by anodic stripping voltammetry in combination with sequentialinjection analysis (ASV-SIA). Furthermore, this article also presents the results ofa special study focused on the supporting electrolyte composition and the effectof selected salts (namely: KNO3, K2SO4, KCl, KBr, and KSCN) on the pre-concentration and stripping processes and the resultant signal-to-noise ratio.

Experimental

Chemicals and Reagents

All chemicals used were of analytical reagent grade and purchased from Sigma-Aldrich or Merck if not stated otherwise. Stock solutions of HCl and acetate buffer(CH3COOH + CH3COONa, 1:1; AcB) were made 1.0 mol l–1 in concentration.Stock solutions of Sb(III) used for film deposition were spectroscopic standardswith a guaranteed content of 1.0000 ± 0.0001 g l–1 Sb(III), acidified to preventhydrolysis in aqueous solutions. Stock standard solutions of 1.0 mg l–1 Zn(II),Pb(II), and Cd(II) were also acidified with HCl (to pH 2.0). The salts KCl, KBr,K2SO4, KNO3 and KSCN purchased from Merck or Wako were made as 0.1 Mstock solutions. Throughout the experimental work, all the solutions were preparedin doubly deionised water, obtained from a Milli-Q Millipore water purificationsystem.

Apparatus

The device for SIA consisted of a syringe pump (Carvo, Sci Inst., USA), an 8-portselection valve (Carvo), PTFE tubing (0.5 mm, i.d.) for flow lines and a holdingcoil (1.5 mm i.d.). The sequential injection system was controlled via PC and therespective software. An electrochemical analyzer (model 12000; CH Instruments,USA) was employed for all the measurements, when voltammetric experimentswere carried out in the square-wave mode (SWV; for details, see below) whenusing a thin-layer flow cell (Bioanalytical System, USA) incorporating the glassycarbon electrode (GCE; the electrode substrate of choice for plating with antimonyfilm), an Ag/AgCl/3M KCl as the reference, and a stainless-steel tube as thecounter electrode, as well as the solution outlet. The proper running of the ASV-FIA assembly is schematised in Fig. 1 and further specified below — underProcedures

In comparative stationary (batch) measurements, and external electrodestand with a three-electrode cell (GCE, Ag/AgCl, and a Pt-wire as the auxiliary


electrode) was used, when the stirring was accomplished with a magnetic barrotated at approx. 300 rpm. The pH was measured using a portable pH-meter inconjunction with a combined glass electrode (both Horiba, Japan), when itscalibration was performed with a set of commercial buffers.

Fig. 1 The assembly for (SW)ASV-SIA measurements. A schematic view

Procedures

Preparation of the SbFE. For batch measurements and experiments in the ASV-SIA regime, the antimony film was deposited in situ; i.e., by potentiostaticelectrolysis performed directly in sample solutions containing Sb(III) species. Forthe batch system, the plating was performed in a solution of 0.01 M HCl + 2.0 mgl–1 Sb(III) (pH 2.0) and at a potential of –1.40 V vs. ref. for 120 s [15], whereas,in the ASV-SIA procedure, the GCE was being plated in a solution of 0.5 M HCl+ 0.75 µg l–1 Sb(III) at –1.50 V, at a flow rate of 0.5 ml s–1 [31,32]. In both cases,the film was removed electrochemically, at a potential of +0.20 V.

Cyclic Voltammetry (CV). The respective measurements were performed at a scanrate of 100 mV s–1, when starting from the initial potential of +0.50 V vs. ref.,going via the selected vertex potential(s) and, returning in the reversed directionto the starting point.


Square-Wave Anodic Stripping Voltammetry in the Stationary Arrangement(SWASV). The batch experiments comprised a potentiostatic pre-concentration(“accumulation step”) at !1.40 V vs. ref., followed by a 15s equilibration periodand the proper anodic scanning (“stripping step”) from !1.40 V up to +0.20 V vs.ref.

The SWV modulation ramp comprised the following parameters: frequency,fSW = 50 kHz; scan rate, vSW = 30 mV s–1, potential increment, iSW = 6 mV;amplitude, )ESW = 50 mV. Prior to each measurement, the electrode surface of thebare (“stripped”) electrode substrate was conditioned / regenerated by applying apotential of +0.20 V vs. ref. for a period of 30 s.

Square-Wave Anodic Stripping Voltammetry Combined with Sequential InjectionAnalysis (SWASV-SIA). The individual sequences can be described by means ofa schematic manual that has comprised a string of the consecutive steps (i = 1-7)with the corresponding experimental/instrumental parameters. Specifically: Step(1) ... Aspirate (i.e., start to suck) the sample solution into the holding coil (HC)and set the following parameters: the valve position (V), 2; pump status (P):“aspirate”; flow rate (vi); 1200 ml s–1; electrode potential (Ei) vs. ref., –1.5 V vs.ref.; time period/duration (ti), 30 s. (2) Aspirate the solution of Sb(III) into HC,when setting: V = 3, P: “aspirate”, vi = 1200 ml s–1, Ei = +0.20 V, ti = 10 s. (3)Dispense the solution of Sb(III) + sample solution into flow cell and perform thedeposition step (accumulation), when having set: V = 4, P: “deliver”, vi = 500 mls–1, Ei = –1.5 V, ti = 100 s. (4) Perform the equilibration step, when setting: V = 4,P: “deliver”, vi = 0, Ei = –1.5 V, ti = 15 s. (5) Perform the stripping step and recordvoltammogram, when having set: V = 4, P: “deliver”, vi = 0, Ei = from –1.5 to+0.10 V (anodic scanning), ti = 15 s. (6) Aspirate the solution of 0.5 M HCl intoHC, when setting: V = 5, P: “aspirate”, vi = 1200 ml s–1, Ei = +0.20 V, ti = 30 s. (7)Dispense the solution of 0.5 M HCl into flow cell and clean the electrode cell,when having set: V = 4, P: “deliver”, vi = 1200 ml s–1, Ei = +0.20 V, ti = 30 s.

Regarding the square-wave modulation, the parameters of choice were thesame as those applied in the stationary SWASV measurements being specified inthe previous section. Finally, when running SVASV-SIA, the electrode cleaninghad already been included at the end of the measuring cycle and without anyregeneration.


The composition of the sample solutions, as one of the key parameters for theresultant quality of antimony films formed, was investigated first. It is known (see,e.g., Refs [7-9]) that the preferably used solution for the preparation of relatedsimilar BiFEs is being based on acetate buffer with pH ca. 4.5, but strongly acidic


-1.0 -0.5 0.0 0.5

5

0

-5

-10

-15 A

I (µA

)

E (V)-1.0 -0.5 0.0 0.5

0

-5

-10

-15 B

I (µA

)

E (V)

solutions of 1 M HCl are occasionally applicable as well [14]. In order tounderstand the electrochemical behaviour of the SbFE, it is wise to check thedeposition/dissolution scheme of the antimony film at the supporting electrodesurface in both media [33,34]. Figure 2 shows the cyclic voltammograms recordedfor the plating solutions of Sb(III) (3 mg l–1) in HCl (pH . 0) (set “A”) and AcB(pH 4.5) (“B”). As indicated by the dissolution peak of antimony, the respectivefilm starts to grow at the potentials more negative than !0.40 V, reaching themaximum at !0.60 V (A) and !0.80 V (B).

Fig. 2 Cyclic voltamograms of 3000 :g l–1 Sb(III) in 0.1 M HCl (pH 2.0, set “A”) andacetate buffer (pH 4.5; set “B”). Experimental conditions: scan rate, 100 mV s–1;initial potential, +0.50 V; vertex potentials, –1.2 V and –1.4 V; final potential,+0.50 V vs. ref.

After reversing the potential scan direction, the total oxidation of antimony(i.e., its dissolution or “stripping”, resp.) can be observed at +0.03 V in the caseof HCl solution. In contrast to this, the reduction peak in AcB has no oxidationcounterpart; i.e., there is no expected dissolution/stripping scheme. Thisphenomenon can be explained by assuming that in the acetate media (due to arelatively high pH), antimony(III) oxide, Sb2O3, can be formed, which is probablythe main reason for the unfavourable behaviour of the electrode in this medium,observed previously for the ASV determinations [20].

In the SWASV-SIA measurements, the shape and intensity of the oxidationpeaks of Cd, Pb and Sb have been found to be significantly influenced by theconcentration of HCl: in the concentration range of 0.01-0.1 M HCl (see Fig. 3A) the oxidation peaks of Pb and Cd are small or missing completely, which canbe attributed to the insufficient amount of metallic antimony at the GCE surface.The increase in the HCl concentration from 0.3 to 0.6 mol l–1 (Fig. 3 B) resultedin an increase in the intensity of the oxidation/stripping signals for Pb(II) andCd(II). This can then be explained in terms of increased solution conductivity anddecreased degree of hydrolysis, yielding an Sb(III) compound that is more easilydischarged, having a favourable effect on the accumulation and stripping processesof both Pb(II) and Cd(II). It should be noted that more negative reduction


-1.5 -1.0 -0.5 0.0

1

2

3

4A

Sb

Cd Pb

1

2

I (µA

)

E (V)

-1.5 -1.0 -0.5 0.00

2

4

6

3

B

Sb

Cd

Pb

2

1

I (µA

)

E (V)

potentials could also cause a decrease in the oxidation peak because of morepronounced effect of the hydrogen evolution.

Fig. 3 Effect of the solution acidity upon the SWASV-SIA signals for 50 :g l–1

Pb(II)and Cd(II) at upon (mol l–1). A: 1) 0.01 and 2) 0.1 M HCl and B: 1) 0.4, 2)0.5, and 3) 0.6 M HCl. Experimental conditions: 750 :g l–1 Sb(III); depositionpotential, !1.50 V vs. ref.; deposition time, 100 s; equlibrium time, 15 s; scanrate, 30 mV. For other parameters, see Experimental

In view of the observation that the oxidation signals of both metals wereparticularly well defined in 0.5 M HCl (Fig. 3 B, curve 2), as well as due to thefact that this positive effect of HCl in flowing streams was also noticed recentlyby other authors [15,21], the half-molar concentration of HCl was finally chosenas the optimum for the determinations in the ASV-SIA regime. It enabled to detectboth Cd(II) and Pb(II), selected as model ions of the two typical heavy metals, atthe low :g l–1 level as proved in a detailed study on the electroanalyticalperformance of the SbFE. It is illustrated in Fig. 4, showing the linear strippingresponses of the two metals of interest in a concentration range of 5-120 Cd(II)and of 4-100 µg l–1 Pb(II), with the resultant r2 = 0.997 and 0.999. The respectiveLOD values (estimated via the 3F criterion) were 1.4 µg l–1 Cd(II) and 1.2 µg l–1

Pb(II), with the corresponding relative standard deviations (RSDs) of ±2.8 % and±2.6 % for six repetitive measurements (n = 6) and the concentrations chosen of50 µg l–1 Cd(II) and Pb(II). Yet lower LODs may perhaps be expected when usingappropriately prolonged deposition times as there were no evident saturation ten-


-1.0 -0.5 0.00

4

8

12

16

Sb

Cd

Pb

I (µA

)

E (V)

Fig. 4 SWASV-SIA measurements with SbF-GCE for different concentrations of Pb(II)and Cd(II). Experimental conditions and instrumental parameters: see Fig. 3 andExperimental

dencies (not shown).A special attention was paid to the effect of some salts in the sample

solution. Specifically, of interest were nearly chemically inert salts of K2SO4 andKNO3, as well as complex-forming halides and pseudohalides: namely, KCl, KBr,and KSCN. All these salts were added at a relatively high concentration (typicallyas 10 mg l–1 X–) in order to ensure their complexing capabilities or even some othereffect. As seen, the individual anions had affected the stripping processes of leadand cadmium — and, eventually, of other metal ions tested — in different ways(Figs 5 and 6). As expected, the presence of K2SO4 had almost no effect and thetwo SWASV signals observed remained unchanged. Furthermore, the presence ofboth KCl and KBr had caused only minimal increases of the oxidation peaks (Fig.5 A). Rather unexpected was the effect of nitrate, when its addition causednoticeable shifts in the peak potentials, giving rise to a coincidence of the Cd-peakwith the first signal of antimony, accompanied by a 10-fold decrease of the Pb-peak and its overlap with the second Sb-peak.

Moreover, the remaining third oxidation peak of antimony significantlyincreased compared to the situation in the absence of KNO3 (see Fig. 5 C). Finally,the presence of KSCN (Fig. 5D) had only little effect upon the signals for Cd andPb; however, having enabled to register also the oxidation peak for zinc, whichcould not be seen in any of the previous cases. (As found in a brand new andparallel study [35], much more pronounced effect of the thiocyanate anion can beachieved in less acidic solutions and at lower concentration of SCN– (of ca. 0.001


-1.5 -1.0 -0.5 0.0

0

2

4

6

8

10A

2

1

I (µA

)

E (V)-1.5 -1.0 -0.5 0.00

2

4

6

8

10 B

2

1

I (µA

)

E (V)

-1.5 -1.0 -0.5 0.00

2

4

6

8

10D

Zn

Cd

Pb

2

1

I (µA

)

E (V)-1.5 -1.0 -0.5 0.0

0

2

4

6

8

10 C2

1

I (µA

)

E (V)

Fig. 5 Effect of the presence of salts in the supporting electrolyte of 0.5 M HCl upon theSWASV-SIA curves obtained in the mixture of 50 :g l–1 Cd(II) + Pb(II) + Zn(II)in the absence (1) and presence (2) of the same concentration of the respectivesalt (c = 10 mg ml–1): A) KCl, B) K2SO4 , C) KNO3 , and D) KSCN. Forexperimental conditions, see Table I and Fig. 3

mol l–1) then, it is possible to enhance also both Cd-peaks and Pb-peaks up to 3-fold increase compared to their original intensity.)

Since the visualisation of the Zn-response had implied some positive effectof KSCN upon the stripping of other electronegative elements, this pseudohalidewas also examined for eventual detection of two hardly reducible species –chromium(III) and manganese(II). As seen in Fig. 6 C and D, some evidence forCr- and Mn-peaks could be noticed due to the effect of KSCN, but both responseswere poorly developed and scarcely applicable in practical analysis at this stageof experimentation. Nevertheless, this study together with the above-mentionedinvestigations of similar focus [35] have shown clearly that there is stillunexploited field in using the complex-forming reagents for the overallimprovement of electroanalytical performance of both SbFEs and BiFEs.

Last but not least, the multi-elemental analysis of a model mixture of toxicmetal ions, including the addition of 50 µg l–1 Hg(II), is illustrated in Fig. 7. Theresultant voltammogram demonstrates that the presence of bivalent mercury hasinfluenced the potential window available, when changing also the respective


-1.5 -1.0 -0.5 0.00

2

4

6

8

10A

1

2

Sb

Cd

PbI (

µA)

E (V)-1.5 -1.0 -0.5 0.0 0.5

0

2

4

6

8

10B

2

1

Cd Pb

Sb

I (µA

)

E (V)

-1.5 -1.0 -0.5 0.0 0.5

0

2

4

6

8

10 D

2

1

Mn

Cd

Pb

Sb

I (µ

A)

E (V)-1.5 -1.0 -0.5 0.0

0

2

4

6

8

10

2

1

C

SbCr

Cd

Pb

I (µA

)

E (V)

Fig. 6 Effect of KBr (A and B) and of KSCN (C and D) upon the SWASV-SIAresponses in 0.5 M HCl containing the mixtures of 50 :g l–1 Cd(II) + Pb(II) +Mn(II) (A and C) or Cd(II) + Pb(II) + Cr(III) (B and D) in the absence (1) and inthe presence of 10 mg ml–1 salt (2). For other conditions and parameters: see Fig.3 and Experimental

responses at the SbFE; especially, the re-oxidation signal for cadmium havingincreased markedly.

One of possible explanations can be associated with the beneficial effect ofthe amalgamation between Hg and Cd, which is a spontaneous process withmarkedly higher effectiveness compared to a concurrent alloy formation with Sb[8,15,19].

Conclusion

In this article, the electroanalytical performance the SbF-GCE operated in situ hasbeen optimised in an effort to outline its practical application in the SWASV-SIAmeasuring regime and for the determination of lead(II) and cadmium(II) at the low:g l–1 concentration level and, possibly for some other metal ions like Zn(II),Mn(II), or Cr(III).

In this respect, the basic characterisation for the determination of Cd(II) and


-1.5 -1.0 -0.5 0.00

2

4

6

8

Sb

Cr

Cd Pb Hg

I (µA

)E (V)

Fig. 7 SWASV-SIA detection of 50 :g l–1 Hg(II) in the presence of 50 :g l–1 Cd(II) +Pb(II) + Cr(III) in the supporting electrolyte of 0.5 M HCl. For other conditionsand parameters: see Fig. 3 and Experimental

Pb(II) has been successfully completed and the optimal experimental conditionsfound (among others, the plating regime, the supporting electrolyte/samplesolution composition, plus other experimental conditions associated). Furthermore,the effect of some selected anions has also been studied, finding that weaklycomplexing halide salts, namely KCl and KBr, had had a favourable effect uponthe shape and intensity of the re-oxidation signals for both Cd and Pb. Moreover,the presence of KSCN in the plating solution has led to a visualisation of somehighly electronegative elements, thus allowing us to examine a multi-targetanalysis — a simultaneous detection of of Cd(II) and Pb(II) together with Zn(II),Mn(II), and Cr(III).

It can be concluded that the SWASV-SIA/SbF-GCE combination usedthroughout the work seems to be beneficial with respect to the following features.The assembly proposed permits a simple in-situ preparation of the SbF-GCE,together with its effective regeneration and renewal, reduces the risk of samplecontamination, spares the amounts of solutions and samples, and mainly opens upthe possibilities for full automation of the measuring process. Due to the initialcharacter of the entire study, all these benefits will have to be proved in practicalanalysis of real samples.

Acknowledgements

Financial support from the Global COE “Science for Future Molecular Systems”MEXT of Japan is greatly acknowledged. Also, V. G. is thankful for financialsupport from the HTMT, I. Š from the Czech Ministry of Education, Youth, andSports (projects LC 06035 and MSM0021627502), and both for a CEEPUS-IInetwork (CII-CZ-0212-04-1011).


References

[1] Wang J.: Analytical Electrochemistry, 3rd Ed., Wiley-VCH, Hoboken, 2006.[2] De Castro M.D.L., Izquierdo A.: Electroanalysis 3, 457, (1991).[3] Gharib Naseri N., Baldock S.J., Economou A., Goddard N.J., Fielden P. R.:

Anal. Bioanal. Chem. 391, 1283 (2008).[4] Rodriguez J., Ibarra I.S., Galan Vidal C.A., Vega M., Barrado E.:

Electroanalysis 21, 452 (2009).[5] Graabæk A.M., Jeberg B.: Inter. Labor. 22, 33, (1992).[6] Wang J.: Ass. Chem. Res. 35, 811, (2002).[7] Wang J.: Electroanalysis 17, 1341 (2005).[8] Economou A.: TrAC – Trends Anal. Chem. 24, 334 (2005).[9] Kokkinos C., Economou A.: Curr. Anal. Chem. 4, 183, (2008).[10] Švancara I., Prior C., Hočevar S. B., Wang J.: Electroanalysis 22, 1405

(2010).[11] Hutton E. A., Ogorevc B., Smyth M. R.: Electroanalysis 16, 1616 (2004).[12] Sattayasamitsathit S., Thavarungkul P., Kanatharana P.: Electroanalysis 19,

502 (2007). [13] Economou A., Voulgaropoulos A.: Talanta 71, 758 (2007).[14] Chuanuwatanakul S., Dungchai W., Chailapakul O., Motomizu S.: Anal. Sci.

(Japan) 24, 589 (2008).[15] Hočevar S. B., Švancara I., Ogorevc B., Vytřas K.: Anal. Chem. 79, 8639

(2007).[16] Tesařová E., Baldrianová L., Hočevar S. B., Švancara I., Vytřas K., Ogorevc

B.: Electrochim. Acta 54, 1506 (2009).[17] Pauliukait� R., Kalcher K., in: 8th Young Investigators’ Seminar on

Analytical Chemistry (YISAC’01), Book of Abstracts, pp. 10-11, Universityof Pardubice, Pardubice, 2001.

[18] Švancara I., Florescu, M., Stočes M., Baldrianová L., Svobodová E., in:Sensing in Electroanalysis – Vol. 5 (K. Vytřas, K. Kalcher, I. Švancara,Eds.), pp. 109-125, University of Pardubice, Pardubice, 2010.

[19] Švancara I., Hočevar S. B., Baldrianová L., Tesařová E., Vytřas K.: Sci. Pap.Univ. Pardubice, Ser. A 13 (2007) 5-19.

[20] Pauliukait� R., Metelka R., Švancara I., Królicka A., Bobrowski A., NorkusE., Kalcher K., Vytřas K.: Sci. Pap. Univ. Pardubice, Ser. A 10, 47 (2004).

[21] Tesařová E., Vytřas K.: Electroanalysis 21, 1075 (2009).[22] Toghill K., Xiao L.-H., Wildgoose G., Compton R.: Electroanalysis 21, 1113

(2009).[23] Jovanovski V., Hočevar S.B., Ogorevc B.: Electroanalysis 21, 2321 (2009).[24] Kokkinos C., Economou A., Efstathiou C. E.: Electrochem. Commun. 11,250

(2009).[25] Jiang X.-C., Sun Q., Zhang J.-L., Wang B.-Z., Du X.-Y.: Sensor Lett. 7, 91


(2009).[26] Zhang J.-L., Shan Y.-J., Ma J., Li X., Du X.-Y.: Sensor Lett. 7, 605 (2009).[27] Slavec M., Hočevar S. B., Baldrianová L., Tesařová E., Švancara I., Ogorevc

B., Vytřas: Electroanalysis 22, 1617 (2010).[28] Urbanová V., Vytřas K., Kuhn A.: Electrochem. Commun. 12, 114 (2010).[29] Sopha H., Baldrianová L., Tesařová E., Hočevar S. B., Švancara I., Ogorevc

B., Vytřas K.: Electrochim. Acta 55, 7929 (2010).[30] Guzsvány V., Nakajima H., Soh N., Nakano K., Imato T., Sensing in

Electroanalysis, Vol. 4 (K. Vytřas, K. Kalcher, I. Švancara; Eds). pp. 121-134, University of Pardubice, Pardubice, 2009.

[31] Guzsvány V., Nakajima H., Soh N., Nakano K., Imato T.: Anal. Chim. Acta658, 12 (2010).

[32] Guzsvány V., Nakajima H., Soh N., Nakano K., Švancara I., Vytřas K.,Bjelica L., Imato T.: Electroanalysis, submitted (2010).

[33] Švancara I., Baldrianová L., Vlček M., Metelka R., Vytřas K.:Electroanalysis 17, 120 (2005).

[34] Švancara I., Baldrianová L., Tesařová E., Vlček M., Vytřas K., SotiropoulosS., in: Sensing in Electroanalysis – Vol. 2 (K. Vytřas, K. Kalcher, Eds), pp.35-58. University of Pardubice, Pardubice, 2007.

[35] Bartoš K.: Deposition of the Bismuth and Antimony Films from AtypicalSolutions with Possible Applicability in Electroanalysis. Diploma Thesis(MSc), University of Pardubice, Pardubice, 2010.

1 To whom correspondence should be addressed. 2 Based on the continuing research collaboration with Dept. of Analytical Chemistry.

33



16 (2010)

NEW TYPE OF PVC-MEMBRANE ION-SELECTIVEELECTRODES AND THEIR APPLICABILITY

TO DETERMINE SOME ANTIDEPRESSANT DRUGS

Anwar A. WASSEL1 and Shabaan ABDULLATIFNational Organization of Drug Control and Research (NODCAR),

Cairo, Egypt2


The construction and electrochemical characterisation of potentiometric sensorsof the PVC-membrane type are described being developed for the determinationof selected antidepressants; namely: Olanzapine, Oxazepam, and Lorazepam. Thesensing PVC-membranes incorporate the ion-associates of the drug cations withammonium reineckate as selective materials dispersed in 2-nitrophenyl-octylether(NPOE) or dibutyl sebacate (DBS) as the plasticizers of choice. These ISEs exhibitrapid, stable and nearly Nernstian response over a relatively wide concentrationrange of 1×10–7 mol l–1-0.01 mol l–1 for Olanzapine and 1×10–6-0.01 mol l–1 forOxazepam and Lorazepam, in mild acidic solutions with pH 4.5. No interferencescaused by either inorganic or organic species were found. The three electrodesdeveloped have been tested for potentiometric titrations of all three drugs, as wellas for their direct determination via the respective calibration curves. Theindividual experiments then resulted in average recovery rates in an interval of

34 Wassel A.A., Abdullatif S./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 33–55

98.4-99.2 %, with the R.S.D. of ca. ±1.5 % (n = 3) for all three analytes of interest.Finally, the method proposed has been applied to the determination of the threedrugs in pharmaceutical formulations and urine, when the results obtained werecomparable to those obtained with the reference HPLC measurements.

Introduction

Benzodiazepine drugs (BZD) belong to a group of substances that are known fortheir sedative, hypnotic, tranquillizing, and anticonvulsant properties and are beingprescribed worldwide as therapeutics to treat anxiety, sleep disorders, statusepileptics, insomnia, frequent nocturnal awakenings, early morning awakeningsschizophrenia, and related disorders. In addition, they are used as muscle relaxantsfor alleviation of panic attacks and as induction agents in anesthesiology [1].

Olanzapine (Olan — see Scheme 1) is an antipsychotropic agent thatbelongs to the thienobenzo-diazepine class, a relatively new benzodiazepine whichhas been found useful to treat schizophrenia and some other psychosis [2-4].Olanzapine is a crystalline solid substance of yellow colour and with a meltingpoint of 263.2 ºC [5]. Its molecular formula is C17H20N4S, corresponding to amolecular weight of 312.44 g mol–1.

Scheme 1 Chemical formula of Olanzapine (2-methyl-4-(4-methyl-1-piperazinyl)-10H-thieno[2,3-b][1,5]benzodiazepine)

Oxazepam (aka Oxazem or Serax-tablets, Oxaz — see Scheme 2) is an anti-anxietyagent of the benzodiazepine type used primarily for treating mild to moderateanxiety. It is also prescribed as a sedative, hypnotic, anticonvulsant, and musclerelaxant [6-9].

This drug is frequently encountered in clinical and forensic toxicology,having been featured in an increasing number of misuses and abuses over the pastyears [10,11]. Finally, it is also used to treat irritable bowel syndrome. Oxazepamis a creamy white or pale yellow crystalline substance, odourless, and with meltingpoint of 205 ºC. Its molecular formula is C15H11ClN2O2 for both possibleenantiomers, corresponding to a molecular weight of 284 g mol–1.

Wassel A.A., Abdullatif S./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 33–55 35

Scheme 2 Chemical formula of Oxazepam (7-chloro-2,3-dihydro-3-hydroxy-5-phenyl-1H-1,4-benzodiazepin-2-one)

Lorazepam (Lora — see Scheme 3) is classified pharmacologically as a shortacting benzodiazepine and under the trademark Ativan® (1 mg or 2 mg of activesubstance) is a generic representative for the popular anti-anxiety medication,indicated for generalized anxiety disorder (GAD), panic disorder, anxietyassociated with depression and insomnia. Also, it is sometimes used as a sleepingagent, anti-convulsant, and for alcohol detoxification [12-15]. Lorazepam/Ativan®

is a white — or, almost white — microcrystalline powder with a melting point of245 ºC. Its molecular formula is C15H10Cl2N2O2, corresponding to a molecularweight of 321.2 g mol–1.

and enantiomerScheme 3 Chemical formula of Lorazepam (7-chloro-5-(2-chlorophenyl)-1,3-dihydro-

3hydroxy-1H-1,4-benzodiazepine-2-one)

In this article, a simple and rapid method for the determination of the three,above described pharmaceuticals has been developed based on the employment ofion-selective electrodes (ISEs) whose PVC-membranes are saturated with thecorresponding BZD/reineckate ion-associate and that can be used for indicationin direct potentiometry, as well as in potentiometric titrations.


Experimental


All reagents and chemicals used throughout the work were of analytical-reagentgrade and solutions were prepared with deionised distilled water (DDW). Bothpolyvinylchloride (PVC) powder and ammonium reineckate (A-RE; purity: 98 %)used as reagent and titrant were obtained from Aldrich. Dibutyl sebathete (DBS;a plasticizer of purity ~ 99 %) was purchased from BDH (British Drug Houses,Pool, England); 2-Nitrophenyl octyl ether (o-NPOE; a plasticizer of purity ca. 99%) from Fluka. Tetrahydrofuran (THF; a solvent with purity of about 99 %) waspurchased from Aldrich.

An acetate buffer solution of pH 4.5 was prepared by dissolving 13.6 mgsodium acetate with approx. 5 ml glacial acetic acid in 1000 ml deionised anddoubly distilled water. A phosphate buffer solution of pH 4.5 was prepared bymixing the appropriate quantities of 0.1 M H3PO4 with 0.2 M Na2HPO4 in 500 mldeionised distilled water.

In interference studies, selected inorganic cations, amino acids, and sugarswere examined via 1×10–3 M solutions prepared in acetate buffer (pH 4.5), whencontaining the following species: Na+, K+, Mg2+, Ca2+, Sr2+, Cu2+, Zn2+, anddiphenhydramine. The remaining solutions of urea, starch, glucose, and maltosewere 0.01 mol l–1.

Standard Solutions and Precipitates for Membrane Modification

Standard Solutions of the Drugs: The stock solutions of Olanzapine, Oxazepam,and Lorazepam (or Olan, Oxaz, and Lora, resp.) were prepared as 0.01 mol l–1, bydissolving the proper weight of solid substance into suitable amount of 0.002 MHCl added in droplets under continuous stirring till a complete dissolution of thedrug was achieved.

The resultant solution was made up to 100 ml in a measuring flask using de-ionized water. Working diluted standards were then prepared in a series ofcalibration solutions in the concentration range from 1×10–7 up to 0.01 mol l–1 forOlan and from 1×10–6 to 0.01 mol l–1 for Oxaz or Lora; all being made as solutionsin acetate buffer (pH 4.5). Both stock solutions and diluted standards were kept indark and in a refrigerator (at 4 oC).

Standard of ammonium reineckate; i.e., ammonium tetrathiocyanato-diammine chromate(III), NH4[CrIII(SCN)4(NH3)2]AH2O; (A-RE): A 0.01 M standardsolution was prepared by dissolving 2.6300 g in 100 ml of deionised distilledwater.Preparation of the BZD / Reineckate Ion-Associates: Each BZD / RE ion-pair


complex was prepared by mixing the equal volumes of 0.01 M A-RE with 0.01 MOlan, Oxaz or Lora; when the latter being added drop-wise under continuousstirring. The resultant precipitates were left in contact with their native liquorovernight to ensure the complete coagulation, then filtered off through aWhatman® filter paper No. 42, washed thoroughly and repeatedly with distilledwater, left to dry at room temperature for at least 24 hours, and finally ground tofine powder.

Apparatus and Other Equipment

All potentiometric measurements were performed at 25±2 °C using a pH meter(model H-8417, Hanna; for further specification, see Ref. [16]) with pH sensitivityof ± 0.05 pH units with Drug-PVC matrix membrane sensors in conjunction witha double-junction Ag/AgCl reference electrode (model 900200, Orion) containinga solution of (10 %, w/w) KNO3 in the outer compartment. The combined glasselectrode (model F-22E, Horiba [16]) was used for all the pH measurements aftercalibration with the use of a set of commercially available buffers.

The electrochemical system used can be represented as follows: Ag|AgCl|KCl (saturated, ca. 3 mol l–1)||Salt bridge||Sample, it means: Reference electrodehalf-cell|Membrane|Filling solution|Ag/AgCl|Ion-selective electrode half-cell;where the filling solution is a mixture of 0.01 M solution for each of investigateddrugs and KCl.

Procedures and Samples

Membrane Preparation: A 10 mg portion of drug/reineckate ion-pair complex wasthoroughly mixed in a glass Petri dish (5 cm in diameter) with 190 mg of PVCpowder and 350 mg of 2-nitrophenyl octylether (NPOE) or dibutyl sebacate(DBS).

The resultant cocktail was dissolved in 5 ml THF, covered with a filterpaper and left overnight to evaporate to dryness at ambient temperature. Using thisprocedure, a semi-transparent and plastic-like membrane with a thickness 0.1 mmcould be obtained. Then, the parts of the membrane material were cut using a corkborer and, when plasticizing with THF, an exchangeable PVC-tip was clipped ontothe end of the electrode glass body. The tip was left to stand for 24 hours to allowcomplete evaporation of THF and, finally, the fabricated electrode was soaked into0.01 M solution of the corresponding drug for 24 hours before use (and also beingstored in the same solution when not in use).

The constructed sensor was washed with deionised distilled water andblotted with tissue paper between measurements. The effect of membrane


composition was studied for Olanzapine-modified sensor which was plasticizedwith NPOE — by weighing three different amounts of ionophore (5 mg, 10 mg,or 15 mg, respectively), and mixing with 190 mg PVC + 350 mg NPOE + 5 mlTHF in a glass mortar. To obtain homogeneous and uniform thickness of themixture [17], the membranes were left to dry free (not less than 24 hours).

Measurement and Construction of Calibration Curves: 10 ml aliquots of eachsample solutions — in acetate buffer of pH 4.5 — having concentrations of1×10–7-0.01 M for Olan or 1×10–6-0.01 M for Oxaz and Lora were transferred into50ml beakers. The potential in mV of each sample solution was directly measuredusing the respective (drug/reineckate ion-pair modified) ISEs based on DBS orNPOE as plasticizer as mentioned in the previous section.

The solutions were stirred and the potential readings were recorded from thelow to the high concentration after stabilization to ±0.3 mV [18]. The potentialswere plotted as a function of –log (drug concentration). These graphs were usedfor the subsequent determination of unknown concentrations of correspondingdrug. The theoretical (aka “lower”) detection limit was determined at the point ofintersection of the extrapolated linear segments of the corresponding drugcalibration curve.

Studies on pH-Effect and Response Time: The effect of pH of the test solution onthe potential values of the electrode system in solutions of different concentrationsof pharmaceutical compounds was tested using two concentrations of each drugion (1×10–3 and 1×10–4 M) over a range of pH (2-11) and recording the potentialreadings of the studied sensors by immersing the Ross combination glasselectrode, corresponding drug-sensor and a double junction Ag/AgCl referenceelectrode in 50 ml beakers containing 25 ml aliquots of 1×10–3 and/or 1×10–4 Mdrug-aqueous solutions. The pH of each solution was gradually changed by addingsmall aliquots of dilute sodium hydroxide and/or hydrochloric acid solutions. ThemV vs. pH profile of each drug concentration was plotted for each electrodesystem. The potential readings that were insensitive to pH changes were obtainedfrom the mV–pH plots. The dynamic response time of the ISEs could also becharacterised by the fact that the respective measurements required to reach asteady-state potential within ±0.3 mV only, even if the analysed concentrationsdiffered also in one order.

Determination of Sensor Selectivity: The influence of some inorganic cations,sugars and amino acids, used as additives or binders, on the response of theelectrodes towards their respective drugs was tested. The separate solution method[19], which requires two potential measurements: first, the potential is measuredin a solution containing a known activity of the ion for which the electrode isselective, second, the potential is measured in a solution containing the interfering


ion, was used to determine the selectivity coefficient value, Kd,i — see later inTable III. Separate drug primary ion (d) and interfering secondary ion (i) solutionshaving equal concentrations were prepared. Their potentials Ed and EI weremeasured using prepared ISEs in the previously mentioned cell. Selectivitycoefficients were calculated using either one of the following equations

(1)

(2)

where EI and Ed are the electrode potentials of 0.01 M solution of each of theinvestigated drug and interfering cation, I, respectively. Equation (1) is used formono-valent secondary ions whereas Eq. (2) is used for divalent or higher ones.zd , zi are the charges on the primary or secondary ions, and S is the slope ofcalibration curve for the primary ion.

The individual interfering inorganic ions and organic species used in thestudies on the selectivity of the different membranes were already specified above.For the proper measurements, the respective drug solution was transferred into a50 ml beaker containing 9.0 ml acetate buffer (pH 4.5). The drug sensor inconjunction with a double junction Ag/AgCl reference electrode was immersedinto the solution and the potential read (in mV). Then, the solution with interferingspecies was transferred into a 50 ml beaker containing 9.0 ml the same buffer andthe potential change was again recorded (as EI). Selectivity coefficients werecalculated as shown before.

Potentiometric Titrations. The individual aliquots of 0.01 M BZD were transferredinto 50 ml titration cell and diluted to 10 ml with the acetate buffer of pH 4.5. Theresulting solutions were stirred and titrated against 0.01 M A-RE according to thecorresponding ion-exchanger used in constructing the electrode. The propertitration curves, i.e., the electrode potential vs. the titrant volume (E-vs.-V) plotsand their end-points were evaluated conventionally.

Pharmaceutical Samples of the Drugs Studied. Five tablets of the drugformulations were weighted and finely powdered in a small mortar. An accuratelyweighted portion of powdered drug, equivalent to 1.562 mg Olan, 1.433 mg Oxaz,or 1.606 mg Lora, was transferred into 5ml measuring flask and dissolved in theminimum volume of de-ionized distilled water and few drops of 0.002 M HCl. Thesolution was filtered into a 50ml calibrated flask, then diluted to the mark with theacetate buffer of pH 4.5 and shaken well.


A series of 10, 15, and 20 ml aliquots of each drug solution were transferredinto 25ml volumetric flasks and made up to the mark with acetate buffer. Thus,solutions having concentrations of 12.5, 18.75 and 25 :g ml–1 Olan, 11.46, 17.2and 22.92 :g ml–1 Oxaz, or 12.85, 19.30 and 29.70 :g ml–1 Lora were made and thepotentials of the resultant drug solutions could be directly measured at thecorresponding ISE.

Biological Fluids (Drugs in Urine): Solutions of 5×10–3 M Olan (31.24 :g ml–1),Oxaz (28.66 :g ml–1) or Lora (32.12 :g ml–1), were prepared by dissolving theappropriate amount of pure substances — specifically: 1.562 mg, 1.433 mg, and1.606 mg — in 50 ml in the order and using solvent as described above. Aliquotsof 10, 15, and 20 ml (pipetted from the respective drug solutions) were transferredinto 25ml measuring flasks; each containing 5 ml urine, and filled up to the markwith solvent. The actual electrode potentials of the individual solutions weremeasured in the same way as above.


Principles of the Electrode Functioning

According to the literature (see, e.g., Ref. [20]), the respective drug moiety can beprotonised and such a cation can readily react with the anionic form of reineckate,following the scheme(s)

(3)

(4)

(5)

where H+-(OLAN)2 or H+-OXAZ and H+-LORA, respectively, denote the protonatedform of the respective drug and the compounds on the right-hand side the resultantion-associates.

The Operability of the Three Electrodes Studied

Olan-, Oxaz-, and Lora-ISEs with the electroactive material chosen, i.e.,H+-DRUG / reineckate were constructed using PVC as a matrix and NPOE or DBSas the plasticizers. The ISEs were investigated in order to compare their electrodeperformance and the results are summarized in Tables I and II. The corresponding


calibration graphs, given in Figs 1, 2 and 3, indicate that for all the electrodestested, having nearly Nernstian cationic response slopes (26.12 and 24.22 mVdecade–1 with a limit of detection 1.2×10–7 and 0.9×10–7 M for Olan-ISEplasticized with NPOE and DBS, respectively, and 52.10- 54.30 mV decade–1 forboth Oxaz-ISE and Lora-ISE. These slopes also confirm the pathways of the

Table I Potentiometric responces characteristics for Olan-RE/membrane ISE with NPOE orDBS as plasticizer

Parameter evaluatedOlanzapine-RE sensor

NPOE DBS

Concentration range, mol l–1 10–7-10–2 10–7-10–2

Slope, mV decade–1 26.12 24.22

Intercept 162.8 167.7

R.S.D, % 0.62 0.77

Limit of detection, mol l–1 3.2×10–7 4.2×10–7

Correlation coefficient (r2) 0.9939 0.9933

Actual pH (range) 4.5 4.5

Response time, s 20 20

Life span, weeks 4 4

Table II Potentiometric responces characteristics for Oxaz- and Lora-membrane ISEs withNPOE or DBS as plasticizer

Parameter evaluatedOxa-RE sensor Lora-RE sensor

NPOE DBS NPOE DBS

Concentration range, mol l–1 10–6-10–2 10–6-10–2 10–6-10–2 10–6-10–2

Slope, mV decade–1 54.30 52.10 55.01 54.43

Intercept 354.8 355.4 343.9 354.5

R.S.D, % 4.6×10–6 5.3×10–6 3.7×10–6 4.8×10–6

Limit of detection, mol l–1 0.56 1.20 0.74 1.30

Correlation coefficient (r2) 0.9989 0.9978 0.9979 0.9978

Actual pH (range) 4.5 4.5 4.5 4.5

Response time, s 30 30 40 40

Life span, weeks 40336 40336 40336 40336


above-described reactions (1) and (2), and the stoichiometry of the correspondingion-pairs formed.

Fig. 1 Calibration curve of Olan-membrane ISEs with NPOE or DBD as plasticizer

Fig. 2 Calibration curve of Oxaz- and Lora-membrane ISEs with NPOE as plasticizer

The linearity range extended down to 1×10–7 mol l–1 for Olan and 1×10–6

mol l–1 for Oxaz or Lora. Also, the detection limits that were estimated accordingto a widely adopted criterion “three sigma” [21,22] could be evaluated from thecalibration curves. Although the three different electrodes exhibited certaindifferences between linearity ranges and response slopes, the resultant nuanceswere found insignificant.


Fig. 3 Calibration curve of Oxaz- and Lora-membrane ISEs with DBS as plasticizer

The Electrodes and Their Selectivity Characteristics

The selectivity coefficient is the main source of information concerninginterferences on the electrode response. In analytical applications, the selectivityfor the analyte must be as high as possible, i.e. the selectivity coefficients must bevery small, so that the electrode exhibits a Nernstian dependence on the primaryion over a wide concentration range. The selectivity of the ion-exchanger ofmembrane electrodes depends on the selectivity of the ion-exchange process at themembrane test solution interface and the mobility of the respective ions in thematrix of the membrane. The hydrophobic interactions between the primary ionsand the PVC membrane are reflected by the values of the Gibb’s energy of transferfor cations between the aqueous and membrane phases. The response of theelectrodes towards different substances and ionic species was checked.

Table III Potentiometric selectivity coefficients, , for three proposed electrodes based onNPOE (where i denotes mediator and j enterrfering species

Interfering species (ion/substance) Olan-sensor Oxa-sensor Lora-sensor

Na+ 8.5×10–4 1.6×10–3 7.2×10–4

K+ 7.7×10–4 4.4×10–3 5.6×10–3

Mg2+ 5.3×10–4 5.8×10–3 4.1×10–3

Ca2+ 8.5×10–3 7.3×10–4 6.3×10–4

Sr2+ 1.2×10–2 6.6×10–3 3.3×10–3


Table III — Continued

Interfering species (ion/substance) Olan-sensor Oxa-sensor Lora-sensor

Cu2+ 6.3×10–2 5.2×10–3 6.1×10–3

Zn2+ 2.3×10–2 8.7×10–3 4.3×10–3

Diphenhydramine 3.4×10–3 2.3×10–4 1.2×10–4

Urea 8.2×10–4 2.2×10–4 2.4×10–4

Starch 2.4×10–4 4.3×10–4 8.1×10–4

Glucose 3.6×10–4 3.6×10–4 7.4×10–4

Maltose 5.4×10–4 6.2×10–4 6.6×10–4

Olanzapine - 0.7×10–2 3.2×10–2

Oxazepam 3.2×10–2 - 3.1×10–2

Lorazepam 1.3×10–2 4.2×10–2 -

As mentioned in the Experimental part, the selectivity coefficients of theinterfering cations were determined by the separate solution method, which is avery simple approach providing a reasonable measure of the degree of interferenceof foreign species that may be present in the test solution [23,24] — see Table III.The mechanism of selectivity is mainly based on the stereo-specificity andelectrostatic environment and it is dependent on how much fitting is presentbetween the locations of the lipophilicity sites in the two competing species in thedrug solution side and those present in the receptor of the ion-exchanger.

The inorganic cations do not interfere because of differences in ion size,mobility and permeability. Also, the smaller the energy of hydration of the cation,the greater is the response of the membrane. The electrodes exhibit good tolerancetoward sugars, amino acids, and urea as the presence of these species up to 4- oreven 5-fold excess did not affect significantly the potential reading. This hasindicated that the presence of these species can be tolerated to a high extent. Sucha high tolerance can then be attributed to the differences in polarity and lipophilicnature of their molecules relative to those of the drugs under investigation. ThoughOlan, Oxaz and Lora mutually interfere, these three similar substances so far donot exist in one pharmaceutical preparation.


Effect of pH and of the Type of Buffer

The influence of pH on the response of the proposed drug-selective electrode waschecked by measuring the potential displayed by 1×10–3 and 1×10–4 M drug testsolutions over the range of pH 2-11. Small volumes of dilute sodium hydroxideand/or HCl were used for adjustment. It is apparent from the potential-pH profilesthat the response of all three sensors was fairly constant in acetate buffer solutionsof pH 3.5-6.0. The potential did not vary over this range by more than ± 3mV andthe electrode could be applied for the determination of investigated drugs. Aconsiderable decrease in the potential above pH 6 was observed, which wasprobably due to the decreased concentration of the protonated form of the drug,whereas — at higher pH — a deviation in the potentials reading might occur dueto the penetration of the hydroxonium ion into the membrane layer. On the otherhand, two types of buffers, i.e., acetate and phosphate media, were investigated onOlan-ISE in order to ensure that the electrode would be functioning [25]. Theeffect of the two buffers was not critical and no marked difference was observedas shown in Fig. 4 and Table IV. However, acetate buffer of pH 4.5 was found tobe more suitable because of the fact that the sensor potentials were almost constantand stable within ±3 mV.

Fig. 4 Calibration curve of Olan-electrode obtained in acetate and phosphates buffers

Effect of the Plasticizer Used

In this work, polyvinyl chloride (PVC) membranes plasticized with DBS or NPOEand containing the lipophilic salt, Olan-Re, Oxaz-Re or Lora-Re, were prepared.Potentiometric responses of the electrodes based on neutral ionophores and


modified with the investigated drugs were greatly influenced by the polarity of themembrane medium, which was in turn defined by the dielectric constants of theincorporated drug moiety and plasticizer (solvent mediator) used. Regarding thelatter, the two dielectric constants were Di(DBS) = 4.01 and Di(NPOE) = 23.6, indicatinga high sensitivity and nearly Nernstian slope for the ISEs with NPOE-plasticizedmembrane, whereas the ISEs with DBS-plasticized membranes showed anomalousresponse to the corresponding drug.

Table IV Potential response of Olan-ISE with NPOE plasticizer and in buffer of choice

Parameter evaluated Acetate buffer Phosphate buffer

Concentration range, mol l–1 10–6-10–2 10–6-10–2

Slope, mV decade–1 26.92 26.33

Intercept 175.6 184.4

Limit of detection, mol l–1 1.0×10–6 0.9×10–6

R.S.D., % 2.80 2.76

Correlation coefficient (r2) 0.9986 0.9983

Composition of the Membranes

In plastic membranes of ISEs studied, the amount of the ion-pair exchanger shouldbe sufficient to obtain reasonable ionic exchange at the gel layer-test solutioninterface, which is responsible for the membrane potential and its changes. Also,the amount of the plasticizer should be in the extent that produces a membrane of

Fig. 5 Effect of composition of ionophore on response of Olan-sensors


good physical properties and, at the same time, plays efficiently its role as asolvent mediator for the ion-exchanger. The composition of each of thesemembranes was varied in such a way as to reach the optimum compositionexhibiting the best performance characteristics (calibration-curve slope, rectilinearconcentration range, and the response reproducibility).

Furthermore, it comprised the different compositions of ionophorecooperating in the membrane and the slope (mV decade–1) of the obtainedcalibration graphs for Olan-sensor plasticized with NPOE. The membrane wasprepared three times and the relative standard deviation (R.S.D.) calculated for theslope values was always fairly low, indicating good reproducibility of thepreparation process and comparable slope values of the calibration graphsobtained with ISEs differing in composition. As shown in Fig. 5, it is evident thatthe change of composition of ionophore has significantly affected the response ofthe electrode. The results proper are gathered in Table V, showing that the optimalcomposition of ionophore is 10 mg, offering an ISE with the best performance.

Table V Effect of ionophore on response of Olan-ISE with NPOE as plasticizer

Parameters evaluatedComposition, mg

5 10 15

Slope, mV decade–1 20.75 26.91 22.38

Intercept 156.7 186.3 173.7

R.S.D., % 5.1 3.83 4.6

Limit of detection, mol l–1 6.2×10–7 3.1×10–7 4.6×10–7

Correlation coefficient (r2) 0.9839 0.9988 0.9923

Regeneration of the Ion-Selective Electrodes Prepared

From the above discussion, it is clear that soaking of the electrodes in theirrespective drug solutions for a long time has a negative effect on the response ofthe membrane towards the ions investigated. The same effect appears afterworking with the electrode for a long time. The regeneration of such “exhausted”electrodes was tried by simply re-loading of the ion-exchangers of the Olan-RE,Oxa-RE, and Lora-RE type, in the external gel layer of the membrane [26]. Thiswas accomplished by soaking the drained electrodes for 24 hours in a solution of0.01 M A-RE, followed by saturation in a solution of the respective drug. Asfound, the soaking in 0.01 M solution of each drug for ca. three hours wassufficient for regeneration and, seen from the slopes of the renewed electrodes,increasing from 21.32, 45.70, and 43.33 up to 24.51, 52.16 and 50.01 mV decade–1


for Olan-, Oxaz-, and Lora-ISEs, respectively. Generally, it has been found thatthe life time of regenerated electrodes is limited up to few hours (60-180 min.).This was due to the ease of leaching of lipophilic salts from the gel layer at theelectrode surface compared to a situation when they had been firmly attached tothe PVC network through the solvent mediator with strong binding forces.

Choice of the Method for Quantitative Analysis and Potentiometric Determinationof the Drugs of Interest Using BZD/RE-modified PVC-Membrane ISEs.

In fact, three principal methods could be considered as potentially applicable tothe quantitative analysis with the three Drug/RE-ISEs developed during thesestudies. Such a choice comprised: (i) direct calibration measurements, constructionof the corresponding plot and the calculation of the concentration from the Nernstequation; (ii) potentiometric titration involving the use of counter ions as thetitrant which seemed more accurate, depending essentially on the use of the ISEas end-point detector, but more time consuming (in an estimate, at least 20-30 min.for the average analysis); (iii) the standard addition(s) method, which is often mostreliable, due to the most effective suppression of matrix effects, and therefore quitefrequently applied in practical determinations with ISEs.

Direct Potentiometry: The individual drugs contained in commercialpharmaceutical preparations had represented the first part of practical samples thatwere analyzed and their concentrations determined with the respective ISE whenusing the standard addition method. The results obtained are listed in Table VI. By applying the least-squares methodfor five replicate determination (i.e., n = 5) at a 95 % confidence level, thecorresponding regression equations can be written as follows

(6)

(7)

(8)

when all three regression plots have exhibited nearly ideal correlation (with r2 =0.9999) and where X is the average reference assay and Y is the average found bythe proposed electrode / method. Also, the RSDs obtained were highlysatisfactory, ranging from ±0.8 to ±2.1 %, and the same can be stated about therecovery rates varying from 97.93 to 99.27 %, with the variation coefficientswithin an interval of 0.81-1.66. Finally, there is also a fine agreement with theresults of the official reference method [27].


The results show that the proposed electrodes can be used to determineOlanzapine, Oxazepam and Lorazepam in pure samples or pharmaceuticalpreparations with both high accuracy and high recovery; all without time-consuming pre-treatment procedures required, e.g., to minimize the undesirablematrix effects. Also, the target drugs were successfully determined in spiked urine

Table VI Quantification of content of Olanzapine, Oxazepam and Larazepam in pharmaceuticalpreparations using respective ISE with NPOE as plasticizer

Method proposed Reference method*

Taken:g ml–1

Recovery%

R.S.D.C.V.

t-value Recovery%

±R.S.D%

F

Olanzapine, Zyprexa tablet, 10 mg

12.5 98.4 0.20 0.6 99.11 0.56 1.14

1.62

18.75 98.66 0.15 0.85 98.32 0.41 0.13

0.81

25.00 98.42 0.41 2.34 100.1 0.37 1.22

1.66

Oxazepam, Serenid-D (Weyth) tablet, 15 mg

11.46 98.22 0.43 1.10 100.02 0.33 1.69

0.83

17.20 98.58 0.47 1.18 97.12 0.40 1.38

1.27

22.93 99.27 0.36 0.86 98.65 0.25 2.07

1.51

Lorazepam, Lorazem tablet, 2 mg

12.85 97.93 0.26 1.80 99.54 0.22 1.39

1.20

19.30 98.44 0.66 1.18 100.21 0.54 1.49

1.42

29.70 98.83 0.44 0.76 98.35 0.34 1.67

1.49*Reference method for Olanzapine


samples with the mean recoveries within 97.67-99.31 %, with the RSDs in an interval of 0.16-0.53 %, respectively. The recovery and the RSDs characterisingthe method tested are summarized in Table VII, surveying also other useful details.These results have shown that all the electrodes can detect the drugs in the spikedurine samples with high accuracy and precision, as well as with s high recovery;again, without any additional pre-treatment procedures.

Table VII Potentiometric determination of Olanzapine, Oxazepam and Lorazepam in urinesamples using respective ISE with NPOE as plasticizer

Method proposed Reference method*

Taken:g ml–1

Recovery%

±R.S.D.%

t-value Recovery%

±R.S.D%

F

Olanzapine

12.50 97.67 0.50 0.27 98.11 0.56 0.79

18.75 97.78 0.41 0.53 98.32 1.03 0.15

25.00 98.23 0.47 2.36 100.21 0.25 3.52

Oxazepam

11.46 97.68 0.26 1.72 99.42 0.33 0.63

17.20 98.12 0.32 0.22 98.00 0.23 1.93

22.93 99.31 0.49 0.68 98.65 0.36 1.85

Lorazepam

12.85 98.24 0.16 1.37 99.54 0.22 0.53

19.30 99.12 0.40 1.15 100.21 0.34 1.38

29.70 98.00 0.53 0.41 98.35 0.34 2.42

*Reference method for Olanzapine

Potentiometric Titrations: Though the determination by potentiometric titrationsmay be time consuming, it offers the advantage of high accuracy and precision,when the end-point can easily be determined via the sharp potential break. Inaddition, partially exhausted electrode can also be employed because theactual/exact potential value during titration and at its end-point is of secondaryinterest [28]. Representative titration curves for the determination of all threedrugs using the respective electrodes are shown in Figs 6, 7, and 8. Notable is thefact that as the concentration of the drug had increased, the inflection of the break-point became sharper than those for lower concentrations of the drugs. Then, these


electrodes could be successfully used to indicate properly the potentiometrictitrations of the drugs investigated.

Fig. 6 Typical potentiometric titration curves of (A) 2.0 ml; (B) 4.0 ml and (C) 6 ml0.01 M Olanzapine with standard solution of 0.01 M ammonium reineckate (A-RE) as titrant and using Olan-ISE as indicator electrode

Fig. 7 Typical potentiometric titration curves of (A) 2.0 ml; (B) 4.0 ml and (C) 6 ml0.01 M Oxazepam with standard solution of 0.01 M A-RE using Oxaz-ISE

Statistical Evaluation of Both Methods for Quantitative Analysis

The results of conventional F-and t-tests are shown in Table VIII. The obtainedvalues were substantially lower than theoretical (tabulated) values; i.e., the


methods applied do not exhibit significant differences compared to those obtainedby the official method [27].

Fig. 8 Typical potentiometric titration curves of (A) 1.0 ml; (B) 3.0 ml and (C) 5 ml0.01 M Lorazepam with standard solution of 0.01 M A-RE using Lora-ISE

Table VIII Statistical treatment of data obtained from determination of three drugs at electrodesproposed in comparison with reference method employing HPLC.

Parametersevaluated

Reference method(based on HPLC)

Olanzapineelectrode

Oxazepamelectrode

Lorazepamelectrode

Working rangemol l–1

0.5-18 :g ml–1 *

12-40 :g ml–1 ** 10–7-10–2 10–6-10–2 10–6-10–2

20-130 :g ml–1 ***

LOD ***mol l–1

0.12 :g ml–1 *

3.5 :g ml–1 ** 3.2×10–7 4.6×10–6 3.7×10–6

5.0 :g ml–1 ***

Accuracy%

98.89 ± 1.3*

98.56 ± 1.1** 98.40-98.66 98.22-99.27 97.93-98.83

98.41 ± 1.3***

* Reference method (used) for Olanzepine (see Ref. [27])** Reference method for Oxazepam and Lorazepam [27]*** Limit of detection; average of three determinations


To check the accuracy, the new potentiometric method was compared withstandardized procudure with the aim to find if there is any significant differencewhich could reveal the extent into which potential deviation(s) may affect theapplicability of the new method(s) compared to the already existing method(s). Asfound out, the method examined did not exhibit any significant differences, asconfirmed by the data in the last Table.

Conclusion

The new ISEs based on the functioning of the BZD/RE ion-associate in the PVC-membrane have been shown to be applicable in potentiometric analysis of threepharmaceuticals of the benzodiazepine family, allowing one to determine as lowas 1×10–7 M Olanzapine and 1×10–6 M Oxazepam or Lorazepam. The electrodesproposed offer the advantages of low cost, ease of fabrication, high stability, fastresponse over a wide concentration of the analyte(s) being almost independent ofpH, as well as adequate selective in the presence of related species.

The whole procedure has also been found sufficiently reliable for quantitativeanalysis, when combined either with the calibration curve method or with thestandard addition(s) procedure. Last but not least, the drugs of interest can bedetermined either in pure (powdered) form or in the commercial dosage (tablet)forms without any special pre-treatment and the employment of all three ISEsdeveloped seems to be feasible also in analyses of turbid or intensively colouredsample solutions.

It can be concluded that the procedure described in the previous sections isa valuable contribution to the palette of similar methods already developed foranalysis of pharmaceuticals with familiar structures, showing also thatpotentiometric indication may offer a possible alternative to more commonvoltammetric measurements (see, e.g., Refs [29,30]).

Acknowledgements

The authors would like to thank Prof. Dr. Ivan Švancara for his interest in thearticle, kind help in its final editing, and — last but not least — for enabling ourpaper to appear in the Sci. Pap. Univ. Pardubice.

References

[1] Gilman N., Rosen P., Earley J., Cook C., Todaro L.: J. Am. Chem. Soc. 112,3969 (1990).


[2] Dossenbach M., Folnegovic V. S., Hotujac L., Uglesic B., Tollefson G.,Grundy, S., Friedel P., Jakovljevic M.: Prog. Neuropsychopharmacol. Biol.Psychiatry 28, 311 (2004)

[3] Heresco-Levy U., Ermilov M., Lichtenberg P., Bar G., Javitt D.: Biol.Psychiatry 55, 165 (2004).

[4] Sinclair D.A., Skaer T.L., Robinson L.M., Dickson W.M., Markowitz J.S.,De Vane C. L.: Eur. Neuropsychopharmacol. 13, S281 (2003).

[5] Clarkes J.: Analysis of Drugs and Poisons. Wiley, New York, 2005[6] Moller H.J.: J. Clin. Psychopharmacol. 19, 2S (1999).[7] Lader M. H.: Eur. Neuropsychopharmacol. 9, 399 (1999).[8] Ashton H.: Textbook of Drugs. Elsevier, Amsterdam, 1994.[9] Argyropoulos S.V., Nutt J.: Eur. Neuropsychopharmacol. 9, 407 (1999).[10] Verwey P., Eling H., Wientjes F., Zitman G.: J. Clin. Psychiatry 61, 465

(2000).[11] Pelissolo G.: Encephale 22, 187 (1996).[12] Blin N., Simon E., Jouve M., Habib D., Gayraud A., Durand, B., Bruguerolle,

P. Pisano: Clin. Neuropharmacol. 24, 71 (2001).[13] Barr J., Zomorodi K., Bertaccini E. J., Shafer S. L., Geller, E.:

Anesthesiology 95, 286 (2001).[14] Houghton J.: Brain J. Anaesth. 55, 767 (1983).[15] Jacaz A., Burtin P.: Clin. Pharmacokinet. 31, 423 (1996)[16] Hassan S.M., Abu Sekkina M.M., El Ries M. A.: J. Pharm. Biomed. Anal.

30, 175 (2003). [17] Authors team: British Pharmacopoeia. HMSO Publication, London, 2007.[18] Auterhoff H., Knabe J., Holtje H.D. and Lehrbuch D.: Pharm. Chem. Acta 1,

481 (1999).[19] Robert W., Arthurk C., Nils F., Wolf R., Andrze K., Willem G., Clin. Chan.

Lab. 38, 6 (2000).[20] Maand T.S., Hassan S.S.M.: Organic Analysis Using Ion-Selective

Electrodes, Vol. 1 and 2, Academic Press, London, 1982.[21] Sokalski T., Ceresa A., Zwickl T., Pretsch E.: J. Am. Chem. Soc. 119, 11347

(1997).[22] Skoog D.A., Holler E.J., Nieman T.A.: Principles of Instrumental Analysis,

5th Ed., Harcourt & Brace, Orlando (1998).[23] Cosofret V.V., Buck R.P.: Analyst 109, 1321 (1984).[24] Gadzekpo V.P.Y., Christian G.D., Anal. Chem. Acta 164, 279 (1984).[25] Bates M.M., Cardwell T.J., Cattrall R.W., Deady L.W., Gregorio C.G.:

Talanta 42, 999 (1995).[26] Shoukry A.F., Badway S.S., Issa Y.M.: Anal. Chem. 59, 1078 (1987).[27] Pistos C., Stewart J.T.: J. Pharm. Biomed. Anal. 33, 1135 (2003); and United

States Pharmacopoeia", USP-NF27; U. S. Pharm. Convention Inc.,Rockville Md., (2008).


[28] Fritz F.S., Schenk G.H.: Quantitative Analytical Chemistry, 4th Ed.; Allyn &Bacon, London, 1975.

[29] El Maali N.: Bioelectrochem. 64, 99 (2004).[30] Radi A.E.: Curr. Pharm. Anal. 2, 1 (2006).


2 Based on the continuing research collaboration with Department of Analytical Chemistry, University of Pardubice.

57



16 (2010)

INFRARED SPECTROSCOPIC ANALYSISOF POLYMORPHISM IN DIPHENYL CARBAZIDE

F. El-KABBANYa, S. Taha HASSANb1 and M. HAFEZa

aPhysics Department, Cairo University, Egypt2,bPhysics Department, Fayoum University, Egypt


Infrared (IR) analysis is used here to investigate the changes in N-N, N-H, andC=O bonds of thermally treated diphenyl carbazide (DPC) during the variationfrom room temperature up to 160 oC. Polymorphism in the DPC compound hasbeen studied by detecting the changes in some IR spectroscopic parameters (e.g.,mode shift, band contour) during the elevation of temperature. Beside IRspectroscopy, also differential scanning calorimetry, X-ray diffraction, NMR, andatomic mass spectra were used as supplementary identification techniques. All ofthe vibrations of DPC were found to be due to the ionic fundamentals 3311 cm–1,3097 cm–1, 3052 cm–1, 1677 cm–1, 1602 cm–1, 1492 cm–1, 1306 cm–1, 1252 cm–1, 887cm–1 and 755 cm–1. The results revealed for the first time that the thermally treatedDPC traverse four different phase transformations at 50, 90, 125, and 140 °C. Thecrystal structure was found to be amorphous, monoclinic, tetragonal,

58 El-Kabbany F. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 57–73

orthorhombic and amorphous in the temperature range of 30-160 °C. The X-raydiffraction patterns support the results obtained by IR.

Introduction

Diphenyl carbazide (DPC; C13H14N4O) is one of the important organic compoundsmainly used as an artificial donor material. Also, it is frequently utilised inanalytical chemistry as a sensitive reagent for some metal ions, especially Hg2+,Cd2+ and [1], when the latter can be determined in the form of the complexwith DPC in real samples with the aid of calorimetry [2], colorimetry [3-7], orelectrochemical stripping analysis [8]. The DPC molecule has an orthorhombicstructure with space group pnma, characterised by the lattice parameters of a =5.7171 Å, b = 8.4121 Å, and c = 25.6982 Å. The two phenyl hydrazide groups arelocated on either side of a crystallographic symmetry plane, passing along thebond direction of the carbonyl group [9].

The present work is considered as an extension of our previous work onDPC, in which differential thermal analysis (DTA), differential scanningcalorimetry (DSC), and some electrical measurements were carried out [10,11].The previous work on DPC compound has shown the presence of an exothermicphase transformation at a temperature of about 90 °C. Nowadays and after morethan twelve years, instrumental resolution, sensitivity and accuracy have beenwidely improved. Besides that, most of the techniques have been computerized.Thus, it is very interesting and promising to perform thermal scanned programs forstudying the effect of heat treatment on the spectroscopic properties of the heattreated DPC. Then, the molecule of DPC can be subjected to an infrared (IR)spectroscopic analysis under conditions similar to that performed previously inthermal and electrical study [11]. The nature of the temperature dependence of thevibrating groups is of great interest, as its role is fundamental to explain many ofthe physical properties of this complex organic compound.

Herein, we report on the important IR changes accompanying newpolymorphic transformations in the DPC molecule during the controlled heattreatment of the samples. Also, a thermal histogram will be suggested fordescribing what happened in DPC during the transformation to severalpolymorphic states. Beside the IR analysis itself, X-ray diffraction (XR), nuclearmagnetic resonance (NMR) and mass spectroscometr (MS) are also used ascomplementary identification techniques. The present study aims to reveal thatthere is not only one, but several phase transformations in DPC during the heattreatment from ambient conditions up to a temperature of 160 °C.

El-Kabbany F. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 57–73 59

Experimental


Diphenyl carbazide (DPC) used for analysis was in powdered form and of ultra-pure grade (obtained from the British Drug House, Lab. Chem. Division; UK).KBr used for FTIR and the solvent DMSO for NMR measurements; both beingpurchased from Sigma-Aldrich.

Instrumental Techniques

IR spectra were recorded using the FT-IR spectrometer (model “Satellite 2000”,Perkin Elmer), when the respective experimental procedure has already beenreported [12]. The samples used were prepared in the form of discs by mixing 20mg of DPC with one gram of KBr in a cylindrical die of 10 mm diameter. The diewas then evacuated to ensure dryness and the sample was pressed. The spectrawere recorded within the wave-number region of 4000-400 cm–1 and run by meansof a computer-controlled system.

NMR analysis was performed in dimethyl sulphoxide (DMSO) selected asthe solvent of choice [12] and the spectra recorded using a commercial instrumentby Varian (model “Gemim 20”, operating at 200 MHz). Finally, the mass spectra(MS) of heat-treated DPC were obtained using a quadrupole mass spectrometer(GC-MS type, model “QP 1000 EX”; Shimadzu, Japan). The ions were formed inthe electron impact (EI) mode at 70 eV and with a temperature of the ions of about250 °C. The work took place in vacuum, and depending on the vapour pressure ofthe sample, the operational pressure was in the order of 10–4 Torr.


Survey and Commentary of the Experiments Performed

In previous investigations, the DPC compound was studied in amorphous form,including its characterization via the electrical and thermal properties when thelatter gave rather inconclusive results. In the present study, however, the powderedform of DPC yielded the thermograms that have clearly indicated that there areapparent phase changes at 50, 90, and 140 °C — see Fig. 1.

Also, obvious sudden changes have been observed in electrical properties(dielectric constant, d.c. resistivity, pyroelectric current, etc.) at 125 °C. Thus,according to both techniques, probable polymorphs at 50, 90, 125, and 140 °C inDPC can be expected. IR spectroscopic analysis could be used here to confirm the


Fig. 1 Differential scanning calorimetric curve obtained by analysing the compound ofinterest (diphenyl carbazide, DPC; C13H14N4O)

polymorphism in thermally treated DPC. The spectral change accompanying thepolymorphic transitions are obtained from 400 cm–1 up to 4000 cm–1 for originalsamples of DPC and thermally treated (previously melted DPC) from roomtemperature (R.T . 30 °C) up to 160 °C melting point (m.p. of DPC). IR analysisused here covered this temperature range.

Figure 2 shows the IR spectra of DPC samples at room temperature (beforeheat treatment Fig. 2A), and after melting and cooling to room temperature, Fig.2 B. One can easily observe a slight change in the mode frequencies betweenspectra A and B.

Also, there is a remarkable variation in the transmittance. This may be dueto a change in the crystal structure after melting. According to our previous workon DPC [10-13], the crystal structure of the R.T. phase of DPC (before heattreatment) is orthorhombic while that after heat treatment is amorphous. Thenormal modes of vibrations of these two phases are due to the ionic fundamentalsand given in Table I. The bands obtained are due to ionic fundamentals of N–H,C–H, N–N, and C=O.

In fact, however, there is also an alternate interpretation that can besummarized as follows

! 3359 – Stretching N–H vibration (in the case of secondary amides there is only1 band and it is impossible to have symmetric and asymmetric bands, which is


Fig. 2 IR spectroscopy of DPC; a) before heat treatment (on top) and b) after coolingdown to room temperature (below)

possible only for primary amides).! 3118 – Band belonging to the group of aromatic C–H stretching bands together

with the band at 3044 cm–1.! 3044 – Correctly interpreted as aromatic C–H stretching. ! 1659 – Band called Amide I. It is combination band of stretching C=O and

bending N–H vibrations.! 1592 – Band called Amide II. It is combination band of stretching C–N and

bending N–H vibrations.! 1486 – Band belonging to the aromatic C=C stretching system of bands.! 1287 – Band called Amide I. It is combination band of stretching C=O, bending

N–H and stretching C–N vibrations.! 748 – Out of plane C–H bending indicating the monosubstituted benzene.

Moreover, this mono-substitution is showed by two bands (748 and 691 cm–1


as one can see in Fig. 2). ! Finally, it is impossible to assign bands 1239 and 882 cm–1 to any characteristic

groups because of the fact that these bands are typical for the finger printregion of FTIR spectrum, i.e. these bands are joined with the vibration of theDPC molecule.

Table I Ionic fundamental vibrations of the DPC compound before and after melting

Normal vibrations of DPC due to ionicfundamentals

Orthorhombic DPC(at R.T.)

Amorphous DPC (atR.T.)

Asymmetric stretching of N–H 3359 3311

Symmetric stretching of N–H 3118 3097

Bending or deformation mode of N–H 1659 1677

Aromatic C–H stretching 3044 3052

Out of plane C–H bending 882 887

N–N stretching symmetric vibrations 1592 1602

N–N astretching symmetric vibrations 1486 1492

C–H stretching vibrations 1287 1306

CùO stretching vibrations 1239 1252

Monosubstituted benzene 748 756

Fig. 3 Detailed IR contour variation of asymmetric and symmetric stretching of N–Hand aromatic C–H bonds in consequence of polymorphism in the molecule ofDPC


Fig. 4 Detailed IR contour variation of bending mode of N–H, asymmetric andsymmetric stretching vibrations of N–N during the polymorphism in DPC

In order to support one of these hypotheses, it should be necessary to re-evaluate or even re-measure some data again. In this respect, the authors remainopen for further discussion; nevertheless, the following text will be consistent withthe first interpretation presented by means of Tables I and II.

The nature of temperature dependence of vibrating groups is of great


interest since its role is fundamental in explaining many of the physical propertiesof this organic compound. The spectrum shown in Fig. 2 A, indicates that thestrongest bands are at 3359 cm–1 and 3276 cm–1. In the amorphous phase Fig. 2 B,the strongest band is that at 3311 cm–1. It is due to the asymmetric stretching ofN–H, while the mode 3097 cm–1 indicates the symmetric stretching of N–H. Thearomatic C–H stretching mode appears here at 3044 and 3052 cm–1 for bothorthorhombic and amorphous phases respectively. The variation of the bandcontours of the three modes 3311, 3097 and 3052 cm–1 at temperatures 30, 80, 110,130 and 150 °C are those corresponding to different phase states of DPC. As

Fig. 5 Detailed IR contour variation of C-H stretching vibrations and C=O stretchingvibration during the polymorphism in DPC


shown in Fig. 3, the frequency and shape changes of these modes at the differenttemperatures are very good indication of the polymorphic changes in DPC duringheating.

Also, it is well known that the distinct shape of a transmittance band forparticular molecule relates to the number of ions that undergo the various energytransitions. Thus, the most intense absorbance line is that which represents thelargest number of ions undergoing a particular transition and the band envelopeas a whole is an indication of the total number of ions involved. As thetemperature of ions increases the band contour will change.

Figure 4 shows detailed variation of the band contours of bending mode ofN–H, symmetric and asymmetric stretching vibrations of N–N as DPC temperatureincreased from 30 °C to 160 °C. The change in mode frequency here is very little,but the change of the mode shape is very clear. The band contours also changesystematically during the temperature elevation and accompany the successivephase variation in DPC. Figure 5 shows the band contour variation of C–Hstretching vibrations and C=O stretching vibration during polymorphic change.Clearly, the mode shift is very little, but the change in band shape is obvious.

Figure 6 shows the variation of the bond contour of the out-of-plane C–Hbending mode during the elevation of temperature between 30 °C and 160 °C.Here the change in band shape is very clear. The out-of-plane C–H bending modeis a very sharp one. Its peak height changes alternatively during phase changes.This indicates that the C–H group may be responsible for phase change, i.e., theoriented motion of this group in space is responsible for phase transition in DPC.

Fig. 6 Detailed IR contour variation of the out-of-plane C–H bending vibration duringthe polymorphism in DPC

Figure 7 shows the variation of the band contour of mono-substituted benzeneduring a temperature variation from 30 °C up to 160 °C. The change in modefrequency and band shape for this band is very limited. The effect of heating onten modes of vibrations of DPC is summarized in Table II.

This effect is clearly notified in the first two stretching modes of(asymmetric and symmetric) vibrations of N-H. In the other modes of vibrations,


Fig. 7 Detailed IR contour variation of monosubstituted benzene moiety during thepolymorphism in DPC

Table II Normal mode vibrations during polymorphic changes in thermally treated DPC

Thermally treated UntreatedDPC

Phase number R.T.Phase

IV III II I

Crystal structure 1 2 3 4 1 4

Temperature, °C 30 80 110 130 150 R.T.

Asymmetric stretching of N–H 3311 3280 3297 3320 3325 3359

Symmetric stretching of N–H 3097 3099 3107 3125 3150 3118

Bending or deformation modeof N–H

3052 3048 3053 3049 3048 3044

Aromatic C–H stretching 1677 1676 1679 1679 1679 1659

Out of plane C–H bending 1602 1600 1601 1600 1599 1592

N–N stretching symmetricvibrations

1492 1488 1489 1489 1489 1486

N–N astretching symmetricvibrations

1306 1300 1302 1301 1299 1287

C–H stretching vibrations 1252 1248 1251 1249 1248 1239

CùO stretching vibrations 887 885 886 884 884 882

Mono-substituted benzene 755 753 755 755 754 748

1 – amorphous; 2 – monoclinic; 3 – tetragonal; 4 – orthorhombic


the effect of heating seems to be little. Such behaviour of N–H group indicates thatN–H group may be responsible for the polymorphism in DPC. This means that theoriented motion of N–H in space may control the phase state of DPC. Now, animportant question arises after these five polymorphic transitions: what is themolecular formula of the final compound? Is it still DPC or this reagent haschanged into a new organic compound? In order to put answer such a question,

Fig. 8 MS spectra of re-heated DPC (analysed at 70eV)


NMR and mass spectra of DPC were obtained by electron impact at 70 eV. Basedon the results of the microchemical analysis, Table III could be assembled andFigs 8 and 9 drawn.

Fig. 9 NMR spectra of re-heated DPC


Table III The relative abundance for the mass spectra of DPC by electron impact (at 70eV)

Mass Int, % Mass Int, % Mass Int, % Mass Int, %

50 10.3 51 32.0 52 11.3 53 6.6

54 2.2 55 0.8 56 0.4 57 1.0

59 0.1 60 2.3 61 1.6 62 3.1

63 6.1 64 6.1 65 31.8 66 17.4

67 1.2 69 0.4 70 0.7 71 0.2

73 0.7 74 2.1 75 0.2 76 1.8

77 70.1 78 15.5 79 5.1 80 4.4

81 1.4 82 0.6 83 0.7 84 0.5

85 0.1 87 1.0 88 0.2 90 1.1

91 8.4 92 32.7 93 100.0 94 13.3

95 1.5 97 1.2 105 9.7 106 11.3

107 27.9 108 35.0 109 2.7 110 0.5

111 0.6 117 0.4 120 0.7 121 0.4

133 0.8 134 5.9 135 1.0 149 0.8

150 2.6 151 8.5 152 6.5 153 0.4

167 0.8 169 1.5 240 1.0 242 4.8

243 1.5

The general molecular formula of the compound under investigation can begiven as C13H14N4O. This means that after all polymorphic transitions, DPC at 50,90, 125, and 140 °C still has the same molecular weight of 242 g mol–1, which isthe same as that of the original unheated DPC. But it belongs to another phasemodification and a new crystal structure. It seems that under the presentexperimental conditions the formation of this new crystal structures is an end ofthe thermodynamic phase transformation of the original DPC subjected to heattreatment processes, as shown in Fig. 10, illustrating the effect of heat and re-heattreatment of DPC and the resulting polymorphic phase changes.


Fig. 10 Effect of heating and re-heating treatment upon the polymorphism of DPC; Ascheme

Interpretation of the Results and Discussion on the Principal Observations

Polymorphic transitions in organic compounds are of particular interest especiallyfor pharmaceuticals and light sensitive compounds. However, polymorphism [14]is the property exhibited by certain substances changing, at certain temperature,their crystal symmetry class or structure. This is accompanied by changes inphysical characteristics [15]. In the framework of phenomenological theory, onecan describe the anomalies in thermodynamic quantities and also, determine thenature of the changes that occur in the spectrum of the elementary excitations ofthe crystal of DPC in the neighbourhood of the phase transitions. The orderparameter is related to the change that occurs in the positions of the atoms in the


crystal lattice during phase transition. The changes in the vibrational (phonon)spectrum of the crystal in the neighbourhoods of the critical point of such phasetransition are experimentally the most noticeable. A change in structure occurs indisplace – type phase transition as a result of the displacement of the ions (atoms)in the crystal lattice.

Here the IR spectra of thermally treated diphenyl carbazide are due to latticevibrations of the individual ions and internal vibrations of molecules. Since thevibrations occur through the lattice and are not connected with a single unit cell,these vibrations are often observed as broad peaks which represent a compositeband of several vibrations. In fact, the stability of DPC and its critical propertiesdepend on the extent of changes which occur in the hydrogen bonding of theamino ions (N–H) to the carbonyl (C=O) ions.

In a polyatomic molecule like DPC, each atom has three degrees of freedomin three directions which are perpendicular to one another. Consequently, apolyatomic molecule like that requires three times as many degrees of freedom asthe number of its atoms. In the case of DPC and according to the character ofvibration, the normal modes of vibrations of DPC can be divided into twoprincipal groups; stretching vibration and bending vibration. In the first type theatoms move essentially along the bond axis, so that the bond length increases ordecreases periodically, i.e., at regular intervals.

In the second type, there occurs a change in bond angles between bondswith a common atom or there occurs the movement of a group of atoms withrespect to the remainder of the molecule without movement of the atoms in thegroup with respect to one another. The first type appears at high frequencies whilethe second type appears at lower frequencies. In fact, in a polyatomic moleculelike DPC, the same bond can perform stretching and bending vibrationssimultaneously.

Conclusion

It can be stated that, according to De Ranter et al. [9], the DPC studied possessesan orthorhombic structure at room temperature. After melting (at . 160 °C), it thentransforms to the amorphous structure. Now, re-heat treatment of amorphous DPCcauses a set of phase transformations according to the following scheme


Thus, the effect of heat treatment on DPC is a full circle — it starts in theamorphous state and ends with amorphous state. The successive re-heat treatmentof DPC for several times, repeat the same phase transition cycle and always endswith amorphous state whatever the number of heating cycles passed. To checkbehaviour like that, an X-ray diffraction pattern for DPC has been performed,which is shown in Fig. 11, depicting the state after the first heat treatment cycle(A) and after the second cycle (B).

Fig. 11 X-ray diffraction patterns for DPC samples after A) first and B) second run ofheating

Both of these two X-ray diffraction pattern indicate amorphous structure forthe starting phase and final phase whatever the number of heating cycles passed.Such behaviour has already been depicted above — in Fig. 10.

Acknowledgements

The authors would like to thank Prof. Dr. Ivan Švancara and A. Prof. Dr. MartinAdam for their interest in the article, kind help in the final editing, and the formeralso for making this paper possible to be published in the Sci. Pap. Univ.Pardubice.


References

[1] Svehla Gy.: Vogel’s Qualitative Inorganic Analysis, 7th Edition; Longman,Singapore, 1996.

[2] Hassanein M., Fouda M.F.R., Amin R.S.: Thermochimica Acta, 132, 265(1988).

[3] Sundari D.S., Raghavendra A.S.: Photosynthesis Res. 23, 95 (1990).[4] Prasad S.M., Singh J.B., Rai L.C., Kumar H.D.: FEMS Microbiol. Lett. Fed.

Eur. Microbiol. Soc. 82, 95 (1991).[5] Melies A., Nemson J.A., Hanison M.A.: Biochim. Biophys. Acta 110, 312

(1992).[6] Mishra U., Kashyap A.K., Pande J.: Environ Technol. 14, 373 (1993).[7] Verma K., Singh D.P.: Current Microbiol. 3, 373 (1995).[8] Panigua A.R., Vazquez M.D., Tascon M.L., Batanero P.S.: Electroanalysis

5, 155 (1993).[9] De Ranter C.J., Blaton N.M., Peeters O.M.: Acta Cryst. B 35,1295 (1979).[10] El-Kabbany F., Taha S., Mansey F.M., Shehap A., Yousef M: J. Phys. Chem.

Solids 58, 449 (1997).[11] El-Kabbany F., Taha S., Shehap A., El-Naggar M.M.: J. Phys. Chem. Solids

59, 1614 (1998).[12] El-Kabbany F., Badr Y., Tosson M.: Physica Stat. Solidi A 63, 699 (1981).[13] El-Kabbany F., Taha S., Mansey F.M., Shehap A.: Infrared Phys. Technol.

38, 169 (1997).[14] Papon P., Leblond J., Meijer P.H.E.: The Physics of Phase Transitions;

Springer, Berlin, 2002.[15] Bernstein J.: Polymorphism in Molecular Crystals, Oxford Univ. Press,

2002.


75



16 (2010)

OPTICAL PROPERTIES OF Tl-DOPED Bi2Se3 SINGLE CRYSTALS

Petr JANÍČEKa, Čestmír DRAŠARa1, Anna KREJČOVÁb, Petr LOŠŤÁKc

and Jiří NAVRÁTILd

aInstitute of Applied Physics and Mathematics,bInstitute of Environmental and Chemical Engineering,

cDepartment of Inorganic Chemistry,dJoint Laboratory of Solid State Chemistry,

The University of Pardubice, CZ–532 10 Pardubice,


Bi2–xTlxSe3 single crystals with the Tl content of cTl = 0 to 5.2×1024 atoms m–3

prepared from the elements of 5N purity by means of a modified Bridgman methodwere characterized by the measurements of infrared reflectance andtransmittance. The values of the plasma resonance frequency Tp, opticalrelaxation time J, and high-frequency permittivity g4 were determined by fittingthe reflectance spectra using plasma-resonance formula. It was found that thesubstitution of Tl atoms for Bi atoms in the Bi2Se3 crystal lattice leads to adecrease in the wp values implying a decrease in the free carrier concentration.This effect is explained within a model of the point defects in the crystal lattice ofBi2–xTlxSe3. The dependences of the absorption coefficient K on the energy ofincident photons were determined from the transmittance spectra. The optical

76 Janíček P. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 75–85

width of the energy gap decreases with the increasing Tl content. The values of theexponent $ from the relation of K ~ 8$ for the long-wavelength absorption edgerange within the interval from 2.1 to 2.5 indicating the dominant scatteringmechanism of free current carriers is the scattering on acoustic phonons.

Introduction

Bismuth selenide Bi2Se3 belongs to the family of narrow-band semiconductors (where A = Bi, Sb and B = Se, Te) with tetradymite structure (space group

). It is the component of the solid solution, for example Bi2Te3–xSex andBi2–xSbxTe3–ySey that form the base materials for n-type room temperaturethermoelectric (TE) devices having dimensionless figure of merit ZT = FS2T/6 (F,S, 6 and T stand for electrical conductivity, Seebeck coefficient, thermalconductivity, and temperature, respectively) value close to 1 [1]. Since thephysical quantities are dependent on free carrier concentration, the influence ofdopants on the free carrier concentration is one of active research areas. Despiteconsiderable research on this material [2], there is a couple of papers on theinfluence of IIIB elements (Ga, In, Tl) on the properties of Bi2Se3 single crystals.There are a few papers concerning the doping with Ga [3] and In [4,5] and itsinfluence on the optical properties of Bi2Se3. While Ga increases the concentrationof free electrons N, In increases N just at low indium content whereas In leads toa decrease of N at higher concentrations. The effect of thallium doping was studiedonly in our previous paper [6], where the results of measurement of the transportcoefficients (Hall coefficient, electrical conductivity and Seebeck coefficient) werepublished. Tl doping results in a decrease of the free electron concentration as wellas in an increase of mobility of the free electrons in Bi2Se3 matrix. The effect ofTl doping on the optical properties of Bi2Se3 single crystals has not been studiedyet.

The aim of this paper is to describe the preparation of the single crystalsBi2–xTlxSe3 for x = 0-0.1 and their characterization by the measurement of thereflectance in the plasma-resonance frequency region and the transmittance ininfrared region. The observations are explained in terms of the point-defect model,which is based on the idea that the Tl impurities create substitutional defects andinfluence the concentration of native defects in Bi2Se3 as well.

Experimental

The single crystals of Bi2–xTlxSe3 with nominal x values between 0 and 0.1 havebeen grown using the Bridgman method. The polycrystalline precursor wassynthesized by heating of stoichiometric mixtures of 99.999 % pure Bi, Se and

Janíček P. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 75–85 77

TlSe. The synthesis of TlSe was carried out by heating stoichiometric mixtures of5N purity Se and Tl at the temperature of 800 K for 1 hour in evacuated quartzampoules. This material was powdered and combined with Bi and Se in the ratiocorresponding to the stoichiometry Bi2–xTlxSe3 (x = 0, 0.01, 0.02, 0.06 and 0.10)and synthesis of the polycrystalline product was carried out in evacuated conicalquartz ampoules in a horizontal furnace at the temperature of 1200 K for 48 hours.Then the samples were annealed at the temperature of 1100 K for 24 hours in thevertical Bridgman furnace and lowered into the zone with the temperature gradientof 125 K cm–1 with the rate of 1.3 mm h–1.

The obtained crystals were 60 mm long with 8 mm diameter and could beeasily cleaved. Their trigonal axis c was always perpendicular to the pullingdirection so that the (0001) plane was parallel to the ampoule axis. The orientationof the single crystal was carried out using the Laue back-diffraction techniquewhich confirmed that the planes were always (0001).

Samples for the reflectance and transmittance measurements were cut outfrom the central part of the crystals. Thin samples (with the thickness of several:m) for the transmittance measurements were obtained using an adhesive tape forcleaving thin layers from the crystal.

Spectral dependences of the reflectance R in the plasma resonancefrequency range were measured at room temperature in unpolarized light onnatural (0001) cleavage faces using an FT-IR spectrometer Bruker IFS 55. Thegeometry of the experiment was such that the electric field vector of theelectromagnetic radiation was always perpendicular to the trigonal c-axis, i.e.

.The measurements of the transmittance spectra on thin samples were carried

out at room temperature using the same spectrometer. The radiation wasunpolarized, but the orientation of the samples with respect to the incidentradiation always fulfilled the condition .

The actual concentration of thallium (cTl) in the samples for measurementsof reflectance and transmittance was determined using the atomic emissionspectrometry (AES).


Reflectance Spectra

The reflectance spectra are depicted in Fig. 1. For lucidity the curves of the dopedsamples are shifted by the constant values 0.3, 0.6, 0.9 and 1.2 for Bi1.99Tl0.01Se3,Bi1.98Tl0.02Se3, Bi1.94Tl0.06Se3, and Bi1.9Tl0.1Se3, respectively. It is evident that thereflectance curves reveal distinct minima which confirm a good singlecrystal quality. The positions of the minima are given in Table I. One can see


500 1000 1500 2000

0.5

1.0

1.5

2.0

5

4

2

1

R (

-)

ν (cm-1)

3

Fig. 1 Reflectance spectra of Bi2–xTlxSe3 single crystals (samples are labeled accordingto Table I). Symbols represent the experiment; the solid lines are fits accordingto plasma-resonance formula

Table I Optical parameters of Bi2–xTlxSe3 single crystals

SampleNo.

Nominalcomposition

cTl (AES)1024 m–3

Actualx

g4

1 Bi2Se3 0 0 771 29.0

2 Bi1.99Tl0.01Se3 0.2 3.2×10–5 752 29.5

3 Bi1.98Tl0.02Se3 1.1 1.6×10–4 734 28.0

4 Bi1.94Tl0.06Se3 4.4 6.41×10–4 696 26.5

5 Bi1.9Tl0.1Se3 5.2 7.69×10–4 677 25.5

Tp

1014 s–1J

10–14 sN/mz

1026 m–3Kmin

cm–1

1 Bi2Se3 1.373 4.75 1.72 913

2 Bi1.99Tl0.01Se3 1.340 4.50 1.67 886

3 Bi1.98Tl0.02Se3 1.285 4.40 1.45 770

4 Bi1.94Tl0.06Se3 1.215 4.50 1.23 653

5 Bi1.9Tl0.1Se3 1.210 4.95 1.17 621

the shift of the minima to lower wave numbers, e.g. higher wavelength withincreasing Tl content.

In order to obtain information on the changes in the free carrierconcentration, associated with the incorporation of Tl into Bi2Se3 lattice, theexperimental curves were fitted using equations for the real (g1) and theimaginary (g2) parts of the complex dielectric function [7]


(1)

(2)

where n is the index of refraction, k index of extinction, J optical relaxation time,g4 the high-frequency permittivity, and Tp the plasma resonance frequency. Forone type of carriers, the last quantity is given by the relation

(3)

where mzm0 is the free carrier effective mass in the direction perpendicular to thetrigonal axis c, N is the concentration of free carriers, and g0 is the permittivity ofthe free space.

It is necessary to mention that relation (3) is not consistent with theconclusion of paper [8] obtained from the Shubnikov de Haas effect measurementsat the temperature of liquid He, showing the splitting of conduction band of Bi2Se3into subbands ()E ~ 40 meV). However our measurements performed at roomtemperature cannot determine the influence of the conduction band splittingbecause the value )E is comparable with the energy kT.

Approximate trial values of g4, J and Tp were inserted in Eqs (1) and (2),and a suitable computer program was used to minimize the function

(4)

where RN is the experimental reflectance and R is the calculated value, given by therelation

(5)

In Fig. 1 experimental values of reflectance RN are represented by points,theoretical dependence R acquired by the best fit of experimental values is


represented by lines. This procedure gives nice fits to the experimental results, ascan be seen from Fig. 1. The values of N/mz were calculated from Tp usingEq. (3).

From Table I it is evident that the values of N/mz decrease with increasingTl content in the crystals (provided that mz is constant it applies also for N).Incorporation of Tl atoms into the Bi2Se3 matrix results in the suppression of thefree electron concentration. The conclusion presented above on the change of thefree carrier concentration caused by Tl atoms entering the Bi2Se3 lattice issupported by the results of the measurements of the transport parameters presentedin our previous paper [6]. We observed that with the increasing content of Tlatoms in the Bi2–xTlxSe3 samples the absolute values of Hall coefficient and Seebeck coefficient increase.

Transmittance Spectra — Absorption Coefficient

As an illustration of the transmittance measurements, the plots arepresented for two samples with the highest content of Tl in Fig. 2 and 3. Well-developed interferences and a relatively high transmittance indicate a good opticalquality of the prepared samples.

Fig. 2 Transmitance spectra of Bi1.9Tl0.1Se3 single crystal

To determine the absorption coefficient K as a function of the incidentphoton energy the transmittance T and reflectance R values corresponding to thelocations of the interference maxima in the plot were used in the formula

(6)

1000 20000

10

20

T (

%)

ν (cm-1)

Bi1.9Tl0.1Se3

d = 8.2 µm


0.1 0.2 0.3

6.5

7.0

7.5

8.0

lnK

(cm

-1)

hν (eV)

1 4 5

1000 20000

10

20

T (

%)

ν (cm-1)

Bi1.94

Tl0.06

Se3

d = 11.2 µm

Fig. 3 Transmitance spectra of Bi1.94Tl0.06Se3 single crystal

Fig. 4 Spectral dependences of absorption coefficient K of Bi2–xTlxSe3 single crystals(samples are labeled according to Table I)

where d is the sample thickness. For lucidity the results are shown only for x = 0,0.06 and 0.1 in Fig. 4. Values of K for samples with x = 0.01 and 0.02 are veryclose to pure Bi2Se3.

A decrease in the free carrier concentration with increasing Tl content in theBi2–xTlxSe3 samples, following from the above interpretation of the reflectancespectra, should result also in the decrease in the value of absorption coefficient Kwith increasing Tl content. This effect is evident in Fig. 4. The value of Kmin in theminimum of the dependence is equal to 913 cm–1 for pure Bi2Se3,


2.0 2.5

6.5

7.0

7.5

lnK

(cm

-1)

lnλ (µm)

Bi2Se

3

Bi1.94

Tl0.06

Se3

Bi1.90

Tl0.10

Se3

whereas Kmin = 621 cm–1 for the Bi1.9Tl0.1Se3 (see also Table I).It is further evident from Fig. 4 that the presence of thallium impurities

results also in a shift of the short-wavelength absorption edge towards lowerenergies. This implies that the incorporation of Tl atoms into the Bi2Se3 latticeleads to a decrease in the optical gap . Also this fact is in a good agreementwith the conclusion that the thallium impurity in the Bi2Se3 lattice decreases theconcentration of the free charge carriers. In accordance with the Moss–Bursteineffect, i.e. with the shift of the short-wavelength absorption edge resulting fromthe change of the free carrier concentration N, with decreasing N we expect a shiftof the edge towards lower energies, hence a decrease in the values of .

By the extrapolation of the short wavelength edge (K2 vs. photon energy)to the K = 0 for pure Bi2Se3 one gets the value of the energy gap ~ 0.24 eV(not shown here). This value is in good accordance with previously publisheddata [9].

Fig. 5 Plots of ln K vs. ln 8 for Bi2–xTlxSe3 single crystals

It is known that the shape of the long-wavelength edge gives informationon the scattering mechanism of the free charge carriers. With the objective todetermine the values of the exponent $ in the relation K ~ 8$ we have plotted thecurves lnK vs. ln 8 in Fig. 5. The values of $ obtained from these plots range from2.1 to 2.5. This indicates that in Bi2–xTlxSe3 single crystals at temperatures close to300 K the dominant scattering mechanism of the free carriers is the scattering bythe acoustic branches of lattice vibrations. The fact that the obtained values of $are higher than 3/2, which in covalent crystals correspond to the scattering byacoustic phonons [10], can be explained assuming that in our mixed crystals wedeal with a mixed scattering mechanism by acoustic phonons and ionizedimpurities. This view is in full agreement with the conclusion of the paper by


Stordeur et al. [11] i.e. that at room temperature the crystal lattice of Bi2Se3 basedmaterials is characterized by the mixed anisotropic scattering mechanism of thefree carriers by acoustic phonons and ionized impurities.

Point Defects in Bi2–xTlxSe3 Single Crystals

As mentioned above, doping with thallium leads to the decrease of the freeelectron concentration in Bi2–xTlxSe3 single crystals. In this paragraph we willqualitatively discuss the findings. Our consideration is based on the followingpoints. The single crystal of Bi2Se3 prepared from stoichiometric melt shows anexcess of bismuth over the stoichiometric composition [12]. According to the ideaspresented in [13], this is the reason why the crystal lattice contains negativelycharged antisite (AS) defects of bismuth atoms on selenium sites andpositively charged vacancies in the selenium sublattice. The concentration of

vacancies [ ] exceeds the concentration of AS defects [ ] and the Bi2+*Se3

crystals exhibit n-type conductivity, the concentration of free electrons [e–] isgiven by the difference of the concentration of both native defects, i.e.

(7)

The results of XRD analysis presented in [6] reveal that the volume of theunit cell decreases monotonously with the increasing thallium content. We assumethat thallium atoms enter the sublattice of bismuth forming TlBi substitutionaldefects. Since indium is presented in valence state +3 in Bi2Se3 [5], we assumeinitially, by analogy, that thallium is also present in this valence state, i.e. it formsuncharged defects

(8)

( and SeSe stand for bismuth vacancy and selenium atom on ordinarysite). It is evident that this way thallium produces no free carriers. Howeverthallium in state +3 can be easily reduced to valence +1 which is more stable formof thallium from the chemical point of view, i.e. it can capture two electrons fromconduction band and reduce the free electron concentration

(9)

The comparison of the actual and nominal concentration of thallium in thesamples gives another potential explanation of the change of the free carrier


concentration. The actual concentration of thallium is smaller by two orders ofmagnitude than the nominal one (see Table I). Actually we deal with dopedcrystals not solid solutions as perhaps suggested by the notation of the nominalcomposition. Only a small percentage of thallium was incorporated into the hoststructure, the rest was segregated on the top of the single crystal in the form of apolycrystalline lump [6]. The X-ray diffraction revealed two structures in thelump, Bi2Se3 and TlBiSe2 [6]. The stoichiometry of the later implies that the singlecrystals grew from a nonstoichiometric melt, namely from the melt that showexcess of selenium over the stoichiometric composition. Such growth conditionslead to a crystal with suppressed concentration of selenium vacancies .According to equation (7) this fact can account for the observed decrease in thefree electron concentration.

We note that both the formation of func and the descending vacancies concentration can account for the decrease of the free electron concentration.

The experimental results do not allow us to specify the dominant mechanism.However, in the view of a noticeable increase in free carrier mobility [6] webelieve that the suppression of selenium vacancies is significant. A suppressionrather than increase in the defect concentration is consistent with the increase inthe free carrier mobility.

Conclusion

The single crystals of Bi2–xTlxSe3 (cTl = 0-5.2×1024 atoms m–3) were prepared usingthe Bridgman technique. The optical properties were measured at roomtemperature. The results obtained from the measurement of reflectance in theplasma resonance frequency region and transmittance in the IR region ofBi2–xTlxSe3 single crystals lead to following conclusions:1. Substitution of Bi atoms by Tl atoms in the Bi2Se3 crystal structure results in

a decrease in the concentration of the free electrons.2. The suppression of the free electron concentration can be ascribed to the point

defects — type and to a decrease of concentration of selenium vacancies.

3. The shift of the short-wavelength absorption edge to the lower energies, i.e.decrease in the optical gap as well as decrease in the absolute value of theabsorption coefficient K with increasing Tl content confirm the decrease in theconcentration of free electrons.

4. Mixed anisotropic scattering mechanism of the free carriers by acousticphonons and ionized impurities was determined from the shape of long-wavelength absorption edge.


Acknowledgements

The research was supported by the Ministry of Education, Youth and Sports of theCzech Republic under project No. MSM 0021627501.

References

[1] Nolas G.S., Sharp J., Goldsmid H.J.: Thermoelectrics, Basic Principles andNew Materials Developments, p. 123, Springer, Berlin, 2001.

[2] Krost A.: in Landolt-Börnstein (New Series), Group III, Vol. 17f, p. 268,Berlin/Heidelberg, New York/Tokyo, 1989.

[3] Horák J., Lošťák P., Montaner A.: Phys. Stat. Sol. (b) 119, K17 (1983).[4] Lošťák P., Novotný R., Urbanová E., Horák J.: Phys. Stat. Sol. (a) 113, 615

(1989).[5] Lošťák P., Beneš L., Civiš S., Süssmann H.: J. Mater. Sci. 25, 277 (1990).[6] Janíček P., Drašar Č., Beneš L., Lošťák P.: Cryst. Res. Technol. 44, 505

(2009).[7] Madelung O.: Handbuch der Physik, Vol. 20, (Ed. S. Flügge) p. 210,

Springer, Berlin, 1957.[8] Kulbachinskii V.A., Miura N., Nakagawa H., Arimoto H., Ikaida T., Lošťák

P., Drašar Č.: Phys. Rev. B 59, 15733 (1999).[9] Hashimoto K., Fyc. Sci-Kyushu Univ. B 2, 141 (1958).[10] Yakovlev A.: Fiz. Tverd. Tela 2, 1621 (1960).[11] Stordeur M., Ketavong K.K., Priemuth A., Sobotta H., Riede V.: Phys. Stat.

Sol. (b) 169, 505 (1992).[12] Offergeld G., Cakenberghe J.: J. Phys. Chem. Solids 11, 301 (1959).[13] Sklenář A., Drašar Č., Krejčová A., A., Lošťák P.: Crys. Res. Technol. 35,

1069 (2000).



87



16 (2010)

DETERMINATION OF COPROPORPHYRIN IAND III IN HUMAN URINE USING HPLC WITH

FLUORESCENCE DETECTION

Roman KANĎÁR1, Pavla Žáková, Martina FRÖHLICHOVÁ,Xenie ŠTRAMOVÁ and Soňa BENCOVÁ

Department of Biological and Biochemical Sciences,The University of Pardubice, CZ–532 10 Pardubice,

Received June 17, 2010

The determination of selected porphyrins, especially of coproporphyrin I and IIIis presently the most reliable direct approach to the screening of porphyrias. AnHPLC method for the simultaneous measurement of coproporphyrin I and III inurine samples has been developed and evaluated. Separation of coproporphyrinI and III from interfering substances was achieved on reversed phase HPLC withgradient elution and fluorescence detection. Analytical performance of thismethod is satisfactory for both coproporphyrin I and coproporphyrin III: theintra-assay and inter-assay coefficients of variation were below 10 %.Quantitative recoveries from spiked urine were between 85.0 and 101.1 %. Thelimit of detection was 3.0 and 2.3 :g l–1, respectively. The preliminary referenceranges of coproporphyrin I and III in a group of donors are 5.0-51.6 :g l–1

88 Kanďár R. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 87–98

(0.59-3.20 :g mmol–1 creatinine) and 14.9-65.4 :g L-1 (0.98-8.21 :g mmol–1

creatinine). The presented method is inexpensive and suitable for screening ofporphyrias.

Introduction

Porphyrins are cyclic tetrapyrroles derived from porphin by substitution of theperipheral positions with various side-chains (Fig. 1). The measurement ofporphyrin levels in urine, blood, feces and tissues is very important for thediagnosis of the porphyrias, diseases due to enzyme deficiencies in the hembiosynthetic pathway [1-4]. It is necessary to point out that abnormal porphyrinmetabolism is also found in alcoholics [5], lead and chemical intoxication [6,7],iron deficiency, chronic infection, inflammation and malignancy [8]. A greatnumber of different methods for the measurement of porphyrin levels in biologicalsamples have been described [9]. The most powerful of published methods seemsto be HPLC with fluorescence detection [10-15]. On the other hand, only a fewmethods use HPLC with UV/Vis [16] or electrochemical [17] detection. Theporphyrins may be separated by adsorption chromatography as methyl esters [18,19] or as free acids [20,21] by reversed-phase chromatography. Methods for thedetermination of methyl esters require using purification steps prior to HPLCanalysis. This problem can be avoided by measuring of the porphyrins as free acids[22].

This article describes a sensitive HPLC with fluorescence detection andquantification of coproporphyrin I and III in human urine using very rapid andsimple procedure of sample preparation.

Fig. 1 Structures of coproporphyrin I (A) and coproporphyrin III (B). R: -CH2-CH2-COOH

Kanďár R. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 87–98 89


Reagents and Chemicals

Uroporphyrin I dihydrochloride, uroporphyrin III octamethyl ester, coproporphyrinI dihydrochloride, coproporphyrin III tetramethyl ester, hydrochloric acid, sodiumdihydrogenphosphate, sodium hydrogenphosphate, sodium acetate, andammonium acetate were obtained from Sigma Chemical Company (St. Louis, MO,USA). HPLC-gradient grade ethanol, methanol, and acetonitrile were from MerckKgaA (Darmstadt, Germany). Lyophilized urine porphyrins controls, Level I andII, Lot 433, were from Recipe (Munich, Germany). All the other chemicals wereof analytical grade.

Since porphyrins are light sensitive, precautions were taken to minimizeexposure of standard and urine samples to daylight by using either amber plastictubes or plastic tubes protected with aluminum foil. Stock solutions of selectedporphyrins were prepared in methanol (about 100 mg l–1), evaporated to drynessunder nitrogen (Linde Gas, Prague, the Czech Republic), and stored at –80 °Cuntil used.

Instrumentation

The chromatographic analysis was performed with a liquid chromatograph(Shimadzu, Kyoto, Japan), LC-10ADvp solvent delivery system, SIL-10ADvpautosampler, CTO-10ASvp column oven, RF-10Axl fluorescence detector andSCL-10Avp system controller. Data were collected digitally with Claritychromatography software (DataApex, Prague, the Czech Republic).

Subject and Samples

Samples of urine were obtained from a group of donors (n = 40, 20 women in theage 21-41 years, mean age 33 years, and 20 men in the age 21-53 years, mean age39 years). None of the studied subjects exhibited renal, hepatic, gastrointestinal,pulmonary or oncological diseases. A written informed consent was obtained fromall participants before starting the protocol and the Hospital Committee on HumanResearch (Regional Hospital of Pardubice, the Czech Republic) approved thestudy.


Sample Preparation

Dry residues of porphyrin stock solutions were dissolved in 200 :l hydrochloricacid (6.0 mol l–1) and incubated at 25 °C for 24 h in dark. The concentrations ofindividual porphyrins were verified on a Shimadzu UV-1700 PharmaSpecspectrophotometer (Shimadzu, Kyoto, Japan), using molar absorptivities at 406 nm(1 cm path-length) of 4.2×105 l mol–1 cm–1 for uroporphyrins and at 401 nm of4.7×105 l mol–1 cm–1 for coproporphyrins.

Twenty-four-hour urine samples were protected against daylight. To 500 :lurine sample or standard, 50 :l concentrated hydrochloric acid was carefullyadded. The mixture was then filtered through a 0.20 :m nylon filter, 4 mmdiameter (Supelco, Bellefonte, PA, USA) and transferred into 0.2 ml crampedamber vial.

Fig. 2 Gradient of mobile phase: mobile phase B percent vs. time (offset 2.2 min toallow for gradient front reaching the mixer of the column). Mobile phase A,mixture of 100 mmol l–1 ammonium acetate-methanol (90:10, v/v), pH 6.5 ± 0.1;mobile phase B – 100% methanol

Chromatography Analysis

Selected porphyrins were separated using a LiChroCART 125×4 mm i.d.,Purospher STAR RP-18e, 5 :m analytical column fitted a LiChroCART 4×4 mmi.d., Purospher STAR RP-18e, 5 :m guard column (Merck KgaA, Darmstadt,Germany). All separations were performed at 37 °C using the gradient shown inFig. 2. Mobile phase A was a mixture of 100 mmol l–1 ammonium


acetate-methanol (90:10, v/v), pH 6.5 ± 0.1. Mobile phase B was 100 % methanol.Prior to use, both phases were vacuum filtered and degassed using ultrasound. Theflow rate was kept constant at 0.5 ml min–1. The optimum response of measuredporphyrins was observed when the excitation and emission wavelengths were394 nm and 624 nm, respectively. The amount of porphyrins was quantified fromthe corresponding peak area using Clarity chromatography software (DataApex).The concentration of porphyrins in the urine samples was determined from thecalibration curve.

Additional Analyses

Creatinine in the urine was measured with the set Creatinine Flex® by standardprocedures using an automatic biochemistry analyzer Dimension® RxL Max®

(Siemens Healthcare Diagnostic Ltd., Deerfield, IL, USA).

Statistical Analysis

The data are presented as mean values ± S.D. Differences between women andmen were analyzed with the use of the Mann-Whitney Rank Sum Test and analysisof correlation was carried out using Spearman Rank Order Correlation (softwareQCexpert, Trilobyte, Pardubice, the Czech Republic and software SigmaStat,Systat Software, Inc., Point Richmond, CA, USA). A p < 0.05 value wasconsidered statistically significant.


We describe here an HPLC assay for selected porphyrins. This work is focused ondescribing our efforts at optimizing sample preparation with regard to samplestability and recovery of the analytes. We found that porphyrins are optimallymeasured by using HPLC with fluorescence detection after acidification withconcentrated hydrochloric acid.

Sample Preparation for HPLC of Porphyrins and Stability

A suitable sample preparation technique is essential for accurate analysis ofporphyrins. Uroporphyrins and coproporphyrins can easily be measured by directinjection of cleaned urine. Then a guard column has to be used to protect theanalytical column. Moreover precipitates, such as calcium salts, tend to adsorb


porphyrins. With a view to dissolve the precipitated material we have added theconcentrated hydrochloric acid to urine. Acid treatment also prevents theformation of metalloporphyrins. With regard to very low levels of other porphyrinsas are hepta-, hexa-, and pentacarboxyl porphyrins in human urine and thepresence of other metabolites with similar structures and properties, it is necessaryto use purification steps prior to their determination. These steps have been basedon either SPE or liquid-liquid extraction. This techniques, however, does not givegood recovery of porphyrins [19,22,23]. The stability of the selected porphyrinswas studied under various conditions during 12 hours. We found out thaturoporphyrins and coproporphyrins in the samples of acidified urine were stableon a cooled autosampler (4 °C, dark) for at least 12 hours.

Fig. 3 An HPLC chromatogram of uroporphyrin I (2.4 :g l–1), uroporphyrin III(3.3 :g l–1), coproporphyrin I (40.1 :g l–1) and coproporphyrin III (58.4 :g l–1) inhuman urine. Peaks: (1) uroporphyrin I, (2) uroporphyrin III, (3) coproporphyrin Iand (4) coproporphyrin III. HPLC conditions: a gradient elution [mobile phase A:mixture of 100 mmol l–1 ammonium acetate-methanol (90:10, v/v), pH 6.5 ± 0.1;mobile phase B: 100 % methanol]; stationary phases were a columnLiChroCART 125×4 mm i.d., Purospher STAR RP-18e, 5 :m and a guardcolumn LiChroCART 4×4 mm i.d., Purosper STAR RP-18e, 5 :m. The flow ratewas kept constant at 0.5 ml min–1, separation ran at 37 °C and porphyrins weremonitored at excitation and emission wavelengths of 394 nm and 624 nm,respectively


HPLC Analysis of Coproporphyrin I and III

An HPLC chromatogram of selected porphyrins in a urine sample is shown in Fig.3. The analytical parameters of coproporphyrin I and III analysis are shown inTable I and II. To determine the within-day precision, the urine sample wasanalyzed 10 times on the same day under the same conditions. Similarly, thebetween-day precision obtained on the same urine sample was analyzed on10 different days. Precision and recovery studies were carried out using normaladult urine for coproporphyrin I and III, aliquoted and stored at –20 °C. Sampleof this urine was spiked with coproporphyrin I and III to obtain levels of theseporphyrins in pathological range. Coefficients of variation were below 10 %. Thespike recoveries ranged between 85.0 and 101.1 %. The calibration curve (13-point for a determination of analytical parameters and 8-point for a routineanalysis was linear in the whole range tested: 5.0-500.0 :g l–1 (Fig. 4 and 5). The

Table I Precision and recovery of coproporphyrin I for urine samples analysis. *Mean oftriplicate assays is recorded

A) Precision (within-day)

n Mean ± S.D.:g l–1

CV%

10 44.1 ± 2.1 4.8

10 214.5 ± 4.9 2.3

B) Precision (between-day)


CV%

10 46.8 ± 3.4 7.3

C) Recovery

Added, :g l–1 Observed:g l–1*

Recovery%

0.0 44.3 ± 2.0

12.7 55.1 ± 2.1 85.0

63.7 100.5 ± 3.6 88.2

127.0 163.6 ± 5.6 93.9

317.5 365.3 ± 11.3 101.1

Mean 92.1 ± 7.1

CV 7.7


calibration curve values obtained from the combination of five standard curves areshown in Table III. The lowest concentration that can be quantified withacceptable accuracy and precision was 5.0 :g l–1 for both coproporphyrin I andcoproporphyrin III. Furthermore, the limit of detection for coproporphyrin I andIII, defined as a signal-to-noise (S/N) ratio of 3:1, was 3.0 and 2.3 :g l–1.

Porphyrins were separated on a reverse-phase column using a gradientsystem of methanol and ammonium acetate. The mobile phase was optimized inorder to obtain the best separation of the analytes in the shortest time. Standardsolutions of porphyrins as well pooled urine were used for a study of the mobile

Table II Precision and recovery of coproporphyrin III for urine samples analysis. *Mean oftriplicate assays is recorded



CV%

10 48.2 ± 2.9 6

10 252.2 ± 6.8 2.7



CV%

10 51.5 ± 4.0 7.8

C) Recovery

Added, :g l–1 Observed:g l–1*

Recovery%

0 51.4 ± 2.2

11.3 61.3 ± 2.6 87.6

56.5 104.3 ± 4.1 93.6

113 160.2 ± 5.8 96.3

282.5 337.0 ± 9.8 101.1

mean 94.7 ± 5.6

CV 5.9

phase composition and a gradient program. Several mobile phases (namelyphosphate and acetate buffers containing methanol, ethanol and acetonitrile) andseveral gradients were assayed and the best results were obtained for theconditions described in the experimental. The retention behavior was studied in


Table III Calibration curve values for the HPLC method. *Thirteen-point for a determinationof analytical parameters and eight-point for a routine analysis

Standard Regression equation Mean slope95 % confidence

interval

Intercept**

:mol l–1

95 % confidenceinterval

Correlationcoefficient

COP-I* y = 20.535 x – 19.488 20.535 0.9 0.9998

(20.155 -20.915)

(0.5-1.1)

COP-III* y = 22.201 x – 16.341 22.201 0,7 0.9999

(21.877-22.525) (0.4-1.0)** x-intercept (:mol l–1) is point at which the line crosses x-axis (where y value equals 0)

Fig. 4 Typical standard curve for HPLC quantification of coproporphyrin I Regressionequation: y = 20.535x – 19.488; R = 0.9998; A = 20.535 (S.D. = 0.155);B = –19.488 (S.D. = 30.803)

dependence of pH value of the mobile phase A in the range of 4.0-7.0. The optimalpH 6.5 was chosen for the best separation and detection of selected porphyrins.Column temperature was adjusted from 25 to 45 °C. Optimal temperature intervalwas from 35 to 40 °C. The criteria were the resolution, the stability of thefluorescence intensity and the analysis speed. Pursuant to records, we can establishthat presented method for the determination of coproporphyrin I and III in humanurine is highly robust.


Fig. 5 Typical standard curve for HPLC quantification of coproporphyrin III .Regressionequation: y = 22.201x – 16.341; R = 0.9999; A = 22.201 (S.D. = 0.132);B = –16.341 (S.D. = 25.568)

Determination of Coproporphyrin I and III in Human Urine of Donors

The preliminary reference ranges of coproporphyrins I and III in a group of donorsare 5.0-51.6 :g l–1 (0.59-3.20 :g mmol–1 creatinine) and 14.9-65.4 :g l–1

(0.98-8.21 :g mmol–1 creatinine). Our normal reference intervals for urinecoproporphyrins are similar to those of others [20,24,25]. We found no significantdifferences in both coproporphyrin I and coproporphyrin III concentration betweenwomen and men (0.95 ± 0.27 :g mmol–1 creatinine vs. 1.02 ± 0.23 :g mmol–1

creatinine, p = 0.389 and 1.82 ± 0.83 :g mmol–1 creatinine vs.1.91 ± 0.89 :g mmol–1 of creatinine, p = 0.411, respectively). We observed nosignificant correlation between age and concentration (R = 0.142, p = 0.695).

As coproporphyrins predominate both in normals and patients withporphyrias (especially in acute intermittent porphyria, porphyria cutanea tarda,porphyria variegata, hereditary coproporphyria, and erythropoetic protoporphyria)[24], we have measured only them. In contrast to uroporphyrin I and III,coproporphyrin I and III were identified in all donors.


Conclusion

We developed a rapid, simple, and very selective HPLC method with fluorescencedetection for the determination of coproporphyrin I and III in human urine. Thismethod provides an inexpensive alternative to those using time-consumingpurification steps prior to HPLC analysis. It affords excellent sensitivity, precision,and accuracy in follow-up samples and is suitable for both research and clinicaltrials, especially within porphyrias screening.

Abbreviations used: SPE – solid phase extraction

Acknowledgements

The Ministry of Education, Youth and Sports of the Czech Republic (MSMT0021627502) and (SGFChT 07) supported this work.

References

[1] Macours P., Cotton F.: Clin. Chem. Lab. Med. 44, 1433 (2006).[2] Nordmann Y., Puy H.: Clin. Chim. Acta 325, 17 (2002).[3] Kaupinnen R.: Lancet 365, 241 (2005).[4] Gross U., Hoffmann G.F., Doss M.O.: J. Inherit. Metab. Dis. 23, 641 (2000).[5] Teshke R., Goys-Könemann C., Daldrup T., Goerz C.: Biochem. Pharm. 36,

1133 (1987).[6] Lamola A.A., Joselow M., Yamane T.: Clin. Chem. 21, 93 (1975).[7] Sinclair P.R., Granick S.: Biochem. Biophys. Res. Commun. 61, 124 (1974).[8] Rossi E., Garcia-Webb P.: Biomed. Chromatogr. 1, 163 (1986).[9] Zaider E., Bickers D.R.: Clin. Dermatol. 16, 277 (1998).[10] Božek P., Hutta M., Hrivnáková B.: J. Chromatogr. A 1084, 24 (2005).[11] Hahn M.E., Gasiewicz T.A.: J. Chromatogr. 563, 363 (1991).[12] Schoenfeld N., Mamet R.: Clin. Chem. 40, 417 (1994).[13] García-Vargas G.G., Hernández-Zavala A.: Biomed. Chromatogr. 10,

278 (1996).[14] Zuijderhoudt F.M., Koehorst S.G., Kluitenberg W.E., Dorresteijn-de Bok J.:

Clin. Chem. Lab. Med. 38, 227 (2000).[15] Respaud R., Benz-de Bretague I., Blasco H., Hulot J.S., Lechat P., Le

Guellec C.: J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 837, 3893(2009).

[16] Magi E., Ianni C., Rivaro P., Frache R.: J. Chromtogr. A 905, 141 (2001).


[17] Li F., Lim C.K., Peters T.J.: Biochem. J. 243, 621 (1987).[18] Straka J.G., Kushner J.P., Burnham B.F.: Anal. Biochem. 111, 269 (1981).[19] Kotal P., Jirsa M., Martásek P., Kordač V.: Biomed. Chromatogr. 1,

159 (1986).[20] Schreiber W.E., Raisys V.A., Labbé R.F.: Clin. Chem. 29, 527 (1983).[21] Sakai T., Niinuma Y., Yanagihara S., Ushio K.: Clin. Chem. 29, 350 (1983).[22] Lim C.K., Li F., Peters T.J.: J. Chromatogr. 429, 123 (1988).[23] Malina L., Miller V., Magnus I.A.: Clin. Chim. Acta 83, 55 (1987).[24] Hindmarsh J.T., Oliveras L., Greenway D.C.: Clin. Biochem. 32, 609 (1999).[25] Elder G.H., Smith S.G., Smyth S.J.: Ann. Clin. Biochem. 27, 395 (1990).


99



16 (2010)

SIMULTANEOUS DETERMINATIONOF VANILLYLMANDELIC AND HOMOVANILLIC

ACIDS IN URINE USING LC WITH COULOMETRICELECTROCHEMICAL DETECTION BASED

ON SPE SAMPLE PREPARATION

Roman KANĎÁR1and Pavla ŽÁKOVÁaDepartment Biological and Biochemical SciencesThe University of Pardubice, CZ–532 10 Pardubice

Received January 5, 2010

The aim of the study was to develop a simple and selective LC method for thesimultaneous measurement of vanillylmandelic acid (VMA) and homovanillic acid(HVA) in human urine. A method for the measurement of VMA and HVA using LCwith coulometric electrochemical detection and investigation into the extractiontechniques with regard to stability and recovery are described. The analyticalperformance of presented method is satisfactory. The presented method uses arelatively rapid SPE procedure. It can be used both for research and for clinicaltesting purposes.

100 Kanďár R., Žáková P./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 99–109

Introduction

Vanillylmandelic acid (VMA) and homovanillic acid (HVA) are the majormetabolites of catecholamines. Epinephrine and norepinephrine are metabolizedto VMA by the combined action of catechol-O-methyltransferase and monoamino-oxidase. HVA represents the main metabolite of dopamine [1-3]. Urine VMA andHVA are often measured in diagnostics of neuroblastoma, pheochromocytoma andneurological diseases [4, 5].

Numerous analytical methods for the measurement of VMA and HVA inurine have been described [6-13]. With regard to the presence of other metaboliteswith similar structures and properties, it is necessary to use purification steps priorto their determination. These steps have been based on either SPE or liquid-liquidextraction (LLE). SPE and LLE are often time-consuming and a source of errors.Final quantification is most often performed using GC and LC. GC methods needa sample derivatization step, which is time-consuming [7,14]. Different LCmethods have been described, mainly with electrochemical [8,15,16] orfluorescence detection [17]. LC with electrochemical detection, using eitheramperometric or coulometric electrodes, can measure VMA and HVA directly.Moreover, these techniques avoid typical problems associated with derivatizationprocedures. Companies Chromsystems (Munich, Germany) and Recipe (Munich,Germany) have marketed commercial LC reagent sets for the determination ofVMA and HVA. Although LC methods for the simultaneous determination ofVMA and HVA have been substantially improved and simplified, they are stillexpensive hence unsuitable for routine analyses. Recently the application of LC-MS or LC-MS-MS methods for the determination of VMA and HVA has beenreported [12,13]. The advantage of LC-MS methods is a rapid assay withoutsample pretreatment; the disadvantage is mainly their expensiveness. Forscreening of large numbers of samples are available enzyme immunoassay (EIA)methods (6). The immunoassays are problematical in cross-reactivity, andmonoclonal antibodies are expensive.

In this paper, we describe a sensitive LC with coulometric electrochemicaldetection for simultaneous determination of VMA and HVA using an internalstandard and a relative rapid and accuracy SPE procedure. Moreover, this methodis inexpensive and applicable to routine analyses.


Reagents and Chemicals

Vanillylmandelic acid (4-hydroxy-3-methoxymandelic acid, VMA), iso-vanillyl-mandelic acid (3-hydroxy-4-methoxymandelic acid, iVMA), homovanillic acid (4-

Kanďár R., Žáková P./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 99–109 101

hydroxy-3-methoxyphenylacetic acid, HVA), 5-hydroxyindoleacetic acid (5-HIAA), hydrochloric acid, sodium hydroxide, orthophosphoric acid, acetic acid,sodium acetate, ammonium acetate and sodium hydrogenphosphate were obtainedfrom Sigma Chemical Company (St. Louis, MO, USA). AG 1-X8 resin, 100-200mesh, chloride form, was purchased from Bio-Rad Laboratories (Hercules, CA,USA). HPLC gradient-grade methanol, ethanol, and acetonitrile were from MerckKgaA (Darmstadt, Germany). Lyophilized urine endocrine controls (normal range,Lot No.: 195; pathological range, Lot No.: 155) were from Chromsystems(Munich, Germany). All the other chemicals were of analytical grade.

VMA, iVMA and HVA solutions were prepared daily in 0.1 mol l–1

hydrochloric acid (5-HIAA in water) and stored at 4 °C until used.

Instrumentation

The liquid chromatograph consisted of LC-10ADvp solvent delivery system(Shimadzu, Kyoto, Japan), injection valve Rheodyne 9725i (Rheodyne, L.P.,Rohnert Park, CA, USA), with 10-:l loop and Coulochem® III electrochemicaldetector (ESA Laboratories, Inc., Chelmsford, MA, USA). The data were collecteddigitally with Clarity chromatography software (DataApex, Prague, the CzechRepublic).

Subject and Samples

Samples of urine were obtained from a group of donors (n = 40, 20 women in theage 23-49 years, mean age 36 years, and 20 men in the age 21-52 years, mean age38 years). None of the studied subjects exhibited renal, hepatic, gastrointestinal,pulmonary or oncological diseases. A written informed consent was obtained fromall the participants before starting the protocol and the Hospital Committee onHuman Research (Regional Hospital of Pardubice, the Czech Republic) approvedthe study.

Sample Preparation

iso-Vanillylmandelic acid was used as an internal standard because it has similarproperties to measured VMA and HVA.

Twenty-four-hour urine samples, protected against daylight, were acidifiedwith concentrated hydrochloric acid.

One milliliter of acetate buffer (0.1 mol l–1, pH 6.1 ± 0.1) containing internalstandard (1.5 :mol l–1 of iVMA) was carefully added to 50 :l urine or spiked urine


standard. After vigorous mixing (20 s) and pH adjustment (between 5 and 7), themixture was applied on the SPE column.

Solid-Phase Extraction

We have used SPE procedure for removal of interfering compounds.Polypropylene SPE tube with a frit and a column volume of 3 ml (Supelco,

Bellefonte, PA, USA) were filled with 0.5 ml of AG 1-X8 anion-exchange resin.The sorbent was tightened with another frit and washed with deionized water untila negative reaction to chloride ions, which was tested by the addition of 300 :l 0.3mol l–1 AgNO3 to 1 ml effluent. The column was equilibrated stepwise (by 2 ml)with 4 ml acetate buffer (0.1 mol l–1, pH = 6.1 ± 0.1) at a flow-rate of 1 ml min–1.This flow-rate was achieved by centrifugation (780 × g, 2 min, room temperature).Then the mixture (1 ml) was applied on the equilibrated column. The SPEprocedure was performed according to the sequence shown in Table I. Collectedeffluent was diluted with a mobile phase (50 :l of effluent and 450 :l of mobilephase) and filtered through a 0.20-:m nylon filter (4 mm diameter, Supelco,Bellefonte, PA, USA). Columns were regenerated stepwise (by 2 ml) with 6 mlHCl (1.0 mol l–1), which removed bound anoins. After about 20 ml washing withdeionized water and testing of negative reaction to chloride ions, the columns wereready for the next run.

Table I The solid-phase extraction procedure

Our SPE Commercial SPE

Column equilibrated withacetate buffer (0.1 mol l–1,pH = 6.1±0.1) [2×2 ml]

- -

Diluted urine applied tocolumn

Diluted urine applied tocolumn

Discard effluent

Column washed withethanol-2 mol l–1, acetic acid(1:1) [2×2 ml]

Column washed with Buffer I[3 ml]

Discard effluents

Column washed with acetatebuffer (0.1 mol l–1, pH =6.1±0.1) [2×2 ml]

Column washed with Buffer II[2×3 ml]

Discard effluents

VMA, iVMA and HVAeluted with 0.5 mol l–1

H3PO4 [2×1 ml]

VMA, iVMA and HVA elutedwith Elution Buffer

[2 ml]


Solid-Phase Extraction Using Set from Chromsystems

We have compared our SPE method with a commercial SPE method developed byfirm Chromsystems (Munich, Germany). After addition of internal standard (1.0ml) to urine or spiked urine standard (50 :l) and vigorous mixing (20 s), themixture (1 ml) was applied on the SPE column. The SPE procedure was performedaccording to the sequence shown in Table I and the flow-rate (1 ml min–1) wasachieved by centrifugation (720 × g, 2 min, room temperature). Collected effluentwas diluted with a mobile phase (50 :l effluent and 450 :l mobile phase) andfiltered through a nylon filter.

Chromatographic Analysis

Chromatography of VMA and HVA was accomplished using an isocratic elutionon a Discovery® HS C18, 250 × 4 mm i.d., 5 :m analytical column (Column#:54951-01; Bonded Phase Lot#: 5084; Silica Lot#: 020419BQ) fitted Discovery®

HS C18, 20 × 4 mm i.d., 5 :m guard column (Supelco, Bellefonte, PA, USA) at thetemperature of 30 °C. The mobile phase consisted of 5 % methanol in 25 mmol l–1

Na2HPO4 (v/v), pH = 6.0 ± 0.1. The flow-rate was kept constant at 0.5 ml min–1.VMA and HVA were detected following LC separation with a Coulochem® IIIdetector equipped with a dual analytical cell (model 5010) and a guard cell (model5020). The guard cell was connected in line before the injector and was used toremove oxidizable impurities in the mobile phase. The dual analytical cell containstwo porous graphite electrodes in series. A carbon filter was placed before a guardcell and between the injector and guard column, a PEEK filter between theanalytical column and analytical cell. For optimum detection of VMA and HVA,the electrode potentials for the guard cell, E1, and E2 were set at +450 mV, +180mV, and +400 mV, respectively. The gain range was 200 nA. The hydrodynamicvoltammogram analysis was performed to optimize conditions for the accuratedetermination of VMA and HVA. It was developed by injection of 10 :l mixedsolution of VMA and HVA (0.1 mmol l–1) and measuring the current produced byVMA and HVA at the electrodes. Before injection of the first sample, the potentialat the electrodes (except the guard cell) was increased stepwise from +100 mV inabout 40 mV increments to the final working potential, and LC system wasequilibrated for approximately 2 hours with the mobile phase at a flow-rate of 0.5ml min–1. When not running samples overnight, flow-rate of mobile phase was setat 0.1 ml min–1 with voltages +100 mV, +200 mV, and +250 mV, respectively.After about 50 injects, the electrode potentials were set first at +600 mV, then at–400 mV, at each electrode for 1 hour with the mobile phase at a flow-rate of 0.5ml min–1, followed by 20-min water rinse and by 60-min methanol rinse at a flow-rate of 0.5 ml min–1 (electrodes off). This procedure was used for removal of


impurities from electrodes and achieving electrode sensitivity and baselinestability.

Statistical Analysis

The data are presented as mean values ±S.D. Differences between women and menwere analyzed with the use of the Mann-Whitney Rank Sum Test and analysis ofcorrelation were carried out using Spearman Rank Order Correlation (softwareQCexpert, Trilobyte, Pardubice, the Czech Republic). A p < 0.05 value wasconsidered statistically significant.


The Effectiveness of SPE, Stability

We have tested the effectiveness of our and commercial SPE procedure. Althoughcommercial SPE procedure is as rapid and simple as our method, it has lessrecovery and VMA was with difficulties separated from interfering compounds.SPE of the samples was studied using a strong AG 1-X8 anex. The volume, flow-rate and composition of solutions were studied with the aim of removing theinterferences with minimum loss of VMA, iVMA and HVA. The volumes assayedwere from 1 to 3 ml at a flow-rate of 1-3 ml min–1. The optimum volume and flow-rate were 2 ml and 1 ml min–1, respectively. The column was equilibrated withacetate buffer (0.1 mol l–1, pH = 6.1 ± 0.1); it has similar composition as thediluted urine. Different mixtures of ethanol and acetic acid (2.0 mol l–1) wereassayed within clean-up step . The 1:1 ethanol-acetic acid mixture provided thebest results, removing the interferences and causing practically no loss of VMA,iVMA and HVA. A solution of 0.5 mol l–1 orthophosphoric acid provided the bestelution of the analytes. The VMA and HVA content in samples extracted usingSPE was stable at 4 °C for at least 12 hours.

In comparison with the commercial SPE procedure, our SPE procedure isinexpensive. With careful regeneration of resin, we used the same SPE columnmore than ten times (intra-assay with CV below 5 % and inter-assay with CVbelow 10 % both VMA and HVA).

Some LC methods use time-consuming extraction techniques [15] oranalyze only diluted urine samples [16,18]. When crude samples are injected onan LC column, resolution and column life decrease. Compared to some publishedextraction techniques, our SPE procedure uses eco-friendly extraction reagents.Organic solvents such as ethyl acetate [7,16] or methanol [12,13] are consideredas significant pollutants.


LC Assay of VMA and HVA

Representative LC chromatogram of VMA and HVA in human urine is shown inFig. 1. The analytical parameters of VMA and HVA analysis are shown inTables II and III. To determine the within-day precision, the urine samples wereanalyzed ten times on the same day under the same conditions. Similarly, thebetween-day precision obtained on the same urine samples, were analyzed for12 consecutive days. Precision and recovery studies were carried out using normalurine for VMA and HVA, aliquoted and stored at the temperature of –20 °C.Samples of this urine were spiked with VMA and HVA to obtain levels of theseanalytes in the pathological range. The calibration curves were linear in the wholerange tested: (2.0-200.0) :mol l–1. The regression lines obtained from thecombination of ten standard curves were y = 0.0181x – 0.0066 for VMA and y =0.0199 x – 0.0094 for HVA. The mean slope, intercept, and correlation coefficient(r) for the calibration curves were 0.0181 (95 % confidence interval,0.0156–0.0207), 0.4 :mol l–1 (0.2–0.7 :mol l–1), and 0.9992 for VMA, 0.0199 (95% confidence interval, 0.0141–0.0258), 0.5 :mol l–1 (0.3-0.8 :mol l–1), and 0.9993for HVA. The lowest concentration that can be quantified with acceptableaccuracy and precision was 2.0 :mol l–1 [47.6 fmol inject–1] (intra-assay with CV= 7.8 %) for VMA and 2.0 :mol l–1 [47.6 fmol inject–1] (intra-assay with CV = 9.7%) for HVA. Furthermore, the limit of detection for VMA and HVA, defined assignal-to-noise (S/N) ratio of 3:1, was 0.4 :mol l–1 (9.5 fmol inject–1) and 0.8 :mol

Fig. 1 Representative LC chromatogram of VMA (38.6 :mol l–1) and HVA (40.2 :moll–1) in human urine. Peaks: VMA (6.7 min); iVMA (9.1 min) and HVA (25.6min). Peak (22.5 min) corresponds to 5-hydroxyindoleacetic acid. LC conditions:isocratic elution (5 % methanol in 25 mmol l–1 Na2HPO4, pH 6.0 ± 0.1).Stationary phase was an analytical column Discovery® HS C18, 250 × 4 mm i.d.,5 :m. The flow-rate was kept constant at 0.5 ml min–1 and separation ran at 30the temperature of °C


Table II Precision of VMA and HVA for urine samples analysis


VMA HVA

N Mean ± SD:mol l–1

CV%

Mean ± SD:mol l–1

CV%

10 22.5 ± 0.9 4.0 25.3 ± 0.8 3.2

10 79.4 ± 2.9 3.7 97.3 ± 2.7 2.8


VMA HVA

N Mean ± SD:mol l–1

CV%

Mean ± SD:mol l–1

CV%

12 21.6 ± 1.7 7.8 24.4 ± 2.0 8.2

12 78.1 ± 5.8 7.4 93.1 ± 7.1 7.6

Table III Recovery of VMA and HVA for urine samples analysis. *Mean of triplicate assays isrecorded

Recovery

VMA HVM

Added:mol l–1

Observed:mol l–1*

Recovery%

Observed:mol l–1*

Recovery%

0 23.5 ± 0.9 - 27.1 ± 1.2 -

2 25.3 ± 1.1 90.0 28.9 ± 1.2 90.0

5 28.1 ± 1.2 92.0 31.7 ± 1.3 92.0

20 42.1 ± 1.9 93.0 45.9 ± 2.1 94.0

50 71.5 ± 2.9 96.0 76.4 ± 3.1 98.6

100 120.8 ± 4.7 97.3 129.5 ± 4.9 102.4

Mean 93.7 ±2.7

Mean 95.4 ±4.5

CV 2.8 CV 4.7

l–1 (19.0 fmol inject–1), respectively.VMA, iVMA and HVA were separated on a reverse-phase column using an

isocratic system of methanol and Na2HPO4. The mobile phase was optimized in


order to obtain the best separation of the analytes in the shortest time. Standardsolutions of VMA, iVMA and HVA as well urine were used for study of themobile phase composition. Several mobile phases (namely different bufferscontaining methanol, ethanol and acetonitrile) were assayed and the best resultswere obtained for the conditions described in the experimental. The columntemperature was varied from 25 to 45 °C. Optimal temperature interval was from28 to 33 °C. The criteria were the resolution, electrode sensitivity, baselinestability and the analysis speed.

Determination of VMA and HVA in Human Urine

The levels of VMA and HVA in a group of donors are (19.7 ± 6.9) :mol d–1 and(23.3 ± 14.8) :mol d–1, respectively. We found no significant differences in VMAand HVA concentration between women and men (18.2 ± 10.3 :mol d–1 versus21.7 ± 9.1 :mol d–1, p = 0,137 and 20.8 ± 17.1 :mol d–1 versus 24.7 ± 15.1 :mold–1, p = 0.226). We observed no significant correlation between VMA and HVAconcentrations and age (R = 0.085, p = 0.443; R = 0.067, p = 0.489).

The aim of this study was to develop and validate LC method withcoulometric electrochemical detection for the simultaneous determination of VMAand HVA in human urine. The major advantages of electrochemical detection areespecially the simultaneous determination of VMA, HVA and 5-HIAA and arelatively rapid sample preparation without time-consuming pre-columnderivatization. Coulometric electrochemical detection offers several advantagesover the commonly used amperometric electrochemical detection [15]. First of all,in the amperometric mode about 5 % of the analyte is oxidized on the surface ofthe electrode; with a coulometric detector, close to 100 % of the analyte isoxidized in dual flow-through porous graphite electrodes. Within past 3 years wehave carried out more than 1500 analyses. By virtue of careful clean-up of urinesamples and a high sensitivity of flow-through porous graphite electrodes we caninject a small amount of enough diluted sample. Under these conditions, thelifetime of the analytical column is at least 1500 injects. The presented method issuperior to similar methods in that it is inexpensive (about 2500$ per1000analyses; Costs of guard cell, analytical cell, analytical column, mobile phase, SPEtubes, anion-exchange resin, nylon filters and chemicals for sample preparationcosts) and suitable for clinical tests.

Conclusion

We developed a simple and very selective LC method with coulometricelectrochemical detection for simultaneous determination of VMA and HVA inhuman urine. Careful pretreatment of human urine samples allow using of a simple


and inexpensive LC method. SPE procedure is relatively rapid and accurate.

Abbreviations used: VMA – vanillylmandelic acid; iVMA – isovanillylmandelicacid; HVA – homovanillic acid

Acknowledgements

This work was financially supported by the Ministry of Education Youth andSports of the Czech Republic (MSMT 0021627502) and (SGFChT 07).

References

[1] Kopin I.J.: Pharmacol. Rev. 37, 333 (1985).[2] Kagedal B., Goldstein D.S.: J. Chromatogr. 429, 177 (1988)[3] Eisenhofer G., Aneman A., Hooper D., Rundqvist B., Friberg P.: J

Neurochem. 66,1565 (1996).[4] Young J.R. Jr., Ries L.G., Silverberg E., Horm J.W., Miller R.W.: Cancer

58,598 (1986).[5] Tuchman M., Morris C.L., Ramnaraine M.L., Bowers L.D., Krivit W.:

Pediatrics 75,324 (1985).[6] Taran F., Bernard H., Valleix A., Créminon C., Grassi J., Olichon D.,

Deverre J.R., Pradelles P.: Chim. Acta 264, 177 (1997).[7] Fauler G., Leis H.J., Huber E., Schellauf C., Kerbl R., Urban C., Gleispach

H.: J. Mass Spectrom. 32, 507 (1997).[8] Radjaipour M., Raster H., Liebich H.M.: Eur. J. Clin. Chem. Clin. Biochem.

32, 609 (1994).[9] Tokuda T., Tokieda T., Anazawa A., Yoshioka M.: J. Chromatogr. 530, 418

(1990).[10] García A., Heinanen M., Jiménez L.M., Barbas C.: J. Chromatogr. A 871,

341 (2000).[11] Adcock J.L., Barnett N.W., Costin J.W., Francis P.S., Lewis S.W.: Talanta

67, 585 (2005).[12] Magera M.J., Thompson A.L., Matern D., Rinaldo P.: Clin. Chem. 49, 825

(2003).[13] Lionetto L., Lostia A.M., Stigliano A., Gardelli P., Simmaco M.: Clin. Chim.

Acta 398, 53 (2008).[14] Adami A., Guarnieri V., Margesin B., Mulloni V., Vincenzi D.: Biosens.

Bioelectron. 20, 1968 (2005).[15] Parker N.C., Levtzow C.B., Wright P.W., Woodard L.L., Chapman J.F.: Clin.

Chem. 32, 1473 (1986).


[16] Mashige F., Ohkubo A., Matsushima Y., Takano M., Tsuchiya E., KanazawaH., Nagata Y., Takai N., Shinozuka N., Sakuma I.: J. Chromatogr. B Biomed.Appl. 63, 63 (1994).

[17] Wielders J.P., Mink J.K.: J. Chromatogr. 310, 379 (1984).[18] Bonfigli A.R., Coppa G., Testa R., Testa I., De Sio G.: Eur. J. Clin. Chem.

Clin. Biochem. 35, 57 (1997)


111



16 (2010)

USE OF NEW REACTIVE DYES FOR DYEINGOF POLYAMIDE 6 AND WOOL

Vlasta LIŠKOVÁa1, Tiago A. D. PINTOb and Ladislav BURGERTa

and Radim HRDINAc

aInstitute of Chemistry and Technology of Macromolecular Materials,bDepartment of Chemistry, The University of Minho, PT–471 057 Braga,

cInstitute of Organic Chemistry and Technology,The University of Pardubice, CZ–532 10 Pardubice

Received January 5, 2010

This paper describes the research of newly designed and prepared reactive dyesfor dyeing of polyamide 6 and wool. For this purpose, disazo reactive dyes withsulfatoethylsulfonyl group have been synthetized by the sequence of diazotationof aromatic amines and coupling reactions of formed diazonium compounds withsecondary components (1-naftylaminesulfonic acids). In order to increase somecolouristic properties of reactive dyes, such as fixation, exhaustion and wetfastness, the reactive group of the dyes was modified with sarcosine in severalcases to form N-methyl-N-carboxymethyl-2-aminoethylsulfonyl group. In addition,dichloro-triazine dyes in modified and unmodified form have been tested as well.It was find out that the modification of reactive group with sarcosine improvedsignificantly the colouristic properties of some reactive dyes applied to both

112 Lišková V. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 111–132

investigated textile materials. Nevertheless, this method was not found to beuniversal for all the types of investigated reactive dyes. What can further be seenfrom the presented results is that among all the of investigated dyes some havebeen found to show especially high values of fixation that can overcome veryfrequent and common problems of low fixation occurring during the applicationof reactive dyes to dying of polyamide.

Introduction

At present, reactive dyes play an important role in textile chemistry. The termreactive dye normally refers to a day applicable to cotton, because the technicalimportance of reactive dyes is essential for cotton. As the consequence ofecological limitations, reactive dyes are also applied to the wool and nylon tosubstitute metal complex dyes.

Commercial reactive dyes for wool and nylon are usually based on thereactive aliphatic double bond joined with an electron acceptor group, wheredouble bond reacts with a nucleophilic group of a substrate “followingantimarkovnikov rule”. These are for example, derivatives of vinylsulfonyl group(-SO2CH2CH2-X = -SO2CH=CH2 + HX) or the reactive group 2-bromoacryl amidyl(-NHCOCH(Br)=CH2) and its precursor 2,3-dibromopropionamidyl (-NHCOCH(Br)CH2Br), where the elimination of HBr takes place.

The vinylsulfonyl reactive dyes (VS) are well known as demonstrated by thepatent publications [1-8]. The vinyl sulfone form is the proper reactiveintermediate stage in the alkaline application of $-substituted ethyl sulfone dyes.VS dyes in the protected form (for example as β-sulfatoethyl sulfones) wereoriginally developed for the reactive dyeing of wool by Hoechst chemists [9].They are storage stable under standard conditions, and the reactive vinylsulfonyldye (chromophore-SO2CH=CH2) is liberated by the reaction with an alkali in thedyebath. Vinylsulfonyl dye reacts subsequently with a nucleophilic group of afibre by Michael addition. Hoechst also invented derivatives of VS dyes formingthe reactive vinylsulfonyl group in an acid dyebath. A water soluble secondaryamine such as N-methyltaurin is used for the protection of vinylsulfonyl group(commercial name, for example, Procilan E dyes of ICI) and the fixation reactionwas described by Zollinger [10].

Precursors of vinylsulfonyl dyes can be generally described by the formuladye-(SO2CH2CH2-Y)n, where Y represents a substituent that can be split off. Forthe acid dyeing this Y respectively HY can be every secondary aliphatic amine.Secondary amines without water-soluble group (such as diethylamine) [11] arehygienically dangerous because they volatilise with steam. Thus HY are watersoluble secondary aliphatic amines of the formula HN(R)-(CH2)n-A, where Arepresents the sulfonic or the carboxylic acid group or a their water soluble salt[12]. Patent literature also describes the dyeing mixtures for the reactive dyeing

Lišková V. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 111–132 113

of synthetic polyamides, where water soluble aliphatic amines containing sulfonicor carboxylic acid groups are used as “additives” to VS dyes [13]. Similarly, thereaction of vinylsulfonyl dyes with secondary aliphatic amines containing sulfonicor carboxylic acid groups is known [14].

The aim of this work was to prepare and study reactive dyes for dyeing ofpolyamide 6 and wool and to increase some colouristic properties such as fixation,exhaustion and wet fastness of the prepared VS dyes of the formula chromophore–SO2CH2CH2N(CH3)CH2COOH where N-methylglycine (sarcosine) was used assecondary amine. Their evaluations are compared with other VS dyes.

Experimental

Instrumentation

All the dyeing experiments were carried out using laboratory dyeing machineScourotester Computex (Budapest, Hungary) and Labomat Mathis (Mathis,Switzerland). Absorbance values were measured using UV/VIS spectrophotometerHelios Gamma Thermospectronic (UNICAM, England) working in a single beamarrangement within spectral range from 190 to 1100 nm and with the slit width2 nm.

Reagents

Propan-1-ol, propan-2-ol, ethyl acetate, NaCl, Na2CO3, NaHCO3 and NaNO2, allof pro analysis grade (p.a.), were obtained from Lachema, Pty. (Brno, the CzechRepublic). N-methylpyrrolidon (p.a.) and PCl3 (p.a.) were obtained from Merck(Darmstadt, Germany). Hydrochloric acid (35 % w/v, p.a.) from Lach-Ner, Ltd.(Neratovice, the Czech Republic), acetic acid (99 % w/v, p.a.) and NaOH (p.a.)from Penta, Ltd. (Chrudim, the Czech Republic) were also used. Other chemicalreagents used for synthesis of dyes are listed in Table I.

Synthesis

The disazo dyes were prepared by the sequence of diazotization of aromaticamines and coupling reactions of formed diazonium compounds with secondarycomponents (1-naphtylaminesulfonic acids). The next step was the diazotizationof monoazo dye and the second coupling reaction with a secondary component.Reactive dyes contained 2-sulfatoethyl-sulfonyl (–SO2CH2CH2OSO3Na) reactivegroup capable to liberate reactive vinyl group in the dye-bath under alkaliconditions. The third synthetic step was the preparation of N-methyl-


NH2

SO2CH2CH2OSO3H

OH NH2

SO3HHO3S

N-carboxymethyl-2-aminoethyl-sulphonyl (–SO2CH2CH2N(CH3)CH2COOH)reactive group capable to liberate reactive vinyl group in the application dye bathunder acidic conditions.

Table I Reagents and chemicals used for preparation of dyes

Acronym Formula Supplier

PAFSES

2-[(4-aminophenyl)sulfonyl]ethylhydrogensulfate

Synthesia, Ltd.(Pardubice, the Czech

Republic

MAFSES

2-[(3-aminophenyl)sulfonyl]ethylhydrogensulfate

Spolchem, Pty. (Ústí nadLabem, the Czech

Republic)

H-acid

4-amino-5-hydroxynaphthalene-2,7-disulfonic acid

Synthesia, Ltd.

Cleve acid(1,6)

5-aminonaphthalene-2-sulphonic acid

Synthesia, Ltd.

Schaefferacid

6-hydroxynaphthalene-2-sulphonic acid

Synthesia, Ltd.

BON acid

3-hydroxy-2-naphtoic acid

Synthesia, Ltd.


NN

VS

NN

OH

SO3Na

VS = SO2CH2CH2OSO3Na

Sarcosine Merck

Preparation of Reactive Dye 1

Diazotization of PAFSES: PAFSES (C8H11NO6S2; M = 281.3 g mol–1; 96.6 %,0.02 mole; 5.82 g) was mixed with 50 ml distilled water and 5.14 ml concentratedHCl (35 %). The dispersion was externally cooled (ice and acetone) to 0-5 °Cunder intensive stirring and then 4.1 ml 5 M NaNO2 was added. The reaction timewas approximately 20 min and the reaction was accelerated using an ultrasoundbath (120 W, 1 min of dispergation). When the reaction was completed (checkedby iodine starch paper) a small excess of HNO2 was removed by addition ofNH2SO3H. The dispersion of diazonium compound was immediately used for thenext coupling reaction.

First Coupling Reaction: 1-Naphthylamine (C10H9N; M = 143 g mol–1; 0.02 mole;2.86 g) was mixed with 30 ml water and 2 ml concentrated HCl (35 %) and thetemperature was maintained at 0-5 °C (external cooling) under intensive stirring.Then the diazonium compound was gradually added and the reaction pH 5 wasadjusted by an addition of 5 M NaOH. The reaction time was approximately2 hours. When the reaction was completed (checked by the reaction with H-acid,drop reaction on filter paper), the monoazo dye MAD 1 (brownish red) wasfiltered off and the press cake was used without drying for the next diazotization.

Diazotization of MAD 1: The monoazo dye MAD 1 was mixed with 80 ml waterand 5.2 ml concentrated HCl (35 %). The dispersion was externally cooled to 0-5 °C and 4 ml 5 M NaNO2 were added under intensive stirring. The reaction timeof this diazotization was approximately 3 hours. When the reaction was completed,the dispersion of diazonium compound was used for the next coupling reaction.

Second Coupling Reaction: Schaeffer acid (C10H8O4S; M = 224.2 g mol–1;0.02 mol; 4.48 g) was mixed with 45 ml water and 4 ml 5 M NaOH to formsolution. After that the temperature was set at 0-5 °C (external cooling) and thediazonium compound from the previous step was added under intensive stirring.


NN

VSA

NN

OH

SO3Na

VSA = SO2CH2CH2NCH3

CH2COONa

NN

VS

SO3Na

NN

OH

SO3Na


pH of the reaction mixture was adjusted at 8 by an addition of 5 M NaOH. Thereaction time was approximately 30 min. When the reaction was completed, pHof the mixture was adjusted at 5 by an addition of concentrated HCl (35 %). Afterthat, 40 g NaCl was added (precipitation of dye was checked by drop test on filterpaper) and the mixture stirred for approximately 20 min. Finally, the dye wasfiltered off and carefully dried at 40 °C.

Thin layer chromatography (TLC) was used to check the composition of theprepared dye. (ALUGRAM SIL G/UV254, mobile phase: propan-2-ol/propan-1-ol/ethylacetate/water = 2/4/1/3 by vol.).

Preparation of Reactive Dye 1A

The reactive dye 1 (C28H19N4O10S3Na2; M = 714.7 g mol–1; 0.01 mole, 7.14 g) wasmixed with 50 ml water and with an excess of N-methylglycine (C3H7NO2; M =89.1 g mol–1; 0.02 mole; 1.78 g) which was added under intensive stirring. Then,pH of the reaction mixture was set at the value of 8 (5M NaOH), and the reactionmixture was stirred at the temperature of 60 °C for 3 hours. When the reaction wascompleted (checked by TLC, solid phase: ALUGRAM SIL G/UV254, mobile phase:propan-2-ol/propan-1-ol/ethylacetate/water = 2/4/1/3 by vol.), the final dye wassalted out by an addition of 20 g NaCl and carefully dried (40 °C)



VSA

NN

NN

SO3Na

OH

SO3Na

VSA = SO2CH2CH2NCH3

CH2COONa

Diazotization of PAFSES: PAFSES (C8H11NO6S2; M = 281.3 g mol–1; 96.6 %;0.02 mole; 5.82 g) was diazotized in the same way as in the case of the reactivedye 1.

First Coupling Reaction: Cleve acid (C10H9NO3S; M = 223.2 g mol–1; 0.02 mole;4.46 g) was mixed with 40 ml water and 4 ml 5M NaOH, and the temperature wasmaintained at 0-5 °C (external cooling) under intensive stirring. Then thediazonium compound was gradually added and pH of the reaction was adjusted at5 by an addition of 5M NaOH. The reaction time was approximately 2 hours.When the reaction was completed (checked by the reaction with H-acid, dropreaction on filter paper), the pH of the reaction mixture was increased to 8 by anaddition of 5M NaOH and the red solution of monoazo dye MAD 2 was used forthe next reaction step (diazotization).

Diazotization of MAD 2: The solution of monoazo dye MAD 2 was mixed with5.2 ml concentrated HCl (35 %). The very fine suspension was externally cooledto 0-5 °C and 4.1 ml 5M NaNO2 was added under intensive stirring. The reactiontime of this diazotization was about 2 hours. When the reaction was completed, thesuspension of diazonium compound was used for the next coupling reaction.

Second Coupling Reaction: Schaeffer acid (C10H8O4S; M = 224.2 g mol–1;0.02 mole; 4.48 g) was mixed with 45 ml water and 4 ml 5M NaOH. After that, thetemperature was set to 0-5 °C (external cooling) and the diazonium compoundfrom the previous step was added under intensive stirring. Thereafter, pH of thereaction was adjusted at 8 by an addition of 5M NaOH. The reaction time wasapproximately 30 min. When the reaction was completed, pH of the mixture wasadjusted at the value of 5 by addition of concentrated HCl (35 %), and NaCl (40g) was added (precipitation of dye was checked by drop test on filter paper). After20 min of stirring the dye was filtered off and carefully dried (40 °C).

The purity of prepared dye was determined by TLC (solid phase:ALUGRAM SIL G/UV254, mobile phase: propan-2-ol/propan-1-ol/ethyl-acetate/water = 2/4/1/3 by vol.).



NN

NN

OH

SO3NaVS

SO3NaVS = SO2CH2CH2OSO3Na

The reactive dye 2 (C28H19N4O13S4Na3; M = 816.7 g mol–1; 0.01 mole; 8.16 g) wasmixed with 50 ml water and an excess of N-methylglycine (C3H7NO2; M =89.1 g mol–1; 0.02 mole; 1.78 g) was added under intensive stirring. After that, pHof the reaction mixture was maintained at 8 (5M NaOH) and the reaction mixturewas stirred at the temperature of 60 °C for 3 hours. When the reaction wascompleted (checked by TLC, solid phase: ALUGRAM SIL G/UV254, mobile phase:propan-2-ol/propan-1-ol/ethylacetate/water = 2/4/1/3 by vol.), the final dye wassalted out by the addition of 20 g NaCl and carefully dried at 40 °C.


Diazotization of MAFSES: MAFSES (C8H11NO6S2; M = 281.3 g mol–1; 0.02 mole;5.63 g) was mixed with 50 ml distilled water and with 5.14 ml concentrated HCl(35 %) and the dispersion was then cooled externally to 0-5 °C using a mixture ofice and acetone. After that, 4.1 ml 5M NaNO2 was added under intensive stirring.The reaction took approximately 20 min; ultrasound bath was additionally used fordispergation (120 W, 1 minute of dispergation). When the reaction was completed(checked with iodine starch paper) a small excess of HNO2 was removed by theaddition of NH2SO3H. The dispersion of diazonium compound was immediatelyused for the next coupling reaction.

First Coupling Reaction: Cleve acid (C10H9NO3S; M = 223.2 g mol–1; 0.02 mole;4.46 g) was mixed with 40 ml water and 4 ml 5M NaOH, and the temperature ofthe reaction mixture was maintained at 0-5 °C (external cooling) under intensivestirring. Then the diazonium compound was gradually added and pH 5 of thereaction mixture was adjusted by an addition of 5M NaOH. After 2 hours, whenthe reaction was completed (checked by the reaction with H-acid, drop reaction onfilter paper), pH of the reaction mixture was increased to 8 by addition of5M NaOH, and the red solution of monoazo dye MAD 3 was used for the nextreaction step (diazotization).

Diazotization of MAD 3: The solution of monoazo dye MAD 3 was mixed with5.2 ml concentrated HCl (35 %). The very fine suspension was externally cooled


NN

NN

OH

SO3NaVSA

SO3NaVSA = SO2CH2CH2N

CH3

CH2COONa

to 0-5 °C and 4.1 ml 5M NaNO2 was added under intensive stirring. When thereaction was completed (after 2 hours), the suspension of diazonium compoundwas used for the next coupling reaction.

Second Coupling Reaction: Schaeffer acid (C10H8O4S; M = 224.2 g mol–1;0.02 mole; 4.48 g) was mixed with 45 ml water and 4 ml 5M NaOH. Then thetemperature of the mixture was set at 0-5 °C (external cooling) and the diazoniumcompound from the previous step was added under intensive stirring. After that,pH of the reaction was adjusted at 8 by addition of 5M NaOH, and the mixture wasleft to react for approx. 30 min. When the reaction was completed, pH 5 wasadjusted by addition of concentrated HCl (35 %) and NaCl (40 g) was added(precipitation of dye was checked by drop test on filter paper). Finally, after20 min of stirring the dye was filtered off and carefully dried (40 °C).

In order to check the purity of prepared dye, TLC was used (solid phase:ALUGRAM SIL G/UV254, mobile phase: propan-2-ol/propan-1-ol/ethyl-acetate/water = 2/4/1/3 by vol.).


The reactive dye 3 (C28H19N4O13S4Na3; M = 816.7 g mol–1; 0.01 mole; 8.16 g) wasmixed with 50 ml water and with an excess of N-methylglycine (C3H7NO2; M =89.1 g mol–1; 0.04 mole; 3.56 g) which was added under intensive stirring. Then,pH of the reaction was set at the value 8 (5M NaOH) and the reaction mixture wasstirred at the temperature of 60 °C for 3 hours. When the reaction was completed(checked by TLC, solid phase: ALUGRAM SIL G/UV254, mobile phase: propan-2-ol/propan-1-ol/ethylacetate/water = 2/4/1/3 by vol.), the final dye was salted outby an addition of 20 g NaCl and carefully dried (40 °C).


NN

NN

OH

VS

SO3Na

COONa


NN

NN

OH

VSA

SO3Na

COONa

VSA = SO2CH2CH2NCH3

CH2COONa


Diazotization of MAFSES: MAFSES (C8H11NO6S2; M = 281.3 g mol–1; 0.02 mole;5.63 g) was diazotized in the same way as that described for dye 3.

First Coupling Reaction: Cleve acid (C10H9NO3S; M = 223.2 g mol–1; 0.02 mole;4.46 g) was coupled with diazonium compound by the procedure described for thepreparation of dye 3.

Diazotization of MAD 4: The solution of mono azo dye MAD 4 was diazotized inthe same way as that described for the preparation of reactive dye 3.

Second Coupling Reaction: BON Acid (C11H8O3; M = 188.18 g mol–1; 0.02 mole;3.76 g) was reconstituted with 40 ml water and 4 ml 5M NaOH. The solution wascooled externally to 0-5 °C under intensive stirring, and then solution ofdiazonium compound was added. The reaction mixture was kept at pH 8 (additionof 5M NaOH), and the reaction time was approx. 1 hour. When the couplingreaction was completed, the dye was precipitated by the addition of 40 g NaCl.The dye 4 was filtered off and carefully dried (40 °C).



NN

NN

OH

SO3Na

NHO

VS

VS


COOH

OH

NH2

VS

+PCl3

OH

NH

O

VS

Fe (IN 1)

The reactive dye 4 (C29H19N4O12Na3S3; M = 780.6 g mol–1; 0.01 mole; 7.81 g) wasmixed with 50 ml water, whereupon an excess of N-methylglycine (C3H7NO2;M= 89.1 g mol–1; 0.04 mole; 3.56 g) was added under an intensive stirring. Afterthat, pH of the mixture was set at 8 by addition of 5M NaOH, and the reactionmixture was stirred at the temperature of 60 °C for 3 hours. When the reaction wascompleted (checked by TLC, solid phase: ALUGRAM SIL G/UV254, mobile phase:propan-2-ol/propan-1-ol/ethylacetate/water = 2/4/1/3 by vol.), the final dye wassalted out by an addition of 20 g NaCl and carefully dried at 40 °C.


Preparation of IN1: MAFSES (C8H11NO6S2; M = 281.3 g mol–1; 0.02 mole; 5.63 g)and BON Acid (C10H8O3; M = 188.18 g mol–1; 0.02 mole; 3.76 g) were mixed with20 ml N-methylpyrrolidone, 2 ml pyridine and powdered iron (0.1 g). Thenphosphorus trichloride (PCl3; M = 137.33 g mol–1; 0.02 mole; 1.74 g) was addedunder intensive stirring. The reaction mixture was refluxed (200 °C) for 3 hours.When the reaction was completed (checked by TLC, solid phase: ALUGRAM SILG/UV254, mobile phase: propan-2-ol/ammonia = 2/1 by vol.), the final reactionsolution was mixed with 100 ml cold water (mixture with crushed ice) andintensively stirred. The precipitated intermediate IN1 was filtered off and mixedwith 100 ml acetone. Then activated carbon (0.5 g) was added and the mixture wasfiltered (removing of carbon and impurities). The filtrate was diluted with waterand the precipitated intermediate was filtered off and carefully dried (40 °C).

Diazotization of MAFSES: MAFSES (C8H10NO6S2H; M = 281.3 g mol–1;0.02 mole; 5.63 g) was diazotized similarly as described for reactive dye 3.


VS NN

SO3Na

NN

NH2

VS


OH NH

N

N

N

Cl

Cl

NaO3S SO3Na

NN

SO3Na

First Coupling Reaction: Cleve acid (C10H9NO3S; M = 223.2; 0.02 mole; 4.46 g)was coupled with the diazonium compound in the same way as shown for reactivedye 3.

Diazotization of MAD 5: The solution of mono azo dye MAD 5 was diazotized inthe same way as that described for reactive dye 3.

Second Coupling Reaction: Intermediate IN1 (C19H17O8S2; M = 451.46 g mol–1;0.02 mole; 9.02 g) was dissolved in 40 ml water and 4 ml 5M NaOH. Then themixture was cooled externally to 0-5 °C under intensive stirring, and the solutionof diazonium compound was added. After that, pH of the reaction was set at 8 bymeans of 5M NaOH, and the reaction took approximately 1 hour. When thecoupling reaction was completed, the dye was precipitated by addition of 40 gNaCl. The dye 5 was filtered off and carefully dried at 40 °C.

Structures of other investigated reactive dyes employed in this study aresummarised below.

Reactive dye 6

Reactive dye 7


OH NH

N

N

N

Cl

N

NaO3S SO3Na

NN

SO3Na

CH2COONa

CH3

NaO3S

SO3Na

NN

NH N

N N

Cl

NH

SO2CH2CH2OSO3Na

H2NOC-HN

NaO3S

SO3Na

NN

NH N

N N

Cl

NH

SO2CH2CH2N

H2NOC-HN

CH2COONaCH3

O

O

NH2

SO3Na

NH

CH3

CH3

SO3Na

NH

CH3N

N

N

Cl

Cl

Reactive dye 7A

Reactive dye 8 (C.I. Reactive Yellow 145)

Reactive dye 8A

Reactive dye 9


O

O

NH2

SO3Na

NH

CH3

CH3

SO3Na

NH

CH3N

N

N

Cl

N

CH2COONa

CH3

OH NH

N

N

N

SO3NaNaO3S

NNNaO3SO-CH2-CH2-SO2

Cl

Cl

OH NH

N

N

N

SO3NaNaO3S

NNN-CH2-CH2-SO2

Cl

NCH2COONaCH3

NaOOC-H2C

CH3

NNOH

CH3NN

CH2=CH-SO2

SO2CH=CH2

Reactive dye 9A

Reactive dye 10

Reactive dye 10A

Reactive dye 11a


Reactive dye 11b

Reactive dye 11A

Reactive dye 12 (C.I. Reactive Black 5)

Reactive dye 12A


Reactive dye 13

Reactive dye 13A

Reactive dye 14

Reactive dye 15

Reactive dye 16


% dye

2% acetic acid

25°C

98°C

60 min85°C

15 min

1g l-1

NaHCO3

Reactive dye 17

TLC Identification of Sarcosine in Dyebath

Identification of sarcosine in dyebath was performed by thin layerchromatography. For this purpose, ALUGRAM SIL G/UV254 and the mixture ofpropan-1-ol with NH4OH (25 %) in 2:1 ratio (vol. parts) were used as the solidphase and mobile phase, respectively. Detection reaction was performed byspraying the chromatogram with ninhydrin (0.05 % solution in acetone) sinceninhydrin forms with sarcosine violet colour. A solution containing 3 % ofsarcosine in aqueous ethanol (1:1 by vol.) was used as reference sample.

NMR spectroscopy: The 1H (360.13 MHz) and 13C (90.556 MHz) NMR spectra ofcompounds were recorded at 300 K on Bruker AMX 360 spectrometer equippedwith a 5 mm broadband probe and Silicon Graphics Indy computer using theUXNMR software. The NMR spectra were measured in hexadeuteriodimethylsulphoxide. The 1H and 13C chemical shifts were referred to the signal of solvent(δ = 2.55 ppm). Due to very large data quantity the NMR spectra are not shownhere.

Dyeing procedure: Dyeing was carried out in pots placed in laboratory dyeingmachine using a liquor ratio 1:20 for nylon fabric and 1:40 for wool. Dyeing ofnylon fabric was performed at 2 % owf (wool 4% owf) in the presence of aceticacid (2 %) with the dyeing profile shown in Fig. 1.

Fig. 1 Dyeing profile


Analysis of dyeing: Absorbance measurements of the original dyebath and theexhausted bath were carried out using UV/VIS spectrophotometer. Using apreviously established absorbance/concentration relationship at the λmax of the dye,the quantity of the dye in solution was calculated and the extent of dye exhaustion(E %) achieved was determined using Eq. (1), where C0 and C are the quantitiesof dye initially in the dyebath and of residual dye in the dyebath, respectively. Calkal

is a concentration of the dye after alkalization step. The calculation of thepercentage covalent fixation was carried out as follows

(1)

(2)

Table II Absorption wavelengths (λmax) of all investigated dyes together with calibration rangesand regression parameters

Reactive dye 8max

nmShade Regression parametersa Calibration

rangemg l–1

1, 1A 501.0 Red A = 0.0376 c; R2 = 0.9997 0-20

2, 2A 540.0 Red A = 0.0227 c; R2 = 0.9993 0-25

3, 3A 535.5 Red A = 0.0180 c; R2 = 0.9999 0-25

4, 4A 548.5 Red A = 0.0346 c; R2 = 0.9997 0-20

5 664.0 Brown A = 0.0031 c; R2 = 0.9999 0-200

6 659.0 Brown A = 0.0018 c; R2 = 0.9997 0-500

7 533.5Red

A = 0.0265 c; R2 = 0.99990-25

7A 538.5 A = 0.0188 c; R2 = 0.9993

8 414.0Yellow

A = 0.0118 c; R2 = 0.99930-50

8A 422.0 A = 0.0117 c; R2 = 0.9991

9 580.5Blue

A = 0.0134 c; R2 = 0.99990-50

9A 587.0 A = 0.0066 c; R2 = 0.9998aestimate of intercept was found to be insignificant


Table II — Continued

Reactive dye 8max

nmShade Regression parametersa Calibration

rangemg l–1

10 516.5Red

A = 0.0166 c; R2 = 0.99960-50

10A 499.5 A = 0.0108 c; R2 = 0.9994

11a 380.5 A = 0.0286 c; R2 = 0.9997

11b 383.0 Yellow A = 0.0308 c; R2 = 0.9999 0-25

11A 382.0 A = 0.0290 c; R2 = 0.9994

12 597.5Blue

A = 0.0277 c; R2 = 0.99910-30

12A 598.5 A = 0.0293 c; R2 = 0.9999

13 349.5Yellow

A = 0.0250 c; R2 = 0.99920-30

13A 378.5 A = 0.0119 c; R2 = 0.9993

14 520.0 Red A = 0.0306 c; R2 = 0.9998 0-25

15 432.0 Yellow A = 0.0635 c; R2 = 0.9993 0-10

16 412.0 Yellow A = 0.0224 c; R2 = 0.9992 0-30

17 522.0 Red A = 0.0144 c; R2 = 0.9995 0-50aestimate of intercept was found to be insignificant


All the investigated dyes prepared in this study were applied to dying ofpolyamide 6 and wool. The wash fastness at 40 °C and 60 °C was tested accordingto the ISO 105-A10S/C01 (1994) and ISO 105-A10S/C03 (1994) procedurerespectively. The wash fastness - potting - was tested according to the ISO 105-A10S/E09 (1994) procedure. The values obtained for dyeing of polyamide andwool are shown in Table III and Table IV, respectively. It can be seen from theresults presented, that modification of reactive group by sarcosine improvedsignificantly the colouristic properties of some of the synthetized dyes for dyeingof polyamide 6 and wool. The modification increased significantly coulouristicresults on wool in the case of reactive dyes 1 and 1A, 2 and 2A, 7 and 7A, 10 and10A, 12 and 12A. The modification of reactive dyes 1, 2 a 12 applied topolyamide 6 was successfull as well.

What can be further seen from the presented results is that regarding thedyeing of polyamide 6 the best colouristic properties were achieved in the pre-


Table III Exhaustion, fixation and fastness properties of reactive dyes on PAD 6

Reactive dye Exhaustion, % Fixation, % Washing 40 °C Washing 60 °C1 97.4 89.6 4-5a / 1-2b, 3-4c 4 / 1, 21A 90.7 99.3 4-5 / 4-5, 4-5 4-5 / 1-2, 4-52 37.3 97 4-5 / 5, 4 4-5 / 5, 42A 99.9 99.8 4-5 / 5, 4 4-5 / 5, 43 85.7 96.4 4-5 / 5, 4 4-5 / 5, 43A 35.5 81.3 4-5 / 5, 4 4-5 / 5, 44 98.2 76.7 4-5 /5, 4 4-5 / 5, 44A 73.3 80.8 3-4 /5, 3 3-4 / 4-5, 2-35 61.4 53.4 4 / 4-5, 4-5 4 / 4-5, 4-56 75.2 74.1 4-5 / 4-5, 4-5 4-5 / 4-5, 57 66.8 41.2 4-5 / 5, 5 4-5 / 5, 57A 69.4 39.8 4-5 / 5, 5 4-5/ 5, 48 100 91.2 5 / 5, 5 5 / 5, 58A 100 81.3 5 / 5, 5 5 / 5, 49 91.1 82.2 5 / 5, 5 5 / 5, 59A 93.3 78.2 4-5 / 5, 5 5 / 5, 510 92.1 83.8 4-5 / 5, 5 4-5 / 5, 510A 46.4 45.3 5 / 5, 5 4-5 / 5, 511a 97.6 82.1 5 / 5, 5 5 / 4, 411b 96.6 96.4 5 / 5, 4 5 / 5, 411A 65.2 59.5 5 / 5, 5 5 / 5, 512 79.4 63.6 5 / 5, 5 5 / 5, 512A 90.3 80.5 5 / 5, 5 5 / 5, 513 89.4 63.4 5 / 5, 5 5 / 5, 513A 98.0 48.4 5 / 5, 5 5 / 5, 3-514 80.7 67.0 4-5 / 4-5, 4-5 4-5 / 4-5, 4-515 87.8 98.9 4 / 4-5, 4-5 3 / 4-5, 4-516 95.2 99.0 4 / 4-5, 4-5 3 / 4-5, 4-517 89.5 88.4 4-5 / 4-5, 4-5 4 / 4-5, 4-5

a colour change, b staining on polyamide, c staining on cotton

sence of dyes 8, 8A, 9, 9A, 10, 11a, 11A, 12, 12A, 13 and 13A. In the presence ofthese dyes typical values of exhaustion better than 90 % were obtained with theexception of those obtained for dyes 11A and 12 for which the exhaustion of only60 and 80 %, respectively, was achieved. On the other hand, for both dyes, i.e.11A and 12, excellent wetfastness values were obtained similarly as for the otherabove mentioned dyes (8, 8A, 9, 9A, 10, 11a, 12A, 13 and 13A). Fixation valuesof 80−90 % were achieved for the most of the above listed dyes; howeverregarding the employing of the dye 2A and 3 much better values (close to 100 %)were obtained. On the other hand, worse wetfastness characteristics for both thedye 2A and 3 were achieved in contrast to those obtained in the presence of otherslisted dyes. Regarding the dyeing of wool the best results were obtained in thepresence of dyes 2A, 8, 8A, 9, 9A, 10, 10A, 11a, 11b, 12A and 13A. The


application of these reactive dyes gave excellent values of exhaustion close to100 % and of fixation better or close to 90 % (reactive dye 13A). It is clearlyevident, that reactive dyes 2A, 8, 8A, 9, 9A, 10, 10A, 11a, 11b, 12A a 13A ensuredexcellent colouristic characteristics for dyeing both investigated materials. Thesedyes thus may be universally recommended for dyeing of both polyamide andwool.

Table IV Exhaustion, fixation and fastness properties of reactive dyes on wool

Reactive dye Exhaustion, % Fixation, % Washing 40 °C Washing 60 °C1 80 91 4a / 1b, 1-2c 4-5 / 1, 2-31A 98.4 99.5 4 / 1, 3 4 / 1, 32 79.0 98 4-5 / 5, 3-4 4 / 1-2, 42A 100 98.8 4-5 /, 4-5, 4 4-5 / 1-2, 4-53 56.4 96.9 5 / 4-5, 4 5 / 1, 4-53A 92.1 89.6 4-5 / 4-5, 2-3 4-5 / 1, 44 63.7 90.7 4-5 / 2-3, 2 4-5 / 1, 44A 90.2 67.0 4-5 / 3, 2 4-5 / 2, 4-55 56.4 91.8 4 / 4, 4 2-3 / 2-3, 4-56 44.6 83.5 3 / 4, 4-5 3 / 2, 4-57 83.5 72.6 4 / 5, 4 4 / 1, 47A 100 80.9 4 / 5, 3-4 4 / 1, 48 100 92.7 5 / 5, 4 5 / 3-4, 4-58A 100 99.3 5 / 5, 3-5 4-5 / 3, 4-59 100 94.8 4-5 / 5, 5 5 / 1-5, 49A 100 94.0 4-5 / 5, 5 5 / 1-2, 410 99.4 90.0 4 / 5, 3-4 5 / 2, 410A 100 96.5 4-5 / 5, 3-4 4-5 / 3,411a 99.2 96.8 4-5 / 5, 5 4-5 / 4-5, 511b 99.8 99.0 5 / 5, 5 5 / 4, 511A 81.1 79.8 5 / 5, 5 5 / 4, 512 81.0 56.1 4-5 / 4-5, 4-5 4-5 / 3, 4-512A 99.9 98.8 4-5 / 4-5, 4-5 4-5 / 3-4, 4-513 99.5 97.2 4-5 / 5, 4 3-4 / 2-3, 3-413A 99.6 87.9 3-4 / 4, 3-4 4-5 / 3-4, 414 35.6 85.7 2-3 / 2, 3 3-4 / 1, 315 89.9 92.7 2 / 1-2, 2 3-4 / 1, 3-516 68.4 98.3 4 / 4, 4-5 4 / 1-2, 4-517 60.3 98.7 3 / 4, 4 3-4 / 1-2, 4-5

a colour change, b staining on wool, c staining on cotton


Conclusion

It can be concluded from the presented results that reactive dyes for polyamide 6offer real potential for producing dyeings of higher fixation and wetfastness thanthose produced from acid and disperse dyes without limitation regarding shade.The prepared reactive azo dyes based on N-methyl-N-carboxymethyl-2-aminoethyl-sulfonyl group are suitable for the reactive dyeing of both polyamide 6and wool. The dyeing procedure is the sequence of acid dyeing at pH 3, alkalitreatment at pH 8 and washing. Due to high fixation and very good wetfastnessvalues achieved with some of these newly designed reactive dyes their successfulapplication in routine practice may be expected.

Acknowledgements

The study was performed with financial support of the Ministry of Education,Youth and Sports of the Czech Republic (project No. 0021627501).

References

[1] U.S. Patent, 4,336,190.[2] U.S. Patent, 4,492,654. [3] U.S. Patent, 4,064,754.[4] U.S. Patent, 4,577,015.[5] U.S. Patent, 4,049,656.[6] U.S. Patent, 3,359,286.[7] U.S. Patent, 3,268,548.[8] U.S. Patent, 3,385,843.[9] Heyna J, Schumacher W.: DBP 965, 902, Hoechst, 1949.[10] Zollinger H.: Textilveredlung 6, 57 (1971).[11] DAS 1,103,883, Hoechst, 1958.[12] BP 880,886, ICI, 1988.[13] EP Patent, 0, 277,624 A2, Hoechst, 1988.[14] U.S. Patent, 3,802,837.


133



16 (2010)

PREPARATION OF SURFACE MODIFIED KAOLINESBY SILICONE HYDROPHOBIZATION

Eva MATĚJKOVÁ1, Andréa KALENDOVÁ and Štěpán OBRDLÍKInstitute of Chemistry and Technology of Macromolecular Materials,



This work deals with preparation of silanized kaolines for application in organiccoatings. Coatings have a crucial task in protection of metal materials againstrust. Leafy fillers improve anticorrosive properties (they participate in barriermechanism), stability in atmospheric influences and adhesion to base plate.Organic silicate substances are able to create hydrophobic layer on a surface ofdifferent materials. Hydrophobization reduces a possibility of osmotic blisterscreation. Four samples were prepared from four types of silanes(polydimethylsiloxane, methyltrimetoxysilane, aminoalkyl siloxane andaminopropylsilane) and two different kaolines. The following samples wereprepared: A- Al6Si2O13 (mullite, S1),B- SiO2 (quartz, S2),C- Al6Si2O13 (mullite,S3),D- SiO2 (quartz, S4). Properties which are important for application incoatings were astimated by means of the following tests: structure identificationby X-ray diffraction analysis, determination of the specific mass, oil consumptiondetermination, specific surface area determination, determination of particle sizeand distribution, determination of the morphology of particles. Determination of

134 Matějková E. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 133–143

methanol (MN) and water (WN) number is very important for evaluatinganticorrosion properties of silanized samples in coatings. The prepared pigmentshave diverse chemical composition and content of structural phase quartz andmullite. The characterization of pigments shows good properties of testedpigments for production of coatings. Mechanical and anticorrosive properties willbe investigated.

Introduction

The protection of metals against corrosion is a complicated and very expensiveproblem. Steel is exposed to various corrosive environments from low up to ultra-high aggressivity. The aim of paint and coating industry is to reduce the price ofanticorrosive protection, which is possible by development of more efficientbinders, pigments, fillers and additives.

The development is influenced by ecological requirements focused onreduction of toxicity in the field of organic coatings and their applications. Astrong toxicity leads to prohibition of pigments as red-lead, lead and chromatesubstances. These were replaced by phosphates, borates, silicates, molybdates andothers which are nontoxic. Attention of this research is devoted to leafy fillerswhich may notably improve anticorrosive properties, stability in atmosphericinfluences and adhesion to base plate [1].

Pigments are basis material for production the anticorrosive coatings besidebinders. They play important part in coatings and lead to paint films propertieswhich improve their quality and field of application [2].

Fillers are compounds whose primary aim is to achieve desired pigmentvolume concentration of solids (fillers and pigments) in coating system. Secondaryaims of fillers are optimization definite properties and specific properties of thissystem. Kaolines belong to fillers with non isometric (flaky, lamellar) shape ofparticles which participate in barrier mechanism of metals protection. Kaoline(china clay) is sedimentary mineral whose basic component is kaoliniteAl2(Si2O5)(OH)4, i.e. hydrated aluminosilicate [3,4]. Kaolinite is white but rawkaoline can be coloured due to foreign materials. It embodies monoclinic crystalgeometry like mica. Kaolinite is made by aerating alkaline feldspar

(1)

Kaolinite loses crystal water by calcination near 600 °C and changes intokaoline [5]

(600 °C) (2)

Matějková E. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 133–143 135

At temperatures around 1,400 °C final products are cristobalite (SiO2) andmullite (Al6Si2O13), and/or amorphous phases.

Kaoline is further used as a delusterant. Other uses of calcinated kaolineconsist in application to decorative paints. Irregular shape of particles increasesstrength of coatings. It is also used in paint to extend titanium dioxide (TiO2) andmodify gloss levels. The largest use is in paper production, as a filler to a paperand for coating surface modifications of paper. Kaoline increases brightness andwhiteness of paper and decreases its price. Another application of kaoline is inrubber and plastic industry, as a filler for cosmetics, in pharmaceutics and inconstruction materials as a fire retardant. Kaolin is used in medicine, as a foodadditive, in toothpaste, as a light diffusing material in white incandescent lightbulbs. It is generally the main component in porcelain. Porcelain is a ceramicmaterial made by heating raw materials, generally including kaolin, in a kiln totemperatures between 1,200 °C and 1,400 °C. Kaolinite has also seen some use inorganic farming, as a spray applied to crops to deter insect damage. Kaoline is asupplementary cementitious material. Metakaolin affects the acceleration ofPortland cement hydration when it was added to concrete [6-9].

Organic silicate substances are able to create hydrophobic layer on a surfaceof different materials. This layer is in principle polymeric silicon microfilm whichmagnifies contact angle surface with water and decreases friction coefficient.Water cannot make up continuous film on such surface, it does not wet the surfaceand it easily streams down. The orientation of all organic groups of silicon chaintoward material surface explains the hydrophobic effect of silicon layer. Thethickness of silicon film is imperceptible (ca, 10–6 cm). The hydrophobic layer isof very high durability, and it is spoiled by drastic recourses which simultaneouslydestroy material surface. A silicone film does not block interstices of material anda hydrophobized material is able to “breathe”. Water vapour goes through itsinterstices, however, the water is not retained [10].

Fig. 1 Scheme of material surface which is hydrophobized by methyl silicon polymer[9]


Experimental

The prepared pigments were characterized by the following procedures. The tests,which characterize properties important for application in coatings, are describedbelow. All the measurements were carried out under ambient conditions, i.e. 23±2°C and 50±5 % relative humidity according to CSN EN 23270.

Structure Identification by X-ray Diffraction Analysis

The structure identification was performed by X-ray diffraction analysis usingdiffractometer X'pert PRO MPD 1880 (PANanalytical, The Netherlands). Whenan X-ray beam bombards a crystalline lattice in a given orientation, the beam isscattered in a definite manner characterized by the atomic structure of the lattice.The Bragg law is useful for measuring wavelengths and for determining the latticespacing of crystals. Modern X-ray crystallography provides the most powerful andaccurate method for determining single-crystal structures.

Determination of Specific mass

The determination was carried out using Micromeritics AutoPyknometr 1320(Micromeritics Instrument Corp., USA) that measures the volume of a samplebased on measuring the volume of gas (helium) displaced by a measured powderpigment.

Determination of Oil Consumption

The oil consumption means the amount of flax oil in grams that makes a paste withdefined properties from 100 g pigment. This determination was carried outaccording to CSN 67 0351 using the pestle-bowl method.

Determination of Specific Surface Area

The specific surface area of kaoline solid particles (m2 g–1) was determined bymeans of a Micromeritics ASAP 2000 Physic/chemisorption unit (MicromeriticsInstrument Corp., USA). Specific surface area is a material property of solidswhich measures the total surface area per unit of mass. It is a derived scientificvalue that can be used to determine the type and properties of a material. It isdefined as the by surface area divided by mass (with the units of m2 g–1).


Determination of Particle Size and Distribution

The size and distribution of particles was determined using Mastersizer 2000(MALVERN Instruments Ltd, UK). Particles and size distribution means thepercentages of each fraction into which a granular or powder sample is classified,with respect to particle size, by number or weight.

Determination of Morphology of Particles

The surface and shape of pigment particles were studied by means of a JEOL-JSM5600 LV scanning electron microscope (Jeol, Japan) with accelerating voltage 15kV. The SEM is designed for direct studying of the surfaces of solid objects byscanning with an electron beam that was generated and focused by the operationof the microscope. The SEM allows a greater depth of focus than the opticalmicroscope. For this reason the SEM can produce an image that is a goodrepresentation of the three-dimensional sample.

Determination of Methanol Number (MN)

Methanol number is the concentration of aqueous methanol solution (in percentby wt.) In which one half of kaolin sample immerses into the liquid. The MN valuereflects the strength by which silane is attached to the surface of kaolin. The higherthe MN value, the stronger bond is. Procedure: 40 ml aqueous methanol of theconcentration series 5, 10, 15 ...100 % MeOH is placed in a beaker, and 10 mlhydrophobized kaoline is added; The mixture is stirred at the speed of 800 rpm for5 min. Then the mixture is transferred into a 100-ml graduated cylinder and leftto stand 10 min to sediment. The proportion of sedimented and floating materialis estimated on the cylinder scale. The MN value is obtained when the amounts offloating and sedimented material are equal.

Determination of Water Number (WN)

Water number is the amount of pigment that is wetted by water and immerses. Theproportion of kaolin that is not coated with silane can be estimated from WN.Thus, the proportion of non-silanized kalin can be deduced. The lower the WN,the better the silanization is. Procedure: 40 ml water and 10 ml pigment suspensionare placed in a beaker. The mixture is stirred at the speed of 800 rpm for 5 min andthen transferred into a 100-ml graduated cylinder and left to sediment for a periodof 24 hours. The proportion of sedimented and floating material is estimated on


a cylinder scale. The WN is the amount of the sedimented part in ml.

Results

The preparation of these types of pigments can be divided according to the way oftheir origin and a type of used silicone and kaoline. The temperature of preparationwas 1,150 °C. Two types of kaolines were chosen for preparing the samples.Kaoline K1 is composed mainly of mullite, kaoline K2 is quartz. A summary ofprepared hydrophobized kaolines is given in Table I.

Table I Summary of prepared kaolines modified by silane hydrophobization

Sample Kaoline Silane

A Al6Si2O13(S1) K2 S1*

B SiO2(S2) K2 S2*

C Al6Si2O13(S3) K1 S3*

D SiO2(S4) K2 S4*

* The codes of silanes ares explained below

Hydrophobization by Polydimethylsiloxane

Silane S1 is a polysiloxane emulsion containing 60 % by wt. poly(dimethyl-siloxane) [(CH3)2SiO]n and 1-3 % by wt. Decylvinylether, the rest being water.Procedure: 83 g of this polysiloxane emulsion and 1670 g kaoline K2 are mixedin 4,170 ml water, and this mixture is heated at the temperature of 60 °C withvigorous stirring for a period of 4 hours, whereupon the suspension is filtered ona Büchner funnel. The filter cake is dried in hot-flue drying kiln at the temperatureof 105 °C for a period of 24 hours. The dried material is homogenized in agrinding mortar and reduced to a fine powder. Sample A – Al6Si2O13 (mullite, S1)was prepared in this way.

Hydrophobization with Methyltrimethoxysilane

The starting materials: isobutyl alcohol, ethanol, kaoline K2 and silane S2-methyl-trimetoxysilane CH3Si(OCH3)3 were placed a into sulphonation flask ofcapacity 2.5 l with a stirrer, two thermometers and a reflux condenser. Reactionproceeded 4 hours in a heating mantle at the temperature of 60 °C. The suspensionwas filtered on a Büchner funnel. The filter cake was dried at the temperature of


100 °C for a period of 6 hours. The mixture was comminuted in a mortar andmixed to fine powder. Sample B - SiO2(quartz, S2) was prepared in this way.

Hydophobization with Aminoalkyl Siloxane

Silane S3 is aminoalkyl siloxane [H2N(CH2)RSiO0.5]n. Only the technology ofpreparing differs: sample C- Al6Si2O13 (mulite, S3) and sample D- SiO2 (quartz,S4).

Silane S3, distilled water, kaoline K1, HCl for acidification were placed intoa sulphonation flask with a stirrer and a thermometer. The mixture was kept at thetemperature of 80 °C for 2 hours. The suspension was filtered, and the thixotropicfilter cake was dried at the temperature of 105 °C for 24 hours. The mixture wasgrounded to fine powder.

Hydrophobization by Aminopropyl Silane

Distilled water, aqueous solution Silane S4- [H2NCH2CH2CH2SiO0.5]n and kaolineK2 were placed into 10-l pot and everything was fully mixed. The suspension wasdried in a spray drier which was heated by burned charges of natural gas. Thesuspension was disintegrated to small droplets by means of a of rotary reel(atomizer) at a speed 30 000 rpm. The droplets came into contact with dryingmedium — burned charges of natural gas and hot air. Water was quicklyevaporated from the surface of the droplets, and the suspension was transformedto dry powder. The dried material was separated from drying medium in a cycloneseparator and the burned charges left to a chimney. Sample D- SiO2(quartz, S4)was prepared in this way.

Characterization of Prepared Kaolines

A – Al6Si2O13(mullite, S1)

Structural phases: mullite (Al6Si2O13), quartz (SiO2)Specific mass: 2.68 g cm3

Oil consumption: 34.5 g/100 g pigmentSpecific surface area: BET-isotherm 0.84 m2 g–1

D10 2.89 :mD50 7.07 :mD90 31.21 :mMN 50WN 0


Fig. 2 SEM micrographs of sample A

B – SiO2(quartz, S2)

Structural phases: quartz (SiO2), mullite (Al6Si2O13)Specific mass: 2.72 g cm3


D10 2.31 :mD50 3.19 :mD90 6.46 :mMN 40WN 0

Fig. 3 SEM micrographs of sample B

C – Al6Si2O13(mullite, S3)

Structural phases: mullite (Al6Si2O13), quartz (SiO2)Specific mass: 2.74 g cm3



D10 0.95 :mD50 3.46 :mD90 30.34 :mMN 0WN Everything was dipped

Fig. 4 SEM micrographs of sample C

Fig. 5 SEM micrographs of sample D

D – SiO2(quartz, S4)

Structural phases: quartz (SiO2), mullite (Al6Si2O13)Specific mass: 2.71 g cm3


D10 2.60 :mD50 8.82 :mD90 18.37 :m


MN 0WN Everything was dipped

Discussion and Conclusion

Samples A – Al6Si2O13(mullite, S1) and C– Al6Si2O13(mullite, S3) have the samechemical composition, i.e. they contain mainly mullite. Samples B – SiO2(quartz,S2) and D – SiO2(quartz, S4) are predominantly composed of quartz. Chemicalcomposition was determined by RTG analysis. The highest value of oilconsumption was measured with sample D – SiO2(quartz, S4) – 52.7 g/100 g ofpigment. The largest value of specific surface area was found in the case of sampleC – 2.63 m2 g–1, which is perceptible from SEM micrograph in Fig. 4. SEMmicrographs (Figs 2-4) show lamellar shape of particles. All the samples hadsimilar sizes and distribution of particles. Sample A – Al6Si2O13(mullite, S1) hadhigher abundance of large particles, so there are clumps of particles in the SEMmicrograph Fig. 2.

The highest value of methanol number was measured in the case of samplesA – Al6Si2O13(mullite, S1) and B – SiO2(quartz, S2) which were silanized withsilane S1 and S2. These samples had simultaneous low value of water number.High degree of silanization predestines samples A and B to good anticorrosiveproperties because hydrophobization declines the possibility of osmotic blisterscreation. Sample A is mainly composed of mullite in contrast to sample B whichis formed by quartz. Anticorrosion tests are being carried out to show, if the degreeof silanization or the chemical composition is crucial for anticorrosiveprotection.The characterization of pigments reveals good properties of the testedpigments to create coatings. The prepared pigments are going to be applied toalkyd resin. Mechanical and anticorrosive properties will be investigated.

Acknowledgments

This work was supported by the Ministry of Education of the Czech Republic(MSMT 0021627501) and the project IGA (SFGChT03).

References

[1] Rožan J., Vaníček O.: Pigments-Powder Paints, pp. 28-32, SNTL Praha,1959.

[2] Buxbaum G.: Industrial inorganic chemie, pp. 190-202, WILEY-WCHVerlag GmbH, 1998.


[3] Rissa K., Lepistö T., Yrjölä K.: Prog. Org. Coat. 55, 137 (2006).[4] Al-Turaif H., Bousfield D.W., LePoutre P.: Prog. Org. Coat. 44, 307 (2002).[5] Murray H.H.: Appl. Clay Sci. 17, 207 (2000).[6] Kislenko V.N., Verlinskaya R.M.: J. Colloid Interf. Sci. 244, 405 (2001).[7] Murray H.H., Kogel J.E.: Appl. Clay Sci. 29, 199 (2005).[8] Veselý D., Kalendová A.: Prog. Org. Coat. 68, 173 (2010).[9] Ece O.I., Nakagawa Z.: Ceramics International 28, 131 (2002).[10] Schätz M., Starch J., Kolář O., Dyk A., Hix P.: Technical Application of

Silicones, pp. 81-108, SNTL Praha, 1959.


145



16 (2010)

ANTICORROSION PROPERTIES OF ION-EXCHANGABLE PIGMENTS IN EPOXIDE

COATINGS

Štěpán OBRDLÍK1, Andréa KALENDOVÁ and Eva MATĚJKOVÁInstitute of Chemistry and Technology of Macromolecular Materials,



Hexavalent chromium compounds (chromates) have been widely used as inhibitivepigments in the formulation of anticorrosive paints. However, their high toxicityand carcinogenic effects are forcing the development of effective chromate-freeorganic coatings. One such alternative, which is very attractive from a scientificpoint of view, is the use of ion-exchangeable pigments (IEPs). The few studiesconducted with this type of pigment are not conclusive about their anticorrosiveefficiency and controversy surrounds their functioning mechanisms, interchangecapacity and anticorrosive performance. In the present research, which focuseson the anticorrosive protection of this type of pigment, calcium/silica (Ca/Si) IEPswere synthesized in first step, next epoxide paint coatings were formulated. IEPshave been applied on low carbon steel panels. The effect of these non-toxicpigments on the protective properties of coatings has been tested by means ofnatural and accelerated corrosion tests (humidity, salt spray and SO2).

146 Obrdlík Š. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 145–151

Experimental results have shown that IEPs are suitable to formulate anticorrosivepaint films with improved anticorrosive effects.

Introduction

Paints initially insulate a metal substrate from the water and oxygen required forenvironmental corrosion. However, all organic film-forming polymers are to agreater or lesser extent permeable. Although paint coatings have been used forcenturies to protect metals against atmospheric corrosion, their protectionmechanisms have not yet been fully explained and for many years have been thesubject of controversy. Anticorrosive protection by paints has recently beendescribed as being a combination of a physical barrier, a chemical inhibitor andan electrical rezistor [1]. IEPs are relatively high surface area inorganic oxidesloaded with ionic corrosion inhibitors by ion exchange with the surface hydroxylgroups. The oxides are chosen for their acidic or basic properties to provide cationor anion exchangers; thus, silica is used as a cation support and alumina foranions. So, calcium-exchanged silica [2], Al–Zn–decavanadate hydrotalcite [3],and cerium(III) and calcium(II) cation exchanged bentonites [4] have beenincorporated as pigments in anticorrosive paints.

Cations permeating the paint film will release calcium from the pigment andat the same time be immobilised. The implications of such a release mechanismare that the inhibitor species is only released when the paint film is beingpermeated by corrosive species and the amount released is dependent on theseverity of the environment. Calcium, released from the silica surface, migrates tothe paint–metal interface where a thin, inorganic layer is deposited which isessentially impermeable to moisture/ions and hence prevents the corrosionreaction from proceeding. According to Goldie [5] the presence of this layer hasbeen detected by X-ray photoelectron spectroscopy on painted metal platesexposed to a variety of environments. Other evidence supporting this layer theoryhas been obtained from electrochemical studies [5]. So, the anticorrosive effect ofCa/Si pigment can be described considering two mechanisms: the exchange ofcalcium ions for cations (H+, etc.) penetrating the coating from the environment,neutralising the acid compounds and forming on the substrate a protective layerconsisting of calcium and iron silicite [5,6].

Obrdlík Š. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 145–151 147

Experimental

Preparation and Characterization of Pigment

IEPs were prepared by high-temperature synthesis based on reaction of silicateswith calcium oxide. Calcium carbonate and silicates with different layeredstructure were chosen as starting raw materials. The first step was thermaldecomposition of CaCO3 to CaO and CO2 according to Eq. (1). The reactionproceeded in solid phase. The resulting pigment had the silica matrix with CaO onthe surface. The amount of CaO in the final pigment was 20 % by wt.

(1)

The following silicates were chosen: Bentone SD2, Mica SFG 20, Talek A-60. The starting raw materials were weighed with an accuracy of ±0.001 g. Themixture of starting materials was homogenized in porcelain mortar for a period of30 min. The second step was calcination in electric oven at the temperature of1150 °C with 2-hour time lag at the peak temperature and the temperatureincreasing from the start at the rate of 5 °C min–1.

The prepared calcinate was ground by wet way in planetary ball mill for aperiod of 5 hours at the rate of 400 rpm. Corundum balls were used as the millballs.

The ground pigment was washed with distilled water to remove the non-reacted particles of starting materials that would negatively affect the pigmentcharacteristics, whereupon the pigment was heated at the temperature of 110 °Cto constant weight.

The powdery pigment was characterized by the following parameters:specific mass, oil consumption, pH value, and specific conductivity of aqueousextract of the pigment. The values of critical pigment volume concentrations(CPVC) were calculated.

Preparation of Coatings

Model paints based on epoxy emulsion (producer Spolchemie a.s., dry mass = 55%) and epoxy resin (producer Spolchemie a.s., dry mass = 75 %) containing thesynthesized pigments were prepeared in order to assess properties of the pigment.For estimation of physical-mechanical and anticorrosion properties the cjosenvolume concentration of pigment in PVC-based coating was 10 %.

The prepared paints were coated on a steel panel (approx. dimensions152×102×0.8 mm) by means of a double-layer applicator blade with 200 :m slot.The second layer of paint was applied (after sufficient curing of first layer) by


means of an applicator blade with the slot producing a film of approx. 100 :mfinal thickness. For the corrosion tests the panels were cut (approx. 8 cm cutsexcept the steel basis in the distance of 1 cm from the edges).

Corrosion Tests

The tested coatings were exposed to corrosive environment chambers with generalcondensation of water, corrosive environment combination chambers with salt fogand corrosive environment condensation chamber containing SO2. The generalanticorrosion efficiency (A) was derived from the test evaluating the corrosion offlat basis (A1) according to ASTM D 610- 85), the test evaluating the corrosion incuts (A2) according to ASTM 1654- 92), and the test evaluating production ofblisters in the coating (A3) according to ASTM D 714- 87). The samples wereexposed to corrosive environment for a period of ca. 500 hours (paints based onepoxy emulsion) or ca. 2000 hours (paints based on epoxy resin).

The value of general efficiency (A) is in the interval of 0-100. A highernumeric value means higher value of anticorrosive efficiency — Eq. (2). Overallassessment was derived after Heubach scale. Equation for general anticorrosiveefficiency A for corrosion test in chamber with NaCl content, condensationchamber with general evaporation of water and in chamber with SO2 content readsas follows

(2)


Properties of Pigments

Parameter Mica Bentone Talek

Specific mass, g cm–3 2.78 2.51 2.86

Oil consumption, g/100 g pigment 64.6 75.9 44.4

CPVC, % 34.1 32.76 42.1

pH of aqueous leach 11.3 9.76 12.31

Note. In the table, there are properties of finished IEPs pigments. The names ofthese pigments are identical with the names of silicates.


Anticorrosion Properties

In the chambers with salt fog, all coatings on the basis of epoxy resin had verygood anticorrosion efficiency. The best of all was the coating pigmented withMica SFG + 20 % CaO. The paint with Talek A60 + 20 % CaO had the minimumanticorrosion efficiency in epoxy emulsion coatings. The other two coatings hada similar, higher than average anticorrosion protection. The results are presentedin Diagram 1.

Diagram 1

In chambers with general condensation of water, epoxy resin coatingsshowed very high anticorrosion protection. The coating pigmented with Talek A60+ 20 % CaO had almost maximal protection. The coating on the basis of epoxyemulsion had a different anticorrosion efficiency. The best of all was the coatingwith Talek A60 + 20 % CaO. This paint showed 100 % efficiency. The results arepresented in Diagram 2.

In the chamber containing SO2 all the coatings (in both types of binders),except coatings pigmented with Bentone SD2 + 20 % CaO, showed higher thanaverage anticorrosion protection. The paint containing Mica SFG + 20 % CaO hadthe best anticorrosion efficiency in epoxy emulsion coatings . Talek A60 + 20 %CaO had the best protection against corrosion in coatings on the basis of epoxyresin. The results are presented in Diagram 3.


Diagram 2

Diagram 3

Conclusion

Several IEPs of calcium/silica (Ca/Si) type were synthesized. The pigments inpowdery state were characterized by the following parameters: specific mass, oil


consumption, pH value and specific conductivity of aqueous extract of thepigment. Coatings based on aqueous epoxy emulsion and on solvent epoxy resinwere formulated and prepared. The IEPs were applied on low carbon steel panels.The effect of these non-toxic pigments on the protective properties of coatings wastested by means of natural and accelerated corrosion tests (humidity, salt spray andSO2). Generally, all the pigments showed very good anticorrosion protection.Especially coatings on the base of epoxy resin showed extreme efficiency. Thebest of all pigments was Mica SFG + 20 % CaO.

Acknowledgments

This work was supported by the Ministry of Education of the Czech Republic(MSMT 0021627501) and the project IGA (SFGChT03).

References

[1] Chico B., Simancas J., Vega J.M., Granizo N., Díaz I., de la Fuente D.,Morcillo M.: Progress in Org. Coat. 61, 238 (2008).

[2] Zubielewicz M., Gnot W.: Prog. Org. Coat. 49, 358 (2004).[3] Buchheit R.G., Guan H., Mahajanam S., Wong F.: Prog. Org. Coat. 47, 174

(2003).[4] Bohm S., McMurray H.N., Powell S.M., Worsley D.A.: Mater. Corros. 52, 896

(2001).[5] Goldie B.P.F.: JOCCA 9, 257 (1988).[6] Fletcher T.E.: Proceedings of the 11th International Corrosion Congress,

Florence, Italy, April 1990, AIM vol. 2, 265 (1999).


153



16 (2010)

CONTAMINATION OF ATMOSPHEREWITH VOLATILE HYDROCARBONS

DURING HANDLING OIL PRODUCTS OF PETROLAND DIESEL FUEL TYPES

Jaromíra CHÝLKOVÁa1, Renáta ŠELEŠOVSKÁa, Martin STUCHLÍKa and Jaroslava MACHALÍKOVÁb

aInstitute of Environmental and Chemical Engineering,bDepartment of Transport Means and Diagnostics,


Received March 19, 2010

The paper describes monitoring of contamination of atmosphere with organicsubstances during ordinary operation of a petrol station, namely during pumpingof fuel by a consumer. The measurements were performed by means of a portableanalyser ECOPROBE 5 equipped with an IR detector. The analyser wascalibrated using standard mixtures of petrol vapours and air as well as by meansof gaseous standards of volatile portions of Diesel fuel. The calibration curvesobtained were used for determination of real concentrations of the analysedsubstances in the units of mg m–3. The practical measurements have shown that thereverse exhaust of petrol vapours cannot prevent contamination of the atmosphereunless the dispensing nozzle is properly inserted during filling the tank.

154 Chýlková J. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 153–165

Introduction

Volatile organic compounds (VOC) are air pollutants whose importance has beenincreasing recently. These substances play an important part in physico-chemicalprocesses in troposphere because they significantly contribute to formation ofozone and other photochemical oxidants. They can also represent a serious healthhazard to population and to the environment because of the toxicity of some ofthem (e.g., benzene and buta-1,3-diene) [1]. In particular, VOC include the crudeoil products, i.e. substances and preparations containing aliphatic, alicyclic andaromatic hydrocarbons possessing low viscosity and low surface-tension values.Besides the crude oil itself they include the products of crude oil treatment, suchas petrol, kerosene oil, Diesel fuel, and mineral oils. Nowadays, in connection withdevelopment of industry, agriculture and transport, it is impossible to avoidsituations that introduce VOC into the environment. In particular, such situationsare connected with:

! escape during handling crude oil products in the cases of their transport,pumping, storage, filling into tanks of machinery, repairs of machinery,exchanges of filling, application of crude oil products as solvents, cleaningaids, grease removers etc.,

! escapes of emergency nature, such as damage of tanks in machinery, rupture ofhosepipes and pressure pipelines in machinery, rupture of crude oil pipelines,during traffic accidents of vehicles transporting these products, accidents ofstorage and handling equipment and their pipe manifolds and fittings, duringcriminal acts and sabotages concerning these types of equipment etc.,

! escapes connected with improper disposal of used packaging materials andcontainers of crude oil products.

Suitable methods for determination of pollution of the atmosphere withVOC include the gas chromatography combined with, e.g., flame-ionizationdetector, thermal conductivity detector, or photo-ionization detector and, inparticular, mass detector [2]. Automatic portable analysers for direct quick and/orcontinuous field analyses of air quality are subject of significant developmentbecause of the necessity of obtaining — as rapidly as possible — the most accuratedata possible about the actual situation of contamination of the atmosphere withVOC, the necessity of taking proper measures in situations of emergency escapeof VOC into the atmosphere or, on the other hand, long-term monitoring of the airquality, e.g., near potential sources of pollution. The most widespread variant ofportable analysers are gas chromatographs developed as a scaled-down simpleralternative to laboratory models [3-6]. Although various types of detectors areapplied, the most frequent combination is GC + mass detector [7-11]. New typesof mass spectrometers were also developed for the purposes of on-line analysis.One of them is the method of Membrane introduction mass spectrometry (MIMS),

Chýlková J. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 153–165 155

which adopts an entry membrane (made of, e.g., poly(dimethylsiloxane)) forconcentrating the sample: the membrane is much more pervious for the organicanalysed substance than for the matrix, which allows reaching of a very lowdetection limit [12]. Another possibility is the so-called Selected ion flow tubemass spectrometry (SIFT-MS), which makes use of reactions of analysedsubstance with precursors (e.g., , , ), the products of these reactionsbeing subsequently analyzed by means of the differentially pumped quadruple MS-ion-counting system [13]. However, the field analysers based on the principle ofgas chromatography and/or mass spectrometry may suffer from the drawback ofthe necessity of relatively demanding instrumentation: in particular, they may needvarious carrier gases and a source of vacuum, which makes the manipulation moredifficult. An important point also is the way of collecting and concentrating ofsamples before the analysis proper [14].

A significant role in the determination of contamination of the atmospherewith VOC also belongs to the portable analysers based of spectrophotometricmethods. Apart from the mass spectrometers [15-20], which have already beenmentioned above in the context of gas chromatography, literature describes, e.g.,an analyser working on the principle of UV spectrophotometry, namely withapplication of various adsorbents, e.g., powdered amorphous SiO2 [21,22] or finelyporous silicate powder [23,24]. Further possibility is offered by the IR analysers,which allow reaching of very low detection limits (less than mg m–3) and are alsosuitable for long-term continuous monitoring of contamination of the atmospherewith VOC [25,26]. This group of analysers also includes the mobile analyserECOPROBE 5 [27], which was used for analysis of the atmosphere in the presentwork. A large advantage of these instruments can be seen in the fact that thesample is sucked directly into the cell volume, where the measurement takes place,and no additional special equipment for collecting and concentrating of analysedsubstance is needed.

The aim of this work was to calibrate the portable analyser ECOPROBE 5by means of gas chromatography and then monitor contamination of theatmosphere with VOC from fuels for motor vehicles.

Experimental

Refinery Products Used. Preparation of Gaseous Mixtures

The portable VOC analyser was calibrated using model standard gaseous mixturesof air with the following types of petrol: Natural 95, Natural 98, Special 91(Paramo, Pardubice, the Czech Republic), Special 100 (OMV Refining, Wien,Austria), and Shell V-Power Racing (Shell Chemicals, Hamburg, Germany), andalso a mixture of air with Diesel fuel (Slovnaft, Bratislava, Slovakia). The


mixtures were prepared by addition of a defined amount of the liquid fuel into agas-tight bag (Linde Gas, Praque, the Czech Republic) whose inner surface wascovered with aluminium foil. The addition was carried out by means of a 500-:lhypodermic syringe (Hamilton), whereupon a defined volume of air was pumpedinto the bag by means of membrane pump M 401. The mixture obtained washeated for a short time period and then left to homogenize at room temperature((25 ± 2) °C). Formation of homogeneous mixture was checked by means of gaschromatography: if homogeneous, the gaseous sample gave a constant response,expressed by the height of the representative peak dominating in the sample;subsequently, the concentration of the component was determined by the samemethod. The standard serving for quantification of unknown petrol samples wasthe gaseous sample prepared by adding 5 :l liquid substance into a gas-tight glasscontainer of defined volume, which was closed by means of silicone septum.Perfect vaporization of the liquid phase was ensured by short heating with hot air.The concentration of this mixture was determined by calculation. In the case ofpreparation of standards of volatile hydrocarbons from Diesel fuel, which is avaried mixture of substances with boiling points in the range of 180-370 °C, it wasnot possible to calculate the concentration on the basis of the whole amount of theliquid sample added, because less volatile components remain in liquid phase atroom temperature or slightly above. The content of volatile components in Dieselfuel (hereinafter TSN) was determined experimentally in the following way: anopen glass vessel was charged with 200 :l Diesel fuel having the density of0.8299 g/cm3, whereupon the vessel was kept at the temperature of 80-85 °C, andthe decrease in the mass of liquid phase was determined by weighing at regulartime intervals. Figure 1 shows the time dependence of the mass decreasedetermined in three repeated experiments.

Fig. 1 Time dependence of mass decrease of Diesel fuel at the temperature of 80-85 °C


From Fig. 1 it is obvious, that in the 15-min period from the beginning ofheating the decrease in volatile components is considerable, about 50 % TSNbeing released. The residual mass is stabilized after ca 100 min. Then the wholedecrease in volatile components reaches 80 % of the initial amount of Diesel fuel.These findings were used in preparation of standards and calculation ofconcentrations of gaseous components of Diesel fuel, which then served forquantitative evaluation of both the model and unknown gaseous mixtures bymeans of gas chromatography.

Instruments and Apparatus Used

The field analyses of gaseous mixtures of air with volatile crude oil products wereperformed with the portable analyser ECOPROBE 5 (RS Dynamics, Prague)equipped with IR detector. The scheme of this instrument is presented in Fig. 2.The detector contains a source of IR radiation, whose rays are directed into thethrough-flow cell, where the IR radiation is absorbed in accordance with theabsorption bands of the given components. The remaining radiation passes throughfour optical filters and reaches four sensors. Three filters transmit only specificwavelengths of the radiation characteristic of carbon dioxide, methane andhydrocarbons. The fourth sensor and its optical filter have reference function. Themeasuring range of the indication sensors lies in the range of 0-500,000 ppm withthe detection limit of 50 ppm. The sensor providing information about the totalcontent of hydrocarbons has been calibrated for methane by the manufacturer.

Fig. 2 Scheme of IR detector in portable analyser ECOPROBE 5

The concentrations of model gaseous mixtures of air with crude oilhydrocarbons were also determined by means of gas chromatography using aChrom 5 apparatus equipped with a flame-ionization detector (Laboratornípřístroje, Prague). Argon was used as carrier gas. Separations were performed onpacked column (DC 200). For analysis of petrols the temperature of injector was160 °C, temperature of oven was 150 °C and temperature of detection was 120 °C.


Analysis of Diesel Fuel was performed at the temperature of injector 180 °C,temperature of oven 180 °C and temperature of detection 150 °C.


Calibration of Analyser ECOPROBE 5 for Gaseous Mixtures of Motor Fuels

Since the IR detector channel that measures the total content of hydrocarbons hadbeen calibrated with methane by the manufacturer, it was necessary beforepractical measurements to calibrate this sensor with a real mixture of hydrocarbonswhich are to be monitored in the field measurements. Therefore, the calibrationcurves were determined for both petrol and volatile substances present in Dieselfuel in order to represent the dependence of values provided by ECOPROBE 5 inppm regime upon the real concentration of analyzed substance in mg m–3. In thecases of the petrol types Natural 95, Natural 98, Special 91, Special 100 and ShellV-Power Racing, the gaseous mixtures of various concentrations of hydrocarbonswere prepared by addition of 200 :l liquid substances into a bag which containeda small amount of pure air. After subsequent short heating, air was added into thebag, and the mixture was left to homogenize for a period of 5 min, the temperaturebeing adjusted at (25 ± 2) °C. The mixture prepared was analysed quantitativelyby means of gas chromatography, whereupon the bag content was analyzed withthe ECOPROBE 5 apparatus. Further gaseous mixtures were prepared by dilutingthe gas in the bag with pure air, and then they were analyzed by both above-mentioned methods. Plotting of the dependence between the response of IRdetector in ppm regime and the real concentration of petrol in the gaseous mixtureprovided the calibration curves given in Fig. 3.

Figure 3 shows that there are only small differences between the individualregression curves; these differences may be due to experimental error of the twoanalytical methods adopted. Therefore, it was decided to create a single calibrationcurve taking into account all the experimental points and apply it to evaluation ofconcentrations of petrol vapours of all the petrol types examined. The equation ofthis summary calibration curve reads as follows

where y stands for the IR detector response presenting the total content ofhydrocarbons in ppm units on the basis of the manufacturer’s calibration usingmethane, x is the real concentration of petrol in mg m–3 and R is the regressioncoefficient.

In order to verify correctness of the above-mentioned equation, a series ofmodel gaseous samples of individual sorts of petrol was prepared; also prepared


Fig. 3 Calibration dependence between values obtained from ECOPROBE 5 in ppmunits and real concentration of petrol vapours in mg m–3 units

was a mixture of Natural 95 and Special 91 in the ratio of 94.2 : 5.8 chosen on thebasis of statistical data about consumption of petrol in the first half of 2007 [28].First, their concentration was determined by means of gas chromatography in therange of 3 051-13 578 mg m–3 and then by means of the portable analyser givingthe readings in ppm units (calibrated with methane); the later data weretransformed into real concentrations of petrol in mg m–3 units using the above-given calibration equation. The results are presented in Table I, where it can beseen that the concentrations determined by means of the analyser ECOPROBE 5differ from those obtained by means of gas chromatography by less than 10 rel.percent, which can be considered as acceptable in gas analysis.

The model gaseous mixtures of air with volatile components of Diesel fuelto be analyzed with ECOPROBE 5 were prepared by adding various amounts(100-300 µl) of Diesel fuel into a bag containing a certain volume of air,whereupon the volume of bag was made up with air. Since the formation ofequilibrium concentration of volatile components is very lengthy at roomtemperature, the actually formed mixtures were analyzed by both methods: always0.5 ml was withdrawn from the sample sucked by the analyzer, and this aliquotwas submitted to gas chromatography. The graphical record produced by theanalyzer (see Fig. 4) shows that the gaseous mixture has a constant compositionthroughout the period of measurements (tens of seconds). Plotting of the dataobtained from ECOPROBE 5 against the real TSN concentrations [mg m–3]obtained from gas chromatography gave the calibration curve whose equationreads as follows


Table I Comparison of results of petrol vapour concentrations obtained from gaschromatography with those measured by means of analyser ECOPROBE 5 andtransformed using above given regression equation

Petrol sort Sample No. Concentrationdetermined bymeans of GC

mg m–3

Concentrationdetermined by

means of analyzermg m–3

Error%

Natural 95

1 9280 8791 –5.28

2 6992 6683 –4.41

3 3051 2989 –2.03

Natural 98

1 7756 8362 7.81

2 5654 6103 7.93

3 4132 4296 3.96

Special 91

1 8830 8146 –7.74

2 7811 7172 –8.18

3 5604 5635 0.56

Special 1001 9840 10699 8.73

2 7741 7908 2.15

3 5183 5478 5.70

Shell V-Powerracing

1 13578 13197 –2.81

2 8147 7802 –4.24

3 5356 4945 –7.68

Note: Results are calculated from three repeated analyses

where y stands for the IR detector response giving the total content ofhydrocarbons in ppm on the basis of the manufacturer’s calibration using methane,x is the real concentration of TSN in mg m–3.

This relationship enables transformation of the analyzer ECOPROBE 5readings into correct estimation of contamination of the atmosphere with volatilecomponents of Diesel fuel, which can be seen in the results presented in Table II.The table shows that the results obtained from ECOPROBE 5 and the above-givencalibration equation differ from those obtained from gas chromatography by lessthan 12 rel. percent, which can be considered satisfactory in gas analysis.


Fig. 4 Graphical record showing constant composition of mixture of air and Diesel fuelin bag during analysis with ECOPROBE 5. ABC ! the name of Locality; Inc, X,Y ! GPS parameters; O2 – oxygen concentration;T – temperature; P ! absoluteambient pressure; PID ! result of photoionization detector, concentration (unitsppm); Meth ! result of IR Methane channel, concentration (units ppm); T.P. !IR Total Petroleum channel, concentration (units ppm); CO2 ! IR CO2 channel,concentration (units ppm)

Table II Comparison of concentrations of volatile components of Diesel fuel determined bymeans of gas chromatography with those obtained from ECOPROBE 5 and abovegiven calibration equation

Sample No. Concentrationdetermined bymeans of GC

mg m–3

Concentrationdetermined by means

of analyzermg m–3

Error%

1 7124 6712 –5.79

2 3979 3502 –11.99

3 2842 3021 6.31

4 3032 2915 –3.84

5 2577 2299 –10.79

Note: Results are calculated from three repeated analyses


Practical Applications

After the analyser ECOPROBE 5 had been calibrated, it was used for monitoringof the volatile components escaping from motor fuels, particularly during theirpumping into tanks of cars. The samples were collected at the distance of ca 10 cmfrom the tank opening. The data obtained were automatically stored into thememory, whereupon the analyser ECOPROBE 5 was reset (zeroized) and thusready for next measurements.

Table III Results of analyses in practical application of analyser ECOPROBE 5 during operationof petrol station

MeasurementNo.

Site of samplecollecting

Atmosphericpressure

kPa

Temperature°C

ccalibr.eq.

mg m–3

1 Background ofpetrol station

97.57424 25.77 Below LOD

2 b 97.57957 25.21 c

3 a 97.56224 25.55 620

4 a 97.55024 26.00 4457

5 a 97.53691 26.01 8811

6 a 97.55158 26.26 17198

7 a 97.55158 26.34 Below LOD

8 Tank opening 97.67556 20.61 c

9 5 cm from tankopening

97.68890 20.96 14212

10 a 97.67823 20.82 10744

11 15 cm fromtank opening

97.67556 20.69 7557

12 a 97.53558 26.01 c

13 a 97.54891 26.11 11254

Notes: Samples from 2 to 11 – petrol Natural 95 and samples from 12 to 13 – diesel fuel; a – 10cm from tank opening; b – 10 cm from canister opening; c – values exceed range of calibrationcurves; LOD – limit of detection of device used

The concentrations of petrol and Diesel fuel were evaluated from the datameasured in ppm regime and transformed by means of the above-given calibrationequations. The results obtained are presented in Table III. Since the dispensing


nozzle of petrol station is equipped with efficient exhaust of petrol vapours, itcould be presumed that the escaping of volatile substances during pumping themotor fuel would be negligible. Our measurements have shown that thisexpectation is only fulfilled if the dispensing nozzle is properly inserted into thecar tank opening, which was the case in the measurements No. 3 and No. 7. Theconcentrations measured here were minimal. In the other cases (Nos. 4-6 and 12),the drivers inserted the dispensing nozzle improperly and the contamination valuesmeasured were high. In the measurements Nos. 2, 8, and 12, the concentrationvalues are not given because the IR detector response exceeded the range ofcalibration curve. The table also shows that during filling of motor fuel into a 5-liter canister (measurement No. 2), when the analyzed gas was sucked at thedistance of 10 cm from the canister opening, the amount of escaping vapours ofNatural 95 was so high that it exceeded the measuring range of analyserECOPROBE 5. The reason was probably in the fact that the pumping of motor fuelwas not continuous, and the dispensing nozzle did not fit tightly in the canisteropening. In this case the exhausting of vapours was almost inefficient.

The measurements Nos. 8-11 simulated the situation when the car tankopening remains open for a certain period of time, and the volatile hydrocarbonsescape into the atmosphere. The analyser analysed samples taken directly at the cartank opening and at the distances of 5, 10, and 15 cm from it. The IR detectorresponses and therefrom subsequently evaluated concentrations decreased withincreasing distance, which confirmed the fact that the results measured stronglydepend upon the choice of site for collecting the samples.

The measuring instrument ECOPROBE 5 was successfully applied topractical analyses of the atmosphere at petrol stations. On the basis of precedingcalibration of the analyser, it was possible to determine real concentrations ofrefinery products in gaseous samples.

Conclusion

Mobile analytical techniques enable not only checking of quality of theenvironment, but they also can fulfil preventive tasks and/or help operationallydeal with ecological accidents. Their application requires serious attention to begiven to the selectivity of measurements and correctness of the results obtained.One of possible ways of calibration is described above: it uses a real mixture andcan be performed in advance and used if needed.


Acknowledgements

This communication has been elaborated with financial support provided byMinistry of Environmental Protection SP/4i2/60/70 and by Ministry of Education,Youth and Sports MSM 0021627502.

References

[1] Ras-Mallorquí M.R., Marcé-Recasens R.M., Borrull-Ballarín R.: Talanta 72,941 (2007).

[2] Santos F.J., Galceran M.T.: Trac-Trends Anal. Chem. 21, 672 (2002).[3] Dewulf J., Van Langenhove H.: J. Chromatogr. A 843, 163 (1999).[4] Helmig D.: J. Chromatogr. A 843, 129 (1999).[5] Grall A.J., Sacks R.D.: Anal. Chem. 71, 5199 (1999).[6] Zhou L., Zeng Y., Hazlett P.D., Matherne V.: Anal. Chim. Acta 596, 156

(2007).[7] Eckenrode B.A.: J. Am. Soc. Mass Spectrom. 12, 683 (2001).[8] Syage J.A., Nies B.J., Švand M.D., Hanold K.A.: J. Am. Soc. Mass

Spectrom. 12, 648 (2001).[9] Chiriac R.E., Lornage R., Fine L., Care J., Gass J.L., LagierT.: Int. J.

Environ. Anal. Chem. 87, 43 (2007).[10] Na K., Kim Y.P.: Atmos. Environ. 35, 2603 (2001).[11] Later D.W., Lee M.L.: J. Am. Soc. Mass Spectrom. 19, 1425 (2008).[12] Ketola R.A., Kotiaho T., Cisper M.E., Allen T.M.: J. Mass Spectrom. 37, 457

(2002).[13] Smith D., Španěl P.: Mass Spectrom. Rev. 24, 661 (2005).[14] Ras M.R., Borrull F., Marcé R.M.: Trends in Analytical Chemistry 28, 347

(2009).[15] Syage J.A., Hanning-Lee M.A., Hanold K.A.: Field Anal. Chem. Tech. 4,

204 (2000).[16] Diaz J.A., Giese C.F., Gentry W.R.: Field Anal. Chem. Tech. 5, 155 (2001).[17] Patterson G.E., Guymon A.J., Riter L.S., Everly M., Griep-Raming J.,

Laughlin B.C., Ouyang Z., Cooks R.G.: Anal. Chem. 74, 6145 (2002).[18] Blain M.G., Riter L.S., Cruz D., Austin D.E., Wu G., Plass W.R., Cooks

R.G.: Int. J. Mass Spectrom. 236, 91 (2004).[19] Riter L.S., Peng Y., Noll R.J., Patterson G.E., Aggerholm T., Cooks R.G.:

Anal. Chem. 74, 6154 (2002).[20] Ouyang Z., Wu G., Song Y., Li H., Plass W.R., Cooks R.G.: Anal. Chem. 76,

4595 (2004).[21] Ueno Y., Horiuchi T., MorimotoT., Niwa O.: Anal. Chem. 73, 4688 (2001).[22] Ueno Y., Horiuchi T., Niwa O.: Anal. Chem. 74, 1712 (2002).


[23] Ueno Y., Horiuchi T., Tomita M., Niwa O., Zhou H., Yamada T., Honma I.:Anal. Chem. 74, 5257 (2002).

[24] Ueno Y., Horiuchi T., Niwa O., Zhou H., Yamada T., Honma I.: Sens. Mater.15, 393 (2003).

[25] Ketola R.A., Kiuru J.T., Tarkiainen V., Kokkonen J.T., Rasanen K., KotiahoT.: Anal. Chim. Acta 562, 245 (2006).

[26] Goldthrop M.D., Lambert P.: Journal of Hazardous Materials 83, 135 (2001).[27] RS Dynamics, Earth Science & Security Equipment. Available from

http://www.rsdynamics.com/.[28] <http://download.mpo.cz/get/32636/ 36211/403913/ priloha001.pdf>,

downloads in April, 2008.


167



16 (2010)

ASPECTS CONCERNING ATTRACTIVENESSOF COMPANY AS EMPLOYER

Marie BEDNAŘÍKOVÁa1, Martina LINHARTOVÁa

and Jaroslava HYRŠLOVÁb

aDepartment of Economics and Management of Chemical and Food Industry,The University of Pardubice, CZ–532 10 Pardubice,

bCollege of Economy and Management, CZ–158 00 Prague


The article discusses selected aspects of attractiveness of a company acting in thecapacity of an employer. The company attractiveness is determined by corporateculture, which defines the specific nature of each company and its activities. Thebasic element of corporate culture is corporate strategy; some theories viewcorporate culture as an effective instrument of strategic management. Theattractiveness of a company as an employer is further affected by a number offactors from the area of the human resources management strategy (e.g. theprocess of recruiting and selecting employees, employee stimulation methods,company educational system, opportunity for teamwork, level of communicationin a company, health and safety at work) and the utilization of voluntaryinstruments (activities), which are used by a company to declare itsenvironmentally friendly approach and respect of sustainable developmentprinciples in the course of its business activities. This article also includes basic

168 Bednaříková M. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 167–178

results of research performed in companies, which are members of the CzechAssociation of Cleaning Stations (CACS).

Introduction

High-quality human capital is very valuable and important for each company andits success. It can be utilized by a company to create the very importantcompetitive advantage on the market. Companies should pay a great deal ofattention to their existing as well as potential employees. To acquire new andretain existing key employees, a company must be attractive for them. It generallyapplies that not only companies choose their employees but employees also choosecompanies, for which they are going to work. Naturally, applicants areconsiderably disadvantaged during the periods of higher unemployment andeconomic crisis; however, if a company comes under the impression of animmediate advantage, it will suffer from this in the future.

The objective of this article is to identify and discuss selected aspectsconcerning attractiveness of a company as an employer. Potential employees selecttheir employer based on the perception of such employer on the outside, the wayit presents itself, the references of existing and former employees, etc. Theattractiveness of an employer is assessed according to certain criteria, whichinclude, for example, stimulation and remuneration method of employees,possibility of career growth, approach of a company to safety at work, promotionof teamwork, and many others. Undoubtedly, these also include an overall levelof corporate culture. It is obvious that there are certain elements within thestructure of corporate culture which are easily noticeable (the most noticeable areartifacts, less noticeable are values and positions of a company) on the outside ofthe company (i.e. by potential applicants as well), and such elements may play acrucial role when assessing attractiveness of a company as an employer. If anemployee takes the first job in spite of the fact that the company is not tooattractive for such employee (e.g. due to the fact that no other job opportunityarose), we can hardly assume that he/she would stay in the company longer thanabsolutely necessary. A company with “unhealthy” level of corporate culture willsooner or later lose its employees. The article also includes basic results ofresearch relating to attractiveness of a company as an employer performed incompanies, which are members of the Czech Association of Cleaning Stations(CACS).

Bednaříková M. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 167–178 169

Selected Aspects Concerning Attractiveness of Company as Employer

The significant element affecting the attractiveness of a company as an employeris, without any doubt, corporate culture. The company management should realizethat the company actually competes with other employers on the labor market forthe sympathy of employees (both existing and potential) and that the level ofcorporate culture may represent a significant advantage, which could ensure thatthe company wins over a high-quality employee. With regard to existingemployees, the company should manage and develop corporate culture in a mannerwhich would result in a satisfied and loyal employee at every single workingposition.

Lukášová [1] states that the impact of corporate culture within a companyis not isolated as it interacts with other subsystems, which mainly includecorporate strategy and organizational structure. The present turbulentenvironment often requires companies to conduct in a flexible manner and to haveflexible strategies in order to flexibly react to changes in the environment and toforesee or even actively create such changes. The relation of corporate culture,strategy, and environment may be described as follows (see Fig. 1).

Fig. 1 Relation of strategy, culture, and environment [1]

It is clear that corporate culture affects not only the formation of the strategyitself, but also the content and implementation thereof. Dytrt [2] states that thehigher correspondence of corporate culture and selected corporate strategy, thebetter results a company may achieve. However, according to Lukášová [1], anabsolute identity is not suitable, because certain degree of conflict creates roomfor the formation of new strategies. On the other hand, it is obvious that if acompany has strong corporate culture, which is not consistent with the selectedstrategy, it would be very difficult for such company to achieve its strategic goals.It is necessary to understand that there is a mutually determining relationshipbetween strategy and corporate culture. The content of culture affects theformation, content, and execution of strategy; the content of strategy leads to theformation of certain type of culture.

Specialized literature more and more presents an opinion that the human

Strategy Culture

Environment


resources management strategy is a key element for companies’ success. Livianand Pražská [3] mention three models of the human resources management:

! The first model, which had dominated for a long time, emphasizes themanagement instruments. In this situation, the human resources managementrelies on rational decision-making, which is ensured by general directors andby the company’s economic strategy. This model was mainly applied to themanagement of socialist enterprises. However, it is necessary to point out thatit still, in a way, persists in the management of many newly formed companiesin the post-communist countries.

! The second model focuses on negotiation and decision-making of managersand seeks compromises between social and economic goals. This model wasbeing developed in 1970s in North-American and European companies. Withregard to the Czech Republic, it is possible to mention Tomáš Baťa, whoapplied this model in his company in Zlín in 1930s.

! Finally, the third model respects the fact there may be various contradictionsand tensions in respect of economic, social, and human goals.

According to Stýblo [4], the planning of human resources ensures the rightpeople in the right place within an organization; therefore, the content of thehuman resources strategy should comprise the following areas:

! Selection process; ! Education of employees; ! Motivation and stimulation of employees, with an emphasis on their

development; ! Teamwork; ! Communication;! Health and safety of employees at work.

Urban [5] provides the following rules for healthy corporate culture: acompany should operate as a team, should promote open communication, andshould create strong motivation and organizational identity. Therefore, thecompany should focus on the management of such areas in order to have satisfiedand loyal employees, who are an essential premise of healthy corporate culture;this naturally results in higher value for customers. It is clear that in the case acompany wishes to have healthy corporate culture, which is one of the pillars ofa company’s success, it must sufficiently focus on the management of individualareas of the human resources management strategy. According to Bednaříková [6],it is likely that an unsatisfied employee will finally “lead” to an unsatisfiedcustomer. In the case a customer meets an emotionally negative employee in anylink of the supply chain, we can hardly assume the customer will receive theexpected value. This can frustrate all efforts of a company.


Armstrong [7] states that the general goal of acquiring and selectingemployees should be to recruit such number and such quality of employees, whichare desired for a satisfactory fulfillment of the company’s need for humanresources, all this while expending minimum costs.

According to Stýblo [4], it is first necessary to identify the necessarycapabilities for a correct selection of employees; companies should selectemployees according to their attitude and then according to their skills, becauseindividual’s conformity with the values and culture of an organization is crucial— i.e. the so-called “psychological contract” between an individual and acompany. In the case a company correctly organizes and prepares the employeeselection process, it can find out what an applicant expects from his/her future joband what he/she intends to offer to his/her employer. Furthermore, this will assistthe company in revealing the harmony/disharmony between the views of anemployee and the corporate culture. Thorough selection process may bedemanding in respect of time and financial resources. However, if the risk ofwrong selection is high, such resources are necessary.

Stachová [8] states that the data necessary for the selection of employees aremainly collected with the use of psychological methods (e.g. graphology analysisof an applicant’s written application, analysis of his/her report card, analysis of theformer employer’s evaluation etc.) and psycho-diagnostic methods (e.g. sensory,personal and/or kinetic tests).

Crucial part of the employee selection process is an interview with anapplicant. The company management should bear in mind that this primary contactplays an important role not only for the employer, but especially for a potentialemployee. For this reason, the interview should take place within a pleasantenvironment, positive atmosphere, and on a professional level. According toBednaříková [6], the right employee selection process represents the first step inpreventing undesired fluctuation.

The key objective of education of employees is for a company to increaseits human capital in value as it is a company’s most valuable asset. It is necessaryto realize that company education is one of the employee benefits, becauseeducation results in the fulfillment of the needs of both the company and itsemployees, as the satisfaction and competitiveness of employees increase both onthe intra-company and external labor market. According to Bednaříková [9], thedevelopment of employees is not just one of many forms of employee care. Today,it also takes on the role of information/knowledge creator and provider of dynamiclevel of education. In order to create, manage, and allocate knowledge where it canbe applied in the best possible manner, it is necessary to establish such corporateculture, which is based on information sharing and permanent learning.

The principal condition for success and effective performance of a companyis, without any doubt, qualified and educated top management (as a source of thecompany’s competitive advantage), which promotes educational system of all


employees based on their capabilities. For this reason, some companies investconsiderable resources in the area. According to Lošťáková [6], the success ofcompanies depends on their ability to convert acquired knowledge, skills andexperience of staff into competitive products and services. The situation in someCzech companies in the area of employee education is not satisfactory.

If companies ask the highest possible performance from their employees, itis necessary to pay a great deal of attention to the most suitable methods ofemployee stimulation using such tools, which will result in the desired motivation.Only highly motivated employees will identify with company goals and will bewilling to adapt their personal goals to them. According to Armstrong [7], themotivation theory explains why people behave in certain manner at work, whythey make specific efforts. Herzka [10] states that most negative causes, which arereflected in the work performance, are related to motivation. The causes ofbehavior of employees represent a difficult issue, which requires more and moreattention of both economists and psychologists.

Stachová [8] provides the following diagram of the motivation process (seeFig. 2).

Fig. 2 Motivation process model [8]

The stimulation process is much more difficult than managers of individualcompanies sometimes realize. It would be absurd to believe that a single approachto stimulation will suit all employees. According to Khelerová [11], for example,salary tends to be less important for higher and top management. They prefer thepossibility for self-fulfillment, achievement of success, career, and job content. Itis not that salary is unimportant; however, relevant remuneration is implied forhigher positions. Specialized literature offers several theories dealing withmotivation, such as Herzberg’s theory, Maslow’s theory, theory of justices, andmany others.

According to Bednaříková [6], effective stimuli for other categories arecommunication and overall awareness of employees, evaluation of employees’complex performance, education, working relations or provision of correspondingworking conditions (e.g. health and safety at work, acceptable working hours,

1. Motive

2. Goal-oriented behavior

Brings

3. Fulfillment of a goal

Invokes Affects


etc.). Hyršlová and Bednaříková [12] state that approach, form of communication,applied management style, and the time and effort of management and personnelspecialists dedicated to other employees of a company will reflect in theirmotivation and overall performance.

To establish healthy (or rather ideal) corporate culture in a company, thecompany employees must operate as a team. The presumption of this systemfunctioning is that team members trust each other, support each other, and togetherstrive to achieve company goals. Individual employees must understand their rolewithin the company. Achievement of the situation, in which a company operates,is first and foremost the task of its management. According to Urban [5], teamorganization and utilization of teams have lately been more and more common.However, a group of people does not always behave as a team, even though it islabeled as one.

Implementation of teamwork makes sense if it can increase the performanceof employees working individually, decrease costs, accelerate communication andmake it more effective, improve decision-making etc. Appropriately introducedteamwork may serve as a tool for decreasing workloads, increasing entrepreneurialthinking, and work devotion and satisfaction of employees themselves.

A team represents a certain group of people, which is formed deliberatelyfor fulfilling certain goal, usually for a specific period of time. However, even apermanent operation of teams within a company is not an exception. Teammembers are appointed irrespectively of their position within a company andshould have an equal position in a team. Team members are responsible forcontributing to the achievement of set goals on the basis of their knowledge andskills.

Company communication plays an irreplaceable role in forming thestandpoints of employees in respect of a company. Company communication ismainly used as a tool, which can be utilized by managers to affect work positions,activity, and behavior of employees with the use of their power and authority,appropriately applied management style, effective methods of stimulation, andremuneration. The level of company communication is based on the level ofcorporate culture. According to Urban [5], healthy corporate culture presupposesopen communication, timely and constructive resolution of any conflicts, listeningto others, focusing on the subject matter and ideas (not on their originator) andwillingness to compromise. The author views the effort of employees to cooperaterather than compete as a display of healthy communication. It is very important forthe healthy communication to exist not only among employees on the same levelbut also among superiors and inferiors [6].

According to Armstrong [7], two-way communication is mainly necessaryso that the management could continuously inform employees about individualareas of corporate policy and company plans relating to such employees, andemployees could immediately share their opinions relating to the intentions and


measures of the management. The intra-company communication strategy shouldbe based on the analysis of what the management wants to say, what employeeswant to hear, and problems existing in the course of providing and receivinginformation. Such analyses may be used to identify which systems are to beformed and what educational programs are necessary for these systems to operate.According to Armstrong [7], company communication systems include electronic,written (company magazines, newspapers, bulletins, notice boards, etc.) and verbal(meetings, group briefings, etc.) systems.

The health and safety at work policies and programs are aimed at protectingemployees and other individuals affected by the company’s production/operationsfrom the danger related to their work or their association with the company. Thehealth at work programs deal with prevention of health damage due to workingconditions. According to Armstrong [7], it is necessary to achieve the highestpossible level of health and safety at a workplace, because elimination or at leastminimization of health and safety risks represents a moral as well as legalobligation of employers. In the case the top management wants to prove itaddresses the protection of company’s employees it is necessary to prepare awritten health and safety at work policy. Furthermore, it must ensure all employeesare aware of such policy — i.e. it must communicate it well. Such policycomprises [7]:

1. Declaration of objective; 2. Definition of means utilized to execute the given objectives; 3. Rules applicable to all employees.

The company’s fulfillment of values and goals in this area is examined bythe health and safety at work audit.

Finally, the attractiveness of a company as an employer may be affected byan involvement of a company in various voluntary activities (programs) aimed atemphasizing the corporate social responsibility (see Ref. [13]). This includes, forexample, implementation and utilization of quality management systems,environmental management systems, or the management systems of health andsafety at work. Companies also take part in the Corporate Social Responsibilityinitiative (or the Responsible Care program). Through the participation in theaforementioned programs and initiatives, a company demonstrates to its existingand potential employees (as well as other stakeholders) its approach tomanagement and continuous improvement of quality, environmental protection,health and safety at work, and other social aspects of business activities.Companies thus declare their position on sustainable development. Long-term,systematic application of responsible conduct in all three areas (economic,environmental, and social) brings many benefits for the company, e.g. highereconomic growth, greater transparency of business operations, lower riskmanagement costs, better relations with external entities and other stakeholders,


as well as higher loyalty and productivity of employees. Therefore, it concernslong-term investments in the company development and in higher attractivenessof a company as an employer.

Basic Results of Research Performed in Member Companies of the CACS The research aimed at corporate culture from the perspective of the contemporarymanagement was carried out in the first half of 2010. It also addressed the area ofimpact of corporate culture on the attractiveness of a company as an employer aswell as other aspects, which — in the view of respondents — affect theattractiveness of a company as an employer. The research was performed with theuse of questionnaires. It approached the representatives of management incompanies, which are members of the Czech Association of Cleaning Stations(hereinafter the “CACS”). The mission of the CACS is to contribute to higherquality of products transported in cisterns, containers, and other bulk packaging.The association is a full-fledged member of the European Federation of TankCleaning Organizations (EFTCO) as well as a collective member of theAssociation of Chemical Industry of the Czech Republic (SCHP ČR). Thefollowing member companies of the association took part in the research [14]:

! BOHEMIA CARGO, s. r. o.; ! ESA, s. r. o.; ! GS SOKOTRANS, s. r. o.; ! HARVIS, s. r. o.; ! Karel Nedorost – AQUTRUCK; ! KOVOPROGRESS, spol. s r. o.; ! Moody International, s. r. o.; ! Pražské služby, a. s.; ! SPETRA CZ, s. r. o.; ! UNIPETROL DOPRAVA, s. r. o.; and ! VADS, s. r. o.

The research resulted in the following basic conclusions:

! Respondents stated that in their view, corporate culture significantly affects thecompany’s performance. 60 % of respondents confirmed that the corporateculture formation process is systematically driven by the company’smanagement and described corporate culture as strong – i.e. strongly affecting“life” within an organization. 80 % of respondents believe corporate culture isin line with the company’s strategy. However, most companies (80 %) do notperform the corporate culture audit – i.e. do not have any feedback on themanagement effectiveness of the corporate culture formation process.


! Nine respondents believe they are attractive employers; however, only 4companies have implemented a mechanism for assessing their position as anemployer (through questionnaires or personal interviews). All respondentsbelieve that corporate culture is one of the elements which can affectattractiveness of a company as an employer.

! With regard to the human resources management strategy, respondents believestimulation and remuneration method play the most important role in theattractiveness of a company as an employer. Also important are:communication, health and safety at work, and education process.

! 80 % of respondents confirmed that, in their view, a company’s involvementin voluntary activities and program positively affects the attractiveness of acompany as an employer.

Conclusion

One of the important factors affecting the attractiveness of a company as anemployer is, without any doubt, corporate culture. The significance of corporateculture is ever increasing in today’s highly turbulent environment. It is a factor,which unambiguously affects the company’s performance and managementeffectiveness of the given organization. It is one of the relevant pillars ofsuccessful business operations. Generally speaking, it is a practical execution ofcorporate identity. All employees of a company, including the top management,take part in the formation of corporate culture. Corporate culture is one the softtools of management; it comprises certain basic elements, such as values,standpoints, behavior standards, and artifacts. It can be subdivided into severallevels depending on the extent to which individual levels are visible for anexternal observer. It is necessary to emphasize that corporate culture does notalways have to benefit a company. Sometimes, it can be a strong drag — e.g.during a change management process, implementation of innovations etc.Corporate culture must be seen in the context of the company’s strategicmanagement. Each company must realize that corporate culture affects not onlyexisting but also potential employees.

To increase the attractiveness of a company as an employer, a companymust pay sufficient attention to individual activities, which are part of the humanresources management strategy. These activities affect the quality of thecompany’s human capital. It is apparent from the performed research that thecompanies in the area under review are well aware of this. They consider thefollowing aspects to be the most important in terms of the attractiveness of acompany as an employer: method of stimulation and remuneration of employees,form of communication in a company, health and safety at work, and education ofemployees. Also relatively important is the companies’ involvement in voluntary


activities and programs, which are aimed at environmentally friendly operationsand at respecting other principles of sustainable development.

Acknowledgements

This work was supported by the Ministry of Education, Youth and Sports of theCzech Republic under project No. 1M0524 (Research Centre for Competitivenessof Czech Economy).

References

[1] Lukášová R., Nový I..: Organizational Culture (in Czech),Grada Publishing,Prague, 2004.

[2] Dytrt Z., Brodský Z..: Company Culture (in Czech), Univerzita Pardubice,Pardubice, 2008.

[3] Livian Y.F., Pražská L.: Human Resource Management in Europe –Comparison with Czech Republic (in Czech), HZ System, Prague, 1997.

[4] Stýblo J.: Contemporary and Future Management (in Czech), ProfessionalPublishing, Prague, 2008.

[5] Urban J.: Personnel Management in Organization – Personnel Dimensionof Management (in Czech), ASPI, Prague, 2003.

[6] Lošťáková H., Bednaříková M., Branská L., Dědková J., Janouch V.,Jelínková M., Nožička J., Pecinová Z., Somová J., Vávra J., Vlčková V.:Differentiated Relationship with Customers (in Czech), Grada Publishing,Prague, 2009.

[7] Armstrong A.: Human Resource Management (in Czech), Grada Publishing,Prague, 2002.

[8] Stachová A.: Personnel Management (in Czech), Slezská univerzita Opava,Karviná, 1997.

[9] Bednaříková M.: Social Aspects of Sustainable Development and TheirConnection on the Employee Turnover (in Czech). In: Proceedings fromInternational Scientific Conference Management of Human Potential inCompany, pp. 93-99, Technical University of Zvolen, Zvolen, 2008.

[10] Herzka P., Fuksová N.: MANEKO 2, 126 (2009).[11] Khelerová V.: Communication and Business Skills of Manager (in Czech),

Grada Publishing, Prague, 2006.[12] Hyršlová J., Bednaříková M.: Social Aspects of Sustainable Development

and a Corporation, Proceedings from International Scientific ConferenceSustainability Accounting and Reporting at Macroeconomic andMicroeconomic Levels, pp. 54-61, CES, Prague, 2007.


[13] Spiring K.: Social Performance and Competitiveness, A Socio-CompetitiveFramework. In: Schaltegger S., Wagner M.: Managing the Business Case forSustainability, The Integration of Social, Environmental ans EconomicPerformance, Greenleaf Publishing, Sheffield, 2006.

[14] CACS [on line], [cit. 2010-7-15]. Available from http://www.cacs.cz/.


179



16 (2010)

INCREASING OF EQUIPMENT RELIABILITYBY RCM APPLICATION

Lenka BRANSKÁ1 and Kateřina ŠILHAVÁaDepartment of Economy and Management of Chemical and Food Industry,



Over the past twenty years, maintenance has changed, perhaps more than anyother management discipline. Managers seek new methods that can beimplemented in their current maintenance systems or they strive for radicalchange of the maintenance system. In recent years, TPM (Total ProductiveMaintenance) and RCM (Reliability Centred Maintenance) have beenrecommended as appropriate methods to improve maintenance. While the TPMmethod has already been described in the literature and some examples of itspractical applications can be found, the second method (RCM) has not beensufficiently clarified even in theory and so far it has not been fully implementedin any of the Czech companies. It therefore seems helpful to state benefits of RCMby way of the example of the selected maintenance system and to outline the basicprinciples that should be observed in its application.

180 Branská L., Šilhavá K./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 179–190

Introduction

In the highly competitive environment, to be successful and to achieve world-class-manufacturing, organizations must possess both effective manufacturingstrategies and efficient maintenance [1]. To support production, maintenance mustensure equipment availability in order to produce products at the required quantityand quality levels. This support must also be performed in a safe and cost-effectivemanner [2].

Over the past twenty years, maintenance has changed, perhaps more thanany other management discipline. The changes are due to a huge increase in thenumber and variety of physical assets (plant, equipment and buildings) which mustbe maintained the world over, much more complex designs, new maintenancetechniques and changing views on maintenance organization and responsibilities.Rapidly grows awareness of the extent to which equipment failure affects safetyand the environment and awareness of the connection between maintenance andproduct quality, too.

Maintenance people are having to adopt completely new ways of thinkingand acting, as engineers and as managers. Managers everywhere are looking fora new approach to maintenance [3]. They seek new methods that can beimplemented in their current maintenance systems or they strive for radical changeof the maintenance system. The main goal of these changes is the improvement ofthe maintenance performance in an enterprise. In recent years, TPM (TotalProductive Maintenance) and RCM (Reliability Centred Maintenance) have beenrecommended as appropriate methods to improve maintenance; they can be usedseparately or simultaneously. While the TPM method has already been describedin the literature and some examples of its practical applications can be found, thesecond method (RCM) has not been sufficiently clarified even in theory and so farit has not been fully implemented in any of the Czech companies. It thereforeseems appropriate to state benefits of RCM for improving maintenance by way ofthe example of the current maintenance system in a selected company and tooutline the basic principles that should be obeyed in its application.

The main objective of this article is to define the possibilities ofimprovement of company maintenance system by application of principles ofRCM. To achieve this primary objective, the following partial goals were defined:

! describe different approaches to the maintenance of production equipmentwith a focus on new methods used in maintenance, in particular TPM andRCM,

! cover the maintenance system currently used in selected manufacturingfacilities in the company of chemical industry and evaluate the currentapplication of the approaches to its improvement,

! propose a modification of current maintenance system, applying the principlesof RCM.

Branská L., Šilhavá K./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 179–190 181

The targets thus identified will be achieved partly on the basis of literatureresearch (analysis of literary data sources and the subsequent synthesis ofestablished facts) and partly using results from primary researches undertaken inspring 2010. This qualitative survey was conducted in a company of the chemicalindustry.

The main objective of the primary researches undertaken was:

! describe the product produced on a selected manufacturing equipment and itscustomers,

! identify the major problems arising from unplanned downtime of productionequipment,

! map simultaneously carried out activities of planned maintenance on aselected manufacturing equipment including their strategic and tacticaloperational management method,

! analyze repairs after a failure in terms of type, frequency and causes and! identify the most serious failures, including possibilities of their elimination.

Primary data were collected using an interview method according to aprepared interview script. The respondents were managers of sales, production andmaintenance departments and other workers with responsibility for planning andrealization of maintenance of the selected production equipment.

Theory

Approaches to the maintenance and development of maintenance systems can bedivided into several developmental stages. Firstly unplanned (Reactive)Maintenance or breakdown repair is the practice of caring for equipment when andonly when it is not functioning properly, and there is no particular person ordepartment responsible for it. Workers take action only if machine is broken andcannot continue production. In such case production system capacity is reduced,workers are idle which causes direct labor costs to rise. Because of urgency inrepairing equipment overall costs will increase [4-6]. This system is calledCorrective Maintenance.

The first scientific approaches to maintenance management date from the1950s and 1960s. At that time preventive maintenance was advocated as a meansto reduce failures and unplanned downtime. In many companies large time-basedpreventive maintenance programs were set-up [7]. Planned Maintenance systemis always organized and always a person/department is responsible for keepingrecords and taking action in case of breakdown. The necessary spare parts havealready been purchased and kept in inventory. Maintenance procedures andmanuals are always available. In this case machine downtime and the overallmaintenance costs are less than unplanned maintenance because it tends to reduce


worker and machine idle time [5,6,8]. Preventive Maintenance (PM) is the practiceof tending to equipment so it will not break down and will operate according torequirements. It entails understanding and maintaining all physical elements ofmanufacturing so they consistently perform at the level required by design.Preventive maintenance is the work activity that has been programmed on aregular basis to inspect a system, uncover potential problems and make whateverrepairs are necessary to ensure that the system does not fail during normaloperation. The costs include personnel, inventory of parts and the lost time whenequipment is down for repairs [4-6,8]. In the 1970s condition monitoring cameforward, focusing on techniques which predict failures using information on theactual state of equipment (e.g., luboil debris analysis, vibration monitoring) [7](Dekker 1996). This approach is called Predictive Maintenance. This proved to bemore effective than the large time-based preventive maintenance programs. In the1980s the computer was brought to the maintenance function. Initially mostattention was paid to facilitating administrative processes, later on by makingmanagement information readily available (e.g., the registration of the causes forovertime); yet their influence on decision making was limited. An importantapproach worth mentioning is Reliability Centred Maintenance (RCM) [7]. RCMis defined by SAE Standard “Evaluation Criteria for Reliability-CenteredMaintenance (RCM) Processes” as “… a specific process used to identify thepolicies which must be implemented to manage the failure modes which couldcause the functional failure of any physical asset in a given operating context.” Inthe RCM approach, maintenance is carried out at the component level and themaintenance effort for a component is a function of the reliability of thecomponent and consequence of its failure under normal operation. The approachuses failure mode effects analysis (FMEA) and to a large extent is qualitative[9,10]. Any RCM process shall ensure that all the following seven questions areanswered satisfactorily and are answered in the sequence shown as follows:

1. What are the functions and associated desired standards of performance of theasset in its present operating context (functions)?

2. In what ways can it fail to fulfill its functions (functional failures)?3. What causes each functional failure (failure modes)?4. What happens when each failure occurs (failure effects)?5. In what way does each failure matter (failure consequences)?6. What should be done to predict or prevent each failure (proactive tasks and

task intervals)7. What should be done if a suitable proactive task cannot be found (default

actions)? ([3]).

After these questions are answered, maintenance activities are optimized,i.e. maintenance strategy is determined for each fragment of production facilities,and activities are defined to ensure maximum reliability of the various parts of the


production equipment (and in consequence of it, the equipment as a whole).According to Moubray, RCM thus serves to find the economically optimal way ofmaintenance of each manufacturing equipment in the long term [3]. At the same time, the Japanese evolved the concept of Total ProductiveMaintenance (TPM) in context of manufacturing [10,11]. Total ProductiveMaintenance (TPM) is a well defined and organized maintenance program, whichplaces a high value on team work and continues improvement. Specific actionsrequired for restoring equipment to a like-new condition, having operatorsinvolved in maintenance, training the labor force and, using the overall preventivemaintenance effectively [5,6]. A goal of TPM is to upgrade equipment so itperforms better than new ones. In TPM the maintenance responsibility is spreadover many departments such as production, engineering, and maintenance, and toa range of people, especially operators and shop workers. In TPM, operatorsperform basic equipment repairs and team of maintenance staff redesign andreconfigure equipment to make it more reliable and easier to maintain [4,6,8]. BothRCM and TPM view maintenance in the broader business context and take intoaccount the link between component failure and their impact on the businessperformance [10].

Experimental

The primary research has studied the recently used maintenance system in selectedmanufacturing equipment in the chemical industry. Manufacturing equipmentprocesses the output of the previous production process of the enterprise andproduces a product for sale to its customers (manufacturers of products forconstruction industry). The product is made by mass, process production. Themain part of the manufacturing equipment is the reactor and accessories (pumps,ventilators, gas piping, inlet and outlet pipes for the material, compressors andcoolers).

Maintenance of the given production equipment is relatively important forthe enterprise because its failure means emergence of some serious problems,especially:

! The need to shift production to other production equipment, originallydesigned to produce special products. However, it will cover only 25%capacity of the monitored manufacturing device. Consequently, meetingcustomer requirements is at risk.

! The need to deliver products to customers in the required time and amountusing pre-sale from a competing manufacturer. This has a direct economicimpact on corporate profits, while the enterprise simultaneously puts at riskits share of expenditures with individual customers.

! Establishment of stock at the entry to the production equipment and the needto shutdown the previous process when reaching the storage capacity.


The enterprise assesses the importance of individual production facilities,by the criteria of irreplaceability, importance of the product in the product line andthe frequency of failures. Based on this evaluation, the given manufacturingequipment is assigned with a maintenance strategy. The monitored productionequipment is rated as indispensable and, consequently, the decision is taken on theapplication of preventive maintenance. Within the maintenance, activities areperformed such as:

! cleaning (after shutdown), ! replacement of oils and lubrication (by the operators on the manufacturing

equipment),! technical inspections

• (partly by production equipment operators in which monitor noise andleakage is monitored and vibration is assessed subjectively and

• partly by external maintenance workers, under which also minormaintenance tasks are carried out, such as replenishment and replacementof oils and tightening of valves and pumps seals).

! Then, overhauls are carried out and ! preventive repairs within the shut-down, based on findings during the removal

of production equipment and the results of predictive maintenance conductedby company diagnostician.

In the maintenance, the company utilizes both its own staff (operators on themanufacturing equipment and maintenance staff), and maintenance outsourcing.External service workers provide approximately 80 % of the maintenance.

Preventive maintenance is planned; the planning is carried out in thehorizon of strategic, tactical and operational. In planning, it is necessary to respectfinancial constraints though, and therefore the business makes use of deferredrepairs risk assessment methodology. If it is necessary to reduce the cost ofplanned maintenance, repairs are excluded which pose only a negligible risk.Subsequently, a plan of repairs is made, from which tasks are then directly derivedfor both internal and external maintenance staff. If on the basis of appliedpredictive maintenance a need arises of an originally unplanned repair, the plan iscorrected. The planned preventive maintenance activities then accumulate until theshutdown of production equipment. Its activities are planned and regulated witha shutdown schedule that is processed for each manufacturing equipment and forindividual departments. Maintenance tasks performed within the plannedpreventive repairs are allocated in the form that also serve for the operationalmaintenance records. Repairs check has several forms. Quality control isperformed of the work done as well as repair time process inspection. Also, areport is compiled on the operation stop process, maintenance budget check andthe overall assessment of corporate maintenance using performance indicators formaintenance.


Corrective maintenance is performed after it is reported, either immediatelyor it is planned into activities of the nearest shut-down for repair. The mostsignificant failures on the manufacturing equipment (in terms of absolutefrequency) were identified by the company to be pump filter basket clogging andsteam pipe defects. The company strives to reduce the number of repairs after afailure through the elimination of the aforementioned defects. A reserve pump wasacquired, so in the case of failure a repairing method can be applied of“interchangeable manner”. Defects in the sealing could be solved by replacing thesealing with the best commercially available seals. This seal, however, presentsrisks in terms of health and safety at work. Therefore, the enterprise in the case ofthe pipe rather applies the predictive maintenance, but the problem is that thedefect may occur at a place different from that where the diagnostics of the pipeperformed.


If we evaluate the maintenance system of the given production equipment, we canstate that it is based on a combination of theoretical approaches outlined. It isbased on the preventive maintenance, which also includes elements of predictivemaintenance. At he same time, principles are also applied of the Total ProductiveMaintenance approach, especially the autonomous maintenance (consisting of theinvolvement of operators in simpler maintenance activities). Despite thesophistication of the entire maintenance system of the selected manufacturingequipment, however, the company does not avoid unplanned repairs, i.e. it is alsonecessary to simultaneously apply the corrective maintenance.

The company strives to eliminate unplanned shutdowns, but only throughthe elimination of most frequent failures. However, evaluation of the importanceof individual failures should be comprehensive. It should be aimed at assessing theoverall impact of each failure on the company (in a similar way as it is done in thecontext of enterprise-wide strategy in setting maintenance for each productionequipment). Overall, it can be concluded that the system of maintenance ofproduction equipment includes all the principles of theoretically cited approachesto maintenance. Differentiation of the maintenance work is carried out only in theenterprise-wide context, i.e. it concerns individual manufacturing facilities as awhole. Application of RCM method could contribute to improving themaintenance system of both the selected manufacturing equipment and thecompany.

Applying the RCM principles for maintenance of the selected productionequipment would appear very desirable because of its indispensability. This wouldmean to convert the current maintenance system based on the standards of repairs(which determine the frequency, scope and time of the repair) to the repair system


favouring a response to the current status of the production equipment (measuredby the predictive maintenance).

In modifying the current method of maintenance (of the selectedmanufacturing equipment), the production equipment should be first identified asa set (in terms of maintenance) of individual parts. Subsequently, failure rate isassessed of these components and importance is examined of each failure. Thisimportance is not given by the frequency of failures, but by their overall impacton the business. The objective of this evaluation is to estimate the risk posed byeach of the failures. In the evaluation, it is possible to use the Risk AssessmentMatrix. For each defect is determined:

! the overall impact on the firm (in terms of safety, damage to property,environment and the company goodwill) and

! the probability that a failure occurs.

Table I An example of Risk Assessment Matrix

Consequence Probability

1 2 3 4 5

A B C D E

People Assets Environment Reputation

No healtheffect / Injury

No damage No effect No impact

Slight Slight Slight effect Slight

Health effect/ injury

Damage Impact

Minor healtheffect / injury

Minordamage

Minor effect Local impact

Major healtheffect / injury

Localiseddamage

Localized effect Regionalimpact

Permanenttotaldisability

Majordamage

Major effect NationalImpact

1 or morefatalities

Extensivedamage

Massive effect World wide

1 – Never heard of in industry; 2 – Heard of in industry; 3 – Incident had occured in our company; 4 –Happens several times per year in our company; 5 – Happens several times per year in location

In the impact of the failure on the goodwill of the company, the impactshould be considered on the most important key stakeholders, especially suppliersand customers as well as the public in the neighbourhood of the business. Theexample of Risk Assessment Matrix is shown in Table I.

If the size of risk is identified for each defect, appropriate activities are


defined to eliminate the most serious failures (or to lessen their consequences).Simultaneously, an optimal maintenance strategy is determined for each part of theproduction equipment. For parts of the production equipment where there is a lowrisk it is possible to use Corrective Maintenance System. In case of parts with themost serious risk, it is appropriate to apply a maintenance system leading tomaximum reliability.

If it is desirable to provide maximum reliability for specific parts of theproduction equipment, the repair must be performed before the limit wear isreached. To determine the actual date of correction, it is necessary to use:

! past data on the prevalence of individual failures (and to process it throughmathematical and statistical methods) and

! information about the current state, which can be identified using predictivemaintenance.

The enterprise has already experienced predictive maintenance and couldtherefore relatively easily define the technical parameters for monitoring the rateof wear and also a place in which to perform this measurement. By the way ofexample of the predictive maintenance application in the piping it is clearlyevident that the suitability of the parameter and the number of points at which themeasurement is made must be continuously reassessed.

A fixed term of the repair should ensure high reliability of the productionequipment by exercising only the necessary amount of money for repairs. Therepair is carried out only when it is strictly necessary, but before the failure of themanufacturing equipment.

Providing a different maintenance strategy for each part of the productionequipment that leads to different schedules of repairs would, however, involveextending the total repair time (and reduce the time available for production). Itis therefore necessary to synchronize the performance of maintenance on thecomponents. This synchronization is provided in merging terms of repairs to oneterm, or to a certain maintenance cycle. Generally, it is possible to adapt to theshortest time, or to the maintenance cycle, and to plan a repair before the first limitwear is reached. But it is also possible to apply the Theory of Constrain approachand fix the date of repairs under the main part and to synchronize otherrecommended deadlines with it. In this case, however, it is necessary to considerthe consequences of postponing the repairs, as postponing the recommended termmay lead to repair costs increase as a result of the damage that would have beenprevented by an earlier repair. It is therefore necessary to carry out a thoroughanalysis of the results of synchronization of terms to repair individual parts withthe main part and then to apply measures in which it is possible. These measuresare to allow extending or shortening the time for repair. In an effort to extend, thebasic question is: What should be done to prevent the production equipment or itspart from exceeding the limit wear? For example, the material from which the


manufacturing equipment is made may be replaced with higher-quality one, ordoubling of the device may be performed. Conversely, if it is desirable to make arepair before it is recommended for the part, one may ruminate contrariwise andmake use a low quality material form which the spare part is manufactured.

An extremely important part of the maintenance management is providingthe spare parts and materials for the maintenance. For the growth of maintenanceperformance it is possible to apply a new system of inventory management ofspare parts which are not used by the enterprise yet. Especially for strategic spareparts, Quick Response and Collaborative Planning, Forecasting andReplenishment may be used as methods aimed at JIT-based stock replenishment.The basis of these stock replenishment systems is the supplier’s response to thecurrent level of inventories. The supplier and buyer share the information onplanned maintenance, allowing suppliers to prepare the future need for spare parts.They also share information on the current level of stocks and its approximationto the ordering limit. At the instant it is reached, the customer makes an order orthe inventory is replenished directly by the supplier. Spare parts that are notdesignated as strategic may be replenished based on demand forecasting.Implementation of this method of inventory management of spare parts allowsreadiness not only for the planned maintenance, but also for correctivemaintenance, whilst reducing maintenance costs.

Conclusion

Based on the results of the primary research, it can be stated that the company usesa maintenance system that is primarily aimed at preventing and combines severaltheoretically described systems (Preventive Maintenance, Predictive Maintenanceand Total Productive Maintenance). By the example of selected productionequipment it is clear that despite the elaborateness of the current maintenancesystem it is not possible to ensure full reliability of the production equipment.Therefore, the enterprise must also apply corrective maintenance and then addressthe problems resulting from the failure of the production equipment.Implementation of the RCM method should significantly improve the reliabilityof the production equipment.

If it is successfully implemented, it is possible to expect not only greaterreliability of the production equipment, but consequently also greater continuityof production processes and less loss of production, improvement of the supplier-customer relations, reduction of the impact on the environment and increase ofsafety at work. Another benefit for the enterprise will be a greater awareness of themanufacturing equipment and involvement of more company workers into the careof the production equipment.

On the contrary, one can not say unequivocally that the implementation of


RCM will bring significant cost savings in maintenance and if that happens, thenonly in the long term. When considering the economic impact of the introductionof the method, one should also take account of costs related to the actualimplementation of the new system. Generally, it is therefore recommended wheredowntimes of the manufacturing equipment bring about extreme or significantimpact on the business, such as in companies seeking corporate and inter-companyinterconnection of material flows, for example on the basis of Quick Response, orCollaborative Planning, Forecasting and Replenishment or in companies with alarge number of environmental accidents. On the contrary, it is not suitable forbusinesses with high share of the production equipment that is subject to revision,since their maintenance is determined by legislative action.

Nevertheless, an appropriate implementation of RCM, especially incombination with TPM, contributes to improved corporate maintenanceperformance and the company as a whole. It can therefore be expected that in thecoming years it will see expansion also in Czech, not just chemical, companies.

References

[1] Moubray J.: Twenty-first century maintenance organization: Part I - theasset management model. February 2003. http://www.mt-online.com/component/content/article/201-february2003/1089-21st-century-m a i n t e n a n c e - o r g a n i z a t i o n - p a r t - i - t h e - a s s e t - m a n a g e m e n t -model.html#backtoarticle (24.6.2010)

[2] Pintelon L., Gelders L.: Eur. J. Oper. Res. 58, 301 (1992).[3] Moubray J.: Reliability-Centered Maintenance, 2nd Edition, Industrial Press,

New York, 1997.[4] Nicholas, J.M.: Competitive Manufacturing Management, Irwin/McGraw

Hill Publishing Co., New York, 1998.[5] Walker, D.L.: Operations Management, a Supply Chain Approach,

International Thomson Business Press, Boston, MA, 1999.[6] Moayed F.: Comparison of Maintenance Operations in Lean vs. Non-Lean

Production Systems, Proceedings IIE Annual Conference. 1102 (2009).[7] Dekker R.: Reliab. Eng. Syst. Safe. 51, 229 (1996).[8] Wireman, T.: Developing Performance Indicators For Managing

Maintenance, Industrial Press, Inc., New York, 1998.[9] Moubray J.: Reliability Centered Maintenance, Butterworth/Heinemann,

Oxford, 1991.[10] Murthy D.N.P., Atrens A., Eccleston J.A. Strategic Maint. Manage. 8, 287,

(2002).[11]Tajiri M., Gotoh F.: TPM Implementation, McGraw-Hill, New York, 1992.[12] Itakura S., Nikola S., Magori H., Iba K., Chen L., Shirai G., Yokoyama R.:


A Strategic Reliability Centered Maintenance for Electrical Equipment in aChemical Plant, 9th International Conference on Proabilistic MethodsApplied to Power Systems KTH, Stockholm, Sweden – June 11-15, 2006.http://www.labplan.ufsc.br/congressos/PMAPS/files/pdf/3.2/3.2_niioka.pdf

[13] Ahuja I.P.S., Khamba J.S.: Int. J. Qual. Reliab. Manage. 25, 709 (2008).[14] Šilhavá K.: Maintenance in Company and Possibilities of Its Improvement

(in Czech), Diploma work. University of Pardubice, Pardubice, 2010.


191



16 (2010)

COST BENEFIT ANALYSIS AND THE SOCIAL TIMEPREFERENCE RATE TO ESTIMATE SOCIALDISCOUNT RATE IN THE CZECH REPUBLIC

Petr FRANC and Liběna TETŘEVOVÁ1

Department of Economy and Management of Chemical and Food Industry,The University of Pardubice, CZ–532 10 Pardubice

Received July 1, 2010

The choice of the discount rate is a crucial issue for evaluating projects with long-term impacts. The paper deals with a calculation of a social discount rate for theCzech Republic based upon the Social Time Preference Rate (STPR), from whichperspective the critical components of the STPR are: the elasticity of the marginalutility of consumption, the growth rate of per capita real consumption, andmortality based utility discount rate. Estimates turn out to be 1.36; 2.9 and 1.31% respectively yielding an overall figure of 5.25 %.

Introduction

Cost-benefit analysis for project and policy evaluation proceeds in two steps. First,one needs to estimate the costs and benefits of a project or policy at each point in

192 Franc P., Tetřevová L./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 191–201

time. Second, one needs to compare these costs and benefits across time. Timediscounting in the public sector remains a source of confusion and some academiccontroversy. The very concept of a “social” discount rate, not revealed by themarket, is rejected by mainstream financial economics. Elsewhere the setting ofpublic sector discount rates equal to the commercial return on private investmentcontinues to have wide appeal. Both these approaches are flawed. More widelyfavoured by experts in the field today is a rate derived as the sum of pure timepreference for marginal utility and a factor reflecting the decline in marginal utilityof income as per capita income increases. However, controversy continues aboutpure time preference, especially in the absence of empirical data on people’s social(as opposed to individual) preferences.

Cost-Benefit Analysis

The presence of market failures is usually considered, along with redistribution,as the main rationale for public sector involvement in the economy. For instance,whenever competition is imperfect, production or consumption generateexternalities, non-excludability and non-rivalry make impossible or undesirablecharging users for the provision of a good, then the government intervention canin principle result in a more efficient allocation of resources thereby potentiallyenhancing social welfare. When there is a case for public involvement, the costsand the benefits of the envisaged intervention should be carefully identified andcompared in order to ascertain whether the latter are likely to outweigh the former.This is the main aim of CBA as an evaluation tool to assist decision-makers tomake rational choices about public resources allocation. [1]

CBA is a policy assessment method that quantifies in monetary terms thevalue of all policy consequences to all members of society. The net social benefitsmeasure the value of the policy. Social benefit (B) minus social costs (C) equal netsocial benefits. [2]

Social Discount Rate

When evaluating government policies or projects, analysts must decide on theappropriate weights to apply to policy impacts that occur in different year.

The social discount rate, which measures the relative value of communalconsumption at different points in time, is one of the most critical parameters inCBA and it is not surprising, therefore, that so much controversy has centred onthe concept of social discounting over the years [3].

Different discounting practices by governments have resulted in theapplication of some widely divergent social discount rates (SDR) across European

Franc P., Tetřevová L./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 191–201 193

countries. In 2002, for example, the French rate, based on the marginal product ofcapital, was 8% while the German rate, based on recent values of the real long-term government bond rate, was just 3 %. The official rate for the UK, acompromise between cost of capital and time preference considerations, was 6 %.The French government followed suit in 2005 by reducing its official rate to 4 %.There is near convergence now between the official discount rates of threeimportant EU member countries. [4] The Czech Republic has not decided tofollow any of existing concepts and usually follows recommended value by theEuropean Commission which does not necessarily have to precise.

A major reason why the quality of CBA varies widely is inconsistent use ofthe SDR. What is the foundation of a communal rate? There is a long-windeddebate in economic literature about this issue. There are, basically, four ideas [5]:

! the market rate of interest;! the government borrowing rate;! the social opportunity cost rate (SOCR);! the social time preference rate (STPR).

Social Time Preference Rate

Long-lived projects typically involve a sacrifice of consumption by the presentgeneration in order to generate benefits for future generations. To decide whetherthe sacrifice is warranted, society must weigh the current loss in consumptionagainst the future gains. Ramsey, Marglin and Arrow argue that society shouldtreat all generations’ welfare equally but should consider that future generationswill likely have higher per capita consumption than the current generations due toongoing economic growth. Consequently, assuming that consumption hasdeclining marginal utility, consumption by a future generation should have a lowerweight than consumption by present generations, where the rate at which theweights decline over time is proportional to the growth rate of per capitaconsumption - the higher the growth rate, the higher the SDR. [2]

The use of STPR as the social discount rate, supported by Marglin,Diamond and Kay, is based on the argument that public projects displace currentconsumption, and streams of costs and benefits to be discounted are essentiallystreams of consumption goods either postponed or gained. Two alternativemethods have been suggested for empirical estimation of STPR. One is toapproximate it by the after-tax rate of return on government bonds or other low-risk marketable securities. The second, more usual, is Ramsey formula.

Definition of Ramsey formula is basically based on CBA outcomes inrelation to Social Welfare concept. Primary CBA financial indicator is NPV (NetPresent Value). NPV can be defined as [6])


(1)

where b is project benefits, c project costs, r SDR and t time.For further analysis we assume that at any time t the net benefits from a

project are available for consumption (Ct – consumption per capita).Project NPV is then rewritten as

(2)

In equation (2) the discount rate r receives a second interpretation. It is the rate atwhich the value of a small increment of consumption falls as time changes and,hence, it is called consumption discount rate (CDR). Note now, that the socialdiscount rate — the rate with which we should discount to evaluate a publicproject — is equal to the consumption discount rate.

To connect NPV with social welfare analysis there can be used additiveParetian social welfare function (W), expressed by integration function orweighting function of consumption in different periods, and consumption growthin one period without fall in other period is considered improvement [7]

(3)

or

(4)

where U(t) is a time invariant utility function and D is the utility discount rate(UDR). Utility function is defined by the equation

(5)

where 0 (0 # 0 < 4) is the elasticity of marginal utility with respect toconsumption.

The weight of utility from consumption declines over time with rate D. UDRis the rate at which the value of a small increment of consumption falls as time


changes. This brings out clearly that the CDR and UDR are different concepts witha certain relation called Ramsey formula

(6)or

(7)

where g or is the growth rate of consumption per capita.

Utility Discount Rate (DDDD)

There is a great deal of discussion and controversy over what value(s) D shouldtake. It represents the rate at which society discounts future generations’ welfare,even if all generations have equal consumption per capita (i.e. g = 0). It is usuallyunderstood as a concept involving two components [8]

(8)

where * is Pure Time Preference Rate – PTPR and L is Changing Life Chance.The first component in formula (8) is *, reflecting the rate at which individualsdiscount future consumption over present consumption, on the assumption that nochange in per capita consumption is expected [9].

According to Spackman [10] some authors (Ramsey, Pigou, Solow, Kula,Price, Broome and Cline) are of the opinion that * = 0, for the reason that positive* gives future generation benefits less value than the current one and such ideaconsider ethically indefensible. Other authors (e.g., Arrow) object that a zero rateof pure time implies a patently unrealistic level of investment. It implies,regardless of the return on investment (provided the return is positive), a savingsrate of 1/0. A plausible value for 0 of around 1.5 % thus implies a savings rate ofabout 2/3. [10] This was also Ramsey’s original outcome but such savings rate isunacceptable [11].

Official methodical approach in Great Britain [9] and other authors (e.g.Scott) suggest that long-run savings behaviour in the UK is consistent with a valueof * of 0.5 %. This component of the social time preference rate is the leastamenable to empirical analysis, but the literature suggests that the range is 0.0-0.5% [12].

The second component, catastrophe risk (L), is the likelihood that there willbe some event so devastating that all returns from policies, programs or projectsare eliminated, or at least radically and unpredictably altered. Examples are


technological advancements that lead to premature obsolescence, or naturaldisasters, major wars, etc. The scale of this risk is, by its nature, hard toquantify [9].

Thus some authors, such as Kula, look at the increasing risk of death for anindividual as they get older. While this will certainly be an important risk of deathfor an individual to favour early consumption over later, it is far from clear whatrole this should play in discussions about the discount rate. Another Newberyfocus on a death rate dividing by the population [8]

(9)

Estimates of the parameter D differ among authors and depend on the method ofcalculation. Some studies use decomposition (* and L), others understandparameter D as the whole.

Elasticity of Marginal Utility (0000)

The parameter 0 represents a social evaluation of the intergenerational distributionof income. It summarizes the key value judgment about how quickly the marginalutility of consumption (dU/dC) declines as average consumption rises [2].

We assume a positive but strictly decreasing marginal utility of consumption(diminishing marginal utility). Formally [8]

(10)

but

(11)

0 then measures the percentage rate at which the marginal utility falls for everypercentage increase in consumption. Formally

(12)

The empirical work on 0 involves three fundamentally different approaches:direct survey methods, indirect behavioural evidence and revealed social values.For the purpose of this work, method revealed social values will be analyzed and


used for SDR calculation. Although other methods have been used worldwide tocalculate 0, revealed social values have been found by the authors the mostunderstandable and easy to apply in practice.

A suitable value for 0 may be revealed through government spending or taxpolicies. For example, the extent of progressiveness in a country’s personalincome tax rates can be viewed as a reflection of the government’s degree ofaversion to income inequality (a measure of 0). According to Evans [13] Stern,Cowell and Gardiner have produced estimates of 0 for the UK using personalincome tax data. In both cases, the tax structure is assumed to be based on theprinciple of “equal absolute sacrifice of satisfaction” and, in common with mostresearchers, iso-elastic utility functions are assumed. The model is set out formallybelow [13]

(13)

where T(Yt) is the income tax function reflecting the tax liabilities of andindividual, K is the constant and Yt taxable income.

Furthermore, if utility functions are typically iso-elastic, then

(14)

Taking the total differential and logs of equation gives

(15)

where t is the marginal tax rate and T(Yt)/Y is the average rate of income tax. Strengths of this approach are its conceptual simplicity and measurability,

and that it may also include concern about fairness as well as marginal utility; butit has two evident limitations. The first comes from the idea that social concernabout contemporary inequality might differ from that about inequality over time[10]. The second concerns the fact that de jure statutory tax rates may be quitedifferent from de facto paid tax rates. This is likely to be the case in countrieswhere tax evasion is a serious issue, particularly if individuals with differentincome levels have different chances to evade [14].


Other problem is when tax system is reformed too often and makes acalculation of 0 more complicated for the reason of insufficiency of exact date ina row in the timeline. For example, Czech tax reform held in 2008 brought to anexistence so called “super gross wage” which will influence a calculation ofmarginal tax rate and make remarkable modification in calculation necessary.

Annual Rate of Per Capita Real Consumption Growth (g)

It is not easy to predict an annual economic growth for period of 20 and moreyears for any country. It is possible to base this prediction on the real economicgrowth involving enough data to involve as many as middle-term economic cyclesas possible. Such analyses might be then a basis for a reasonable method of anannual real consumption per capita rate tendency.

To estimate values of g, regression analysis is an ideal tool

(16)

where Ct is real per capita consumption growth in t year, t is years and B constant.A result of the regression analysis is a value g and a correlation coefficient or anindex of determination in the case of non-linear trends.

SDR Estimate for the Czech Republic

A calculation of D is based on formula (8) and as such will be understood as a sumof two components, * and L. Estimate of * will be taken from the literature in theamount of 0.25 as the middle value of the most common interval. According to theCzech Statistical Office a death rate between 1997 and 2007 amounted to 10.6 (per1.000 persons), i.e. L = 1.06. Based on formula (8), D = 1.31.

Parameter 0 will be calculated using Revealed Social Value method for itsdata availability and possibility to compare results with European Commissionofficial recommendations coming from this approach as well. According to theOECD statistics in the Czech Republic:

a) average wage was 250 262 CZK;b) annual taxable income (Yt) was 218 979 CZK; c) the income tax: T(Yt) was 33 157 CZK.

Annual taxable income slightly exceed an interval (121 200-218 400 CZK)for the marginal tax rate 20 % and therefore this rate can be regarded as t. Usingformula (15), 0 amounted to 1.36. This value reaches average value estimate 1.35by Evans [4] for 20 OECD countries including the Czech Republic for which the


interval was set to 1.22-1.36.The Czech economy underwent a dynamic transformation period from the

beginning of the nineties. Due to this process there is not a sufficiency of relevantdata of a consumption development as in many other countries especially in theEU-15. Formula (16) will be used to estimate parameter g, because it betterrepresent a trend and straightens variations than evolution or a division betweenlogarithms of the last and first value of population consumption. A referentialperiod is 13 years, 1995-2007. A coefficient of correlation R2 amounted to 0.9536and shows a strong linear trend. A value of parameter g got from the regression atvalue of 2.9 % can be considered (in respect to a quality of disposable data)relevant, see Fig. 1.

Fig. 1 Calculation of the g parameter with linear regression method

Based on formula (6): r = 1.31 + 1.36×2.9 = 5.254, i.e. 5.25 %

Social discount rate for the Czech Republic based on STPR approach,Optimal Growth Rate model and method of Revealed Social Values for the 0estimate (2006 data) was estimated as 5.25 %.

Conclusion

The choice of the social discount rate plays a critical role in cost-benefit analysisand project evaluation, and has been a subject of intense debate for the last severaldecades. In a perfectly competitive world without market distortions, the marketinterest rate is the appropriate SDR. In the real world where markets are distorted,there are at least four alternative approaches in the choice of the SDR. Economists

ln C(t) = 0,029t+ 13,79

R2 = 0,9536

13,60

13,70

13,80

13,90

14,00

14,10

14,20

14,30

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Ln C(t)


have not reached a consensus as to which is the most appropriate.This work focused on STPR approach. That is because the European

Commission put an emphasis on this method in the process of evaluation whensubjects of public and private sector apply for the EU funds financial support. Asthe major components of the STPR were indentified: the utility discount rate(applied a risk of death indicator to estimate parameter L), the elasticity of themarginal utility of consumption (applied Reveal Social Value method), and thegrowth rate of per capita real consumption (applied regression analysis on the 13years referential period). Final estimates turn out to be 1.31; 1.36 and 2.9 % withoverall figure of 5.25 %.

SDR 5.25 % is much closer to recent studies of SDR in the EU and issupported by the European Commission recommendation.

References

[1] Mairate A., Angelini F.: Cost-Benefit Analysis and EU Cohesion Policy[online], [cit.: 2010-28-06],available at URL: <http://www.economia.unimi.it/uploads/wp/MAIRATEANGELINI2006_34.pdf>.

[2] Boardman A., Greensberg D.H., Vining A.R., Weimer D.L.: Cost-BenefitAnalysis: Concept and Practice, 2nd edition, Prentice Hall, New Jersey, 2001.

[3] Kula E.: Regional Welfare Weights in Investment Appraisal – The Case ofIndia [online], [cit.: 2009-03-17], available at URL: <http://www.jrap-journal.org/pastvolumes/2000/v32/32-1-6.pdf>.

[4] Evans D.J.: Fiscal Studies 26, 197 (2005).[5] Kula E.: Social Discount Rate in Cost-Benefit Analysis [online], [cit.: 2008-

10-02], available at URL: <http://www.economia.unimi.it/uploads/wp/KULA-

2006_19.pdf>.[6] Zhung J., Liang Z., Lin T., Guzman F.: Theory and Practice in the Choice

of Social Discount Rate for Cost-Benefit Analysis: A Survey, AsianDevelopment Bank, Manila, 2007.

[7] Creedy J., Guest R.: The Economic Record 84, 109 (2008).[8] Pearce D., Ulph D.: A Social Discount Rate for the United Kingdom,

CSERGE, Norwich, 1995.[9] HM Treasury: The Green Book – Appraisal and Evaluation in Central

Government [online], [cit.: 2007-03-24], available at URL:<http://greenbook.treasury.gov.uk/index_terms.htm>.

[10] Spackman M.: Social Discount Rates for the European Union: An Overview[ o n l i n e ] , [ c i t . : 2 0 0 9 - 1 5 - 0 2 ] , a v a i l a b l e a t U R L :<http://www.economia.unimi.it>.


[11] Arrow K.: Intergenerational Equity and the Rate of Discount in Long-TermSocial Investment, IEA, Stanford, 1995.

[12] Oxford Economic Research Associates: A Social Time Preference Rate forUse in Long-Term Discounting, Oxford Economic Research Associates,Oxford, 2002.

[13] Evans D.J.: Social Discount Rate for the European Union [online], [cit.:2010-02-14],

available at URL: <http://www.economia.unimi.it/uploads/wp/EVANS-2006_20.pdf>.

[14] Lopez H.: The Social Discount rate. Estimates for Nine Latin AmericanCountries [online], [cit.: 2010-15-04], available at URL: <http://www-wds.worldbank.org/external/default/WDSContentServer/IW3P/IB/2008/06/03/000158349_20080603084938/Rendered/PDF/wps4639.pdf>.


203



16 (2010)

QUALITY COSTS MONITORING – PILOT PROJECTIN LASSELSBERGER’S PRODUCTION PLANT

Jaroslava HYRŠLOVÁa, Marie BEDNAŘÍKOVÁb1

and Patricie TESNEROVÁc

aCentre for Economic Studies of the College of Economy and Management,CZ–158 00 Prague,

bDepartment of Economy and Management of Chemical and Food Industry,The University of Pardubice, CZ–532 10 Pardubice,

cDepartment of Economy and Management of Chemical and Food Industry,Institute of Chemical Technology, CZ–166 28 Prague


The costs of quality play an important role in the management system, as theyindicate the level of management in relation to quality, reveal potentialpossibilities for savings and help to raise the effectiveness of the decision-makingin the area of quality management. The aim of this article is to present the pilotproject of the quality costs monitoring system in the production plant of acompany that produces ceramic tiling. The company is the largest manufacturerof ceramic tiles and paving in the Czech Republic and one of the biggest Europeanmanufacturers of tiling materials.

204 Hyršlová J. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 203–217

Introduction

Quality costs are perceived in various ways (with regard to historical development,authors’ approach and the requirements of company management). Nenadál speaksabout quality costs as financial measurements in quality management systems. Inhis approach, the quality costs are all financial resources that the supplier orcustomer has to spend on the processes of the securing and/or improving of thequality of its products [1].The definition provided by Wong corresponds to thisapproach: The costs of quality are all costs spent by the company to ensure thatthe overall concept of the product provided to the customer really meets theirrequirements [2]. On the other hand, Crowley states that the costs of quality arethe difference between the current revenues and the revenues at the moment whenall customers are always satisfied [3]. The essence and importance of themonitoring of the costs of quality as a management tool (and not justquantification of value) was captured by Atkinson, who regards quality costs asthe link between quality management and financial targets and companytargets [4].

The content of quality costs — i.e. their internal structure (see, e.g., Refs [5-7]) — is also perceived differently. As a rule, the formulation of the structure ofquality costs is based on the following prerequisites [3,6]:

! Low quality costs the company money, while good quality earns money for thecompany.

! It is usually cheaper to provide high quality products. The costs ofimprovement are spent only once, while the costs of the removal ofinsufficiencies and/or defects are spent repeatedly.

! Each flaw (defect) has its cause; these causes may be removed – flaws(defects) may be prevented. Prevention is always cheaper.

! The cost incurred by the customer may be substantially higher than the costof the rectification of the defect; to satisfy the customer’s requirements, it isnot sufficient to monitor the company’s costs and revenues, but the costs thatthe customer will have to spend on the product over its entire life cycle mustbe monitored as well.

In order to be able to play the role of a management tool (to provideinformation that supports decision-making), the cost monitoring system has tocover all substantial costs related to the quality management system in thecompany, the costs of internal and external defects (i.e. costs of non-quality), aswell as the costs incurred by the users of the product. The system set in this wayforms an important part of the quality management (and management in general).It enables the setting of an optimal quality management system, contributestowards the elimination of costs related to low quality production and helps toidentify the possibilities for improvement and to optimise company activities

Hyršlová J. et al./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 203–217 205

(processes). In order for the cost monitoring system to be of benefit for thecompany, significant cost items must be identified and relevant responsibilitiesmust be defined, with the aim of using the acquired data for the improvement ofindividual corporate processes [3,5,8].

The company may use various models within the quality costs monitoringsystem (see, e.g., Ref. [1]):

! PAF (Prevention, Appraisal, Failure) model. This model focuses on themonitoring of the costs of internal defects, the costs of external defects,control costs and prevention costs.

! COPQ (Cost of Poor Quality) model. The model is based on the prerequisitethat the failure to meet the requirements causes substantial economic lossesto producers; the model recommends monitoring the costs of internal defects,the costs of external defects, the costs of wasted investments and environmentdamages (the costs incurred in relation to the non-compliance withenvironment protection laws and its return to the original condition; theseinclude also the costs of the treatment of job-related illnesses etc.).

! Process costs model. The model works with the costs of compliance (i.e. realcost for the transformation of inputs into outputs in the most effective manner)and the costs of non-compliance (wasted time, material and capacities —related to the creation of non-compliances within processes).

! Lifecycle costs model. This model focuses also on the costs incurred by theuser; this model makes sense only by a limited group of products withforeseen usage period of more than a year, where the costs of assembly,operation and maintenance are not negligible in comparison to the acquisitionprice of the product.

The quality costs monitoring system is usually implemented in the followingsteps [6,1,2]:

1. Appointment of team and proposition of the basic concept of the quality costsmonitoring system.

2. Presentation of the concept to top management.3. Creation of implementation plan.4. Selection of a part of the company for the pilot project.5. Presentation of the aim of the system to the management of the part of the

company selected for the pilot project.6. Identification of major quality costs and a proposal of their internal structure.7. Collection of required data (available from the existing company information

system).8. Proposal of the system output format.9. Proposals of adjustments of the information system in the company so that

additional relevant data may be obtained.


10. Selection of a period for which the data will be collected and evaluated.11. Compilation of a quality costs report (incl. assessment) and presentation of the

report to the management.12. Modification of the monitoring system and cost reporting so that it supports

decision-making.13. Implementation of the system into the entire company.14. System maintenance (regular processing and reporting of the detected

information).

The entire quality costs monitoring system is useful only if the acquiredinformation is evaluated and used by the management to support the decision-making processes in the company.

The following text focuses on the system of quality costs monitoring in acompany that produces ceramic tiles. It presents the existing method used for themonitoring of quality costs and proposals of adjustments, so that the systemdepicts all major cost items related to the quality and represents an effectivemanagement tool. The article presents the pilot project of costs monitoring in oneof the company’s production plants. The attention focuses on the identification ofmajor quality costs, the proposal of their internal structure and collection of therequired data (see steps 4-7 of the implementation process).

Quality Costs Monitoring in the Company – Current State

LASSELSBERGER is a family company owned by Austrian group LASSELS-BERGER GmbH Pöchlarn, which does business in the production of ceramic tiles,as well as the mining and adaptation of raw materials and production of buildingmaterials. LASSELSBERGER is currently the largest manufacturer and supplierof ceramic tiles on the Czech market. Its products are produced and sold under twobusiness brands: RAKO and OBJECT objektová keramika. In 2008, the companysupplied over 27 mln m2 of ceramic tiles and pavings on all markets. Almost 11.4mln m2 of ceramic materials were supplied to the Czech market. Europe remainsthe traditional export market — Germany, Austria, France, as well as theNetherlands and Scandinavian countries. The company’s aim is to maximise theeffort at satisfying the growing customer requirements.

The quality management system in LASSELSBERGER complies with therequirements of the ČSN EN ISO 9001:2009 norm. The company has created,documents, applies and adheres to a quality management system and is continuallyimproving its effectiveness. The company has defined processes that secure theactivities of the quality management system, defined how they are applied, andintroduced their monitoring, if and where these processes may be measured andanalysed.


Systematic monitoring of quality costs is non-existent in the company. Onlythe costs of complaints proceedings are regularly monitored and reported in thecompany (they are included in the Quality Information Summary). The complaintsare sorted into complaints related to production flaws and complaints related toshipment, i.e. logistic complaints (shortages, product exchanges, break-ups etc.)The complaints are monitored acc. to period, plant, claim admittance, type ofdefect, country, warehouse (in the case of logistic complaints) and customers. Theinformation is quantified in the amount of defects, m2 as well as CZK. The reportis discussed on management meetings.

Pilot Project of Quality Costs Monitoring System (Selected PRODUCTIONPLANT)

One of the company’s production plants was selected for the pilot project, in linewith the progress of the implementation of the quality costs monitoring system(see above).

Characteristics of the Production Process, Quality Management and CostMonitoring System

In order to be able to process a proposal for the monitoring of quality costs in theselected part of the company, we had to become acquainted with the productionprocess in the plant, the methods used for quality management within productionand the existing production costs monitoring system.

The production process in the plant consists of the following steps (seeFig. 1):

1. Ensuring of input raw materials and their storage. 2. Wet milling in mills – homogenisation of raw materials. From the mills, the

homogenised liquid mass is drained into reservoirs, from which it istransported through pipes for further processing.

3. Drying in a spray drier. Granulate with water content of cca 5.5 % is created.This granulate is then taken from the drier into a reservoir.

4. Press moulding. The granulate is transformed into a tile. The semi-finishedproducts are transported via conveyors from the pressing machines for furtherprocessing.

5. Preparation of engobes and glazes. Engobes and glazes are readied in aseparate production step. The amount of the waste created within thepreparation of the glazes is not known and/or controlled. It is estimated thatthe waste amounts to ca. 9.5 % of the glazes and engobes put into the millsand subsequently 5 % of the suspensions produced in the glaze preparation


Fig.1 Production process scheme [9]

facility. 6. Glazing on glazing lines. Here the engobe, glaze and print are applied to the

tile semi-finished product. All in all, the waste rate for the pressing andglazing production steps is reported to be 2 % of the total production. Theproduction is monitored in the company in m2 and in tons.

7. Firing in gas furnaces. The firing process takes 40-50 minutes. 8. Product control and sorting. The products are divided into three categories,

which correspond to EN CSN 14411 (1st class, 2nd class and waste). The lossfrom the quality check amounts to 5 % of the total size of the surface of theproduction output.

9. Packing and subsequent dispatch to product warehouse.

The non-existence of waste in the classic sense of the word is a specificfeature of ceramic production. All non-quality (defective) products (semi-finishedor finished products) are recycled and put back into the first production phase.

The entire production process is monitored through 18 checkpoints(locations), where the product parameters are monitored and recorded and low-quality (defective) products are excluded. The nineteenth checkpoint (location) isthe company laboratory. The location of the checkpoints within the productionprocess and their functions are depicted in Fig. 2 and Tables I-III.

The monitoring of the flow of the material through the production iscurrently done through ERP (Enterprise Resource Planning) within the SAPsystem. According to the existing corporate management accounting system, theproduction is divided into three cost centres (see Fig. 3):

! Mass (raw materials) preparation – includes raw material warehouse, millingand drying;

! Glaze preparation and ! Production – includes pressing, glazing, firing, quality check and sorting and

packing.

1. Raw material warehouse

9. Packing 8. Quality check, sorting

7. Firing

2. Milling

5. Preparation of glazes

3. Drying 4. Press moulding

6. Glazing


Table I Checkpoints – input control [9]

Operation Checkpoint No. Value Performed by

Plastic rawmaterials

Raw materialswarehouse

1 Chemical analysis Companylaboratory

Moisture Companylaboratory

Firing shrinking

Absorbability

Non-plastic rawmaterials

Raw materialswarehouse

2 Chemical analysis Companylaboratory

pH Plant laboratory

Frita Glaze warehouse 3 Look Companylaboratory

Microscope

Alkali

Leaches

Colouring devices Colouring deviceswarehouse

4 Look Companylaboratory

Glaze Glaze warehouse 5 DilatationChemical analysis

Companylaboratory

Fig. 2 Checkpoints (locations) within the production process [9]

1. Raw material warehouse

9. Packing 8. Quality check, sorting

7. Firing

2.Milling

5. Preparation of glazes

3. Drying 4. Press moulding

6. Glazing

1 2

3 4 5

6 7 8 9 10

11

12

14

13

15 16 17 18


Table II Checkpoints – interoperational check [9]


Preparation ofmasses

Raw material shot 6 Weight Device operator

Calcitemanagement

7 Litre weight Device operator

Preparation ofmasses

8 Sieve residue Shift foreman,mate

Litre weight

Spray dryer (SD) 9 Chemical analysis Companylaboratory

Moisture / SD SD operator

Moisture / shot SD operator

Glaze preparation Raw material shot 10 Weight Mill operator

Glaze operatinglaboratory

11 Sieve residue Plant laboratory,operator

Litre weight Plant laboratory

Glazing

Litre weight Paste plantoperator

Flow Paste plantoperator

Sieve residue Paste plantoperator

Pressing plant 12 Strength control Device operator

Wedge-shapedness

Surface and edgecontrol

Dimensions

Rigidity Plant laboratory

Penetrometrics

Moisture ofpressing / driedout pressing

Dimensions


Table II – Continued


Glazing Glazing plant 13 Litre weight Glazing plantoperator

Weight of water

Weight of glazing

Weight of engobe

Flow

14 Layer

15 Visual control

Firing Oven 16 Dimensions Device operator

Curving

Look

Temperature curve Operator

Plant laboratory 17 Parameters Plant laboratory

Table IIICheckpoints – output control [9]


Palette inspection Sorting room 18 Complaints Output control

Inspection offeatures

Laboratory 19 Qualities acc. toEN 176, EN 159,PZN

Companylaboratory

Identification of Major Quality Costs and Proposal of Their Internal Structure

Given the nature of the production, the PAF model was selected for the monitoringof the costs; the costs are sorted in classification to prevention costs, control costs,costs of internal defects and costs of external defects. The content of individualquality costs categories (incl. the characteristics of activities falling withinindividual areas) is apparent from Table IV.

The system collects and registers data about individual cost categories,monitors the development of total quality costs, as well as individual cost items,in the monitored period, and compares the share of individual cost groups on totalcosts. The system provides information supporting decision-making, with the aimof improving production and other company processes.


Raw materials warehouse

Packing Check, sorting

Firing

Milling

Glaze preparation

Drying

Pressing Glazing

Cost centre PRODUCTION

Cost centre GLAZE PREPARATION

Cost centre RAW MATERIALS PREPARATION

Fig. 3 Cost centres within monitored plant [10]

Table IV Quality costs structure [9]

Prevention(prevention costs)

Control(control costs)

Internal defects(costs of internaldefects)

External defects(costs of externaldefects)

Quality informationsystem anddocumentationmaintenance

Input, intraoperationaland output control

Irreparable rejectsminus utilisable waste

Irreparable externalrejects

Quality managementsection

Laboratory tests Internal rejectsreparable (reworkingand adjustments)

External rejectsreparable (reworkingand adjustments)

Training andeducational programs

Metrology Deficits and damages Damage liability

Development of newcontrol and testmethods

Expert opinions Discounts on non-compliant products

Compliants solution

Procurement ofservices from externaltestrooms andlaboratories

Removal ofirreparable defectiveproducts

Minus acceptedcompensations fromemployees, suppliers,sales agents, insurers

Production of samplesfor destruction tests

Minus acceptedcompensations fromemployees, suppliers,sales agents, insurers


Collection of Required Data

The quality costs for 2008 and 2009 were calculated within the pilot project. Themain source of data was the existing management accounting system. The pilotproject used also information from intracompany quality reports (information oncomplaints) and data from production reports.

Notes on the determination of individual cost items:

! The costs of the quality information system and the maintenance of thedocumentation and the costs of the quality management section weredetermined by an expert estimate; this is a portion of the cots of the relevantcompany sections that carry out the monitored activities for all productionplants of the company.

! The costs of the quality management section include also the cots of thetraining and education programs and the costs of the development of newcontrol and test methods.

! The costs of the input, interoperational and output control are a part of thecosts of the Technologies centre. This centre carries out the technologicalaspects of the production preparation; it prepares, e.g., the technologicaldocuments (bills of material, work procedures) for the given production linesor the production processes, cooperates on the setting of the productionequipment, tests and evaluates the production samples, cooperates on theselection and assessment of input raw materials.

! The laboratory tests are carried out in laboratories which perform this activityfor all production plants in the company (they are a separate cost centre). Thecosts of the laboratory tests allocated to the monitored production plant weredetermined by estimation (acc. to the plant production/total companyproduction volume ratio).

! The costs of laboratory tests include also the costs of metrology, expertopinions and purchased external services (external test rooms andlaboratories).

! No products are produced specifically for destruction tests. Destruction tests,as well as other tests, are performed on finished products in the laboratories(the used amount is negligible). The losses on finished products due to testsare a part of the total losses of finished products and the cots of these lossesare part of the cost item Irreparable rejects (see costs of internal defects).

! Certain losses are incurred in each production step of the tile production. Fourtypes of defective products were identified within the entire productionprocess, all of them return back into production (as input raw materials); theflows are depicted in Fig. 4 with a dashed line. Flow 1 (between the glazingand pressing processes) consists of pressed, wet and unglazed defective tiles.Flow 2 consists of defective tiles with a glaze layer. Flows 1 and 2 are createdthrough the sorting of non-quality products on the belt, or they are tiles that


were used for the purpose of quality control. Flow 3 consists of finished firedtiles that do not comply with the rigidity parameters. Flow 4 consists ofcompleted fired tiles that were excluded due to low quality. The costs relatedto these flows are classified as the so-called costs of internal defects.

Fig. 4 Depiction of the creation of non-quality products for the purpose of calculationof the costs of internal defects [9]

! The calculation of the costs of internal defects must be based on the data fromthe cost centres that monitor the production process. The MFCA calculationcan be used with an advantage (see Refs [10,11]). The costs of internal defectsdo not include the costs of the raw materials used, as non-quality products arereturned back into the production process (as input raw materials).

! Deficits and damages are monitored directly in the accounting system.! 1st and 2nd class tiles, as well as non-quality products, are created within the

production of tiles (see above). As the plant’s aim is to produce 1st classproducts, the discounts on 2nd class products are included into the costs ofinternal defects (2nd class products are sold for half the price of 1st classproducts).

! The calculations of the costs of external defects (costs related to complaints)were taken over from the intracompany quality report.

Quality Costs Calculation

Within the pilot projects, the quality costs in the selected part of the company(selected production plant) for 2008 and 2009 were calculated using themethodology described above (see Table V).

1. Raw material warehous

9. Packing

8. Quality check,

7. Firing

2. Milling

5. Glaze preparation

3. Drying 4. Pressing

6. Glazing

1 2 3 4


Table V Quality costs in 2008 and 2009 (in thousands of CZK)

Cost item Year 2008 Year 2009

Prevention costs 3 091.4 2 294.8

Control costs 3 340.1 4 052.1

Costs of internal defects 76 636.5 57 342.9

Costs of external defects 556.3 522.9

Total quality costs 83 624.3 64 212.7

Production volume, mln m2 6.6 6.1

The total quality costs dropped by almost CZK 19.4 mln between the yearsof 2008 and 2009; the production in the same period decreased as well, though(from 6.6 million m2 to 6.1 million m2 — i.e. almost by 8 %).

Figure 5 depicts the structure of quality costs in the monitored period. It isapparent from the figure that the structure of the costs has changed. The share ofthe costs of internal defects has decreased, while the control costs on total qualitycosts has increased.

Fig. 5 Structure of quality costs in 2008 and 2009 (in %) [9]

Conclusion

The quality management system is an essential part of the management of everycompany. The article focused on the monitoring of quality costs, which forms an

Quality costs structure in %

3,7% 3,6%4,0% 6,3%

91,6% 89,3%

0,7% 0,8%

0,0%

10,0%

20,0%

30,0%

40,0%

50,0%

60,0%

70,0%

80,0%

90,0%

100,0%

2008 2009

Ext.defects

Int. defects

Assessment

Prevention


integral part of quality management and may represent a very effectivemanagement tool.

The quality costs monitoring system in the selected Lasselsberger plant hasthe aim of providing information about major cost items that the company’smanagement will be using to support its decision-making process. The proposalwithin the pilot project is based on the PAF model. Quality costs are sorted intoprevention costs, control costs and the costs of internal and external defects. Thebasic source of the relevant data is the management accounting system and theERP system. The pilot project involved the determination of the costs of qualityin the 2008-2009 period. The collected data clearly indicate that the mostsignificant cost item is the costs of internal defects, which account for ca. 90 % ofall quality costs. In order to be able to cut the costs, one must focus on individualproduction operations (input raw materials, production recipes, tuning and settingof production equipment). The lowest cost item is the cost of external defects.

The main advantage of the quality costs monitoring system is the fact thatthe level of quality of company activities, products and services (incl. qualitymanagement) is quantified in monetary units. As a rule, each operation that is notcarried out in appropriate quality leads to the creation of a non-quality product; thesystem quantifies the economic losses incurred by the company. This maycontribute to a change in the perception of the errors within the company processesby the company’s management, and primarily employees. Thanks to the obtainedinformation, the places where the biggest economic losses are incurred may beidentified, and, on this basis, rectifying measures may be proposed andimplemented – their efficiency may be defined very precisely. The aim of themeasure is to improve individual company processes, and thus also the company’seconomic results. The implemented measures are systemic and systematical.

For a successful implementation of the quality costs monitoring system, thesystem must be supported by the company’s top management, the implementationmust be handled by a professional interdisciplinary team, the system mustgradually be expanded to the entire company and, in many cases, adjustments tothe company’s information system are required. It should also be noted that thesystem in itself does not solve the problems with the quality, does not offerspecific solution and does not eliminate the company’s costs.

Acknowledgements

This work was supported by the Ministry of Education, Youth and Sports of theCzech Republic under project No. 1M0524 (Research Centre for Competitivenessof Czech Economy).


References

[1] Nenadál J.: Measurements in Quality Management Systems (in Czech), 2nd

ed., Management Press, Prague, 2004.[2] Wong L.: Quality cost. [online] [cited Feb 10 2010] Available from internet

URL: http://ictlab.tyict.vtc.edu.hk//~tsangkt/PQM/QM/qm\_ppt/3\_Quality\%20 cost.ppt.

[3] Crowley D.: Cost of Quality. [online] [cited Feb 06 2010] Available frominternet URL: http://www.sasqag.org/pastmeetings/CostOfQuality.ppt.

[4] Atkinson H., Hamburg J., Ittner, Ch.: Linking Quality to Profits, New Jersey,Institute of Management Accountants, 1994.

[5] Dale B.G., Plunkett J.J.: Quality Costing. 2nd ed., Chapman & Hall, London,1995.

[6] Harrington H.J.: Poor-Quality Costs. 2nd ed., DEKKER, New York, 1987.[7] Schiffauerova A., Thomson V.: Int. J. Qual. Reliab. Manage. 23, 2006.[8] Kohl M., Miller I.: Quality Standards in Practice (in Czech), Modern

Management, Prague, 2007.[9] Tesnerová P.: Quality Costs in Lasselsberger Ltd. Plant (in Czech), Diploma

thesis, Prague Institute of Chemical Technology, Prague, 2010. [10] Hyršlová J., Vágner M., Palásek J., Bednaříková M.: Sci. Pap. Univ.

Pardubice, Ser. A, 15, 199 (2009).[11] Palásek J.: Application of Material Flow Cost Accounting in a Plant (in

Czech), Diploma thesis, Prague Institute of Chemical Technology, Prague,2009.


219



16 (2010)

TIME SERIES FORECASTING AS A TOOLOF OPERATIVE MANAGEMENT

Michal PATÁK and Vladimíra VLČKOVÁ1

Department of Economy and Management of Chemical and Food Industry,The University of Pardubice, CZ–532 10 Pardubice


Present quickly varying business environment requires changes in traditionallogistic management of companies of chemical and food industry operating onB2C markets. Usage of recent logistic technologies however is made impossibledue to the position of Czech companies on the markets mentioned. This articlepresents the problems which these companies solve on the operative level, andpoints out possibilities of time series analysis utilization as a tool of demandplanning. Possible application of this tool in operative management and itsincidences are described on a chosen company of food industry.

Introduction

Global markets on which companies operate nowadays are mainly characterizedby quick and hardly foreseeable changes, by supply overpressure (overhang) abovedemand, by rapid development of information and communication technology and

220 Paták M., Vlčková V./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 219–228

by well-informed and demanding customers. As a consequence, there is strongorientation on customer and development of systems of differentiated CRM basedon customer value [1]. This is the reason why maximum satisfaction of evenindividual customer requirements mainly in the field of individually providedlogistic services are main aims for companies. Time plays important, if not keyrole in it. This means for companies that they must be able of quick and adequatereaction on customer’s requirements.

This can be assured only on condition of high flexibility and effectivenessnot only of business processes but even of the whole supply chains, of which thecompany is a link. It results in the development of Supply chain management andmany logistic technologies. Ability of all supply chain links to share neededinformation and knowledge is the basic condition for implementation of thesetechnologies. Primarily, it is sharing of joint demand forecast and its unifiedutilization in planning process of all links of chain — demand planning. It requiresnot only technical and software securing of these information transfers, which areoften shared on-line, but also confidence between individual links and theirwillingness to cooperate and to share information. Under these conditions methodsas JIT (Just in Time), QR (Quick Response), ECR (Efficient Consumer Response),CRP (Continuous Replenishment Planning), RMR (Retail ManagementReplenishment), VMI (Vendor Managed Inventory), CFaR (CollaborativeForecasting and Replenishment), CPFR (Collaborative, Planning, Forecasting andReplenishment) can be successfully implemented. Many of companies in theCzech Republic are a part of supply chains, in which there is for many reasons nowillingness to cooperate in the field of information sharing about demands andsales [2]. Those are namely manufacturing companies, which supply their productsinto foreign retail chains, where also companies of chemical and food industrybelong. Sales forecast derived on the basis of analysis of sales time series is oneof the tools which can help in solution of these problems.

Time Series Forecasting and Application in Operative Management

Time series forecasting is based on analysis of data recorded over a period of time,discovering of the pattern in the historical data, and extrapolate that pattern intothe future. The business series follow various patterns. Study of the types of datapatterns is an important step in selecting an acceptable time series model.Operative management is working with sales data rarely older than several months.That collection of data forms short-term time series with horizontal pattern. Ahorizontal pattern presents data values fluctuating around a constant mean. Arandom fluctuation around constant value is typical for stable processes withstochastic defects, products with stable sales or most of the short-term time seriesof sales. Thus these time series often miss secular trend, seasonal and cyclical

Paták M., Vlčková V./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 219–228 221

variation. There are two basic techniques used in time series forecasting withhorizontal data pattern in the literature (see, e.g., Refs [3-6]):

! Simple Moving Average,! Single Exponential Smoothing.

The moving-average method can be useful in removing the randomfluctuations for forecasting. Although moving averages are centered, it is moreconvenient to use them to predict the following period of time series. Then theformula for a simple moving average is

(1)

where Ft is forecast for the coming period, n is number of period to be averaged,At–i is actual sales i-periods ago.

Exponential smoothing is the most used of all forecasting techniques. Theformula for single exponential smoothing results from modified Eq. (1) and ismathematically represented as follows

(2)

where " is smoothing constant which determines the level of smoothing and thespeed of reactions to differences between forecasts and actual occurrences.

The value of sales forecast is the most important information in theoperative management. It is especially used in production planning and inventorycontrol. The forecast accuracy can facilitate operative decision making. Allforecasts contain some error because of the interaction of many indefinable factorsin the model. Forecast error is definition for differences between the forecast valueand what actually occurred. The common terms used to describe the degree oferror are mean absolute deviation (MAD) and mean absolute percentage error(MAPE) defined as follows

(3)

(4)

where t is period number, A is actual sales for the period, F is forecast sales forthe period, and n is total number of periods. When the errors that occur in theforecast are normally distributed (the usual case), the mean absolute deviationrelates to the standard deviation as [3]


(5)

where F is standard deviation. Knowledge of standard deviation determinesutilization of historical data to quantify the volume of the safety stock [4].

Characteristics of Companies

Manufacturing companies of food industry equally as the majority of companiesof chemical industry producing consumer goods get significant part of produce tofinal customers through retailers. The present era is characterised by growingbargaining strength of retail chains mainly as consequence of the fact thatmanufacturers’ offer significantly exceeds demands of end users. If manufacturerswant to hold a market, they must be able to realize in a very short time period veryquickly changing requirements of retailers and to agree to the disadvantageousterms of supplier-customer relations. Retailers are not willing to share informationabout consumer demand or behaviour. Though, the customer service lead time hasto be a far smaller interval than the lead time required by the company to produceor distribute the product. It reflects on increasing requirements of operativemanagement of manufacturing, purchase, but also on other logistic activities alongthe whole supply chain. Thus, demand planning is one of the few instruments howto control processes in this case. Possibilities of forecasting exploitation will bedemonstrated on a specific company, which cannot be disclosed due to the datasensitivity [7].

The company chosen for analysis is a dairy with long tradition of producingdairy products. It occupies strong position on Czech market above all in thesegment of butter-type spreads, cream spreads, cream yogurts and cottage cheeses.Considering existence of a great number of tastes, range of products of thecompany contains about thirty various kinds of products. Portfolio of productsdiffers also by the size of consumer package (several consumption packages,gastro production). The company is selling these products also under a few privatetrade marks of Czech and foreign retail chains in comparable quality, but indifferent consumer packages. The company gets more than two thirds of produceto final customers through retailers. The company’s position in the marketplaceanalysis revealed all problems stated in the previous paragraph. The companyposition is also complicated by short-term usable life of the products (several daysor weeks). Although key raw material for all products manufacturing is creamwhich is purchased from farmers with one week lead and average production cycleis three days, the customer service lead time is usually one day. In addition,customers accept only delivery of products with full usable life or only with its


partial expiration. Key customers make use of fixed-time period model orderingand withdraw products several times on the basis of the order from previous day.

The fluency of material flows in the company is ensured by managers withlong-time experiences. However, there is a little correspondence betweenoperational plan and operative management. Even managers find some steps ofoperative management uneconomical. For this reason, operative planning andcontrol was analyzed in the company. Following processes which influenceprocesses effectiveness mostly were identified by the analysis:

! weekly demand planning,! weekly purchase of cream,! weekly production scheduling,! safety stock of final production assessment.

Weekly forward sales are estimated in the company on the basis of executedsales over the last period. This estimation has low accuracy but it presents basicinformation for purchase of cream which has to be ordered a week beforeconsumption. The production is daily controlled by executed sales currently.Production scheduling is complicated because of unpredicted fluctuations in salesand stint of purchased cream. Safety stock of final production is not quantified butaverage reserve/stock in store of finished goods corresponds to roughly averageweek sale. Thus it is possible to assume, that safety stocks make about one half ofaverage week sales.

It is obvious that this kind of management based on innocence of accuracyand reliability exactly ascertained forecast, requires big amount of operativeinterferences. This management is in addition related to skills and experiences ofconcrete people, with their departure this non systemic management is not in thelong term maintainable. Time series analysis could be one of the few instrumentshow to control processes at the operative level.

Results of Time Series Analysis and Discussion

Time series analysis was focused on the possibility to predict weekly salesproviding primary information for operative control. Sales of model product whosetime series were not misrepresented by promotion were used to research. Timeseries of sales executed by three characteristic retailers (retailer A, B and C) wereanalysed during forty weeks in 2008. Retailer A and B represents foreign retailers(retailer A is the key customer of company), and retailer C is a representative ofCzech retailers. Time series were smoothed by moving-average method and singleexponential smoothing. Forecast error was described by MAPE. In view of theproduction cycle and purchase of cream it was necessary to modify both offorecasting equations as follows


0

5

10

15

20

25

30

0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0

Smoothing Constant ��

MA

PE

[%]

Retailer A

Retailer B

Retailer C

(6)

(7)

that returns forecasts a week before models (1) and (2). Those sales forecasts arereally usable in operative management in this case.

It is important to select the best period n for the moving average (6). Thegreater random fluctuations are smoothed by the longer moving-average period.But there is the need to have a large amount of historical data and there is aproblem that medium-term changes in possible trend can be lost. Number ofperiods was chosen from 2 to 9 and the influence on forecast accuracy wasmonitored.

In the single exponential smoothing (7), alpha is given a value between 0and 1. The low value of alpha keeps information from data many periods ago,whereas the higher value prefers the actual demand to historical data. Adjustingthe value of the constant can also help with reaction to changes in possible trend.Therefore, the smoothing constant was optimized by minimisation of MAPEbecause the choice of alpha influences the forecast accuracy extremely. The valuesof MAPE depending on choice of smoothing constant are shown in Fig. 1.

Fig. 1 Smoothing Constant-MAPE Diagram

Comparison of the MAPE values obtained by application of forecasttechniques and by predicting in the company is shown in Table I. The forecasterror was the lowest in the single exponential smoothing with optimal smoothingconstant ("OPT). Time series analysis based on this technique is shown in Fig. 2.

Utilization of exponential smoothing would reduce forecast error by morethan 7 % in sales of the key customer. Low values of optimal smoothing constants


0 5 10 15 20 25 30 35 40

Time (Weeks)

Sal

es

Sales

Forecast (Single Exponential Smoothing)

Retailer A

Retailer B

Retailer C

as well as descending forecast error with simultaneous ascending number of period

Table I Forecast Techniques Comparison by Forecast Accuracy (MAPE)

Retailer A Retailer B Retailer C

MAPE of company prediction, % 19.5 11.2 25.2

"OPT 0.062 0.007 0.090

MAPE of single exponential smoothing, % 12.1 7.2 20.1

MAPE of single movingaverage, %

n = 2 16.5 9.6 23.9

n = 3 15.4 8.7 22.1

n = 4 15.1 7.6 21.8

n = 5 14.4 7.2 20.4

n = 6 14.3 7.6 21.0

n = 7 14.3 7.5 21.0

n = 8 13.7 7.3 20.6

n = 9 13.7 7.5 20.5

Fig. 2 Utilization of Single Exponential Smoothing to Time Series Analysis

in moving-average methods point out that data values really fluctuate around aconstant indeed. This information can be decisive for capacity planning and fixedproduction scheduling without necessity of bigger operations intervention.Constant mean of weekly sales determines weekly production and weeklypurchase of cream. Random fluctuations of sales values could be covered by safety


stock of final production.High variance of sales data of key customer (retailer A) is decisive for safety

stocks quantity because sales variance of the others is insignificant in absolutepoint of view (see Fig. 2). Safety stocks value based on time series analysisamounts 50 % of average weekly sales. Carrying of high inventory would notreduce costs. But the company can profit through integration of processes viaunified joint forecasts within the whole company. High level of safety stocks offinal product would have other negative effects. In regard of short-term usable lifeof some products, part of the produce would be sometimes damaged or cut under.In the case of repeated exhaustions of safety stocks it could happen that productioncapacity could be not sufficient for their replenishment, and in some cases,significant part of week capacity would have to be hold for the case of excessiveexhaustion of stocks. The degree of impact of these implications, however, can befind out only after full analysis of time series of all sales realized by the company.Such analysis has not been carried out yet, because the company does not disposewith database of historical sales enabling bulk data processing for all items ofselling assortment.

At present, the company deals with implementation of suitable informationtechnologies. They would enable forecasting based on time series analysis and itsexploitation in another company functions and namely particularly in planning andmanagement of production/manufacturing. Integration of company functions,roofed-over by in-house information system, represents one of the fundamentalconditions not only of effective exploitation of forecasting in operationsmanagement, but also of the whole intellectual conception of demand planning [8].

Conclusion

Orientation of companies on customers requires changes in management not onlyof individual companies, but of the whole supplier — customer chains. Situationon markets with fast turning consumer goods though inhibits logistic technologies’implementation. This is the reason of very frequent undesirable operativeinterference into production plans or of disproportionate increase of finishedproducts' reserves. Operative management within the company can be streamlinedthrough integration of business process via sharing of joint forecast in all businessfunctions.

It has been shown on the above mentioned company, that basic techniquesof time series analysis can be successfully applied for needs of operativemanagement. Prediction models can be surely modified for particular needs ofeach company, whereas gained forecasts can result in significant specification ofsale judgment. In the case of stated example the forecast at the biggest customerwas improved from 19.5 % on 12.1 %. Achieved error of forecast cannot be


removed through analysis of this type time series, because it is caused byunpredictable behaviour of customers. By analysing time series it has been foundout that values of former sales fluctuate around a constant average value, whichis not changed significantly from the long-term point of view. It allowsimplementation of integrated control, because of which significant decrease ofinterferences in operations management should take place.

In spite of the fact, that interconnection of demand forecasting process withall planning activities within the company can become competitive advantage, thisinterconnection evokes frustration of majority workers. One from many otherreasons is the fact that demand forecast is based on probability and from itsprinciple it can never be taken as fully reliable. Potentiality of forecastexploitation, however, does not depend on the level of its reliability only. Eachassessed forecast is an effective tool for decision making, because each decisioncomes out from certain judgment of the future. Additional reasons whyquantitative methods generally are not applied in practice are apprehensions fromnecessary changes in management [9]. These changes are, in addition, oftenconnected with investments mainly into information technology. But if companieswant to keep their position within supply chain, they cannot avoid changes inmanagement.

References

[1] Lošťáková, H.: Principle of Differentiated CRM Strategy According toCustomer Value to Company (in Czech), In Lošťáková H., Branská L.,Dědková J., Gros I., Grosová S., Honzáková I., Jelínková M., Pecinová Z.,Simová J., Vávra J., Vlčková V.: Differentiated CRM Strategy According toCustomer Value for Company, pp 77-86, The University of Pardubice, 2006.

[2] Vlčková V.: Vedecké listy 6, 122 (2007).[3] Chase R. B., Aquilano N. J.: Production and Operations Management, 7th

edition, McGraw-Hill Book Company, New York, 1995.[4] Gros, I.: Mathematical Models for Managerial Decision Making (in Czech),

1st edition, Institute of Chemical Technology, Prague, 2009.[5] Hindls R., Hronová S., Seger J.: Statistics for Economists (in Czech), 8th

edition, Professional Publishing, Prague, 2007.[6] Levine D. M., Stephan D., Krehbiel T. C., Berenson M. L.: Statistics for

Managers Using Microsoft Excel, 4th edition, Prentice Hall, New Jersey,2005.

[7] Paták M.: Analysis of Demand Planning in Chosen Company (in Czech),Diploma Thesis, The University of Pardubice, Pardubice, 2009.

[8] Vlčková V., Paták M.: Role of Demand Planning in Business ProcessManagement, In Proceedings of the 6th International Scientific Conference


“Business and Management 2010“, Vilnius, 2010.[9] Vlčková V., Machač O., Munzarová S.: Scope and limitations of quantitative

methods in supply chain management, In Proceedings of the 6th InternationalConference “Financial and Logistic Management 2009“, Malenovice, 2009.

229



16 (2010)

TRIPLE HELIX MODELIN THE CZECH TERTIARY EDUCATION

Veronika SABOLOVÁDepartment of Economy and Management of Chemical and Food Industry,



The article is devoted to the issue of triple helix model and its application in theCzech tertiary education. Introduction describes the crucial provisions of theWhite Paper on Tertiary Education. Then the triple helix model is described basedon the foreign literature search. Particular attention is focused on the importanceof the triple helix model in the White Paper on Tertiary Education. Finally,possible subjects of the cooperation between university, industry and governmentare indicated.

Introduction

The Czech tertiary education is being reformed at present. Some parts of reform(especially those essential) are already approved and others will follow. Thisreform of tertiary education is based on the concept outlined in the White Paperon Tertiary Education. This book, among other things, outlines the future need foruniversities to work more closely with industry (application practice) under theauspices of the government. Scientific literature refers to the cooperation

230 Sabolová V./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 229–236

(interconnection) of universities, government (public administration) and industry(enterprises) as to the “triple helix” which is considered an evolutionary model ofinnovations [1,2]. The merit of this three-entity cooperation model, which topopposing vertices, is relatively simple, nevertheless efficient. The role ofuniversities rests in provision of scientific, expert and partly also technical andmaterial capital. This model counts on industry to be a main guarantor of financialcapital and, in a way, imaginary university customer. The government thenfacilitates necessary conditions (particularly legislative) for this kind ofcooperation (or eliminates obstacles in cooperation between universities andcompanies).

White Paper on Tertiary Education

The White Paper on Tertiary Education is a conceptual and strategic documentsetting the direction of the tertiary education in the Czech Republic for the horizonof 15-20 years. This White Paper, however, is not a technical manual for executionof the changes leading towards the set aim [3]. It serves merely as a concept for thereform of the tertiary education in the Czech Republic. The reform spirit shouldbe in compliance with the philosophy of human resources being one of the mainpillars of competitiveness in the current development of the Czech society. Themain aim of the reform will rest in change of management and funding principlesof the tertiary education system. Figure 1 describes fundamental principles of thetertiary education reform.

Fig. 1 Fundamental principles of the tertiary education [4]

GREATER OPENNESS

COMPETITIVENESS

DY

NAM

ICS

JUS

TIC

E

FUNDING

I NST

ITUI

ON

S

STU

DE

NT

DIVERSIFICATION

R ESE

AR

CH

AND

DE

VE

LO

PME

NT

MA

NA

GE

ME

NT

Sabolová V./Sci. Pap. Univ. Pardubice Ser. A 16 (2010) 229–236 231

According to the White Paper the universities are supposed (especially) toengage stakeholders in their own activities and management, and thus shouldensure a more effective feedback and support elements of management control.Meeting this target means the enhancement of accountability and efficiencyexercised towards external environment. Stakeholders of the university are eitherthose who contribute to the university operation or demand the universityoutcome. The university operation is indirectly supported by every taxpayer, i.e.the public. Other stakeholders are the companies, first, from the position ofemployers demanding university graduates, second, from the position of researchand development clients. Last but not least, the stakeholders include studentsdemanding education as the essential university outcome. In the case of aparticular university stakeholders involve also representatives of regionalgovernments and enterprisers. This group, however, may be extended withrepresentatives of research and cultural institutions, non-profit sector and,naturally, alumni. The White Paper on Tertiary Education sets the partial target interms of creating more favourable conditions for cooperation with thesestakeholders by deeper engagement of both sides in particular.

The White Paper suggests the engagement of stakeholders into strategicdecision-making of universities should be reached by extending the operation ofthe board of trustees. The White Paper authors’ intention was to make theuniversities more open to their external environment from both the institutionaland strategic points of view. The stakeholders’ engagement into universityoperation and management should not be limited to the decision-making power butthey should also bring private funding into the sector which is traditionallyfinanced from public money. Meeting this aim will mean enhancement ofmultiple-source funding university operations. The core idea rests in the conceptwhere the power to express oneself freely in terms of university problem issues isonly fair if it is counterbalanced by their financial involvement in tertiaryeducation.

Triple Helix Model

In the triple helix development model, government devolves decision making tocollaborations with regional and local authorities and other actors. Industryengages in endogenous innovation as well as transfer. Universities play aninnovative role in society, active in translational research, entrepreneurial trainingand community development, as well as, traditional tasks. These nascenttransformations have fundamentally changed the development landscape, makingtriple helix actors the central development partners [5].

The triple helix is based on the premise that the university plays anenhanced role in development in concert with government and industry, the two


traditional leading institutional spheres. Higher education institutions are virtuallyeverywhere and their flexible nature opens them to fill a variety of roles, wellbeyond traditional missions. Traditional missions of teaching and research embeda knowledge transfer capability in any society. In the training of human capital forall sectors of society, the university, through its alumni, provides the basis forenhanced interaction. The prominent role of the university in the triple helix hasmade this model especially relevant to developing countries where universities arepresent and industry is either making strides, relatively weak or largely lacking [5].Critics have argued that the university systems in most developing countries areacademically oriented and industries are either non-existent or too weak andgovernments too bureaucratic to play respective roles envisaged by the triple helixmodel. However, the problem as noted by Konde [6] does not lie with the model,but the fact that, in many countries, the triple helix entities seem to be weakbecause their elements tend to work in isolation [5].

The triple helix model comprises three basic elements [7]:

1. a more prominent role for the university in innovation, on a par with industryand government in a knowledge-based society;

2. a movement towards collaborative relationships among the three majorinstitutional spheres in which innovation policy is increasingly an outcome ofinteraction rather than a prescription from government;

3. in addition to fulfilling their traditional functions, each institutional sphere“takes the role of the other” in some regards (institutional spheres overlap andcollaborate and cooperate with each other, as you can see in Fig. 2 [8]). Thismay take the form of a university taking government’s role of initiatingdevelopment projects or industry’s role of firm formation. Universities,traditional providers of human resources and knowledge, are now criticalsocio-economic development actors. The institutional spheres still performtheir traditional functions but increasingly assume the task of advancinginnovation and development.

Fig. 2 The triple helix model [8]

STATE

IND

USTR

Y A

CAD

EMIA


Many universities have expanded their organizational capabilities to engagein knowledge transfer and development. In addition, universities are alsoextending their teaching capabilities from educating individuals to shapingorganizations by using the incubators. The incubator was essentially a means totrain a group of individuals to operate as an organization. The incubator modelwas extended from an earlier emphasis on forming high-tech firms to creating low-tech firms as well as cooperatives that make it possible for excluded populationsto collectively enter the labour market as service providers contracting with publicand private sector organizations for cleaning and other tasks [9].

A triple helix development model is based on the following trends [5]:

1. The transition from an industrial society to a knowledge-based society inwhich knowledge producing institutions, like universities, potentially play agreater role in innovation and development.

2. The supersession of large scale physical technologies that mandatebureaucratic forms of organisation to increasingly flexible smaller scale hightechnologies that can be utilized by smaller scale organizations.

3. The emergence of polyvalent knowledge, in such areas as biotechnology,computer science and nanotechnology, that is at one and the same timetheoretical and practical; capitalizable and publishable.

4. The rise of new university formats that incorporate a classic ivory tower focuson discipline development with a culture of entrepreneurship, innovation andtechnology transfer.

A triple helix development model contrasts with others that place greateremphasis on “state-led, market-led or community-led development” [10].Although this is laudable, it omits a critical agent of knowledge-baseddevelopment – a university that is capable of undertaking socio-economicdevelopment initiatives in cooperation with teaching and research. The triple helixdevelopment model focuses on creating intermediary mechanisms that play abroader role than in developed environments. They not only fill the gaps betweenindustry and university and between discovery and application but also in someinstances they substitute for weak or missing actors. Moreover, in the triple helixdevelopment mode there is strong emphasis on interactions, linkages andcollaborations [5].

This new way of development thinking, revolving around the crucialknowledge actors, strengthens diversity and represents a radical departure from theconventional development models that have separated the three institutionalspheres, most often placing universities in a peripheral role in developmentstrategies and policies. Thus, the triple helix refocuses the development fieldmaking it possible for technology transfer to play a residual role in support of thedevelopment of indigenous technological capability [11].


Triple Helix Model in the White Paper on Tertiary Education

All three participants of the triple helix model are defined in the White Paper onTertiary Education. In this document, their cooperation is seen as crucial for apositive outlook on the future of tertiary education system in the Czechia andwhole Czech economy as a knowledge-based society. Industry and government(whether at regional or national level) are in the White Paper described asstakeholders of the university, with which in the future should forge strongerrelationships than ever before. Proposal to tertiary education reform encouragesthe use of the triple helix model to improve the quality of tertiary education systemand the human capital that is crucial for economic development.

Tertiary education reform tries in the field of cooperation university,industry and government to resolve the current situation outlined in Fig. 3. Theoriginal image of triple helix model represents an ideal situation where the contactsurfaces (i.e. notional amount of cooperation) between particular elements are thesame size and occupy a large enough area for effective cooperation. Nevertheless,in the following picture the sphere of industry is shifted to level, where itscooperation with the university as well as with the government represents a muchsmaller area. The reason of this situation is unattractive offers on cooperation fromstate and universities. In many cases it is only a request for funding that is notbalanced with opportunity to engage in creative activities and decision-making.White Paper tries to change this unpleasant fact through the tertiary educationreform. It tries to involve industry to decision-making in the board of trustees ofuniversity and to the funding of research, development and also education. For thiscooperation it is necessary to create a favorable environment in whole economyand this must the government ensure.

Fig. 3 The triple helix model in the Czech Republic

STATE

ACA

DEMIA

INDUSTRY


The role of individual elements of the triple helix model will depend on thespecific subject of their cooperation. However, the government will have to secure(irrespective of subject cooperation) not only a favorable environment, butespecially right direction of joint action of the university-industry-governmentcooperation. The subject of cooperation between university and industry mayrelate either education or research and development.

In the area of education the university should cooperate with industrialenterprises when preparing curriculum. Not only the list of study subjects withinthe study field but also the contents of these subjects should reflect therequirements of practice. Students can gain the awareness of practice requirementsduring their working practice, or as the case may be, during their study visitorganised by the partner industrial entity in its facilities. Students may also learnat field trips in the company. Based on their experience acquired in theenvironment of an industrial company students may produce their theses. Theresults contained in the theses may serve as grounds for further research at theuniversity or be used directly by the partner industrial entity for its own benefit.The industrial company may also organise both for the students and universitystaff training sessions in various areas, where indeed, also the company staffattend. On the other hand the university may also train the company staff or createindividual study plans for their staff if they are interested in studying an accreditedsubject field. Both the partners may hold conferences where both a scientific(theoretical) and application (practical) points of view are presented. They mightalso utilize their infrastructures, e.g. lecture halls, accommodation, catering,leisure time facilities, special laboratories or computers. The main benefit of theuniversity and industrial company interconnection would, indeed, rest in creatingjob opportunities for university graduates in the industrial company.

In research and development the university and industrial entityinterconnection might mean, e.g., joint projects. The results of projects could bethen theoretically modelled in academic environment and consequently tested inpractice of the industrial company. The university could also provide variousstudies, reviews or expertise reports. The last but not least comes the universityexpert consulting provided to the company.

Conclusion

The general idea — the university would in future have to work more closely withindustry (application practice) under the auspices of the government — is logicalat first glance. Its implementation, however, is not so simple. It depends on way,how it creates the connection and on what principles will be basing its operation.Triple helix model is recognized by the scientific community an innovativeconcept, through which can be achieved the development of not only the education


system, but also a society which has undergone a transformation to a knowledge-based economy. If reform of tertiary education is really done in the spirit of triplehelix model, the concept could become for the Czech Republic the way from theglobal economic crisis (some economists contend that the crisis is only theepiphenomenon ongoing transformation of the economy).

References

[5] Dzisah J., Etzkowitz H.: Triple helix circulation: the heart of innovation anddevelopment. Int. J. Technol. Manage. Sust. Dev. 7, 101 (2008).

[7] Etzkowitz H.: The Triple Helix: University-Industry-Government Innovationin Action, Routledge, New York, 2008.

[8] Etzkowitz H.: The Triple Helix of University-Industry-Government:Implications for Policy and Evaluation [online], [cit. 2010-09-01], availableat URL: <http://www.sister.nu/pdf/wp_11.pdf>.

[9] Etzkowitz H., Carvalho de Mello J.M., Almeida M.: Res. Policy 34, 411(2005).

[1] Etzkowitz H., Leydesdorff L.: Research Policy 29, 109 (2000).[6] Konde V.: Int. J. Technol. Manage. 27, 440 (2004).[10] Kothari U., Minogue M.: Development Theory and Practice: Critical

Perspectives, Palgrave, Houndmills, 2002.[2] Leydesdorff L.: Research Policy 29, 243 (2000).[3] Matějů P., Ježek F., Münich D., Slovák J., Straková J., Václavík D.,

Weidnerová S., Zrzavý J.: White paper on tertiary education [online], [cit.2010-09-01], available at URL: <http://www.msmt.cz/uploads/bila_kniha/schvalena_bktv/White_Paper_on_Tertiary_Education_fin.pdf>.

[4] Matějů P., Václavík D., Münich D.: White paper on tertiary education:expected role of post-secondary professional education (in Czech) [online],[cit. 2010-09-01], available at URL: <http://www.asociacevos.cz/zapisy/BK_koncept_VOS.ppt>.

[11] Saad M., Zawdie G.: Technol. Anal. Strategic Manage. 17, 89 (2005).

ISBN 978 - 80 - 7395 - 386 - 7 (Online)ISBN 978 - 80 - 7395 - 368 - 3 (Print) ISSN 1211-5541

Documents

Scientific Papers 16 2010 University ědeckých prací ... · solid and high sugar content food have been reported [16-20]. Comfrey (Symphytum officinale L.) belongs to the borage