Investigating the effect of pH on the surfacechemistry of an amino silane treated nano silica

Mehran Rostami and Mohsen Mohseni

Department of Polymer Engineering and Colour Technology, Amirkabir University of Technology, Tehran, Iran, and

Zahra RanjbarDepartment of Surface Coatings and corrosion, Institute for Colorants, Paint and Coatings (ICPC), Tehran, Iran

AbstractPurpose – The purpose of this paper is to investigate the effect of different PHS on the surface chemistry of fumed silica treated withaminopropyltrimethoxysilane (APTMS).Design/methodology/approach – The reaction conditions involved variation of pH ranging from acidic to alkaline. Different analytical techniquesincluding FT-IR spectroscopy, thermogravimetric analysis (TGA), CHN and Zeta potential analyses were employed to investigate the surface chemistry oftreated particles. In addition, the stability of silanised silica dispersions were studied using turbidimetric and rheometric measurements.Findings – It was revealed that in all conditions silica was more or less chemically grafted by the silane. When the pH of treating bath was adjusted to1-2 prior and during the reaction, 58 percent grafting was observed, as obtained by CHN and TGA analyses. At very alkaline conditions, however, thegrafting content declined to 29 percent. The variations in grafting were dependent on the silane hydrolysis and its further condensation with the silicasurface. Zeta potential measurements showed a drastic change from 27.1 mv to þ18.01 mv (at pH 7) for the untreated particle and the one with thehighest grafting, respectively. The dispersion stability of differently treated particles varied in solvents with different Hansen solubility parameters (HSP).Moreover, due to the variations of surface chemistry of particles, their rheological behaviours were significantly influenced.Originality/value – The results obtained in this work showed that the surface chemistry of fume silica could be tuned with treating method.The highest content of grafting led to a better dispersion in solvents having greater hydrogen bonding component and to an inferior dispersion insolvents with higher polar component.

Keywords Surface treatment, Surface chemistry, Rheology, Silica, Wettability

Paper type Research paper

Introduction

Fumed silicas, due to their very fine particle size and high

specific surface area, have shown to be one of the most

important reinforcing fillers for polymers and coatings (Robertsand Bergna, 2006; Sun et al., 2005). Apart from the geometry,

surface chemistry of silica also greatly affects the degree ofreinforcement. As silica contains a large number of silanol

groups on its surface in the formof vicinal, geminal and isolated,

it can be considered highly polar (Chen et al., 2005). Therefore,it is less compatible in non-polar media (Jesionowski and

Kkrysztafkiewicz, 2001). Moreover, the surface silanol groups

have a great affinity to form hydrogen bonding with each other,resulting in a strong filler-filler interaction (Chen et al., 2005;Jesionowski and Kkrysztafkiewicz, 2001; Jiang et al., 2007).This can be prevented by the aid of coupling agents, thereby

modifying the hydrophilic nature of silica surface. Such a

surface modification not only improves the wettability of silicain organic media, but also functionalises the particle to interact

chemically with coating media. The formation of chemicalbonds between the inorganic and organic is expected to

guarantee a durable interaction between the two incompatible

phases. For themodification of silica surfaces, organosilanes are

most commonly used (Chen et al., 2005; Sun et al., 2005).Recently, much attention has been given to the use of these

materials which are able, to some extent, to couple filler to an

organic media (Jesionowski and Kkrysztafkiewicz, 2001;

Jiang et al., 2007; Iijima et al., 2007; Wieczorek et al., 2004).

Generally, silane coupling agents may possess two functional

active groups, i.e. an alkoxy group capable of reacting with

surface silanol, and an organic functionality, generally having

amino, epoxy and acrylic groups in its structure. These may

participate in the curing system leading to a possible strong

linkage between the silica and the polymer (Jiang et al., 2007).Consequently, a silane coupling agent can function as a bridge

between silica and polymer, thereby enhancing interaction

between these two. Many types of silane coupling agents,

varying in chemistry and reactivity are available, ofwhich amino

silanes are among those usually used for silica modification

(Chen et al., 2007; Jesionowski and Kkrysztafkiewicz, 2001;

Wieczorek et al., 2004).The action of alkoxysilanes starts with hydrolysis. The rate

of hydrolysis significantly depends on the pH as well as on the

type of organo- and silicon-functional groups (Roberts and

Bergna, 2006; Xhanthos, 2005). Different factors can affect

the reaction of silane with inorganic particles, among which,

the natural pH of the silane, isoelectric point of the particle,

ratio of water to silane as well as the pH of the reaction media,

and the method of dispersion can be considered as the more

important ones (Xhanthos, 2005). At high and very low pH

values, the rate of hydrolysis is higher than that of the silane

natural pH, at which silanes are most stable. For example,

The current issue and full text archive of this journal is available at

www.emeraldinsight.com/0369-9420.htm

Pigment & Resin Technology

40/6 (2011) 363–373

q Emerald Group Publishing Limited [ISSN 0369-9420]

[DOI 10.1108/03699421111180509]

363

the rate of reaction of a monomeric trialkoxysilane in acetic

acid solution increases by a factor of ten upon reducing the

pH value from 4 to 3.The overall purpose of surface treatment of silica is to

render the surface functionalised and hydrophobic comparedwith the pristine particle. There are various reports on the

surface treatment of nano silica with silane coupling agents,especially using amino silanes. For example, Ishida and his

co-workers treated E-glass fibres and silica powder with

aminopropyltriethoxy silane. They varied the pH of treatingbath and reported that the maximum number of molecules

was adsorbed when silica was treated with a silane solution atits natural pH (10.6) (Wieczorek et al., 2004). Other

researchers have also reported the treatment of silica with

different types of organo functional silanes including amino,vinyl, methacryloxy and mercapto.Jesionowski and Kkrysztafkiewicz (2001) and Krysztafkiewicz

et al. (1981) used, however, acidic conditions for treatment of

amino silanewith silica. It seems that adjusting the pHof treatingbath during the reaction remains a challenge.Moreover, the role

of isoelectric point of silica on silane reaction has not been taken

into consideration systematically. This latter property maygreatly affect the state of reaction of amino silane with silica.

In addition, in previous works in this field, the ratios of couplingagent to silica were varied between 1 and 10weight parts per 100

parts silica. These seem to bewell below the stoichiometric value

of coupling agent to react with the existing surface silanols.In the work reported here, we aimed at treating fume silica

with aminopropyltrimethoxysilane (APTMS) at differentconditions in which the pH value was varied prior to and

during the reaction. The properties of the treated silica wereinvestigated using thermogravimetric analysis (TGA), FT-IR

spectroscopy and CHN analysis. Also, particles were

characterised by measuring their pH, zeta potential anddensity. The dispersion stability of the particles in different

solvents was also studied and the rheological behaviour of thetreated particles was investigated. It was attempted to discuss

the results of these analyses based on the chemistry of treated

particle at different process parameters.

Experimental

Materials

Hydrophilic fumed silica (Aerosil 200) having a specific

surface area 200m2/g and particle diameter of 12 nm wasobtained from Degussa Co. The pH of a 4wt% silica

dispersion (according ISO 787-9) and the number of hydroxylgroups of silica were 3.7-4.7 and 4.6 OH.nm21, respectively,

(Degussa web site). Aminopropyltrimethoxysilane (APTMS),97 percent in ethanol, was purchased from Degussa Co. All

other chemical reagents, including ethanol, hydrochloric acid

and caustic soda were purchased from Merck Co. Allmaterials were used as received.For the rheological studies, an acrylic polyol resin was used.

The resin was kindly provided by Tak Resin Co. (Iran) and

contained55wt%solid andhadanacidvalueof5-10mgKOH/g.

Method for silica treatment

Samples were refluxed in a glass round-bottom flaskequipped with a Graham condenser and a heating oil bath.

The amounts of silica and saline were calculated according toa stoichiometric value (unity), and hydrolysis ratio of 3

(h ¼ 3) as follows (Mrkoci, 2001):

M ¼ 3 £ SSSi £mSi £MSilane £OH:No £ 1019

NA

where M is the amount of silane (in grams), SsSi is the surface

area of silica, mSi is weight of silica, andMSilane is the molecularweight of silane. OH.No is the number of hydroxyl groups

per nm2 on silica surface, NA is the Avogadro number and 1019

is a conversion factor. All the reactions were performed athydrolysis ratio of 3 (the ratio of water to silane). Silane was

poured drop-wise at the rate of 0.5 g/min. The amounts of

materials in the reaction vessel are shown in Table I.To investigate the performance of the treated particles, six

different reaction conditions were conducted, as shown in

Table II.A mechanical homogeniser (IKA) was used for all samples at

19,000 rpm. All reactions lasted for two hours. The resultingslurries were first centrifuged at 5,000 rpm, followed by washing

for three times with ethanol. In each washing stage, the

suspensions were centrifuged. Then, the sedimented particleswere dispersed inwater:ethanol (50:50) solution and dried using

a spray dryer at 1108C to collect the variously treated silicas.

Method for preparation of acrylic polyol/silica mixtures

Treated and untreated silica particles were ultrasonicallydispersed in methoxy propyl acetate (MPA) for five minutes

and then added to the polyol. The content of silica was kept

constant at 4wt% for all samples. The dispersion of particlesin resin was achieved in one hour by pearl milling using

0.5mm zirconia beads at 1,200 rpm shear rate. The ratio of

zirconia beads to polyol was 1:1wt/wt.

Method for characterisation of treated particlesMeasurement of pHpH of each powder in the mixture of ethanol and water was

measured according to ISO-787-9 standard.

Measurement of densityApparent densities were recorded using a helium picknometer.

FT-IR spectroscopic analysisTo further study the attachment of silane coupling agent tosilica, FTIR spectroscopy was performed. The IR spectra of

samples were recorded using an FTIR Perkin Elmer

Spectrum One with KBr-sample discs.

TGATo investigate the efficacy of surface treatment, TGA wasperformed at N2 atmosphere with a heating rate of 108C

min21 from room temperature to 6008C by a TGA-DTA

Pyris Diamond SII Analyzer.

Elemental analysisElemental analysis for carbon, hydrogen and nitrogen content

of the modified and unmodified silica samples was performedaccording to ASTM- D5291 by a Foss-Herueus CHN-D-

Rapid Analyzer.

Table I Amount of materials in treatment reaction

Materials g

Amino silane 4.1

Silica 5

Absolute ethanol 85.44

Water 3.56

Effect of pH on an amino silane treated nano silica

Mehran Rostami, Mohsen Mohseni and Zahra Ranjbar

Pigment & Resin Technology

Volume 40 · Number 6 · 2011 · 363–373

364

NTU measurementTurbidity or dispersion stability of dispersed silica powders ina solvent can be considered as an indication of its surface

polarity. To this end, a HACH 2100AN Turbidimeter wasused to study the dispersibility of treated and untreatedsamples in solvents differing in Hansen solubility parameter(HSP) components according to ASTM-D4046. Table III

lists the solvents used together with their corresponding polar,disperse and hydrogen bonding HSP values. These solventswere used due to the similarity of gd (Hansen, 2000). Theclarity of upper phase and the rate of sedimentation ofparticles were also recorded as an indication of dispersibility.

Measurement of viscosityThe rheological studies were conducted using a ModularCompact Rheometer MCR300 on 4wt% silica dispersion atthe shear rate of 0-50 s21 at ambient temperature.

Zeta potential analysisTo discuss the different surface treatment approaches on thesurface charge of each sample, a ZETA Sizer MALVERN3000HS was used to measure the zeta potential at different

pHs. To this end, 0.15wt% silica was dispersed in a mixtureof ethanol: H2O (50:50wt/wt).

Characterisation of the morphology of particles

For morphological studies, a JEOL JEM2010 TEM operated

with a thermionic emission gun at accelerating voltage of 200keV was used. Samples were prepared by depositing a verydiluted suspension of particles in ethanol on carbon coatedcupper grids and further drying at ambient.

Results and discussion

Chemical modification of silica surface using alkoxysilanes hasbeen reported by many researchers (Culler et al., 1986;

Jesionowski and Kkrysztafkiewicz, 2001; Naviroj et al., 1984;Pham et al., 2007; Vansant et al., 1995; Wieczorek et al.,2004). Low concentration of amino and epoxy silanesolutions were often used to modify silica surfaces(Kang et al., 2001; Shen et al., 2004; Vejayakumaran et al.,2008). The report by Vansant et al. (1995) showed that

low concentration of silane coupling agents has prevented the

occurrence of multi-layer adsorption of alkoxysilanes on

the silica surface. Moreover, it has been discussed that a

high concentration of alkoxysilanes has led to a multi-modal

size distribution and an increase in the average particle

size (Jesionowski and Krysztafkiewicz, 2002). Therefore,

a stoichiometric amount of APTMS was used in the present

study and the reaction time of two hours was utilised to react

silica with the silane.In treating method S1, APTMS was added to the slurry of

silica and ethanol at pH 7. At this pH, silane has low reactivity

(Xhanthos, 2005). Silane was added drop-wise (0.5 g/min)

and the slurry pH reached 9-10. At this condition, the

reactivity of silane is high, so it could be hydrolysed into

silanol and then condensed with silica. At this condition, the

condensation of silane seemed very fast and both silica surface

and silanol groups in silane are negatively charged, so they

may not show affinity to each other. It may be expected that

silica hydroxyl groups and hydrolysed silane interact with each

other through hydrogen bonding. The reactivity of silanol

groups at pH 9-10 was high, resulting in condensation and

formation of Si–O–Si networks. Therefore, the increased

viscosity of slurry may be explained due to this reactivity.In method S2, silica was added to the mixture of ethanol

and water according to Table II. It was dispersed by a

homogeniser for about 30 minutes and then 1ml hydrochloric

acid was added to change pH from 7 to 1.24. At this pH,

silica surface is positively charged. The silane was then added

drop wise. The pH of slurry changed from 1.24 to 9 and the

viscosity increased rapidly.In method S3, silica was first added to the mixture of

ethanol and water. Three microlitre acid was then added to

lower the initial pH to 1. At this condition, silica surface has

positive charges. Silane (diluted in ethanol) was then added

drop wise. In this case, silane has high reactivity and form

silanol. The pH of slurry reached two and the viscosity

increased. After a few minutes, 3ml caustic soda was added

and the pH of reaction changed to 4.5.In method S4, silane was first added to the mixture of

ethanol and water. The pH of ethanol changed from 7 to 10.

The initial pH of the mixture before the addition of silica was

ten. At this condition silica has negative charges and Si–O2

groups may not show affinity to silica surface. At this pH,

silane has high reactivity and changes into silanol, with no

obvious change in viscosity. At this condition, compared to

method S1, silane may have enough time to condense with

itself and form oligomers. Then silica was added and the

viscosity of slurry increased rapidly.In method S5, silane was first added to the mixture of

ethanol and water, to which was added 3ml HCl to change

Table II Conditions of treatment

Samples Components prior to treatment pH (initial)

Added component

for treatment

PH (in the course

of reaction) Final pH

S1 H2O þ EtOH þ Silica 7 Silane 9-10 9-10

S2 H2O þ EtOH þ Silica þ HCl(1cc) 1-2 Silane 8-9 8-9

S3 H2O þ EtOH þ Silica þ HCl(3cc) 1-2 Silane 1-2 4-6 NaOH (2cc)

S4 H2O þ EtOH þ Silane 9-10 Silica 9-10 9-10

S5 H2O þ EtOH þ Silane þ HCl(3cc) 1-2 Silica 1-2 4-6 added NaOH (2cc)

S6 H2O þ EtOH þ Silane þ HCl(3cc) 1-2 Silica 1-2 1-2

Table III HSPs of solvents

Y(MPa0.5)

Solvent YH Yd Yp

Butyl acetate 6.3 15.8 3.7

2-Propanol 16.4 15.8 6.1

MEK 5.1 16 9

Effect of pH on an amino silane treated nano silica

Mehran Rostami, Mohsen Mohseni and Zahra Ranjbar

Pigment & Resin Technology

Volume 40 · Number 6 · 2011 · 363–373

365

pH from 10 to 1.4. At pH 10, silane has high reactivity andchanges to silanol. Silanol reactivity is high at pHs lower andhigher than three. Silica was then introduced to this mixture.About 2ml of 5wt% caustic soda was added. At thiscondition, the pH of reaction was 4.4, and the protonatedamino groups may have formed. These positively chargedgroups have affinity to surface.Method S6 is similar to method S5 but in the absence of

caustic soda. The pH of treating bath was adjusted at 1-2using 3 ml HCl. At this condition, the hydrolysis of silane isgreater. In addition, silica contains positive charge with whichsilane groups more easily react. The pH was kept acidicduring the course of reaction. This led to lowered affinity ofprotonated amino groups to react with silica.All observations made above can be discussed based on the

hydrolysis conditions of silane and its ability to react withsilica particle depending on the pH. Silanols are mostly stableat around pH 3, and their reactivity is higher at the pH lowerthan 1.5 or higher than 4.5. Silanols condense to formoligomers and, ultimately, two- and three-dimensionalnetworks may form. When considering silane hydrolysis andcondensation, a different reactivity in different pH ranges canbe expected. At very low pH, silanes hydrolyse very quickly.The formed silanols are relatively stable and, over time, formcoordinated networks. At neutral pH, however, silaneshydrolyse very slowly to silanols, which are unstable andtherefore condense. Thus, in both cases, there is still a slowreaction in the transition from silanes to Si–O–Si networks(Li et al., 2006; Wypyeh, 1999; Xhanthos, 2005). At pH . 8,silanes become highly reactive again and form silanols veryquickly. These silanols are very unstable and condense to giveuncoordinated Si–O–Si networks. Therefore, it is predictedto observe different behaviour for the above treated silicas asprepared according to Table II. It may be expected that thetreated sample with method S6, in which pH adjustment oftreating procedure is kept at 1-2, may contain greater graftedsilane. It should be noted that the isoelectric point of silica isthree, meaning that the possibility of reacting negative Si–O2

groups with the positive charges of surface may be higherbelow this pH. These may mean that the sample preparedusing method S6 probably contains higher amount of graftedsilane. On the other hand, as the silane contains aminofunctional groups, the existence of greater amount of siloxanestructure will coincide with the presence of more aminofunctionalities.

Particles characterisationpH and density measurementsTo further study the behaviour of treated particles, pH valuesof each sample together with that of the untreated one weremeasured, as shown in Table IV.The untreated fumed silica has a pH of 4.1 according to

Table IV, the pHs of treated fumed silicas are higher thanthat of the untreated one (S0). This means that treatedfumed silicas have become more alkaline, the extent ofwhich depends on the presence of amino functionalities at

the surface. As can be seen from Table IV, sample S6 has the

highest pH among the other samples. This is in agreement

with the hypothesis made above suggesting that sample S6may be composed of a higher amount of silane grafted.

In turn, sample S4 in which the pH of treating bath is very

alkaline, may have lower amount of silane grafted, judging

from its pH being closer to that of untreated silica. According

to methods S5 and S3, amino groups in these samples have

probably protonated and reacted with negatively charged

silica surface. So, the pH values of samples S5 and S3 are

lower than S6. This may suggest that the attachment of silane

to silica has occurred from the amino side rather than from

silanol (upside down). These more ionic silanes attached to

silica may be easily washed away.The density of each particle was also measured by Helium

Piknometer. The density of all treated particles has decreased

(Table IV) because of the presence of an organic layer and the

increased particle volume. Accordingly, the mass per unit

volume (r) of treated particles decreased. Sample S6 has the

lowest density. This is again in agreement with the result of

pH measurement. The decreased density, however, may not

seem very significant. But, with respect to treating procedures

adopted, it may reveal that the particles which densities are

closer to that of S0 may have possibly lower amount of

grafting silane at the surface and lesser organic functionalities.

The lowest density is seen for sample S6, meaning that, the

silane grafting is the highest among other samples.

FT-IR studies

Figures 1 and 2 show the FTIR spectra of the untreated (S0),

and sample S6. The characteristic IR absorption peaks for pure

and modified samples are also summarised in Table V (from

Socrates, 2001). A sharp and strong Si–O–Si stretching peak

(1,010-1,190 cm21) is observed for both samples, indicating

that the main structure has not changed by the modification

reactions (Naviroj et al., 1984; Plueddmann, 1982; Socrates,

2001). The peak at 800-900 cm21 corresponding the silanol

groups can be observed for both samples, confirming that after

the reaction between the silica surface and APTMS, not all

silanol groups are consumed. This also suggests that the

modification reactions occurred primarily on the silica surface

rather than in the internal structure. The peak at 670-760 cm21

and 1,170-1,250 cm21 corresponding to Si–C groups can be

observed in Figure 2. The latter is suppressed by the presence of

Si–O–Si peak at around 1,110 cm21. The broadness of this

peak for the treated sample, compared to S0, may also confirm

the presence of Si–C peak which is absent in Figure 1.The peak observed at 1,600-1,630 cm21 may indicate the

presence of physically adsorbed water.The FTIR spectrum of sample S6 shows an absorption

band at ,2,920 cm21, which is attributed to –CH2 due to

the presence of propyl groups in APTMS. The peak observed

at 1,430-1,500 cm21 corresponds to NH2 deformation.

The peak attributing to NH2 stretching has overlapped with

the stretching vibrations of O-H groups at 3,200-3,500 cm21.

Table IV pH value and density of treated and untreated fumed

Samples S0 S1 S2 S3 S4 S5 S6

PH value 4.1 8.4 8.8 8.2 7.1 8 9.6

Density(g/cm3) 6 0.0001 2.1405 2.0614 2.0302 2.0373 2.0806 2.0188 1.9621

Effect of pH on an amino silane treated nano silica

Mehran Rostami, Mohsen Mohseni and Zahra Ranjbar

Pigment & Resin Technology

Volume 40 · Number 6 · 2011 · 363–373

366

All these observations may reveal the existence of a chemically

attached APTMS. To compare the structure of samples S1-S6using FT-IR spectra, the intensity of the peak corresponding

to the NH2 deformation in APTMS grafted samples can be

used as a reasonable indication of grafting yield. As can be

seen in Table V, the peak corresponding to NH2 deformation

is appeared at 1,420-1,550 cm21. This peak, which can only

exist in APTMS grafted silicas, was taken as an indication of

grafting. Therefore, the intensity of this peak to that of CH2

can be estimated as the extent of grafting. The higher is this

ratio, the greater is APTMS at the silica surface. The

proportional intensities of NH2-deformation/CH2 of treated

sample are shown in Table VI. According to this table, sample

S6 has the highest ratio, which means that the presence of

amino groups is greater. This may correspond to a higher

content of the silane at the surface. The reason for such an

observation is that in sample S6 the pH used for hydrolysis

is more effective. This pH is lower than that of the pH of

Figure 1 FTIR spectrum of untreated silica sample

0.1

Si-OHSilanol

Si-O-Si

3,430

OH (Molecular Water)

1,630

1,110

811

Si-OHSilanol

888684828078767472

Tra

nsm

isio

n (%

)

706866646260585654525048

5.04,000 3,600 3,200 2,800 2,400 2,200 1,800 1,600

Wavenumber, cm–1

1,400 1,200 1,000 800 600

Figure 2 FTIR spectrum of sample S6

2.480

75

70

65

60

55

50

45

40

35

30

25

20

15

10

51.0

Tra

nsm

isio

n (%

)

4,000 3,600 3,200 2,800 2,400 2,200 1,800 1,600

Wavenumber, cm–1

1,400 1,200 1,000 800 600

OH-NH2

Si-C

Si-O-Si

CH2

NH2

OH Molecular Water

3,4193,043

1,615

804

701

Si-C

Si-OH

1,104

1,494

C-N

Effect of pH on an amino silane treated nano silica

Mehran Rostami, Mohsen Mohseni and Zahra Ranjbar

Pigment & Resin Technology

Volume 40 · Number 6 · 2011 · 363–373

367

isoelectric point of silica, at which the surface is enriched ofpositive charges. Meanwhile at pH 1-2 the hydrolysed silanemolecules form silanol groups.These can easily react with the silanol at the silica surface.

Hence, as the condition for this reaction is facilitated, the

grafting occurs more effectively. Consequently, at somecondition the silane may not be easily hydrolysed (e.g.,sample S4) or the condition is such that pH is above theisoelectric point of silica (lower affinity to graft). Theprotonation of amino groups at pH 4-6 is also a concernbecause the formation of a zwitter ion through self-reaction ofsilanol and protonated amine can take place. Alternatively, thesame reaction may occur between positively charged aminogroups and silica surface silanols. At these two possibleconditions, the bonding between silane and silica seemsweaker and can be desorbed from the surface during washing.

The overall result is that the grafting content reduces.

TGA and CHN analyses of particles

The APTMS-modified nano silica particles and untreatedfumed silica were further characterised by TGA, as shown inFigure 3. For treated samples, two and three stages of weightlosses can be observed. The weight loss at low temperatures

(,2008C) could be attributed to the release of adsorbedwater and volatiles such as alcohol, while the weight loss inthe temperature range 200-6008C are resulted from the silanolcondensate groups and the decomposition of the graftedAPTMS molecule. The decomposition of aminopropylgroups occurs slowly at 200-4008C, followed by a rapid

degradation above 4008C. The decomposition of amino silane

leads to formation of ammonia, ethylene, hydrazine and

methane (Li et al., 2006). Table VII reports the results of

TGA analysis for different samples.According to Table VII, sample S6 has the greatest total

weight loss. Accordingly, it may be predicted that grafting of

silane on the surface of this sample may be the highest one.

This is also in accord with the result of pH and density

measurements (Table IV). It can be observed that the weight

loss of all samples is higher than that of the unmodified silica.

The ash content of samples at 6008C has also been shown in

Table VII. It was also observed that when samples burned at

5008C the residual colour of sample S6 is darker compared to

that of other samples. This means that the organic content of

this sample is higher as revealed by the ash content in

Table VII. In turn, samples S1 and S4 in which the conditions

were less in favour of grafting of silane to silica, showed a

brighter residue when fired, indicating lesser amounts of

organics presented at the surface. This can also be seen from

the weight losses of these samples from Table VII. It means

that the pH variations used in samples S1 and S4 lead to

particles, which their adsorbed silica layer is more physical

rather than chemical. The lower ratios of NH2 deformation to

that of CH2 (as shown in Table VI) for these samples are in

good agreement with the thermal analysis.To quantify the extent of silane adsorbed on the surface of

particles, CHN analysis was done.Table VIII gives the elemental analysis of treated and

untreated samples. The results show that highest amount of

organo functional groups corresponds to sample S6. The C, H

and N contents for S0 sample (#0.5) is negligible which can

be considered as the instrumental error. For better

comparison, the total weight losses at 6008C, as shown in

Table VII, are also accompanied by the elemental analysis

results. As can be seen, except for the S0, the presence of C, H

and N on other samples is more significant. Sample S4 has the

lowest total content of CHN and sample S6 shows the

greatest. A simple presumption was conducted to calculate

the percentage of grafting as follows.Each mole of silane contains one mole nitrogen. Therefore,

by dividing the N contents in Table VIII to the atomic

number of nitrogen (14) one obtains the moles of nitrogen at

the treated silica surface which is in turn equal to the number

of moles of silane molecules grafted. The initial mole of silane

at the starting bath (based on 4.1 g silane for 5 g silica) is

0.458 for 100 g silica. This is assumed to be the stoichiometric

amount (One mole silane to each mole hydroxyl group). By

dividing moles of nitrogen to 0.458, the percentage of grafting

is obtained. This grafting percentage is seen in the last column

of Table VIII. The results revealed that the grafting

percentage was varied between 29 and 58, meaning that at

different conditions used in Table II, not more than 58 percent

of grafting can be reached. Interestingly, this corresponds to

sample S6 and is again in agreement with the results shown

previously. This indicates that the percentage of grafting at

various conditions used in Table II is highly pH dependent. In

other words, the condition at which the treating bath remains

acidic enough (pH , 3) prior and during the reaction will

lead to a greater grafting. For samples in which the pHs

distance from 3 (get closer to pH 7 and above) the grafting

tends to decline. This has been observed for samples S5

and/or S3 (pH 4-6) and S1, S2 and S4 (pH 8-10), respectively.

Table V IR absorption of main functional groups for pure and modifiedsilica nanoparticles

Functional group Wave number (cm21) Intensity

Si–OH 3,200-3,700 M

1,020-1,040 W-M

800-955 S

O-H (molecular water) 1,600-1,630 M

Si–O–Si 1,010-1,190 VS

2CH2 2 2,850-2,945 M

NH2 deformation 1,420-1,550 M

Si–CH2 2 670-760 M

1,175-1,250 W

NH2 800-850 M

1,600-1,700 M

3,300-3,500 M

Notes: S ¼ Strong, VS ¼ very strong, M ¼ medium, W ¼ weakSource: Socrates (2001)

Table VI The ratio of intensities of NH2/CH2 vibration frequency

Samples NH2/CH2

S1 1.087

S2 1.159

S3 1.018

S4 1.013

S5 1.064

S6 1.27

S0 0

Effect of pH on an amino silane treated nano silica

Mehran Rostami, Mohsen Mohseni and Zahra Ranjbar

Pigment & Resin Technology

Volume 40 · Number 6 · 2011 · 363–373

368

Surface properties of particles

Surface chemistry of untreated and treated silica can be

evaluated by observation of the suspension and/or

sedimentation behaviour of particles in different liquids.

Those liquids with stronger interaction can suspend solids

more effectively or retard sedimentation.To compare the dispersibility of treated and untreated

samples they were dispersed in butyl acetate, 2-propanol and

MEK. The HSP of solvents is shown in Table III. The

Turbidity and/or rate of settling (dispersion stability) of each

slurry were recorded by a turbid meter. The NTU

(Nephelometric Turbidity Unit) is related to size, shape and

concentration of particle. In this method the concentration of

particles in solvent is similar. The rate of sedimentation is

related to density and surface chemistry of particles. If

particles have a good interaction with solvent, they have low

rate of settling. The results are shown in Figure 4. It should be

noted that the higher dispersibility of particle in solvents is

directly related to the interaction of particle with its media.

Therefore, in turbidimetric measurements those particles are

considered to have more interaction with the solvent that have

higher NTU values. It means that the slope of reduction in

NTU can more reliably be used as an indication of dispersion

stability. Accordingly, although treated samples have almost

higher NTU in MEK at the start of measurement, as shown

in Figure 4, the slope is much higher than that of S0 sample,

meaning that the stability of dispersion by these particles are

weaker in MEK. This in turn can be discussed by surface

chemistry of treated particles as a result of silanisation.It can be seen in Figure 4 that, except for untreated sample,

all treated ones tend to sediment after ten minutes, making

the NTU lower. According to Table III, MEK has low gH and

high gP. The S0 has higher polarity and more hydroxyl groups.

To further observe the behaviour of samples in solvents with

different gH and gP, the NTU values of all particles were also

studied in butyl acetate (not shown here). The very low

amounts of NTU (lower than ten) for almost all samples

dispersed in this solvent means that the stability of dispersion

can be considered very low (all samples showed the same

settling behaviour). This arises from the very low gP of butyl

acetate compared with that of MEK. However, the gH of the

former is slightly higher than that of the latter (6.3 vs. 5.1).

The conclusion from this observation is that for better

dispersion stability the hydrogen bonding of HSP (gH) is as

Figure 3 TGA of treated and untreated samples

80

8284

86

8890

9294

96

98100

102

0 100 200 300 400 500 600 700

Temperature (°C)

%R

esid

ual W

eigh

t

S1

S2

S3

S4

S5

S6

S0

Table VII TGA of sample

Samples

Decomposition

step

Temperature

range (8C)

Loss

(wt%)

Total

weight loss

(wt%)

S0 1 0-200 0.25

2 200-400 0.26 0.8

3 400-600 0.26

S1 1 0-200 2.13

2 200-400 3.43 8.9

3 400-600 3.48

S2 1 0-200 1.69

2 200-400 2.15 7.7

3 400-600 3.89

S3 1 0-200 2.24

2 200-400 7.8 13.2

3 400-600 3.28

S4 1 0-200 2.35

2 200-400 1.65 9.83

3 400-600 5.92

S5 1 0-200 1.76

2 200-400 6.94 12.86

3 400-600 4.33

S6 1 0-200 1.9

2 200-400 8.96 15.8

3 400-600 5.18

Table VIII Elemental analysis of samples

Elements (mass %)

Samples Carbon Hydrogen Nitrogen SCHN

Weight

loss

at 6008C

Percentage

of grafting

based on

N content

S1 8 2 2.9 12.9 8.9 47

S2 8.5 2.2 3.2 13.96 7.7 51

S3 6.8 2.1 2.8 11.7 13.2 45

S4 5 1.2 1.8 8 9.83 29

S5 6.9 1.7 2.3 10.9 12.86 37

S6 9.5 2.9 3.6 16 15.8 58

S0 #0.5 #0.5 #0.5 #1.5 0.8 <0

Effect of pH on an amino silane treated nano silica

Mehran Rostami, Mohsen Mohseni and Zahra Ranjbar

Pigment & Resin Technology

Volume 40 · Number 6 · 2011 · 363–373

369

important as the polar (gP) one. The fact that MEK showed

to be a better solvent for S0 is due to the higher gP as

explained above. However, when the samples were dispersed

in 2-propanol, the S0 has lower initial NTU value which

means that the interaction between this particle is lower

compared with the treated ones. The slopes in NTU curves

after 20 minutes seem similar for all samples dispersed in

2-propanol. The lowerNTU is an indication of less interaction.

MEK has gH equal to 5.1 while that of 2-propanol is 16.4

which is 3.2 times greater. The gP of MEK is nine compared to

6.1 in 2-propanol. In spite of having a lower gP in 2-propanol

(6.1 vs 9), the gH of 2-propanol is much higher than that of

MEK (16.4 vs 5.1).Therefore, the better dispersibility of treated samples in

2-propanol is most likely due to the greatest hydrogen bonding

component rather than gP. One can conclude that as the silica

is silanised using APTMS, the contribution of gH shows a key

parameter to affect the dispersion stability. It means that the

treated samples can interact better in solvents with high gH and

gP. It can be predicted that the amino groups on the treated

surface can better interact with the media having high gH

compared with silanol groups in S0. The reason for such an

observation can be explained by the higher nitrogen contents of

S6 sample, in which a higher grafting has occurred. In addition

the lowest grafting corresponding to sample S4 is further

confirmed by its inferior interaction with 2-propanol (lower

NTU) which is closer to the behaviour of S0. Therefore, the

results of turbidimetric measurements have good agreement

with the results shown previously.This can be further explained by the presence of an

adsorbed layer of siloxane on the treated silica surface, thick

enough to prevent the aggregation of particles. It should also

be noticed that if one only compares the gP of a media without

taking into account the value of gH, the prediction of

dispersion stability seems difficult.Other methods for evaluation of surface properties of

silanised silica have also been reported. These methods were

mainly based on the heat of immersion in water and benzene.

The principle of surface silanisation using amino silane is the

decrease or ablating of hydrophilic nature of pristine silica.

Figure 4 NTU values of samples in MEK and 2-propanol

00 10 20 30 40 50 60 70

50

100

150

200

250

300

350

Time (min)

0 10 20 30 40 50 60 70

Time (min)

NT

U (

in2-

Pro

pano

l)

S1 S2 S3 S4 S5 S6 S0

0

50

100

150

200

250

NT

U (

in M

EK

)

S1 S2 S3 S4 S5 S6 S0

Effect of pH on an amino silane treated nano silica

Mehran Rostami, Mohsen Mohseni and Zahra Ranjbar

Pigment & Resin Technology

Volume 40 · Number 6 · 2011 · 363–373

370

The presence of amino groups at the surface provides it with

new organo functional groups while reduces its hydrophilicity.

Jesionowski and Kkrysztafkiewicz (2001) measured the heatof immersion of treated silica with amino silane. They found

that the heat of immersion of amino treated silicas increased inboth water and benzene. However, the authors reported that

the degree of hydrophobisation of amino silane treated silicahad increased. This means that the presence of amino silane

has two roles. The first role results from the siloxane and

alkyl structure (hydrophobic) and the second comes fromthe amino functionality (hydrophilic). The overall result is

reducing hydrophilic properties of silica and imparting surfacefunctional groups. Further studies with these treated particles

are being performed in our laboratory to investigate the effectof these particles in a polyurethane film.

Rheological studies of particles suspension

The rheological properties of the fumed silica dispersions areinfluenced by both the dispersion state and the interaction

between particles and the medium (Naviroj et al., 1984).However, it is difficult to discuss the respective contribution

of these effects, as the dispersion state is also related to thesurface interaction. To study the effect of surface chemistry on

the rheological behaviour of samples on acrylic based polyol

was used. Figure 5 shows the viscosity of treated anduntreated silicas at different shear rates.The amount of adsorbed polyol on the silica powders with

strong interaction and/or homogeneous dispersion, play an

important role in the viscosity of polyol/silica. According toFigure 5, the viscosity of polyol with modified silica are shear

thickening compared to that of S0 sample. In other words,

filled resins containing treated particles have shear thickeningbehaviour while the resin with S0 particle shows a thinning

effect. This arises from the interaction of the adsorbed layerwith the media. According to Figure 5, sample S6 has the

highest shear thickening effect. This observation again is inagreement with the results shown previously. Therefore, the

higher percentage of grafting leads to a greater interaction

with the media. This is due to the increased hydrogenbonding between the surface treated samples and the polyol.

As the moles of amino silane increases the possibility of

making hydrogen bonding with polyol increases. It should be

noted that the rate of increase in viscosity reach to a constant

value at shear rates exceeding 10 s21. This may mean that theinternal structure resulted from hydrogen bonding cannot

withstand the shearing applied greater than 10 s21. At shearrates lower than 10 s21, with respect to the concentration

used in this experiment (4wt%), the results shown in Figure 5reveal that the surface chemistry of treated particles have

considerably changed. This leads to a better interaction

between the particles and polyol due to the effect of hydrogenbonding. The comparison between different samples shows

that the viscosity of sample S6 is higher than that of othersample. This again can be explained because of a greater

presence of amino silane at the surface and higher interactionwith the resin.

Zeta potential studies

Measurement of the electrokinetic potential can be used incolloidal systems. Moreover, electrokinetic measurements are

increasingly successful in characterising solid surfaces. Electrokinetic phenomena can be observed by the contact of a solid

surface with a polar liquid medium, because of the existenceof an electric double layer at the solid–liquid interface. The

potential at this surface is called the electro kinetic or zeta (z)

potential (Alkan et al., 2005). Zeta potential occurs at theslipping plane of a solid immersed in a water solution can be

directly related to the magnitude and sign of the surfacecharge (Hunter, 1981). The zeta potential is therefore used

for measuring the potential of solid at the shear plane tocharacterise its electrokinetic behaviour, because the surface

potential is not directly measurable.The surface charge on treated and untreated particles was

measuredusing zeta potential of 0.15percent suspensions at pHs

2, 4, 7 and 9. Figure 6 shows the results of such measurements.It was revealed that the S0 has a surface potential 27.1mv (at

pH 7). As known, S0 has an isoelectric point at pH 3. Therefore,the observation of a negative value in zeta potential at pH7 seems

reasonable, because at this pH the untreated silica is basic.

However, the zeta potentials of treated samples at pH 7 wereall positive. The values measured for samples S3, S4, S5 and S6are 12.95, 8.5, 11.1 and 18.01mv, respectively, (at pH 7).

Figure 5 Viscosity of treated and untreated silicas at different shear rates

00 10 20 30 40 50

0.05

0.1

0.15

0.2

0.25

0.3

Shear Rate (S–1)

Vis

cosi

ty (

Pa.

S)

S1 S2 S3 S4 S5 S6 S0

Effect of pH on an amino silane treated nano silica

Mehran Rostami, Mohsen Mohseni and Zahra Ranjbar

Pigment & Resin Technology

Volume 40 · Number 6 · 2011 · 363–373

371

This shows that at pH 7 the treated particles show acidic

behaviour.Hence, it can be expected that the isoelectric points of

treated silicas are all above pH 7. This means that particles

becomemore basic upon treatingwith amino silane, as expected.

A more positive value for zeta potential means a more basic

character, as a result of high silane grafting. As shown previously

the pH of treated samples varied based on the chemistry of

particles (Table IV). It can be seen that there is a good agreement

between the pHof particles S3, S5 and S6 (in Table IV)with their

zeta potentials. The pHs of these samples are 8, 8.2 and 9.6,

respectively. Therefore, as seen in other experiments the higher

presence of grafted amino silane at the surface coincides with a

more basic particle enabling to interact through hydrogen

bonding with less polar media. These observations confirm the

chemical attachment of silane to the surface. At proper pH

adjustment, grafting seems more effective due to the amino

groups being up side up (Bellmann et al., 2002).Finally, for comparing the morphological studies of untreated

and treated particles a transmission electron microscope was

used. Figure 7 shows the typicalTEMmicrographs of samples S0andS6, inwhich particleswith average sizes of 12-16nmare seen.

It canbeseen that the shapeandsizeof the treated samplewith the

highest amount of grafting is almost similar to that of the pristine

silica.Theparticle size of sampleS0 is ca.12nmas reportedby the

manufacturer. So, it can be concluded that the morphologies of

treated do no change considerably upon grafting. However, the

chemistry of particles has altered significantly.

Conclusion

The purpose of this work was to study the surface chemistry ofmodified silica using aminopropyl trimethoxy silane. It was

observed that the pH of treating bath as well as the order ofadding reactants affected the grafting of silane at the surface.The surface chemistry of treated particles was studied usingdifferent analytical techniques. It was concluded that thehighest grafting corresponded to the method in which silane is

first hydrolysed at pH 1-2 followed by addition of silica andmaintaining the pH at 1-2. Other conditions result in obtainingsilica with lower grafting. This was attributed to the greaterability of silane to hydrolyse and to chemically adsorb to the

silica. All treated particles were analyzed by FT-IRspectroscopy. The presence of higher content of carbon-Nitrogen group together with a greater value of particle pH anda lower density were also observed for the sample treated at very

acidic pH during the course of reaction. The graftingpercentages, however, varied between 28 and 58. The highestcontent of grafting lead to a better silica dispersion in solventshaving greater hydrogen bonding component and inferiordispersion in solvents having higher polar component. The zeta

potential of particles also changed at the different treatments.

Figure 6 Zeta potential of particles in different pH

–27

–17

–7

3

13

23

1 2 3 4 5 6 7 8 9 10

pH

Zet

a P

oten

tial (

mv)

S3 S4 S5 S6 S0

Figure 7 Transmission electron micrographs of untreated (S0) and S6 samples

S6 S0

20 nm 20 nm

Effect of pH on an amino silane treated nano silica

Mehran Rostami, Mohsen Mohseni and Zahra Ranjbar

Pigment & Resin Technology

Volume 40 · Number 6 · 2011 · 363–373

372

The unmodified particle showed a zeta potential of 24.6mvwhile this was þ18.01mv for the highest grafted particle atpH 7. This increased potential corresponded to a more basiccharacter as a result of amino functionality at the surface. Therheological behaviour of treated particle also changed fromshear thinning to shear thickening for untreated and treatedsamples, respectively. The overall conclusion is that the surfacechemistry highly depends on the treating procedure.

References

Alkan, M., Tekin, G. and Namli, H. (2005), “FTIR and zetapotential measurements of sepiolite treated with someorganosilanes”, Microporous and Mesoporous Materials,Vol. 84 Nos 1-3, pp. 75-83.

Bellmann, C., Klinger, C., Opfermann, A., Bohme, F. andJurgen, H.P. (2002), “Evaluation of surface modification byelectrokinetic measurements”, Progress in Organic Coatings,Vol. 44 No. 2, pp. 93-8.

Chen, G., Zhou, S., Gu, G. and Wu, L. (2007), “Modificationof colloidal silica on the mechanical properties of acrylicbased polyurethane/silica composites”, Colloids and SurfaceA: Physicochemical and Engineering Aspects, Vol. 296 Nos 1-3,pp. 29-36.

Chen, G., Zhou, S., Gu, G., Yang, H. and Wu, L. (2005),“Effects of surface properties of colloidal silica particles onredispersibility and properties of acrylic-basedpolyurethane/silica composites”, Journal of Colloid andInterface Science, Vol. 281 No. 2, pp. 339-50.

Culler, S.R., Ishida, H. and Koeing, J.L. (1986), “The silaneinterphase of composites: effect of process conditions on g

aminopropyltriethoxy silane”, Polymer Composites, Vol. 7No. 4, pp. 231-8.

Hansen, C.M. (2000), Hansen Solubility Parameters, A UserHandbook, CRC, Boca Raton, FL.

Hunter, R.S. (1981), Zeta Potential in Colloid Science: Principlesand Applications, Academic Press, London.

Iijima, M., Tsukada, M. and Kamiya, H. (2007), “Effect ofparticle size on surface modification of silica nonoparticlesby using silane coupling agents and their dispersion stabilityin methylethylketone”, Journal of Colloid and InterfaceScience, Vol. 307 No. 4, pp. 418-24.

Jesionowski, T. and Kkrysztafkiewicz, A. (2001), “Influence ofsilane coupling agents on surface properties of precipitatedsilicas”, Applied Surface Science, Vol. 172 Nos 1-3, pp. 18-32.

Jesionowski, T. and Krysztafkiewicz, A. (2002), “Preparationof the hydrophilic/hydrophobic silica particles”, Colloids andSurfaces A, Vol. 207 Nos 1-3, pp. 49-58.

Jiang, Z.X., Meng, L.H., Huang, Y.D., Liu, L. and Lu, C.(2007), “Influence of coupling agent chain lengths oninterfacial performances of polyacrylacetylene resin andsilica glass composites”, Applied Surface Science, Vol. 253No. 9, pp. 4338-43.

Kang, S., Hong, S., Choe, C.R., Park, M., Park, R.S. andKim, J. (2001), “Preparation and characterisation of epoxycomposites filled with functionalized nanosilica particlesobtained via sol-gel process”, Polymer, Vol. 42 No. 3,pp. 879-87.

Krysztafkiewicz, A., Wieczorek, W. and Dmka, L. (1981),

Polimery, Vol. 26, p. 175.Li, X., Cao, Z., Zhang, Z. and Dang, H. (2006), “Surface-

modification in situ of nano-SiO2 and its structure and

tribological properties”, Applied Surface Science, Vol. 252

No. 22, pp. 7856-61.Mrkoci, M.I. (2001), “Influence of silica surface

characteristics on elastomer reinforcement”, M.Sc. thesis,

Queen’s University Kingston, Ontario Canada.Naviroj, S., Culler, S.R., Koenig, J.L. and Ishida, H. (1984),

“Structure and adsorption characteristics of silane coupling

agents on silica and E-Glass fiber; dependence on Ph”,

Journal of Colloid and Interface Science, Vol. 97 No. 2,

pp. 308-17.Pham, K.N., Fullston, D. and Sagoe-Crentsil, K. (2007),

“Surface modification for stability of nano-sizes silica

colloids”, Journal of Colloid and Interface Science, Vol. 315No. 1, pp. 123-7.

Plueddmann, E.P. (1982), Silane Coupling Agents, Michigan

Plenum Press, New York, NY.Roberts, W.O. and Bergna, H.E. (Eds) (2006), Colloidal

Silica, Fundamentals and Applications, CRC Taylor

& Francis, Boca Raton, FL.Shen, G., Horgan, A. and Levicky, R. (2004), “Reaction of

N-phenyl maleimide with aminosilane monolayer”, Colloidsand Surfaces B, Vol. 35 No. 1, pp. 59-65.

Socrates, G. (2001), Infrared and Raman Characteristic GroupFrequencies, Table and Charts, 3rd ed., Wiley, Chichester.

Sun, Y., Zhang, Z. and Wong, C.P. (2005), “Study on mono-

dispersed nano-size silica by surface modification for

underfill applications”, Journal of Colloid and InterfaceScience, Vol. 292 No. 1, pp. 436-44.

Vansant, E.F., van der Voort, P. and Vrancken, K.C. (1995),

Characterisation and Chemical Modification of the SilicaSurface, Elsevier Science, New York, NY.

Vejayakumaran, P., Rahmana, I.A., Sipaut, C.S., Ismail, J.

and Chee, C.K. (2008), “Structural and thermal

characterisations of silica nanoparticles grafted with

pendant maleimide and epoxide groups”, Journal ofColloid and Interface Science, Vol. 328 No. 1, pp. 81-91.

Wieczorek, M., Krysztafkiewicz, A. and Jesionowski, T.

(2004), “Influence of modification by N-2-(aminoethyl)-

3-aminopropyltrimethoxysilane on physicochemical

properties of bentonite”, Journal of Physics and Chemistryof Solids, Vol. 65 Nos 2/3, pp. 447-52.

Wypyeh, G. (1999), Hand Book of Filler, Chem Tec

Publishing, Torento.Xhanthos, M. (2005), Functional Filler for Plastics, Wiley-

VCH Verlag GmbH, Germany.

Further reading

Degussa (n.d.), web site, available at: www.aerosil.com

Corresponding author

Mohsen Mohseni can be contacted at: mmohseni@aut.ac.ir

Effect of pH on an amino silane treated nano silica

Mehran Rostami, Mohsen Mohseni and Zahra Ranjbar

Pigment & Resin Technology

Volume 40 · Number 6 · 2011 · 363–373

373

To purchase reprints of this article please e-mail: reprints@emeraldinsight.com

Or visit our web site for further details: www.emeraldinsight.com/reprints