Coca Cola Zero

Journal of Chromatography A, 1218 (2011) 29762983

Contents lists available at ScienceDirect

Journal of Chromatography A

journa l homepage: www.e lsev ier .com

Simultaneous analysis of aspartame and its hydroby on-l ddetecti orm

Cheanyea Department o wanb Research Cen d, Chu

a r t i c l

Article history:Received 14 October 2010Received in revised form 1 February 2011Accepted 15 March 2011Available online 8 April 2011

Keywords:Two-dimensiochromatograpReversed-phasLigand-exchanOn-line postcouorescence dUltraviolet detThermal hydroCoca-Cola ZeroAspartameAmino acid en

An innovative two-dimensional high-performance liquid chromatography system was developed for thesimultaneous analysis of aspartame and its hydrolysis products of Coca-Cola Zero. A C8 reversed-phasechromatographic columnwithultraviolet detectionwasused as therst dimension for the determinationof aspartame, and a ligand-exchange chromatographic column with on-line postcolumn derivation u-

1. Introdu

The artimethyl esteeither tablchewing-gumix aspartaand antioxiics [1]. Theindustry. Mcoping with

Corresponversity, 200 Ch+886 3 265339

E-mail add

0021-9673/$ doi:10.1016/j.nal high-performance liquidhye chromatographyge chromatographylumn derivationetectionectionlysis

antiomer

orescence detection was employed as the second dimension for the analysis of amino acid enantiomers.The uorimetric derivative reagent of amino acid enantiomers was o-phthaldialdehyde. The hydroly-sis of aspartame in Coca-Cola Zero was induced by electric-heating or microwave heating. Aspartamewas quantied by the matrix matched external standard calibration curve with a linear concentrationrange of 050gmL1 (r2 = 0.9984). The limit of detection (LOD) and the limit of quantication (LOQ)were 1.3gmL1 and 4.3gmL1, respectively. The amino acid enantiomers was analyzed by the matrixmatched internal standard calibration method (d-leucine as the internal standard) with a linear concen-tration range of 010gmL1 (r2 = 0.99880.9997). The LODs and LOQs for l- and d-aspartic acid and l-and d-phenylalaninewere 0.160.17gmL1 and 0.520.55gmL1, respectively, thatwas 1213 timesmore sensitive than ultraviolet detection. The overall analysis accuracy for aspartame and amino acidenantiomers was 90.299.2% and 90.496.2%, respectively. The overall analysis precision for aspartameand amino acid enantiomers was 0.11.7% and 0.56.7%, respectively. Generally, the extent of aspartamehydrolysis increases with the increase of electro-thermal temperature, microwave power, and the dura-tion of hydrolysis time. d-aspartic acid and d-phenylalanine can be observed with the electro-thermalracemization at the hydrolysis temperature 120 C for 1 day and only d-aspartic acid can be observed atthe hydrolysis temperature 90 C for 2 and 3days. For the microwave induced hydrolysis, only l-asparticacid was detected at the power 560W for 1min and 320W for 3min.

2011 Elsevier B.V. All rights reserved.

ction

cial sweetener, aspartame (l-aspartyl-l-phenylalaniner) has long been commonly used as a substitute for

e sugar or sugar in low-calorie meals, soft drinks,ms, and frozen deserts. Recently, there is a trend tome with other kinds of additive such as preservativesdants to yield synergistic effects in food and cosmet-refore, aspartame is important in food and cosmeticsedicinally, the design of aspartame is partially due todiseases associated with the consumption of carbohy-

ding author at: Department of Chemistry, Chung Yuan Christian Uni-ung Pei Road, Chungli 32023, Taiwan, ROC. Tel.: +886 3 2653322; fax:9.ress: [email protected] (C. Cheng).

drates. In anatomy, aspartame has been used as amodel compoundto identify and recognize the active site of the sweet taste receptor[2].

However, aspartame ishydrolyzednaturallyunder the inuenceof temperature andpH to generate l-aspartic acid, l-phenylalanine,l-phenylalanine methyl ester, l-aspartyl-phenylalanine, etc. [3,4]The hydrolysis products, l-aspartic acid and l-phenylalanine, mayfurther undergo racemization to form d-aspartic acid and d-phenylalanine by high temperature [5,6], pH [6], and enzyme [7].L-form amino acids are nutrients for organisms and occur nat-urally while d-form amino acids are usually few and useless innature. However, in recent years some d-amino acids have beenfound in mammals including humans [8,9] that play importantbiological functions [1013]. Nevertheless, there are inborn hered-itary diseases related with amino acids such as phenylketonuria. Inphenylketonuria, phenylalanine is abnormallymetabolized to toxicphenylpyruvic acid due to the absence of phenylalanine hydrox-

see front matter 2011 Elsevier B.V. All rights reserved.chroma.2011.03.033ine postcolumn derivation uorescenceon coupled two-dimensional high-perf

h Chenga,b,, Shing-Chen Wua

f Chemistry, Chung Yuan Christian University, 200 Chung Pei Road, Chungli 32023, Taiter for Analysis and Identication, Chung Yuan Christian University, 200 Chung Pei Roa

e i n f o a b s t r a c t/ locate /chroma

lysis products of Coca-Cola Zeroetection and ultravioletance liquid chromatography

ngli 32023, Taiwan

C. Cheng, S.-C. Wu / J. Chromatogr. A 1218 (2011) 29762983 2977

ylase by gene defect. The accumulation of phenylpyruvic acid inchild causes mental retardation. It seems that the only effect ofaspartame hydrolysis would be on the patient of phenylketonuria.

Analytical methods [1416] including high-performance liquidchromatogrtive analysidetectionmamino acidacid enantichromatogrwith stationfunctional gthe most ppool of amiof amino acfound difc

Amino areversed-phsensitive ureagents [2acid enantiraphy. In orrange of ancated matrion-line coluphase columsolve this pwas traditioSince the leamino acidsensitive utiomers ana

The purHPLC systeits thermalenantiomeruorescentoping an ocoupled LEenantiomer

2. Experim

2.1. Chemic

The ena99%), l-ph98%), d-leu(99.5%), 2-(St. Louis,from Alfa AphenylalanInc. (OH, UHPLC gradeMerck (Darminum cansupermarketapwaterbyBarnstead u

2.2. Standa

One thotions of amwater. Then

to the desired concentration for use. The sample of Coca-Cola Zerowas degassed in ultrasonic bath for 10min and the degassed sam-ple (pH 2.12) was ltered through a 0.2m pore-size membranelter and stored in the refrigerator for later use.

ino

amidissecahM. Torat

ssiblegent

ectro

thered

n colsiusainlee. Onlls of, 90,mplewas

nwamplealyze

icrow

olum25mon-aThes (32ysisl toto

alysn cal

5mLetricrentresprd a2mMsparttivelamePLC s

EcliGerm4.6mat 5

, v/v. Thhe xaphy (HPLC) [17,18] have been reported for quantita-s of amino acid. However, HPLC coupled with differentethods [1929] has beenwidely used for the analysis ofs. Particularly, chiral HPLC [3034] can separate aminoomers most efciently. Within the category of chiralaphy, ligand-exchange chromatography (LEC) [3335]ary phase containing single amino acid enantiomericroup to complex the metal ions has been shown to beowerful method for the simultaneous separation of ano acid enantiomers. While the simultaneous analysisid enantiomers and dipeptide aspartame with LEC wasult because of the long retention time of aspartame.cids and peptides can be separated very easily by thease chromatography [34] and canbedetectedbyhighlyorescence detector with various uorescent derivative3,24,27,29,3640]. However, the separation of aminoomers is difcult with the reversed-phase chromatog-der to resolve completely and simultaneously a widealytes containing amino acid enantiomers in compli-x, two-dimensional HPLC (2-D HPLC) [4144] that usesmn-switching technique [45] to combine a reversed-n and a ligand-exchange column is found usefully to

roblem. In the literature, the less sensitive UV detectornally used for LEC to analyze amino acid enantiomers.ss sensitive UV detector cannot be applied for traceenantiomer analysis, it is urgent to develop a highlyorescence detection coupled LEC for amino acid enan-lysis.pose of this work aims the development of a 2-Dm for simultaneous determination of aspartame andhydrolysis and racemization products, amino acid

s, in Coca-Cola Zero and the use of the most commonderivative reagent o-phthaldialdehyde (OPA) for devel-n-line postcolumn derivation uorescence detectionC to improve the detection sensitivity for amino acids.

ental

als and materials

ntiomers of amino acid, d-aspartic acid (d-Asp,enylalanine (l-Phe, 99%), d-phenylalanine (d-Phe,cine (d-Leu, 99%), sodium tetraborate decahydratemercaptoethanol (99%) were purchased from SigmaUSA). o-Phthaldialdehyde (OPA, 98%) was purchasedesar (Karlsruhe, Germany). Aspartame (l-aspartyl-l-

ine methyl ester) was obtained from ICN BiochemicalsSA). l-Aspartic acid (l-Asp, 99%), methanol (MeOH) of, copper (II) sulfate (CuSO4, 98%) were all bought frommstadt, Germany). Plastic bottled (600mL) and alu-ned (330mL) Coca-Cola Zero were obtained from localt. Ultrapure water (18.2M) was prepared rst fromreverseosmosis anddistillationonce then treatedwithltrapure system (Dubuque, IA, USA).

rds and sample preparation

usand microgram per milliliter each of standard solu-ino acids were prepared by dissolving CuSO4 in pure, they were stored in the refrigerator or further diluted

2.3. Am

Theday byorate dof 16mof the bthe pothe rea

2.4. El

Theperformthat caof a Cewith stvolumthe we(37, 60was cosamplesolutioThe saand an

2.5. M

A vinto awith noven.powerhydrolto coodilutedHPLC.

2.6. Anadditio

Sixvolumof diffeaddedstandausingdard arespecaspart2-D HZorbaxbronn,(12.5tained(80/20254nmfrom t[46].acid derivation reagent

no acid derivation reagent was prepared freshly eacholving 2.2mg of OPA in 1.0 L of 0.01M sodium tetrab-ydrate buffer (pH=9.5) to make an OPA concentrationhen, 1.0mL of 2-mercaptoethanol was added. The pHe buffer was adjusted by 5mM NaOH solution. To avoiddecay of the derivation reagent by irradiation of light,was kept in a dark bottle.

-thermal hydrolysis of Coca-Cola Zero

mal hydrolysis of pretreated Coca-Cola Zero sample waswith a specially designed electric-heatingmicro-reactorntrol the hydrolysis temperature precisely to one tenthdegree. Theheating-block typemicro-reactorwasmadess steel that contains 20 screw cappedwells of 1.5mL ine milliliter Coca-Cola Zero sample was placed in one ofthe micro-reactor and heated to a specic temperatureand 120 C) for a period of 15days. After hydrolysisted the reactor was cooled to room temperature. Thepipetted, the volume was measured, and 2mM CuSO4s added tomake the sample volume to theoriginal 1mL.was further diluted 5 folds with 2mM CuSO4 (pH 2.86)d by the 2-D HPLC.

ave hydrolysis of Coca-Cola Zero

e of 5mL pretreated Coca-Cola Zero sample was placedL volumetric ask. The volumetric ask was sealeddsorptive cotton and put in a home-used microwavecola sample was hydrolyzed at different microwave0, 560, and 800W) for a period of 1 or 3min. Afterwas completed, the volumetric ask was allowedroom temperature. Then 2mM CuSO4 solution wasthe mark (a 5-fold dilution) and analyzed by the 2-D

is of aspartame in Coca-Cola Zero by standardibration method

of pretreated cola sampleswere put in respective 25mLasks. Six 1000gmL1 aspartame standard solutionsvolumes, 0.0, 0.25, 0.50, 0.75, 1.0, and 1.25mL, wereectively to the six cola samples. Each of aspartamedded cola sample solutions was diluted to the mark

CuSO4 solution. The dilution makes the nal stan-ame concentration to 0, 10, 20, 30, 40, and 50gmL1

y in each of the six cola samples. The six 5-fold dilutedstandard added cola samples were analyzed with theystem by a reversed-phase chromatographic columnpse XDB-C8 (1504.6mm i.d., 5m, Agilent, Wald-any). The guard column was Zorbax Eclipse XDB-C8m i.d., 5m). The column temperature was main-

0 C. The mobile phase was 2mM CuSO4/methanol) at 1.00mL/min. The UV detection wavelength wase aspartame content in the cola sample was found-intercept of the standard addition calibration curve

2978 C. Cheng, S.-C. Wu / J. Chromatogr. A 1218 (2011) 29762983

Fig. 1. (A) Thechromatograpexchange coluligand-exchancolumn-switch

2.7. Simultaenantiomers

The 2-D(Tokyo, Japaa Shimadzuternary puuorescenca Micro-tecof the derivand a Rheo(Cotati, USAaration of ai.d., 5m, AZorbax EcliChirex 3126with a gua(Torrance,aration. Thcolumn we(RheodyneThe chromareversed-phScientic Inand storedchromatogrligand-exchorescent dan AMD-40The detailetaneous deenantiomerFig. 1A.

The opecola samplevalve was isoon after ttion II to cuaspartame t

as all of the amino acid enantiomers were delivered to the LEC col-umn, the switching valvewas switchedback toposition I; at stage3,aspartame in the cola sample was separated by the reversed-phasecolumn and detected by UV detector, the chromatographic run

en stratiowas

ing ts sumalytih col1.0m

alysl cal

ernalaspaith thcolae prurac, 5.0g 5-f theitheachpea

of earve wre.

alysd int

quahe thmatl stationacidylalasoluschematic diagram of the two-dimensional high-performance liquidhy system. Position I: the reversed-phase column and the ligand-mn in parallel, position II: the reversed-phase column and thege column in series. (B) The time and duration for operating theing at different stage and corresponding valve position.

neous analysis of aspartame and amino acidby 2-D HPLC system

HPLC system consisted of a Shimadzu LC-9A pumpn), a ShimadzuSPD-6AUVdetectoroperatedat254nm,CTO-6A column oven, an Aglient 1100 series qua-

mp (Waldbronn, Germany), an Aglient 1100 seriese detector operated at ex of 340nmand em of 450nm,h ultra-plus II pump (CA, USA) as the on-line deliveryation reagent, a 1 m long post-column derivation loop,dyne 7725i injection valve with a 20L sampling loop). The analytical column for the reversed-phase sep-spartame was Zorbax Eclipse XDB-C8 (1504.6mmgilent, Waldbronn, Germany). The guard column was

pse XDB-C8 (12.5mm4.6mm i.d.). The LEC column,d-penicillamine column (250mm4.6mm, i.d., 5m)

was thand dusystemaccordandwatwoanfor botand at

2.8. Anexterna

Extysis ofZero wof themust bsis acc1.0, 2.5by usinEach otimes warea ofgroundsignaltion cusoftwa

2.9. Anmatche

Thefrom tvia theinternaan eluaminod-phenCuSOrd column (30mm4.6mm, i.d.) from PhenomenexUSA), was used for the amino acid enantiomer sep-e reversed-phase column and the ligand-exchangere coupled parallel by a six-port switching valve7000, Cotati, USA) to form the 2-D HPLC system.tographic data processing software for analyzing thease separation of UV detection was obtained fromformation Service Corporation (SISC, Taipei, Taiwan)in a P-III 500 computer (Acer, Taipei, Taiwan). Theaphic data processing software (ChemStation) forange separation of amino acid enantiomers and u-etection was obtained from Agilent and stored in0 computer (Gigabyte Technology, Taipei, Taiwan).d conguration of the 2-D HPLC system for simul-tection and analysis of aspartame and amino acids with UV and uorescence detector is shown in

ration procedures for analysis of thermally hydrolyzedwere as follows: for stage 1, as the column switching

n position I, the cola sample was loaded and injected;he injection, the switching valve was turned to posi-t and deliver amino acids produced by the hydrolysis ofo the LEC column for separation anddetection (stage2);

4the internations for the20.0gmL

2.5, 5.0, andtration of instandard soenantiomerD HPLC syspeak area rinternal station curvesoftware.

3. Results

3.1. Selectioderivation r

The sepaexchange cand uses aion can forthe stationaopped for the next sample injection. The optimal timen for operating the column switching of the 2-D HPLCfound by using the statistical ANOVA procedure [47]

o the variation of peak areas of amino acid enantiomersmarized inFig. 1B.During thechromatographic run the

cal columnsweremaintained at 50 C. Themobile phaseumnswas the same (2mMCuSO4/methanol, 80:20, v/v)Lmin1.

is of aspartame with 2-D HPLC by matrix matchedibration method

calibration method was used for the quantitative anal-rtame in the sample of thermally hydrolyzed Coca-Colae 2-D HPLC system. Owing to the complexity of matrixsample, the external calibration curve for aspartameepared by matching the matrix to assure the analy-y. Therefore, standard aspartame concentrations of 0,, 10.0, 20.0, 30.0, 40.0, and 50.0gmL1 were preparedfold diluted Coca-Cola Zero with 2mM CuSO4 diluent.standard aspartame solutions was then analyzed 4

the 2-D HPLC system of UV detection. The average peakof the standard was corrected with the average back-

k area (signal of 0gmL1 standard). The correctedch of the standard was then used to obtain the calibra-ith the linear least-square regression method of Excel

is of amino acid enantiomer with 2-D HPLC by matrixernal standard calibration method

ntitative analysis of amino acid enantiomers obtainedermal hydrolysis of aspartame in Coca-Cola Zero wasrix matched internal standard calibration method. Thendard used for the analysis was d-leucine which hasretention time roughly central for the four possibleenatiomer products, l- and d-aspartic acid and l- andnine. The cola sample was diluted 5-fold with 2mMtion and was used as the matrix in the preparation ofl standard calibration curves. The standard concentra-four amino acid enantiomers were 0, 2.5, 5.0, 10.0, and1 for ultraviolet detection and 0, 0.25, 0.5, 0.75, 1.25,10.0gmL1 for uorescence detection. The concen-ternal standard in each of the amino acid enantiomerlutions was xed at 4gmL1. Each of the amino acidstandard solutions was analyzed 4 times with the 2-

tem of either UV or uorescent detection. The averageatio of each of the four amino acid enantiomers to thendard was used to obtain the internal standard calibra-by the linear least-square regression method of Excel

and discussion

n of mobile-phase and concentration of postcolumneagent for 2-D HPLC system

ration of amino acid enantiomerswas through a ligand-olumn that contains d-penicillamine stationary phasemobile phase of aqueous Cu2+ ion solution. The Cu2+

m complexes with amino acid enantiomers both onry phase and in the mobile phase. The ligand (amino


Fig. 2. The var t concentration of uorescence derivative reagent OPA: (A) l-aspartic acid,(B) d-aspartic

acid enantiand the mocomplexesenantiomeramino acidenantiomertion for theand l- and dCu2+ ion asi.e. over 20methanol, ction ofmeththe stationareasonabletion of 2mMthis kind oftiomer prodthe choicereversed-physis producone kind oseparation

In orderpostcolumnCu2+ ion coconcentratithe complein the comcentrationCu2+ ion inbe observedtion of OPAis shown inwas small aOPA was inOPA was uamino aciddetector.

ptimal column-switching times for 2-D HPLC system

3 illumplminoL1

tectolarquiclumnC colimeing tiation of the uorescence detection peak area for amino acid enantiomer at differenacid, (C) l-phenylalanine, and (D) d-phenylalanine.

omers) can exchange between the stationary phasebile phase. However, due to different stability of theformed with different amino acids and their l- and d-s, the ligand exchange rate was different for differentenantiomers, thus, a separation of different amino acids can be achieved for analysis. Although a best separa-four amino acid enantiomers (l- and d-aspartic acid-phenylalanine) can be obtained with the use of 2mMthe mobile phase, the retention time was too long,

0min for phenylalanine. The use of organic modier,an decrease the retention time, however, the composi-anol is limited to 30%by volume so as not to deterioratery phase. In order to protect the column and obtain aretention time for analysis, a mobile phase composi-

3.2. O

Fig.cola safour a4gmwas deSince pelutedtial cothe LEthree tswitchCu2+ ionmethanol (80:20, v/v) was employed. Withmobile phase composition, the four amino acid enan-ucts can be well separated within 70min. In addition,of this kind of mobile phase is also suitable for thease separation of aspartame and its thermal hydrol-ts (amino acids) in cola sample. Therefore, there is onlyf mobile phase applied for the two different kinds ofcolumn of the 2-D HPLC system.to make successful uorescence detection with theuorescent derivation of the amino acid enantiomer-mplexes after the LEC separation, we found that theon of the OPA must be high enough to compete withxing agent Cu2+ ion for the amino acid enantiomersplexes. Experimental results showed that as the con-of OPA is 4 times larger than the concentration of thethe mobile phase, an obvious uorescence signal can. The uorescence signal increases as the concentra-increases for the four amino acid enantiomers whichFig. 2. However, the increase of the uorescence signals the concentrationof theuorescentderivation reagentcreased to a concentration of 16mM. Therefore, 16mMsed for the postcolumn uorescent derivation of theenantiomers to produce a highly sensitive uorescent

Fig. 3. The revthe 2-D HPLCl-aspartic acidwithout any sature: 50 C; mrate: 1mLminAsp= l-aspartl-Phe= l-phenstrates the reversed-phase separation for 5-fold dilutede spiked with 20gmL1 aspartame and each of the

acid enantiomers (l-Asp, d-Asp, l-Phe, d-Phe) andd-leucine (the internal standard). The chromatogramed with UV detector at the switching-valve position I.l- and d-aspartic acid (k =0.486 and N=421plates) arekly out from the reversed-phase column, the best ini--switching time used to transfer the aspartic acid toumn with the switching-valve position II was tested atpoints (1.1, 1.2, and 1.3min) and the second column-ime was xed at 5.8min. The second column-switchingersed-phase chromatography of the 5-fold diluted cola sample withat switching valve position I: (a) with standard addition of d- and, d- and l-phenylanaline, d-leucine, and aspartame; (b) plain sampletandard addition. Column: Zorbax Eclipse XDB-C8; column temper-obile phase: 2mM copper(II) sulfatemethanol (80:20, v/v); ow

1; ultraviolet detector at =254nm. Peak: Apm=aspartame, l-ic acid, d-Asp=d-aspartic acid. d-Leu=d-leucine, Med= cola medium,ylalanine, d-Phe=d-phenylalanine.


time makes the switching-valve back to position I and allows l-and d-phenylalanine (k =2.52 and N=768 plates) be transferredto the LEC column but the rest of the components of cola sampleincluding aspartame (k =12.72 and N=1982 plates) remained inthe reversed-phase column for separation. Comparison of averagepeak areas and their standard deviations for the four amino acidenantiomers with statistical ANOVA procedure [47] at the threetested column-switching times indicates that the initial optimaltime for changing the valve position I to position II was 1.2min.

Similarly, after xing the initial column-switching time at1.2min, the optimal second column-switching time that switchesthe valve position II back to position I and allows aspartameremained in the reversed-phase column for separation was testedat three different times (5.7, 5.8, and 6.0min). The results testedwith ANOVA procedure showed that the second optimal column-switching time was 5.8min. The column-switching time and itscorrespond

3.3. LOD an

The mawas used fohydrolyzedchromatogrthematrixmrespondingThe standar050gmLtication (Lcurve of ve10.0gmL

The quafrom the thwas mademethod. Thtion uoresamino acidinternal staof standarddetection aThe concen4gmL1.bration curand the LOthe sensitivbetter thanity was smawith uoresthat was inProbably, thLEC shouldAlthough thtion couple

Fig. 4. The reversed-phase chromatogram of the residue aspartame in thermaled Coge: (ay, anddescr

vityd-phhat t

alysmicr

ore tt ofrd adnd Us (diero2.2wasrtama-Coltledme.4 illme

ramse ofof h

sultsomeriodsth an electro-thermal temperature 120 C, however, this sit-did not happen at lower temperatures such as 90, 60, and

5 illustrates the production of amino acid enantiomers fromdrolysis of aspartame for four different temperatures with

Table 1The linear equ calibration curve of aspartame and the matrix matched internal standardcalibration cur and limit of quantication.

Analytes Fluorescence detector

Linear equation r2 LOD(gmL1)

LOQ(gmL1)

Aspartamel-Aspartic ac y=0.4826x0.0230 0.9997 0.2 0.6d-Aspartic a y=0.5021x0.0466 0.9994 0.2 0.6l-Phenylalan y=0.4357x0.0051 0.9990 0.2 0.5d-Phenylala y=0.4301x0.0153 0.9988 0.2 0.5ing valve position are summarized in Fig. 1B.

d LOQ for aspartame and amino acid enantiomer

trix matched external standard calibration methodr the quantitative analysis of aspartame in thermallycola sample which is separated by the reversed-phaseaphy with ultraviolet detection. The linear equation foratched external standard calibration curve and its cor-linear correlation coefcient (r2) are listed in Table 1.d calibration curve had a linear concentration range of1. The limit of detection (LOD) and the limit of quan-OQ) found for aspartame from the standard calibrationlowest concentrations of standard (0, 1.0, 2.5, 5.0, and1) [47]were1.3gmL1 and4.3gmL1, respectively.

ntitative analysis of amino acid enantiomers producedermal hydrolysis of aspartame in cola sample with LECby the matrix matched internal standard calibrationedetectionmethodwaseitherpost-columnOPAderiva-cence detection or UV detection. The LODs and LOQs ofenantiomers were estimated from the matrix matchedndard calibration curve by ve lowest concentrations, 0. 0.25, 0.5, 0.75, and 1.25gmL1, for uorescencend 0, 2.5, 5.0, 10.0, and 20.0gmL1, for UV detection.tration of the internal standard, l-leucine, was kept atThe linear equation of the four internal standard cali-ves, their corresponding linear correlation coefcients,Ds and LOQs were listed in Table 1. It is obvious thatity of LEC with uorescent detection was 1213 timesthat of the UV detection. The improvement of sensitiv-ll as compared to the sensitivity of amino acid analysiscent detection by the reversed-phase chromatographygeneral about 103 folds better than the UV detection.e quenching effect of Cu2+ ion in the mobile phase of

be mostly counted for the aforementioned observation.e enhancement of sensitivity for uorescence detec-d LEC was not as large as we expect for the common

hydrolyzthird stafor 5-dasame as

sensitireverseprove tto LEC.

3.4. Anheated

Befcontenstandaumn abatcheCola Z89.9sampleof aspaby Coctic botasparta

Fig.aspartamatogincreasdegreeysis reenantient perday wiuation37 C.

Fig.the hy

ation and linear correlation coefcients of the matrix matched external standardves of the four amino acid enantiomers and their corresponding limit of detection

UV detector

Linear equation r2 LOD(gmL1)

LOQ(gmL1)

y=7981.6x3401.5 0.9984 1.3 4.3id y=0.2151x0.1566 0.9915 2.5 8.4cid y=0.2021x0.1214 0.9915 2.5 8.5ine y=0.2052x0.1525 0.9911 2.6 8.6

nine y=0.2007x0.1528 0.9912 2.6 8.6ca-Cola Zero with the 2-D HPLC at switching valve position I of the) 120 C for one-day reaction, (b) 90 C for 3-day reaction, (c) 60 C(d) 30 C for 5-day reaction. The chromatographic conditions are theibe in Fig. 3.

of uorescent detection for amino acid analysis withase chromatography, this is the rst time technically tohe more sensitive uorescent detection can be coupled

is of aspartame thermal hydrolysis with electricallyo-reactor by 2-D HPLC system

he thermal hydrolysis of cola sample, the originalaspartame in cola sample was determined by thedition calibration method using reversed-phase col-V detection. The analytical results for four differentfferent manufacturing date) of plastic bottled Coca-sample were 79.63.7, 117.61.6, 81.53.5, andgmL1 and for one aluminum canned Coca-Cola Zero86.61.9gmL1. The results show that the contente in cola sample maybe adjusted for regions or timesa Company. For the following study, we used the plas-Coca-Cola Zero sample containing 117.61.6gmL1

ustrates the result of the electro-thermal hydrolysis ofin cola sample with the 2-D HPLC system. These chro-show that the extent of hydrolysis increases with antemperature while the time needed to get a certainydrolysis is shorter. Table 2 shows the detailed anal-for thermal hydrolysis of aspartame and its amino acidproducts at different reaction temperatures and differ-. Aspartame was almost entirely hydrolyzed in just one


Table 2The quantitative analysis results for the electric heating thermal hydrolysis of aspartame and the thermal racemization of its amino acid enantiomeric products in Coca-ColaZero.

Reactiontemperature(C)

Reactiontime (day)

l-Aspa d-Aspa l-Phea d-Phea Aspartamea Asp Phe

Conc.(gmL1)

RSD(%)

Conc.(gmL1)

RSD(%)

Conc.(gmL1)

RSD(%)

Conc.(gmL1)

RSD(%)

Conc.(gmL1)

RSD(%)

Yield(%)

Yield(%)

120b 1 2.7 0.1 3.7 9.3 0.2 2.2 2.2 0.1 4.5 2.6 14.0120 1 14.1 0.1 0.7 8.6 0.1 1.2 22.8 0.2 0.9 2.2 0.1 4.5 19.3 21.390 1 17.7 0.1 0.6 7.3 0.1 1.4 20.6 0.2 1.0 15.1 6.290 2 21.9 0.1 0.5 3.2 0.1 3.1 11.1 0.2 1.8 14.0 0.1 0.7 21.3 9.490 3 17.6 0.1 0.6 4.0 0.1 2.5 8.4 0.1 1.2 6.0 0.1 1.7 18.4 7.160 1 3.1 0.1 3.2 81.5 0.5 0.6 2.6 60 2 4.3 0.2 4.7 69.2 0.6 0.9 2.8 60 3 5.6 0.1 1.8 2.2 0.1 4.5 58.9 0.2 0.3 2.5 1.860 4 7.7 0.1 1.3 2.9 0.1 3.4 44.5 0.1 0.2 6.5 2.560 5 10.9 0.2 1.8 2.9 0.1 3.4 37.0 0.2 0.5 9.3 2.537 3 1.5 0.1 6.7 99.3 0.1 0.1 1.3 37 4 2.0 0.1 5.0 95.0 0.1 0.1 1.7 37 5 2.5 0.1 4.0 90.3 0.1 0.1 2.1 a Number of measurement =4 (n=4).b Reaction of 100gmL1 standard aspartame aqueous solution.

different reaction periods. The enantio-separation of LEC for l- andd-aspartic acid and l- andd-phenylalanine is efcient. The enantio-selectivity () and the resolution (RS) between l- andd-aspartic andbetween l- and d-phenylalanine calculated from chromatograma were 1.33 and 1.51 and 1.37 and 4.16, respectively. Resultsin Table 2 show that there were obviously more d-aspartic acid(k =1.78 andN=1156) andd-phenylalanine (k =6.40 andN=2869)produced at the highest temperature of 120 C than those low tem-peratures. This phenomenon indicates the energy needed for theracemization of l- and d-amino acid enantiomer is larger than thehydrolysis of aspartame. In addition, the thermal degradation of

amino acids or reactions of amino acid with other substances suchas caramel by Maillard reaction [48] may occur in the cola sample.These side-reactions are also accounted for the lower yields of theracemization. Correspondingly, thermal hydrolysis of 100gmL1

standard aspartame in aqueous solution was performed that indi-cates l-phenylalanine (k =4.67 and N=6623) is more stable thanl-aspartic acid (k =1.33 and N=1244) for thermal degradation andthe thermal racemization of l-aspartic acid is more easily per-formed than l-phenylalanine. It seems that the complexmediumofcola sample may protect the hydrolysis product, amino acid, fromthermal degradation.

Fig. 5. (A) Thefor one day reenantiomer prcolumn tempel-Asp= l-aspaLEC chromatogram of the 2-D HPLC for the amino acid enantiomer products after thermaction, (b) hydrolysis at 90 C for 3-day reaction, and (c) hydrolysis at 60 C for 5-day roducts after thermal hydrolysis of aspartame in Coca-Cola Zero performed at 37 C and vrature: 50 C, mobile phae: 2mM copper(II) sulfatemethanol (80:20, v/v), ow rate: 1mrtic acid, d-Asp=d-asparticacid, d-Leu=d-leucine (internal standard), l-Phe= l-phenylalaal hydrolysis of aspartame in Coca-Cola Zero: (a) hydrolysis at 120 Ceaction. (B) The LEC chromatogram of 2-D HPLC for the amino acide-day reaction period. Column: Chirex 3126 d-penicillamine column,Lmin1, uorescence detector (ex = 340nm and em =450nm). Peak:nine, d-Phe=d-phenylalanine.


Table 3The quantitative analysis results of the microwave heating thermal hydrolysis of aspartame and its hydrolysis amino acid enantiomeric products in Coca-Cola Zero.

Power (W) Reactiontime (min)

l-Aspa d-Aspa l-Phea d-Phea Aspartamea Asp Phe

Conc. RSD Conc. RSD Conc. RSD%)

Conc. RSD Conc. RSD Yield Yield

800800800b

560560b

560320

a Number ob Reaction o

Results (at 90 C anseems to reof cola samcola sampleat temperatformation o

For therat 60 C, whand l-phen35days. Ththe activatiamino acid

The therdone at a lohumanbodacid can bein Table 2 spreserved inworry aboucola drink a

The RSDshown in Tsis by UV aof amino acsis precisioenantiomerysis of aspasystem wasexperimenttion was 90acid and l-very good osystem.

Our resuthat is theincrease oferature, thethe role of cthan d-phenmedium (3for 1 day [5also showedoccurred at

3.5. Analysihydrolysis w

Nowadafoods andtain aspart

aveca-Cs a mithavesumystemsamer, iantic aci320Wmenoraturhermationrtamhenyg.berg

aspa700aspazatio0.19nt frm usultsn colforof c

yzedanal

e aspas lostratmea. Then 91od a(gmL1) (%) (gmL1) (%) (gmL1) (

3 1 1 3 3 1 1.7 0.1 5.9 3 1.9 0.1 5.2

f measurement =4 (n=4).f 80gmL1 standard aspartame aqueous solution.

Table 2) for thermal hydrolysis of aspartameperformedd reacted for 13days implies that l-phenylalanineact more easily with other substances in the mediumple. Therefore, the thermal hydrolysis of aspartame infavors the production of both l- and d-aspartic acid

ure 90 C. Also, at 90 C, which is still high enough, thefd-aspartic acid can be aidedwith long hydrolysis time.mal hydrolysis of aspartame in cola sample performedere the energy is not high enough, only l-aspartic acidylalanine can be accumulated after a reaction period ofe thermal energy at this temperature cannot overcomeon energy of thermal racemization between l- and d-for both aspartic acid and phenylalanine.mal hydrolysis of aspartame in cola sample was alsower temperature of 37 C (a simulated temperature ofy). At this temperature a very small amount of l-asparticdetected for a long reaction period of 35days. Resultshow that most of aspartame in cola sample was stillgood form. This result also means that people may not

t the occurrence of the hydrolysis of aspartame in thet the usual body temperature.values for those replicate quantitative measurementsable 2 were all lower than 1.7% for aspartame analy-nd were from 0.5% to 6.7% for uorescence detectionid enantiomers that demonstrates a very good analy-n for the determination of aspartame and amino acids with the 2-D HPLC system. The accuracy for the anal-rtame and amino acid enantiomers with the 2-D HPLCfound through the percentage recovery of the spikeds [45]. The spike recovery for aspartame with UV detec-.2100.8% and the spike recovery for l- and d-asparticand d-phenylalanine was 90.496.2% that also showsverall analysis accuracy (90.299.2%) for the 2-D HPLC

lts shown here are consistent with the literature [5,6]extent of aspartame hydrolysis increases with an

temperature and the length of time elapsed. In the lit-rate of hydrolysis can be speeded by acid which playsatalyst. There was also more d-aspartic acid producedylalanine produced by the racemization in an acidied

microwtledCoused atame wmicrowTable 3HPLC sin colaHowevacid enaspartieitherphenotemperapid tdegradof aspathan pheatin

Stensis ofunderage ofracemirangedifferemediuthe restame i800Wmatrixhydrol

Theresiduacid wdemonaspartasystembetweevery goand 6M HCl) and at a hydrolysis temperature 110 C]. Using bottled CoCa light as a sample, Chen et al. [6]that hydrolysis and racemization of aspartame can be

room temperature (25 C) and pH 2.0.

s of microwave induced aspartame thermalith 2-D HPLC system

ys there are a lot of microwave heating ready-to-eatdrinks sold in the supermarkets which may con-ame. Therefore, in order to understand the effect of

4. Conclus

A 2-D Hligand-exchchromatogrmal hydrolyis the rstand l- andon-line posanalysis sen(gmL1) (%) (gmL1) (%) (%) (%)

12.8 0.1 0.8 34.4 0.1 0.3 37.6 0.1 0.3 16.0 0.1 0.6 18.0 0.1 0.6 37.4 0.1 0.3 2.1 46.2 0.2 0.4 2.3

heating on the hydrolysis of aspartame, plastic bot-ola Zero containing 81.53.5gmL1 of aspartamewasodel sample to study the thermal hydrolysis of aspar-pulsed microwave heating at three different outputpowers, 320, 560, and 800W for either 1 or 3min.marizes the quantitative results determined by the 2-D. It is obvious that the extent of aspartame hydrolysis

ple increases as the microwave output power increases.n most of the cases, the hydrolysis products, aminoomers were not found. Only a very small amount of l-d can be detected at the microwave output power of

for 3min operation or 560W for 1min operation. Thisn could be due to the instantaneous extremely highe produced by the pulsed microwave power to trigger aal hydrolysis of aspartame and the subsequent thermalof amino acids. In contrast to the thermal hydrolysis

e with electric heating, aspartic acid seems more stablelalanine in Coca-Cola Zero matrix under the microwave

et al. [5] also performed the microwave hydroly-rtame in acidied aqueous medium (3 and 6M HCl)W or 300W for 1, 2, 3, 6, or 10min. The percent-rtame hydrolysis was in the range 82100% and then for l-aspartic acid and l-phenylalanine was in the4.83% and 0.123%, respectively. The results are quiteom ours mainly because the signicant difference ofed for the experiment. In our study, comparison offor microwave induced thermal hydrolysis of aspar-a sample and in aqueous solution performed at either1min or 560W for 3min shows that the complexola sample is able to protect aspartame from being.ysis precision of microwave induced hydrolysis for theartame and the only hydrolysis product of l-asparticwer than 0.8% and 5.9%, respectively. These resultsed a good analysis precision for the determinationnd the trace aminoacid enantiomerswith the2-DHPLCspike recovery for aspartame and l-aspartic acid was

.393.5% and 90.193.5%, respectively, that also showsnalysis accuracy for the 2-D HPLC system.ions

PLC system was developed by on-line coupling aange chromatographic column and a reversed-phaseaphic column which can be used for studying the ther-sis of dipeptide ester aspartame in Coca-Cola Zero. Thisreport on the determination of l- and d-aspartic acidd-phenylalanine using a ligand-exchange column withtcolumn OPA derivation for uorescent detection. Thesitivity of amino acid enantiomers in cola sample with


the uorescence detection coupled ligand-exchange chromatogra-phy was 1213 times better than the traditional UV detection. Theanalytical results of the electro-thermal and microwave inducedhydrolysis of aspartame in cola sample imply that the hydrolysisof aspartame in Coca-Cola Zero increases with the increase of tem-perature and the increase of hydrolysis time and there are smallextent of thermal racemization of l-amino acid to its correspond-ing d-amino acid at high temperature. At body temperature, theextent of aspartame hydrolysis is negligible. Overall, the simulta-neous analysis of aspartame and amino acid enantiomers in colasample by the 2-D HPLC system exhibits high sensitivity, accuracy,and precision.

Acknowledgements

The authors gratefully acknowledge the nancial support fromNational Science Council under the grant number NSC 97-2113-M-033-002 and the nancial support from Chung Yuan ChristianUniversity under the grant number CYCU-98-CR-CH.

References

[1] J.F. Garca-Jimnez, M.C. Valencia, L.F. Capitn-Vallvey, Anal. Chim. Acta 594(2007) 226.

[2] G. Morini, A. Bassoli, P.A. Temussi, J. Med. Chem. 48 (2005) 5520.[3] K. Hayakawa, T. Schilpp, K. Imai, T. Higuchi, O.S. Wong, J. Agric. Food Chem. 38

(1990) 1256.[4] S. Pattanaargson, C. Chuapradit, S. Srisukphonraruk, J. Food Sci. 66 (2001) 808.[5] M. Stenberg, G. Marko-Varga, R. ste, Food Chem. 74 (2001) 217.[6] T.-J. Hsieh, S. Chen, Amino Acids 33 (2007) 123.[7] C. Cheng, S.-C. Wu, J. Chromatogr. A 896 (2000) 299.[8] K. Hamase, A. Morikawa, K. Zaitsu, J. Chromatogr. B 781 (2002) 73.[9] Y. Nagata, M. Higashi, Y. Ishii, H. Sano, M. Tanigawa, K. Nagata, K. Noguchi, M.

Urade, Life Sci. 78 (2006) 1677.[10] H. Wolosker, E. Dumin, L. Balan, V.N. Foltyn, FEBS J. 275 (2008) 3514.[11] A. DAnie[12] G.DAnie

ogy 141 ([13] T. Furuch[14] D. Kazmi

(2004) 17

[15] M. Krizman, I. Virant-Klun, M. Prosek, J. Chromatogr. B 858 (2007) 292.[16] Y. Cheng, L. Fan, H. Chen, X. Chen, Z. Hu, J. Chromatogr. A 1072 (2005) 259.[17] C.L. Ampanella, G. Crescentini, P. Avino, J. Chromatogr. A 833 (1999) 137.[18] J. Lpez-Cervantes, D.I. Snchez-Machado, J.A. Rosas-Rodrguez, J. Chromatogr.

A 1105 (2006) 106.[19] Y.V. Tcherkas, L.A. Kartsova, I.N. Krasnova, J. Chromatogr. A 913 (2001) 303.[20] S. Hou, H. He, W. Zhang, H. Xie, X. Zhang, Talanta 80 (2009) 440.[21] Z. Liu, P.E.Minkler, D. Lin, L.M. Sayre, Rapid Commun.Mass Spectrom. 18 (2004)

1059.[22] Y.-H. Chen, L.-L. Shih, S.-E. Liou, C.-C. Chen, Food Sci. Technol. Res. 9 (2003) 276.[23] Y.V. Tcherkas, A.D. Denisenko, J. Chromatogr. A 913 (2001) 309.[24] X. Zhao, Y. Suo, Chromatographia 67 (2008) 375.[25] X. Zhang, Y. He, M. Ding, J. Chromatogr. B 877 (2009) 1678.[26] P. Kubn, P.C. Hauser, J. Chromatogr. A 1128 (2006) 97.[27] A. Pappa-Louisi, P. Nikitas, P. Agraotou, A. Papageorgiou, Anal. Chim. Acta 593

(2007) 92.[28] P. Agraotou, S. Sotiropoulos, A. Pappa-Louisi, J. Sep. Sci. 32 (2009) 949.[29] D. Kutln, I. Molnr-Perl, J. Chromatogr. A 987 (2003) 311.[30] I.W. Kim, H.M. Choi, H.J. Yoon, J.H. Park, Anal. Chim. Acta 569 (2006) 151.[31] M.H. Hyun, S.C. Han, B.H. Lipshutz, Y.-J. Shin, C.J. Welch, J. Chromatogr. A 910

(2001) 359.[32] K. Kim, K. Lee, Biotechnol. Bioprocess. Eng. 5 (2000) 17.[33] V.A. Davankov, J. Chromatogr. A 1000 (2003) 891.[34] M. Zaher, I. Baussanne, C. Ravelet, S. Halder, M. Haroun, J. Fize, J.-L. Dcout, E.

Peyrin, J. Chromatogr. A 1185 (2008) 292.[35] T.J. Ward, K.D. Ward, Anal. Chem. 82 (2010) 4712.[36] M. Gilar, H. Xie, J.A. Aworski, Anal. Chem. 82 (2010) 265.[37] T. Yokoyama, A. Sakai, M. Zenki, Chromatographia 67 (2008) 535.[38] A. Jmbor, I. Molnr-Perl, J. Chromatogr. A 1216 (2009) 6218.[39] Y. Wang, X.-J. Kang, W.-H. Ge, X.-Z. Sun, J. Peng, Chromatographia 65 (2007)

527.[40] H. Zhang, I. Le Potier, C. Smadja, J. Zhang, M. Taverna, Anal. Bioanal. Chem. 386

(2006) 1387.[41] K. Hamase, A.Morikawa, T. Ohgusu,W. Lindner, K. Zaitsu, J. Chromatogr. A 1143

(2007) 105.[42] K. Hamase, Y. Miyoshi, K. Ueno, H. Han, J. Hirano, A. Morikawa, M. Mita, T.

Kaneko, W. Lindner, K. Zaitsu, J. Chromatogr. A 1217 (2010) 1056.[43] V. Guillen-Casla, M.E. Leon-Gonzalez, L.V. Perez-Arribas, L.M. Polo-Diez, Anal.

Bioanal. Chem. 397 (2010) 63.[44] T. Welsch, C. Schmidtkunz, B. Mller, F. Meier, M. Chlup, A. Khne, M. Lmmer-

hofer, W. Lindner, Anal. Bioanal. Chem. 388 (2007) 1717.[45] C. Cheng, C.-S. Chen, P.-H. Hsieh, J. Chromatogr. A 1217 (2010) 2104.[46] D.C. Harris, Quantitative Chemical Analysis, 7th ed., W. H. Freeman and Com-

y, NeMille

, Pears. Cho,.llo, Brain Res. Rev. 53 (2007) 215.llo, A. Tolino, A. DAniello, F. Errico, G.H. Fisher,M.M. Fiore, Endocrinol-2000) 3862.i, H. Homma, Biol. Pharm. Bull. 28 (2005) 1566.erczak, W. Ciesielski, R. Zakrzewski, M. Zuber, J. Chromatogr. A 10591.

pan[47] J.N.

ed.[48] I.H

431w York, USA, 2007.r, J.C. Miller, Statistics, Chemomertrics for Analytical Chemistry, 4thon Education Limited, Harlow, England, 2000.S. Lee, H.-R. Jun, H.-J. Roh, Y.-S. Kim, Food Sci. Biotechnol. 19 (2010)

Simultaneous analysis of aspartame and its hydrolysis products of Coca-Cola Zero by on-line postcolumn derivation fluoresc...IntroductionExperimentalChemicals and materialsStandards and sample preparationAmino acid derivation reagentElectro-thermal hydrolysis of Coca-Cola ZeroMicrowave hydrolysis of Coca-Cola ZeroAnalysis of aspartame in Coca-Cola Zero by standard addition calibration methodSimultaneous analysis of aspartame and amino acid enantiomers by 2-D HPLC systemAnalysis of aspartame with 2-D HPLC by matrix matched external calibration methodAnalysis of amino acid enantiomer with 2-D HPLC by matrix matched internal standard calibration method

Results and discussionSelection of mobile-phase and concentration of postcolumn derivation reagent for 2-D HPLC systemOptimal column-switching times for 2-D HPLC systemLOD and LOQ for aspartame and amino acid enantiomerAnalysis of aspartame thermal hydrolysis with electrically heated micro-reactor by 2-D HPLC systemAnalysis of microwave induced aspartame thermal hydrolysis with 2-D HPLC system

ConclusionsAcknowledgementsReferences

Documents

Coca Cola Zero