SEARCH

SEARCH BY CITATION

Keywords:

  • aspartame;
  • buffer;
  • chemical stability;
  • kinetics;
  • pH

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusions
  7. References

ABSTRACT:  The kinetics of an acid-base catalyzed reaction, aspartame degradation, were examined as affected by the changes in pH and pKa values caused by adding polyols (sucrose, glycerol) to phosphate buffer. Sucrose-containing phosphate buffer solutions had a lower pH than that of phosphate buffer alone, which contributed, in part, to reduced aspartame reactivity. A kinetic model was introduced for aspartame degradation that encompassed pH and buffer salt concentrations, both of which change with a shift in the apparent pKa value. Aspartame degradation rate constants in sucrose-containing solutions were successfully predicted using this model when corrections (that is, lower pH, lower apparent pKa value, buffer dilution from the polyol) were applied. The change in buffer properties (pH, pKa) from adding sucrose to phosphate buffer does impact food chemical stability. These effects can be successfully incorporated into predictive kinetic models. Therefore, pH and pKa changes from adding polyols to buffer should be considered during food product development.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusions
  7. References

Physical, chemical, and microbial changes all contribute to food deterioration and shelf life loss. Chemical reactivity, in particular, promotes adverse flavor changes, discoloration, and nutrient loss. It is therefore imperative to understand how to control detrimental chemical reactions so that product quality and shelf life can be improved.

The rates of chemical reactions leading to a loss of shelf life depend on numerous parameters such as temperature, water activity, pH, product composition, ingredient interactions, oxygen, and light exposure. Of these parameters, the pH of the food is frequently controlled and monitored in an attempt to limit chemical reactivity. Some examples of chemical reactions influenced by pH include aspartame degradation (Homler 1984; Prudel and others 1986; Bell and Labuza 1991a), ascorbic acid and thiamin degradation (Dwivedi and Arnold 1972; Connors and others 1986; Mauri and others 1992), and sucrose hydrolysis (Kelly and Brown 1978). Another important pH-dependent reaction is the Maillard reaction, which affects flavor, color, and nutritional quality of foods (Friedman 1996; Martins and others 2001). A change or shift in pH would affect the rates of these reactions as well as the shelf life of the product. Therefore, potential changes to the pH of food products should be recognized, and the reasons for pH changes identified.

Food systems are composed of multiple constituents that serve specific functions in the product, many of which directly or indirectly affect the system pH. Citric and phosphoric acids, as well as their salts, are commonly used in foods to control pH and to modify tartness. Ionic compounds, such as sodium chloride, can have a profound effect on pH by interacting with charged buffer species as well as changing the dielectric properties of the solvent (Perrin and Dempsey 1974; Wescott 1978; Galster 1991; McMillan 1994). Polyols, such as sucrose and glycerol, are commonly used as food additives due to their water-binding and humectant properties (Sloan and Labuza 1975; Lindsay 2008). Because of their nonionic characteristics, sucrose and glycerol are generally not expected to influence the pH of the system. However, Bell and Labuza (1992) showed that various sugar types did affect the pH of buffer solutions. For example, the pH of phosphate buffer solutions decreased upon the dissolution of sugars, the magnitude of which increased as the sugar concentration increased (Bell and Labuza 1992). Chuy and Bell (2006) recently reported that dissolving 2 molal sucrose into phosphate buffer caused its titration curve to display a downward shift, indicating a decrease in the apparent pKa value of the buffer in the presence of sucrose. A shift in the apparent pKa would change the ratio of the various buffering species as well as the hydrogen ion concentration (that is, pH), as described by the Henderson–Hasselbach equation. This unanticipated shift in pH and buffer species ratio from adding seemingly inert nonionic polyols may alter the rates of acid-base catalyzed reactions and consequently impact food stability.

The literature has not presented data on whether buffer pKa and pH changes associated with the addition of polyols affect the kinetics of food chemical reactions. Therefore, the objective of this study was to elucidate the extent to which these polyol-induced pKa and pH changes affect chemical stability. Knowledge about buffer–polyol interactions resulting in pH changes could help improve the formulation and shelf-life prediction of related food and beverage products.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusions
  7. References

Aspartame kinetics in 0.1 M phosphate buffer solutions with and without added polyols

Sodium phosphate buffer solutions (0.1 M) were prepared at pH values of 4.8, 5.0, 5.2, 5.4, 5.6, and 5.8 by mixing appropriate volumes of bulk 0.1 M sodium monobasic and dibasic phosphate buffer solutions. At each pH level, the phosphate buffer solution was divided into three 101.2-g aliquots, containing 100 g water and 1.2 g buffer salts. Glycerol (18.4 g) was added to 1 aliquot to give 2 molal glycerol in the buffer solution. Similarly, 68.4 g of sucrose were added to a 2nd aliquot to obtain 2 molal sucrose in the buffer solution. The 3rd aliquot, to which no polyol was added, was the phosphate buffer control. The pH values of the solutions were measured before and after the addition of the polyols. In addition, the water activities of the polyol-containing solutions were measured using an Aqualab CX2 (Decagon Devices, Pullman, Wash., U.S.A.).

Aspartame degradation is a well-studied chemical reaction, which is subject to acid-base catalysis. As such, aspartame was selected as the model reaction to evaluate the effect of polyol-induced pKa and pH changes. Aspartame was added to each solution to obtain a final concentration of approximately 10 mg/100 mL solution. Each solution was passed through a 0.2-μm nylon filter. Aliquots (2 to 3 mL) were placed into screw-capped vials and incubated at 25 °C. The sampling intervals depended on the pH of the solution; single samples were removed once or twice per week for a period of up to 49 d. A total of 8 to 10 samples were analyzed throughout the course of the experiment.

The concentration of aspartame in each sample was determined using the isocratic reverse-phase high-performance liquid chromatography (HPLC) method described by Stamp and Labuza (1989) and modified by Bell and Labuza (1991a). The analysis utilized a 3.9 × 150 mm Nova-Pak C18 column (Millipore Corp., Milford, Mass., U.S.A.). The mobile phase consisted of 80/20 (v/v) deionized water/acetonitrile solution, containing 7 mM sodium heptanesulfonate and 5 mM sodium monobasic phosphate. The pH of the mobile phase was adjusted to pH 3 using 85% phosphoric acid. The sample (20 μL) was injected into the mobile phase flowing at a rate of 1 mL/min. Detection occurred at a wavelength of 214 nm using a UV/Visible detector.

The rate constants for aspartame degradation were determined using pseudo-1st-order kinetics, where the natural log of percent aspartame remaining was plotted as a function of time. The rate constants (that is, the slope) with 95% confidence intervals were determined using computerized least squares analysis, as described by Labuza and Kamman (1983).

Additional aspartame kinetic studies

To obtain a larger data set for kinetic modeling, other aspartame kinetic studies were performed. Six phosphate buffer solutions (0.075 to 0.29 M at pH 6.34 to 7.84) were prepared without added sucrose, to which aspartame was added. Nine additional phosphate buffer/2 molal sucrose/aspartame solutions were also prepared. The buffer concentrations of these solutions, prior to adding sucrose, ranged from 0.02 to 0.2 M while the final pH values ranged from 6.0 to 8.1. The kinetic and analytical methodologies were as described previously. The rate constants for aspartame degradation were again determined using the pseudo-1st-order kinetic model.

Aspartame kinetics in sodium dibasic phosphate solutions

Sodium dibasic phosphate solutions in deionized water were prepared at the following concentrations: 0.01, 0.02, 0.05, and 0.1 M. The pH values of each solution were measured. Aspartame (15 to 20 mg) was added to 100 mL of each phosphate buffer solution. The bulk aspartame/buffer solutions were held at 25 °C. A 1-mL aliquot of each solution was removed at varying time intervals for up to 330 min for analysis. Each sample was immediately acidified to pH 5.2 to 5.8 using an acidic phosphate buffer solution. The samples were analyzed using the HPLC method described previously. The rate constants for aspartame degradation in sodium dibasic phosphate buffer were determined as previously described. This experiment allowed for the catalytic effect of the dibasic anion (HPO−24) to be determined for the kinetic model.

Aspartame kinetics in sodium monobasic phosphate solutions

Sodium monobasic phosphate solutions (0.05, 0.075, 0.10, and 0.125 M) were prepared in deionized water. Aspartame (13.5 to 14.2 mg) was added to 100 mL of each solution, and the pH was measured. The solutions were filtered using a 0.2 μm nylon filter, and 2 to 3 mL aliquots were placed into cryovials, which were incubated at 25 °C. Samples were removed from the incubator approximately every 10 to 22 d, for up to 150 d. Samples were frozen at −80 °C until analysis.

Samples were thawed rapidly in lukewarm water, vortexed thoroughly, and analyzed for aspartame using the previously described HPLC method. The rate constants of aspartame loss were again determined using pseudo-1st-order kinetics. This experiment allowed for the catalytic effect of the monobasic phosphate anion (H2PO4) to be determined for the kinetic model.

Development of the aspartame degradation kinetic model

Aspartame degradation is an acid-base catalyzed reaction that has been modeled using pseudo-1st-order kinetics (Prudel and others 1986; Bell and Labuza 1991a; Skwierczynski and Connors 1993). Aspartame has 2 primary degradation pathways (Homler 1984). Unprotonated aspartame undergoes a cyclization reaction to produce the corresponding diketopiperazine (DKP), which is the major degradation product above pH 5.5 (Bell and Labuza 1991b; Skwierczynski and Connors 1993). Cleavage of the methyl ester yields the other major degradation product, α-L-aspartyl-L-phenylalanine (α-AP), which is favored at pH values less than 3 (Bell and Labuza 1991b). Thus, as pH increases, the proportion of DKP increases, while the concentration of α-AP decreases (Bell and Labuza 1991b). At pH 7, DKP was the only measurable product formed (Bell and Labuza 1991b).

In the current study, aspartame degradation was used as a model acid-base catalyzed reaction to evaluate the effect of nonionic polyols on its kinetics in phosphate buffer solutions. These aqueous solutions of phosphate buffer exist as an equilibrium mixture of several different species (that is, the acid [H3PO4], the monobasic anion [H2PO4], the dibasic anion [HPO−24], and the tribasic anion [PO−34]), present in varying proportions depending on the pH of the medium (Christian 1980). Of phosphoric acid's 3 pKa values, the equilibrium concentrations of the monobasic and dibasic anions are governed by the 2nd pKa value (pK2) between approximately pHs 4.5 and 9. Because buffer type and concentration influence the rates of aspartame degradation (Tsoubeli and Labuza 1991; Bell and Wetzel 1995), it is important to take into account the catalysis by the buffer salts when evaluating the effect of pH and pKa on the rate constant.

Based on the above-mentioned degradation pathways, the pseudo-1st-order rate constant for aspartame degradation, kobs, can be expressed using the following equation for an acid-base catalyzed reaction

  • image(1)

where ko (rate constant for the uncatalyzed reaction) equals 1 × 10−7 min−1, k[H+] (rate constant for catalysis by hydronium ions) equals 0.00146 M−1 min−1, and k[OH−] (the rate constant for catalysis by hydroxyl ions) equals 79.8 M−1 min−1, as previously determined at 25 °C by Bell and Wetzel (1995). In addition, [H+], [OH], [H2PO4], and [HPO−24] are the molar concentrations of hydronium ions, hydroxyl ions, monobasic phosphate ions, and dibasic phosphate ions, respectively. Using the measured pH, [H+] and [OH] are calculated, assuming an activity coefficient (γ) equal to one. Because pH actually measures hydrogen ion activity (aH), γ is required to convert these into hydrogen ion concentrations ([H+]= aH/γ) for the kinetic expressions. From the pH and the apparent pKa value of phosphate buffer, [H2PO4] and [HPO−24] can be calculated using the Henderson–Hasselbach equation. The terms kA1 and kA2 are the rate constants for the formation of α-AP catalyzed by the monobasic and dibasic phosphate ions, respectively, while kD1 and kD2 are the rate constants for the formation of DKP catalyzed by the monobasic and dibasic phosphate ions, respectively. The term fNH2 is the fraction of aspartame having the amine group unprotonated (the required molecular conformation leading to DKP formation). This term is calculated using the pKa value of 7.9 for aspartame's amine group (Skwierczynski and Connors 1993) and the measured pH of the system; algebraic rearrangement of the Henderson–Hasselbach equation yields the following equation

  • image(2)

As mentioned previously and verified experimentally, DKP was the only measurable degradation product in the dibasic phosphate buffer solutions (pH > 7). This finding is consistent with those reported in the literature for solutions having alkaline pH values (Prudel and others 1986; Gaines and Bada 1988; Pattanaargson and Sanchavanakit 2000). The amount of α-AP was negligible, so kA2, the rate constant for the formation of α-AP catalyzed by the dibasic phosphate anion, was taken as zero and omitted from Eq. 1. Thus, the kinetic equation for aspartame degradation in dibasic phosphate buffer becomes

  • image(3)

Defining k′ as (ko+k[H+][H+]+k[OH−][OH]), substituting into Eq. 3, and rearranging the equation, one arrives at

  • image(4)

A plot of (kobsk′) as a function of (fNH2[HPO−24]) gave a straight line with a slope of 0.271 (that is, kD2= 0.271 M−1 min−1) and an R2 value of 0.999. The standard error associated with kD2 was 2.0 × 10−5.

Similarly, the rate constants for the formation of α-AP (kA1) and DKP (kD1) catalyzed by monobasic phosphate buffer were determined using the following equation

  • image(5)

Using nonlinear regression to fit the experimental data to Eq. 5, the values of kA1 and kD1 were determined to be 3.98 × 10−5 and 2.33 × 10−2 M−1 min−1, respectively. The R2 value for this nonlinear regression model was 0.995. The standard errors associated with kA1 and kD1 were 3.4 × 10−6 and 6.7 × 10−3, respectively.

Substituting the values of the rate constants for catalysis by dibasic and monobasic phosphate buffer into Eq. 1, the final mathematical equation for modeling aspartame degradation in phosphate buffer at 25 °C becomes

  • image(6)

It should be pointed out that this equation is unique to aspartame degradation in phosphate buffer at 25 °C. Equation 6 is not applicable to the reaction occurring in a different buffer type or at a different temperature. The above-mentioned kinetic model will be used to evaluate how the polyol-induced changes in buffer pKa and pH values affect aspartame degradation, an acid-base catalyzed reaction; this discussion will be presented in detail subsequently.

The experimentally determined rate constants were compared to those predicted from Eq. 6 using linear regression to determine slopes and R2 values. Pearson's product moment correlation coefficient and the significance of the correlations between experimental and predicted kinetic data were then evaluated.

Results and Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusions
  7. References

Figure 1 shows typical pseudo-1st-order plots for aspartame degradation in phosphate buffer, initially at pH 5.6, with and without added polyols. The excellent linearity confirms the appropriateness of the pseudo-1st-order kinetic model. For visual clarity, this figure was prepared using log base 10 to scale the y-axis; however, pseudo-1st-order rate constants were actually calculated using the slope of the percent aspartame remaining plotted on a natural logarithmic scale versus time, as described previously. Aspartame degradation in phosphate buffer was reduced by adding sucrose, but not glycerol (Figure 1).

image

Figure 1—. Pseudo-1st-order kinetic plots for aspartame degradation at 25 °C in phosphate buffer initially at pH 5.6 before polyol addition.

Download figure to PowerPoint

Figure 2 shows the pseudo-1st-order rate constants for aspartame degradation in various 0.1 M phosphate buffer solutions as a function of the original pH before polyol addition (assuming no pH effect). The 0.1 M buffer concentration was also before adding the polyols. This graph shows that the degradation rate constants in the buffer containing 2 molal sucrose were lower than those for the reaction in buffer alone or buffer containing 2 molal glycerol. Because both the 2 molal sucrose and glycerol solutions had the same water activity (0.95), water activity lowering is not the reason for the increased stability from adding sucrose.

image

Figure 2—. Pseudo-1st-order rate constants for aspartame degradation in phosphate buffer at 25 °C as a function of the initial buffer pH measured before adding the polyols. Error bars represent the 95% CI.

Download figure to PowerPoint

Figure 3 shows the same kinetic data as in Figure 2, but with the correctly measured final pH of the phosphate buffer/polyol solution. Sucrose caused the pH of the phosphate buffer to decrease by approximately 0.3 pH units. Therefore, the curve associated with sucrose in Figure 2 shifted to the left such that all the data are now superimposed. These results suggest that the pH-lowering effect associated with sucrose being added to phosphate buffer was indeed influencing the kinetics of this acid-base catalyzed reaction. However, this experiment involved 1 buffer concentration over a limited pH range. Additional data over a broader pH range at other buffer concentrations are needed to validate and more fully understand these findings.

image

Figure 3—. Pseudo-1st-order rate constants for aspartame degradation in phosphate buffer at 25 °C as a function of the final pH measured after adding the polyols. Error bars represent the 95% CI.

Download figure to PowerPoint

Because aspartame degradation is influenced by both buffer concentration and pH, a broader array of kinetic data cannot be presented graphically on a 2-dimensional pH-rate profile, as shown in Figure 2 and 3. The additional kinetic data from buffer solutions of various concentrations and pH levels were evaluated by comparing the experimental rate constants to those predicted using the kinetic model for aspartame degradation (Eq. 6).

In Eq. 6, the values of [H+] and [OH] were calculated from the measured pH of the solution, again assuming γ= 1. As mentioned, fNH2 was determined from the pKa of the aspartame amine group and the pH of the solution (Eq. 2). The values for [H2PO4] and [HPO−24] were calculated from the initial buffer concentration, the measured pH, the apparent pKa value, and the dilution attributed to polyol addition. The formation of the 2 molal sucrose or glycerol solutions resulted in an increase in total volume by a factor of 1.42 or 1.14, respectively. Over the pH range of this study, pK2 is the appropriate pKa value to use for the kinetic model. The average value for the apparent pK2 of phosphate buffer alone was 6.78 while with the addition of 2 molal sucrose or glycerol, the average apparent pK2 values were 6.49 and 6.76, respectively (Chuy and Bell 2006).

Inserting the appropriate values into the aspartame degradation kinetic model (Eq. 6), the pseudo-1st-order rate constants were predicted for each pH/buffer concentration combination. Figure 4 compares the experimental rate constants to the predicted rate constants. A perfect predictive model would yield a slope of 1.0 and a correlation coefficient of 1.0. As seen in this figure, the slope was 0.962 with an R2 of 0.996. Pearson's product moment correlation coefficient of 0.997 indicates an excellent fit between the predicted and experimental data, which was significant (P < 0.05). The model was also tested by incorporating the standard errors into kA1, kD1, and kD2; even with the inclusion of these errors, the slope varied only from 0.960 to 0.964 and the R2 value remained at 0.996. Therefore, the kinetic model presented in Eq. 6 can be successfully employed.

image

Figure 4—. Correlation between experimentally determined rate constants for aspartame degradation in phosphate buffer at 25 °C with those predicted from the kinetic model.

Download figure to PowerPoint

Figure 5 compares the aspartame data from Bell and Wetzel (1995) with that predicted using Eq. 6. At buffer concentrations below 0.2 M, the kinetic model holds very well, with a slope of 0.92 and R2 of 0.996. However, at high buffer concentrations (0.5 and 1.0 M), the actual rate constants are lower than predicted by the model. At these high buffer concentrations, ionic interactions may reduce the catalytic ability of the various phosphate anions, resulting in lower rates. Thus, the predictive capability of our model deviates at high buffer concentrations.

image

Figure 5—. Correlation between experimentally determined rate constants for aspartame degradation in various concentrations of phosphate buffer at 25 °C with those predicted from the kinetic model. Experimental data are from Bell and Wetzel (1995).

Download figure to PowerPoint

Kinetic data for aspartame degradation in phosphate buffer containing 2 molal sucrose were collected. The addition of sucrose lowered the pH of the phosphate buffer solutions, which changed [H+], [OH], and fNH2 in Eq. 6. Sucrose also caused a change to the pKa value of the buffer and increased the solution volume (that is, dilution), both of which changed [H2PO4] and [HPO−24]. Taking all these effects into account, the predicted rate constants were determined. Figure 6 shows the predictive capability of the kinetic model for the buffer solutions containing 2 molal sucrose. The slope of the plot is 0.76, which is still relatively close to the ideal value of 1. A slope lower than 1 indicates that the actual reactivity was less than that predicted by the kinetic model. This reduction of the slope relative to 1 could be due to interactions between sucrose and the phosphate anions. FTIR measurements have suggested interactions between the hydroxyl groups of sugars and phosphates (Ohtake and others 2004). Such an interaction may lower the catalytic ability of the phosphate anions such that the net reactivity is decreased from that predicted in the model. Despite the slightly lower slope and a bit more scatter of the data, a definite linear relation exists (R2= 0.956), and Pearson's product moment correlation coefficient of 0.978 is significant (P < 0.05).

image

Figure 6—. Correlation between experimentally determined rate constants for aspartame degradation in phosphate buffer containing 2 molal sucrose at 25 °C with those predicted from the kinetic model.

Download figure to PowerPoint

The modeling process was repeated for the glycerol-containing solutions from the 0.1 M phosphate buffer experiments. As mentioned, the dilution factor was 1.14, and the pKa value was 6.76 (Chuy and Bell 2006). Making the appropriate substitutions into Eq. 6, the graph of the experimental rate constants as a function of the predicted rate constants (Figure 7) again showed excellent linearity (R2= 0.997) with a slope of 1.25. This slope, being greater than 1, indicates that the actual reactivity was greater than that predicted by the kinetic model. Seow and Cheah (1985) showed that glycerol could react with glycine in solution at pH 4, resulting in brown pigment formation. Aspartame, being composed of amino acids, may also react with glycerol to cause faster degradation, which the kinetic model (Eq. 6) would not take into account. This enhanced degradation was not seen directly in Figure 1 because while glycerol may react to cause greater aspartame loss, it is simultaneously decreasing the degradation by diluting the catalytic buffer salts. Therefore, glycerol-induced aspartame loss is compensated for by the more dilute phosphate buffer, resulting in the experimental degradation rate constants with and without glycerol being similar.

image

Figure 7—. Correlation between experimentally determined rate constants for aspartame degradation in phosphate buffer containing 2 molal glycerol at 25 °C with those predicted from the kinetic model.

Download figure to PowerPoint

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusions
  7. References

The kinetic model successfully estimated the rate constants for aspartame degradation in phosphate buffer both with and without the presence of polyols. The ability to utilize changes to the buffer properties (pKa, pH, buffer species concentration) attributed to the incorporation of nonionic polyols in the kinetic model suggests that these changes are affecting the kinetic behavior of acid-base catalyzed reactions. Because reaction kinetics depend upon reactant concentration, the pH change from adding sucrose to phosphate buffer appears to result from an increase in the hydronium ion concentration due to a shift in the apparent pKa of the buffer. This conclusion is consistent with that of Chuy and Bell (2006). Thus, the relatively inert nonionic sucrose can affect the kinetics of acid-base catalyzed reactions by changing the apparent pKa value, which causes the buffer pH to change as well as the relative amounts of the buffering species. These potential effects should be recognized during product development so that formulations can be adjusted for optimum product stability.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusions
  7. References
  • Bell LN, Labuza TP. 1991a. Aspartame degradation kinetics as affected by pH in intermediate and low moisture food systems. J Food Sci 56:1720.
  • Bell LN, Labuza TP. 1991b. Potential pH implications in the freeze-dried state. Cryo-Letters 12:23544.
  • Bell LN, Labuza TP. 1992. Compositional influence on the pH of reduced-moisture solutions. J Food Sci 57:7324.
  • Bell LN, Wetzel CR. 1995. Aspartame degradation in solution as impacted by buffer type and concentration. J Agric Food Chem 43(10):260812.
  • Christian G. 1980. Analytical chemistry. 3rd ed. New York : John Wiley & Sons. 643 p.
  • Chuy S, Bell LN. 2006. Buffer pH and pKa values as affected by added glycerol and sucrose. Food Res Int 39:3428.
  • Connors KA, Amidon GL, Stella VJ. 1986. Chemical stability of pharmaceuticals. 2nd ed. New York : John Wiley & Sons. p 20820.
  • Dwivedi B, Arnold R. 1972. Chemistry of thiamine degradation: mechanism of thiamine degradation in model system. J Food Sci 37:8868.
  • Friedman M. 1996. Food browning and its prevention: an overview. J Agric Food Chem 44:63153.
  • Gaines SM, Bada JL. 1988. Aspartame decomposition and epimerization in the diketopiperazine and dipeptide products as a function of pH and temperature. J Org Chem 53(12):275764.
  • Galster H. 1991. pH measurement: fundamentals, methods, applications, instrumentation. Weinheim : VCH. 356 p.
  • Homler BE. 1984. Properties and stability of aspartame. Food Tech 38(7):505.
  • Kelly FHC, Brown DW. 197879. Thermal decomposition and colour formation in aqueous sucrose solutions. Sugar Technol Rev 6:148.
  • Labuza T, Kamman J. 1983. Reaction kinetics and accelerated tests simulation as function of temperature. In: SaguyI, editor. Computer-aided techniques in food technology. New York : Marcel Dekker, Inc. p 71115.
  • Lindsay RC. 2008. Food additives. In: DamodaranS, ParkinKL, FennemaOR, editors. Fennema's food chemistry. 4th ed. Boca Raton , Fla. : CRC Press. p 689749.
  • Martins SIFS, Jongen WMF, Van Boekel MAJS. 2001. A review of Maillard reaction in food and implications to kinetic modeling. Trends Food Sci Technol 11:36473.
  • Mauri L, Alzamora S, Tomio J. 1992. Effect of electrolytes on the kinetics of thiamine loss in model systems of high water activity. Food Chem 45:1923.
  • McMillan G. 1994. pH measurement and control. 2nd ed. Triangle Research Park , N.C. : Instrument Society of America. 304 p.
  • Ohtake S, Schebor C, Palecek SP, De Pablo JJ. 2004. Effect of pH, counter ion, and phosphate concentration on the glass transition temperature of freeze-dried sugar-phosphate mixtures. Pharm Res 21:161521.
  • Pattanaargson S, Sanchavanakit C. 2000. Aspartame degradation study using electrospray ionization mass spectrophotometry. Rapid Commun Mass Spectrom 14:98793.
  • Perrin DD, Dempsey B. 1974. Buffers for pH and metal ion control. London , U.K. : Chapman and Hall. 176 p.
  • Prudel M, Davidkova E, Davidek J, Kminek M. 1986. Kinetics of decomposition of aspartame hydrochloride (Usal) in aqueous solutions. J Food Sci 51(6):13937, 1415.
  • Seow CC, Cheah PB. 1985. Reactivity of sorbic acid and glycerol in non-enzymatic browning in liquid intermediate moisture model systems. Food Chem 18:7180.
  • Skwierczynski RD, Connors KA. 1993. Demethylation kinetics of aspartame and L-phenylalanine methyl ester in aqueous solutions. Pharm Res 10(8):117480.
  • Sloan AE, Labuza TP. 1975. Investigating alternative humectants for use in foods. Food Prod Devel 9(9):7588.
  • Stamp J, Labuza TP. 1989. An ion-pair high performance liquid chromatographic method for the determination of aspartame and its decomposition products. J Food Sci 54(4):10436.
  • Tsoubeli MN, Labuza TP. 1991. Accelerated kinetic study of aspartame degradation in the neutral pH range. J Food Sci 56(6):16715.
  • Wescott C. 1978. pH measurements. New York : Academic Press. 172 p.