Aspartame Optical Biosensor with Bienzyme-Immobilized Eggshell

An aspartame optical biosensor has been fabricated by employing a bienzyme ..... A fluorescent glucose biosensor based on immobilized glucose oxidase ...
6 downloads 0 Views 215KB Size
Download PDF
Anal. Chem. 2002, 74, 863-870

Aspartame Optical Biosensor with Bienzyme-Immobilized Eggshell Membrane and Oxygen-Sensitive Optode Membrane Dan Xiao

College of Chemistry and Chemical Engineering, Hunan University, Changsha 410082, P. R. China Martin M. F. Choi*

Department of Chemistry, Hong Kong Baptist University, Kowloon Tong, Hong Kong SAR, P. R. China

An aspartame optical biosensor has been fabricated by employing a bienzyme system composed of r-chymotrypsin and alcohol oxidase immobilized onto an eggshell membrane and an oxygen-sensitive optode membrane as the transducer. The detection schemes involve the enzymatic reactions of aspartame leading to the depletion of the oxygen level of the medium with a concomitant enhancement of the fluorescence intensity of the oxygensensitive membrane. The scanning electron and transmission electron micrographs show the microstructure of the eggshell membrane which is successfully immobilized with bienzyme. Using this novel immobilization technique, the aspartame biosensor shows extremely good stability with a shelf life of at least 8 months. The rate change of the fluorescence intensity in 4 min is found to be linearly related to the concentration of aspartame. The useful analytical working range of the biosensor is from 0.056 to 3.07 mM aspartame. The effects of temperature, pH, and ionic strength on the response of the aspartame biosensor are investigated in detail. Citric acid, cyclamic acid, D-fructose, D-galactose, D-glucose, hydrogen peroxide, DL-malic acid, L-phenylalanine, saccharin, sodium benzoate, and sucrose show no interferences but ethanol interferes strongly. The aspartame biosensor has been applied to determine aspartame contents in some commercial products. Artificial sweeteners are a staple in the diet of many people. They have been on the market for more than a century and were expected to be a U.S. $1.1 billion business in the United States in 2000. Among these sweeteners, aspartame (N-L-R-aspartyl-Lphenylalanine 1-methyl ester) has dominated the U.S. market with the 1995 estimated consumption at 18.8 million lb. It is rapidly replacing saccharin and cyclamate in the market as they are suspected to cause cancer even though currently there is controversy over the long-term health effects of these sweeteners in humans.1,2 The determination of aspartame can be done by many analytical procedures such as spectrophotometry,3-5 capil* Corresponding author: (e-mail) [email protected]; (fax) +852 2339 7348. (1) Kirschner, E. M. Chem. Eng. News 1997, 75 (April 21), 21. (2) Khan, R.; Konowicz, P. A. Sweeteners. In Encyclopedia of Analytical Science; Townshend, A., Ed.; Academic Press: London; 1995; Vol. 9, pp 5085-5090. 10.1021/ac001097a CCC: $22.00 Published on Web 01/19/2002

© 2002 American Chemical Society

lary electrophoresis,6,7 thin-layer chromatography,8 and gas chromatography.9 High-performance liquid chromatography (HPLC)10-15 is the most commonly used method for the quantification of aspartame. However, these methods usually require lengthy pretreatment of the sample prior to the chromatographic separation. There is still a demand for an inexpensive, quick, and automated method for the determination of aspartame. Recently, research in the development of chemosensors and biosensors has been very fast growing. They represent an active, exciting, and innovative research area in analytical chemistry with more than 1000 papers being published annually.16 Despite the remarkable breadth of biosensor technology, commercial success is still limited. The development of a pocket-sized blood glucose monitor can be considered to be one of the success stories in this field.17 However, only a few potentiometric aspartame sensors18,19 and amperometric aspartame biosensors20-24 have been reported in the literature to date. (3) Lau. O.-W.; Luk, S.-F.; Chan, W.-M. Analyst 1988, 113, 765-768. (4) Hamano, T.; Mitsuhashi, Y.; Aoki, N.; Yamamoto, S.; Tsuji, S.; Ito, Y.; Oji, Y. Analyst 1990, 115, 435-438. (5) Fatibello-Filho, O.; Marcolino-Junior, L. H.; Pereira, A. V. Anal. Chim. Acta 1999, 384, 167-174. (6) Kvasnicka, F. J. Chromatogr. 1987, 390, 237-240. (7) Sabah, S.; Scriba, G. K. E. J. Pharm. Biomed. Anal. 1998, 16, 1089-1096. (8) Gro ¨sz, J.; Jonas, O. J. Planar Chromatogr. 1990, 3, 261-263. (9) Furda, I.; Malizia, P. D.; Kolor, M. G.; Vernieri, P. J. J. Agric. Food Chem. 1975, 23, 340-343. (10) Tyler, T. A. J. Assoc. Off. Anal. Chem. 1984, 67, 745-747. (11) Tsang, W.-S.; Clarke, M. A.; Parrish, F. W. J. Agric. Food Chem. 1985, 33, 734-738. (12) Lawrence, J. F.; Iyengar, J. R. J. Chromatogr. 1987, 404, 261-266. (13) Stamp, J. A.; Labuza, T. P. J. Food Sci. 1989, 54, 1043-1046. (14) Prodolliet, J.; Bruelhart, M. J. AOAC Int. 1993, 76, 268-274. (15) Prodolliet, J.; Bruelhart, M. J. AOAC Int. 1993, 76, 275-282. (16) Forster, R. J.; Diamond, D. Anal. Commun. 1996, 33, 1H-4H. (17) Rouhi, A. M. Chem. Eng. News 1997, 75 (May 12), 41-45. (18) Nikolelis, D. P.; Krull, U. J. Analyst 1990, 115, 883-888. (19) Badawy, S. S.; Issa, Y. M.; Tag-Eldin, A. S. Electroanalysis 1996, 8, 10601064. (20) Renneberg, R.; Riedel, K.; Scheller, F. Appl. Microbiol. Biotechnol. 1985, 21, 180-181. (21) Smith, V. J.; Green, R. A.; Hopkins, T. R. J. Assoc. Off. Anal. Chem. 1989, 72, 30-33. (22) Chou, S.-F.; Chen, J.-H.; Chou, L.-W.; Fan, J.-J.; Chen, C.-Y. J. Food Drug Anal. 1995, 3, 121-126. (23) Chou, S.-F. Analyst 1996, 121, 71-73. (24) Compagnone, D.; O’Sullivan, D.; Guilbault, G. G. Analyst 1997, 122, 487490.

Analytical Chemistry, Vol. 74, No. 4, February 15, 2002 863

Optical sensors for a wide variety of chemical species have drawn considerable attention in the past few decades because they possess certain advantages over conventional electrical and electrochemical sensors or devices. The advantages include immunity to electromagnetic interference, ease of miniaturization, possibility of remote and in situ monitoring, and high informationcarrying capacity. They may also be used safely where sources of ignition are to be restricted.25 To our knowledge, no previous attempts have been made to develop optical biosensors for determining aspartame in food samples. Enzyme-based biosensors are currently one of the hot topics for research and development since essentially all chemical reactions in living systems are catalyzed by enzymes. For a reusable enzyme-based biosensor, most commonly, an enzyme is required to be immobilized on, or in close proximity to, the surface of a transducer. As a consequence, immobilization strategies for enzymes are of paramount importance in order to preserve their biological activity.26 Over the years, many methods have been used to immobilize enzymes for use in biosensing. To date, most enzyme immobilization techniques fall into one of three categories: (1) physisorption,27,28 (2) covalent attachment,29,30 and (3) entrapment.31-35 Sometimes two or three enzyme immobilization methods may be used simultaneously. Often the problem of long-term operational and storage stability presents a crucial hurdle for commercial development of biosensors. The shelf life of an enzyme-based biosensor usually depends on how long the biological activity of the enzyme can be retained, and this may vary from days to months. It has been reported that some biomaterials including silk,36-38 collagen,39,40 and eggshell membrane41,42 were employed as platforms for the immobilization of enzymes and the lifetimes of the immobilized enzymes were much extended. Eggshell membrane, having excellent gas and water permeability, may be an ideal biomaterial for enzyme immobilization. Its thickness is approximately 65-96 µm43 and it is mainly (25) Taib, M. N.; Narayanaswamy, R. Analyst 1995, 120, 1617-1625. (26) Wink, T.; van Zuilen, S. J.; Bult, A.; van Bennekom, W. P. Analyst 1997, 122, 43R-50R. (27) Koncki, R.; Leszczynski, P.; Hulanicki, A.; Glab, S. Anal. Chim. Acta 1992, 257, 67-72. (28) Gilmartin, M. A. T.; Hart, J. P. Analyst 1994, 119, 833-840. (29) Trettnak, W.; Wolfbeis, O. S. Anal. Lett. 1989, 22, 2191-2197. (30) Koncki, R.; Mohr, G. J.; Wolfbeis, O. S. Biosens. Bioelectron. 1995, 10, 653659. (31) Ku ¨ nzelmann, U.; Bo ¨ttcher, H. Sens. Actuators, B 1997, 38-39, 222-228. (32) Chen, Q.; Kenausis, G. L.; Heller, A. J. Am. Chem. Soc. 1998, 120, 45824585. (33) Heller, J.; Heller, A. J. Am. Chem. Soc. 1998, 120, 4586-4590. (34) Vidal, J.-C.; Garcı´a, E.; Me´ndez, S.; Yarnoz, P.; Castillo, J.-R. Analyst 1999, 124, 319-324. (35) Hartnett, A. M.; Ingersoll, C. M.; Baker, G. A.; Bright, F. V. Anal. Chem. 1999, 71, 1215-1224. (36) Demura, M.; Takekawa, T.; Asakura, T.; Nishikawa, A. Biomaterials 1992, 13, 276-280. (37) Qian, J.; Liu, Y.; Liu, H.; Yu, T.; Deng, J. Biosens. Bioelectron. 1997, 12, 1213-1218. (38) Liu, Y.; Chen, X.; Qian, J.; Liu, H.; Shao, Z.; Deng, J.; Yu, T. Appl. Biochem. Biotechnol. 1997, 62, 105-117. (39) George, S.; Chellapandian, M.; Sivasankar, B.; Sundaram, P. V. Bioprocess Eng. 1996, 15, 311-315. (40) Michel, P. E.; Gautier-Sauvigne´, S. M.; Blum, L. J. Anal. Chim. Acta 1998, 360, 89-99. (41) Hu, P.; Fang, Y.; Zhou, T.; Zhu, M. Fenxi Ceshi Xuebao 1995, 14, 51-53. (42) Deng, J.; Yuan, Y.; Xu, J.; Xiao, D.; Wang, K. Fenxi Huaxue 1998, 10, 12571259.

864

Analytical Chemistry, Vol. 74, No. 4, February 15, 2002

composed of biological molecules and protein fibers, which may supply polycations to stabilize the enzymes.33 With the protection of eggshell and eggshell membrane, if not infected by microbes and bacteria, an egg can be kept fresh for a long time. In this article, we explore the possibility of fabricating a long-lived optical aspartame biosensor based on a bienzyme-immobilized eggshell membrane. To the best of our knowledge, this is the first report on an aspartame optical sensor that consists of an immobilized bienzyme eggshell membrane and a highly oxygen-sensitive optode membrane. EXPERIMENTAL SECTION Reagents. Acetone, diethyl ether, N,N-dimethylformamide, 4,7-diphenyl-1,10-phenanthroline, ethanol, ethylene glycol, 50% (w/w) glutaraldehyde solution in water, L-phenylalanine, ruthenium(III) chloride hydrate, poly(vinyl chloride), sodium salt of saccharin, sodium chloride, and tetrahydrofuran (THF) were purchased from Aldrich (Milwaukee, WI). Aspartame, alcohol oxidase (EC 1.1.3.13 from Hansenula sp.) with a specific activity of 20-40 units/mg of protein, R-chymotrypsin (EC 3.4.21.1 from bovine pancreas) with a specific activity of 40-60 units/mg of protein, citric acid, cyclamic acid, and D-glucose were obtained from Sigma (St. Louis, MO). 1,4-Dithio-DL-threitol, D-galactose, and potassium tetrakis(4-chlorophenyl)borate were purchased from Fluka Chemicals (Buchs, Switzerland). A 35% (w/w) hydrogen peroxide solution in water and DL-malic acid were obtained from Acros Organics (Geel, Belgium). D-Fructose was purchased from Merck (Darmstadt, Germany). Sodium benzoate was obtained from BDH Chemicals (Poole, U.K.). Sucrose was from Ajax Chemicals (Auburn, Australia). All other reagents were of analytical reagent grade and used without further purification. The buffer solution for preparing aspartame standards was 25 mM sodium phosphate solution at pH 6.9. All solutions were prepared with deionized water. The dye ion pair, tris(4,7-diphenyl-1,10-phenanthroline)ruthenium(II) ditetrakis(4-chlorophenyl)borate [Ru(dpp)3][(4-Clph)4B]2, was synthesized according to a modified method.44 [Ru(dpp)3][(4Clph)4B]2 adsorbed on silica gel was synthesized as the following procedure: 10 g of Silica Gel 60 (Merck, Darmstadt, Germany) was stirred with 10 mL of a 0.1 M ethanolic solution of [Ru(dpp)3][(4-Clph)4B]2 overnight. The solid residue of [Ru(dpp)3][(4Clph)4B]2 adsorbed on silica gel was filtered and washed successively with THF, diethyl ether, acetone, ethanol, and deionized water in order to remove excess and unadsorbed [Ru(dpp)3][(4Clph)4B]2 dye ion pair. This solid residue was put in an oven at 100 °C for 5 h to evaporate the solvents. The dry [Ru(dpp)3][(4Clph)4B]2 adsorbed on silica gel was then kept in a desiccator for further use. Oxygen-Sensitive Optode Fabrication. Ten milligrams of [Ru(dpp)3][(4-Clph)4B]2 adsorbed on silica gel was evenly dispersed in ∼0.3 g of uncured one part silicone prepolymer (Wu Xi Sealant Factory, Jiangsu, China) thoroughly. By use of a spreading method, the mixture was adhered to the surface of a clean transparent glass plate to form a silicone-based oxygen-sensitive film. The membrane was left at 75 °C for 24 h to completely cure. The thickness of the oxygen-sensitive film was estimated to be (43) Liong, J. W. W.; Frank, J. F.; Bailey, S. J. Food Prot. 1997, 60, 1022-1028. (44) Klimant, I.; Wolfbeis, O. S. Anal. Chem. 1995, 67, 3160-3166.

Figure 1. Schematic diagram of the homemade flow cell positioned with an aspartame-sensitive optode membrane. Key: (1) fixing screw; (2) Poly(tetrafluoroethylene) support; (3) glass support; (4) cell volume, 0.3 mL; (5) surface-immobilized R-chymotrypsin layer; (6) eggshell membrane; (7) surface-immobilized alcohol oxidase layer; (8) silicone rubber; (9) [Ru(dpp)3][(4-Clph)4B]2 adsorbed on silica gel particles; (10) O-ring seal; (11) stainless steel tubing; (12) sample inlet; (13) sample outlet; (14) excitation light beam; (15) emission light beam.

∼10 µm. It was stored in a desiccator at room temperature prior to being used. Immobilization of Bienzyme on Eggshell Membrane. An eggshell membrane was carefully peeled off from a broken fresh eggshell after the albumen and yolk have been removed. It was then cleaned with deionized water. Further cleaning steps of the eggshell membrane were necessary in order to completely remove the albumen from the eggshell membrane. The cleaning procedure proceeded in the following sequences: 1 min

2h

surface rinsing 98 water immersing 98 1 min

2h

1 min

2h

surface rinsing 98 water immersing 98 surface rinsing 98 water immersing 98 1 min

phosphate buffer (pH 6.9) surface rinsing 98 store in phosphate buffer (pH 6.9)

The cleaned eggshell membrane was finally stored in a pH 6.9 phosphate buffer (25 mM) until further use. The eggshell membrane was removed from the phosphate buffer. It was cut into a circle of a diameter of ∼15 mm and then placed on a clean glass slide. One milligram of R-chymotrypsin dissolved in 5 µL of pH 6.9 phosphate buffer (25 mM) was added; after ∼2 min, 1 µL of 25% (w/w) glutaraldehyde solution as the cross-linking agent was dropped onto the surface of the membrane and remained for 5 min. The membrane was then immersed in a phosphate buffer for 20 min. Afterward, the membrane was released from the buffer and placed on a clean glass slide with the fresh surface facing upward. The enzyme immobilization procedure was repeated by using alcohol oxidase instead of R-chymotrypsin. Finally, the bienzyme-immobilized membrane was rinsed with and immersed in a phosphate buffer (pH 6.9) three times alternately to remove the un-cross-linked and unabsorbed enzymes. The bienzyme-immobilized eggshell membrane was kept

in a pH 6.9 phosphate buffer (25 mM) for further assembly of an aspartame optical biosensor. Preparation of Eggshell Membranes for Electron Microscopy. Eggshell membranes (1 × 1 × 5 mm) with and without immobilized enzymes were prepared for scanning electron microscopy and transmission electron microscopy. Membranes were initially fixed with 1% (w/v) osmium tetroxide and 2.5% (w/v) glutaraldehyde solutions in Petri dishes to minimize actual manipulation of the membrane surfaces. The samples were removed from the Petri dishes and sequentially dehydrated by submersion in 50%, 70%, 80%, 85%, 90%, 95%, and 100% (v/v) acetone. Membranes were further immersed in an epoxy resin solution for a few minutes, stained with lead nitrate and uranium acetate, and dried. The membranes were finally coated with a thin layer of gold using a spray gun and observed under a H-600 Hitachi scanning electron microscope (Tokyo, Japan). Aspartame Optical Biosensor. The eggshell membrane with both surfaces immobilized with R-chymotrypsin and alcohol oxidase, respectively, landed on the surface of an oxygen-sensitive optode membrane. A schematic diagram of the experimental arrangement of the aspartame optical biosensor is depicted in Figure 1. The flow cell was positioned in a spectrofluorometer for fluorescence measurements. About 1.0 mL of a freshly prepared standard or sample solution was injected into the homemade flow cell with the use of a 5.0-mL syringe. All measurements were performed in the air-saturated buffer solutions. The fluorescence emission intensity at 612 nm was collected at an excitation wavelength of 468 nm. Unless otherwise stated, all fluorescence measurements were taken under batch conditions at 20 ( 2 °C and at a pressure of 101.3 kPa. When the aspartame optical biosensor was not in use, it was stored in a 10 mM dithiothreitol solution either at 4 or 23 °C. Apparatus. Fluorescence measurements were recorded using a spectrofluorometer consisting of a lamp power supply (model LPS-220), a xenon lamp (model A1010), and a photomultiplier Analytical Chemistry, Vol. 74, No. 4, February 15, 2002

865

detection system (model 710) from Photon Technology International (London, ON, Canada). The pH measurements were taken on an Orion combined pH glass electrode (Chicago, IL). RESULTS AND DISCUSSION Oxygen-Sensitive Optode Membrane. The oxygen-sensitive optode membrane acting as a transducer was employed to measure the rate of oxygen consumption in the enzymatic reactions of aspartame and methanol:4

H2NCH(CH2CO2H)CONHCH(CH2C6H5)CO2CH3 + R-chymotrypsin

H2O 98 H2NCH(CH2CO2H)CONHCH(CH2C6H5)CO2H + CH3OH alcohol oxidase

CH3OH + O2 98 HCHO + H2O2 R-chymotrypsin hydrolyzes aspartame to produce methanol, which is subsequently oxidized to formaldehyde by alcohol oxidase with the consumption of oxygen. The depletion of the oxygen level can then be picked up by the oxygen-sensitive optode membrane. Surely, the analytical performance of the oxygen-sensitive optode membrane can affect the sensitivity and limit of detection of the aspartame biosensor. The optical oxygen sensing is based on collision quenching of the fluorescence of [Ru(dpp)3][(4-Clph)4B]2 molecules by oxygen molecules.45,46 Hence, the biosensor response is composed of a dynamic balance in the diffusion of aspartame into the eggshell membrane and oxygen into the silicone rubber film and the consumption of oxygen in the enzymatic reactions; as a result, there is a steady-state decrease in oxygen level and, consequently, an increase in fluorescence intensity. Quenching can be quantified by intensity quenching measurements. The oxygen quenching process is described by the well-known Stern-Volmer equation:44-46

Io/I ) 1 + KpO2 where I is the fluorescence intensity, the subscript o denotes the absence of oxygen, K is the Stern-Volmer constant, and pO2 is the partial pressure of oxygen. A plot of (Io/I) versus the partial pressure of oxygen should give a straight line with a slope K and an intercept of 1 on the y-axis. The oxygen-sensing film was tested to show good response to various dissolved oxygen concentrations. The results clearly demonstrated that the oxygen-sensitive membrane was very sensitive to the variation of dissolved oxygen concentration in an aqueous buffer. A linear Stern-Volmer plot (Io/I ) 0.0347[O2] + 0.9910; r2 ) 0.9980) was obtained from the oxygen-sensitive film, where [O2] is the dissolved oxygen concentration in micromolar. The ratio of Ide/Iair was 8.5, where Ide and Iair are the fluorescence intensities of the oxygen-sensitive optode membrane on exposure to deoxygenated water and airsaturated water, respectively. To our knowledge, it is one of the most sensitive oxygen-sensitive optode membranes to date. The limit of detection was calculated to be 0.61 µM from the SternVolmer plot as that dissolved oxygen concentration which (45) Demas, J. N.; DeGraff, B. A. Anal. Chem. 1991, 63, 829A-837A. (46) McDonagh, C.; MacCraith, B. D.; McEvoy, A. K. Anal. Chem. 1998, 70, 45-50.

866

Analytical Chemistry, Vol. 74, No. 4, February 15, 2002

produced an analytical signal equal to 3 times the standard deviation of Io/I at zero value. The oxygen-sensitive optode membrane showed good repeatability (Ide/Iair ) 8.5 ( 0.2; n ) 10) when it was alternately exposed to deoxygenated water and air-saturated water. The response of the membrane to the dissolved oxygen is reversible and fast. The response and recovery times of the oxygen optode are less than 30 s and 3 min, respectively. The photostability of the oxygen-sensitive optode membrane is extremely good as well. There was no significant photodegradation when it was irradiated at 468 nm using an xenon lamp set at 70 W for 12 h. The luminescence intensity was stable throughout 12 h without any drift of signal. The compound in the silicone rubber film was extremely stable and could be stored for a long period (>300 days) without any degradation. Bienzyme Immobilization on Eggshell Membrane. Aspartame amperometric biosensors have been established and reported in the literature.22-24 Alcohol oxidase and R-chymotrypsin were immobilized or entrapped in a dialysis or polycarbonate membrane. Unfortunately, the stable lifetimes of their sensors were only 7-50 days due to the loss of activity of the alcohol oxidase. It has been reported that some biomaterials36-42 can be employed as platforms for the immobilization of enzymes, and the lifetimes of the immobilized enzymes were much extended. Therefore, we decided to employ eggshell membrane to immobilize both R-chymotrypsin and alcohol oxidase in order to extend the lifetime of the aspartame optical biosensor. Figure 2 and Figure 3 display the transmission electron micrographs and scanning electron micrographs of eggshell membranes with and without the immobilized enzymes, respectively. The scanning micrographs and transmission micrographs can show the surface and the internal structure of the eggshell membranes, respectively. Some of the enzymes were immobilized on the surface of the eggshell membrane while some entered into the net-veined structure of the eggshell membrane. It is clearly shown that the enzymes were successfully immobilized on the eggshell membrane, which could be used for the fabrication of aspartame optical biosensor. Response Behavior of Aspartame Optical Biosensor. The magnitude of the analytical signal of the aspartame biosensor is determined by the oxygen quenching constant, the oxygen concentration, the aspartame concentration inside the oxygen sensing membrane, the amounts or activities of R-chymotrypsin and alcohol oxidase on the eggshell membrane, and the temperature of the biosensor system. The typical response curve of the aspartame biosensor is shown in Figure 4. The relative signal change of the aspartame biosensor is much greater when the sample solution is at the stop flow mode than at the flowing mode. Thus, the stop flow mode was chosen to detect aspartame in sample solution. The rate change of the fluorescence intensity, R, is defined as

R ) (I - Io)/t

where Io and I represent the detected fluorescence signals from the biosensor exposed to buffer solution and aspartame solution, respectively, and t is the time duration. In most of the applications, t was taken as 4 min since a constant signal could be obtained only after quite a long time. The rate change of the signal is an

Figure 2. Transmission electron micrographs of the eggshell membrane. (a) Fresh eggshell membrane. (b) Eggshell membrane immobilized with enzyme. Key: (1) internal cavity; (2), (4) protein fiber; (3), (6) face boundary between membrane and air; (5) cavity filled with enzyme; (7) membrane surface immobilized with enzyme.

appropriate parameter to relate the concentration of aspartame because all isolated samples, after a sufficient period of time, become thermodynamically equilibrated. Consequently, the total signal change can even become independent of analyte concentration when one of the cosubstrates is not present in sufficient excess.47 A calibration curve is plotted using R against concentration of aspartame (Figure 5). Although the curve has a slight sigmoid slope, it shows a reasonable linearity (R ) 108.8[aspartame] - 5.593; r2 ) 0.9931) by using the linear-square fit method. The limit of detection was calculated to be 32 µM from the calibration plot as that aspartame concentration which produced an analytical signal equal to 3 times the standard deviation of R at zero value. The limit of detection can be lowered if a longer incubation time was used in the detection, but it will, of course, sacrifice the analysis time. The repeatability of the aspartame biosensor was good when it was alternately exposed to buffer solution and 0.57 mM aspartame solution for five times (R ) 58.1 ( 4.1; n ) 5). The response of the biosensor to aspartame is reversible by exposing to a running phosphate buffer for 10-15 min. (47) Wolfbeis, O. S. Anal. Chim. Acta 1991, 250, 181-201.

Figure 3. Scanning electron micrographs of the eggshell membrane. (a) Fresh eggshell membrane. (b) Eggshell membrane immobilized with enzyme. Key: (1) protein fiber; (2) protein fiber immobilized with enzyme.

The response of the biosensor was investigated by variation of the partial pressure of oxygen. A 0.60 mM standard aspartame solution was subject to 10%, 21%, 40%, and 60% oxygenation, respectively, and the rate changes of the fluorescence intensity were recorded. It was observed that the sensitivity of the biosensor increased when the dissolved oxygen content of the standard solution decreased. The response of the biosensor at 60% oxygenated solution was reduced to 6.9% of its value at 10% oxygenated solution. Since the response of the biosensor varies with the dissolved oxygen content of a solution, all the calibration aspartame standard solutions and samples were required to be airsaturated before analysis could be done. Effects of pH and Ionic Strength. The pH effect was studied over the range pH 4.0-9.6. Figure 6 shows the normalized rate change of the fluorescence intensity against pH when the biosensor was subject to 0.57 mM aspartame standard at various pH phosphate buffer solutions. The highest R value is reached when the pH value is 7.7. The optimum pH values are in the range 5.7-9.0. The effect of ionic strength on the response of the Analytical Chemistry, Vol. 74, No. 4, February 15, 2002

867

Figure 4. Typical response curves of the aspartame biosensor at excitation/emission wavelengths of 468/612 nm subject to various concentrations of aspartame at pH 6.9 phosphate buffers (25 mM): (0) 0.000, (1) 0.056, (2) 0.113, (3) 0.170, (4) 0.227, (5) 0.340, (6) 0.453, (7) 0.566, (8) 0.749, (9) 0.960, (10) 1.150, (11) 1.410, (12) 1.920, and (13) 3.070 mM aspartame.

Figure 5. Calibration plot of the aspartame biosensor at various aspartame concentrations. Plot of rate change of fluorescence intensity, R, against concentration of aspartame. Linear regression analysis: slope, 108.8; y-intercept, -5.593; r2 ) 0.9931.

biosensor was investigated by subjecting the biosensor to 0.75 mM aspartame standards containing 0-0.5 M KCl. It was found that the response decreased slightly with an average of -0.045%/ mM KCl upon the increase of KCl concentration. It is possible that increasing the ionic strength of a solution will slightly affect the enzymatic rates of the enzymes. Since the optimum pH values of the aspartame biosensor cover a wide range (pH 5.7-9.0) and the effect of ionic strength is small, deionized water can substitute for the buffer solution in the preparation of standards. Thus, the experimental procedures can be simplified. Effect of Temperature. It is well known that both the analytical performance of the enzyme-immobilized membrane and oxygen transducer are highly sensitive to variations of temperature. Higher temperature would result in a significant drop in 868 Analytical Chemistry, Vol. 74, No. 4, February 15, 2002

Figure 6. Effect of pH on the normalized rate change of fluorescence intensity of the aspartame biosensor upon exposure to 0.57 mM aspartame at various pH phosphate buffers.

lifetime and yield of fluorescence intensity and also decrease the Stern-Volmer quenching constant of the ruthenium(II) complex resulting in a decrease in sensitivity of the aspartame biosensor. However, raising the working temperature has a counteracting effect on the biosensor. The activity of an immobilized enzyme is governed by the kinetics of the enzymatic reaction. The reaction will be increased by raising the working temperature. Thus, a study of the effect of temperature on the aspartame biosensor was carried out over the range 12-44 °C (Figure 7). The normalized rate change of the fluorescence intensity for the aspartame biosensor upon exposure to 0.82 mM aspartame solution (pH 7.0) increased sharply with higher working temperatures. The possible reasons are that the enzymes can acquire higher activities at higher temperatures and subsequently give a faster signal change with oxygen consumption at a faster rate

Table 2. Determination of Aspartame Contents in Commercial Products Using the Proposed Aspartame Biosensor, HPLC, and Spectrophotometric Methods method sample

proposed

HPLCa

spectrophotometricb

Equalc Diet Coke Diet Pepsi Diet Seven-up Diet Sprite

3.31%d 0.62 mM 0.70 mM 2.60 mM 2.33 mM

3.28% 0.59 mM 0.68 mM 2.56 mM 2.27 mM

1.44% 0.52 mM 0.83 mM 0.76 mM 0.53 mM

a Reference 10. b Reference 3. c Low-calorie sweetener. d Declared value on the label is 3.67%.

Figure 7. Effect of temperature on the normalized rate change of fluorescence intensity of the aspartame biosensor upon exposure to 0.82 mM aspartame solution (pH 7.0) at different temperatures. Table 1. Interference Test of the Aspartame Biosensor on Exposure to Various Potential Interferences interference

concn (mM)

relative change of signal (%)

aspartame citric acid cyclamic acid ethanol D-fructose D-galactose D-glucose hydrogen peroxide DL-malic acid L-phenylalanine saccharin sodium benzoate sucrose

0.57 3.16 3.84 1.36 4.20 3.66 10 0.91 1.12 2.91 3.45 7.19 3.49

100 0.0 0.0 >100 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

during the enzymatic reactions of aspartame and methanol. Although the analytical sensitivity was highest at 44 °C, for practical purposes, temperatures lower than 44 °C are recommended to prolong the lifetime of the biosensor since enzyme can easily be denatured at a high working temperature. The response of the biosensor at 44 °C dropped to ∼10% of its initial value after standing overnight. Stability of the Aspartame Biosensor. The long-term stability of the biosensors was tested over 6- and 8-month periods at 23 and 4 °C, respectively. When the biosensor was stored in a refrigerator at 4 °C or ambient conditions at 23 °C and measured intermittently, the rate change of the fluorescence intensity of the biosensor on exposure to 0.57 mM aspartame was found to be above 95% of its initial value over these periods. There was no significant difference in stability between the aspartame biosensor kept at either 4 or 23 °C. By contrast, when a similar biosensor was fabricated by immobilizing the bienzyme onto a poly(vinyl chloride) membrane, the rate change of the fluorescence intensity of the biosensor on exposure to 0.57 mM aspartame dropped to 55% of its initial value over a 5-day duration. Obviously, the remarkable stability of the aspartame optical sensor consisting of R-chymotrypsin and alcohol oxidase immobilized on eggshell membrane is possibly related to the biological compatibility of the eggshell membrane with the enzymes. Eggshell membrane

is mainly composed of biological molecules and protein fibers, which may supply polycations to stabilize the enzymes.33 It can be imagined that three types of immobilized enzymes may occur within the eggshell membrane: The first type is the chemical cross-linking between an enzyme molecule with another enzyme molecule or an enzyme molecule with a protein fiber. The second type is the electrostatic interaction between an enzyme and the macromolecular net-vein of the eggshell membrane. And the third type is enzymes that are not cross-linked to the other enzymes or protein fibers. They move freely within the cavities of the eggshell membrane. Actually, the net-veined structure and the gas-permeable hydrophilic property of eggshell membrane can provide an excellent biological microenvironment for enzymes to stay together with some enzymatic reactions to take place. Interference Test. The interference test was performed with some common ingredients present in foodstuff samples. The purpose was to investigate the effects of the potential interferences on the response of the aspartame biosensor. The relative change of the fluorescence intensity of the aspartame biosensor in 4 min was assessed when it was exposed to the interferences (Table 1). The results showed that the aspartame biosensor had no response to citric acid (3.16 mM), cyclamic acid (3.84 mM), D-fructose (4.20 mM), D-galactose (3.66 mM), D-glucose (10 mM), hydrogen peroxide (0.91 mM), DL-malic acid (1.12 mM), Lphenylalanine (2.91 mM), saccharin (3.45 mM), sodium benzoate (7.19 mM), and sucrose (3.49 mM). However, ethanol exhibited a significant interference on the aspartame biosensor since the enzymatic reaction of the sensor involved the oxidation of the alcohol class of substrates to their aldehydes with a concomitant effect on the depletion of the oxygen level. Aspartame Determination in Foodstuffs. Food samples were bought from a local supermarket and used as testing samples. The pH of the testing sample was adjusted to ∼7.0 on addition of a few milliliters of 25 mM Na2HPO4 solution and then diluted from 2- to 5-fold with suitable phosphate buffer solution to yield a testing sample solution of pH 7.0. The rate change of the signal of each sample solution was measured and compared with that of a set of aspartame standard solutions. The results of the aspartame contents obtained from the proposed biosensor method were also compared with that of a spectrophotometric method3 and an HPLC method10 and are summarized in Table 2. The above results show that the biosensor agrees well with the HPLC method; however, it in general gives higher values compared with the spectrophotometric method. It is not surprising to observe this difference since this spectrophotometric method Analytical Chemistry, Vol. 74, No. 4, February 15, 2002

869

Table 3. Recovery Data for Commercial Products Using the Proposed Aspartame Biosensor concentration (mg/100 mL) sample Equala

Diet Coke

Diet Pepsi

Diet Seven-up

Diet Sprite

a

added

found

recovery (%)

2 5 10 20 2 5 10 20 2 5 10 20 2 5 10 20 2 5 10 20

2.05 4.89 9.95 19.83 2.15 5.12 9.85 21.02 2.11 5.08 10.31 19.84 2.09 4.92 9.75 18.98 2.11 5.21 10.13 21.12

102.5 97.8 99.5 99.2 107.5 102.4 98.5 105.1 105.5 101.6 103.1 99.2 104.5 98.4 97.5 94.9 105.5 104.2 101.3 105.6

Low-calorie sweetener.

usually gives a lower recovery. In addition, the proposed biosensing method gives a value closer (3.31%) to the claimed value (3.67%) than the spectrophotometric method (1.44%) in one of our chosen samples. The results of the recovery test of the samples

870

Analytical Chemistry, Vol. 74, No. 4, February 15, 2002

are summarized in Table 3. We were confident that the proposed aspartame biosensor could provide a reliable and accurate method for the determination of aspartame in alcohol-free food samples. CONCLUSION The aspartame optical biosensor has been successfully applied to determine the concentration of aspartame in some commercial products. It also exhibits excellent stability with a long lifetime of at least 8 months. We anticipate that the lifetime should be much longer as we are still in the progress of monitoring its lifetime monthly. Future work directed to the fabrication of other enzymesbased biosensors using eggshell membrane as the immobilization platform is anticipated to be promising. Similar work is in progress in our laboratory, and the results will be released soon. ACKNOWLEDGMENT The work described in this paper was partially supported by a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (project HKBU 2058/98P). Financial support of this research from HKBU (project FRG/96-97/II-54) is also gratefully acknowledged. This work was presented at the Fifth Asian Conference on Analytical Sciences, Xiamen University, 4-7 May 1999, Xiamen, P. R. China. Received for review September 1, 2000. Accepted March 20, 2001. AC001097A