Anal. Chem. 2003, 75, 6753-6758
Silver-Induced Enhancement of Thiochrome-Based Peroxide Measurements Jianzhong Li, Purnendu K. Dasgupta,* and Guigen Li
Department of Chemistry and Biochemistry, Texas Tech University, Lubbock, Texas 79409-1061 Shoji Motomizu
Department of Chemistry, Faculty of Science, Okayama University, Tsushimanaka, Okayama 700, Japan
Hydrogen peroxide is an unusually important analyte because of the unique role it plays in a variety of areas: wastewater treatment,1 contaminated soil remediation,2 dentistry,3 paper and pulp manufacture,4 and clinical analysis,5,6 to name a few. It is the role of H2O2 and organic peroxides in atmospheric chemistry that has interested us most,7,8 and the topic has been extensively reviewed.9 The most common approach to measuring peroxides involves the oxidation of a substrate by peroxide, mediated by a peroxidase enzyme (or an enzyme mimic), to produce a colored/fluorescent/
electroactive compound. In recent years, thiamine (vitamin B1, TH) has proven to be a very attractive substrate for this purpose;7,8,10 the reasons for this include the availability of TH in a pure form at low cost, its compatibility with classical peroxidase enzymes (e.g., horseradish peroxidase, HRP) as well as various enzyme mimics, the high fluorescence intensity of the oxidized product (thiochrome, TC), the large separation between the excitation and emission maxima of TC, and the ability to excite such fluorescence with inexpensive light-emitting diode (LED) sources. Typically, hydrogen peroxide and organic peroxides react similarly in an assay that uses the HRP-TH reaction system. The discrimination between H2O2 and an organic peroxide such as methyl hydroperoxide (MHP) is provided by some agent that preferentially destroys H2O2. Enzymes such as catalase11 or inorganic surface catalysts such as MnO2 have typically been utilized. The latter is inexpensive, can be used as a disposable packed bed, and has therefore seen extensive use7,8,12 for this purpose. Hydrogen peroxide is also used in high concentrations in many sample digestion procedures. The excess H2O2 must generally be destroyed (or at least reduced in concentration) to alleviate bubble formation problems in measurement systems that handle the digested samples. If trace metal determination is an objective, MnO2 is not generally an acceptable destruction catalyst because of trace metals that leach from commercially available MnO2. One of us has been successfully using Ag2O for this purpose for some time; Ag2O is available in high purity. Although MnO2 does perform well as the catalytic differentiating agent between H2O2 and MHP, the catalyst bed must be replaced after several days of continuous use for atmospheric
* Corresponding author. E-mail:
[email protected]. (1) Wagner, M.; Brumelis, D.; Gehr, R. Water Environ. Res. 2002, 74, 33. (2) Pardieck, D. L.; Bouwer, E. J.; Stone, A. T. J. Contam. Hydrol. 1992, 9, 221. (3) Marshall, M. V.; Cancro, L. P.; Fischman, S. L. J. Periodontol. 1995, 66, 786. (4) Gratzl, J. S. Papier 1992, 46, V1. (5) MacLachlan, J.; Wotherspoon, A. T. L.; Ansell, R. O.; Brooks, C. J. W. J. Steroid Biochem. Mol. Biol. 2000, 72, 169. (6) Yamamoto, K.; Ohgaru, T.; Torimura, M.; Kinoshita, H.; Kano, K.; Ikeda, T. Anal. Chim. Acta 2000, 406, 201. (7) Li, J.; Dasgupta, P. K. Anal. Chem. 2000, 72, 5338. (8) Li, J.; Dasgupta, P. K.; Tarver, G. A. Anal. Chem. 2003, 75, 1203. (9) Jackson, A. V. Crit. Rev. Environ. Sci. Technol. 1995, 29, 175; Lee, M. H.; Heikes, B. G.; O’Sullivan, D. W. Atmos. Environ. 2000, 34, 3475.
(10) Li, J. Z.; Zhang, Z. J.; Li, L. Talanta 1994, 41, 1999. Zhu, Q. Z.; Li, Q. G.; Lu, J. Z.; Xu, J. G. Anal. Lett. 1996, 29, 1729. Zhu, Q. Z.; Zheng, X. Y.; Li, F.; Xu, J. G.; Liu, F. H.; Li, W. Y. Chem. J. Chin. Univ. 1997, 18, 1303. Zhu, Q. Z.; Liu, F. H.; Li, D. H.; Xu, J. G.; Su, W. J.; Huang, J. W. Anal. Chim. Acta 1998, 375, 177. Chen, Q. Y.; Li, D. H.; Zhu, Q. Z.; Zheng, H.; Xu, J. G. Anal. Lett. 1999, 32, 457. Chen, Q. Y.; Li, D. H.; Yang, H. H.; Zhu, Q. Z.; Zheng, H.; Xu, J. G. Analyst 1999, 124, 771. Zhu, Q. Z.; Yang, H. H.; Li, D. H.; Chen, Q. Y.; Xu, J. G. Analyst 2000, 125, 2260. Xu, C. L.; Zhang, Z. J. Anal. Sci. 2001, 17, 1449. Xu, C. L.; Li, B. X.; Zhang, Z. J. Chem. Anal. 2002, 47, 895. Li, Q. Y.; Morris, K. J.; Dasgupta, P. K.; Raimundo, I. M.; Temkin, H. Anal. Chim. Acta 2003, 479, 151. (11) Lazrus, A. L.; Kok, G. L.; Gitlin, S. N.; Lind, J. A. Anal. Chem. 1985, 72, 1053. (12) Dasgupta, P. K.; Hwang, H. Anal. Chem. 1985, 57, 1009. Hwang, H.; Dasgupta, P. K. Anal. Chem. 1986, 58, 1521.
Thiamine is presently one of the most attractive substrates used for sensitive fluorometric measurements of peroxides. Thiochrome (TC), a highly fluorescent product, is formed in enzyme-mediated oxidations. It is assumed that H2O2 is nearly quantitatively converted to TC. The reaction cannot differentiate H2O2 from many other peroxides such as methylhydroperoxide (MHP); to perform differential measurements, H2O2 can first be selectively destroyed by a suitable catalyst such as MnO2. In substituting Ag2O for MnO2 to accomplish the selective destruction of H2O2, we achieved the stated objective but were puzzled by a 3-fold increase in the MHP response in the presence of Ag2O. It was soon discovered that traces of dissolved Ag+ and Hg2+ can dramatically increase the yield of TC in this reaction from either H2O2 or MHP; the normal yield in fact is only 20%. We present here a reaction scheme and kinetic model that adequately describes this behavior and should provide a path to substantially increase the sensitivity of this important assay method.
10.1021/ac0349126 CCC: $25.00 Published on Web 10/29/2003
© 2003 American Chemical Society
Analytical Chemistry, Vol. 75, No. 23, December 1, 2003 6753
analysis because the activity slowly decreases. Whether this is due to poisoning by adventitious material or other reasons is unknown. We wanted to explore whether Ag2O would perform substantially better. This was studied using the HRP-TH reaction system. When used in identical bed dimensions and flow conditions,8 a miniature Ag2O column removes the H2O2 signal completely, as does MnO2. Under the same conditions with an MnO2 bed, an MHP sample produces 86% of the signal that is produced without the bed; i.e., the MnO2 bed destroys only 14% of the MHP.8 When an MHP sample was passed through an Ag2O bed, to our considerable surprise, the signal due to MHP increased by nearly a factor of 3 compared to that without the Ag2O bed. In this communication, we report the details and chemistry of this serendipitous discovery. EXPERIMENTAL SECTION Reagents. Both the hematin (Hmn)-TH and the HRP-TH reaction systems were studied; the first responds to H2O2 only, whereas the second responds to both H2O2 and MHP. The reagent concentrations and preparation and storage methods, including the use of different concentrations of TH in the two systems for optimum sensitivity, and the synthesis, purification, and assay of MHP all were as previously described.8 p-Hydroxyphenylacetic acid (PHPA) and p-cresol stock solutions (10 mM) were prepared by dissolving 152 mg of PHPA (Kodak) and 108 mg of p-cresol (ICN), respectively, in 100 mL of water and were stored under refrigeration. Stock solutions, 1 mM in metal ion concentration, were prepared by separately dissolving 15.6 mg of Ag2SO4, 27.2 mg of HgCl2, 33.1 mg of Pb(NO3)2, 18.3 mg of CdCl2, and 24.9 mg of CuSO4‚5H2O in 100 mL of water. All reagents were analytical grade. Ag2O (Aldrich, 99%) was packed in a 0.87 mm i.d. × 10 mm 20 ga PTFE tube as a miniature column for the selective destruction of H2O2. Thiochrome was obtained from Sigma (T-7891, >99%) and dissolved in pH 8.0 phosphate buffer to make solutions of desired concentrations that were stored under refrigeration. Equipment. A previously described automated flow injection analysis (FIA) system with an LED-liquid core waveguide fluorescence detector7 was used for flow injection studies. Batch fluorescence studies were conducted with a Shimadzu RF-540 spectrofluorophotometer. Methods. Briefly, in the FIA system, flowing streams of TH and HRP (or Hmn) were mixed in a tee, and the mixed stream was merged with a water carrier stream. The merged flow proceeded to the detector via a mixing coil (tR ≈ 3 min). MHP or H2O2 was injected through a six-port loop injector (200 µL) into the carrier. When desired, the Ag2O minicolumn was inserted immediately after the injection valve. For batch measurements, the order of reagent addition was as follows: TH, metal ion solution (when used), buffer solution (50 mM phosphate buffer, pH 8.0), HRP/Hmn, water (for a fixed constant final volume of 10 mL), and finally H2O2/MHP solution. Immediately after the solution was vortex mixed, a stop watch was started, the solution transferred to a quartz cuvette, and the fluorescence intensity measured as a function of time with excitation and emission wavelengths of 375 and 440 nm, respectively (both slits 2 nm). To avoid any photochemistry, discrete readings were taken, and the excitation shutter was closed after each measurement. 6754 Analytical Chemistry, Vol. 75, No. 23, December 1, 2003
Figure 1. Effect of the concentration of added Ag+ on the methylhydroperoxide (MHP) signal in an FIA system using the HRPcatalyzed oxidation of thiamine. Experimental conditions as in ref 8.
RESULTS AND DISCUSSION Silver Ion Enhances Formation of the Fluorescent Thiochrome. When 5 µM H2O2 was injected into the HRP-TH system and the Ag2O column was present, no detectable signal was observed, showing that Ag2O can effectively remove H2O2. When 5 µM MHP was injected into the same system with the Ag2O column present, the signal increased ∼3 times relative to that obtained in the absence of the Ag2O column. The solubility of silver compounds in water has been discussed at length.13 Therein, Maass estimates that the equilibrium solubility of silver oxide/ hydroxide in water should lead to ∼50-120 µM Ag+, depending on the degree of equilibration with CO2 (the CO2 equilibrium calculations are probably not strictly correct, but this is not critical for the present purposes as equilibrium is likely not established by passage through the short Ag2O column). We reasoned that some silver is dissolving in the water from the Ag2O column. We removed the Ag2O column but spiked different concentrations of Ag+ (as Ag2SO4) into the MHP solution before injection into the system. The fluorescence signal increased with the added Ag+ concentration from 10 to 100 µM and then reached a plateau value, as shown in Figure 1. (Because TH hydrochloride was used, there might have been some AgCl precipitation at the higher Ag+ concentrations, but this appears to have had minimal effect on the results, as will be discussed later for the batch method.) A very similar enhancement was observed with H2O2 instead of MHP, showing that the silver-ion-induced enhancement is common to both analytes. In contrast, addition of Ag+ to the desired final product, thiochrome, showed no evidence of fluorescence enhancement. Further, the fluorescence spectra of the product formed in the HRP-TH-H2O2 and the HRP-TH-MHP reactions with or without added Ag+ were identical to that of authentic thiochrome. These results conclusively demonstrated that (a) it is the dissolved silver that is responsible for the signal enhancement; (b) the effect is not due to any enhancement of thiochrome (13) http://www.silver-colloids.com/Papers/Solubility_Products.PDF.
Figure 2. Temporal development of fluorescence for (a) 10 µM Ag+ (0) (b) 10 µM Ag+ + 2 µM H2O2 (9), (c) 2 µM H2O2 (b). All solutions contain 2.9 mM TH and HRP (10 units/mL) as the catalyst, pH 8.0, 50 mM phosphate buffer.
fluorescence by Ag+; (c) the effect is likely due to increased thiochrome formation, as the fluorescence characteristics of the product formed are invariant; and (d) H2O2 is destroyed only by the solid catalyst, as there is no effect of dissolved Ag+, at least on the experimental time scale. Silver Changes the Product Yield. The FIA system basically allows for a fixed time measurement. The batch method was therefore used to monitor the kinetics of thiochrome formation; the results are shown in Figure 2. Initially, we assumed that, perhaps, in our FIA procedure, the reaction is quite incomplete and the enhancement is due to the increase in the rate of the reaction brought about by Ag+ acting as a cocatalyst/catalyst promoter. As should be apparent from Figure 2 (the solid lines shown are simple fits to a first-order rate process), the addition of 10 µM Ag+ actually results in a decrease of the rate constant. It is an increase not in the rate of the process but in the ultimate product yield that is responsible for the enhancement. Silver Exerts the Same Effect on the HRP-TH and Hmn-TH Systems. The Hmn-TH-H2O2 reaction system was affected by Ag+ in a qualitatively and quantitatively similar fashion as the HRP-TH-H2O2 system. These results indicate that the effect of silver is not related to the nature of the enzymatic catalyst; nor is it a matter of promoting the catalyst, as the amount of the final product formed changes. Silver Effect Is Specific to Thiamine. We tested the effect of Ag+ on two other commonly used substrates,14 p-hydroxyphenylacetic acid and p-cresol, with both HRP and Hmn as catalysts. In none of the four catalyst-substrate combinations was there any discernible effect of adding silver, either on the rate or on the final fluorescence intensity (see Figures S1 and S2 in the Supporting Information). Silver shows sensitivity enhancement only with thiamine; the effect is specific to the TH-TC oxidation process. (14) Genfa, Z.; Dasgupta, P. K. Anal. Chem. 1992, 64, 517.
Thiamine Oxidation Can Lead to Either Thiochrome or Thiamine Disulfide. The oxidation of TH (1) results in one of two products, fluorescent TC (3) or nonfluorescent thiamine disulfide (TDS, 6, Figure 3). The exact ratio of TC to TDS is affected by the pH,15 the solvent,16 and the nature of the oxidant. The chemistry of thiamine in basic solutions is complex and involves irreversible steps, and the precise products obtained can also depend on the order in which the different ingredients are added and the time between steps. Consequently, TC vs TDS production has been much debated and discussed.17-20 Hydrogen peroxide is capable of oxidizing TH without any enzymatic mediation; however, the product is almost exclusively TDS, and at the low micromolar levels involved here, the reaction is very slow. Sykes and Todd17 had originally argued that only one-electron oxidants such as ferricyanide can produce TC from TH; multiple electron oxidants, e.g., KMnO4, K2Cr2O7, MnO2, I3-, etc., result in TDS. The reaction with ferricyanide was the preHPLC U.S. Pharmacopeia method for determining thiamine content in vitamin supplements; it produces 67% TC from TH. The ability of Cu2+ to produce TC from TH more quantitatively has thus been touted as a superior alternative.21 However, the best reagents to produce TC quantitatively from TH are CNBr and Hg(II).22 The reactions are conducted in alkaline conditions. As such, it is not likely that Hg(II) is reduced to Hg(I). The number of electrons in the oxidant thus have little to do with the product. Indeed, in recent years, Perez-Ruiz et al.23 have extensively used phosphomolybdic or arsenomolybdic acid to oxidize TH to TC. The oxidation of TH by Hg(II) to TC has been particularly well studied, and kinetic methods for TH20,24 and an equilibrium method for Hg25 have been reported. Does Silver(I) Oxidize TH Directly? The Ag(I)-Ag(0) reduction potential is comparable to that of Hg(II)-Hg(0) and is, in fact, significantly higher than that of the Cu(II)-Cu(I) couple. If Cu(II) and Hg(II) can oxidize TH to TC, can the effect of Ag(I) simply be a direct oxidation? First, when TH is determined using Hg(II) or Cu(II), millimolar levels of the metal ion solutions are used. When low levels of Hg(II) are determined, by the Hg(II)-TH reaction, the reaction requires more than 80 min. Careful examination of Figure 2 will show that the Ag+-only blank does show a small increase with time but it is negligible relative to the signal from micromolar levels of H2O2. Rather than a redox mode, a coordinative mode of reaction of TH with silver that has been described in the literature is more likely. In alkaline solution, TH is known to form an adduct with Ag+, such that TH can be determined by addition of an aliquot of (15) Kawasaki, C. Modified thiamine compounds. In Vitamins and Hormones; Harris, R. S., Ed.; Academic Press: New York, 1963; Vol. 21. (16) Wostman, B. S.; Knight, P. L. Experientia 1960, 16, 500. (17) Sykes, P.; Todd, A. R. J. Chem. Soc. 1951, 534. (18) Nesbitt, P.; Sykes, P. J. Chem. Soc. 1954, 4585. (19) Maier, G. D.; Metzler, D. E. J. Am. Chem. Soc. 1957, 79, 4386. (20) Ryan, M. A.; Ingle, J. D. Anal. Chem. 1980, 52, 2177. (21) Perez-Ruiz, T.; Martinez-Lozano, C.; Tomas, V.; Ibarra, I. Talanta 1992, 39, 907. (22) Fujiwara, M.; Matsui, K. Anal. Chem. 1953, 25, 810; Edwin, E. E.; Jackman, R.; Hebert, N. Analyst 1975, 100, 689. (23) Perez-Ruiz, T.; Martinez-Lozano C.; Tomas, V.; Martin, J. Talanta 2001, 54, 989; Anal. Chim. Acta 2001, 442, 147; Anal. Chim. Acta 2001, 447, 229; Anal. Bioanal. Chem. 2002, 372, 387; Anal. Lett. 2002, 35, 1239. (24) Gonza´lez, V.; Rubio, S.; Go´mez-Hens, A.; Pe´rez-Bendito, D. Anal. Lett. 1988, 21, 993. (25) Holzbecker, J.; Ryan, D. E. Anal. Chim. Acta 1973, 64, 333.
Analytical Chemistry, Vol. 75, No. 23, December 1, 2003
6755
Figure 3. Proposed reaction schematic.
Ag+ followed by determination of the excess Ag+ 26 or by direct potentiometric titration with AgNO3 using a Ag+-sensitive electrode.27 Chemistry. On the basis of the above facts and the extant literature as cited above (especially refs 17-20 and 28), we propose the scenario shown in Figure 3. TH exists in tautomeric forms 1a or 1b. Form 1a can rapidly and reversibly lose a proton to form 2 (which has been variously called dihydrothiochrome17 or the cyclic intermediate19; this can be further deprotonated to form what has been termed the yellow thiol,17,19 but this compound is not involved in the present scenario at the presently used pH of 8 and will not be further discussed). Compound 2 is oxidized to produce TC (3); TC can be reduced to 2 by reducing agents such as dithionite. The 1b form of TH readily and irreversibly undergoes nucleophilic attack with OH- to form the “pseudobase”19 4, which is rapidly deprotonated to form the “colorless thiol”19 5 (an isolatable product). It is 5 that is oxidized to TDS (6). We believe that, as long as 2 is available, enzyme-activated H2O2 preferentially oxidizes 2, leading to the desired product 3. However, the top pathway in Figure 3 (1a f 2) is reversible, but the bottom pathway (1b f 4 f 5) is not. If TH is made alkaline, within a few minutes, very little 2 will actually remain. (26) Srividya, K.; Balasubramanian, N. Chem. Pharm. Bull. 1997, 45, 2100. (27) Hassan, S. S. M.; Elnemma, E.; Talanta 1989, 36, 1011; Feng, J. X.; Song, L. Y.; Li, S. J.; Qie, W. J. Chin. J. Anal. Chem. 1999, 27, 453. (28) Risinger, G. E.; Pell, F. E. Biochim Biophys. Acta 1965, 107, 374.
6756
Analytical Chemistry, Vol. 75, No. 23, December 1, 2003
Indeed, when Hg2+ is used as an oxidant to oxidize TH, compared to when Hg2+ is added before the solution is made alkaline, if the Hg2+ is added only 3 min after the solution is made alkaline,
