Improved Electrodeposited Iridium Oxide pH Sensor Fabricated on

Sol-Gel Deposition of Iridium Oxide for Biomedical Micro-Devices. Cuong Nguyen , Smitha Rao , Xuesong Yang , Souvik Dubey , Jeffrey Mays , Hung Cao ...
1 downloads 0 Views 157KB Size
Anal. Chem. 2003, 75, 1258-1266

Improved Electrodeposited Iridium Oxide pH Sensor Fabricated on Etched Titanium Substrates Sayed A. M. Marzouk

Department of Chemistry, Faculty of Science, United Arab Emirates University, P.O. Box 17551, Al-Ain, United Arab Emirates

In the present paper, the preparation and characterization of an improved solid-state pH sensor are described. The sensor is based on anodically electrodeposited iridium oxide film, as a pH-sensing layer. Merits of the present sensor include (i) excellent adhesion of the pH sensitive layer to the substrate, (ii) excellent reproducibility of sensor fabrication, (iii) faster preparation procedure, and (iv) low cost of the titanium substrate. These advantages are realized by combining acid-etched titanium as the electrode substrate with an optimized electrodeposition solution consisting of IrCl4 as an iridium source, hydrogen peroxide, potassium oxalate, and potassium carbonate. Heating the electrodeposition solution to 90 °C reduced the time required for solution development from ∼3 days to 10 min. The pH-sensing layer is protected with a layer of Nafion and a microporous polyester membrane. The improved sensor showed a super-Nernstian response (-73.7 ( 1.2 mV/pH unit) in the pH range of 1.5-11.5. The present pH sensor, fabricated in a tubular form, is used as a detector in a flow injection analysis (FIA) system for pH measurements. Optimization of the FIA experimental parameters resulted in a linear dependence of peak heights on the pH of the injected samples in the pH range of 2-11. Direct electrochemical deposition of iridium oxide was first introduced by Yamanaka1 for the fabrication of electrochromic display devices involving a procedure based on an alkaline iridium (IV) oxalate electrodeposition solution. Anodic decomposition of such an oxalate complex resulted in a concomitant deposition of iridium oxide at the surface of the anode. Despite the absence of publications dealing with such simple and direct electrochemical deposition of iridium oxide since it was first described,1 a decade later, few publications have been reported in the literature by different research groups.2-5 The author of this paper along with others have previously reported the pH sensing characteristics of such electrodeposited iridium oxide film (AEIROF) on sputtered platinum substrates.2 The fabrication of IrO2 films by electrodeposition has offered * Fax: +971(3)-7671291. E-mail: [email protected]. (1) Yamanaka, K. Jpn. J. Appl. Phys. 1989, 28, 632-637. (2) Marzouk, S. A. M.; Ufer, S.; Buck, R. P.; Johnson, T. A.; Dunlap, L. A.; Cascio, W. E. Anal. Chem. 1998, 70, 5054-5061. (3) Baur, J. E.; Spaine, T. W. J. Electroanal. Chem. 1998, 443, 208-216. (4) Petit, M. A.; Plichon, V. J. Electroanal. Chem. 1998, 444, 247-252. (5) Zhang, J. M.; Lin, C. J.; Feng, Z. D.; Tian, Z. W. J. Electroanal. Chem. 1998, 452, 235-240.

1258 Analytical Chemistry, Vol. 75, No. 6, March 15, 2003

several advantages over other frequently used techniques.6-11 Such advantages include (i) a more convenient and cheaper fabrication process than other techniques, such as reactive sputtering;7 (ii) a low-temperature process suitable for plastic substrates, in comparison to high-temperature treatments;8,9 (iii) potential for using cheaper substrates, in comparison to expensive iridium as base metal;6 and (iv) versatility of sensor shapes and designs made feasible with electrodeposition on substrates of different geometries. These advantages have allowed a convenient method of fabricating pH sensors on flexible substrates that were used successfully for continuous monitoring of extracellular pH of rabbit papillary muscle and whole swine heart physiological preparations during acute ischemia.2,12 In the same period, Baur and Spaine,3 attempted the modification of the original Yamanaka solution.1 However, their electrodeposition solution required larger amounts of expensive iridium salts and suffered from air sensitivity. pH sensors based on such a solution had a limited lifetime and showed severe break in the calibration plots at pH 6. No advantages over the Yamanaka solution1 were presented, and applications were limited to the monitoring of acid-base titrations. Two years later, the same authors with others extended the sensor applications to microscopic pH imaging.13 Petit and Plichon4 described anodic electrodeposition of iridium oxide films on SnO2-coated glass substrates from a deposition solution close to that originally described1 using K3IrCl6 instead of IrCl4 as the iridium source. This solution required 4 days for stabilization at 35 °C. The prepared electrodes were tested for color bleaching characteristics as electrochromic devices and not as pH sensors. On the other hand, no other apparent advantages were offered by this solution. Zhang et al.5 prepared micro-pH sensors by electrodepositing iridium oxide on platinum using a solution and procedure similar to that described by Yamanaka.1 IrCl3 was used as a source of iridium instead of IrCl4 without appropriate justification. The (6) Hitchman, M. L.; Ramanathan, S. Analyst 1988, 113, 35-39. (7) Katsube, T.; Lauks, I.; Zemel, J. M. Sens. Actuators 1982, 2, 399-410. (8) Ardizzone, S.; Carugati, A.; Trasatti, S. J. Electroanal. Chem. 1981, 126, 287-292. (9) Wang, M.; Yao, S.; Madou, M. Sens. Actuators, B 2002, 81, 313-315. (10) Suzuki, H.; Arakawa, H.; Sasaki, S.; Karube I. Anal. Chem. 1999, 71, 17371743. (11) Lee, I.; Whang, C.; Choi, K.; Choo, M.; Lee, Y. Biomaterials 2002, 23, 23752380. (12) Marzouk, S. A. M.; Buck, R. P.; Johnson, T. A.; Dunlap, L. A.; Cascio, W. E. Anal. Biochem. 2002, 308, 52-60. (13) Wipf, D. O.; Ge, F.; Spaine, T. W.; Baur, J. E. Anal. Chem. 2000, 72, 49214927. 10.1021/ac0261404 CCC: $25.00

© 2003 American Chemical Society Published on Web 02/19/2003

prepared sensors were applied to monitor pH changes during the bioceramic coating with calcium phosphate. In the previous works,2,12 the interest and focus were on the characterization of AEIROF/Pt as a pH sensor as well as its application in monitoring pH in physiological studies. Despite the success of such pH sensors in the intended applications, some limitations were encountered, such as (i) the relatively low adhesion of the pH-sensing layer to the smooth platinum substrates, which lowered the reproducibility of the fabrication process; (ii) the use of expensive platinum substrates; and (iii) the excessively long time (∼3 days) required for the preparation of electrodeposition solution. Such limitations were not solved in the parallel work presented by other groups.3-5,13 This conclusion, in addition to the continued and the growing general interest in pH sensors based on iridium oxide layers,14-19 provided the driving force for the present work. Therefore, the aims of the present work were 2-fold. The first aim was to find a substrate that promotes greater adhesion of the AEIROF to produce a more reliable pH sensor suitable for different applications and to enhance the reproducibility of the fabrication process. Such a substrate should also be cheaper than platinum previously used.2,12 The second aim was to shorten the time required to prepare the electrodeposition solution. The realization of the above-mentioned goals and successful application of the improved pH sensor as a detector in a pH flow injection analysis (FIA) system are presented. EXPERIMENTAL SECTION Materials and Reagents. Iridium(IV) chloride (IrCl4‚xH2O) (Catalog no. 12184); titanium rods, 6.4-mm diameter (Catalog no. 10393); titanium wire, 2.0-mm diameter (Catalog no. 10397); titanium foil, 0.89-mm thick (Catalog no. 10399); copper rods, 6.4mm diameter (Catalog no. 14126); gold wire, 0.25-mm diameter (Catalog no. 00725); silver rod, 3.17-mm diameter (Catalog no. 42758); tungsten rod, 1.5-mm diameter (Catalog no. 42233); zirconium wire, 1.0-mm diameter (Catalog no. 00939); tungsten wire, platinum-coated, 0.25 mm (Catalog no. 13630); nickel wire, 2.0 mm (Catalog no. 14188); cobalt wire, 2.0-mm diameter (Catalog no. 43290); nicke-chrome alloy, 6.4-mm diameter (Ni/Cr; 80:20 wt %, Catalog no. 42365); stainless steel wire, 1.0 mm (Catalog no. 40947); Hastelloy gauze, 20 mesh (Ni/Mo/Cr/Fe/W/Co/Mn 57.5:15.5:15.5:6:3.5:1.5:0.5 approximate wt %, Catalog no. 40938); and conductive carbon cement adhesive (Catalog no. 41212) were purchased from Alfa Aesar (Karlsruhe, Germany). All other chemicals were of analytical reagent grade; all solutions were prepared using distilled water. Instrumentation. A scanning potentiostat/galvanostat (Princeton Applied Research, EG&G, model 362) was used in galvanostatic current deposition of AEIROF and in cyclic voltammetry (CV) experiments. An Orion pH/mV meter (model 420) and a combination Ross glass electrode (Orion, Catalog no. 815600) (14) Dexter, S. C.; Chandrasekaran. P. Biofouling 2000, 15, 313-325. (15) Suzuki, H.; Hirakawa, T.; Sasaki, S.; Karube, I. Anal. Chim. Acta 2000, 405, 57-65. (16) Meyer, R. D.; Cogan, S. F.; Nguyen, T. H.; Rauh, R. D. IEEE Eng. Med. Biol. Soc. 2001, 9, 2-11. (17) Wang, M.; Yao, S.; Madou, M. Sens. Actuators, B 2002, 81, 313-315. (18) Lee, I.-S.; Whang, C.-N.; Choi, K.; Choo, M.-S.; Lee, Y.-H. Biomaterials 2002, 23, 2375-2380. (19) Bezbaruah, A. N.; Zhang, T. C. Anal. Chem. 2002, 74, 5726-5733.

were used in pH adjustment of buffer solutions and for reference pH measurements. A two-channel ADC-100 interface card (Pico Tech., U.K.) connected to a PC installed with PicoLog software (Pico Tech., U.K.) was used in digital X-Y and X-t recordings. A custom-made 32-channel data acquisition system was used for simultaneous recordings of up to 32 AEIROF-pH sensors against a Ag/AgCl common reference electrode. The data acquisition system was constructed from 32 high-input-impedance differential amplifiers (model INA116, Texas Instruments) and four 8-channel interface cards (model ADC-16, Pico Tech., U.K.), which provided a total of 32 channels for simultaneous potentiometric monitoring. The ADC-16 cards were connected via four serial ports to a PC installed with PicoLog software (Pico Tech., U.K.) used for real time data acquisition and display. Unless otherwise stated, all experiments were carried out in an air-conditioned laboratory at 23 ( 1 °C. Electrodeposition Solution. The solution was prepared by dissolving 75 mg of IrCl4‚xH2O in 50 mL of distilled water and stirring for ∼10 min in a 100-mL glass beaker. A 0.5-mL aliquot of 30% hydrogen peroxide (Riedel-deHae¨n, Catalog no. 18312) was added, and the solution was stirred for approximately another 10 min. Then a 365-mg portion of potassium oxalate hydrate (RiedeldeHae¨n, Catalog no. 32313) was added, and the solution was stirred for 10 min. The pH of the solution was raised slowly to pH 10.5 by addition of small portions of anhydrous potassium carbonate (BDH, Catalog no. 10196). The solution was then either allowed to stand for 2-3 days in an air-conditioned laboratory or was treated at relatively higher temperatures. The effect of temperature on the acceleration of solution development was tested by heating the freshly prepared solution for a given period on a preheated hot plate. A thermometer was used to monitor the solution temperature. Sensor Fabrication Procedure. Metals and alloys tested as electrode substrates were either in the form of a disk (area 0.32 cm2), wire (5-mm length), sheet (5-mm length, 5-mm width), or mesh (10-mm length, 10-mm height), depending on the form of the purchased metal or alloy. Titanium substrates were tested in different forms, including disks, wires, and sheets, which were connected to copper wires for electrical connection. Except for the electrode active area, all metallic parts of the electrode were isolated with nail polish coatings. The electrode’s surface was polished with fine polishing paper and then rinsed thoroughly with distilled water. For titanium substrates only, the electrode surface was roughened by etching in 70% sulfuric acid at 80 °C for 1-2 min. Teflon tape was used to protect the nail polish and parts of the electrode immersed in the acid except for the electrode active surface to be etched. After etching, the electrodes were thoroughly rinsed with distilled water, the Teflon tape was removed, and the electrode was thoroughly rinsed again with distilled water to remove any acid underneath. Next, the iridium oxide layer was deposited using galvanostatic current deposition in a 75-mL glass cell containing 50 mL of the deposition solution. A Hastelloy gauze alloy, 20 mesh (4-cm length × 4-cm width), was used as the cathode. A constant current density of 2 mA/cm2 for ∼2.5 min proved optimal for the formation of complete and well-adhered layers of dark greenish-blue AEIROF on titanium substrates. The produced pH sensors were thoroughly rinsed with distilled water and air-dried for a few minutes. Unless otherwise stated, the Analytical Chemistry, Vol. 75, No. 6, March 15, 2003

1259

AEIROF was coated with Nafion by addition of 2.5 and 0.75 µL of 5% Nafion solution (Aldrich, Catalog no. 52,708-4) for disk sensors of 6.4 and 3.0 mm diameter, respectively. Wire sensors were coated with Nafion by dipping in 5% Nafion solution twice and were then air-dried. Unless otherwise stated, the disk sensors were further protected with microporous polyester membranes (0.1µm pore size) (Poretics Corp., CA). The membrane was fixed in place by means of a Teflon cap and Teflon tape. Two sensors from each batch were tested for IrO2 cyclic voltammetry (CV) behavior in 0.5 M sulfuric acid. The potential was cycled between -0.25 and 1.25 V vs a saturated calomel electrode (SCE) at a scan rate of 50 mV/s. The sensors used for the CV experiments were discarded, and the rest of the sensors were soaked in a universal buffer of pH 7.0 for 2 days to achieve a drift in potential reading of less than 0.5 mV/hour. Sensor Calibration. The sensors were calibrated by measuring their potential response to a series of different buffer solutions covering the pH range of 2.0-11.0. Buffer solution of pH 2.0 was based on the NaCl-HCl mixture. Solutions of pH 3.0-9.0 were universal buffers containing 10 mM potassium hydrogen phthalate, 10 mM phosphate, and 10 mM Tris. Solutions of pH values 10.0 and 11.0 contained 50 mM sodium carbonate and 10 mM borax. All described buffer solutions contained 140 mM NaCl as ionic strength adjustor and were adjusted to the desired value by addition of HCl, NaOH, or both. Alternatively, sensors were calibrated by addition of HCl and NaOH to a universal buffer solution of composition (10 mM acetate, 10 mM phosphate, 10 mM Tris, 5 mM borax and 140 mM NaCl) in the pH range of 1.5 -11.8. pH-FIA Systems. A tubular pH detector was fabricated by drilling a channel (1 cm in length, 2-mm diameter) in a titanium rod (1 cm in length and 4-mm diameter). The obtained titanium tube was fixed at 90° through a 4-mm-diameter hole drilled into a 2-mm-thick copper foil (6 mm in width, 10 cm in length). The external surface of the tube and the copper connection were protected with Teflon tape, and the internal wall of the channel was etched by immersion in the sulfuric acid solution for 2 min. The pH-sensing layer was deposited onto the internal wall of the channel to form a tubular pH sensor. A Nafion coating was achieved by dipping the tube in 5% Nafion solution, and the excess solution inside the tube was removed by shaking it several times. The entire pH flow injection system was constructed as shown in Figure 1. An HPLC pump (Pye Unicam, model PU 4015, England) was used to propel the carrier solution to the tubular pH detector through PVC tubing at a given flow rate. An injection valve (Rheodyne, model 7125) equipped with a 1.0-mL injection loop was used for manual sample injection. The connection between the valve and the tubular detector was made of PVC tubing (0.8mm i.d., 10-cm length). A Ag/AgCl reference electrode (Orion, model 90-02) was placed downstream at the outlet of the tubular detector in a wall-jet arrangement. RESULTS AND DISCUSSIONS Testing Different Substrates. To realize the intended goals of this work, efforts were focused first on testing different substrates to identify possible alternatives to the expensive platinum. In this phase, the original Yamanaka solution1 was used. A variety of pure metals were tested for this purpose, such as Au, Ag, Ti, Cu, Ni, W, Zr, and Co. Some alloys were tested also for 1260 Analytical Chemistry, Vol. 75, No. 6, March 15, 2003

Figure 1. Schematic diagram of a single-channel-FIA setup: (1) carrier solution tank, (2) pump, (3) injection valve, (4) tubular detector, (5) copper connection, (6) reference electrode, (7) pH meter, (8) interface card, (9) PC, (10) waste solution, and (11) the position of the second injection valve used for injecting washing aliquots. The insert shows the expanded view of the tubular detector fixed by means of a copper foil, 2-mm thick, which also served as electrical contact. The carrier solution, exiting from the tubular detector, ensures the permanent contact between the reference electrode and the tubular detector.

this purpose, such as stainless steel, nickel-chrome and Hastelloy. Moreover, conductive carbon cement was tested for the same purpose as a coating on the copper disk electrodes. The most desired feature of the selected substrate is to promote excellent adhesion to the AEIROF. This should enhance the reproducibility of the sensor fabrication and increase the reliability of the sensor in challenging applications, such as continuous in vivo pH measurements. In this study, the criteria for the appropriate extent of adhesion were the resistance of deposits to peeling off the surface using adhesive tape, resistance to wiping with tissue paper, stability of the cyclic voltammograms known for IrO26 during several successive potential cycles, and the visual examination of blue AEIROF deposits at the conclusion of the experiment. The characteristic blue deposits of the AEIROF could not be obtained on cobalt, silver, and tungsten substrates using current densities in the range of 1-10 mA/cm2 for up to 10 min. Deposits obtained on zirconium at current densities in the range of 1-10 mA/cm2 for up to 20 min were incomplete and easily wiped from the surface using tissue paper. Therefore, no further studies were carried out using such substrates. Deposits on titanium, copper, gold, nickel-chrome, stainless steel, Hastelloy (2 mA/cm2 for 6 min), and carbon cement (10 mA/cm2 for 10 min) were obtained as complete blue layers. Except for titanium, stainless steel, and Hastelloy, the extent of adhesion of the AEIROF was inadequate, because the deposits were removed by wiping with tissue paper or adhesive tape. Moreover, cyclic voltammetry of AEIROF on copper and nickel-chrome resulted in the removal of deposits and substrate dissolution. Although deposits on stainless steel and Hastelloy resisted the peeling test, they did not tolerate the CV experiments in 0.5 M sulfuric acid, as indicated by the disappearance of the blue deposits, and naturally, the characteristic CVs for IrO2 were not obtained. Among all substrates tested, titanium provided a uniquely superior extent of adhesion to the AEIROF. In addition, titanium provided the advantages of being much less expensive than

Figure 2. Cyclic voltammograms for AEIROF in 0.5 M sulfuric acid at a scan rate of 50 mV/s. Oxide layer deposited on titanium disk substrate electrodes (6.4-mm diameter) from a 2-day-old solution: (A) titanium substrates polished with very fine polishing paper, 6-min deposition at current density of 2 mA/cm2; (B) titanium substrate etched in 70% sulfuric acid at 80°C for 1 min, deposition time is 2.5 min only using the same current density as in A.

platinum substrates and easily etched in acid solutions to provide even better adhesion. Moreover, it is commercially available in different forms such as rods, wires, foils, plates, tubes, and gauze, which may suit different sensor designs. The second best substrates, that is, stainless steel and Hastelloy, suffer from relatively limited adhesion, relatively higher price than titanium, not being easily etched to promote surface roughness, and the commercial unavailability of some forms. Therefore, among the studied substrates, titanium is suggested as an undisputed new substrate for the electrodeposition of the AEIROF. Titanium substrates polished with very fine polishing paper (P 800, Gelva Achilles, Holland) were used initially as substrates for electrodeposition of the AEIROF. Deposition for 6 min from a 2-day old solution yielded a dark greenish-blue layer, which showed a stable CV characteristic of the IrO2 layer,6 as shown in Figure 2A. The characteristic CV behavior of IrO2 layers was available in the literature for those layers prepared by electrochemical activation of iridium substrates;6,10 reactive sputtering;7 and to a lesser extent, by thermal treatment.8 In a previous related paper,2 a CV characteristic for IrO2 was not feasible with AEIROF/ Pt because of the limited adhesion to the substrate. However, in the present work, the characteristic CV for IrO2, similar to those reported previously,6,10 is obtained with AEIROF/titanium, as shown in Figure 2A. The quasi-reversible nature of the obtained CV was in agreement with those obtained for IrO2 by other techniques6,10 showing comparable high peak currents, that is, ∼3-4 mA/cm2

To further enhance the adhesion of the AEIROF to the titanium substrates, the roughening of the surface was attempted. The long etching time described for titanium substrates for the deposition of manganese oxide20 was proved inappropriate for the deposition of AEIROF. In the present work, a brief 1-2-min etching step using the previously suggested20 70% sulfuric acid at 80 °C promoted adequate surface roughening, which greatly enhanced the extent of adhesion to the AEIROF layer. A study of the effect of the etching step showed that surface roughening decreased the electrodeposition time required to obtain adequate deposits. The cyclic voltammograms of the AEIROF layers obtained with 6 and 2.5-min deposition on polished and 1-min etched titanium substrates are shown in Figure 2A and B, respectively. The obtained cyclic voltammograms are very similar in terms of the peak current and peak separation. This indicated that similar AEIROF could be obtained on etched titanium substrates at shorter deposition time in comparison to the polished smooth titanium. This can be attributed to a higher current deposition efficiency on the etched electrode surface. In addition to promoting greater adhesion, the etching process was more convenient and more reproducible than polishing especially for tubular detectors. Etching is even more necessary for possible miniature titanium electrodes or sputtered titanium layers where polishing is not adequate. The quasi-reversible nature of the obtained CV (shown in Figure 2) was in agreement with those obtained for IrO2 obtained by other techniques,6,10 showing comparable high peak currents, that is, ∼3-4 mA/cm2. As the thickness of the AEIROF layer increased by increasing the deposition time, the obtained cyclic voltammograms became less reversible, as shown in Figure 3. The peak separation (∆Ep) values were 69, 117, 156, and 222 mV for layers obtained using deposition times of 0.5, 1.5, 2.5, and 3.5 min, respectively. The corresponding ratios between anodic and cathodic peak currents were 1.05, 1.06, 1.1, and 1.3, respectively. A similar behavior was observed with increased thickness of the IrO2 layer by chemical activation of iridium in sulfuric acid.6 Repeated potential scans for 12 cycles resulted in identical cyclic voltammograms (data not shown). This confirmed the great adhesion of the AEIROF to the etched titanium substrates. Although more reversible CVs were obtained at short deposition times, more stable layers, which provided better pH sensing properties, were obtained at longer deposition times and a relatively longer etching time, that is, 2 min. Therefore, unless otherwise stated, 2.5-min deposition at current density of 2 mA/ cm2 on titanium substrates etched in 70% sulfuric acid at 80 °C for 2 min were used throughout the remainder of this work. Electrodeposition Solution. The solution described by Yamanaka1 contained iridium (IV) chloride, hydrogen peroxide, oxalic acid, and potassium carbonate. It was evident that changing the iridium salt was of limited value.3-5 In the present work, replacing IrCl4 with (NH4)2[IrCl6] proved inappropriate, and the solution did not show the characteristic color development from yellow to blue for up to 7 days at room temperature. The nature of the stabilization process of the electrodeposition solution that involves the blue color formation is not exactly known and could be attributed to the formation of an iridium-oxalate complex.1 (20) Izumiya, K.; Akiyama, H.; Kumagai, N.; Kawashima, A.; Hashimoto, K. Electrochim. Acta 1998, 43, 3303-3312.

Analytical Chemistry, Vol. 75, No. 6, March 15, 2003

1261

Figure 3. Cyclic voltammograms for AEIROF obtained by electrodeposition on titanium disk electrodes (6.4-mm diameter) etched for 1 min in sulfuric acid. The deposition solution was developed by heating at 90 °C for 10 min: (O) 0.5-, (b) 1.5-, (0) 2.5-, and (9) 3.5min deposition. Remainder of conditions are similar to those in Figure 2A.

This color development proved2,12 necessary to allow the anodic electrodeposition of iridium oxide to take place. Hence, IrCl4 was used throughout the rest of the work as the iridium source in the electrodeposition solution. The oxalate complexing agent played a critical role in the plating solution. In both the original description1 and all subsequent suggested solutions,2-5,12,13,16 oxalate was added in the form of oxalic acid. However, in the present work, an equivalent amount of oxalic acid salt, that is, potassium oxalate, was used instead of oxalic acid itself. This simple modification reduced the time required for the subsequent slow pH adjustment step from ∼40 min to