Anal. Chem. 2002, 74, 6111-6113
Quantitation of Manganese by Use of an Electron Spin Resonance Method Kayoko Minakata* and Osamu Suzuki
Department of Legal Medicine, Hamamatsu University School of Medicine, 1-20-1 Handayama, Hamamatsu 431-3192, Japan
An electron spin resonance (ESR) measurement is described for the quantitation of manganese in biological materials. The content of Mn was measured at room temperature by using 5 µL of a solution of the tissue wetashed with HNO3. In the ESR spectrum, Mn(H2O)62+ showed the characteristic six lines, and interferences were not detected from other substances in the tissues. For these tissue samples, good agreement was obtained between the values obtained by the present method and those obtained by conventional flame atomic absorption spectroscopy. The concentrations of Mn in urine, blood, and most tissues were so low that enrichment of Mn by coprecipitation with ZnS was applied for these samples. It took, however, only 5 min to quantify Mn in 1 mL of urine. Manganese is an industrial toxin. Chronic poisoning occurs in miners exposed to Mn dust. The disease is characterized by an encephalitic-like picture and progresses to resemble Parkinson’s disease.1 Unfortunately, no report has been found for a colorimetric quantitation method that is sensitive as well as specific to Mn.2,3 The conventional atomic absorption spectroscopy (AAS) method requires 1 g of liver containing 1 µg of Mn/g of wet tissue, which is the highest concentration among soft tissues.4 Mn2+ is paramagnetic and can be detected by electron spin resonance (ESR).5,6 The application of quantitative ESR for measuring Mn2+ and several other metals was introduced in a review article by Burns and Flockhart,7 and the concentration of these metals in (sea-)water was quantitated.8-11 In the present work, application of the ESR method was studied for animal samples. For the
quantitation,7 the double-integrated value of the ESR spectrum (first-derivative signal) of the species is usually compared with that of another species whose spin content is known. Mn2+ solution, however, can be standardized by itself,7,9 because Mn takes only the +2 state in acid solution, although it has several valence states from -2 to +7.
* Corresponding author: (tel/fax) +81-53-435-2239; (e-mail) kminakat@ hama-med ac.jp. (1) Purdey, M. Med. Hypotheses 2000, 54, 278-306. (2) Paynter, D. I. Anal. Chem. 1979, 51, 2086-2089. (3) McCall, K. A.; Fierke, C. A. Anal. Biochem. 2000, 284, 307-315. (4) Shearer, D. A.; Cloutier, R. O.; Hidiroglou, M. J. Assoc. Off. Anal. Chem. 1976, 60, 155-159. (5) Minakata, K.; Suzuki, O.; Horio, F. Clin. Chem. 2001, 47, 1863-1865. (6) Reed, G. H.; Cohn, M. J. Biol. Chem. 1970, 245, 662-667. (7) Burns, D. T.; Flockhart, B. D. Philos. Trans. R. Soc., London Ser. A 1990, 333, 37-48. (8) Hayes, R. G.; Myers, R. J. J. Chem. Phys. 1964, 40, 877-882. (9) Burns, D. T.; Dalgarno, B. G.; Flockhart, B. D. Anal. Proc. 1985, 22, 2425. (10) Burns, D. T.; Dalgarno, B. G.; Flockhart, B. D. Analyst 1986, 111, 725726. (11) Burns, D. T.; Dalgarno, B. G.; Flockhart, B. D. Anal. Chim. Acta, 1989, 218, 93-100.
EXPERIMENTAL SECTION Apparatus. ESR measurement was performed by using a JEOL JES-FE2XG ESR spectrometer. The equipped cavity was ES-UCX2 with TE011 mode, and its sensitivity was 1.5 × 1010 spins (i.e., 2.5 × 10-14 mol)/10-4 T. As shown in Figure 1a, the ESR spectrum of Mn(H2O)62+ (a standard Mn solution in 1 M HNO3) was characterized by six lines.6 The g values of the third and fourth lines from the left were 2.037 and 1.979, respectively, and the hyperfine splitting between them was 9.6 mT. A modulation width of 2 mT, which was the maximum width of the instrument, was adopted to obtain the highest signal,7,11 because peak-to-peak width of first-derivative signal was 2.7 mT. A flat cell of 100 µL was often used for ESR measurement of aqueous samples.7,9-11 In flat cell, however, the signal of Mn2+ in either 0.01, 0.1, or 1 M HNO3
10.1021/ac020345v CCC: $22.00 Published on Web 10/29/2002
© 2002 American Chemical Society
Figure 1. ESR spectrum of 5 µL of 200 ppb Mn standard solution in 1 M HNO3 (a) and that of 5 µL of ashed liver solution (∼1 M HNO3) (b). The gain setting was 2.5 × 103 for both measurements.
Analytical Chemistry, Vol. 74, No. 23, December 1, 2002 6111
Figure 2. Comparison of tissue Mn concentrations determined by the proposed ESR method (y axis) and by the conventional flame atomic absorption spectroscopy method (x axis). Groups A-C are explained in the text.
solution did not increase proportionally to square root of microwave power (MP) even at ∼1 mW, due to dielectric loss7 of MP in the large volume of acid solution. On the contrary, the signal of a 5-µL solution in the hematocrit capillary placed into a 5-mmo.d. quartz ESR tube increased proportionally to the square root of MP at more than 100 mW of MP in either 0.01, 0.1, or 1 M HNO3 solution. The relative ratio of signals in either 0.01, 0.1, or 1 M HNO3 solution was 100, 96, and 80%, respectively, at any MP from 1 to 100 mW. An MP of 65 mW was chosen for the following measurements because noise level increased partially with the increase of MP. Gain setting could be varied from 1 to 104 according to signal height. The accuracy of the ESR method was compared with the AAS method measured on a Shimadzu AA6200. A temperature-controlled heat block was used for wet-ashing. Centrifugation at 6000g and a vortex mixer were used in the process of enrichment of Mn. Polypropylene tubes (0.5 and 1.5 mL) were obtained from Eppendorf (Netheler-Hinz-GmbH), pipets and pipet tips, from 0.1 µL to 1 mL, from Gilson, and a 20-µL quartz hematocrit capillary (60 mm long × 0.8 mm o.d.) from Drummond Scientific Co. For sealing the capillaries, a vinyl plastic putty (Critoseal, Oxford Labware), was used. Sample Preparation. Tissues were obtained from six untreated rats denoted as A, six rats treated with paraquat denoted as B, and six rats treated with paraquat in a mineral restricted diet denoted as C, respectively (Figure 2). Urine and blood were obtained from normal humans. Tissue and blood were wet-ashed with concentrated HNO3, and 5 µL of the diluted ashed solution of liver or kidney was put into a quartz capillary for ESR measurement. Chemicals. All chemicals were of atomic absorption grade or biochemical grade from Wako Pure Chemical Ltd. Ultrapure water having specific resistance of 18 MΩ‚cm was used. All glassware or plastics were soaked in concentrated HNO3 or 0.3 M HNO3, respectively, overnight and rinsed more than 10 times with ultrapure water. RESULTS AND DISCUSSION Comparison with AAS. The ESR spectrum of Mn in ashed solution of liver is shown in Figure 1b and was the same as that of a Mn standard solution (a), indicating no interference from 6112 Analytical Chemistry, Vol. 74, No. 23, December 1, 2002
Figure 3. Ratio of MnS coprecipitated with ZnS from 100 µL of 50 ppb Mn standard solution. Complete coprecipitation of MnS was obtained from 50 ppm Zn solution with 10-50 µL of (NH4)2S, from 100 ppm Zn solution with 10-70 µL of (NH4)2S, and from 200 to 500 ppm Zn solution with 5-100 µL of (NH4)2S, respectively.
other substances in ashed solution. The height of two peaks on the left was used for quantitation. The concentrations of Mn in liver and kidney from 18 rats were measured by the AAS method (x) and ESR method (y), where x and y denote the abscissa and ordinate of Figure 2. Those two concentrations were related with the equation y ) 0.96x + 0.04 with a coefficient of 0.96. The results shown in Figure 2 indicate good agreement between the two results. Enrichment of Mn. When the concentration of Mn was low in urine and diluted ashed solutions of tissues and blood, coprecipitation of MnS with ZnS was applied to enrich the Mn because the solubility of MnS was lower than ZnS and diamagnetic Zn did not interfere with the ESR measurement. Furthermore, we noticed previously by using the ESR method that ZnS contained a minute amount of MnS, indicating a strong affinity of MnS to ZnS. As the source of S, K2S crystal and (NH4)2S solution were compared, and (NH4)2S was chosen because it gave precipitates that were easy to handle. As the source of Zn, Zn solution in 0.1 M HNO3 of atomic absorption grade and Zn(NO3)2 crystals of high-purity grade were examined. Both were confirmed to give precipitates of ZnS with (NH4)2S, which did not contain MnS. When the pH of the Zn solution was between 2 and 6, precipitates of ZnS could be obtained efficiently with the addition of (NH4)2S. Therefore, the pH of the sample was adjusted between 2 and 6 using pH test paper. Only 0.1 µLthe solution was lost due to a single pH measurement. For urine, the pH was set around 2 at first to liberate Mn from bound proteins and cherators in urine. As shown in Figure 3, precipitates of MnS with (NH4)2S were not obtained from dilute Mn solutions without Zn. Using 100 µL of 50 ppb Mn standard solution at pH 2, suitable concentrations of Zn and (NH4)2S were examined by changing the concentration of Zn from 50 to 500 ppm and changing the amount of (NH4)2S from 5 to 100 µL. Complete coprecipitation of MnS was obtained with Zn at concentrations from 50 to 500 ppm and (NH4)2S from 10 to 50 µL. For 100 µL of ashed solution, 15 µL of 5 N NaOH was added first to adjust the pH to 2-6, and then Zn(NO3)2 was added. The recovery of Mn in liver and kidney after coprecipitation with ZnS was more than 95% of that of direct measurement of
ashed solution. Based on these findings, the proposed procedure of enrichment of Mn in ashed solution was as follows. When the sample volume is different from 100 µL, the amounts of other solutions should be changed proportionally. 1. Take 100 µL of either Mn standard solution in 1 M HNO3 or ashed sample solution (∼1 M HNO3), 15 µL of 5 M NaOH solution, and 2 µL of 5000 ppm Zn solution in an Eppendorf tube and mix them for 2 s by a vortex mixer. 2. Confirm the pH of the solution to be between 2 and 6 using pH test paper (If the pH of the solution is out of the range, adjust the pH using either 5 M NaOH or 5 M HNO3). 3. Add 30 µL of (NH4)2S and mix them for 2 s. 4. Centrifuge it at 6000g for 2 min. 5. Remove the supernatant, and add 7 µL of 1 M HNO3 solution to the precipitate and solubilize it. 6. Put 5 µL of the solution in a quartz capillary tube and seal the top and bottom of the capillary with putty. 7. Measure the ESR spectrum of the capillary containing either the standard or sample. For 1 mL of urine, adjust the pH to ∼2 and add 20 µL of 5000 ppm Zn and 200 µL of (NH4)2S. It took only 5 min to quantify Mn in urine and in dilute ashed solutions. The relative standard deviation for 10 replicate determinations was 5%. The recovery of 2 ng of Mn per gram added to either urine or blood and tissue before ashing is 95%. The limit of detection is 50 pg of Mn, and 500 pg (100 ppb, 5 µL) of Mn is required to keep the relative standard deviation for 10 replicates below 5%. The detection limit of Mn in dilute acid solution in a flat cell was reported to be 10-7 M.7,9,11 The sensitivity in a flat cell and that in the capillary placed into the quartz tube were compared for 1 ppm Mn in 0.01 M HNO3. The signal height of 100 µL of the solution in a flat cell at optimal 6.5 mW was the same as that of 2 µL of the solution in the capillary at 65 mW, indicating that the total spins in a flat cell required for the detection were 50 times higher than that in the capillary. Moreover, in the capillary, the signal height in 1 M HNO3 attained was 80% of that in 0.01 M
HNO3, whereas in a flat cell, detection of the signal in 1 M HNO3 was almost impossible. The present method requires 1 mg of wet liver and kidney, a few milligrams of heart, lung, and spleen, 100 µL of blood, and 1 mL of urine for the quantitation. AAS required 20 mL of blood and 1 g of liver.4 GFAAS reported in 1976 required 2 mL of blood,12 but the method was greatly improved in 1979 so that only 2 µL of plasma was enough.2 In the latter method, however, Mn was purified from tissues and blood partly to avoid the interferences from other substances.13 Using ordinal ashed sample containing large amounts of other substances, the quantitation level of GFAAS was 2500 pg (50 ppb, 50 µL).13 ICPES required 25 µL of blood and 1 mg of liver.14 Although our ESR method is comparable to GFAAS and ICPES in its sensitivity, the ESR method does not burn the sample,7 and it can be used for another purpose. The ashed sample is stable for more than a year. As shown in Figure 1, large amounts of other substances in ashed tissues did not interfere with the assay of Mn. Even though several strong cherators might be contained in the samples, Mn is liberated from these cherators in 1 M HNO3 solution. Due to the poor solubility of EDTA, no interference was confirmed for up to 0.05 M EDTA, 0.1 M citric acid, and 0.1 M oxalic acid solutions. We used 1 M HNO3 solution for quantitation because we used concentrated HNO3 for wet-ashing. The peak height of the Mn standard solution in either 1 M HCl, 1 M HClO4, or 1 M H3PO4 was the same but that in 1 M H2SO4 was 80% of that in 1 M HNO3.8 Received for review May 21, 2002. Accepted September 11, 2002. AC020345V (12) Buchet J. P.; Lauwerys, R.; Roels, H. Clin. Chim. Acta 1976, 73, 481-486. (13) Smeyers-Verbeke, J.; Michotte, Y.; Van den Winkel, P.; Massart, D. L. Anal. Chem. 1976, 48, 125-130. (14) Kniseley, R. N.; Fassel, V. A.; Butler, C. C. Clin. Chem. 1973, 19, 807812.
Analytical Chemistry, Vol. 74, No. 23, December 1, 2002
6113
