JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JANUARY 1994 VOL. 9 11 Silicon Measurement in Bone and Other Tissues by Electrothermal Atomic Absorption Spectrometry Huang Zhuoer Guangzhou Environmental Monitoring Center Guangzhou 5 10030 People's Republic of China A method for the determination of silicon in bone and soft tissues using electrothermal atomic absorption spectrometry is described. Small wet biopsy samples are digested with concentrated nitric acid at 90 "C. The atomization signal of silicon in pyrolytic graphite coated graphite tubes is markedly enhanced by the addition of a lanthanum-calcium mixture in the test solution. No L'vov platform was used. Ammonium dihydrogenphosphate is added to the soft tissue digestate solution to eliminate the interferences arising from the biological matrices while tartaric acid disodium salt is used as a chemical modifier for the analysis of bone digest.Concentrations of silicon in the test solutions are determined against an aqueous standard calibration curve. The characteristic mass is 37 pg (integrated absorbance signal equal to 0.0044 s). For the sample digestion liquids the within- and between-run relative standard deviations are < 5.5%. The accuracy of the method is evaluated by determining the recovery of silicon added to the sample solutions and the results are close to 100%. The detection limits for silicon in bone and other tissues are 0.90 and 0.14 pg g-' wet mass respectively. Examples of silicon contents found in bone brain kidney liver spleen and heart of laboratory rats are given.Keywords Silicon; bone and soft tissues; chemical modifier; electrothermal atomic absorption spectrometry After the work of Carlislel and Schwarz and MilneY2 the biological effects of silicon in humans have been under investi- g a t i ~ n ~ particularly in chronic haemodialysis and patients with Alzheimer's disease.+'' It has been demonstrated that silicon levels are elevated in plasma and various tissues in patients with chronic renal failure on haemodialysis.4-8 In Alzheimer's disease aluminium and silicon were found to be co-localized within tangle-bearing neurons and senile plaque c o r e ~ . ~ * ~ ' Some workers have focused attention on the chemis- try of aluminium and silicon associated with the neurological d i s e a ~ e . ' ~ . ~ ~ Some articles have been published on the effect of silicon deficiency on the mineral composition of bone14 and the role of silicon in medicine and biology." With this background it is necessary to measure silicon in biological fluids and tissues in order to study the biological and toxic effects of this element in man and experimental animals.In the literature several methods have been available for the determination of silicon in serum and urine,16" but only a few procedures have dealt with the measurement of this element in t i s s ~ e s . ~ . ~ ~ ~ ~ ~ Determination of silicon in tissues by d.c. arc atomic emission spectroscopy4 or neutron activation analysis20.21 suffers from high detection limits and high relative standard deviation. In addition the radiochemical methods have the drawback of requiring special chemical separation schemes to remove the interfering activities from the other major elements present in the samples.In the past decade the general approach to silicon examination in tissues has been utilization of various micro-techniques including electron probe X-ray microanalysis,22 scanning electron microscopy and energy dispersive X-ray a n a l y s i ~ ~ ~ ' ~ ~ ~ - ~ ~ and imaging ion micro~copy.~~ Although X-ray microanalysis is considered to be specific for the presence of silicon and has been used effectively to localize this element in it is difficult in its present state of development to obtain reliable results for the silicon content of tissue specimens. Electrothermal atomic absorption spectrometry (ETAAS) has become the method of choice for the determination of silicon in biological fluids'618 and should be considered for use in measuring this element in bone and soft tissues.Nevertheless by contrast with serum and urine various tissue specimens need to be digested to produce sample solutions prior to analysis by ETAAS. In the recent past several wet digestion methods have been used for pre-treatment of tissue sample^.^^^^ However little has been published on sample digestion for the determination of silicon in bone and soft tissues. On the other hand the tissue digestion liquids still contain complex organic and inorganic constituents that can severely interfere with the determination of silicon by ETAAS. The method presently developed for determining silicon in serum and urine33 could not be employed immediately in the analysis of tissue digestion solutions especially in the bone digestion samples owing to the presence of suppressive inter- ferences.This paper describes an improved method of sample preparation for accurate determination of trace amounts of silicon in biopsy samples by ETAAS. Experimental Apparatus All the measurements were carried out on a Perkin-Elmer Zeeman 3030 atomic absorption spectrometer equipped with an HGA-600 graphite atomizer an AS-60 autosampler an Anadex Silent Scribe printer and pyrolytic graphite coated graphite tubes. The hollow cathode lamp for silicon was used at a working current of 30 mA 251.6 nm spectral line and 0.2 nm bandwidth. Argon was used as the purge gas. Sample aliquots (30 pl) were injected into the furnace for analysis.The furnace programme is presented in Table 1. Signals were pro- cessed in the peak area mode. Materials and Reagents No glassware was used and doubly distilled water was used throughout the work. Polystyrene tubes (6 or 12 ml) poly- styrene sample cups and automatic pipettes with disposable Table 1 in tissue digestion solution Temperature programming of furnace for analysis of silicon Gas flow Step TemperaturerC Ramp time/s Hold time/s rate/ml min-' 1 120 5 50 300 2 160 5 5 300 3 400 5 5 300 4 1400 10 30 300 5 2500 0 2 0 6 2700 1 3 30012 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JANUARY 1994 VOL. 9 tips were used. All reagents used were of analytical-reagent grade or better. The silicon stock standard solution (1000 mg 1-I) as ammonium hexafluorosilicate in water was obtained from J.T. Baker (Phillipsburgh NJ USA). Diluents Three diluent solutions were prepared with constituents as follows (per litre) (1) 30 mg of La [as La(NO,),] 30 mg of Ca (as CaCl,) 1.5 g of NH4H2P04 0.5 g of Na,EDTA and 1 ml of concentrated nitric acid; (2) 30 mg of La 30 mg of Ca 10 g of NH4H2P04 and 0.5 g of Na,EDTA; and (3) 30 mg of La and 5 g of tartaric acid disodium salt. The diluents ( l ) (2) and (3) were used for the preparation of standards the dilution of soft tissue and bone digestion liquids respectively. These solutions have no effects on the absorbance value of blank or silicon-containing solutions. Contamination Control Because silicon is ubiquitous in the environment it is important to avoid contamination when dealing with samples.All items utilized during sample collection digestion and analysis were checked for contamination with silicon. Disposable 6 ml poly- styrene tubes and sample cups were rinsed with doubly distilled water and dried before use. The pipette tips used for dispensing sample solution were rinsed twice with water and then with sample solution once. Sample Collection and Digestion Samples were obtained from male Wistar rats (body mass 400-45Og) fed in an animal experiment as a control group. The animals were killed and the tissues were removed and washed with doubly distilled water and then weighed and stored in stoppered polystyrene tubes at -20°C prior to analysis. Bone samples were de-fatted with methanol and chloroform after removing the marrow with a strong stream of distilled water.Throughout the work great care was taken to avoid contamination. A wet digestion procedure was employed in destroying the sample tissues prior to analysis by ETAAS. In brief bone specimens (0.1-0.2 g) and soft tissue samples (0.5-1.0 g) were digested with 1 ml of concentrated nitric acid in the polystyrene tubes with screw caps for 4-6 h in an oven at 90°C. Screw caps were pierced with a needle to allow nitrous vapours to escape. The clear digest was diluted to 10 ml with water and then stored at 4 "C. Analytical Procedure Before analysis the sample digestion liquids were further diluted with the relevant diluents. Silicon was measured in soft tissue digestion liquids after 10-fold dilution with diluent (2) whereas the bone digestion solution was diluted 30-fold with diluent (3).The concentrations of silicon in the diluted samples were calculated against a calibration curve prepared by diluting the standard in diluent (1). An intermediate standard containing 200 pg 1-' of silicon was prepared by diluting the stock stan- dard with water. Working standards containing 0 10 20 50 and 100 pg 1-1 of silicon were prepared by diluting the inter- mediate standard with diluent (1). At the beginning of the analysis by ETAAS a blank solution i.e. diluent ( l ) was run 5-8 times to obtain a low and reproducible absorbance value and the spectrometer was zeroed on this value. A pyrolytic graphite coated graphite tube can be used for up to approximately 180 firings with stable sensitivity of the silicon measurement but aged tubes (above 220 firings) would encounter more matrix interference.Results and Discussion Sample Digestion The digestion procedure is an important step in trace element analysis of tissue by ETAAS. According to some biological samples can be digested with nitric acid at tempera- tures of 60-105 "C. In sample digestion for the determination of aluminium however Blotcky and Claa~en,~ have warned of A12C16 subliming from the digested sample matrix at tempera- tures above 80 "C. Precautions against analyte loss should therefore be taken with sample digestion for the determination of silicon in bone and soft tissues. With the proposed digestion procedure no loss of silicon was observed during sample digestion.Water bone and other samples with or without the addition of 5.0 or 10.0 pg of silicon respectively (in 50 pl of aqueous standards) were digested with 1 ml of nitric acid to evaluate the effects of digestion temperatures of 60 70 80 90 and 100 "C on the recovery of added silicon. Analytical recover- ies of added silicon in various biological materials (including bone brain kidney liver spleen and heart) were in the range 98-105% independent of the digestion temperatures tested suggesting that the proposed wet digestion method is suitable for the determination of silicon in bone and other tissues. Instrument Settings The optimum furnace temperature programme for the determi- nation of silicon in bone and other tissues is summarized in Table 1. It is our experience that a charring step at 400°C is necessary to obtain reproducible absorbance values for the same test sample.At this temperature the sample residue after drying could be scorched without sputtering. In the presence of lanthanum and calcium a pyrolysis step at 1400°C for 30 s is sufficient to obtain the maximum absorbance value for a given amount of silicon. The optimum temperature for atomiz- ation was found to be 2500°C. Matrix Interference and Chemical Modification Because of the extremely low levels of silicon in bone and soft tissues digested samples containing relatively high concen- trations of biological matrices are required to attain sufficient concentrations of the analyte. This causes serious problems in the analysis of the digest by ETAAS owing to severe matrix effects.Soft tissues During development work it was found that the presence of tissue matrices resulted in poor sensitivity and poor precision. These interferences are mostly caused by organic matrices which have not been completely destroyed after the digestion process. Although the mixture of lanthanum and calcium has been confirmed to be an excellent chemical modifier for the atomization of silicon in a graphite atomizer,33 interferences from the tissue matrices could not be eliminated by the addition of this mixture to samples. Consequently another chemical modifier must be used. In the determination of silicon in serum by ETAAS ammonium dihydrogenphosphate has been used effectively as a chemical modifier to eliminate interferences from organic matrices.33 This phosphate is also an efficient chemical modifier for use in the measurement of silicon in the digestate of soft tissue by ETAAS and the degree of modifi- cation is directly proportional to the concentrations of the phosphate in the test solutions.With a concentration of 1% m/v of ammonium dihydrogenphosphate the diluted sample containing 10 mg ml-I of tissue matrix can be determined for silicon with no matrix effects. Bone The bone digestate solution contains a high concentration of calcium phosphate. In analysis by ETAAS the presence ofJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JANUARY 1994 VOL. 9 13 considerable amounts of phosphate residue after drying gives rise to the risk of sputtering during thermal treatment. On the other hand incomplete decomposition of the sample matrices before atomization results in suppressive interferences on the silicon signal.Fig. l(b) and (c) shows the shapes of the atomiz- ation signals obtained by silicon added to digested bone samples with the addition of lanthanum. Clearly the bone matrix suppresses the silicon atomization signal and part of the signal is shifted to a higher temperature if the concentration of bone matrix in the solution is rather high. According to the chemistry of silicate and phosphate silicate can enter the phosphate particles in a molten state when treating a mixture of silicate and phosphate at high tempera- tures. Thus it is possible that the interference effect of bone matrix is due to the interaction of calcium phosphate (in considerable amounts) with silicate to form polycompounds (for example phosphorosilicate glass).The phosphorosilicate polycompounds are rather stable at high temperature ( 1400 "C) and have not been dissociated completely before the atomiz- ation cycle resulting in a suppression effect on the silicon atomization signal. Although ammonium dihydrogenphosphate is a good chemi- cal modifier for the determination of silicon in many biological materials by ETAAS as discussed above it is less useful for the digested bone samples. Indeed a suppressed silicon signal will be observed if the concentration of calcium in the test solution is higher than 60mg1-' (in the presence of pho~phate).~~ Previously in the determination of trace elements in sea- 0.5 ( b ) I 0.5 Q) ( C) P 4 0.25 I ( d ) 1 .o 1 0.5 I A 0 \ 0 1 .o Ti m e/s 2.0 Fig.1 Comparison of atomization signals obtained for 100 pg 1-' of Si (20 pl) added in water and bone digestion samples with the addition of 40 mg 1-l of La. (a) Aqueous solution containing 40 mg 1-' of Ca.; (b) and (c) diluted solutions containing 0.66 and 2.5 mg ml-' of bone matrix respectively; and ( d ) and (e) diluted solutions containing 0.66mgml-' of bone matrix with the addition of 1.0% m/v of Na4EDTA and 0.5% m/v tartaric acid disodium respectively. Values for the integrated signals are (a) 0.240; (b) 0.128; (c) 0.133; ( d ) 0.211; and (e) 0.261 s water by ETAAS it has been confirmed that the addition of some hydroxyl-acids can eliminate the suppressive interferences from high salt matrices.34 In the present study tartaric acid disodium salt Na,EDTA citric acid ascorbic acid and ammonium acetate were tested in order to examine their effects on the atomization of silicon added in bone digestate in the furnace in the presence of lanthanum.The results indicated that only the tartaric acid and EDTA salts among the reagents tested could substantially enhance the absorbance signal for silicon added in the digested bone samples. The shapes of silicon signals obtained in these instances are shown in Fig. 1 ( d ) and (e). From these figures it can be seen that the suppression effect of bone matrix on the silicon signal is eliminated by the addition of tartaric acid disodium or Na,EDTA but the signal obtained with the addition of tartaric acid disodium is more regular than that with the addition of Na,EDTA.On the other hand Fig. 2 presents the effects of the concentrations of these two reagents on the sensitivity of silicon measurement in bone digestion solution. The suitable concentrations of Na,EDTA in the sample solutions are in the range of 2-3% m/v whereas 0.5-4% m/v of tartaric acid disodium salt can be used with a higher sensitivity of silicon measurement. These results together indicate that tartaric acid disodium salt is superior to Na,EDTA in the modification effect on bone matrix. The concentration of 0.5% m/v of tartaric acid disodium salt in the diluted samples was selected in order to obtain a negligible blank value. Under this condition the diluted sample contain- ing up to 1.0 mg ml-' of bone matrix can be determined with quantitative recoveries of added silicon. With the observation mirror it was found that the residue of sample solution after drying in the presence of tartaric acid disodium salt or Na,EDTA was scorched at temperatures ranging from 350 to 400 "C and considerable amounts of foam materials were produced during this process.In contrast no similar phenomenon was observed when either citric acid ascorbic acid or ammonium acetate was added to the test solution. From these observations it is believed that the above mentioned modification effect is partly associated with the presence of foam materials during the charring process of organic reagents. The effect of tartaric acid disodium salt or Na,EDTA can be attributed mostly to the complexing reaction of the majority of calcium in the bone digest with the organic reagent instead of with PO or Si03.On the other hand the presence of foam materials during thermo-treatment can prevent the formation of refractory polyphosphates. The combination of these factors leads to the elimination of suppressive interferences from the bone matrix on the determination of silicon by ETAAS. Calibration Sensitivity and Detection Limits With the use of suitable chemical modifiers a variety of biological samples can be analysed for silicon by ETAAS with no interferences and aqueous standards can be used for (0 0.15 \ 1 0 (D e n P 0.10 m U CI E 0.05 e I A L - I I I I 0 1 2 3 4 Concentration (% w/v) Fig. 2 Effects of the concentrations of A tartaric acid disodium salt or B Na4EDTA on the sensitivity of silicon measurements (50 pg 1- of Si 20 pl) in the bone digestion samples (0.66 mg ml-' of bone matrix) containing 40 mg 1-' of La14 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JANUARY 1994 VOL.9 Table 2 Precision for determinations of silicon in bone and soft tissue digestion samples; results given as mean f SD Matrix concentration$/ Tissue mg ml-' Bone 1 0.66 2 0.42 3 0.58 Brain 1 5.3 2 6.2 3 4.8 Liver 1 6.9 2 5.8 3 4.4 Kidney 1 7.2 2 5.1 3 5.6 Within-run* Between-day? CSiISi 1- 9.0 1- 0.3 21.5 k 0.7 42.1 k 1.1 6.5 f 0.2 58.2 k 1.6 105f2.1 10.8 f 0.3 58.1 f 1.3 105? 1.2 16.2 t- 0.6 60.1 f 1.0 111 f 1.8 RSD(%) 3.3 3.3 2.6 3.1 2.7 ' 2.0 2.8 2.2 1.1 3.7 1.7 1.6 [Silt:' Pg 1- 9.1 f0.4 21.8f 1.2 41.8f 1.6 6.6 f 0.3 58.1 f 1.9 106 -t 2.8 11.0 & 0.4 57.8 f 2.3 106 & 2.6 16.5 f 0.8 59.3 f 1.3 112 f 2.8 RSD(%) 4.4 5.5 3.8 4.5 3.3 2.6 3.6 4.0 2.5 4.8 2.2 2.5 * n=10.t n=3 d = 10. 1 In the test solution. 9 1 no Si was added; 2 and 3 5.0 and 10.0 pg of Si were added to the samples before digestion respectively. calibration purposes. This was confirmed by a comparison of the values of the characteristic mass (m,) calculated from the slopes of the standard additions graphs in various biological matrices including bone brain liver lung heart kidney and spleen with that from a calibration graph for the aqueous standards. The measured values of rn were the same as that obtained for silicon in serum and urine,33 i.e. 37 pg gave an integrated absorbance signal with a net area of 0.0044s suggesting that the determinations were completely interference free.In all instances the calibration graphs were linear up to 200 pg 1-l of silicon in diluted solutions. Since the blank values of the diluents used for the prep- aration of standards and the dilution of tissue digestion liquids are negligible and the sensitivity of the silicon measurement in all instances are the same it is not necessary to match the concentrates of chemical modifiers in the standard and sample solutions. In addition the diluent used for the preparation of standards can also be used for the dilution of serum and urine samples33 and the latter diluted samples can be determined for Si against the same calibration graph as used in the present study. The detection limits (3a n= 10) for silicon in the diluted solutions of bone and other tissue digestion liquids were 0.6 and 0.7 pg l-' respectively corresponding to 0.90 and 0.14 pg g-' of silicon in the initial wet specimens of bone (0.2 g) and other soft tissues (0.5 g) respectively.These results were obtained by repeatedly measuring the same diluted solution in each case. These detection limits are low enough for the determination of silicon in biopsy samples. Precision and Accuracy Precision data for the determination of silicon in various biological materials are given in Table 2. The within-run and between-day variations were calculated from the results deter- mined for the sample digestion liquids and the standard spiked liquids over a period of 1 month. The variations are satisfactory at the concentrations studied.The accuracy of the method was more difficult to assess because there was no appropriate reference material available with a certified value for silicon in tissues. Nevertheless the recoveries of silicon added to various tissue specimens have been determined. Analytical recoveries were close to 100% (Table 3). Silicon Contents in Tissues of Laboratory Rats Using the proposed procedure the silicon contents in bone brain kidney liver spleen and heart samples from four labora- Table3 Analytical recovery of silicon added in bone and soft tissue digestion samples Range of matrix Tissue concentration*/mg ml- ' Recovery( YO)? Bone 0.48-0.96 99.2 1.7 Kidney 5.2- 10.1 101 2 1.4 Liver 5.8-9.8 loo+ 1.5 Spleen 5.5-10.2 101 1.6 Brain 4.6- 8.8 1002 1.2 Heart 6.2-9.6 look 1.0 * In the test solution. 7 Mean+SD determined by adding 10.0 or 20.0 pg 1-' of Si to the diluted samples and then reading the concentrations from an aqueous calibration curve; n = 6.tory rats were determined. The results are listed in Table 4. Although the number of animals tested is too limited to perform statistical analysis significant differences could be seen between the mean silicon levels in selected tissues. Briefly the highest concentration of silicon was found in the bone whereas the brain contained the lowest silicon content. Such a result is in agreement with those of previous studies on rats using ~ i l i c o n - 3 1 . ~ ~ ~ ~ Since the laboratory male rats were from the same control group batch in the animal experiment the animal- to-animal variance in the tissue silicon content shown in Table4 is normal and unlikely to be attributable to the analytical error.Conclusions The ETAAS technique provides a sensitive precise and accu- rate method for determining silicon in biological materials even at the extremely low concentrations found in bone and soft tissue specimens. The mixture of lanthanum-calcium- phosphate is a very specific chemical system that can be used Table 4 Silicon levels found in tissues of laboratory rats. Duplicate samples were determined for each tissue; results given in pg g-' of wet mass Animal No. Bone Brain Kidney Liver Spleen Heart 1 13.6 1.23 1.55 1.57 1.75 1.30 2 12.9 1.21 2.71 1.72 1.92 1.27 3 10.9 1.04 2.25 1.26 1.69 1.52 4 13.8 1.24 1.96 1.38 1.56 1.46 Mean value 12.8 1.18 2.12 1.48 1.73 1.39JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JANUARY 1994 VOL.9 15 effectively as the chemical modifier for the determination of silicon in various biological materials by ETAAS after nitric acid digestion or direct dilution (the digestion step is not necessary for serum and urine samples33). For the bone digest in addition it is necessary to add tartaric acid disodium salt to the test solution for the purpose of chemical modification. With the proposed procedure direct aqueous calibration is possible. The detection limits of the method described here are low enough for the determination of trace amounts of silicon in tissues and the method can be employed in the study of silicon metabolism in various organs. The author thanks Kim Solvang for her considerable editorial assistance in the preparation of this manuscript.1 2 3 4 5 6 7 8 9 10 11 12 13 References Carlisle E. M. Science 1972 178 619. Schwarz K. and Milne D. B. Nature (London) 1972 239 333. Silicon Biochemistry Ciba Foundation Symposium 121. Wiley Chichester 1986. Indraprasit S. Alexander G. V. and Gonick H. C. J. Chronic Dis. 1974 27 135. Mauras Y. Riberi P. Cartier F. and Allain P. Biomedicine 1980 33 228. Adler A. J. and Berlyne G. M. Nephron 1986 44 36. Hosokawa S. Morinaga M. Nishitani H. Maeda T. and Yoshida D. Trans. Am. Soc. Artif. Intern. Organs 1987 33 260. Gitelman H. J. Alderman F. R. and Perry S. J. Am. J. Kidney Dis. 1992 19 140. Perl D. P. and Brody A. R. Science 1980 208 297. Candy J. M. Oakley A. E. Klinowski J.Carpenter T. A Perry R. H. Atack J. R. Perry 2. K. Blessed G. Fairbairn A. and Edwardson J. A. Lancet 1986 1(8477) 354. Edwardson J. A. Oakley A. E. Taylor G. A McArthur F. K. Ward M. K. Bishop H. P. and Candy J. M. Adv. Neurol. 1990 51 223. Birchall J. D. and Chappell J. S. Clin. Chem. 1988 34 265. Bilinski H. Horvath L. and Tybojevic-Cepe M. Clin. Chem. 1992 38 2019. 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 Brossart B. Shuler T. R. and Nielsen F. H. Proc. N.D. Acad. Sci. 1990 44 95. Dobbie J. W. Scott. Med. J. 1982 27 1 . Berlyne G. M. and Caruso C. Clin. Chim. Acta. 1983 129 239. Gitelman H. J. and Alderman F. R. J . Anal. At. Spectrom. 1990 5 687. Holden A. J. Littlejohn D. and Fell G. S. Anal. Proc. 1992 29 260. Roberts N. B. and Williams P.Clin. Chem. 1990 36 1460. Velandia J. A. and Perkons A. K. J. Radioanal. Chem. 1974 20 715. Guzzi G. Pietra R. and Sabbioni E. J. Radioanal. Chem. 1976 34 35. McClure J. and Smith P. S. J. Pathol. 1984 142 293. Goligorsky M. S. Chaimovitz C. Nir Y. Rapoport J. Kol R. and Yehuda J. Miner. Electrolyte. Metab.. 1985 11 301. Funahashi A. Schlueter D. P. Pintar K. and Siegesmund K. A. Br. J. Int. Med. 1988 45 14. Jennings D. A. Morykwas M. J. Defranzo A. J. and Argenta L. C. Ann. Plast. Surg. 1991 27 553. Schmitt J. Dietzmann K. and Von-Bossanyi P. Acta Histochem. Suppl. 1992 42 319. Landis W. J. Lee D. D. Brenna J. T. Chandra S. and Morrison G. H. Calcif. Tissue Int. 1986 38 52. Bonilla E. Clin. Chem. 1978 24 471. D’Haese P. C. Van de Vyver F. L. De Wolff F. A and De Broe M. E. Clin. Chem. 1985 31 24. Van Ginkel M. F. Van der Voet G. B. and De Wolff F. A. Clin. Chem. 1990 36 658. Anderson J. R. and Reimert S. Analyst 1986 111 657. Blotcky A. J. and Claassen J. P. Anal. Chem. 1992 64 2910. Huang Z. E. Clin. Chem. submitted for publication. Zhang Z. X. Huang Z. E. Li G. K. and Yang X. H. Bull. Anal. Test. (Chinese) 1991 10(4) 1. Mehard C. W. and Volcani B. E. Bioinorg. Chem. 1975 5 107. Adler A. J. Etzion Z. and Berlyne G. M. Am. J. Physiol. 1986 251 E670. Paper 310341 5 A Received June 14 1993 Accepted August 31 1993