V O L U M E 23, NO. 4, A P R I L 1 9 5 1 interfere with the iodotellurite rnet’hod. Under the conditions described above for the separation of bismuth, the selenite is also reduced and the finely divided metallic selenium is ext,racted and suspended in the nonaqueous layer. Thus both selenium and bismuth may be removed in one step. If there is no bismuth to be removed, it is more efficient to extract the selenium wit,h chloroform than with the esteralcohol mixture. The iodohismuthite is not extracted with chloroform, however. For the determination of telluriuni when both bismuth and selenium are present, the following procedure is recommended.
To a 60-ml. separatory funnel add approximately 50 ml. of a solution containing 0.05 to 0.2 mg. of tellurium and either bismuth or selenium or both. Make the solution slightly acid (about 0.05 X hydrochloric acid) and add 2 ml. of 2 M potassium iodide. Allow to stand for 10 minutes. Extract with 5-ml. portions of amyl alcohol-ethyl acetate ( 3 to 1) until colorless. Transfer the aqueous layer to a 100-ml. volumetric flask. Add 20 ml. of 2 11.1 potassium iodide and 10 ml. of 2 -V hydrochloric acid, dilute to the mark, and read the per cent transmittance at 335 mp. Results of analyses carrird out by the above procedure are shown in Table IV. Iron and Copper. Both ferric and cupric ions oxidize iodide to iodine in acidic medium and thus interfere in the tellurium determination. Tellurium(1T’) may be separated from these ions by rpduction with stannous chloride to the metallic state and subsequent filtration. The tellurium may then be oxidized back to tellurite with nitric a d and determined by the iodotellurite method. To a niist,ure containing 0.05 to 2.0 mg. of telluriuni(1V) and irou and/or copper, add 2 ml. of 1 31 C.P. stannous chloride in 12 &Vhydrochloric acid. Let it stand for 10 minutes to assure complete reduction. Filter off the finely divided tellurium on asbestos with a suitable microfilter (the Kirk calcium-type microfilter, Microchemiral Sperialties Co., Berkeley, Calif., was successfully used in developing the method). Dissolve the met,allic tellurium by adding slowly 5 ml.of boiling vonceiitrated nitric acid to the filter. Collect the acid filtrate in a tuhe at the hottom of the filter, rinse the filter with three 3-ml. portions of 1 &Vnitric acid, and hoil the filtrate for 3 minutes to espel oxides of nitrogen. C‘ool and just nwtralizr to tlic ~~lic~iiolphthalein end point with
653 ammonium hydroxide. Transfer to a 50-ml. volumetric flask, add 10 ml. of 2 M potassium iodide and 5 ml. of 2 N hydrochloric acid, dilute to volume with distilled water, and read the per cent transmittance a t 335 mp. The effectiveness of this separation is shown in Table V. This method should be applicable to separation from a large number of other metals, including bismuth and arsenic.
Table V.
Separation of Tellurium from I r o n arid Copper
(In each case 0.05 m g . T e was taken. Separation was effected by the stannous chloride method and final determination made by the iodotelliirite method. Fe is added as FeCla and Cu as CIISOI) Fr Added. Cu Added. Te I:ound, XIK. >I& ME. 0.00
0.00
0 0500
0.00
1 .oo
0,0505 0.0510
1.00
0 00
0.0510 0.0500
1.00
1 00
0.0505 0.0.500
Other Ions. Ions that oxidized iodidt:, such as nitrite, chromate, and arsenate, should be avoided. Ions that form insoluble iodides, such as silver, lead, and niercurous, should also be absent. .ilthough they form stable iodide complexes, cadmium, mercuric. and zinc were found to be without, effect in concent,rations up to 500 p.p.m. Other ions tested and found to I)(, without effect are nickelous, cobaltous, nitrate, and sulfate. LlTERATURE CITED
Archibald, E. H., “Preparation of Pure Inorgaiiic Compounds,” New York, John Wiley & Sons. 1932.
Dehkio. R. H., ANAL.CHEM.,20, 488 (1948). Steinberg, H. H., Massari, S. C , Miner. A. C., and Rink. R., J . I d . Hug. Tozicol., 24, 183 (1943). (4) Volkov, S. T., Zauodskaya Lab., 5, 1429 (1936). ( 5 ) Zemel, V. K., Ibid., 5, 1433 (1936).
RECLIVED October 9, 1950.
Estimation of Gold in Biological Materials S I M U E L NATELSOSl ~ Y D JOSEPtI L. Z U C K E R M W Jewish Hospital of Brooklyn, Rrooklyn, Y.Y . Gold compounds have been used in the treatment of arthritis. Two cases of gold poisoning during treatment have heen encountered by the authors. Because no method was available for the estimation of gold i n the minute amounts found in biological fluids and tissues, a method w-as devised by the authors. By this procedure 0.2 microgram of gold can be detected. The entire procedure can be
T
HE p u r p o ~of this inwstigation was to develop a method for thc. quantitative estimation of gold in biological fluids, tissues, and feces. Gold Compounds have been used for some time in the treatment of rhrumatoid arthritis (5, 8). 9 o t infrequently, t k c manifestations occur owing to such treatment. During t,he past year two patirnts have been admitted to this hospital suffering from exfoliative dermat,itis following therapy with gold salts. The use of such detoxifying agents as British .intiI,ewisit~has been suggested for thc treatment of t,his condition ( 5 ) . 1
Present address, Rockford 1Iemorial Hospital, Rockford, Ill.
carried out in one test tuhe. ‘The colorimetric reaction of p-diethylaminobenzal-rhodanine w-ith gold salts is employed. Results of tests on the blood, urine, and tissue of patients suffering from exfoliative dermatitis following treatment with gold salts are given. Thus, a method is abailable for following the gold levels in the blood and urine during treatment with gold detoxifying agents.
For effective evaluation of treatment with gold salts and the detoxification of gold in the body, a quantitative indication of gold levels in the blood, urine, and stool is necessary. Gold levels in tissue are also of interest for research and post-mortem studies. The available methods are not sensitive enough to determine gold in microgram and submicrogram quantities. For example, Block ( 1 ) reported a sensitivity in the range of 5 t o 30 microgranis in blood. Jamieson reported a sensitivity of 5 micrograms in urine (6). Consequently, the authors developed a nicthod of gold analysis of greater sensitivity.
654
ANALYTICAL CHEMISTRY
This method depends on the sensitive color reaction reported by Sandell (7). The sensitivity of the reaction waa improved by employing a 5-cm. light path (9)and developing the color in 3.8 ml. of solution as against a 1-cm. light path for 4.5 ml. of solution in the original. These changes increased the sensitivity by a factor of 6. The sensitivity was further increased by investigating the absorption spectra of the color and measuring it a t the wave length of maximum absorption in a spectrophotometer.
Figure 1. Change of Optical Density with Gold Content of Gold-p-Diethylaminobenzal Rhodanine Complex 3.RmI. volume with 5-cm. light path
The method for isolating the gold was improved by using a different ashing technique. The procedure was simplified by employing centrifugation instead of filtration. Thus the whole procedure is carried out in one reaction vessel. This change permits a number of simultaneous determinations to be carried out with relative eass. It requires small amounts of materials-1 ml. of serum or whole blood, or 1 gram of tissue. It is sensitive t,o less than 0.2 microgram of gold. The possibilities of mechanical loss are minimized and only equipment usually found in the laboratory is requiied. Organic inaterial is removed by drying in an oven a t 100° C.; and then ashing with a small amount of sulfuric acid in a muffle furnace a t 550" C. The remaining carbon is removed by adding nitric acid and taking to dryness in an electric oven a t 100" C. The nitric acid that remains is removed by the addition of hydrochloric acid which is t,hen evaporated in the oven, The gold is taken up in hydrochloric acid and bromine water. Tellurium tetrachloride is added, followed by stannous chloride which reduces both the tellurium and the gold salts to the metallic state so that they coprecipitate. This precipitate is centrifuged, t,he supernatant is aspirat,ed, arid the residue is washed with dilute hydrochloric acid. The hydrochloric acid is driven off in the oven and the residue is dissolved in aqua regia. This is then dried and t,he residue is redissolved in dilute acid, the color-forming reagent and sodium fluoride are added, and the color is read in a photoelectric colorimeter after a suitable time interval ( 6 ) . The absorption spectrum of the color produced shows a broad peak between 490 and 510 mp. The color is therefore read a t 500 mp as a convenient, midpoint. The color obeys Beer's law. until a saturated solution of the chelate compound is reached. Sincc: thischelate is only slightly soluble in water, the straight portion of the density against gold curve extends until a concentration of approximately 1 microgram per ml. is reached. 4 t , higher concentrations, the curve bends over so that' increasing the gold concentration does not uniformly increase the density. In order to avoid this difficulty the amount of sample chosen should contain no more than 3 micrograms of gold when the color is developed in 3 ml. of solution. Since the densit'y of the color is sensitive to changes in pH (7) it is necessary to control carefully the concentration of the final acid. The same acid solution must be used t o make up the stand-
ard curve as is used t o dissolve the gold in the final step of the determination. The aqua regia in the final step must be removed as completely as possible in order not to influence the acidity of the final solution. Figure 1 is a curve obtained by plotting the concentration of gold against the reading in a Coleman Model 11 spectrophotometer. REAGENTS
Sulfuric Acid. Concentrated analytical grade. Nitric Acid. Concentrated analytical grade. Hydrochloric Acid. Concentrated analytical grade. For the 2 N solution, concentrated hydrochloric acid is diluted with distilled water. The 0.1 A' solution is made by accurately diluting the concentrated acid. The strength of this acid affects the density of the color. Bromine Water. Saturated solution of bromine in distilled water. Tellurium Tetrachloride Solution. Tellurium (100 mg.) is treated with 2 ml. of concentrated nitric acid and evaporated to dryness. The residue is treated with 1 ml. of concentrated hydrochloric acid and is again evaporated to dryness. The residue from this is dissolved in 10 ml. of concentrated hydrochloric acid and diluted to a total volume of 100 ml. Stannous Chloride Solution. Twenty grams of the dihydrate are dissolved in 100 ml. of 2 S hydrochloric acid. The solution must be filtered. Aqua Regia. One part of concentrated nitric acid is mixed with three parts of concentrated hydrochloric acid. Sodium Fluoride. To prepare a 1% solution, 1 gram of sdliium fluoride is dissolved in 100 ml. of distilled water. Color-Forming Reagent. (This material may be obtained from Adam Laboratories, 106-12 Jamaica Ave., Richmond Hill 18, 5 , Y.) p-Diethylaniinobenzal-rhodanine (15 mg.) is dissolved in 100 ml. of 95% ethyl alcohol. To completely dissolve the dye it is necessary to heat the mixture on a water bath for a few minutes. The dye solution must be kept cold to prevent decomposition. If pure dye is used the solution may be kept in the cold for a t least 2 weeks. The impure dye does not keep for more than 2 days when dissolved. The dye may be recrystallized from glacial acetic acid. Stock Gold Standard. A solution of gold chloride in 0.1 S hydrochloric acid is made up to contain-1 mg. of gold per ml. The solution may be made by dissolving the contents of a standard ampoule of gold chloride crystals (as used in making standards for the "gold number" test) in a suitable amount of acid. The solution should be kept in a brown bottle in the cold. This standard is stable for many m o n t h when kept cold. Working Gold Standard. The stock standard is diluted to contain 0.5 microgram of gold per ml. with 0.1 N hydrochloric acid. This solution is not stable and must be made up immediately before use. This working standard often contains little or no gold within a few hours after it has been made. PRELlMINARY TREATMENT
Whole Blood, Blood Serum, Tissue, or Feces. Whole blood or serum i l ml.) or 1 gram of tissue or dried feces is measured directly into a double thickness, borosilicate glass tube of about, 50- to 75-nil. capacity. A~pproximately 0.1 ml. (2 drops) of concentrated sulfuric acid is added and t,he water is evaporated in an electric oven at 120" C. Urine. 1-rine (100 m1.j is acidified with 10 ml. of concentrated hydrochloric acid. This solution is boiled for 20 minutes t,o hydrolyze any proteins which may be present. Then 5 ml. of tellurium ktrachloride solution are added, followed by 50 ml. of stannous chloride solution. The solution is heated on a steam bath for 0.5 hour, The precipitate is centrifuged in a 250-ml. centrifuge tube. The supernatant is discarded and the precipitate is mixed with 50 ml. of 2 N hydrochloric acid and recentrifuged. After discarding the supernatant,, the residue is treated with 2 drops of concentrated sulfuric acid and the water is removed from the mix by evaporating a t 120" C. Ashing is carried out in the 250-ml. centrifuge tube as described below. ISOLATION OF GOLD
b
The residue from the whole blood, blood serum, tissue, feces, or urine is ashed in a muffle furnace a t 550' C. for about 2 hours or until almost all the carbon has disappeared. When the ashing is completed the material looks grayish white with fiome particles of carbon still present, although it may have a slight pinkish tinge. If ashing is carried out a t too high a temperature, in glass, low TP-
V O L U M E 23, NO. 4, A P R I L 1 9 5 1 sults will b~ obtained. This is probably due t o a reaction between t,he gold and the glass. -1fter ashing, the tube is allowed to cool to room teniperature and concentrated nitric acid is added, 5 ml. if the residue is from blood, serum, tissue, or feces, and 10 ml. if the residue is from urine. The tubes are placed in an electric oven a t 100" C. to dry. Concentrated hydrochloric acid (5 ml.) is added and the evaporation is repeated. All the carbon should now be oxidized and the residue should appear white. The residue is taken up in 3 ml. of concentrated hydrochloric acid, 3 ml. of water, and 2 nil. of saturated bromine water. It is then heated for 0.5 hour on a boiling water bath. This step should dissolve all the material in the tube. A white precipitate of telluric acid in the tubes from urine is occasionally observed. This may be ignored for it does not. hold back the gold. From this point on the procedure is the same for all the biological materials. The solution is cooled to room temperature and 0.2 ml. of the tellurium reagent is added. The solut,ion is mixed and 6 ml. of stannous chloride solution are added. A black precipitate of tellurium forms. The tube is heated on a boiling water bath for 0.5 hour. It is then centrifuged a t 2500 r.p.ni. for 15 minutes and t'he supernatant is aspirated. The precipitate is washed twice \Tith 2 N hydrochloric acid, with centrifuging and aspirating of the supernatant each time. The aspirations should be performed with a rubber bulb attached to a glass tube so that control is maint,ained a t all times. The \\-ashing is done with approximately 10 nil. of acid each time. There is a t'endency for some of the precipitate to float on top of the liquid after centrifugat,ion. Care should be exercised in order not to lose this material. This is easily done hy aspirating from below the surface of the liquid. The precipitate is dissolved i n 1 ml. of aqua regia and dried a t 60" C. In order to save time, the misture is first evaporated a t a temperature of 90" to 100" C. until it is nearly dry. I t is then placed in :in oven at 60" C', uritil completely dry.
655 Table I. No. of Samples 8 8 8 32 8 4
2 2
Results of 72 Recoveries of Gold in Serum and Whole Blood Gold Added, Y 0.50 1.00 1.50 2.00 2.50 1.20 3.80 0.20
Type Sample Serum Serum Serum Serum Serum Whole blood Whole blood Whole blood
Gold Found, Y thv.) 0 48
0 98 1 48 1.92 2 40 I 18 3 84 0 18
Standard Deviation 0.026 0,050 0.033 0.142 0.224 0.022 0 141 0,002
%
Recovery, Av. 96 98 98 96 96 98 101 90
Table 11. Results Obtained on Material from Patients Suffering from Exfoliative Dermatitis Following Treatment with Gold Salts Patient la la
15 la 2 3 a
T y p e Saniple Whole blood Urine Liver from post-niorteni Kidney from post-mortem Whole blood Whole blood
Gold I'ound '.34y/nil. 8 . i r / 1 0 0 nil. 60r/100 grams 3 0 ~ / 1 0 0grariis 1.20r/ml. 0.88.r / 1 1 1 1 .
Patient eupired.
XIicrogranis of gold in sample = reading of unknown X niicrogranis of gold standard reading of standard I t is niorc convenient, however, to r a d thf, cwiicenttxtion of the unknown directly from the standard curvr.
DEVELOPXENT OF COLOR
The residue from this process is dissolved in an accurately nicasured volume of 0.1 .V hydrochloric acid. When the sample dctcrmined contains less than 3 micrograms, 3 ml. of acid should be used. For larger sanipleq, more acid must be used. Concentrations of gold exceeding 1 microgram per ml. produce colors which do not obey Beer's la\$. The 3-ml. solution was the smalloat amount which could be used in the cuvettes chosen. To the 3-ml. acid solution of unknown is added 0.3 ml. of the sodium fluoride solution, and 0.5 nil. of the color-forming reagent. The color is permitted t o develop for 15 minutes and is then read in the Coleman Universal sprctrophotometer a t 500 nip with a I' C . 4 filter over the cell. The cuvettes arc horizontal with a (.tipacity of 3 ml. and light paths 5 em. in length. The blank consists of a tube which has been run through the rntire procedure starting with gold-free water as a specimen. A standard curve is run with each day's determinations. PREPARATIOh OF CALIRRATIOY CHART
-1calibration chart is prepaiyd bl- diluting the working standard to levcls of 0 . 2 , 0.5, 1.0, and 2 0 micrograms per 3 ml. The dilutions HM' made in the samc volume of 0.1 S hydrochloric acid as the unknoxns. For each 3 ml. of these standards, 0.3 ml. of sodium fluoride solution and 0.5 ml. of the color-forming reagent ai'e added. The blank consists of 3 nil. of 0.1 S hydrochloric acid, 0.3 ml. of sodiuni fluoride solution, and 0.5 ml. of the colorforming rclagent. PREPARATION O F RECOVERY SAMPLES
llccovc,ries n.ere made I J adding ~ a suitable amount of the working gold standard directly to the tube containing the unknown and procerding as for the unknown alone.. Table I shows the results of T2 recoveries on wholt~ blood and serum using varying amounts of gold.
DISCUSSION
method requires no s p c i a l equipment not ordinaiily n the routine laboratory. .S. photoelectric colorimeter may he used in place of the spectrophotometer. .-llthough taking 3 days for completion, the method is not difficult. Most of the t8imc is spcnt in ashing and drying p i i o d s which permit the operator to do othrr work. The use of this method permits more careful control in chry,sotherapy and is a h of value in dia.gnosing gold toxicity. Table I1 lists some results obtained on samples from pat,ients with symptoms of exfoliative dermatitis who had been treated with gold compounds. Patient 1 espired. The determination on patient 3 was done 6 months after gold therapy had bern discontinued. Table I1 shows the high levels which were obtained on the patient who died ( 2 ) . Figure 1 shows that if the concentrakion of gold is maintained below 1 microgi~ainper ml. a curve may b(x ohtained which follows Beer's law. Table I indicates that' the method is accurate within 5%. LITERATURE CJTED
(1) Block, TI-. D., and Buchanan. 0. H., J . B i d . Chem., 136, 379 (1940). 12) Freyberg. R. H., Block, W. D., and Le\-y, 8.. J . CZin. Incest., 20, 401 (1941). (3) Hcnch, P. .3.,Bauer, It-.,Crain, D. C . . 1~'xyIxrg.R . H.. Graham, \I-.,Holbrook, IV. P.. Lockie, L. M., l\lcEweri. ('., Rosenberg, E. F., and Ptecher, 1%.M., Ann. Internal M e d . , 28, 66 (1948). (4) Jamieson, A. It., and Watson, R.S.,A m l y s t , 63, 702 (1938). (5) Kurell, W. C . , Pills1,ury. P., and Gellert, S. *4., Stanford M e d . Bull., 5, 197 (1947).
Sandell, E. B., ANAI,. CIHEZI., 20, 253 (1948). (7) Sandell, E. B., "Colorimetric Analysis of Traces of lIrtnls," p. 252, New Tork, Interscience Publishers, 1944. ( 8 ) Short, C. L., Reckman, W. W., and Bauer, K . .Vu Eugl. J . (6)
Med.. 235, 362 (1946). (9) Sobel, A. E., and Snow, S. D., J . B i d . Cheni..171, 617 (1947).
CALCULATIONS
In the portion of the curve which follows Beer'slaw, the amount of gold in the sample may be determined by using the formula,
RECEIYED September 14, 1950. Presented before t h e Division of Biological Chemistry a t the 116th Meeting of t h e L 4 x E R I C A N CHEMICL POCIETY, Philadelphia, Pa.
