phenols do not interfere. Thc estimation can also be carried out in the presence of amine bases if diphenylcarbohydrazide is used as the indicator. Thiophene seriously interferes. ACKNOWLEDGMENT
The authors thaidi tlir Council of Scientific and Industrial Research, India, for financial assistance.
LITERATURE CITED (1) Das, 11. N., -4.1-AL.CHEM.26, 1086
(1954). (2) Das, M.S . ,J . Indian. Ckent. SOC.31, 39 (1954). (3) Killing, 0. W., J . Ani. Chem. SOC.79, 2711 (1957). (4) Kundu, K. K., Das, M. Tu’., Sci. and Culture (Calcutta) 23, 660 (1958). (5)?Palit, 8. R., IND. ENG.CHEM.,. ~ N A L . hD. 18,246 (1946). (6) Palit, S. R., Das, hf. N., Sornayajulu, G. It., “Non-r\qrieous Titration,” Indian
ASRJC. Cultivation of Sci., Calcutta, 1954. (7) Pifer, C. N7., \Vollish, E. G., - 1 ~ 4 ~ . CHEM.24,300 (1952). (8) Zbzd., p. 519. (9) Pifer, C. W-., Wollish, E. G., J . -1m Phnrm. Assoc., Sei. Ed. 42,509 (1953). (10) Sporek, K . F., Analyst 81, 474 (19.56). (11) Tomitek, O., Zukriegelova, >I.,Chent. listy 46, 263 (1952). It13 EIYED for review December 12, 1058. hccepted April 16, 1959.
Estimation of Arsenic in Biological Tissue by Activation Analysis HAMILTON SMITH Department o f Forensic Medicine, Universify of Glasgow, Glasgow, Scotlund
b Activation analysis is quick and accurate for the estimution of arsenic in very small samples of biological materials. After nitric-sulfuric acid digestion of the activated sample, a Gutzeit separation is combined with an estimation using a Geiger tube which accepts liquid samples. Blanks are not required. A single hair weighing 0.5 mg. can b e analyzed and two people can analyze up to 100 samples of hair in 2 days. The arsenic content of hair from over 1000 living subjects has been estimated by this technique.
T
I!HhE problems iii tlic quantitative clrtrrmination of arsenic in biologiea1 tissue are: the destruction of tissue whik keeping the arsenic in the rcaction nirdium, the analysis, and the final detrrmination of arsenic. Of the methods tricrl, a Geiger counter detection of the r:itlioisotopes of arsenic was the most sensitive. -1complete analysis consists of preparation of the specimen and irradiation in the atomic pile, digestion of the s:iinple in nitric-sulfuric acid mixture, Gutzeit separation of the active isotope : i i i d added carrier arsenic, and detection of the activity by a Geiger counter.
REAGENTS AND APPARATUS
J\-lliw possible AnalaR (British Druy IIoiises) reagents were used. IXgestion mixture of concrntratctl iiitric and sulfuric acids (5 to 3). 16- to 22-mesh zinc pellets. Sodium iodide, 15% solution. Stannous chloride, 4001, solution i i i 500/, hydrochloric acid. Mercuric chloride, 1.6% solution. 0.001N iodine solution in 4001, sodium iodide solution. A solution containing 10 y of arsenic per ml. The apparatus was a modification of
that of Thomas and Collier ( 1 2 ) . All joints n-ere ground glass sealed with water. DIGESTION OF SAMPLES
I k a u s e organic material reduces the accuracy of the final separation of arsenic by 10 to 40% ( 3 ) , the sample must undrrgo a digestive process to drstroy or eliminate all organic matcirial. This takes place after activation of the samplr in the pile and can he perfornirtl bj- wet ashing ( 9 ) ) dry :ishing (5-6,12), or ashing in an oxygen bomb ( 2 ) . Recauqc the second and third methods iiivolrc either a length\ procrdure or the uso oi high tcmpt~inturrsantl do not result in a significantly higher arsenic recol-ery ratio. the \$et digestion method n n s IISPd. WET DIGESTION METHOD
?‘lie quick, siniplt., and accurate method of Milton (9) was tried first on the organic samples. A sample is boiled with dilute caustic soda solution t o fix any free arsenic, antl then wetashed using a mixture oi concentrated sulfuric and nitric acids. The results on a standard sample were reproduciblv nithin 4o/c. Because the results were ithin the same region of accuracy 1%hen thc digestions were repeated without thv alkali treatment, this latter digestion was further investigated. The standard sample was a piece of filter paper to which radioactive arsenic had been added, which g a l e approximately the same digestion conditions as in actual practice. The following series of experiments was made: Standard samples were placed in borosilicate glass test tubes measuring (iX 6/* inches and of about 25-ml. (hapacity. Eight milliliters of a mixture
of concentrattd sulfuric and nitric acids in the proportions of 3 to 5 were added t.0 each. The tubes were placed on a digestion stand and heated to boiling, using fine Carborunduni as bubblcr. Tlie boiling was continued until tlic paper was digested and the nitric acid was rvaporated. The digestion was completed in under 1 hour. The vessels were cooled, the cont.ents diluted to a st,aridard volume, and the activity as estimated by counting in an M - 6 Geiger t’uhe accept,ing liquid s:tlllpl(~s. Four additional samples w r e made up and digested in the same manner, with t h r addition of 10 y of inactiye arsenic as a carrier. The results wried from 90 to 08% recovery, iiidrpriident of the addition of carrier arseiiic. Some splasliirig out of the tubr probably ~ o c t ~ u i i tfor ~ ~ dthe loss. Subsrqucntly flasks of n, greater capacity were used. ontl set of digestions was carried out in 50-ml. Kjeldahl digestion fla.sks. The flasks were too large and the digrstion took 3 hours because the x i d refluxed in the necks of the flasks. The results show a large loss probably due to the lengthy digestion and bhe high tefnperature required to r i m w e t,he nitric acid from the flasks. A third set of digestions was carried out in conical-bottomed flasks; with 6inch necks, and of 25-ml. capacit.y. This digestion took less than 1 hour and all samples showed 100% recowry within the statistical counting error (1%). This method was used in all furt.her investigations. Summary of Digestion Method. T h e sample after irradiation with 3 ml. of concentrated sulfuric acid and 5 ml. of concentrated nitric acid is heated in a conical-bottomed flask (as above) until all t h e nitric acid is removed. T h e reaction mixture is cooled and then diluted. T h e reaction should be complete within 3/4 hour, after which inactive arsenic is added as a carrier. VOL. 31, NO. 8, AUGUST 1959
1361
Table 1.
Recovery of Arsenic
Araenic, P.P.M. Calcd. Found 0.415 0,420 1.70 1.68 1.91 1.90 4.76 4.68 9.05 8.98 20.5 20.2 110.0 108.3
Error,
%
1.1 1.2
0.5
1.6
0.8
1.5
1.2
SEPARATION OF ARSENIC
A
modification of Thomas and Collier’s Gutzeit technique (11) was used. The digested sample was washed into a 200-ml. flask. Ten micrograms of inactive arsenic were added to act as a carrier, followed by 2 ml. of concentrated sulfuric acid, 4 ml. of concentrated hydrochloric acid, 5 ml. of a 15% solution of sodium iodide, and 0.4 ml. of a 40% solution of stannous chloride in 50% hydrochloric acid. The solution in the flask was then diluted to about 150 ml. and placed on a boiling water bath for 5 minutes, by which time it had attained the temperature of the bath. Ten grams of 16- to 22-mesh zinc pellets were added and the evolution of hydrogen was allolved to continue for 15 minutes. During this time the arsenic was evolved, passed through a lead acetate on cotton wool filter, and collected in 1 ml. of a 1.6% solution of mercuric chloride. Five milliliters of 0.001N iodine in 40% sodium iodide solution were added to help the solution of any solids formed. The delivery tube was then washed well into the trap. The contents were made up to a standard volume and counted in an &I-6 Geiger tube. Various steps in the analysis were investigated. Evolution Time. A standard solution was prepared by placing 10 y of inactive arsenic in the evolution flask and adding enough active arsenic to give about 10,000 counts per minute. Five milliliters of concentrated sulfuric acid, 4 ml. of concentrated hydrochloric acid, 5 ml. of 15% sodium iodide solution, and 0.4 ml. of the stannous chloride solution Ivere also added and the solution was made up to 150 ml. Each sample was heated in the water bath for 5 minutes before the evolution of hydrogen was started. During this time the temperature of the flask contents attained that of the water bath. In this experiment the water was boiling. Ten grams of 16- to 22-mesh zinc pellets were added and the evolution was continued for varying times. The following results were obtained : Time, Min. 5 10
15
Recovery, Time, Recovery, % Min. 70 95 20 9 9 . 8 , 100.1 99.3
100.1
25
100.2
From this i t was deduced that a t the boiling point the time required for the total recovery of the arsenic was above
1362
ANALYTICAL CHEMISTRY
14 minutes. The following experiments were made to confirm this. Reaction Mixture Temperature. These experiments were conducted using a reaction mixture as described above, and t h e same procedure, but t h e time of evolution was kept constant for 15 minutes and t h e temperature of the water bath, and hence of the reaction mixture, varied, The results obtained were a s follows: ReTemp., covery, % “C. 17 33 32 59 50 90
Temp.,
Recovery,
70 90 99.5
98.5 99.8 100
c.
%
From the results of the two experiments i t was concluded that 100% recovery would be obtained when the solution was given 5 minutes to heat up followed by 15 minutes’ evolution, the process taking place on a boiling water bath. This method was used throughout subsequent experiments. Form of Zinc. The form of the zinc used in t h e Gutzeit test has received much attention. It has been used as sheet, strip, powder, granules, shot, and pellets. Cassil ( 3 ) recommends 20-.mesh zinc pellets which give a steady and rapid flow, useful for paper strip and disk absorption techniques ( 8 ) . I n this method steadiness of flow was not essential. b u t meed of evolution was. Several forms of zinc were used in standard experiments of the above type. T h e results were as follows: Type of Zinc Granulated Pellets 4-mesh 4-8 mesh 8-16 mesh 16-22 mesh 20-30 mesh
Recovery, % 81
i3 87 90-9i 100 100
Ten grams of 16- to 22-mesh pellets were used in the subsequent standardizing and routine experiments. Acid. T h e standard solution was prepared as described b u t t h e volume of t h e acids was changed. The evolution took place at t h e boiling point for 15 minutes. T h e results showed t h a t the acid concentration was not critical -1 ml. of either acid made no difference in t h e recovery. T h e acid was maintained at a total of 5 ml. of concentrated sulfuric acid and 4 ml. of concentrated hydrochloric acid. Use of Sodium Iodide. Sodium iodide was used in t h e evolution flask t o dilute t h e active sodium formed when t h e biological samples were activated and also because sodium iodide is naturally stable, whereas t h e potassium, of potassium iodide, includes one long-lived naturally occurring unstable isotope. Ten milliliters of 40% potassium iodide gave a count rate of about 4 times the
background when placed in an M-6 Geiger tube. Lead Acetate Filter. Small amounts of hydrogen sulfide are liberated from t h e reaction mixture. This was removed on passing the hydrogen, hydrogen sulfide impurities, and arsine through a filter of cotton wool impregnated with lead acetate. No detectable amount of arsenic could be found fixed in t h e filter, so t h e method was safe t o use. Collecting Arsine. T h e arsenic in t h e form of arsine was removed from t h e hydrogen b y passing i t through a t r a p containing 1 ml. of a 1.6% mercuric chloride solution. I n experiments t o detect losses in this trap, only a few showed positive results and of these, all were below 0.1%. ACTIVATION ANALYSIS
When the only stable isotope of arsenic is irradiated with thermal neutrons, a n unstable isotope is produced by neutron capture. This unstable isotope decays with a half life of 26.8 hours, with the emission of @-particles and y-rays. The activation cross section is 4.0 barns and the saturation activity for a pile flux of lo1* neutrons per square centimeter per second is about 2 X 1 O I 2 disintegrations per minute per gram of elemental arsenic. The reaction is represented by : 75 33
As
+01n
4
As
76 33
The sensitivity of determination is lo-’ gram compared with the highest chemical sensitivity of 10-7 to 10-8 gram. The activation method eliminates blanks other than the standardizing blank which gives the activity expected per gram of arsenic. Submicro separations are avoided because inactive carrier arsenic may be added to increase the total weight of arsenic present and other carriers to hold back any other active atoms produced. The accuracy of determination is limited only by the statistical error of counting. Further information may be obtained from Smales’ review (10). Method of Activation Analysis. About 3 t o 6 mg. of t h e samples are weighed into a polythene bag, which is then sealed. A sample of AnalaR arsenious oxide, weighing between 1 and 2 mg., is prepared in the same way and used as a standard. ( T h e weights are t o t h e nearest 0.01 mg.) The sealed bags are placed in a standard aluminum can and sent to a suitable atomic pile (BEPO) for irradiation with thermal neutrons for 1 day. The unit is returned immediately after being taken from the pile. The samples are removed to be digested and processed as described above. The standard sample is dissolved in sodium hydroxide solution and made up to 1 liter. One milliliter of this is taken and made up to 100 ml. and 1 ml. of
this solution is t&en as the standard. From this the count rate expressed ab counts per minute per milligram of elemental arsenic can be calculated. The amount of arsenic in the samples can be found by comparing the activity of the processed samples with the count rate. Detection of Activity. Of t h e two detection methods, Geiger counter (AI-6 tube accepting liquid samples) and scintillation counter, t h e former was more efficient a n d was used in all experiments. T h e number of counts recorded b y a sample in both systems was t h e same, b u t when t h e scintillation counter was used, t h e background was ten times as large as with the Geiger counter. Test of Complete Method. Filter papers were imp1 egnated with knon n weights of arsenic in t h e form of sodium arsenite. T h e arsenic content was calculated and t h e test pieces sent for irradiation. The analysis n-as carried through completely, including all t h e usual steps. Because t h e digrstion and t h e separation v e r e satisfactoiy, it was felt t h a t good results from this test R-ould confirm the accuracy of the method (Table I). The complete method is a relatively rapid simple technique for determining arsenic in biological material with an accuracy above 98%. =is a further check, the activation method was compared with two chemical methods in an investigation of the arsenic content of detergents (7'). The procedures included a Gutzeit separation
Table II.
Thomas and Collier's Method Sample, -is, p,p.m. grams 31.0 1.0 25.0 1.0 13.5 1. 0 18 5 1 0 1 5 1 0 0 85 1 0 1.6 1.0 0.5 6.0
Comparison of Methods
B.D.H. Method Sample, As, p.p.m. gram 30 0.10 25 0.10 14 0.25 17 5 0 25 1 % 1
and titrimetric estimations using the method of Thomas and Collier (11) and a Gutzeit separation and British Drug Houses stain comparison method ( I ) . Results are shown in Table 11. Table I1 shows the close agreement between the results. It also demonstrates the use of very small weights of sample in the activation method. This is important when samples of tissue from a living subject are to be analyzed, and in medico-legal cases where the amount of available material is often limited. Thc nidhod has also been used in investigations of medical and general interest (6). ACKNOWLEDGMENT
The author thanks John Glaister, .J. hl. A. Lmihan. P. R. Peacock, and Edgar Rentoul for advice gii-en during this investigation. LITERATURE CITED
(1) British Drug IHouses Ltd., British Pharmacopoeia (appendix VI), 1953.
0 25
0 25
Activation Analysis Sample, As, p.p.m. gram 35.9 0.00467 24.8 0.00839 13.8 0.00454 18 2 0 00889 1 87 0 00768 0 67 0 00540 1.64 0.00762 0 57 0.00381
(21 Carey, E. P., Blodgett, G.! Satterlee,
s., I N D . E N G . CHEM., A%SAL. ED. 6, 327 (1934). ( 3 ) Cassil, C. C., J . Assoc. 0.fic. .igr. Chemists 20, 171 (1937). t-l) Evans, R. J., Bandemar, S.L . , ~ ~ N A L . CHEM.26,595 (1954). ;5j Fleuret, A. M., Bull. SOC. chitts. 12, 133 (1945). (6) Lenihan, J . 11. A,, Smith, H., International Conference on Peaceful Uses of *4tomic Energy, Paper 69, Geneva, 1968. \ i ) Lenihan, J . M. A,, Smith, H.. Chalmers, J. G., Suture 181, 1463 (1958) (8) Mills, P. H., J . Assoc. 0j%. d g ? . Chetnzsfs 18, 506 (1935). ( 9 ) llilton, R., Ph.D. thesis, University of London, London, England, 1952 (10) Smales, A. h.,Ann. Rept. Progress Chem. (Chem. SOC.,London) 46, 285 (1949). (11) Thomas, 31. D., Collier, T. R., J . Incl. Hyg. Tosicol. 27, 201 (19451. (12) Toung, E . G., Rice, F. A. H.. J . Lab. Clin. Ired. 29, 439 (1944). H.
RECEIVED for review December 29, 1958. Accepted April 13, 1959. Grant supplied by the Medical Research Council.
Dete rmi nation of GIucono I acto ne, G a Ia cto noIactone, and Their Free Acids by the Hydroxamate Method OLIVER G. LIEN, Jr.' Naval Biological laboratory, School of Public Health, Universify of California, Berkeley, Calif.
b The sensitivity of gluconolactone and galactonolactone to hydrolysis under alkaline conditions indicated that the hydroxamates of these compounds could b e formed a t a near-neutral p H reaction rather than under the alkaline conditions which have been used. The data indicate that there is loss of sensitivity for quantitative determination of these compounds under alkaline conditions and that this results from competition between hydrolysis and hydroxamate formation. The effect of variation of temperature and acidity upon the formation of the
lactones from the corresponding free acids is described.
H
reacts with esters and various related compounds to form hydroxaniic acids which react with iron(II1) in acidic solution to give colored complexes. Lipmann and Tuttle (8) used these reactions for the quantitative determination of acetyl phosphate. Hestrin (6) investigated the method in more detail and developed a procedure for the determination of acetylcholine which was also suitable for the determination of the esters, TDROXTLAMINE
lactones, and anhydrides of short-chain fatty acids. The two procedures differed primarily in that Hestrin used an alkaline hydroxylamine reagent (0.75AY with respect to sodium hydroxide) while Lipmann and Tuttle used a hydroxylamine reagent adjusted t o pH 6.4. Goddu, LeBlanc, and Wright ( 4 ) developed an hydroxamate method in which the reactions take place in a n Present address, Department of Chemistry, San Jose State College, San Jose, Calif. VOL. 31, NO. 8, AUGUST 1959
1363