DITHIZONE as an ANALYTICAL REAGENT* WAYNE E. WHITE Western Reserve University, Cleveland, Ohio
The discovery of the compound, diphenylthiocarbazone (2) in 1882 as a part of his work on the structure and (dithizone), and its applications to the qualitative and reactions of phenylbydrazine. The preparation is quantitative analysis of many of the heavy metals are re- relatively simple and direct as indicated by the equafiewed. As yet, its most important usage is in the estima- tions: tion of minute amounts of lead and the various ways i n which these determinations have been carried out are indicated. The sources, preparation, and stability of the reagent are mentioned.
+ + + + + +
Phenylhydradne
Phenyithiaearbazinatc
H
H
S
H
H
T
0 THE casual observer and occasional user of analytical procedures i t may be of no particular significancethat organic chemistry has become an important aUy of its sister branch, inorganic qnalitative and quantitative analysis. Certain organic compounds such as dimethylglyoxime, alizarin, "aluminon," "cupferron," and the pH indicators are so familiar because of long usage that one may not think of them as representatives of a whole class of substances which may be used for similar purposes. But if this same casual observer is made to realize that analytical methods are being simplified, are being speeded up, and, in some cases, are being made more accurate through use of special organic reagents, he will a t once become interested. Anything which will eliminate tedious hours of weighing, of filtering, and of ignition or drying to constant weight is of great importance to every chemist and laboratory. Many of the organic analytical reagents find most application to cases where the concentration of the sought-for element or radical is very low. In these cases the analysis is generally dependent on a distinctive coloration which serves as a qualitative indication, and oftentimes the intensity of this coloration can be used for a quantitative estimation. Other reagents, such as 8-hydroxyquinoline or anthranilic acid, are simply precipitants which have one advantage or another over inorganic precipitating agents. Dithizone is a reagent of the first class: it is most valuable for low concentrations and it is a color-producing substance. It was first used in this country in 1933 ( I ) , and because of the comparative recency of its introduction and the multiplicity of its possible applications i t is worthy of particular consideration. Diphenylthiocarbazone-the complete name for dithizone-was first made and studied by Emil Fiscber
-
Presented a t the 12th Ohio-Michigan Regional Meeting of the American Chemical SocietyinToledo, Ohio, October 19,1935.
On warming of the carbazide in alcoholic potassium hydroxide solution, a mutual oxidation-reduction occurs, thus: H H S H H
a=&-N-C-N-NO H
I I I I I H
+
S
I II OI N - N +- C +- N = Diphenyithioearbazone
H
H
S
UA-A-L-NH~ Phenylthiosemiearbazide
The diphenylthiocarbazone is soluble in the alkali and is precipitated therefrom by dilute acid. In practice, i t seems to be easy to get dithizone but not so easy to get i t in a pure state. When first precipitated i t probably retains some of the semicarbazide. This may be removed by redissolving in dilute aqueous alkali and again precipitating by acid. The precipitated mass is very likely to hold some of the inorganic salt formed upon acidulation of the alkaline solution and must be washed thoroughly with water for purification from this contaminant. Emil Fischer noted the acid nature of dithizone and the formation and color of its zinc salt. Apparently no one attempted to use i t as an analytical reagent, however, until Hellmut Fischer took up the study of the metallic dithizonates and their possible applications in analytical chemistry in 1925. To him goes most of the credit for the present knowledge and usefulness of this organic compound ( 3 ) . In the most comprehensive of his several papers (3f) the following metals
are said to give distinctive reactions with dithizone: Cu, Ag, An, Zn, Cd, Hg, T1 (ous), Sn (ous), Pb, Bi, M n (ous), Co, Ni, Pd, and Pt. It is pointed out that, with the exception of Mn and the Group VIII metals, these elements all fall in the B families of the various periodic table groups. The nature of the reaction which takes place between the dithizone and a metallic salt and the structure of the resulting compound are matters of conjecture. Fischer (3f) gives possible formulas of the compounds as obtained with a mono-, a di-, or a trivalent metal when in acidic and basic solutions. For the present discussion it is sufficient to point out the following as facts of most practical importance: (1) the pH of the solutiou determines the type and the stability of the metallic complexes-generally a dithizonate capable of formation and existence in an alkaline solution will be decomposed by the addition of acid to lower the pH below 7, and conversely, the dithizonates formed in an acid solution are unstable in the presence of a base; (2) the dithizone molecule acts as a monobasic acid usually and the metallic dithizonates are formed by replacement of one of the hydrogens in the group H H
I
l
l
--N-N(3) there may be some type of ring formation in some cases to increase the valence of the dithizone radical to two; (4) the lead complex, however, which is capable of formation and existence in a slightly alkaline solution is known to have the composition PbDz where D is the monovalent dithizoue radical; and (5) with those metallic ions which are capable of easy reduction-such as Hg (ic), Au (ic), and Au (ow--oxidation of the reagent may precede or accompany the usual salt formation. As stated above, the principle involved in the use of this reagent is that of color formation. Dithizone itself dissolves in chloroform (and to a lesser extent in carbon tetrachloride and other inert organic solvents) to give a green color; the metals mentioned above yield complexes which are also somewhat soluble in chloroform and impart to their solutions colors ranging from shades of violet to distinct red and orange shades. Therefore, the basic procedure to be followed in qualitatively analyzing for some particular ion is to shake some of the dithizone solution with a properly prepared aqueous solution. A color change in the chloroform layer indicates the presence of the metal. A little reflection on the various factors involved will lead to an understanding that the sensitivity of the test can be controlled through regulation of (1) the concentration of the dithizone, (2) the relative volumes of the reagent and tested solutions, (3) the thoroughness with which the two immiscible liquids are shaken together, and (4) the pH of the aqueous solutiou. Consequently, an exact statement cannot be made of the minimal weight of a metal which can be detected by dithizone unless the conditions of making the test are defined. It may be stated, however, that by ordinary test-tube technic and without taking any particular precautions
to increase the delicacy of the reaction, one can expect quantities of a metal down to 1 to 5y* to be readily detectable. Assuming 27 of one of the metals reactive toward dithizone to be present in 10 cc. of solution (a convenient volume to use with 1 cc. of reagent solution) and assuming that a distinctive coloration occurs on shaking this with some of the dithizone solutiou, the sensitivity, in terms of concentration, may be said to be 0.2 part per million. The delicacy of the test for a metallic ion may be increased approximately one hundred times by use of a micro-technic employing diminutive test-tubes or spot plates. It has been reported that amounts of zinc as small as 0.0257 can be detected (4),and that the dithizone detection of minute traces of lead in urine is more satisfactory than the spectroscopic method (5). In quantitative analysis of the heavy metals, dithizone has been found useful in two different ways: (1) for a preliminary extraction of the sought-for metal from undesirable impurities or excessive amount of solvent as preparation for determination by some other means, and (2) for the actual determination of the amount or concentration of the metal. The concentrating action of the reagent is obvious upon consideration of the relative volumes of the aqueous and chloroform solutions. One may use 5 cc. of dithizone solution with 100 cc. of water solution of ziuc, for example, and by thoroughly shaking them together he may extract all of the ziuc from the water (provided, of course, the amount of zinc is less than equivalent to the amount of dithizone in the 5-cc. solutiou), and thereby effect a concentration to 5% of the original volume in a small fraction of the time required to boil away the water. If it should then be necessary to get the metal back into a water solution, a little concentrated hydrochloric acid may be added to liberate the dithizone from the metallic complex and to return the metal into the acidic aqueous phase. In order to evaluate the possibilities of extracting one of the metals from others which may interfere in some particular analytical procedure one should refer to H. Fischer's paper (3f) describing the groups into which metals may be placed in accordance with their reactions toward dithizone. Four groups are obtained as follows: (A) Pd, Au, Hg, Ag, and Cu are the metals reacting in an acid solutiou; (B) in a slightly acid solution, buffered to a pH of about 6, Zn is most sensitive to dithizone (large amounts of Co, Ni, Pb, and Cd are said to interfere, however, by producing some coloration also) ; (C) in a slightly alkaline solution containing CN', Sn (ous), T1 (ous), Pb, and Bi are reactive; and (D) in a strongly alkaline solution containing tartrate ion, Co, Ni, and Cd are affected. Within each of these groups Fischer gives ways of distinguishing between the members of the group so that a whole scheme of qualitative analysis for these metals is set up. A review of the literature reveals an uncertainty in regard to the effect of acidity on the reaction of the * The symbol y is used to represent a microgram; i. c., 0.001 milligram.
bismuth ion with dithizone. Although there is no doubt original reagent. This procedure has the obvious adthat it produces a coloration under the same conditions vantage of eliminating one step, and it is said to give as does lead, namely in a slightly alkaline cyanide solu- results which are reproducible to within about l y (9) tion, apparently it also reacts in an acid solution of when used for analysis of lead in spray residues or in pH 2, so that Bi may be extracted from a Pb-Bi mixture biological materials. The third type of colorimetry is considerably differby first using the reagent in the acid solution (6). This being the case, it would seem that Fischer's group ent from the two just described. In each of the other A should also contain Bi, or that possibly this metal modifications, precautions are taken to extract all of might make up another group coming in between A the lead from the aqueous solution and then to remove all excess dithizone from the chloroform layer to leave and B. In the second mode of applying this reagent to quan- a solution of the pure, light-red lead dithizonate. In titative analysis-"for the actual determination of the this third procedure, however, the plan is to leave the amount or concentrationw-two general procedures excess reagent in the chloroform solution so that in are available. Both methods have had application as making color comparisons for finding amount of lead, yet chiefly in the estimation of lead. The first of the two one deals with a series of color shades ranging from pure methods may be designated as titrimetric, and is su- greens (in the absence of any metal) through various perior to the second, which may be called the colori- bluish and purplish shades (resulting from the mixing metric method, when analyses are expected to cover a of varying amounts of green and red) to, finally, a pure wide range of metal concentrations. However, in the cherry red if there is enough lead to react with all the estimation of lead as spray residues on fresh fruits and dithizone. This two-color colorimetry has the advegetables where the results will nearly always fall vantage of permitting fairly exact lead determinations within the range of 0.000 to 0.025 grain of lead per without the use of an expensive colorimeter; the compound the colorimetric method, wherein the colors parisons may be made satisfactorily by the unaided obtained from the sample are compared directly to eye in daylight against a white background. For sprayresidue lead analysis, it has been found satisfactory in standard colors, is more convenient. In brief, the titrimetric method involves the addi- giving results within 0.001 grain per pound* of the cortion of small portions of a dithizone solution of known rect value (1). A possible simplificatiou of the method for spraytiter to the prepared lead solution (buffered to some fixed pH value between 7.5 and 10.0) contained in a residue lead analysis outlined in the preceding paraseparatory funnel. After each dithizone addition the graph is based on a fact mentioned by Fischer (3) but two layers are shaken together and the lower--chloronot previously considered in this discussion. Dithiform-layer is drawn off. When it appears that the zone's great affinity for many of the metals is evidenced, of course, by the completeness with which the ions of lead is being used up from the aqueous solution-this being denoted by a less rapid and complete color these metals are removed from their aqueous solutions. change-the dithizone additions should be very small; But the metal need not be in solution, or even in the endpoint is attained when the reagent remains combination, for dithizone, in the presence of a little green when shaken with the water layer. The total ammonia, reacts readily with solid, metallic lead, lead volume of reagent solution is, of course, the measure sulfate, lead arsenate, or lead urso1ate.t Many of the of amount of lead. This procedure has been success- other metals and their undissolved compounds react fully used in the determination of minute amounts of as well. This reactivity with the slightly soluble lead lead in biological materials and is said to be sensitive compounds has led to an attempted perfection of a procedure whereby the dithizone solution is not only to l y Pb (7). A colorimetric method has also been used for the used to indicate amount of lead but is also used as the determination of lead in biological materials (8). The solvent to remove the lead from the apple or other fruit. particular procedure followed in this work is considered The lead is washed off by the dithizone which changes more cumbersome than other modifications to be de- in color corresponding to the amount of lead removed scribed in the next paragraphs. Briefly described, this (10). method calls for an extraction of the lead from its Another adaptation of dithizone to a practical slightly alkaline water solution by excess dithizone, a analytical problem has been made in identifying highly removal of the excess dithizone from the chloroform by heated milk. I t is said that if as much as 3% to 5% a washing with 0.1 N ammonium hydroxide (ammonium of heated milk be added to a raw milk the addition can dithizonate is soluble in water but insoluble in chloro- be detected by the coloration produced in the dithizone form), and finally conversion of the red lead dithizone solution. - Presumably the effect is one of oxidation; complex back to the green of free dithizone by addition * As used in spray-residue analyses, a difference of 0.001 grain of hydrochloric acid. The green solution is then com- per pound corresponds to an actual differenceof lead in the volume of solution used of about 77. pared to standards by means of a colorimeter. t Ursolic acid is one of the constituents of the natural, waxy. By the second modification, the lead is extracted in protective coating of the apple. It is possible that some of the the same way and excess dithizone is removed but in lead from the lead-arsenate sprays is held on the fruit as lead Cf. COHEE,R. F. AND ST. JOAN,J. L.. I d . Eng. this case the lead complex is kept for colorimetric ursolate. Chem.,26, 781 (1934). and SELL,H.M. AND KREMERS, R . E., comparison rather than being reverted back to the Ind. Eng. Chem., Anal. Ed..7, 105 (1935).
the milk probably gains in oxidizing power on being heated (11). As a h a 1 citation of published work on this versatile reagent, mention is made of its use in zinc analysis. In this procedure an alkaline solution of the reagent is recommended instead of the u s i d chloroform or carbon tetrachloride solution. With the proper precautions, it is claimed that 0.001 to 1.000 mg. of zinc can be determined with an error no larger than about 1% and the method is said to be suitable for determining zinc in biological materials (12). It seems possible that further investigation of alkaline aqueous solutions of dithizone might reveal this reagent to be useful as a precipitating agent for some of the heavy metals and that dithizone might thereby be useful as a "macro" reagent as well as for amounts in the neighborhood of 1to 1000y. It was in an aqueous alkaline solution that this compound was first used (3a),but it might be worth while to go back to this for further experimentation, because of the higher concentration of dithizone obtainable. In the adaptations which have thus far been made, however, i t is believed that chloroform (W.S.P. grade) is the most satisfactory solvent. The reagent appears to be more stable in this solvent than when in carbon tetrachloride and the greater solubility in the chloroform is also a factor of occasional importance. In preparing and using the
reagent solution, care must always be observed in preventing harmful contamination. Of course, the solution must not be allowed to come in contact with metals or their compounds. Contact with rubber must also be avoided, and long exposure to direct sunlight is probably detrimental. Because of the oxidizability of the dithizone, strong oxidizers must be carefully avoided also. If care is taken in making and storing a solution (in a brown, glassstoppered bottle) there should be no appreciable change in six to eight weeks. If it is desired to keep a solution for longer periods i t may be done by covering i t with a layer of sulfurous acid solution to prevent oxidation by atmospheric oxygen. The proper concentration of the reagent is determined by the use to be made of it, but in general it will be best in the range of 15 to 50 milligrams per liter. A preparation of diphenylthiocarbazone was outlined in the first part of this discussion; it is not recommended, though, that one attempt this preparation for the purpose of getting the analytical reagent. More detailed directions must be published before the average chemist will find it as satisfactory to make his dithizone as to purchase it. It is marketed in this country a t present by the Eastman Kodak Company, the G. Frederick Smith Chemical Company, and possibly by other concerns.
LITERATURE CITED
(1)
WI-NN,
H. J., AND omens, 5. Assoc. Official Agr. Chem.,
j. Ibid., 46, 517 (1933). k . Z . anal. Chem., 103, 241-57 (1935)
17, 117-8, 130-5 (1934). EarrL, Ann., 212, 316-24 (1882). (2) FISCHER, (3) FISCRER. HELLWT
General papers a . Wiss. Verdffcntlich.Siemnzs-Konsern, 4, 15G70 (1925); Chem. Abstr., 20, 3660 (1926). b. A n ~ m .Chem... 42. 1025-7 (1929): Chem. Abstr... 24. 567 (1930).
c.
d.
.
.
1.
Mikrochemie, 2 (n. s.), 319-29 (1930). WITH LEOPOZDI, G., Wiss. Veroffentlich. SiemensKonaern, 12, 44-52 (1933); Chem. Abstr., 27, 3418
(4) (5)
(1933).
(6)
a. Angew. Cham., 46, 442-6 (1933). f. Ibid., 47, 685-92 (1934). g. Wiss. Veroffentlich. Siemens-Wcrken, 14, No. 2, 41-53 (1935); Chem. Abrtr., 29, 6553 (1935). Detection of cobalt in nickel h. w i s s . vcriffcntlich. s ~ ~ ~ 6, 147-9 ~ ~(1928); . Chem. Abstr., 23, 1364 (1929). Determination of capper and lead i. WITH LEOPOLDI, G., Angew. C h . , 47, 90-2 (1934). Detection and determination of mercury.
Determination of silver LEOPOLDI, G. AND VON U S L ~H.. , Z. anal. Chem..
101, 1-23 (1935).
.
,.
WITR
(7) (8)
K (9) (10) (11) (12)
Detection of zinc m . ~ T LEOPOLDI, H G., Z. ibid., 97,385-95 (1934. REINACKER. G. AND SCHIPP, W., ibid.W., 94,409-15 (1933). B O H N E ~ A MHP. , AND LINNEWEH, DeUt. Arch. klin. Med., 175, 157-69 (1933); Chem. Abstr., 27, 5680 (1933). WILLOUGHBY, C. E., WILKINS, E.S.. AND KEAEMER.E. 0.. Ind. Eng. Chem., Anal. Ed., 7, 285-6 (1935). WILKINS, E. S., JR. AND OTHERS, Ind. Eng. Chm., Anal. Ed., 7, 33-6 (1935). Ross, J. R. AND Luchs, C. C., Can. Med. Assoc. J., 29, 649-50 (1933); Chem. Abststr., 28, 1732 (1934). WINTER, 0. B. m~ oT=Rs. I d . E w . Chem., Anal. Ed.,7, ~265-71 ~ (1935). ~ ~ ~ , WHITE,W. E.. ibid.. 8,231 (1936). E ~ L EK., AND PPEIFFER. H., Z. Untersuck. Lcbensm.. 68, 307-10 (1934); Chem. Abstr., 29,237 (1935). DECKERT, W., Z. anal. Chem., 100, 386-90 (1935).
ADDITIONAL REFERENCES
Determination of lead
General W., Z. anal. Chm., 96, 1 2 8 3 1 (1934). F~ESENIUS, Determination of bismuth in copper ANON., Amlyst, 60, 554--6 (1935). ANON.,Chem. Trade J.. 97, 31 (1935); Chem. Abslr., (1935).
ALLPORT,N. L. (1932).
29, 5768
Determination of copper in foods MEHURIN, R. M., 1.ASSOC. Oficial Agr. Chem., 18, 1 9 2 4 (1935).
AND
S ~ ~ S H I RG.E H., , Analyst, 57, 440-9
KENT-JONES. D. W. AW HERD,C. W., ibid., 58, 152-3 (1933). LYNCH, G. R., SLATER, R. H.. AND OSLER, T. G., a i d . , 59, 787-806 (1934).
&us. J., Z. ges. erptl. Med.. 95, 4 3 M O (1935). BEHRENS, B. AND TAEGER, H., ibid., 96, 282-303 GARRATT, D. C., Analyst, 60, 817 (1935).
(1935).
CLIFFORD, P. A. AND WICAMANN, H. J., J. ASSOC.O&id Agr. C h m . . 19, 130-56 (1936). Determination of mercury W~LBLING, H. AND STEIGER, B., Angew. Chem., 46, 279-81 (1933).
WINKLER,W. 0.. J. AISOC. O f l c i d Apr. Chenz.. 18, 6 3 8 4 4 (1935). Determination of traces of heavy metals in water HELLER. K., KUHLA, G., AND MACAEK, F., Mikrochnnie, 18, 193-222 (1935).
