Studies with dithizone—XII2

Studies with dithizone—XII2

Tdanta. 1968. Vol. 15, pp. 811 to 821. Pcrgamon Press. Printed in Northern Ireland STUDIES WITH DITHIZONEXII* FORMATION OF THIADIAZOLINES BY ...

944KB Sizes 26 Downloads 143 Views

Tdanta.

1968.

Vol. 15, pp. 811 to 821.

Pcrgamon

Press.

Printed

in Northern

Ireland

STUDIES WITH DITHIZONEXII* FORMATION

OF THIADIAZOLINES BY CONDENSATION ALDEHYDES AND KETONES

WITH

H. M. N. H. IRVING and U. S. MAHNO~ School of Chemistry,

University

of Leeds, Leeds 2, U.K.

(Received 24 November 1967. Accepted 26 January 1968) Summary-The red colour that develops when mineral acids are added to solutions of dithizone (I; 3-mercapto-1,5-diphenylformazan) in certain samples of dioxan is mainly due to the formation of 2-methyl3-phenyl-5-phenylazo-1,3,4_thiadiazoline (II) derived in part from adventitious traces of 2-methyldioxalane. A purple compound of molecular formula C,,H,,N,S is also formed from (I) by an independent and slower reaction. The thiadiazoline (II) is readily prepared from (I) and acetaldehyde, but analogous compounds from formaldehyde benzaldehyde, acetone and ethyl methyl ketone are obtained in better yield by starting from diphenylthiocarbazide. Di-p-tolyldithizone gives similar reactions. Reduction of the thiadiazoline (II) [which has spectra very closely resembling those of 1 :l complexes of (I) with arylmercury (II) cations] with ammonium hydrogen sulphide in ethanol yields diphenylthiocarbazide by opening of the hetero-ring and elimination of an alkyl residue from the intermediate 3-alkylmercapto1,5diphenylformazan. Other examples of nucleophilic displacements from the formazan group by SH- have been investigated.

DURING an extensive investigation of the acid dissociation constant of dithizone (3-mercapto-1,5-diphenylformazan, H,Dz; I, R1 = H, Ar = Ph) and its analogues, absorptiometric measurements of relative concentrations of the strongly coloured species H,Dz and HDz- were carried out on a series of buffer solutions made up in various dioxan-water mixtures. Previous experience with these reagents had led us to expect that decomposition would become increasingly troublesome as the medium became more alkaline,l and in similar measurements in 50% aqueous dioxan Freiser et uZ.~*~ found it necessary to extrapolate their absorbance measurements to zero time. In our hands, however, little difficulty was experienced when measurements in the more alkaline solutions were conducted under anaerobic conditions with very pure reagents. On the other hand, solutions in the more acidic buffers ([H+] > 0.1M) sometimes unexpectedly changed colour from green to red in the course of a few minutes. This colour change is illustrated in Fig. 1 for a solution of di-p-tolyldithizone (I; R1 = H, Ar = p-tolyl) in 70 % (v/v) dioxan-water containing sufficient hydrochloric acid to give a 1.OM solution. Similar results were found with dithizone itself and with di-p-anisylthiocarbazone (I; R1 = H, Ar = p-methyoxyphenyl). Although the rate of reaction decreased when the acidity was reduced to 0*5M, the nature of the mineral acid used appeared to be unimportant; similar changes occurred with * Part XI-H. Irving and A. M. Kiwan, J. Chem. Sot., 1963,428s. t Present address: Dr. U. S. Mahnot, C-76 Savarkar Marg., Tilaknagar, Jaipur-4, India. 811

H. M. N. H. IRVINO and U. S. MAHNOT

812

perchloric and sulphuric acids. There was no obvious change in behaviour when the dioxan content was progressively reduced to 30 %, but 70 % dioxan was used in all the subsequent work in view of its greater solvent powers. The clearly defined isosbestic points (Fig. 1) at 467 and 549 rnp (or at 4.58 and 537 rnp for dithizone itself and at 458 and 576 rnp for itspmethoxy analogue) immediately suggested an equilibrium between the thiocarbazone (&,, 450 and 620 rnp) and some -I 80

60

25

15

20 Frequency,

cm-’

X IO4

l.--Spectra of 1,776 x 1CPM di-~tolyldit~~ne in 70% v/v dioxan-water containing hydrochloric acid ([H+l = 1*0&f). Curves A to Hshow the changes after 1.1, 5, 10, l&20,25, 30 and 63 min respectively. After a further 27 min the spectrum coincided with that labelled H.

Fm.

protonated species with A,,, 500 m,u; if no side-reactions occur this species would have a high molar absorptivity of ~20000. The slowness of the reaction raised doubts as to the correctness of this hypothesis, which was definitely disproved by the observation that the spectrum of a solution of l-214 x lo_SM di-~-toIy~dithizone in 70% dioxan-water at [H+] = l*OM taken 70 min after mixing the components, and which had not changed after a further 70 min, was quite unaffected by the addition of an exact equivalent of alkali, remaining unchanged for at least 16 hr. A freshly prepared solution of dithizone in glacial acetic acid is green with peaks at 440 and 625 rnp, but after several days thecolour changes to purple, and though there is a broad band at-500 rnp (similar to the intense band at 532 my in the purplesolutions of dithizone in cone. hydrochloric acid), this is subordinate to a much more intense band at 312.5 rnp (Fig. 2). Clearly different species are involved in this reaction and

813

Studies with ditbizone-XII

the new species absorbing at 312.5 rnp arises from the “purple compound” (A) described below. The possibility that the red colour of solutions of dithizone in acidified aqueous dioxan was due to oxidation by peroxides in the dioxan used was eliminated by

\ \

II4

I

1

I

3

Frequency,

\

I’ I

1I

I 2

cm-’ Xt04

l%. Z.-Spectrum of 2*47 x 10dM dithizone in redistilled glacial acetic acid. (a) Immediately after preparation; (6) (broken line) after 4 days; (c) after 7 days.

deliberately adding hydrogen peroxide. The peak at 620 rnp due to dithizone then disappeared during some 40 min while the dithizone peak at 450 rnp also decreased in intensity and underwent a hyps~hro~c shift. No peak appeared at ~500 rnp and the spectrum of the fina solution closely resembled that of the “yellow oxidation product” referred to by many workers. p The solution was shown to contain anhydro5-mercapto-2,3,diphenyltetrazolium hydroxide (VII) together with a smaller amount of a substance that regenerated dithizone on standing with an aqueous solution of sodium thiosulphate. Dioxan appeared to be essential to the formation of the substance absorbing at 500 rnp; solutions of dithizone in ethanolic 1*0&f perchloric acid or in 0.16M sulphuric acid in 70 % glacial acetic acid-water mixture decomposed in a few hours and gave no new substances absorbing in the visible region. Isolation of the substance

814

H. M. N. H. IWING

and U. S. MAHNOT

absorbing at 500 rnp was attempted by carrying out the reaction with l-g quantities of dithizone, but the reaction mixture was then shown chromatographically to contain two products, (A) a purple crystalline compound with a coppery lustre, m.p. 180-2” (decomp.), and (B) a more soluble dark red crystalline compound m.p. 99” (decomp.) which dissolved in most organic solvents but not in water or aqueous alkali. When the time of reaction and the relative proportions of acid, dithizone and dioxan were varied systematically, the ratio of the amounts of purple compound and red product appeared to increase with time. That the red compound was a precursor of the purple compound was disproved by a variety of experiments, and the correct explanation is that there is an initial rapid reaction between dithizone and aldehydic impurities liberated by the action of acid on the dioxan-water mixture, and when all these have been consumed the relatively slow reaction between the excess of dithizone and the acid then leads to increasing amounts of compound A. Compound A, of molecular formula C,H,,N,S, had a characteristic spectrum in ethanol (J,,, 314 and 536-538 rnp, emax23.6 x 103and 9.04 x 103)and was identified by mixed melting point and comparison of its absorption spectrum with that of a specimen of the “beautiful reddish-bronze crystalline compound” prepared by heating dithizone or “dehydrodithizone” with glacial acetic acid according to Ogilvie and Corwin’s procedure,6 and rigorously purified. Not all its reactions agreed with those reported by Ogilive and Corwin and its investigation will form the subject of a later paper. The spectrum of compound B in 70% dioxan showed that this was responsible for the colour changes observed in an acidic solution of dithizone in this solvent. Its molecular formula, C,,H,,N,S, (confirmed by vapour pressure osmometry in chloroform) differs from that of dithizone, C,,H,,N,S, by an additional two carbon and two hydrogen atoms which must clearly have been derived from the solvent. Gas chromatography of the dioxan used, which had been purified by Freiser’s method2es revealed the presence of an impurity which was absent from dioxan purified according to Weissburger and Proskauer’s method,6 but present to the extent of about 1% in an untreated commercial sample of dioxan. This impurity was shown by gas chromatography to be 2-methyldioxalane. This would have been eliminated in Weissburger and Proskauer’s procedure which includes heating under reflux with hydrochloric acid to decompose acetals, followed by a stage to eliminate aldehydes. These steps are omitted in Freiser’s procedure which is based on prolonged refluxing over metallic Acid hydrolysis of the 2-methyldioxalane gives sodium followed by fractionation. rise to acetaldehyde which condenses with dithizone (I; R1 = H, Ar = Ph) to give 2-methyl-3-phenyl-5-phenylazo-1,3,4-thiadiazoline @a; R2 = H, R3 = Me, Ar = Ph). This was very readily synthesized by mixing dithizone with acetaldehyde and adding a few drops of cont. hydrochloric acid. Condensation with acetone to give 2,2-dimethyl-3-phenyl-5-phenylazo-1,3,4thiadiazoline (IIIJ; R2 = Ra = Me, Ar = Ph) proceeded similarly as did that with ethyl methyl ketone although the red product in this case was not obtained analytically pure. Similar condensations were carried out with di-p-tolylthiocarbazone (I; R1 = H, Ar =p-tolyl). Preund ‘~8has reported that diphenylthiocarbazide (V) condenses with phosgene or thiophosgene to give the same products (VIII; X = 0 or S) as those obtained from dithizone itself. Pel’kis et al. m” have recently extended this method to the synthesis of a large number of homologues of VIII (X = S). Clearly

815

Studies with dithimne-XII

(Ha) R’ = (II b) R’ = (11~) R’ = @Id) Rz =

H, R3= Me, Ar =Ph R” i= Me, Ar = Ph R’ = H, Ar = Ph

H, R”= Ar =Ph ilIe) R*= H, R”= Me, Ar =p-to@

Ph-NH.NH.CS.NH.NH-Ph

(VII)

WI)

Reagents: 1, HCl; 2, NH8 and HpS in ethanol at 0”; 3, &Fe(CN), or NaOH-H,O, 4, (a) Me1 for R = Me, (6) chloracetic acid for R = --CH% . CO%H; 5,Zn-NaOH 6, boiling NaOMe-MeOH; 7, alkaline solution of dextrose.&

when R1 = H; when R1 = H;

the loss of two hydrogen atoms must occur at some stage in this reaction and we now find that dithizone can be replaced with advantage by diphenylthiocarbazide (V) in condensations with carbonyl compounds. Thus formaldehyde and benzaldehyde gave rise to (IIc), (R2 = R3 = H, Ar = Ph) and (IId) (RZ = Ar = Ph) respectively. ArN-

vm x=c.,

‘S’

j C-N=NAr

NH,.H,S+

Ary-f x=c

\

,C--NH.NHAr

Ix

s

The NNR spectrum of 2-methyl-3-phenyl-5-phenylazo-l,3,4-thiadiazoline @Ia) recorded in deuter~hlorofo~ at 60 MC showed the symmetrical doublet due to CH:, (centered on T = 8.41, the 1:3:3:1 quartet due to CH (centered on T = 3.9) and a complex group due to two sets of 5 equivalent protons; the integrated curves confirm the groupings of 3,1, and 10 protons respectively. Preund et al.‘** have shown that the ketones or thioketones (VIII; X = 0 or S) are readily reduced to the corresponding hydrazo compounds (IX; X = 0 or S) when treated with an ethanolic solution of ammonium hydrogen sulphide. When, however, the red compound @Ia) was reduced under these conditions, the colourless product, obtained in over 90 % yield proved unexpectedly to be diphenylthiocarbazide (V), identified by mixed m.p., infrared spectrum, and oxidation to dithizone. By contrast with the behaviour of the thiadiazohnes (VIII) where the azo-grouping in the side chain is reduced to a hy~azo-soup, the hetero-ring remaining untouched, reduction of (Ha) must have opened up the heterocyclic ring with the subsequent elimination of the alkyl residue -CHRZRS. Had the ring ruptured between C(2) and S(1) an N-alkyldithizone would have resulted, subsequent reductive de-alkylation

H. M. N. H. IRVINGand U. S. MAHNOT

816

of which would seem unlikely in view of the resistance of N-alkylanilines and phenylhydrazines to reduction by ammoniacal hydrogen sulphide at 0”. If, on the other hand, the heterocyclic ring is opened between N(3) and C(2) the primary intermediate would be an S-alkyldithizone (e.g., III). Corwin and 01givie5 have already shown that the reduction of 3-methylmercapto-1,5-diphenyltetrazolium iodide (VI; R = Me) by ammonium hydrogen sulphide in ethanol at 0” gives rise first to S-methyldithizone (III; R = Me) which they were able to isolate, and then to an unstable white crystalline substance which was readily reoxidized to S-methyldithizone on exposure to the air. This substance, tentatively formulated as (IV) (R = Me), gave diphenylthiocarbazide (V) when allowed to stand in contact with ethanolicammonium hydrogen sulphide for 4 hr. 3-Carboxymethylmercapto-1,5-diphenyltetrazolium iodide (VI; R = -CH,CO,H) underwent a similar reductive sequence5 and we now find that the end-product of the reduction of (VII) is likewise diphenylthiocarbazide (V). These processes which involve the elimination of an S-alkyl group must be due to the nucleophilic displacement of RS- by the excess of HS- added, followed by tautomeric rearrangement to the isomeric thioketone. The conversion of a 3-chloroformazan to the corresponding dithizone is a further example of this procesP*12 and in Bamberger’s synthesis of dithizones from 3-nitroformazanP-l5 it is postulated that the 3-amino group obtained intermediately by treatment with alcoholic ammonium hydrogen sulphide is subsequently displaced by SH-. All the new thiadiazolines have very similar spectra (Fig. 3). The effect of the pmethyl substituent in (He) is to increase the intensity of the absorption and to produce a very small bathochromic shift of the main peak in the visible region (Table I). A TABLE I.--ULTRAVIOLET

AND VISIBLE SPEaRA OF 1,3/t-THIADIAZOLINES PHENYLMERCURIC DITHIZONATE FOR COMPARISON

WITH THAT OF

Compound

@aI

206-7

W).

206-7

8

Pheny Zercur y dithizonate

266.5 (13.8 x 10s) plateau

206-7 268 265 (12.3 266 (13.6 (13.0 x 1P) 10s) 1W) 263 (20.7 x lo*) 204-5 (46.6 x 1O5)

shoulder -300 305 (10.7 x 108)

492 (28.6 x 10s) 460 (21.7 x lo*)

310 shoulder (11.2 x 10’) 322 (13.3 x lOa)

498 493 (29.0 496 (29.3 x lOa) (33.3 10’) lo*) 476 (32.5 x lo*)

point of considerable interest is the close similarity between the spectra of these new heterocyclics and the 1: 1 complexes of dithizone with arylmercury(I1) cations1s-18. Most of these show bands at 200, 260-280, -320 and 450-550 rnp; the last three bands have been identified in almost all metal dithizonates although the positions and intensities of the absorption maxima are influenced by the nature of the metal.ls The X-ray crystallographic study of mercury(I1) dithizonate has disclosed a very strong bond between sulphur and mercury together with a co-ordinate bond to mercury from one of the nitrogen atoms co-ordinated to a phenyl residue.2o By analogy the structure of phenylmercury(I1) dithizonate may be represented as (X) and the formal similarity to the 1.3,4-thiadiazoline ring is immediately obvious.

Studies with dithiz.one-XII

(‘,H,-NH-N 4 II C,H,-Hg, ,_$I-N=NC,H, S

817

C,H,-K----N II C,H,--CIH \ ,C-N=NC,H, s WI

m

0.9

O-6

05 ii x 5

04

P

03

02

0. I

5

4

3

Frequency.

cm-’

2

X IO’

FIG.3.-Ultraviolet and visible spectra of 1,3,4_thiadiazolines in ethanol. I-Substance IIa (the red compound B), 2.97 x 1O-5M; Z-substance IIb, 2.63 X 10-6M; 3-substance IIc, 1.75 x 10-hM, 4-substance IId, 2.06X 10-6M; 5phenylmercury(fI) dithizonate, 2.43 x 10-6M in 75 % v/v ethanol-water.

Duncan and Thomas21 ascribe a band at 3139 cm-l in the infrared spectrum of dithizone in carbon tetrachloride solution to the N-H stretching frequency and all metal dithizonates are said to show this vibrational feature in the region 3100-3300 cm-l. The spectrum of diphenylcarbazide (V) in a potassium bromide disc is now found to show four bands in this region, which could well be due to the NH groups in this molecule. On the other hand, the complete absence of any bands in this region in the spectra of the 1,3,4_thiadiazoles IIa-IIe is a clear indication that oxidative coupling must have taken place when they are formed by the condensation of a carbonyl compound with dithizone or diphenylthiocarbazide.

818

H. M. N. H. IRVINGand U. S. MAHN~T

The spectra of all the thiadiazoles show a number of high intensity bands in the 3067-3030 cm-l region, and these are probably due to aromatic =CH stretching. The very sharp and intense band at 2967 cm-l for IIa and IIb and centered at 2976 cm-l for He can be ascribed to the assymetric stretch of the CH, group (lit. 2962 & 10 cm-l); it is, of course, absent in the spectra of TICand III. The weak bands at 2920 and 2857 cm-l in TIC have been assigned to the asymmetric and symmetric C-H stretch of the CH, group (lit. 2926 and 2853 cm-l). A tertiary -C-H stretching frequency would be expected at 2890 x 10 cm- l. No such band appears in the spectrum of the 2,2’-dimethyl compound IIa which shows no structure between 2924 and 1590 cm-l; overlap with neighbouring bands would prevent its identification in the other cases. When potassium bromide discs were used no bands were observed in the region 2000-1650 cm-l but with concentrated solutions in hexachlorobutadiene all the compounds showed typical overtone patterns of substituted aromatics. The 2-phenyl derivative IId showed 12 well-resolved bands which differentiated it from the other compounds studied. The bands of high intensity at 1590 rt 3 and 1520 f 7 cm-l in the spectra of all the compounds and bands of variable intensity at 1495 5 10 cm-l appear to be C=C skeletal vibrations, but those at ~1450 cm-l may overlap the bands due to the asymmetric -C(CH,) group deformation in certain cases. The sharp band at 1466 cm-l in the spectrum of IIc has been assigned to the CH, group deformation. The sharp bands centered at 1380 and 1364 cm-l in the spectrum of III, are due to the -C(CH,), group deformation vibrations (lit. 1385-1380 and 1370-1365 cm-l). The bands in the 1250-1100 cm-r region are doubtless due to N-C,H, vibrations: Meriwether et aLza have assigned bands in this region which are found in the spectra of all metal dithizonates. The characteristic pattern in the region 835-792 cm-l for IIe is typical of 1,4 aromatic disubstitution (lit. 860-800 cm-l). Other assignments and typical spectra are given in detail elsewhere.23 In view of the characteristic spectra and high intensity of absorption (E,,~ = 20000-33000) of the new 1,3,4_thiadiazolines it was hoped that they could be exploited for the quantitative absorptiometric determination of aldehydes and ketones, but the reaction with dithizone or diphenylthiocarbazide was too slow unless the appropriate catalyst was used in homogeneous solution, and even then preliminary concentration of the carbonyl compound was necessary and yields were variable. The reactions discussed above do, however, emphasize one of the problems constantly to be borne in mind when carrying out solution work with low concentrations of reactants: at the 10-SM reactant level as little as 0.1% of foreign material in the solvent (e.g., dissolved oxygen, moisture, or in this case 2-methyldioxalane) may lead to disturbing and unexpected side-reactions. The purple compound (A) may have the bis-azo structure Ph.N=N.CS.N=N.Ph; further work on its reactions and a structure determination by X-ray crystallography will be reported elsewhere. EXPERIMENTAL Reagents

Commercialdioxan was purified*s8by refluxing over metallic sodium for 12 hr followed by fractional distillation immediately before use. This procedure does not remove Zmethyldioxalane; another procedures was used for this when necessary. Demineralized water was used to prepare all dioxan-water mixtures, and isopiestic ammonia and hydrochloric acid. a4 Glacial acetic acid and all organic solvents were redistilled immediately before

Studies with dithizons--XII

819

use. Analar samples of dithizone were purified by methods previously described** and the ratio of the absorbances at 620 and 450 rnp of carbon tetrachloride solutions of the reagent was used as an index of purity. A value of 1.68-l-70 was regarded as satisfactory. Absorption spectra were measured with a Unicam SP 700 recording spectrophotometer or with an SP 500, with matched silica cells. Before use these cells and all glassware needed were rigorously freed from every trace of metallic contamination and from oxidizing matter. 2-Methyl-3-phenyl-5-phenylazo-l,3,4-thiadiazoline (Ha) Concentrated hydrochloric acid (4-5 drops) was added to a paste of dithixone (1 g) and redistilled acetaldehyde (4.5 ml, excess) and the mixture was cooled in an ice-bath and triturated with a glass spatula for 10-15 min. Chloroform (100-150 ml) was then added and the solution was shaken with water in a separatory-funnel. The aqueous phase was rejected and the organic phase washed twice with water and then with dilute ammonia to remove unreacted dithizone: it was then dried over anhydrous sodium sulphate and the solvent was distilled off under reduced pressure. The product was recrystallized from methanol to give dark red needles, m.p. 99-100” decomp. Yield 44%. A poorer yield (20 %) was obtained when the condensation of diphenylthiocarbaxide (V; 2.6 g) with acetaldehyde (1 ml) in presence of cont. hydrochloric acid (O-2 ml) was carried out slowly (2 hr) in ethanol (50ml). (Found: C, 63.8%; H, 5.2%; N, 19.8%; S, 11.5% m.w. (vapour pressure osmometry in chloroform) 289: CIIH1aN,S requires C, 63.8 %, H, 5.0%; N, 19.9 %; S, 11.3 %; m.w., 283.) 2,2’-Dimethyl-3-phenyI-5-phenylazo-l,3,4-thiadiazoline (IIb) Dithizone (1.0 g) in acetone (50 ml) containing cont. hydrochloric acid (1.0 ml) was shaken mechanically for 1.5 hr. After removal of the solvent in a stream of air the sticky residue was dissolved in chloroform and transferred to a separatory-funnel, washed with water (3 x 50 ml) and the solution dried over anhydrous sodium sulphate. Chromatography on alumina with benzene as the eluent gave ultimately dark reddish-orange crystals which after two recrystallixations from methanol had m.p. 77-78’ decomp. Yield 47%. (Found: C, 64.7%; H, 5.4%; N, 19.1%: C,,H,,N,S requires C, 64.9 %; H, 5.4%; N, 18.9 %.) 3-Phenyl-5-phenylazo-1,3&thiadiazoline

(IIc)

Concentrated hydrochloric acid (@4 ml) and 36% w/v aqueous formalin (2 ml) were added to a suspension of diphenylthiocarbazide (V, 2.6 g) in ethanol (50 ml) and stirred mechanically for 3 hr. Dark red crystals separated (0.25 g) and a further amount (1.87 g) was obtained by removing the solvent in a stream of air. A chloroform solution of the total crude product was transferred to a column of alumina and chromatographed with benzene; dark red crystals (0.6 g) were obtained on evaporation. After two recrystallizations from benzene these had m.p. 164-5” decomp. Yield 23 %. (Found: C, 62.5%; H, 4.7%; N, 21.0%: C1dHIaNNISrequires C, 62.7 %; H, 4.5 %; N, 20+9x.) 2,3-DiphenyI-5-phenylazo-1,3&thiadiazoline

(IId)

Diphenylthiocarbazide (V, 2.6 g), benzaldehyde (2 ml, excess), concentrated hydrochloric acid (0.5 ml) in ethanol (70 ml) were shaken mechanically for 2.5 hr. A dark red solid (0.25 g) which separated, together with the red residue left after evaporation of the solvent in a stream of air, was taken up in chloroform, filtered from an unidentified white substance (0.58 g, m.p. 278-280” decomp.), and treated with a solution of sodium bisulphite to remove excess of benxaldehyde. Chromatography on alumina with a 1% solution of methanol in benzene as the eluent gave a prominent red band from which the thiadiazoline was isolated as red crystals, m.p. 144-5” decomp. Yield 23 %. (Found: C, 69.7 %; H, 4.7%; N, 16.3 %: ClOH1,N,S requires C, 69.5%; H, 47%; N, 16-3x.) 2-Methyl-3-p-tolyl-5-p-tolyIazo-l,3,4-thiadiazoIine (He) This compound was prepared from acetaldehyde and di-ptolylthiocarbazone as described for the phenyl analogue. After two recrystallimtions from methanol the red crystalline product had m.p. 156-7” decomp. Yield 20%. (Found: C, 65.6%; H, 5.7%; N, 18.2%: C1,HlsN,S requires C, 65.8 %; H, 5.9%; N, 18.1%) Isolation of the compounds A and B In a typical experiment, dithizone (1 g) suspended in a mixture of redistilled commercial dioxan (70 ml), concentrated hydrochloric acid (10 ml) and distilled water (20 ml) was shaken mechanically for 16 hr and then poured into water (100 ml). The reddish-brown solution obtained on extraction with chloroform (50 ml) was washed repeatedly with water to free it from dioxan. After being tied over anhydrous sodium sulphate the chloroform was allowed to evaporate at room temperature and the red residue (O-6 g) was taken up in chloroform (15 ml) and chromatographed on a column of neutral alumina. The less strongly adsorbed reddish-orange band was eluted with 1% ethanol in

820

H. M. N. H. IRVING and U. S. MAHNOT

benzene and on removal of the solvent gave substance B, (al5 g). The slower moving dark purple band was eluted with chloroform and cave O-35 a of substance A. Substance A gave deep purple crysta% with a coppery lustre, m.p. ISO-182”, from aqueous ethanol. Found: C,61*5%; H, 4.0%; N, 22.0%. m.w. by vapour pressure osometry in chloroform, 246: C,oHlONIS requires C, 61.4%; H, 3.9%; N, 22.0%; m.w. 2543). Substance A was identified with the “isomeric dehydrodithizone” described by Ogilvie and Corwin,6 by comparison with authentic samples prepared by heating (a) dithizone or (b) the syndnone (VII) under reflux with glacial acetic acid. After purification bv chromatomaphv and recrystallization from aqueous ethanol all the samples had ident& melting pocnts and abs&$ion spectra: The purple compound A was also obtained by boiling diphenylthiocarbazide (V, 2 g) in glacial acetic acid (15 ml) under reflux. The initially colourless solution changed to dark green (5 min) and then to purplish red (30 min). The reaction mixture was cooled, poured over ice and extracted with chloroform; sticky purple crystals were obtained (O-95 g) on removing the solvent in a current of air. The crude material was triturated with a small volume of chloroform and filtered from a greyish white solid (a29 g) which after two recrystallizations from boiling ethanol formed light st&&coloured crystals, m.p. 200-201” decomp. This substance was identied as p-phenylthiosemicarbazide by mixed m.p. and the identity of its infrared spectrum with that of an authentic specimen.lsb Investigation of the impurity in dioxan by gas-liquid chromatography A 5-foot column packed with 10% dinonyl phthalate on Celite was used with a hydrogen fiame detector. Chromatograms were taken of (i) commercial dioxan, (ii) dioxan purified by Frciser’s method,aq* (iii) dioxan purified by Weissburger and Proskauer’s method,8 (iv) 2-methyl-1,3-dioxalane prepared according to Hibbert and Timma and carefully fractionated, (v) purified commercial acetaldehyde diethylacetal (l,l’-diethoxyethane), and (vi) redistilled acetaldehyde. With a constant flow-rate of carrier eas the main neaks of i, ii and iii had retention times of 71.3, 74.7 and 78-l set respectively but saGpIes i and ii had subsidiary peaks at 51.0 and 54.3 set respectively, which were identified as due to 2-methyldio~I~e (retention time 53 sets in iv). Samples i, ii and iii were shown to be free from l,l’-~e~oxyet~e (retention time 67-6 set) and a~~dehyde (2@9 set). By comparison with chromatograms containing known amounts of pure dioxan (iii) and Zmethyldioxalane (iu) it was estimated that the commercial dioxan or a specimen purified by Freiser’s method contained about 1 y0 of 2-methyldioxalane. Reduction of 2-methyl-3-phenyl-5-phenylazo-l,3,4-thiadiazoline (IIa) with ammonium hydrogen @hide in ethanol A sus~n~on of the trampoline (Ha; 0.56 g) in ethanol (25 ml) was cooled to 0” and saturated with ammonia gas (10 min). Hydrogen sulphide was then bubbled in (10 min) whereupon ammonium hvdroeen sulnhide separated as shinine leaflets, No colour change occurred at tit but after 1.5 hr at’ roo”m temgrature g yellow solutionyesulted which was poured>lowly over cracked ice; diphenylthiocarbazide (V) then separated as shining colourless platelets (0.47 g) which were collected, washed first with water and then with alcohol, and dried in a vacuum desiccator. After recrystallization from ethanol, the substance when heated changed colour to light green at 145-150” arid decomposed at 156-158”. Identification as cv) was confirmed by analysis, mixed m.p., infrared spectroscopy, and by oxidation with methanolic potash to dithizone. Diphenyt~~b~de (V) was also found to be the product of the prolonged action of ammonium hydrogen sulpbide on ethanolic solutions of S-methyldithizone (III; R = Me) and the sydnone (VII), confirming previous statements by Ogilvie and Corwin.6 Acknowledgement-U. S. M. thanks the Association of Commonwealth a Commonwealth Scholarship.

Universities for the award of

Za~~~Die rote Farbe, die sich beim Zusatz von Mineralsguren zu Lasungen von Dithizon (I; 3-Mercapto-1,5-diphenylformazan) in gcwissen Dioxanproben bildet, riihrt im wesentlichen von der Bildung von 2-Methyl-3-phenyl-5-phenylazo-1,3,4-thiadiazolin (II), das sich teilweise von zuftilig anwesenden Spuren von 2-Methyldioxolan ableitet. Eine purpurrote Verbindung der Summenformel C$,H,,N,S bildet sich such aus (I) durch eine unabhkgige langsamere Reaktion. Aus (I) und Acetaldehyd 1st sich das Thiadiazolin (II) leicht herstellen; analoge Verbindun~ von Form~dehyd, Bcnzaldehyd, Aceton und Athylmethylketon erhHlt man in besserer Ausbeute, wenn man von Diphenylthiocarbazid ausgeht. IX-p-tolyldithizon gibt iihnliche Reaktionen. Die Reduktion des Thiadiazolins (II)

Studies with dithizons_XII

821

(das Spektren zeigt, die denen der 1: I-Komplexe von (I) mit Arylqueck&er (II)-l&ionen sehr lhnlich sindj mit Arnoniumhydroeensulfid in Athanol zibt Dinhenvlthiocarbazid durch c)ffnen des Heterorings und Elimi~erung*ein& Alkylrestes aus dem Zwischenprodukt 3-Alkylmercapto-1,5diphenylformazan. Andere Beispiele nucleophiler Verdriingungen durch SH- von der Formazangruppe wurden such untersucht. R&nn&La coloration rouge qui se developpe quand des acides mineraux sont ajoutts a des solutions de dithizone (I, 3-mercapto 1,5-diphenylformazan) dans certains t?chantillons de dioxane est principalement due a la formation de 2-methyl-3-ph6nyl-5-phenylazo1,3,4-thiadiazoline (II) provenant en partie de traces fortuites de 2-mtthyldioxalane. Un compose pourpre de formule mol&daire C,,H,,N,S se forme aussi a partir de (I) par une reaction indtpendante et plus lente. On prepare ais6ment la thiadiazoline (II) a partir de (I) et d’acttaldehyde, mais les composes analogues des formaldehyde, benzaldehyde, acetone et mtthylbthylcetone sont obtenus avec un meilleur rkndement en partant hu diphenylthiocarbazide. La di-ptolvldithizone donne des reactions semblables. La reduction de la thkdiazoline (II) [qui a un spectre ressemblant exactement a ceux des complexes 1 :l -de (I) avec ]es cations arylmercure (II)] par le sulfure acide d’ammonium en ethanol donne le dinhenylthiocarbazide par ouverture de l’het6rocycle et elimination d’k r&idu alkyle du 3alkyl-mercapto 1, 5-diphenylformazan intermediaire. On a etudie d’autres exemples de deplacements nucleophiles du groupe formazan par SH-. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.

REFERENCES H. Irving, S. J. H. Cooke, S. C. Woodger and R. J. P. Williams, J. C/rem. Sot., 1949, 1847. K. S. Math, Q. Fernando and H. Freiser, Anal. Chem., 1964,36,1762. H. Freiser, R. G. Charles and W. D. Johnston, J. Am. Chem. Sot., 1952,74,1383. S. S. Sahota, Ph.D. Thesis, Leeds, 1964, and refs. therein. J. W. Ogilvie and A. H. Corwin, J. Am. Chem. Sot., 1961,83,5023. A. Weissburger and E. Proskauer, Organic Solvents, Physical Constants, and Methoa!s of Purification, p. 139. Pergamon Press, Oxford, 1935. M. Preund and M. Kuh, Ber. 1890,23,2821. M. Preund, ibid., 1891,24,4178. M. Z. Peretvazhko and P. S. Pelkis, Zh. Obshch. Khim., 1961, 31, 3726. R. G. Dubcnko and P. S. Pelkis, ibid., 1963, 33, 2682. H. M. Irving and C. F. Bell. J. Chem. Sot.. 1953. 3538. R. Fusco an’h R. Romani, darz. Chim. Itai, 1946,76,419. E. Bamberger, R. Padova and E. Omerod, Annalen, 1926,446,260. E. Bamberger, Ber., 1894,27, 155. E. Bamberger, 0. Schmidt and H. Levenstein, ibid., 1900,33,2043. (b) EL Fischer and A. Besthorn, Annalen, 1882,212, 316. H. Irving and J. J. Cox, J. Chem. Sot., 1961, 1470. A. M. Kiwan, Ph.D. Thesis, Leeds, 1965. H. Irving and J. J. Cox, J. Chem. Sot., 1963,466. Idem, Proc. Chem. Sot.. 1959, 360. M. hk. Harding, J. Chekr. Sol., 1958 4136. J. F. Duncan and F. G. Thomas, ibid., 1960, 2814. L. S. Meriwether, E. C. Breitner and C. L. Sloan, J. Am. Chem. Sot., 1965, 87, 1441. U. S. Mahnot, Ph.D. Thesis, Leeds, 1966. H. Irving and J. J. Cox, Analyst, 1958,83,526. H. Hibbert and J. A. Timm, J. Am. Chem. Sot., 1924,&I, 1286.