the electrophilic character of bunsen's cacodyl disulfide, me2as(s)-s-asme2

DOI: 10.1002/zaac.200700279

The Electrophilic Character of Bunsen’s Cacodyl Disulfide, Me2As(S)-S-AsMe2

Panayiotis V. Ioannou*, Dimitris G. Vachliotis, and Theodore D. Sideris

Patras/Greece, Department of Chemistry, University of Patras

Received March 22nd, 2007; revised May 29th, 2007.

Abstract. The behaviour of Bunsen’s cacodyl disulfide, Me2As(S)-S-AsMe2 towards Lewis bases of group 15 of the Periodic Tablewas studied mainly by 1H NMR. While Ph3N did not react, 4-dimethylaminopyridine and triethylamine isomerized the disulfideto Me2AsSSAsMe2. Ph3P, (PhO)3P, (MeO)3P, (EtO)3P and Ph3Asdesulfurized the disulfide to Me2AsSAsMe2. (PhS)3P also desulfur-ized the disulfide but Me2As-S-SPh was also produced. Me2As-S-SPh was the main product with (PhS)3As nucleophile. In most sys-

Introduction

In 1843 Robert Bunsen prepared a compound named caco-dyl disulfide and formulated it as Me4As2S2, by reducingcacodylic acid, Me2AsO2H, with hydrogen sulfide [1]. Healso prepared the same compound by reacting cacodylsulfide, Me2AsSAsMe2 (1), with octasulfur (see Table 1 forformulae). Because of the reducing properties of hydrogensulfide and the oxidizing properties of octasulfur, theBunsen disulfide was thought to have the structure ofbis(dimethylarsenic) disulfide, Me2As-S-S-AsMe2 (2).However, X-ray analysis [2, 3] revealed that it was anAsV/AsIII compound dimethylarsino dimethyldithio-arsinate, Me2As(S)SAsMe2 3. A probable mechanism offormation is discussed later on. Analogous compoundshave then been prepared, e.g. Pr2As(S)SAsPr2 [4],Me(Ph)As(S)SAs(Ph)Me [5], Ph2As(S)SAsPh2 [6],Me2As(S)SAsPh2 [7], and R2P(S)SAsR2 (R � Me, Ph) [8],by these two methods or by reacting R2AsS2Na withR�2AsCl or R�2PCl.

Bunsen studied the thermal behaviour of 3 and its reac-tion with metal salts (Cu(NO3)2, AuCl3, Pb(AcO)2, SbCl3and Bi(NO3)2) where the metal salts (CuI, AuI, PbII, SbIII

and BiIII) of dimethyldithioarsinic acid were obtained.Since then, only a few chemical reactions using 3 have beendescribed in the literature. Thus, with excess CF3I/100 °C/5 days Me2(CF3)As and Me2(CF3)As�S were isolated [9].Heating 3 with 1/8S8 and Na2S the sodium salt of dimethyl-dithioarsinic acid was prepared [10]. More recently,oxidation of 3 with t-butyl hydroperoxide gaveMe2As(S)SAs(O)Me2 and Me2As(S)OAs(S)Me2 [11].

* Prof. Dr. P. V. IoannouDepartment of Chemistry,University of PatrasPatras/Greecee-mail: [email protected]

Z. Anorg. Allg. Chem. 2007, 633, 2077�2084 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 2077

tems, cationic and anionic species were formed. The results are bestinterpreted by initial attack of the nucleophile on the electrophilicAs�S sulfur.

Keywords: Dimethylarsino dimethyldithioarsinate (cacodyl disulf-ide); Bis(dimethylarsenic) disulfide; 4-Dimethylaminopyridine;Triethylamine; Triphenylphosphine; Triphenylarsine; Phosphites;Triphenyl trithiophosphite; Triphenyl trithioarsenite; Sulfenium ion

In 1971, Zingaro et al. [12] found that dissolution of pure3 in non-protic, e.g. CHCl3, and protic, e.g. MeOH, sol-vents gave 1H NMR spectra (three singlets) indicative of anequilibrium between 3 and 2. Such an equilibrium was laterfound for other similar compounds [7, 8]. However, dissol-ution of 3 in basic solvents (pyridine or aniline) gave onlytwo singlets with the same peak intensities but havingchemical shifts different from those in non-protic or proticsolvents and the authors commented that such behaviorwas not understood [12].

We were studying the room temperature oxidationof AsIII thioesters (ArS)3As [13], (RS)3As [14], andL-As(SPh)2 (L � R or Ar) [15] by octasulfur in the presenceof triethylamine as an activator of S8 [16]. Extending ourstudies, we used L2As-SPh (L� Me or Ph) as substrateshoping to get 3 and Ph2As(S)SAsPh2 under milder con-ditions than by reacting L2As-SPh with molten octasulfur[4]. However, the reaction was complicated and we sus-pected that either triethylamine or activated octasulfur orboth reacted with the products, one of which might havebeen 3. We, therefore, studied the electrophilic behaviour of3 equilibrated with 2 in the presence of the nucleophilesPh3N, Ph3P, Ph3As, N,N-dimethylaminopyridine (DMAP),Et3N, (PhO)3P, (MeO)3P, (EtO)3P, (PhS)3P, and (PhS)3Asand in this paper we present our results. In other communi-cations from this laboratory we shall present our results onthe nucleophilic and nucleophilic/electrophilic behaviourof 3.

As naturally occuring sulphur-containing arsenic com-pounds have been recently discovered [17, 18] and investi-gators are trying to identify sulphur-containing dimethyl ar-senicals as arsenic metabolites [4, 19, 20], one of whichmight be 3 [3], our results may be of some interest.

Results

The indentification of the speciesThe identification of the compounds was done by 1H NMRand TLC, Table 1.

P. V. Ioannou, D. G. Vachliotis, T. D. Sideris

Table 1 1H NMR values of methyl protons in Me2AsIII and Me2AsV compounds and Rf values used to identify the products of thereaction of Bunsen’s cacodyl disulfide 3 with equimolar amounts of various nucleophiles. During the reaction some singlets shift upfieldas shown. Our δ values have a range of less than ± 0.002 ppm of the quoted values.

1H NMR: δ values (ppm) of methyl protons TLC (silica gel)Compound Solvent Me2AsV Me2AsIII Compound Solventα Rf

Me2AsSAsMe2 (1) CDCl3 � 1.33a, 1.399 1 B 0.13CD3OD/D2O 3:1 � 1.377

Me2AsSSAsMe2 (2) CDCl3 � 1.40b, 1.36c, 1.399 2 B 0.13CD3OD/D2O 3:1 � 1.37d, 1.377

Me2As(S)SAsMe2 (3) CDCl3 2.16b, 2.11c, 2.164�2.058e 1.55b,c, 1.544�1.529e 3 B see textCD3OD/D2O 3:1 2.14d, 2.167�2.134f 1.50d, 1.527

Me2As-SSPh (7) CDCl3 � 1.353 7 A 0.35Me2AsSPh (8) CDCl3 � 1.336Me2As(S)OH (9) CD3OD/D2O 3:1 2.05g, 2.056 �Me2As(O)OH CD3OD 1.921 � S8 A 0.87

D2O 2.11h, 1.50i � B 1.00Me2As(S)OCD3 CD3OD/D2O 3:1 2.167�2.134j � Ph3N A 0.36

B 1.00Me2As-S-AsPh3

� (6a) CDCl3 � 2.060�2.030k DMAP B 0.00Me2As-S-NEt3

� (6b) CDCl3 � 2.127�2.058l Ph3P B 0.95Me2As-S-DMAP� (6c) CDCl3 � 2.143�2.110m Ph3P�S B 0.58Me2As-S-pyridine� (6d) pyridine � 1.83c Ph3As B 0.97

Ph3As�S B 0.58Me2As-S� (5) CDCl3 � 1.980-1.891n PhSSPh A 0.37

pyridine � 1.10c (PhS)3P A 0.13aniline � 1.10c (PhS)3As A 0.11

Me2As(S)S� (4) CDCl3 2.246 � (PhO)3P B 0.90D2O 1.99b, 1.95c � (PhO)3P�S B 0.18

Me2As(O)O� CD3OD 0 �

a) neat, external TMS [12]; b) ref. [3]; c) ref. [12]; d) in MeOH [12]; e) the singlet progressively shifts in the presence of Et3N, DMAP, and Ph3As; f) the singletprogressively shifts in the presence of H2S; g) 2.05 in D2O (referenced to MeOH) [20]; h) ref. [20]; i) at pH 7.0 [21]; j) coincides with the singlet of 3 in thepresence of H2S; k) in the presence of Ph3As; l) in the presence of Et3N; m) in the presence of DMAP; n) in the presence of Ph3As stays at 1.980; in the presenceof Et3N moves 1.907 � 1.891; in the presence of DMAP moves downfield 1.970 � 1.988 ppm; o) 1.633 ppm in the presence of 0.5 mol Et3N; 1.567 ppm inthe presence of 3.0 mol Et3N per mol of cacodylic acid.α : A � petroleum ether; B � ether/petroleum ether 1:1

TLC could not separate compounds of similar structureand 1H NMR became the main analytical tool. Moreover,3 seems to partly decompose on TLC giving S8 and twoother species.

The singlets due to Me2AsSSAsMe2 (2) and toMe2AsSAsMe2 (1) coincide, while the singlets ofMe2As-S-SPh (7), and Me2As-SPh (8), differ slightly intheir resonance positions. These three singlets do not shiftin the presence of other compounds (charged or not) in thesolution, resonating as shown in Table 1, having a range of±0.002 ppm.The singlets of Me2As(S)-S-AsMe2 (3) when inequilibrium with 2 are at 1.544 ± 0.001 and 2.164 ± 0.002.However, because of binding of other species (e.g. Et3N,DMAP, (PhS)3As) to 3 the singlets progressively shift up-field as shown in Table 1 and in Figure 1. The largest shiftwas observed in the presence of Et3N (0.106 ppm) and thesmallest in the presence of Ph3As (0.004 ppm). It is note-worthy that the shifting singlet of Me2As(S)S- of 3 usuallycontained one positively charged AsIII compound (e.g.Me2As-S-NEt3

� (6b)) as the integrations coupled with elec-troneutrality showed. The chemical shifts of the chargedspecies showed variation with time probably indicatingequilibria (e.g. equation (2)).

The presence of charged species, other than Z (seeScheme 2), in an NMR solution of, e.g., 3�2 � Et3N,comes from the observation that when 3 has totally reacted,Et3N is positively charged. This is taken as an indication of

www.zaac.wiley-vch.de 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Z. Anorg. Allg. Chem. 2007, 2077�20842078

the presence of Et3N bound to a cation. Then, for the solu-tion to be electroneutral, an anion, e.g. Me2As-S� (5), aris-ing from the fission of 3 should be present. The assignmentof the resonance position of the Me2As group in a cationand in an anion was based on the general trend that a posi-tively and a negatively charged species of a molecule res-onate downfield and upfield, respectively. Therefore, somespecies were identified based on these considerations.

Reaction of 3 with non-fissionable nucleophiles

An equilibrated mixture of 2 and 3 (abbreviated 2�3) inCDCl3 showed the expected singlets [12], Table 1 and Fig-ure 1A. This mixture did not react with triphenylamine dur-ing 3 days probably because the lone pair of the nitrogenatom is on a p atomic orbital, as it was deduced from itssolid state structure [22], and, being involved in pπ interac-tion with the phenyl groups, is not available for donation toa Lewis acid. In fact, salts of this amine with HF, HCl andHClO4 cannot be prepared [23]. On the contrary, the solidstate structures of Ph3P [22, 24] and of Ph3As [22, 25] aretetrahedral with the lone pair available for reaction.

Adding Ph3P to 2�3 a very fast (5 min) desulfurizationtook place giving 1, the 1H NMR signal of which coincidedwith that of 2 and a very disagreeable odor developed dueto cacodyl sulphide 1. The singlet at 7.3 ppm of Ph3Pchanged to the pattern of Ph3P�S, whose production was

The Electrophilic Character of Bunsen’s Cacodyl Disulfide

also verified by TLC. Therefore, the main reaction was asin equation (1). Three small singlets at 1.800, 1.874 and1.986 ppm with relative integration ratio of 0.10, 1.00 and0.54, not seen in the case of Ph3As, could not be assigned.A fast internal desulfurization of R2PSAs(S)R2 toR2P(S)SAsR2 has been previously observed [8], probably in-volving a 4-center intermediate. This strengthens the pro-posed by Zingaro et al. [12] mechanism for the spontaneousisomerization of 3 to 2 which involves attack of AsIII onAs�S sulfur.

Ph3As reacted with 2�3 much slower than Ph3P, de-veloping the same offensive smell, and giving Ph3As�S and1, equation (1). The reaction was not complete in 6 days atroom temperature. The Ph3As�S was detected by TLC.Also three small peaks developed slowly at 1.980, 2.052,and 2.245 ppm having a relative ratio of 1:2.6:1.6. Thesepeaks should be attributed to charged species Me2As-S�

(5), Me2As-S-As�Ph3 (6a), and Me2As(S)S� (4), respec-tively.

The free and protonated forms of triethylamine in CDCl3show the CH3 protons at 1.03 and 1.45 ppm, while the CH2

protons are at 2.51 and 3.18 ppm, respectively [26a]. Addingtriethylamine to an equilibrated mixture of 2�3, producedthe changes shown in Figure 1B. The signals of the triethyl-amine moved downfield and we propose that Et3N is boundas in 6b and, probably, as in the zwitterion Z of equation(1). The Me2As- protons of the zwitterion Z and theMeCH2N protons of triethylamine are broad indicating re-stricted rotations or interactions between Me2As- andMeCH2N. The 1.40 ppm peak increased, i.e. Et3N facili-tated the isomerization of 3 to 2 being unable to desulfurize3 to 1 because nitrogen cannot expand its octet. For thesolution to be electroneutral when 6b is formed, we postu-late the presence of an anion like 5. In 30 min, Figure 1C,more Et3N was bound as the cation Me2As-S-N�Et3 6b(38 %), the Me2As- protons of 3 were becoming sharper,whereas the protons of Et3N were broad. These dataindicate that there is an exchange, equation (2), producinga stabilized sulfenium ion [27, 28]. By 53 h, Figure 1E, thecation 6b is 75 %. Upon cooling, the exchange rate slowsdown and the splitting of the Et3N protons is restored, Fig-

Z. Anorg. Allg. Chem. 2007, 2077�2084 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.zaac.wiley-vch.de 2079

Figure 1 The equilibrated system 2�3, A, in the presence of anequimolar quantity of triethylamine after 3 min, B, 30 min, C,90 min, D, and 2 days, E, at room temperature, and after 11 days(run at �10 °C), F.

ure 1F. Transient appearance of the anion 4 was noticedand by the 11th day a strong smell due to 1 was evident.Attempts at the preparation of a solid 6b·5 by reacting stoi-chiometric amounts of 3 and Et3N in dry ether gave anoily product.

DMAP as a free base in CDCl3 has the methyl, H-3,5and H-2,6 protons at 2.93, 6.46 and 8.21 ppm respectively,while in its protonated form the shifts in DMSO-d6 are3.20, 6.98 and 8.21 [26b]. Five minutes after the addition of


an equimolar quantity of DMAP to 2�3, its Me2N- singletresonated at 3.03 ppm and it was somewhat broad. There-after, all the singlets in the spectrum were sharp. The Me2N-one resonated at �3.07 ppm and the bound species shouldbe Me2As-S-DMAP� 6c, stabilized by resonance in the aro-matic ring [29, 30] and, therefore, a sulfenium ion is notproduced in this case. As in the case of Et3N, DMAP facili-tated the isomerization of 3 to 2. After 3 days at room tem-perature, the solution contained 21 3, 32 2, 23 6b, 23 5, and�2 mol% 4.

Reaction of 3 with fissionable nucleophiles

Although (PhO)3P gives the Arbuzov reaction only at hightemperatures, it is included in this sub-section. This phos-phite is not reactive towards S8 [31], but it desulfurizes 3 in< 2 days at room temperature giving 1 and (PhO)3P�S. Theformation of the thiophosphate was verified by 31P NMR[32] and TLC. This result indicates that the As�S sulfur in3 is quite activated as an electrophile. Trimethyl and triethylphosphites desulfurized 3 in less than 5 min giving 1 and(RO)3P�S (R � Me, Et) verified by 1H [33�35] and 31P[32�34] NMR. Isomerization of 3 to 2 was not observed.The formation of the Arbuzov products, Me2As-S-P(O)(OR)2, could not be verified from the 1H NMR spec-tra, but a very small peak at �35 ppm in the 31P NMRspectra may be due to them.

Triphenyl trithiophosphite desulfurizes 3 giving(PhS)3P�S and 1, as in the case of Ph3P and (PhO)3P, butits action was slower (�10 h). A singlet at 1.350 ppm,slowly increasing after 2 h, is due to Me2As-S-SPh (7) (seebelow).

Triphenyl trithioarsenite (or tris(phenylthio)arsine),As(SPh)3, in the solid state has a C3 symmetry [36] while insolution has a C3ν symmetry [37, 38] with the lone pair onarsenic being somewhat “caged” by the three phenylgroups. It was not nucleophilic towards amsylaltes,[RO-SO2-Ph-NMe3]�[CF3SO3]� [39], but it showed nucleo-philic activity towards 3. In various NMR runs we observedthe formation of a yellow suspension after 0.5-5 h, thesupernatant of which contained octasulfur (by TLC) andPhSSPh (by 1H NMR). The canary-yellow solid was notAs2S3 (by color and IR), contained varying amounts ofAs2O3 (by IR) and dissolved with decomposition in D2O/NaOH, to a colorless solution containing thiophenolateand 4·Na� at 1.96 ppm, Table 1. The color and the insolu-bility point towards a polymeric material. An apparent mo-lecular weight of �1.500 was calculated for one experimentas described before [13]. Colored, insoluble solids were ob-tained by the action of S8/Et3N on (ArS)3As [13] and(RS)3As [14], and we could not characterize them.


From the changes observed in the colorless solution, i.e.before the yellow solid appears, it is clear that As(SPh)3

does not desulfurize 3 because the sulfide 1 is not produced.In the AsIII region (1.3-1.6 ppm) two compounds areformed: a dominant one with the methyl protons at1.353 ppm, due to Me2As-S-SPh (7), and a minor oneat 1.336 ppm due to Me2As-SPh (8). In the AsV regionfor neutral molecules a singlet at 2.034 ppm, due to anunknown compound, appeared, while 3 had not totallyreacted. The major product could not be isolated in a purestate by column chromatography, but it was found to beMe2As-S-SPh (7) by molecular weight estimation from1H NMR and from the chemical shift of the reaction of 3with thiophenol where the expected 7 resonates at1.354 ppm (Sideris et al., unpublished results). Attempts atfinding the molecular weight of the main product byMALDI mass spectrometry were not successful. The massspectra of known arsenic compounds, e.g. 3 [12], (ArS)3As[40], and other organoarsenic compounds [41], reveal thatthe molecular ions are of low relative abundances and thespectra are quite complicated showing unexpected fragmen-tations.

Stability of 3 in the presence of MeOH/H2O and H2S

Bunsen [1] reported that when an alcoholic solution ofcacodylic acid is diluted with water a high amount ofMe2As-S-AsMe2 (1) was produced together with the disulf-ide 3, and he wrote the equation (3).

In order to see whether the sulfide 1 was produced fromthe disulfide 3, we dissolved 3 in CD3OD/D2O �3:1 (morewater precipitates out the disulfide), and ran its spectrumafter 2 h. While Zingaro et al. [12] reported that anequilibrated solution of disulfide in methanol contains onlythe disulfides 2 and 3, we find that in CD3OD/D2O 3:1during 2 h some 3 was isomerized to 2. Also, anotherpeak appeared at 2.056 ppm which can be attributed toMe2As(S)OH (9) Table 1. Moreover, the peak at 2.167 ppmattributable to Me2As(S)S- of 3 contained another com-pound based on the integration of the peaks. This new AsV

compound is probably Me2As(S)-OCD3. It cannot beMe2As(S)SH because it plus Me2As(S)OH (9) have moresulfur atoms than the starting compound 3. Also, the freeacid Me2As(S)SH is not stable and gives 3 [10, 11] (seeScheme 1).

Bubbling H2S for 5 min to the mixture, slowly (12 days)resulted in disappearance of Me2AsSSA3Me2 (2). SinceMe2As-S-AsMe2 (1), was not detectable by 1H NMR at1.35-1.37 ppm, it can be said that 1 does not arise fromthe formed disulfides 2 and 3 but from a precursor of thedisulfide 3.


Discussion

The reaction of cacodylic acid with H2S

The disulfide 3 is most conveniently prepared from dimeth-ylarsinic acid (cacodylic acid) and hydrogen sulfide ratherthan from the sulfide 1 and octasulfur [1, 12]. The reactionof dry cacodylic acid and hydrogen sulfide to give 3 is veryexothermic, so a solvent like ethanol was used to moderatethe exothermicity, while in ethanol/water the reaction gave3 and the sulfide 1 [1]. The formation of 3 evidently involvesthe formation of the protonated dimethyldithioarsinate,4·H�, because when sodium cacodylate is used in the placeof cacodylic acid, then 4·Na� is obtained [42]. We proposethat the formation of 4·H� should involve the concertedreaction of cacodylic acid with H2S, Scheme 1, which will

Scheme 1 The formation of Bunsen’s cacodyl disulfide (3), inethanol containing a little water, and cacodyl sulphide (1) (alongwith 3) in ethanol containing more water. In the presence of chlori-des of heavy metals, complexes of 4 are obtained [10]. When so-dium cacodylate is used, then 4 ·Na� is obtained [42].

give the unstable [20] Me2As(O)SH which rearranges to themore stable [20] dimethylarsinothioic acid (9), and withmore H2S the free dimethyldithioarsinic acid (4·H�) is


formed. The acid 4·H� does not dimerize by expulsion ofH2S, but rather, by a concerted reaction gives 10 (but see[3]). The pentacoordinated AsV, having suitable leavinggroups, is unstable and by expulsion of disulfane gives theproduct 3. Expulsion of disulfane [15] should be much eas-ier than expulsion of HSS�. The disulfane is metastable anddecomposes thermally [43] to H2S � 1/8S8 via a non-radicalchain mechanism [44].

Because 3 in the presence of H2O and H2S does not give1, the formation of 1 (together with 3) from cacodylic acidand hydrogen sulfide in ethanol/water [1], most likely arisesfrom 4·H� via 10� as suggested in Scheme 1.

The preparation of 3 from 1 and octasulfur on warming[1] should involve nucleophilic attack of AsIII on activatedS8 [15].

Chemical properties of Bunsen’s cacodyl disulfide 3

Apart from its novel isomerization, 3 to 2 [12], which isan intramolecular redox reaction, 3 shows nucleophilic (e.g.reaction with metal cations [1]), electrophilic (e.g. reactionwith S8/S2� [10], and with pyridine and aniline [12]), prob-ably both electrophilic and nucleophilic (e.g. reaction withF3C-I [9]), and reductive (e.g. reaction with t-butyl hydro-peroxide [11]) aspects. Most reactions are accompanied byfragmentation of 3.

Our results clearly demonstrate the electrophilic charac-ter of 3 towards a variety of nucleophiles. From the non-fissionable nucleophiles, Ph3P and Ph3As desulfurized 3 bythe straightforward mechanism shown in equation (1).However, even for these simplest cases, other reactions cantake place to a small degree which, in the case of Ph3P, wecould not conjecture. In the case of Ph3As, minor pathwayslead to isomerization of 3 to 2 and the production of 1 and4 as shown in Scheme 2 for the case of Et3N. The rates ofdesulfurization of 3 by Ph3P and Ph3As, compared to therates of their reaction with octasulfur [45, 46], indicate thatthe As�S sulfur is quite activated as an electrophile.

The results on the action of Et3N and DMAP on an equi-librated mixture of 2�3 which gives mainly 2 and splits 3into two other species, Figure 1, can be explained as shownin Scheme 2. Attack of the nitrogen nucleophiles shouldtake place on As�S sulfur by analogy of the phosphorusnucleophiles (equation (1)) to give the zwitterions Z(R3N � Et3N, DMAP). Since nitrogen cannot expand itsoctet to give R3N�S, the zwitterions Z decompose into theanion 5 and the cations 6b and 6c. Their, somewhat slow,recombination gives 2. After all 3 has been consumed andpart of Et3N is still bound, it follows that one of the down-field singlets, Figure 1, should be due to the cation 6b. Forelectroneutrality, 5 is the most likely anion. The anion 5should be nucleophilic towards 3 giving the anion 11. Thiscan collapse by the probable routes shown in Scheme 2 togive 1 and 4. The production of 1 by these pathways ex-plains the puzzling desulfurization of 3 by Et3N. However,we cannot offer a simple mechanism for the consumptionof 4 (Figure 1 D, E). While DMAP and Et3N do not differ


Scheme 2

in the nature of the products of their reaction with 3, theydo differ in the stabilization of the cations. In 6c there is aresonance stabilization of the cation in the aromatic ring[29, 30], but in 6b such stabilization is not possible and theresults can be explained by the formation of a stabilized[27, 28] sulfenium ion, equation (2). It is the stabilization ofthe cations 6b and 6c the reason for their slow recombina-tion with 5 to produce 2. Zingaro et al. [12] by using pyri-dine as reactant and solvent did not observe any isomeriz-ation of 3 to 2 but only formation of the anion 5 and thecation 6d.

Triphenyl, trimethyl and triethyl phosphites reacted with3 giving (RO)3P�S and 1, equation (1), without isomeriz-ation to 2. Reaction in the Arbuzov way could not be estab-lished with certainty.

Triphenyl trithiophosphite, (PhS)3P, desulfurized 3, equa-tion (1). However, a small amount of Me2As-S-SPh (7), wasproduced by a mechanism analogous to that of (PhS)3As.With (PhS)3As as the nucleophile, 1 and the anion 5 are notproduced. Instead Me2As-S-SPh 7, and very small amountsof Me2As-SPh 8 are formed in solution before the yellowsolid appears. These two products should be produced fromthe zwitterions Z (L � (PhS)3As) (equation (1)) by intra-molecular attack of Me2As-S- sulfur and arsenic, respec-tively, on the PhS- group of (PhS)3As�. What causes thespontaneous formation of the yellow solid in the system(PhS)3As and 3 is not known.

While 3 equilibrated with 2 in methanol is stable [12], thepresence of water hydrolyses 3 to a small extent producingMe2As(S)OH, 9, and Me2As-SH. The latter, as a free acid,is autoxidized probably via the peroxyacid Me2As(S)OOH[47] to 9, which, in the presence of methanol, should be inequilibrium [48, 49] with its methyl ester.


Conclusions

From the experimental results presented herein, it is clearthat Bunsen’s cacodyl disulfide 3 has a dual personality:electrophilic and nucleophilic. For its electrophilic behaviorthe proposed routes to the observed products can also ex-plain the literature results [10, 12] where the nucleophile(�S-S7-S�, pyridine, aniline) attacks on the As�S sulfur.

Experimental Section

General

Triphenylamine (Merck), triphenylphosphine, triphenyl, trimethyland triethyl phosphites, triphenylarsine, and 4-dimethylaminopyri-dine (Aldrich) were used as received. Bunsen’s disulfide 3 was pre-pared in 65-70 % yields from cacodylic acid (Serva) and hydrogensulfide according to Zingaro et al. [12], m. p. 68�69 °C (lit. [12]70�71 °C). TLC (Et2O/petroleum ether 1:1: Rf 0.40 (brown spot),Rf 0.95 (yellowish spot) for S8; streaking from Rf 0.0 to 0.40 withspots visible at Rf 0.18 and 0.06. Triphenylphosphine sulfide wasprepared from triphenylphosphine and equivalent amount of octas-ulfur in chloroform and triturated with boiling methanol. Yield93 %, m. p. 162�164 °C. Triphenylarsine sulfide was prepared fromtriphenylarsine and equivalent amount of octasulfur in the presenceof 30 mol% triethylamine in refluxing absolute ethanol till a clearsolution was obtained (28 h). Cooling at �4 °C and centrifugationgave the product in 89 % yield, m.p. 169 °C (lit. [46] 160 °C). Tri-phenyl trithiophosphite was prepared from phosphorus trichloride,thiophenol and pyridine in dry ether, contaminated by 8 % tri-phenyl phosphorotrithioate (Tsivgoulis and Ioannou, in prep-aration) and having m. p. 76�77 °C (lit. [50] 77�78 °C). Triphenyltrithioarsenite was prepared according to Serves et al. [51].

Silica gel 60 H (Merck) and silica gel H (Merck) were used for thinlayer (TLC) and column chromatography, respectively. Spots on


TLCs were made visible by iodine vapours. Spraying with sulfuricacid and charring did not produce spots. IR spectra were obtainedon a Perkin-Elmer model 16PC FT-IR spectrometer. 1H NMRspectra (400 MHz) were recorded on a Bruker DPX Avance Spec-trometer in CDCl3 with internal TMS (0.000 ppm) as standard. 31PNMR spectra (162 MHz) were referenced to (PhO)3P�O in ace-tone-d6 having a shift of 17.3 ppm upfield relative to 85 % H3PO4

[52]. The 31P chemical shifts having been adjusted to 85 % aqueousH3PO4, differed by �5 ppm from those reported in the literature[32�34]. The MALDI mass spectra were obtained using an Ap-plied Biosystems MALDI TOF-TOF 4700 Proteomics spec-trometer using α-cyano-4-hydroxycinnamic acid as matrix.

Reaction of 3�2 with various nucleophiles

In a NMR tube, 3 (17.2 mg, 62.5 µmol) was dissolved in CDCl3(0.8 ml) and allowed to equilibrate, by forming 2, for about 2 h[12]. An equimolar quantity of the nucleophile was then added andthe progress of the reaction was followed by 1H NMR and TLC,Table 1. The smell of 1 was detectable even from the stopperedNMR tubes. The results for the triethylamine nucleophile areshown in Figure 1.

Reaction of 3�2 with As(SPh)3

The progress of the reaction of 3�2 and As(SPh)3 in the colorlesssolution is described in the Results section. In order to identify theproduct giving the 1.352 ppm peak in the 1H NMR spectrum, thesupernatant from a preparative run (0.45 mmol scale) was chroma-tographed (silica gel 35 g in petroleum ether) eluting with a) petro-leum ether (150 ml), b) ether/petroleum ether 1:5 (150 ml), c) ether/petroleum ether 1:1 (50 ml), and d) ether (100 ml), collecting 50 mlfractions. Fractions 1-5 gave 24 mg of a solid which was mostlyPhSSPh, and fraction 6 gave a mobile, very faint yellow, nearlyodorless oil (186 mg) which by 1H NMR (Table 1) was a mixture ofPhSSPh (�20 %), 7 (�70 %), 2 (�3 %), and an unknown impurity(�10 %). In the aromatic region, the ortho hydrogens of 7(7.41�7.45 ppm) are well separated from the ortho hydrogens ofPhSSPh (7.48-7.51 ppm). The spectrum was identical (except inintensities) in the aromatic region, to the spectrum of an equimolarmixture of Me2As-SPh and PhSSPh, prepared by reduction ofcacodylic acid with thiophenol [15, 47] but differed in the Me2As-singlets: Me2As-SPh (8) at 1.336 ppm, while Me2As-S-SPh (7) at1.353 ppm. The singlet at 1.353 was seen in the reaction of 3 withthiophenol where 7 is expected to be a product (Sideris et al., inpreparation). The MALDI mass spectrum of the mixture underinvestigation showed peaks at m/z 402 (35), 379 (60), 374 (100), 335(35), 294 (70), and 212 (52 % relative abundance) from which wecould not directly deduce the molecular weight of 7. From the1H NMR spectrum of known quantities of the mixture and DMAP,a molecular weight of 274 was calculated. This is close to the mo-lecular weights of Me2AsSSPh (7) (246), Me2AsSSSPh (278), andMe2As-SPh (8) (214). The last two compounds do not fit the 1HNMR spectrum of the mixture in the Me2As- region. The yield of7 was 121 % assuming that 1 mol of 3 gives 1 mol of 7 and theyield of PhSSPh was 62 % calculated for a reaction of 1:1 of 3and (PhS)3As.

Acknowledgment. We thank Professor G. Sindona (Universita dellaCalabria, Italy) for running the MALDI mass spectrum, andAssistant Professor G. M. Tsivgoulis (University of Patras) for veryuseful discussions.


References

[1] R. Bunsen, Justus Liebigs Ann. Chem. 1843, 46, 1.[2] N. Cameron, J. Trotter, J. Chem. Soc. 1964, 219.[3] M. W. Fricke, M. Zeller, H. Sun, V. W.-M. Lai, W. R. Cullen,

J. A. Shoemaker, M. R. Witkowski, J. T. Creed, Chem. Res.Toxicol. 2005, 18, 1821.

[4] L. S. Sagan, R. A. Zingaro, K. J. Irgolic, J. Organomet. Chem.1972, 39, 301.

[5] C. A. Applegate, E. A. Meyers, R. A. Zingaro, A. Merijanian,Phosphorus Sulfur 1988, 35, 363.

[6] L. Silaghi-Dumitrescu, M. N. Gibbons, I. Silaghi-Dumitrescu,J. Zukerman-Schpector, I. Haiduc, D. B. Sowerby, J. Or-ganomet. Chem. 1996, 517, 101.

[7] L. Silaghi-Dumitrescu, I. Silaghi-Dumitrescu, I. Haiduc, Rev.Roum. Chim. 1989, 34, 305.

[8] L. Silaghi-Dumitrescu, I. Haiduc, J. Organomet. Chem.. 1983,252, 295.

[9] W. R. Cullen, Can. J. Chem. 1963, 41, 2424.[10] A. T. Casey, N. S. Ham, D. J. Mackey, R. L. Martin, Aust. J.

Chem. 1970, 23, 1117.[11] L. Silaghi-Dumitrescu, S. Pascu, I. Silaghi-Dumitrescu, I.

Haiduc, M. N. Gibbons, D. B. Sowerby, J. Organomet. Chem.1997, 549, 187.

[12] R. A. Zingaro, K. J. Irgolic, D. H. O’Brien, L. J. Edmonson,Jr., J. Am. Chem. Soc. 1971, 93, 5677.

[13] M. N. Haikou, G. M. Tsivgoulis, P. V. Ioannou, Z. Anorg. Allg.Chem. 2004, 630, 1084.

[14] G. M. Tsivgoulis, T. C. Fotopoulou, P. V. Ioannou, Phos-phorus, Sulfur Silicon 2006, 181, 413.

[15] A. Terzis, P. V. Ioannou, Z. Anorg. Allg. Chem. 2004, 630, 278.[16] P. D. Bartlett, G. Lohaus, C. D. Weis, J. Am. Chem. Soc. 1958,

80, 5064.[17] H. R. Hansen, R. Pickford, J. Thomas-Oates, M. Jaspars, J.

Feldmann, Angew. Chem. 2004, 116, 341; Angew. Chem. Int.Ed. 2004, 43, 337.

[18] E. Schmeisser, A. Rumpler, M. Kollroser, G. Rechberger, W.Goessler, K. A. Francesconi, Angew. Chem. 2006, 118, 157;Angew. Chem. Int. Ed. 2006, 45, 150.

[19] K. T. Suzuki, B. K. Mandal, A. Katagiri, Y. Sakuma, A. Ka-wakami, Y. Ogra, K. Yamaguchi, Y. Sei, K. Yamanaka, K.Anzai, M. Ohmichi, H. Takayama, N. Aimi, Chem. Res.Toxicol. 2004, 17, 914.

[20] H. R. Hansen, A. Raab, M. Jaspars, B. F. Milne, J. Feldmann,Chem. Res. Toxicol. 2004, 17, 1086.

[21] N. Scott, K. M. Hatlelid, N. E. MacKenzie, D. E. Carter,Chem. Res. Toxicol. 1993, 6, 102.

[22] A. N. Sobolev, V. K. Belsky, I. P. Romm, N. Yu. Chernikova,E. N. Guryanova, Acta Crystallogr. 1985, C41, 967.

[23] R. D. W. Kemmitt, R. H. Nuttall, D. W. A. Sharp, J. Chem.Soc. 1960, 46.

[24] J. J. Daly, J. Chem. Soc. 1964, 3799.[25] A. N. Sobolev, V. K. Belsky, N. Yu. Chernikova, F. Yu.

Akhmadulina, J. Organomet. Chem. 1983, 244, 129.[26] C. J. Pouchert, J. Behnke, The Aldrich Library of 13C and 1H

FT NMR Spectra, Aldrich Chemical Company, 1993. a) vol.1, pp. 481, 482; b) vol. 3, p. 236.

[27] D. N. Harpp, D. K. Ash, R. A. Smith, J. Org. Chem. 1979,44, 4135.

[28] G. K. Helmkamp, D. C. Owsley, W. M. Barnes, H. N. Cassey,J. Am. Chem. Soc. 1968, 90, 1635.

[29] E. F. V. Scriven, Chem. Soc. Rev. 1983, 12, 129.


[30] G. Höfle, W. Steglich, H. Vorbrüggen, Angew. Chem. 1978, 90,602; Angew. Chem. Int. Ed. Engl. 1978, 17, 569.

[31] J. R. Lloyd, N. Lowther, G. Zsabo, C. D. Hall, J. Chem. Soc.,Perkin Trans. II 1985, 1813.

[32] E. Fluck, H. Binder, Z. Anorg. Allg. Chem. 1967, 354, 139.[33] D. I. Loewus, F. Eckstein, J. Am. Chem. Soc. 1983, 105, 3287.[34] P. C. J. Kamer, H. C. P. F. Roelen, H. van den Elst, G. A. van

der Marel, J. H. van Boom, Tetrahedron Lett. 1989, 30, 6757.[35] K. Takahashi, J. Yamasaki, G. Miyazima, Bull. Chem. Soc.

Jpn 1966, 39, 2787.[36] G. C. Pappalardo, R. Chakravorty, K. J. Irgolic, E. A. Meyers,

Acta Crystallogr. 1983, C39, 1618.[37] G. Distefano, A. Modelli, A. Grassi, G. C. Pappalardo, K. J.

Irgolic, R. A. Pyles, J. Organomet. Chem. 1981, 220, 31.[38] G. C. Pappalardo, K. J. Irgolic, R. A. Pyles, E. Montoneri,

Spectrochim. Acta 1982, 38A, 363.[39] G. M. Tsivgoulis, D. N. Sotiropoulos, P. V. Ioannou, Phos-

phorus, Sulfur Silicon 1998, 141, 97.[40] T. B. Brill, N. C. Campbell, Inorg.Chem. 1973, 12, 1884.[41] V. Nischwitz, S. A. Pergantis, Anal. Chem. 2005, 77, 5551.[42] M. Förster, H. Hertel, W. Kuchen, Angew. Chem. 1970, 82,

842; Angew. Chem. Int. Ed. 1970, 9, 811.


[43] M. Schmidt, W. Siebert, in: Comprehensive Inorganic Chemis-try, J. C. Bailar Jr., H. J. Emeleus, R. Nyholm, A. F. Trotman-Dickenson (eds), Pergamon Press, Oxford, 1973, vol. 2, pp.831, 833.

[44] F. Feher, H. Weber, Z. Elektrochem. Angew. Phys. Chem. 1957,61, 285.

[45] P. D. Bartlett, G. Meguerian, J. Am. Chem. Soc. 1956, 78,3710.

[46] R. A. Zingaro, R. E. McGlothin, R. M. Hedges, Trans. Fara-day Soc. 1963, 59, 798.

[47] T. D. Sideris, P. V. Ioannou, Phosphorus, Sulfur Silicon 2006,181, 751.

[48] G. S. Levskaya, A. F. Kolomiets, J. Gen. Chem. USSR 1967,37, 855.

[49] V. S. Gamayurova, V. I. Savdur, B. D. Chernokal’skii, J. Gen.Chem. USSR. 1979, 49, 153.

[50] T. Koenig, W. D. Habicher, K. Schwetlick, J. Prakt.Chem.1989, 331, 913.

[51] S. V. Serves, Y. C. Charalambidis, D. N. Sotiropoulos, P. V.Ioannou, Phosphorus, Sulfur Silicon 1995, 105, 109.

[52] N. Muller, P. C. Lauterbur, J. Goldenson, J. Am. Chem. Soc.1956, 78, 3557.

the electrophilic character of bunsen's cacodyl disulfide, me2as(s)-s-asme2

Documents