“62
CHEVESWALLIXG A N D MYRNA SCHMIDT PEARSON [COKTRIBUTIOS FROM
THE
VOl. 86
DEPARTMENT OF CHEMISTRY, COLUMBIA UNIVERSITY, NEW YORK27, N. U.]
Some Radical Reactions of Trivalent Phosphorus Derivatives with Mercaptans, Peroxides, and Olefins. A New Radical Cyclization1 BY CHEVES \
l
TA N D ~ MYRNA ~ SCHMIDT ~ ~PEARSOS~ ~ ~ ~
R E C E I V EJDA N U A R8, Y 1961 In the radical reaction between n-butyl mercaptan and triethyl phosphite, thiyl radicals are shown to react with phosphite a t rates intermediate between their addition to styrene arid cyclohesene. By suitable adjustment of reagent concentrations the n-butyl radicals formed may be caused t o alkylate styrene and methyl acrylate in low yield. Reaction of 6-tnercapto-1-hesene with triethyl phosphite produces the 6-hexen-1-yl radical which cyclizes chiefly to met1i)lc~clopentane. This “unexpected” direction of radical ring closure and the failure of the 5-penten-1-yl radical t o cyclize is suggested to have a steric origin. By competitive reactions it is shown t h a t t-butox\- radicals attack triethyl phnsphite 500-1000 times as rapidly as they abstract hydrogen from typical hydrocarbons, and some 82 times as rapidly as they attack triphenylphosphine. Relative reactivities of thivl radicals toward trivalent ohomhorus eive an order of reactivity P(n-C$Hg)?> P(OCZH:,):i> P(CsH:)., >
Previous work from these laboratories3 has shown t h a t the reaction between trialkyl phosphites and mercaptans
+ l’(OEt)s +R H + SP(OEt)3
RSH
+ I’(OEt)a fit_
[RSJ!(OEt)a]
--+
R.
+ SP(OEt)3
k7
RS.
(1)
first described by Hoffmann, et u / . , is ~ a radical chain process involving the propagating steps RS.
additional elementary steps are possible in the chain sequence leading to two additional products
k-
+ P(OEt)3 +RSR + S P ( 0 E t ) a
(2) (3)
+ RSSR +RSR
t RS.
(5)
and that alkoxy radicals from t-butyl and cumyl peroxides are rapidly deoxygenated by phosphites to kr
RO.
t- I’(OEt)s + [ROP(OEt)s]+R .
+ OP(OEt)$
(9)
(io)
Sequence 7 and S amounts simply to the radical addition of a mercaptan to a double bond to give a sulfide b u t 2 , 9, and 10 gives the over-all reaction RSH
(4)
in which reaction 3 is replaced by R.
(8)
k8
I t was also found that a comparable reaction occurs between disulfides aiid trialkyl phosphites (again using triethyl phosphite as an example) RSSR
d
klc
+ RSH d K H + RS.
(7)
7
+ RSH ks RSCH2CH2R’+ RS. R . + CH~=CHR’ + R C H ~ H R ’ R C H ~ C H R !+ RSH +R C H X H ~ R ‘+ RS.
RSCHkHR‘
ka
R.
+ CH,=CHR’ 2R S C H ~ C H R ’
+ P(0Et)a + CH2=CHR’
+
+
R C H I C H ~ R ’ SP(OEt)3 (11)
a possible technique for free radical alkylation. We have investigated the relative rates of these processes by treating n-butyl mercaptan with triethyl phosphite in the presence of styrene and cyclohexene a t 70’ using azobisisobutyronitrile (AIBN) as initiator. Results are shown in Table I . In all reactions, sulfide
(6)
give phosphate and hydrocarbon by coupling or disproportionation of R. radicals. Similar reactions were also found to take place with trialkyl- or triarylphosphine^.^ These processes are interesting, both as among the first examples of radical addition to phosphorus to yield the phosphoranyl radical with a valence shell expanded to nine electrons,6and because they provide a method of generating carbon radicals of known structure from the corresponding mercaptan, disulfide, or peroxide. The work described here was undertaken to determine the relative rates of the intermediate steps involved and to see whether carbon radical additions t o olefins could be introduced into the chain sequence to yield over-all reactions of synthetic utility. Mercaptan-Phosphite-Olefin Reactions.-If an olefin is introduced into a mercaptan-phosphite systein, four (1) T a k e n from the P h I > , Thesis of hl. S. Peat-son, Columbia U n i v e r s i t y , IP63 Partial s1ippot.t of this wol-k b y g r a n t s f r o m t h e National Science Fnundation i n gi-atefully acknouledged ( 2 ) I)u Pont teaching fellow 1960-1R61, Esso F e l l o w , 1961-1962. ( 3 ) C Walling and R . Rabinoivitz, J . A m Chcm. .Soc., 79, ,3326 (1957); 81, 1243 (1959) ( 1 1 1’ W . H n f f m a n n . R J Isss. T. C. Simmons, and K. S. Hanzel. i b i d t i 4 1 4 (l!lzfi) C \4‘alling, 0 H . Basedow, and E . S Savas, %bid , 8!?, 2181 (19601. F liamiiez and pi. NcKelvie, i b i d . , 79, ,5829 (19.57).
TABLE I REACTIOSS OF T R I E T H YPHOSPHITB, L X - B U T YMERCAPTAN, L A N D OLEFIXS Quantitics iti millimoles, all a t 70”, 1-21 hr. reaction in the presence of A I B S unless indicated‘ [P[Olefinla [RSHIo (0Et)alo
[sp10Et)aIf [Sulfidelf
kdkz
1.5 2.3 1.9 4.5 4 5 2.1 1.9 2.3 1.3 4.5 3.5 4.3 2.5 1.9 16 4.3 4.3 4.1 1.37 4.4 0.52 9 5 29.2 17 4 1 02 4.9 0.43 9 . 5 29.2 26 2 18 0 15b 73 100 Cyclohesene 100 100 0 66 26 4.05 9 5 29 2 l!) 0 0.87 .2fj 3.14 29.2 9 5 ‘18.5 a Subscript 0 and f indicate initial and final concentration, respectively. At GO”. Styrene
was a conspicuous product, but n o alkylated olefin could be detected. Under these circumstances, yields of sulfide and phosphorothionate, SP(OEt)a, should be given by the relation
providing the reversibility’ of 7 can be neglected. ( 7 ) C Walling and U’ Helmieirh rbrd , 81, 1144 (1959)
June 5 , 1984
A NEW RADICALCYCLIZATION
Values of k7!lk2 for styrene have been calculated from the integrated form of 12 and show some scatter with an average value of 3.2. If 7 is significantly reversible, the true value would be somewhat larger and observed ratios would vary with mercaptan concentration, and i t may be noted that the lowest value calculated, 1.9, is in fact from an experiment in which mercaptan was completely consumed. The absolute value of k, has been calculated by Sivertzs as 8 X IOs a t 25' and, as a very low activation energy process, should be quite temperature insensitive. Thus kS must be approximately 2.5 X lo8,indicating that thiyl radical attack on phosphite is an extremely rapid, low activation energy process. For cyclohexene a t 70°, k7,'kz = 0.26. Qualitatively such a smaller ratio is expected since cyclohexene is in general less reactive toward radical addition than is styrene. However, comparison of the two values gives a relative reactivity styrene : cyclohexene of 12.3, significantly lower than the results of direct competition toward the structurally similar dodecanethyl radical a t 60' for which a ratio of G8 has been r e p ~ r t e d . ~ Since no alkylated olefin was detected in the experiments of Table I , it was evident t h a t a change in reaction conditions would be required to suppress the competing formation of sulfide and alkane. Taking into account the possible reversibility of i the competitions involved become olefin
A
RSH
RSCHzbHR' ---+sulfide
+ RS.
+ R. , /
SP(0Et)a
olefin\
RSH
R C H 2 6 H R ' -alkylated
prod.
+ RS.
2263 TABLE I1
ALKYLATION YIELDS~ IN SLOW ADDITIONEXPERIMEXTS Alkylated olefin,
Olefin
Conditionsb
Sulfide,
7%
%
0 B 0 Styrene A 0 B 7 Methyl acrylate B 10-15 All yields relative to SP(0Et)a produced. A, equimolecular quantities of reagents, 60-io", A I B S initiator; B, 507, excess phosphite, 120-130°, DTBP initiator. Cyclohexene
A
25
R'OPR~-+
R'OPR?
+R
(19)
Reaction 19 and the more hypothetical lS point up the fact t h a t a t least two modes of decomposition of phosphoranyl radicals are available : P-scission as in reactions 2 and 6 and an a-scission w'hich amounts to the reversible addition of a radical to trivalent phosphorus (although the leaving radical may be different from that which adds). Buckler has noted that the competition between the two modes is quite solvent dependent, and additional examples of a-scission ( 1 7 ) S. A. Buckler, ibrd.,
84, 3093 (1062)
CHEVESWALLING A N D MYRNA SCHMIDT PEARSON
2266
have been noted in the electrolytic reduction of R 4 P + to R3P and R . , ‘ * and as a step in the cooxidation of PC13 and hydrocarbons. l9 The possibility t h a t @-scissionmay involve one of the groups initially attached t o phosphorus rather than the attacking radical has been shown by the isolation of 3% toluene from the reaction of n-butyl mercaptan with benzyl diethyl p h ~ s p h i t e . ~The use of thiophenol provides a better case, since cleavage of the phenylsulfur bond should be relatively difficult. In this system we now find 10% toluene and only 47, benzene among the products obtained with A I B X a t GO’. Experimental Materials except as indicated were commercial reagents, fractionally distilled before use when appropriate and purity checked by g.1.c. or physical constants. Sulfides.--Authentic samples of n-butyl cyclohexyl sulfide and n-butyl P-phenylethyl sulfide were prepared by heating equimolar quantities of cyclohexene and styrene, respectively, with n-butyl mercaptan a t 60” in the presence of AIBN. 6-Mercapto-l-hexene.-Thiolacetic acid (0.40 mole) was added dropwise with stirring t o 0.60 mole of 1,5-IiexadienezQcontaining 1-2% AIBN in a stirred flask under nitrogen a t room temperature, and the mixture held a t 50’ overnight. Distillation yielded 207, 6-hexen-1-yl thiolacetate, b.p. 8 5 9 0 ” ( 7 mm.), nZ7D1.4792, together with unreacted diene and a high boiling residue. The material had an infrared spectrum consistent with the expected structure and was used without further purification. For hydrolysis, 13 g. of the thiokdcevdte was refluxed under nitrogen for 2 . 5 h r . with 30 ml. of water and 60 nil. of 20% methanolic K O H . After distillation of the methanol, additional water was added and the mercaptan extracted with ether, washed, dried, and distilled. The product, 6-mercapto-1-hexene, b.p. 45-50” (14 mni.), ~ 2 2 . 51.4694, ~ was obtained in 50y0 yield. I t showed a single peak on g.1.c. analysis and an infrared spectrum consistent with a n -SH group and a terniinal double bond. Due to its facile polymerization it was stored a t -25O in the dark under A-2. 5-Mercapto-l-pentene.-5-Bromo-l-pentene ( 10 g.), thiourea (5 g . ) , and 40 mi. of absolute ethanol containing 0.1 g. of 4-tbutylcatechol were refluxed under 1-2. Addition of ether and cooling t o -25” gave 11.39 g. (i’lyO) of the isothiuroniuin hydrobromide as whitc needle-like crystals, m.p. 91-93’. Hydrolysis of 10 g. of the salt was accomplished by stirring under nitrogen in the dark a t room temperature with 5 tnl. of 502, aqueous S H I O H containing 0.1 g. of hydroquinone. .At the end of 2 hr. two layers of liquid were visible. The lower aqueous layer was acidified with dilute HCI, and extracted with ether. The upper layer was n.ashct1 witli water, dried, and distilled together with the ether extract to give 2.5 g. of crude mercaptan, b . p . 37-39’ (11 m m . ) , which was stored under S2 a t - 2 5 ” . Although both infrared and 11 .m.r. spectra showed the expected -SH and vinyl groups, g.1.c. analysis indicated two contaminants I and I1 and a purity of about 80‘;lc. The three components could be separated by g.l.c., but reanalysis of the collected major peak showed t h a t the two impurities were reformed during collection. Heating the crude mercaptan or exposing i t t o air also led to conversion t o I , 11, and polymer. Compound I was ~
Hot-ne$-a n d A . h l e n t r u p A n i t , 646, 65 (1D61), h1 Finkelstein, J . O r e , C h ~ m, 27, 4076 (1962) (I!)) I‘ K . M a y o , I,. 1)urham. and K. S . Origgs, J . A m . Chrin. Soc , 8 6 , :
