COMLICNICATIONS T O THE EDITOR
Feb. 5 , 1960
DIALKYLALKYLTHIOBOROXATES AXD TRIALKYLBORATES
CHARACTERISTICS O F
Yield, Compound
% 45 38 29 72 63 58
B.p. (mm.),
c.
749
U'ith regard to the infrared absorption spectrum of polyvinyl chloride, the ratio of the absorbance a t 635 cm.-l to that a t 692 ern.-' has been used as a measure of crystalline syndiotactic units in the polymer.2-6 Table I gives the D635,/D692 ratio for polyvinyl chloride prepared in bulk at high and low temperatures and in n-butyraldehyde at high temperature.
Ethylene n-butyl thioboronate 84 (6) Ethylene i-butyl thioboronate 83 (6.5) Di-n-amyl-n-butyl thioboronate 132 ( 0 . 5 ) Di-n-amyl-l-butyl thioboronate 122 (0 2) Di-n-butyl-s-butyl thioboronate 105 ( 0 . 8 ) Di-n-butyl-n-hexyl thioboronate 119 ( 0 . 3 ) Tri-n-amyl thioborate 65 164 ( 0 . 4 ) Tri-n-butyl thioborate 70 117 (1 0) ROHM & HAASCOMPANY REDSTONE ARSEXALRESEARCH DIVISION HUNTSVILLE, ALABAMA M FREDERICK HAWTHORNE RECEIVED DECEMBER 17, 1959
TABLE I INFRARED .4BSORPTIOs Polyvinyl Chloride Prepared in
Bulk a t +50" Bulk a t -70" %-Butyraldehyde a t $50"
DATA Dsis/D~s?
1.5 2.6 4.3
An increase in the absorbance ratio is indicative of increased crystallinity. The infrared absorption CRYSTALLINE POLYVINYL CHLORIDE measurements were made using films cast on Sir: potassium bromide discs from tetrahydrofuran or We have prepared crystalline polyvinyl chloride cyclohexanone solutions; for the polymer prewith conventional free-radical catalysts a t 50' using pared in n-butyraldehyde the film was cast from a aliphatic aldehydes as polymerization media. X1- cyclohexanone solution a t 120'. Based on the though all aliphatic aldehydes tried have produced above evidence, we have assigned the syndiotactic this stereoregulating effect, n-butyraldehyde and structure to this polyvinyl chloride. 2-ethylhexylaldehyde appear to give the best The stereoregulating effect of aliphatic aldehydes results. As is to be expected from the chain trans- on the free-radical polymerization of vinyl chloride fer activity of aldehydes, very low molecular weight is not yet understood. Study of the mechanism polymers are produced, e.g., approximately 5,000. of this effect is in progress. The reaction is carried out a t 50' using a mole to Acknowledgments-The author is grateful to mole ratio of monomer to aldehyde and a milliI. E. Smiley for the X-ray diffraction data and J. mole of free-radical catalyst per mole of monomer. A typical polymerization using 72.0 g. J. Mannion for the infrared absorption measureof n-butyraldehyde, 62.5 g. of vinyl chloride, and ments. DEPARTMENT 0.16 g. of azodiisobutyronitrile in 9 hours yields RESEARCH P. H. BURLEIGH r j I A M O X D hLKALI COMPASY 9.13 g. of polyvinyl chloride. The crystalline PAIXESVILLE, OHIO material in tetrahydrofuran gives a cloudy soluRECEIVED DECEMBER 4, 1959 tion which does not clear up on heating. The same effect is noted with cyclohexanone but on heating to 120' the solution becomes clear. A NOVEL SYNTHESIS OF BENZOCYCLOBUTENE DERIVATIVES' The crystallinity, and hence stereoregularity, of the polymers was investigated using X-ray Sir: diffraction and infrared absorption methods. The About a year ago we reported2 the formation of regular sequences in polyvinyl chloride crystallize a stable dipositive carbonium ion from trichlororeadily so that changes in crystallinity are directly methylpentamethylbenzene (I). We now wish related to changes in regularity. The unoriented to record another unusual and useful reaction of this X-ray pattern of the polyvinyl chloride prepared and related substances. in n-butyraldehyde possesses sharp lines with I is a colorless, crystalline solid, m.p. 94.5"d" spacings a t 5.16 (m), 4.66 (s), 3.65 (m), 95.0'. It was observed that a t or slightly above 3.44 (m), 2.84 (m), 2.56 (w), 2.28 (m), 2.07 (w), its melting point, I evolved hydrogen chloride. 1.78 (yw), 1.59 (vw), 1.40 (vw), 1.25 (\7w), 1.14 In a quantitative experiment, 1.074 g. (4.05 (vw) A. (letters in parentheses indicate relative mmoles) of I was heated a t 110-125' in a dry nitrointensities where s = strong, m = medium, w = gen stream which subsequently passed through weak, vw = very weak). These values agree well standard sodium hydroxide. In the first hour, with those reported by Fordham, Burleigh and 3.76 mmoles of hydrogen chloride was collected. Sturm.' The lines with "d" spacings at 3.65, Two more hours netted 0.50 mmole, and after 1.40, 1.25, and 1.14 A., however, have not been three additional hours, only 0.03 mmole more of reported previously. hydrogen chloride was produced (a total of 105.970 of the theoretical for one mole). On cooling, there (1) J. W. L. F o r d h a m , P. H . Burleigh, a n d C. L. S t u r m , Abstracts 135th A.C.S. Meeting, Boston, April, 1959, was isolated a crystalline solid 11, m.p. 73-74' (2) S. Mizushima, T. Shimanouchi, K . N a k a m u r i , M. H r y a s h i , (from pentane). Calcd. for CleH14C12: and S . Tsuchiya, J . Chcm. Phys., X6, 970 (1957). A NEW METHOD FOR THE PREPARATION OF
(3) R. J. Griaenthwaite and R. F. Hunter, Chem. & I n d . , 433 (1959). ( 4 ) S. Krimm, A . R. Berenn, V. L. Folt a n d J . J . Shipman, i b i d . , 433 (1969). (5) T. Shimanouchi, S . Tsuchiya a n d S. Mizushima, J. Chem. P h y s . , 30, 1365 (1959). (6) S . Krimm, Society o j Plasfic Engineers Journal, (Sept.) 797
(1959).
(1) Acknowledgment is made t o t h e donors of t h e Petroleum Research F u n d , administered b y t h e American Chemical Society, for s u p p o r t of t h i s research, under grant 488-C. (2) H. H a r t and R. W. Fish, THIS JOURNAL, EO, 5894 (1958). A detailed report of this work is in preparation. (3) All analyses by Spang Microanalytical Laboratory, P. 0. Box 1111, Ann Arbor, Michigan.
750
COMMUNICATIONS TO THE EDITOR
1701.
82
Its synthetic utility and the chemistry of the prod-
I
1
2 70
a,50%
IN
CC14,
55"
(6) T h e results of a preliminary deuterium exchange experinient with 111 are of some interest. 111 (0.397 g . , 1.50 mmolesi dissolved in 4 ml. of DzO and 6 mi. of ethanol was refluxed for t w o hours. A 4-ml.
z5 )
A i
165
I m, 25$
I ucts are being e ~ p l o r e d . ~ , ~
I N CC14,
53"
sample was quenched in 20 ml. of ice-cold distilled water, a n d extracted immediately with three 10-ml. portions of pentane. Recovery of I11 was nearly quantitative; t h e crystalline product showed considerable C - D absorption a t 4 . 4 5 ~ ;t h e carbonyl a t 5 . 7 ~ was unaltered. (7) We are indebted t o M r , F r a n k J. Chloupek f o r stimulating discussions of this work.
I\
THEKEDZIECHEMICAL LABORATORY MICHIGAN STATE UXIVERSITP HAROLD HART EASTLAXSING, MICHIGAS RICHARD 1%'. FISH RECEIVED DECEMBER 23, 1959
Fig. l.-Nuclear magnetic resonance spectra a t 60 Mc.: numbers are cps. with respect to benzene as an external reference,
ALLYLCOBALT CARBONYLS
Sir :
We wish to report the formation of a new type of cobalt carbonyl compound, allylcobalt tricarbonyl. When excess NaCo(CO)1 reacts with allyl bromide in ether a t 25', exactly one mole of carbon monoxide precipitate with aqueous-alcoholic silver nitrate; is liberated per mole of bromide. Distillation after removing the silver chloride, work-up of the yields a reddish yellow, air-sensitive liquid which filtrate, or alternatively by direct hydrolysis can be crystallized from pentane to give a yellow of I1 with aqueous acetone, there was obtained in crystalline solid, m.p. -33 to -32'. a4naZ.Calcd. for CjHbOdCo: CO, 45.G; Co, 32.0. excellent yield a crystalline ketone 111, m.p. 153-154' (from ethanol). Anal. Calcd. for CIS- Found CO (by treatment with Iz in methanol), H140: C, 82.72; H, 8.10. Found: C, 82.57; 44.8; Co, 31.5, 31.1. The compound was shown H, 8.16. Oxime, m.p. 173-175', 2,4-DNP, m.p. to be diamagnetic by e.p.r. The n.m.r. spectrum I11 had a carbonyl band a t 5 . 6 7 ~ of the allyl compound shows three types of hydro282-283'. and bands in the ultraviolet (ethanol) a t 306 mp gen in the ratio 2:2:1. Although the spectrum (log E = 3.47) and 265 r n M (log E = 4.18). The was not sufficiently well resolved to differentiate n.m.r. spectra of I1 and I11 are shown in the f i g ~ r e . ~unequivocally between the unsymmetrical strucThese data are best interpreted in terms of the ture I and the symmetrical structure 11, the prereaction scheme
C, 62.89; H , 6.16; C1, 30.95; mol. wt., 229. Found: C, 62.93; H , 6.27; C1, 30.88; Rast molecular weight (camphor), 201. I1 gave an immediate
HCH2
HC,l,Co(C0)3
4W '\
HCi
;Co(CO)j
\.,
C ' H:
I
111
x
cH z I1
ponderance of evidence would appear to favor the symmetrical structure. Whereas alkylcobalt tetracarbonyls are quite unstable thermally,213allylcobalt tricarbonyl appears to be stable indefinitely a t room temperature. The equilibrium mixture of 1-bromo-2-butene and 3-bromo-1-butenereacts with NaCo(C0)4 in an entirely analogous fashion. One mole of carbon monoxide is evolved and distillation yields a redyellow oil identical with the reaction product of butadiene and HCo ( CO)r. ,5 The compound is presumably 2-butenylcobalt tricarbonyl.6 If the reaction of allyl bromide with XaCo(C0)r is carried out a t 0' in the presence of carbon monoxide, about 0.5 mole of carbon monoxide is absorbed. At this point the infrared spectrum of the solution shows a band a t 5.5 p, indicating the presence of an acylcobalt c o m p o ~ n d . ~ Gas is
From mesitylene, one can obtain a cornpound presumably analogous to 11, m.p. 55-57' (from hexane). Anal. Calcd. for C1OHIOCl:!: C, 59.72; H, 5.01; C1, 35.26. Found: C, 59.90; H, 5.10; C1, 35.27. In a preparative scale experiment, without purification of the intermediate 11, an 89% yield of I11 was obtained from 5 g. of I. This constitutes, then, an excellent preparative procedure for benzoW e are indebted t o D r . J . C. W. Chien for t h e cyclobutenones and compounds derived t h e r e f r ~ m . ~and(1)interpretation of t h e e . p r . a n d n . m r. spectra. (4) W e are indebted t o M r . James C. Woodbrey for determining these spectra. ( 5 ) For other benzocyclobutene syntheses, see M. P. Cava a n d A. A. Deana, THISJ O U R N A L , 81, 4266 (1959), a n d earlier papers b y Cava's group; L. IIorner, W. Kirmse a n d K . M u t h , Chem. Bcr., 91, 430 (1D58); A. P. terBorg and A. F. Bickel, Proc. Chem. Soc., 283 (1958); and IC. R . Jensen and W. E. Coleman, J . Org. C h e m . , 23, 869 (1958).
determination
(2) W. Hieber, 0. Vohler and G. Braun, Z . Salurfororsch., lSb, 192 (1958). (3) R. F. Heck and D. S. Breslow, unpublished work. (4) 13. B. Jonassen, R. I. S t e a m s , J . K e n t t a m a a , D . W. Moore a n d A. 0 . Whittaker, THISJ O U R N A L , 80,2586 (195s). ( 5 ) a'.TV. Prichard, U. S . Reissue 24,653 (1959). ( 6 ) H. B. Jonassen independently has arrived a t t h e same conclusion (London Conference on Coordination Chemistry, April, 1959).
