July 20, 1960
COMMUNICATIONS TO THE EDITOR
3795
reaction is exothermic but easily controlled. These known phosphonous dichlorides thus have fiK& Xmsi ( m r ) (log e) e been prepared: p-chlorophenyl-, b.p. 132-133' I1 9 . 1 f 0.2" 410-425 (3.88)d a t 20 mm. (reportedj6133' a t 20 mm.), 37% crude 111 1 1 . 1 =k 0 . 2 b ; 1 0 . 8 f 0.2' 510 ( 3 89) yield; 9-tolyl-, b.p. 109-110' a t 11 mm. (reported,' IV 1 1 . 5 & 0.2"; 1 1 . 6 + O . l b ; 107-110" a t 10 mm.), llY0 yield. Both com1 1 . 3 i.0 . 1 " 533 ( 3 . 8 0 ) pounds were analyzed and gave the correct values. By titration in acetonitrile with dioxane-perchloric acid, Additional proof of their identity was obtained by according to the method of H. K. Hall, J . Phys. Chem., 60, 63 (1956). *Spectrophotometric, in dioxane, using piperi- hydrolysis to the corresponding phosphonous acids, dine-piperidine perchlorate as internal buffer, according to each of which had melting points in agreement the method of L. P. Hammett and A . J . Deyrup, TH!S with literature values. JOURNAL, 54, 2721 (1932). Spectrophotometric, as in To demonstrate the versatility of the new footnote b , in acetonitrile. Shoulder. e In isooctane. method, two new and otherwise difficultly obtainCompound I1 gives a colorless solution (Amax able phosphonous dichlorides were prepared : m 342) no shoulder a t 410-425 mp) in ethanolic per- chlorophenyl-, b.p. 124-125' a t 18 mm., 33% chloric acid, signifying conversion to the conjugate yield (Anal. Calcd. for CGH4ClaP: C, 33.76; acid ; the change is quantitatively reversed by H , 1.89; P, 14.51. Found: C, 34.04; H, 2.06; alkali, even after storage of the acidic solution. P, 14.65). p-cyanophenyl-, b.p. 127-128' a t The unstable crystalline perchlorate of I1 is ob- 4.5 mm., 25% yield (Anal. Calcd. for C7H4tained as needles (from ethyl acetate-ethanol) C1,NP: C, 41.21; H, 1.98; P, 15.19. Found: which decompose above about 80'. A n d . Calcd. C, 41.42; H , 2.03; P, 15.17). Each was hydrofor CI1H,,O4NC1: C, 51.26; H, 4.69. Found: lyzed to the phosphonous acid : m-chlorophenyl-, C, 51.50; H, 5.20. This substance shows no recrystallized from carbon tetrachloride, m.p. P: =+N-H absorption in the infrared (KBr disc). 90.5-91.5" ( A n d . Calcd. for C ~ H ~ C ~ O Zc, Presumably, protonation occurs in the five-mem- 40.80; H, 3.43; P, 17.54. Found: C, 41.17; H , bered ring. 3.65; P, 17.11). p-Cyanophenyl-, recrystallized Further properties and reactions of I1 and re- from ethanol-carbon tetrachloride, m .p. 166-167' lated substances are being investigated. (Anal. Calcd. for C7H6h'OzP: c, 50.31; H, DEPARTMENT OF CHEMISTRY JEROME A . BERSON 3.62; P, 18.54. Found: C, 50.61; H, 3.87; P, USIVERSITYOF SOUTHERN CALIFORSIA EARLM. EVLETH 18.20). Los ASGELES 7 , CALIFORNIA ZACHARIASHAMLET The following experimental procedure is typical. RECEIVED JUNE 10, 1960 All operations were conducted in a nitrogen atmosphere. To 200 ml. of dry ethyl acetate, 0.20 mole of p-cyanobenzenediazonium fluoroborate and 1.7 g. A NEW SYNTHESIS OF ARYLPHOSPHONOUS of cuprous bromide, was added 0.23 mole of phosDICHLORIDES' phorus trichloride. After stirring for 30 min., the Sir: mixture was warmed to 40°, a vigorous reaction X few years ago,, Doak and Freedman announced commencing. When complete, 0.20 mole of magan aromatic phosphonic acid synthesis destined to nesium was added slowly with cooling to hold the become the most widely useful of the several temperature a t 40-50'. The reaction mixture methods known.3 Readily prepared aryldiazo- was freed of solvent and the residual semi-solid nium fluoroborates are the starting materials ; mass extracted with three 100-ml. portions of benthese are treated with phosphorus trichloride or zene-heptane mixture, 1: 1 v./v., The extract was tribromide in a dry solvent in the presence of a flash-distilled to remove solvent and the residual cuprous halide as catalyst. The mixture is then liquid then vacuum-distilled. There was obtained hydrolyzed and the phosphonic acid obtained first a 2-g. fraction, b.p. 44-61' a t 6 mm., solidifygenerally in 30-5070 yield. The intermediate of ing a t room temperature. This product is indethis reaction has not been subjected to study but pendent of the structure of the starting diazonium has been suggested by Crofts" to be ArPXa+BF4-. fluoroborate and variable in quantity; it is a fuming, We reasoned t h a t the intermediate, if hydrolyz- phosphorus-free substance, and is believed to be able to a phosphonic acid, may well bear chemical the boron trifluoride-ethyl acetate complex (b.p. similarity to an aryl tetrahalophosphorane, and 123' a t 772 mm., m.p. 3 7 . 5 O g ) . The phosphonous hence might be reducible to an arylphosphonous dichloride (10 g.) was collected a t 130-137' a t 6 dichloride, XrPClz.6 IVe have indeed found that mm. It was redistilled through a 6 in. packed magnesium shavings added to the diazonium fluoro- column, and after a small forerun the main cut borate-phosphorus trichloride reaction product in was collected a t 127-128' a t 4.5 mm.; it provided ethyl acetate solvent effect this reduction. The the analytical data listed above. An alternative (1) Supported b y a g r a n t from t h e Duke University Research Counisolation procedure is to distill the solvent-free cil. reaction product directly, eliminating the extrac( 2 ) G. 0 . Doak a n d L. D. Freedman, THIS JOURNAL, 73, 5658 tion. The large solid residue, however, makes this (1951). distillation somewhat difficult. (3) L. D . Freedman a n d G. 0. Doak, Chem. Reu., 67,479 ( 1 9 5 7 ) . (4) P. C. Crofts, Q U Q Y ~ . Reviews, 12, 341 (1958). It appears t h a t we have here a new method of ( 5 ) L. D . Quin a n d C. H. Rolston, J . Ovg. Chem., 23, 1603 (1958), synthesis of arylphosphonous dichlorides, and and references cited therein. E . P. Komkov, K. U. K a r a v a n o v and TABLE I
Compound
S. 2. E v e n , Zhlrr. obshchei Khim., 28, 2963 (1958), have reported a somewhat similar preparation of alkylphosphonous dichlorides by reducing with metals t h e complex salts R P C l a t AlCla- (J. P. Clay, J . O v g . Chem., 16, 892 (1951)).
(6) D. R . N i j k , Rec. frau. chim., 41,461 (1922). (7) B. Buchner a n d L. B. Lockhart, Jr., THISJOURNAL, 73, 755
(1 951). (8) G.T.Morgan a n d R . Taylor, J . Sac. Chem. Ind., 10, 869 (1931).
3796
COMMUNICATIONS TO THE
hence of arylphosphonous acids, of potentially as wide versatility as the Doak-Freedman phosphonic acid synthesis. \Vork is continuing to define the scope of the synthesis and establish optimum reaction conditions. DEPARTMEXT O F CHEMISTRY LOUIS D. QUIX J. STEVENSOS HUMPHREY, JR. DUKEUKIVERSITY SORTH CAROLIXA DURHAM, RECEIVED M A Y 20, 1960 STEREOCHEMISTRY OF SUBSTITUTION AT SILICON. REACTIONS OF T H E SILICON-OXYGEN BOND WITH INVERSION AND RETENTION OF CONFIGURATION Sir :
For the first stereochemical studies of the siliconoxygen bond we have used the optically active compounds R3Si*-O--, all of which contain the a-naphthylphenylmethylsilylgroup (a-NpPhhleSi-). Reaction (1) below is a n unusual case in stereochemistry. Formation of optically pure potassium silanolate (11) rigorously proves retmtion of configuration as the stereochemical path.
!aID
-7O"(II)
The optically active disiloxane (I) was synthesized from I1 and (f)RBSi*C1, and I1 was obtained from reaction of (+)R&*OH with either @assium hydroxide or potassium. R3Si*-OCH3 [a]D
-16'
KOH (s)
-----+
(111)
xylene
R8Si*-OK
[a]D
+68" ( I v )
Retention of configuration in ( 2 ) is rigorously demonstrated by formation of I11 from IV and dimethyl sulfate, a reaction which does not involve the asymmetric center. R3Si*-O-COCH3
KOH (s) ___f
xylene
R$Si*-OK
(3)
[a]D+60° (VI) CHBOH R3Si*-O-COCH3 ----f R,Si*-OCH3 (4) pentane; amine [ a ] D $16" (Is-) [ a ] D -12' (VII) [a]D
+18" (V)
Predominant inversion in reactions (3) and (4) is demonstrated rigorously by formation of (-) R3Si*-OCOCH3 from VI and acetyl chloride, a reaction which does not involve the asymmetric center. Sensitivity of stereochemistry t o variations in leaving group and solvent is apparent. For more basic leaving groups (methoxy and siloxy relative t o acetoxy) poor ionizing solvents favor retention probably by favoring SNi reaction of relatively undissociated complexes.
EDITOR
1-01. s2 CHJ
\0 _ _ _ - _ _ - K /
, R $ i ' _ _ _ _ _ _ _OH Ssi retention i n xylene
The change to a good ionizing solvent results in inversion reactions even for the methoxysilane (111). Thus, I11 is rapidly racemized in metham1 solvents by small concentrations of inethoxide ion. Furthermore, I11 is hydrolyzed by base in aqueous ~ with preacetone to (+)R3Si*OH, [ a ] +a", dominant inversion in a reaction competitive with silanol racemization. For the less basic leaving groups, inversion seems to be favored in both types of solvents for the CYnaphthylphenylmethylsilyl compounds. RsSi*OH,' [ D I D +26', in xylene, shaken with powdered potassium hydroxide (ROH, 85yc; H?O, 13-147,) gave 11. (Hydrolysis of I1 under controlled conditions yields the original silanol.) Reaction I1 with (+)R3Si*Cl2 gave I (m.p. SS89'; found: Si, 11.1; characteristic Si-0-Si infrared maximum a t 9.35 microns). Reactants were heated (steam-bath) for one hour in reactiors (1) and (2). Reaction of IV with dimethyl sulfate gave rapid formation of 111,' [ a ] D - 1 G " , 1n.p. (i4". Reaction of I1 with excess acetyl chloride gave L? (found: Si, 9.2; s a p . equiv., 306). Reaction ( 3 ) is rapid a t 23" and gave LTIin a reaction a t least S570 stereospecific. Reaction (1) gaVe VI1 in 27'3 yield in a reaction a t least 9Oc/;, stereospecific. We thank Dow Corning Corporation for generous support. (1) I,. H. Sommer and C . L. Frye, THISJ O U R N A L , in press. ( 2 ) L. H . S u m m e r and C. L. Frye, ibiil , 81, 1013 (1959)
LEO €I. SOMMER DEPARTMENT O F CHEMISTRY THEPENSSYLVAXIA STATE USIVERSITT CECILL. FRVE PARK, Ps. UNIVERSITY RECEIVED MAY28, 1960 T H E MECHANISM OF T H E VON RICHTER REACTION
Sir: The transformation of aromatic nitro compounds t o carboxylic acids through the agency of alcoholic potassium cyanide was first described in 1871 by von Richter.' Since the reaction was invariably attended by loss of the nitro function, von Richter initially assumed that cyanide ion, from which the carboxyl group derived, displaced the nitro group directly from its position on the aromatic nucleus. The concurrent structural investigations of Meyer, Wurster, Solkowsky and especially of Griess and Korner2 ultimately compelled von Richter to revise the structures assigned to his acidic products and, in 1875, in the last of his papers on the subject, he concluded that the carboxyl function must take up a position on the aromatic ring ortho to t h a t vacated by the nitro group. This formulation of the reaction has since been amply (1) V. van Richter, Bey., 4, 21 (187l),andsucceeding papers, ibid., 4, 459, 553 (1871); 7 , 1146 (1874); 8, 1418 (1875). Chim. I f d . , 306 (2) P . Griess, ibid., 5 , 192 (1872); W Korner, GQZZ. (1874).
