284
The problem in adapting the photolysis of aromatic iodo compounds for synthetic purposes is to establish conditions which promote the scission of the carbon-iodine bonds without causing subsequent reactions of products, or undesired side reactions. Thus, the use of a quartz housing and a widespectrum light source lead to various side reactions.a However, by conduct'ing the reaction at or near room temperature and using an ultraviolet source which provides energy a t a wave length which essentially cleal-es only the carbon iodine bond, the side reactions are minimized, The synthesis of 2-hydroxybiphenyl in one step, in 60-70oJo yields of chromatographically pure product, is carried out as follows: o-Iodophenol (lg) in 50 ml. pure, dry benzene, was irradiated in a "Vyc0r~~-7100 tube, by a helical cold-cathode, low pressure mercury lamp (manufactured by Dallons Laboratories, Los Angeles, California). After twenty hours of irradiation, the release of iodine, conveniently measured by titration with thiosulfate, was complete. Isolation of the product from the organic layer and purification by two passes through an alumina column gave 0.46 g. (60% yield) of pure 2hydroxyhiphenyl, fully characterized by its m.p.4 and infrared Bpectrum. By similar procedures, in synthetically useful yields, we obtained the following products: diphenyl (from iodobenzene irradiated in benzene); 4hydroxybiphenyl (from p-iodophenol irradiated in benzene) ; 4nitrobiphenyl (from Piodo-nitrobenzene irradiated in benzene) ; a separable mixture of 2-methyl-4'-nitrobiphenyl and 4methyl4'-nitrobiphenyl (by irradiation of 4iodonitrobenzene in toluene) ; and 2-methoxybiphenyl (by irradiation of iodobenzene in anisole). The irradiation of 1g. of 2,6-di-iodo-p-cresol in benzene led to 712 mg. of a highly crystalline, iodine-free product, m.p. 66-7", the infrared epectnrm of which showed a phenolic group and polyphenyl absorption, but which has not yet been fully characterized. The structure of the fully-characterized products obtained are those to be expected for free radical attack on the substrates used.6 These reactions, which occur a t the wave lengths where iodoaromatic compounds show continuous absorption, can thus be considered as occurring, undoubtedly, by free radical mechanisms. The effective, low temperature generation of aryl radicals, and specifically of hydroxyaryl radicals, by this means, is novel. The use of diiodoarenes in this synthesis is also illustrated by the formation of pterphenyl from pdiiodobenzene and of p quaterphenyl from 4,4'-diiodobiphenyl, irradiated in benzene. We are continuing studies to determine the scope of these reactions, both as regards sources for aryl radicals, from iodo compounds, as well as solvents (4) 0. Honigschmid, Monatsh. 22,568 (1901).
VOL. 26
COMMUNICATIONS
(5) G. H.Williams, Romolytie Aromatie Substihtia, p. G', Pergamon Press. London (1960).
capable of capturing the radicals generated by the photochemical technique.
Achmowledgment. We are indebted to the'Nationa1 Institutes of Health, Grant A-703, and to the U p john Company, Kalamazoo, PvIich., for support of this study. DEPARTKENT OF CHEMISTRY WALTEB WOLF UNIVEBBITY OF SOUTHERNCALIFOENIA Los ANGELES7, CALIF. NORMAN KEAWBCH Received October 26, 1960
The Phenoxphosphinic Acid Ring System Sir: Only a few phosphinic acids are known in which the phosphorus atom is a member of a ring system.' In particular, the synthesis of phenoxphosphinic acid (I) or its derivatives has not been previously accomplished. Although phenoxarsinic acid is easily prepared by refluxing pheny1 ether with arsenic trichloride and a small quantity of aluminum
D
oHr'IoH chloride and then oxidizing the resulting chloroarsine with bromine water,) the analogous reaction with phosphorus t,richloride yields pphenoxyphenylphosphonic acid. It is not clear why-arsenic trichloride preferentially attacks the %position of phenyl ether, while phosphorus trichloride attacks the 4-position. Numerous attempts in this laboratory to prepare phenoxphosphinic acid by the intramolecular dehydration of o-phenoxyphenylphosphonic acid have met with failure.' This compound is unaffected by po1"yphosphoric acid a t 130" and is apparently sulfonated by concentrated sulfuric acid a t 100". The method of Campbell and Way,6 whereby 2biphenylylphenylphosphinic acid was cyclized to 9-phenyl-9-phosphafluorene 9-oxide by heating with an excess of phosphorus pentachloride in nitrobenzene, also failed with bphenoxyphenylphosphonic acid. We can not explain the resistance of ophenoxyphenylphosphonic acid to cyclization, but (1) Cf. L.D. Freedman and G. 0. Do&, J. Ozs. Chem .~ ,24. 638 (1959) (2) W L Lewis, C.D. Lowry, and F. H. Bergeim, J . Am. Chem. SOC.,43,891 (1921). (3)W. C. Davies and C. J. 0. R. Morris, J. C h . SOG., 2880 (1932). (4)'Dehydrstion is often an excellent way of preparing heterocyclic arsenic and antimony compounds; cf- F. G. Mann in J. W. Cook's Progress in Organic Chemistry, Vol. 4, Butterwortha Publications, London, 1958, p. 218. (5)I. G. M. Campbell and J. I(.Way, PTOG. Chem. SOC., 231 (1959).
JANUARY
1961
COIUMUNICATIONS
285
formed by nitrating fructose by different methods and arise from anhydridization, polymerization or degradation.' -a A striking demonstration of the extreme ease with which chemical changes occur in the fructose molecule, even when treated under very mild conditions, is given by treating the sugar with ethanol. Ten grams of analytically pure, dried D-fructose (the purity of the material was checked chromatographically) was suspended in 10 ml. of absolute ethanol and dissolved by gentle heating to a temperature not exceeding 76". The solution was kept a t this temperature for 12 hr. under careful avoidance of superheating on the walls of the vessel. After cooling, the clear colorless solution was subjected to paper chromatography to yield a spectrum of a t least eight spots besides the original spot of fructose. Of these, four had higher and four smaller Rf-values than fructose. By comparing them chromatographically with reference-substances,we were able to identify seven of these spots : methyl glyorcrtll; difructose anhydride 111'; methylglyoxal fructoside' ; difructose anhydride 14; diheterolevulosan I14-6; diheterolevulosan 14, @-ethylfructupyranoside-the latter has also been isolated in substance. The same results could be obtained in shorter time (in about 2-3 hr.) by boiling fructose with 95% ethanol, which caused immediate solution. The chromatogram in this caae was identical with the former one. A third method for effecting transformations of fructose even without heating, was to keep h e l y powdered, dried fructose with absolute ethanol a t room temperature for some months. Eventually, complete solution occurred and the solution yielded a chromatogram which had besides the eight spots obtained by the first two methods, two additional spots which seemed to represent glucose derivatives. It follows that under all these conditions, which exclude the action of bases or acids and avoid high temperatures, fructose undergoes anhydridization, VENEFLEAL DISEASE EXPERIMENTAL LABORATORY dimerization, isomerization and degradation to COMMUNICABLE DISEASE CENTEB U. S. PUBLICHEALTHSERVICE (&fragments. SCHOOL OF PUBLIC HEALTH LEOND. FREEDMAN None of these effects could be observed in UNIVEESITY OF NOETH CAEOLINA G . 0. DOAK CHAPEL HILL,NOBTH CN~OLINA JOYCER. EDMISTEN aqueous solution of fructose. Alcoholic solutions of glucose also failed to show any change under the Received October 19, 1960 conditions of our experiments.
the failure of kbiphenylylphosphonic acid to undergo cyclodehydration has been previously noted.6 We M y succeeded in preparing a derivative of phenoxphosphinic acid by heating ptolyl ether with phosphorus trichloride in the presence of aluminum chloride, and then hydrolyzing the reaction mixture. Since thc 4positions of ptolyl ether are blocked, the phosphorus trichloride attacked the 2-position.' The lo-chlorophenoxphosphine which was undoubtedly formed as an intermediate was presumably oxidized during the Friedel-Crafts reaction and subsequently hydrolyzed to 2,&dimethylphenoxphosphinic acid. No attempt has yet been made to isolate these intermediate chloro compounds. 2,8-Dimethylphenoxphosphinic acid ww characterized by analysis and ultraviolet absorption. The spectra of phenoxarsinic and 2,gdimethylphenoxphosphinic acids are remarkably similar. The slight bathochromic shift exhibited by the phosphinic acid can probably be attributed to the methyl groups [phenoxarsinic acid: A, (95% ethanol) 214.5 (e 34,400), 240.5 (a 12,700), 275 (e 3,560), 294 (e 5260); 2,&dimethylphenoxphos phinic acid A, 218 (e 35,700), 246 ( E 19,300), 297 (e 5,000), 304 mp (e 5,6301.1 In a typical experiment, 9.9 g. of ptolyl ethef, 17.4 ml. of phosphorus trichloride, and 8.5 g. of anhydrous aluminum chloride were placed in a %necked flask equipped with a sealed stirrer and a reflux condenser protected with a drying tube. The mixture was stirred and refluxed for about 22 hr. On pouring the reaction mixture over 400 g. of cracked ice, an oil was obtained which solidified rapidly. The solid was removed by atration and washed thoroughly with water. After reprecipitation from 5% sodium hydroxide solution, the solid was recrystallized from 95% ethanol to give a 73% yield of pure 2,&dimethylphenoxphosphinic acid; m.p. >300". Anal. Calcd. for C1$IlaOaP:P, 11.90; neut. equiv., 260.2. Found: P, 11.76; neut. equiv., 259.2.
(6) L. D. Freedman and G . 0. Doak, J . Osg. Chem., 21, 238 (1956). (7) J. Reilly snd P. J. Drumm, J . Chem. Soe., 2814 (1927), have iound that Friedel-Crafts acylation of ptolyl ether yields Zsubstituted compounds.
Auto-Transformation of D-Fructose
Sir: The structural lability of fructose and the variability of its derivatives are well known. We have described a great number of compounds, which are
DEPAR~N OFTBIOLOGICAL CHEMISTRY THE HEBREWU N I V E ~ ~ I T Y A. H. SHAMGAR JEBUSALEM,ISRAEL J. LEIBOWITZ Received August 2, 1960 (1) A. 11. Shamgar and J. Leibowitz, Bull. Rea. Counc. Israel, ?A, 34 (1958). (2) M. Sarel-Imber and J. Leibowitz, J . Org. Chem., 24, 1897 11959). ~(3) A. H. Shamgar and J. Leibowita, J . Org. Chem.,25, 430 (1960). (4) E. J. MaDonald and B. K. G a s , A d N . Chem.,24, 422 (1952). (5) N. Albon and D. Gross, T k Analyst, 76,287 (1951). . - - - - I
