PREPARATION OF ARYLBORONIC ACIDS
July 20, 1938
3GOR
TABLE I1 PHYSICAL
Formula
Density. g./ml; a t 20
Ref. index, n20D
PROPERTIES O F
Mole refraction Calcd. Found
THE
ETHYLTIN CHLORIDESn
Melting point , OC.
Calcd. Ei p . , "C.a t 1 atm.
Vapor pressure eqn. cnnstantsb
A
B
Calcd. heat o f va]mr~zatiitn. cal 'mulr
Trouton's crinit , cdI 'mrilr/ deqree
CsHjSnCL 1.965* 1.5408* 40.22 40.61* -10 2554* 8.377* 192 11872* 25.1 IC2H&SnCL 84 2753* 8.385* 227 12581' 25.2* (CnHs)aSnC1 1.429 1.5055 50.54 50.15 15 2652* 8.416* 206 12110* 25,3* For previously reported values, see E. Rrause and A. \-, Grosse, "Die Chemie der Metall-Organischeii T-crbindungrti," Photolithoprint, Edwards Bros., Ann Arbor, Mich., 1943, pp. 336, 310. 5 Constants for the equation log I' = - n / T 8.
Table 111. To obtain the values reported in the
-AHan
15in
+ 292
+
where n is the number of carbon atoms in the TABLE I11 molecule. HEATS O F COMBCSTION AND HEATS O F FORMATION OF The heats of formation given in Table I11 were SEVERAL TIN TETRAALKYLS AT 300'K. A N D 1 XTM. calculated from the heats of combustion, taking the - AHcomb , Calcd AHform, heats of formation of H20(1),CO?(g) and Sn02(s) Compound kcal :mole - AHcomh kcal /mole to be 68.317, 94.052 and 138.8 kcal. per mole, 1934 i.3 920 +9 respectively. l6-lS i 9 0 4 i 10" Comparison of the heats of combustion observed (CH&3nC?Hi io34 i 5 1077 +53 in this work with those of previous investigators"' 1221 i 5 1234 $28 (CH,)?Sn(C?H.,)? reveal t h a t the results obtained in this investigation 1358 i 1 1391 +54 CH& (C?HS) 1 are consistently larger in magnitude. The principal source oi error in experiments of this type, where one of the combustion products is a solid, is (CzHd2Sn(C4Hg)2 2170 f 2 2176 -51 failure to obtain complete combustion. This would lead to low values for the observed heats of combustion. This fact, in addition to the checks (CsHdSn 4060 i 3 4060 +119 Data of Lippincott and T ~ b i n . "~ Data of Jones, et aZ.4 between individual determinations and the agreement, within the limits of experimental error, of table, the following corrections were applied : (1) the observed heats of combustion with a single emall weighings were reduced to weight in vacuo; pirical equation, supports the higher values ob(2) the bomb process was corrected to constant tained in this investigation. temperature (27') ; (3) the combustion reaction Acknowledgment.-The authors are grateful to was corrected from one a t constant volume to one Mr. Samuel Von Winbush and Miss Essie Shelton a t a constant pressure of one atmosphere. The who performed the analyses for carbon, hydrogen, latter two corrections were made according to the tin and chlorine. Financial help received from the method of Washburn.lS The data show the ex- Office of Naval Research is also gratefully acpected increase in the observed heats of combustion knowledged. with increased size of substituent alkyl radical. D. D. Wagman, J , E. Rilpatrick, W. U. Taylor, IC. S Pitrer With the exception of compounds containing and(l(i) F. D. Rossini, i b i d . , 3 4 , 143 (1945). methyl radicals, the heats of combustion may be (17) E. J. Prosen, R . S. Jessup and F. D. Rossini, ibirl., 33, 117 represented by the empirical equation (1944). (15) E W Washburn, J
R e s ~ n i r h Y'ntl Bziv Slondouds, 10, 525
(18) J. Moose and S. W. Parr, THIS J O C R X A I . , 46, 2030 (1921).
XASHVILLE 8, TENXESSEE
(1933)
[CONTRIBUTION FROM
THE
DEPARTMENT O F CHEMISTRY, I O W A STATE COLLEGE]
Some Studies on the Preparation of Arylboronic Acids1 BY HENRY GILMANAND LEONARD 0. MOO RE^ RECEIVED JANUARY 16, 1958 T h e yields of boronic acids from the reaction of various organic derivatives of the Group 11-B metals with tri-n-butyl borate are not as good as those obtained from the reaction of Grignard reagents with tri-n-butyl borate or from the reaction of organomercury compounds with boron trichloride. The course of the reaction of boron trichloride with organomercury compounds is dependent upon the groups substituted on the aromatic radical attached to the mercury.
An interest in boronic acids has developed recently because of the proposed use of boron-containing compounds in brain tumor therapy. The treatment involves the irradiation of carcinogenic
tissue which has absorbed preferentially a coinpound containing boron-10.3 Azo dyes which cnntain boron have been shown to be promising for this t r e a t n ~ e n t but , ~ the work of several has shown
(1) This work was supported in p a r t by the United States Atomic Energy Commission under Contract No. AT(11-1)-59. ( 2 ) Du P o n t Teaching Fellow, 195G-1957.
(3) Iiaturally occurring boron contains 18.83cI, of this isotope. See NBS Circular h-o. 499, p. 7. (4) H. Gilman, L. Santucci, D. R . Swayampati and R . 0 Iinnck. THISJ O U R N A L , 79, 28!)8 (1957).
t h a t any boron-containing compound may be useful.> 'The reaction of Grigiiartl reageiits with borate esters is a conventional procedure for the preparation of aroniatic boronic acidsfi.'; however, inany boronic acids cannot be prepared by this method because of side reactions. This study was undertaken to explore t h e possibility of using less reactive organometallic compounds than the Grignard reagents, which might allow the preparation of new boronic acids. Such boronic acids could conceivably be synthesized with many substituted groups which might aid the preferential absorption of the compound, or reduce undesirable secondary effects.% A series of experiments was performed to compare the reactions of the organic derivatives of zinc, cadmium and mercury with tri-n-butyl borate. This study indicated that the order of reactivity of these three metals was: Cd > Zn > Hg. The greater reactivity of cadmium was quite unexpected since previous studies on relative reactivities with active hydrogen had indicated t h a t zinc was iiiore reactive than c a d n i i ~ m . ~ No explanation is available for the apparent reversal of the order of reactivity for these two metals in these two different reactions. Since organornercury compounds did not react with tri-n-butyl borate, attempts were made to use a more reactive boron compound. The reaction of diarylniercury compounds with boron halides a t elevated temperatures and pressures is known.10-13 The high temperature and pressure reported by others earlier are not necessary, since the reaction of boron trichloride with diphenylmercury occurs very readily a t room temperature and atmospheric pressure. The reaction is not instantaneous since work-up of the reaction mixture immediately after completion of the addition of the boron trichloride gave a lower yield than did the usual procedure of allowing the mixture to stir for 30 minutes after the addition. A 100% excess of boron trichloride appears to be necessary for the best yields. The yield of benzeneboronic acid, after hydrolysis, is about the same as t h a t which can be obtained from the reaction of tri-n-butyl borate with phenyllithiumI4 or phenylmagnesium halide. (.i) Sre ref 4 for a bibliography on earlier work on the use of boron crimpounds i n brain tumor therapy. ( t i ) 11. 12. Lappert, Chenz. Reus.. 56, 959 (1950). (7) 11.S . Kharasch and 0. Reinmuth, "Grignard Reactionsof Nonmetallic Substances," Prentice-Hall, Inc , Xew York, N . Y . , 1934, p. 1:3:35. (8) 'The reduction of undesirable biological effects of organic compounds h y substituted groups is %vel1 established. T h e reduction of t h e carcinogenic activity of certain azo dyes by the presence of such groups a s bromo, chloro, hydroxy, nitro and trifluoromethyl is a n example. See J . P. Greenstein, "Biochemistry of Cancer," 2nd ed., Academic Press, Inc.. New York, S.Y . . 1954. I ! ) ) J . F.Nelson, Iow'1t Stale, Coil J . S c i . , 12, 145 (1937). f l 0 ) A . Michaelis and P . Becker, Be?., 13, 58 (1880). ( 1 1 1 A . Michaelis and P, Becker. ibid., 15, 180 (1882). ( 1 2 ) A . 1Iichaelis and 11,Behrens, ibid., 27, 244 (1891). (13) A. Rlichaelis, A w n . , 315, 19 (1901). (1.1) Yields of 70-75y0 of benzeneboronic acid have been obtained from phenyllithium a n d tri-n-butyl borate [H. Gilman and J. J. G o i d m a n , unpublished studies]; 5O-G0% from phenyl Grignard and 64, tri ii butyl borate [F. R. Bean and J. R. Johnson, THISSOURKAL, 441,; (1932); J. R . Johnson, private communication, Oct. 11, 19541; and 80:; from phenyl Grignard and trimethyl borate [P. A. Mc-
Before hydrolysis of the reaction mixture, the mercuric chloride which has been produced must be filtered froin the solution of the arylboron dichloride. Mercury salts react with arylboroiiic acids to yield boric acid and the diarylniercury C O I I 1 pound.15s1fiThis reactioii is so iicarly complete t h a t i t has been developed into an analytical 1)rocerlurcfor Groups attached to the aryl radical of the organomercury compound have a distinct effect upon thc reaction. The hydroxy and amino groups, which aid electrophilic substitution reactions, not only seem to aid the displacement of mercury by boron, b u t also the displacement of boron by a protoii. The reactivity in compounds containing these groups is so great t h a t no arylboronic acid could bc isolated under our experimental conditions. T h e three isomeric hydroxybenzeneboronic acids, ~vliich are not extremely unstable mheii pure, have becii prepared by other nieans.lS The carboxyl group has an effect opposite to t h a t of the hydroxy anti amiiio groups. The carboxyl group decreases t h r reactivity of the aryl radical so t h a t the iiicrcury is not displaced by boron even a t elevated teniper:itures.
Experimental Reaction of Tri-n-butyl Borate with Organic Compounds of Group 11-B Metals.-To an ether solutio11 of 0.1 equixralent of an organometallic reagent (0.1 inole of a r ~ l n i e ~ d l l i c halides and 0.05 mole for diarylmetallic compounds) under a n atmosphere of dry nitrogen in a three-iiecked, rountlbottomed flask, was added 0.25 mole of tri-n-butyl br)rate'* (15046 excess). The reaction mixture wac stirred for 6 hours and then was hydrolyzed by the addition of sufticient IO'/", sulfuric acid t o obtain a clear aqueous layer. Thc layers were separated and the water layer extracted with 200 ml. of ether in three portions. T h e combined ether solution was then extracted with 200 ml. of l(Jyopotxssiuin hydroxide in three portions, and the basic scilutioi~acidified with lo?;. julfuric acid to Obtain the boronic acid. TABLE
BETZENEBOROSIC . i C l D
FROM
1
CjRG.4SIC IIERKATIVE O F (>R:lUP
h r 11
i c s N i )2 ~ t 1 ( C6H3)2Z~1 iCtjHi) ZIIC1 (CsHJuCd (CsI1;)?Cda ( CgHj) pCd (CgHa)nCd I'CsHj) CdCl ( CsHa)nH g
GHJ b
BORATEAX11 11-B 1 1 E T A I . S
TRr-n-nv-rYl,
-1
'Temli., O C .
I~1eld.I ; ,
25 - 70 2 ,-) 35 23 33 -I' 25
2i 1,s 14 >; 1
:;*5 110
17
I) (I
Reacted for 18 hours with 0.5 mole of tri-z-butyl borate. Reacted for 144 hours in refluxing toluene.
Table I shows the results that were obtained from a series of reactions. The temperature is the temperature a t which t h e tri-n-butyl borate was added t o the organometallic rcCusker, private communication, I'eh. 2 3 , 11333; Ruigh, Ericksnn, Gunderloy and Sedlak, Wright Air Development Center Technical Report 5 5 - 2 0 , P a r t 11, p. 59 (1955) I . (15) A. D. Ainley and F. Challenger, J . Chena. Soc., ?li1(1930). (16) H. R. Snyder and C . Weaver, THIS J O U R N A L , 70, 232 (1948). (17) T. S . West, .LI?luiiwgia, 4 7 , S i (1953) IC.A . . 47, 4791 (1953)l. (18) H. Gilman, L. Santucci, D. K. Swayampati and R . 0 . Ranck, THISJOCRNAL, 79, 3077 (1967). (19) P a r t of t h e tri-wbutyl borate used in these studies was kindly provided by t h e Pacific Coast Borax Co.
July 20, 1958
ARYLBORONIC ANHYDRIDES AND AMINE COMPLEXES
agent and which was maintained until the mixture was hydrolyzed. The percentage yield is based upon the crude product. In all cases the purity of the crude acid seemed to be the same, since the melting points were nearly the same. Reaction of Boron Trichloride with Organomercury Compounds.-The organomercury compound (0.05 mole) was suspended in 500 ml. of chlorobenzene in a threenecked, round-bottomed flask equipped with a stirrer, a Dry Ice-acetone condenser attached to a mercury pressure relief valve, and an inlet tube which extended below the level of the solvent. The inlet tube was connected through a mercury trapz0to a tank of boron trichloride. Boron trichloride was bubbled into the mixture until the boron trichloride tank had decreased in weight the amount desired for the reaction. After stirring for the desired length of time, the misture was filtered to remove the mercuric chloride and the filtrate was hydrolyzed by the slow addition of ice. The hydrolysis appeared to be catalyzed by hydrogen chloride and so once started was autocatalytic. The rate of hydrolysis could be controlled only by the amount of water added. The acid which was formed was extracted with 250 ml. of 1074 potassium hydroxide in four portions. The aqueous solution \\\as washed with 100 ml. of ether and then acidified. The cream colored solid which separated was filtered and recrystallized from water. ( 2 0 ) This trap served t o remove any chlorine which was present in the boron trichloride as well a s t o allow t h e rate of addition t o be observed carefully.
[CONTRIBUTION FROM
THE S O Y E S
3611
The results of several reactions are summarized in Table I-
ll.
TABLE I1 REACTIONO F BORONTRICHLORIDE WITH ORGANOMERCURY Yield,
%*
70.5 55.0
52.0 57 , 0 70.0 0 . 0d 0.oe 0.Of 0 . of This represents the time after the addition of the boron trichloride was complete. The yield is based upon t h e recrystallized boronic acid. Used 0.2 mole of organomercury compound. There was separated 547, of phenol and 927, of mercuric chloride. Isolated nearly 1007, of mercuric chloride. Recovered the starting mercury compound quantitatively. 0 This reaction was run a t 120'.
'
AMES, IOWA
CHEMICAL LABORATORY, UNIVERSITY O F
ILLISOIS]
Aryl Boronic Acids. 11. Aryl Boronic Anhydrides and their Amine Complexes' BY H. R. SNYDER, MILTONS. K O N E C K AND Y ~ W. J. LENNARZ RECEIVEDFEBRUARY 1, 1958 The change of an a r j lboronic acid to its anhydride, which sometimes occurs when the acid is simply warmed in an anhydrous solvent, can be detected b y examination of the infrared absorption spectrum of the sample. As a result of the change, the hydroxyl absorption band of the acid disappears and a new absorption near 700 cm.-', characteristic of the anhydride, appears. The bromination of p-tolueneboronic acid in carbon tetrachloride is found to produce a mixture of p-bromomethylbenzeneboronic acid and its anhydride. The yield in the bromination is increased by prior conversion of p-tolueneboronic acid to its anhydride. The formation of complexes between arylboronic anhydrides and amines is studied further, and the use of the complexes in the purification of the boronic acids is suggested.
The syntheses of boronic acid analogs of azo dyes,3ajbs4diaminoacridine dyes5 and certain amino acids6 are of interest because the substances may be of use in the boron-disintegration therapy of brain tumors proposed by Kruger.' In this therapy, tumor cells are destroyed by the dissipation of the high energy of products resulting from the neutron-induced disintegration of the boron isotope 10 which has been localized within the neoplasm in the form of some suitable organic derivative. In the course of recent studies in this Laboratory devoted to the synthesis of these boronic acid analogs, the ease with which aryl boronic acids form the corresponding anhydrides, a phenomenon noted (1) P a r t of this work was supported by a grant [AT(11-1)-314] from the Atomic Energy Commission. (2) Texas Co. Fellow, 19%-1957. ( 3 ) (a) H. R. Snyder and C. Wea\.er, THISJ O U R N A L , TO, 232 (1918); (b) H. R . Snyder and S. L. Xeisel, i b i d . , T O , 77-1 (19-18). (4) H. Gilman, L. Santucci, D . Swayampati and R . Ranck, i b i d . , 79, 2598 (iRS7). i.?) R I . S. Konecky, Thesis, Doctor of Philosophy, University of Illinois, 1958. (6) H. R . Snyder, A . J. Reedy and W. J. Lennarz, THISJ O U R N A L , 80, 835 (1958). (71 P. G. Kruger, R a d i a t i o n Research, 3, 1 (1955).
by Gilman, et ~ l . for , ~ o-hydroxybenzeneboronic acid, has been observed. With the objective of developing a criterion for distinguishing a boronic acid and its anhydride from each other and from mixtures of both, the infrared spectra of several aryl boronic acids and the corresponding anhydrides were studied. As a result of this study a region of absorption, invariably strong and sharp, between G80 and '705 cm.-' has been empirically assigned to the aryl boronic anhydride structure. This absorption is absent from the spectrum of a pure aryl boronic acid, while the infrared absorption attributed to the hydroxyl group of a boronic acid is not detected in the spectrum of a pure anhydride. The infrared spectrum of a mixture of an acid and anhydride exhibits both of these absorption bands in diminished intensity. Table I shows the region of absorption in the infrared for the hydroxyl group of the boronic acid together with the position of the anhydride band for the corresponding anhydride. The absorption band which invariably occurs near 1330 cm.-' with both aryl boronic acids ( 8 ) H. Gilman, L. Santucci, D. Swayampati and R . Ranck, THIS JOURNAL, 79, 3077 (1957).
