CYCLIC BENZENEBORONATE ESTERS
May 20, 1958
(although the N(CH3)3+ group generally does fit). It is likely, therefore, that charged substituents involve additional complications. The nature of the inductive effects of charged substituents will be considered further in later Gublications. VALUES
O F p i FOR Acid
1. Cs&NHa+ 2. C~HISH
3. 4. 5.
CaHaOH
TABLE I11 ACID IONIZATIONS IN WATER,25' Acid
PI
2.93 2.30
8.
1.05
CMHIAS(OH)X 0-
CsHaB(0H)z CaHaCOOH 0 CaH6P(OH)z 0-
0.76
0'95
t 10. CaHaSe-OH 11. C I H ~ C H ~ N H J + 12. 13.
CIHICHZCOOH CsHsCHxCHzCOOH
0.91 0.72 +0.49
+0.21
Table I11 lists PI values obtained for a number of acid ionization equilibria in aqueous solution. The values given for acids 6-13 are actually the Hammett p values given by JaffC, since too few data are available in most of these reaction series to apply equation 7. The pr and p values for these acids are expected to be quite similar, although when comparison is possible the relationship discussed in the following paragraph is frequently followed more closely by PI values. The important point illustrated in Table I11 is that to a rough approximation PI depends only upon the position with respect to the benzene ring a t which ionization occurs in the side-chain. Acid ionizations (1-3) in which the formal charge on the first atom of the side-chain is decreased by one unit have PI values of 2.3 to 2.9. If the unit decrease in formal charge acts through an additional atom, PI values of 0.72 to 1.05 are obtained (acids 5-11). Acids 12 and 13 provide further examples of the applicability of the equation: PI (2.8 0.5)1-i, where i = the number of saturated atoms between
*
[CONTRIRUTION FROM
the benzene ring and the atom a t which the unit decrease in formal charge takes place. This relationship is followed in a manner roughly independent of the charge type of the acid or the kind of atom involved, and conforms to the Branch and Calvin scheme for treating inductive effects in acid ionizaation e q ~ i l i b r i a . ~ ~ The PI value of phenylboric acid (4) is of special interest. Although the proton leaves from the
OH group, the conjugate OH base is probably adequately represented by the second atom of the -B(
THE
'
) -, so that 3I:< the $rst atom loses nearly a unit formal charge on ionization of the acid. Accordingly, the PI value is nearer to that for acids 1-3 than 5-11. Having used equation 7 to demonstrate the general applicability of the CJI inductive scale, and to determine the inductive reaction constant, 191, we are now in a position to evaluate the total resonance effect by applying equation 1 in the form
major resonance form, (CsH6-B -I
0 0
1 6.
PI
0T
2.83
2443
total effect of reson.
= log ( k p / k o ) - U I ~for I @-substituent
total effect of reson. =log (km/ko) -
u ~ p for :
(9)
m-substituent
Equation 9 can be used to examine the general applicability and limitations of the UR resonance scale, and to establish the nature of the dependence of resonance effects on reaction type. The next paper in this series will deal with such an examination. Acknowledgment.-The authors wish to express their appreciation to Professor H . H. JaffC for the benefit of valuable discussions. (24) Reference Bb, pp. 193-200. NOTE A D D E D I N PROOF.--The ionization of pyridinium ions in H*O, B o with , p = +5.69, (H. H. JaE6 and G. 0. Doak, THISJOURNAL, 7 7 , 4441 (1950)) also conforms to this rough scheme. The formal charge of the nitrogen atom of the ring is decreased b y one in the ionization, corresponding in th.e above formulation t o i = -1.
UNIVERSITY PARK,PENNA.
DEPARTMENT O F CHEMISTRY, UNIVERSITY OI' UTAHJ
Cyclic B enzeneboronate Esters BY JAMES M. SUGIHARA AND CARLOS M. BOWMAN RECEIVEDXOVEMBER 18, 1957 Reaction products of berizeneboronic acid with cis- and tran~-cyclopentane-l,2-diol, cis- and truns-cyclohexane-l,2-diol, pyrogallol, methyl b-D-glucopyranoside, 2,3-butanediol, 1,3-butanediol, 1,4-butanediol, 3,4-di-0-benzoyl-~-mannitola n d galactitol are described in addition to the previously prepared tribenzeneboronates of D-mannitol and D-glUCitOl. Ring structures containing 5-, 6- or 7-members are possible. Polyols tend to form completely substituted benzeneboronates. An explanation, based upon ease of hydrolysis and method of preparation, is offered for the latter observation.
Cyclic boronate esters of the compounds D-mannitol, D-glUCitOl, pinacol, pentaerythritol, diethyl D-tartrate, cis-indane-1,2-diol and catechol are reported.' Reaction was effected in aqueous media in all cases. Benzeneboronate esters of pentoses and two G-deoxyhexoses were prepared2 by fusing (1) H. G . Kuivila, A. H. Keough and E. J. Soboczenski, J . O r g . Chem., 19, 780 (1954). ( 2 ) M . L. WolIrom and J . Solms, i b r d . , 21, 815 (1956).
the two reagents. However, knowledge concerning steric requirements for ring formation and preferred ring size is largely lacking. Accordingly, this investigation was initiated in the hope of obtaining some information in this direction. Cyclic diols were selected as model compounds to study the possibility of formation of cyclic boronate esters, since the relative positions of the hydroxyl groups are essentially fixed. Refluxing of
2444
JAMES
M.
SUGIHARA AND CARLOS
benzeneboronic acid with cis-cyclopentane-1,2-diol and with cis-cyclohexane-l,2-diol in anhydrous acetone followed by distillation (procedure A) gave crystalline benzeneboronates (I, 11) of low melting points. Since the esters of the tram-isomers were
?d.~ O \ V l v f A N
VOl.
so
dure6 of benzoylation, hydrolysis and acetylation. The only isolable product was methyl p-D-glucoindicating essentially comside tetraacetate ('ilyO), plete blocking of all the hydroxyl groups. Application of the same procedure to an equimolar mixture of the reactants gave a reaction mixture, separable b y chromatography, containing methyl p-D-glUC0side tetrabenzoate (3770) and the tetraacetate (41yo),These observations indicate that nearly one-half of the glucoside reacted completely a t the site of all of the hydroxyl groups with benzeneboronic acid, and a substantial amount of the re1,n = 1 111, n = 1 IV, n = 2 11, 12 = 2 mainder of the reagent did not react. Since the crystalline compounds with higher melting points, chromatographic procedure has been demonstrated6 they were purified by recrystallization (procedure to separate any diacetate dibenzoates, i t is likely B). On the basis of elemental analyses and molecu- that the total amount of monobenzeneboronate lar weight determinations, the ester of frnns- esters in the reaction mixture is small. dlkanediols were allowed to react with benzenecyclopentane-1,2-diol has been assigned structure I11 and the ester of the cyclohexane homolog, struc- boronic acid to deterniine any preference for ring size. 2,3-Butanediolj 1,3-butanediol and 1,4ture IV. The difference in behavior of the cyclopentane- butanediol formed distillable products in yields of 81, diol isomers is interpretable by the knowledge that SS and 71Yo6,respectively, and assigned structures the carbon-oxygen bonds in the cis isomer are es- V, lrI and VII, respectively. Under the same sentially coplanar while those in the trans isomer are R, distinctly not. The proposed seven-membered V,R1 = R P = CHJ; n = 0 1 V I , R1 = H ; Rz = CH3; n = 1 HC-0 ring structure is possible, as demonstrated by a V I I , R1 = RP = H; n = 2 I \ molecular model. (CH2)n B-C~IIS The variance in reactivity of the two cyclohexI / anediols has been reviewed. The energy requireHC-0 ments for cis functions to move into a coplanar I Rz position are small, since the movement involved resembles the partial inversion of a chair form into experimental conditions, 1,S-peritanediol gave no a boat form. The forcing of two trans groups into distillable or crystallizable product. Thus foressentially coplanar positions reduces the distance mation of five-, six- and seven-membered rings was between axial atoms and increases strain in the ring. indicated but not an eight-membered ring. From Assuming that benzeneboronate ester formation percentage yields the five- and six-membered ring requires a coplanar intermediate, the difference in compounds appear to form somewhat more readily type of products formed from cis- and trans-cyclo- than the seven-membered ring. hexane-1,2-diol is understandable. The tribenzeneboronate ester of D-mannitol has Another polyhydroxy1 compound, which ap- been previously described.' However, the size of peared to be of interest, was pyrogallol, since all of the boronate ring system has not been established. its oxygen atoms are coplanar with the ring. .I In order to obtain information which might provide pyrogallol monobenzeneboronate readily was ob- a more precise structural assignment, the hexitol tained. Benzoylation of this compound yielded was treated with benzeneboronic acid in various the nionobenzoate. Hydrolysis of the benzene- molar ratios. Reaction was effected in anhydrous boronate ester group was then effected by the ad- acetone. Petroleuni ether extractions of the solids dition of D-mannitol to an aqueous solution of the derived provided high yields (9~17~ with three moles compound and neutralizing the complex formed. of benzeneboronic acid) of the crystalline product The pyrogallol benzoate obtained is apparently the (procedure C). lT7ien this procedure was followed same as the compound previously r e p ~ r t e dand , ~ is using two moles of benzeneboronic acid per mole probably the 1-benzoate, since the compound is of D-mannitol, the crystalline triboronate was isoreadily oxidizable. The attempted methylation of lated in an amount accounting for 92y0 of the benthe monobenzoate with diazomethane failed. Xce- zeneboronic acid. \Vhen the amorphous reaction tylation of pyrogallol benzeneboronate gave pyro- mixture was benzoylated, followed by hydrolygallol triacetate. Apparently the ester linkage is sis of boronate ester groups, D-mannitol hexanot stable toward acetylation conditions. benzoate was the only isolable crystalline coni. Since methyl @-D-glucopyranoside had been pound, accounting for 25Yoof the original hexitol. If found to be a useful substrate in studying the reac- the triboronate ester formation should consume tion of boric acid with poly0ls,5 the compound was all CJf the benzeneboronic acid, one-third of the manapplied in this investigation. The reaction of two nitol should remain umeacted and therefore yield moles of benzeneboronic acid per mole of the gluco- the hexabenzoate. side yielded an amorphous product, whose characIn the above procedure C, solubilities of reacter was studied by the previously described proce- tants and product in acetone might tend to favor (3) D. H. R. Barton and R. C. Cookston, Quart. Rev., 1 0 , 81 (1926). the formation of the triboronate ester, since D(4) A. Einhorn and F. Hollandt, A n n . , 301, 95 (1898). mannitol has low solubility in this solvent and ben( 5 ) J . M. Sugihara and J. C. Petersen, THISJ O U R N A L , 78, 17GO zeneboronic acid is highly soluble. ,.2 modification (1956).
May 20, 1958
CYCLIC BENZENEBORONATE ESTERS
2445
98-99', truns-l,2-cyclopentanediol,*m.p. 50°, and cis1,2-cyclohexanediol,@b,p. 105-110° (15 mm.), were prepared by methods described in the literature. 3,4-Di-O-benzoyl-~-mannitol.-Asolution containing 20 g. of 1,2 :5,6-di-O-isopropylidene-~-mannitol~O in 100 ml. of anhydrous pyridine was cooled t o 0" and 12.5 ml. of benzoyl chloride was added while agitating. The resulting *$uspension was allowed t o stand a t room temperature for 15 hr. and then poured over ice and water. The mixture was vigorously stirred for 0.5 hr. and then extracted with 300 ml. of ether. The ether extract x-as washed twice each with 50ml. portions of 2 N hydrochloric and 2 N sodium hydroxide, and once with 75 ml. of w a t q . The ether solution was dried over anhydrous sodium sulfate. Evaporation of solvent under reduced pressure left a sirup (37.7 g.), which failed to crystallize. A solution of 8.0 g. of the sirup, 20 ml. of acetone, 0.5 ml. of 12 N hydrochloric acid and 2 ml. of water was refluxed for 5 hr. Barium carbonate (5 9.) was added to neutralize the acid. The solids were removed by filtration. The filtrate was evaporated on a steam-bath under reduced pressure, leaving a mhite amorphous solid, which was crystallized from absolute ethanol. Recrystallization from ethanol gave 2.8 g. (44%) of 3,4-di-O-benzoylD-mannitol, m.p. 183-155", [aIz8D $.1.12' (c4.02, pyridine). Anal. Calcd. for C20H2208: C, 61.53; H, 5.68. Found: C,61.41; H, 5.49. Benzeneboronic Esters of Alcohols, Phenols and Amines. -Four general procedures are described for the preparation of compounds listed in Table I. Procedure A.-A solution of 0.043 mole of a diol (aminoalcohol, aminophenol or diamine), 0.013 mole of benzeneboronic acid and 20 ml. of anhydrous acetone was refluxed for 4 hr. The acetone was removed under reduced pressure on a steam-bath. The remaining clear viscous liquid was transferred t o a Claisen flask and distilled under reduced pressure. In any system containing water, the point of equilibProcedure B .--The same reaction mixture, as described rium lies considerably to the left. However, in procedure A, was refluxed for 4 hr. Upon cooling the water insolubility of the triboronate ester of D- acetone solution, crystallization occurred. Recrystallizafrom acetone yielded purified samples. mannitol permits its formation even in an aqueous tion Procedure C.-A solution of 0.0055 mole of a poly01 (or medium.' I n procedure D, little reaction might phenol), 0.011 (0.0055 or 0.0165) mole of benzeneboronic be expected because of the water present. Re- acid and 20 ml. of anhydrous acetone was refluxed for 4 hr. in moval of acetone would bring about ester forma- an anhydrous system. Removal of the acetone under reduced pressure over a steam-bath left a white amorphous tion and its precipitation as a result of its water solid. Extraction of the boronate ester \vas effected by reinsolubility. The mono- and disubstituted deriva- fluxing this residue with three 7.5-1111. portions of petroleum tives of D-mannitol would be expected to have ether (b.p. 110-125') for 25 min. each time. LVhen the greater water solubility and would not, therefore, combined petroleum ether extracts were cooled, crystallization occurred. Purification was effected by recrystallizabe isolable. tion from petroleum ether. An additional compound, D-mannitol 3,4-diProcedure D.-A solution of 0.0055 mole of a poly01 in benzoate, previously undescribed, was allowed to 20 ml. of acetone and 10 ml. of water was heated to the rereact with benzeneboronic acid. A crystalline di- flux temperature over a steam-bath. A solution containing 0.0055 (or 0,011) mole of benzeneboronic acid, 20 ml. of boronate ester was obtained, and assigned the acetone and 10 ml. of water was added slowly to the reaction structure D-mannitol 3,4-dibenzoate 1,2:j,G-diben- mixture over a period of 0.5 hr. After this addition was zeneboronate. completed, the solution was refluxed for 4 hr. longer. ReReaction products of 2-aminoethanol and o- moval of acetone under reduced pressure over a steam-bath amorphous solid, which was treated in the same aminophenol with benzeneboronic acid were ob- left a white as described in procedure C. tained but were found to be unstable. Bromina- manner Tribenzeneboronates of hexitols n ere obtained by either tion and acylation of o-aminophenol benzeneboro- procedure C or D, when molar ratios of hexitols to benzeneboronic acid of one t o one, one to two, and one to three were nate failed to yield expected products. applied; the following data mere obtained: (hexitol, yield, One (C.M.B.) of us gratefully acknowledges sup- melting point) D-mannitol 75-94%, 135" (reported' 66%, port from a University Research Committee 133-135'); n-glucitol, 81-95%, 19400(reported' 6692, 157Fellowship. 189'); and galactitol, 90%, 162-163 Anal. Calcd. for C24H230&: C , 65.53; H, 5.27; B, Experimental6 7.38. Found: C,65.00; H, 5.27; B, 7.29. Benzeneboronic acid,' a . p . 215-217', trans-1,2-cycloBenzoylation of Pyrogallol Benzeneboronate .-A r,olution hexanedio1,g m.p. 102-103 , cis-l,2-cyclohexanediol,~m.p. of 2 g. of pyrogallol benzeneboronate and 10 ml. of anhydrous pyridine was cooled t o 5" and a solution of 4.0 g. of (6) All melting points are corrected. C and H analyses were made benzoyl chloride and 10 ml. of anhydrous pyridine was by Drs. G. Weiler and F. F. Straus, Oxford, England. Boron was deadded slowly while agitating. The resulting suspension was termined by hydrolysis of esters and titration of the released benzeneboronic acid (G. E. K. Branch, D. L. Yabroff and B. Bettman, THIS allowed t o stand for 8 hr. a t room temperature. Absolute ethanol (30 ml.) was added in portions, and the resulting JOURNAL, 66, 937 (1934)). Molecular weights were determined cryosolution was allowed t o stand for an additional hour. All scopically in benzene. (7) F.R. Bean and J. R. Johnson, THISJOURNAL, 64, 4415 (1932). the solvents were evaporated by passing a jet of air over the solution to leave a white, crystalline residue. Recrystalli( 8 ) A. Roebuck and H. Adkins, "Organic Syntheses," Coll. Vol.
of this procedure was applied to maintain an excess of D-mannitol in the reaction mixture (procedure D). An acetone-water solution of benzeneboronic acid was added to an aqueous acetone solution of Dmannitol. Petroleum ether extraction of the reaction residue yielded only the triboronate ester. Application of procedures C and D with D-glucitol and galactitol gave the tribenzeneboronate esters only in high yields. The ease of hydrolytic cleavage of benzeneboronate esters of several sugars is reported.2 A study of the ease of hydrolysis of D-mannitol tribenzeneboronate appeared desirable, since the compound is preparable in an aqueous medium.l The hydrolysis of 0.5 g. of the ester in 50 ml. of water was incomplete, as shown by the incomplete recovery of benzeneboronic acid. A lOy0 hydrochloric acid solution effects a greater degree of hydrolysis. Optical rotation data indicated that the ester is stable in anhydrous solvents but readily hydrolyzed in moist acetone or pyridine. These observations obtained using D-mannitol may be rationalized by assuming that a readilyreversible equilibrium reaction is involved.
.
111. John Wiley and Sons, Inc., New York, N. Y., 1955,p. 217. (9) N. A. Milas and S. Sussman, THIS J O U R N A L , 69, 2345 (1937).
(10) E. Baer and H. 0. L. Fischer, J . B i d . Chem., 128, 468 (1939).
2446
JAMES
TABLE I BENZENEBORONATES OF ALCOHOLS, Iieageut
Vol.
M. SUGIHARA AND CARLOS AI. B O ~ W A N
B.p. 'C. Mm.
M . p . , OC.
Formula
PHENOLS AKD A M I N E S Analyses, 70 Calculated C H B C
Found
H
B
so
ProI-ield, cedure %
cis-1,2-Cyclohexanediol 95 0 . 3 42-43 CIPHISO~B 71.32 7 . 4 8 5.35 71.84 7 . 2 8 5.32 A 97 trens-1,2-Cyclohexanediol 134-136 Ci8HZ003BZ4 70.65 6 . 5 9 7.07 70.82 6.64 6 . 9 1 B 60 cis-1,2-Cyclopeiitanediol 80 82 I 15-17 CllH1302B 70.26 6 . 9 7 5 . 7 5 7 0 . 0 3 7 . 0 3 5 . 6 4 X 95 trens-1,2-C~~clopentaxic~~iol 127-128 Cl,Hp,OaB$ 69.69 6.19 7 . 3 9 70.14 6 . 3 8 7 . 2 4 B 74 Pyrogallol 166 C12H903BC 67.98 4.28 5.10 6 8 . 0 4 4 . 3 8 4 . 9 2 C 90 2,3-Butanediol 75-77 1 ClOHl3O2B 68.23 7.44 6 . 1 5 G8.44 7 . 5 1 6.05 A SI 1,3-Butanediol 85-86 1 CloHltOzB 68.23 7.44 6 . 1 5 68.69 7 . 5 2 5 . 9 9 A 88 1,4-Butanediol 90-95 1 CloHlaOzB 68.23 7 . 4 4 6 . 1 5 6 8 . 5 5 7.60 6 . 0 3 A 71 1 A 0 1,s-Butanediol >300 3,4-Di-O-benzoyl-~-niannitol 149-150 C22H280sB~d 68.38 5 . 0 2 3 . 8 5 67.99 5 . 0 8 3 . 8 2 C e 83 A 21 (Decomposes) 80-85 2' 2-Aminoethanol .4 81 99-101 (Decomposes) o- Aminophenol 140-145 2' o-Phenylenediamine (Decomposes upon distillation) aMolecular weight: calcd. 306; found 310. Ir Molecular weight: calcd. 292; found 297. Gives a green-colored soluCrystallized and recrystallized from ethanol. f I)i+ [ a I z s ~+2.35' ( c 2.52, pyridine). tion in alcoholic ferric chloride, tillation in nitrogen atmosphere. zation from absolute ethanol gave 1.9 g. (65%) of pyrogallol hydrolysis and acetylation. Five grams of the glucositlc benzeneboronate benzoate, m.p. 169-170". yielded 6.7 g. (71%) of methyl 8-D-glucoside tetraacetnte, ilnal. Calcd. for C19H13041B: C, 72.18; €1, 4.14; B, m.p. 103-105", alone or admixed with a n authentic sample. T h e same series of reactions using equimolar amounts of 3.42. Found: C, 72.43; H , 4 . 1 8 ; B,3.36. Pyrogallol Benzoate.-Pyrogallol benzeneboronate hen- methyl fl:D-glucopyranoside (0.5 9.) and benzeneboronic acid zoate (4.0 g . ) was dissolved in 25% ethanol. D-Mannitol gave a SlruP. Chromatographic separation5 gave 0.37 g. (4.0 g.) was added t o this solution, and 2.0 g. of sodium bi- (41%) of methyl 6-D-glucopyranoside tetraacetate, m.p. carbonate was introduced in portions to the resulting inixture 103-105", alone or admixed with a n authentic sample, and to neutralize the mannitol-benzeneboronate complex. The 0.59 g. (37% )oaf methyl p-D-glucopyranoside tetrabenzoate, par- m.p. 160-162 , alone or admixed with a n authentic sample. extracted three times with resulting solution D-Mannitol Benzeneboronate.-The amorphous solid obtions of &her. ~h~ combined ether solution lvas dried Over sulfate and then evaporated to a sirup, tained by the reaction of two moles of benzeneboronic acid anhydrous which crystallized upon adding a small volume of chloroform, Per mole of D-mannitol(1.0 g.) was benzoylated in the usual manner. Chromatography of the resulting sirup gave 0.19 giving 2.1 g. (72%) of pyrogallol benzoate, m.p. 138-139". Two recrystallizations from chloroform raised the melting g. (25%) of D-mannitol hexabenzoate, m.p. 149-150", alone or admixed with an authentic sample. point to 141°, reported4 140'. T h e attempted methylation of pyrogallol benzoate with Hydrolysis of D-Mannitol Tribenzeneboronate.-,4 susdiazomethane yielded only the starting phenol. Pyrogallol pension of 0.5 g. of D-mannitol tribenzeneboronate and 50 benzoate reacted readily with Tollens reagent. ml. of water x a s heated on a steam-bath for 2 hr., and then Acetylation of p y r o g ~ l o lBenzeneboronate.--A solution cooled and extracted twice with 25-1111. portions of ether. of anhydrous T h e combined ether solution was dried over anhydrous snof 3.15 g. of p y r o g a ~ l benzeneboronate, o~ 30 pyridine and 15 of acetic anhydride allowed to stand dium sulfate and evaporated. T h e residue was recrl-stalovernight. Absolute ethanol (75 ml.) u,aS added to the k e d twice from toluene t o give 0.20 g. (48%) of benzeneallolved to stand boronic acid, m.p. 215-217", alone or admixed with an ausolution solution, and the for a n additional hour. Evaporation of the solvents by thentic A mixture of 0.5 g. of D-mannitol tribenzeneboronate, 30 passing a jet of air Over the solution left a sirup. T h e latter ,vas dissolved in a minimum volume of absolute ethanol. ml. Of 10% hydrochloric acid and 30 1111. of ether v a s agiCrystallization from ethanol yielded 3.47 g. (93%) of pyro- tated, and the ether Phase separated. The aqueous Phase was re-extracted with 15 ml. of ether. T h e ether solution gallol triacetate, m.p. 164-165" (reported" 165'). ~ ~ calcd. ~ for l c . ~ ~ c, H 57.14; ~ ~H, ~4.79.~ pound: : yielded 0.34 g. (82%) of benzeneboronic acid, 111.p.2 1 . 5 217O. C , 57.12; H , 4.65. Methyl p-D-~lucopyranoside ~ ~ ~ ~ ~ ~ Rotations ~ (25", b n-line, ~ c~ 1.O) ~of D-mannitol ~ ~ tribcnzenet ~ tTvo moles of benzeneboronic acid per mole of methyl p-Dboronate in various solutions were measured and thc followg~ucopyranoside was treated as given in procedure C, an ing d a t a obtained: (solvent, rotation) anhydrous acetone, +58'; moist acetone, 0 ' ; amorphous solid \vas obtained. This \+.asprocessed by tile 4-50? anhydrous pyridine, general technique applied to borate esters6 of benzoylation, moist pyridine, 0'; absolute ethanol and acetone (1 : 1, l)y vol.), f 4 8 O . (11) Knoll and Co., German Patent 308,240, Feh., 1888; Chrm. Zenlv.. 711, 270 (1900).
SAT.TLAKECITY 12, UTAII
,
-
~
~
