3024
COMWJNICATIONS TO
droxide ion with an ester, i t would appear in the reaction of other nucleophiles with t h a t ester. Since the special effect of the o-hydroxyl group does not appear in the latter reactions, i t must be concluded that i t also does not operate in the former reaction and further that the former reaction must be interpreted as the reaction of water with the ionized form of the ester. If one interprets the pH-independent reaction in the alkaline region in terms of the reaction of water with the ionized form of the ester, the question may be raised as to the reason for the facile reaction of this species. The most straightforward explanation is that the phenoxide ion may act as an intramolecular general basic catalyst for the reaction of water with the ester group, as suggested above. The mechanism may be depicted as in eq. 8, or alternatively as the corresponding mechanism
0
-
I1 a
C-0 & (H
+ HOR
(8)
in which the internal base removes a proton from the addition compound of water and the ionized ester, as discussed by Jencks and Carriuolo for general basic cataly s i ~ . It~ appears, ~ however, that in this instance one
THE
Vol. 83
EDITOR
can rule out those mechanisms which are kinetically general basic catalysis, but which are mechanistically general acid-hydroxide ion reactions. 37 Previous results in the literature concerning catalysis by neighboring hydroxyl groups can all be interpreted according to the above m e ~ h a n i s m . ~ -Bruice ~ and FifeRconsidered eq. 8 as the mechanism for the effect of a neighboring hydroxyl group, but discarded this mechanism on the grounds of an argument involving deuterium oxide solvent isotope effects.38 The deuterium oxide isotope effects in our hands are compatible with both possible mechanisms of general acid-specific hydroxide-catalyzed hydrolysis and general basecatalyzed hydrolysis (see above). It is not mandatory, of course, t h a t all examples of neighboring hydroxyl group catalysis occur by the same mechanism. Some may operate oia eq. X involving general basic catalysis by the alkoxide ion while others may operate via general acidic catalysis by the un-ionized hydroxyl group. However, since the hydrolysis of salicylate esters, which when calculated as the hydroxide ion reaction of the un-ionized ester has the most significant neighboring hydroxyl group catalysis, appears to involve the fphenoxide ion and not the phenolic group, i t would appear t h a t the other examples of neighboring hydroxyl group participation should be re-examined in this light. ( 3 7 ) W . P. Jencks and J. Carriuolo, J . A m . Chem. Soc , 8 3 , 1713 ( I W i I ) . (38) T h e equation of Bruice and Fife6 for calculating the isotope effect does not agree with eq 5 , h u t a revised equation from Professor T C Bruice (personal communication) does agree with eq 3 . When their isotopic d a t a are calculated according t o eq 5 , the resulting isotope effect does not support general base catalysis Their isotopic d a t a are, however. in accord with nucleophilic catalysis b y hydroxide ion.
COMMUNICATIONS T O T H E EDITOR Electrophilic Substitution. Electronic Effects in S E ~ mechanisms intermediate between the limiting cases l a , l b , and IC. Reactions
Siv: There are many reactions in which a carbon-l.' bond is broken by the attack of an electrophilic group or atom where the usual result is retention of configuration a t the center of displacement. Examples are: base catalyzed H-D exchange in carbon acids,' electrophilic cleavage of organometal bonds, * oxidation of organoboron compounds with hydrogen peroxide, and Beckmann, Lossen, and hydroperoxide rearrangem e n t ~ . All ~ these reactions can be represented5 by ( I ) D. J C r a m , D. A . Scott, and Q', D . Nielsen, J A m Chem. S a c , 83, 3896 (19crl). ( 2 ) (a) S U'instein, T . G . Traylor, and C. S Garner, ibid., 77, 3741 (1955), (b) H . B. Charman, E. D . Hughes, and C. K . Ingold, J . Chem. S o c . , 2523 (1959); (c) F. R. Jensen and 1- H . Gale, J. A m . Chem. S o c . , 81, 1261 (1959). (3) H C. Brown, "Hydroboration," W. A . Benjamin, Inc., New York, N. Y ,1902, p . 07. (4) See, e . g , J. A . Berson and S Suzuki, J . A m Chem Soc., 81, 4088 (19.59) for references (5) T h e terms SEI refers t o dissociations into ions or ion pairs. Sc2 is conceived t o have a transition state in which two electrophiles are attached t o the carbon orbital without appreciable rehybridizatirin or development c,f charge on carbon. I n this respect it is analogous t o t h e S N mechanism. ~ T h e term S * E ~representing , u h a t is usually called aromatic substitution, is introduced t o specify an intermediate of the Pfeifer-Wizingerl type ( i e , rr-complex) denoted by * instead of to differentiate an intermediate from a transition state Thls terminology allows for S E I and S E aromatic ~ substitution, the latter demonstrated herein, and for the nonaromatic PfeiferN'izinger intermediates which are required t o explain some accelerated electrophilic reactions a t vinyl and cycloprop>-l groups 9 d ((5) E D Hughes and C . K Ingold, J C h e m Soc , 244 (1933) ( 7 ) P Pfeifer and R . 1Vizinger. A1277., 461, 132 (1928).
*
SE1
-
R1R2R&Y
R1R2R3C-
+
Y+zR1RzR3CE (la)
r
-.*
L
J
These formulations represent rearrangement to an electron-deficient E if an E-Y bond is present in the reactant and typical electrophilic substitutions otherwise. When a ligand connects E with I.' in the transi-
/ tion state
\
\'/ /
c
E/" '\\
//'
1
the mechanism is generally
Z called SEi. However, since this process differs so little from s E 2 we are including it in the designation S E % . The electronic effects can be predicted for such formulations. Electron supply by R will accelerate
COMMUNICATIONS TO THE EDITOR
Oct. 5 , 1963
sE2 and s*E2 and retard SEI reacti0ns.O Such effects have been amply demonstrated for mechanisms near S E I ' and S * E ~ .However, ~ there is wide disagreement concerning effects of structure on rates of reactions which are reported to be S E (or ~ SEi) . g \Ye have therefore sought a type of carbon-metal bond
TABLE I RATESOF REACTIONS O F BORONIC ACIDSWITH HYDROGEN PEROXIDE A K D RELATIVE RATESOF MIGRATION FROM BORON TO OXYGEX R
IR B ( O W 2I
~ could be that the S E reaction sufficiently polar studied over a full range of structure without contribution from S * Ewhile ~ still avoiding the SElmechanism. Such a system is shown in eq. 2 . R--B
/OK
+
OOH-
K
[RB(OH)200M]-
(1)
'OH
R
+
+
OH -
This reaction is not accompanied by protolysis and it occurs with retention of configuration in water. The oxidation is therefore not an SEI reaction. Kuivila and Xrmourloa have inferred from kinetic studies that the reaction of arylboronic acids with hydrogen peroxide is not S * E ~ Our . work confirms their proposal that no x-participation is involved. Reaction 2, whose rate k, reflects the effect of structure on this "sE2" process, is preceded by a rapidly established equilibrium." Therefore, K must be estimated in order to convert kobsd into relative k m . \Ye have measured the acidities of our subject boronic acids (eq. 3 ) and, with the reasonable assumption12 RB(OH)2
+ 2 H20
RB (OH)8-
+ H30+
(3)
that K a Ka, have obtained relative k m from the relation kmrel = kobserved 'Ka. The reactions of alkylboronic acids with hydrogen peroxide are cleanly second order over the observed HO range (- 1.5 t o 6).
The pH-rate profile for n-butylboronic acid is like that published for phenylboronic acid, lob exhibiting both specific acid and base catalysis and a rate minimum a t pH -3. Unlike arylboronic acids it has no term that is second order in alkylboronic acid, and the rate minimum is much deeper. Typical rate data for both acid ( k 2 ' ) and base ( k z ) catalyzed reactions are shown in Table I along with the acidities and relative values of k , which appear as k z / K , or Aiz'/Ka. The first striking result in Table I is that t-butylboronic acid is a stronger acid than n-butylboronic acid. This is the reverse of the order expected from electronic ( 8 ) E . S . Gould, "Mechanism and Structure in Organic Chemistry," H o l t - D r y d e n , S e w York, S Y.,1959, Chapter 11. (9) ( a ) C . S Marvel and H 0. Calvery, J . A m . Chem Soc., 45, 820 (1923); (bi S Winstein a n d T . G T r a y l o r , ibid , 77, 3747 (1955), I C ) M M Kreevoy and R . 1, Hansen, ibid , 83, F20 ( 1 9 6 1 ) , (d) R . E Dessy, G F Reynolds, and J Y K i m , ibid., 81, 2083 (1959); (e) G A . Russell a n d K . I. K a g p a l , Tetrahedron Lelters, 421 (1901). (10) ( a ) H G. Kuivila and A. G Armour, J . A m . Chem Soc., 7 9 , 5059 (1957), ( b ) H. G. Kuivila, ibid, 77, 4014 (19.5.5). (11) J 0 . Edwards, ibid , 76, 0154 (1953) (12) Some justification of this assumption is available J H . Polevy'a has used literature ratelob and acidity" d a t a for arylboronic acids t o plot Also, we log ( k o h s d / K e ) ZIS. u obtaining a n excellent line with p = - 2 . 0 3 have observed t h a t relative over-all rates of reaction of cyclohexylboronic acid and n-butylboronic acid are t h e same with hydrogen peroxide or t butylhydroperoxide (13) J H . Polevy, P h . D Thesis, University of New Hampshire, 1960, p. 63 (14) D L. Yabroff. G E. K Branch, and J J . Almquist, J A m , Chem Soc.. 66, 2940 (1933)
3025
CHan-Busec-Bu-
t-Bu-
ki, kz', I./mole sec., 1 /mole sec., Ho = 5.23a Ho = 1.48'
0 000127 00480 0233 ,0718 .ll ,016'
0.00024 00724
kz'/Ka
lO"K,
ki/K, (relative)
2 52d 1 83 2 5 4 32 3 05 138d 32 73
(1) 52 185 330 680 2 3 4 2 24
(1) 42
(relative)
,094 229 ... LBicycloheptyl 2.4 ,0319' CeH5,0068 ... Vinyl ,0875 , . . CeHsCHza A buffer 0.158 M in sodium acetate and 0.042 111 in acetic acid. 3.77 M perchloric acid. Data of Kuivila. Other kinetics were determined by- his method". See ref. 10b. Determined by the method of Yabroff, Branch, and Almquist. See ref. 14. e All data a t 25.0".
considerations and therefore implies that less steric compression is experienced in tetrahedral boron [RB(OH),-] than in trigonal boron [RB(OH),J. Consequently, the observed rate sequence, bridgehead > 3" > 2" > l o ,is not due to relief of steric strain as was suggested for a similar rate order in the BaeyerVilliger reaction. Both SEI and S*E2 mechanisms are excluded by the extreme sluggishness of the cleavage of aryl-, vinyl-, and cyclopropyl-boron bonds. Thus, even in this very polar carbon-metal bond the R group does not migrate as a "free" R- as has been suggesttd. lfi The third suggestion for this rate sequence in the Baeyer-Villiger reaction, i e . , development of cai bonium ion character in the migrating group, lSb is considered very unlikely in the present reaction because the group departs from a negatively charged boron atom. The transition states for the acid and base catalyzed reactions can therefore be formulated for all structure;
in R. This is the first instance of electrophilic substitution or migration of R to an electron-deficient center in which no change in mechanism occurs when R is varied. This reaction is therefore considered to be a good model for an 932 reaction (including SEI) and the order observed, t-Bu > sec-Bu > n-Bu > vinyl phenyl > Me, is suggested to be the electronic sequence for the S E process. ~ l7 Recently, reports of the opposite sequence, 1" > 2" > 3", for electrophilic substitution have a ~ p e a r e d . ~ ~ - ~ In these cases, additional factors such as changes in the leaving interfering side reactions, and stringent steric requirementsge make the extrapolation of their results to other electrophilic substitutions tenuous. We, therefore, conclude that purely S E (or ~ SEI) reactions will show the rate sequence 3" > 2" > 1' > CsH5,vinyl, cyclopropyl 2 Me and that the extent of involvement of the s * E 2 mechanism can be measured by the observed position of phenyl, vinyl, and cyclopropyl groups in this series.
-
(1.5) ( a ) hf. F. Hawthorne, W D. Emmons, and K . S.M c C a l l u m , d i d 80, 0393 (1958); (bl N' D. Emmons and G. B. I,ucas, ibid , 77, 2287 (195.5) (10) S. L. Friess and S . F a r n h a m , ibid , 73, 3318 f l Y A O ) \Ve shall demon (17) T h e structural effects in this reaction are small s t r a t e a similar order with large effects on C-B bond cleavage rates i n t h e following communication.
3026
Vol. 85
COMMUNICATIONS TO THE EDITOR
Acknowledgment.-We are grateful to Professor Henry G. Kuivila for enlightening discussions and to the National Science Foundation for financial support (Grant GP-242).
cleavage of boronic acids might attain the five-niembered transition state I11 rather than the three-membered 11. Structural effects on transition states such as I (and we imply 11) are not strongly dependent on the nature of the leaving group. Therefore, from the (18) International Christian University, M i t a k a , Tokyo, Japan HIROSHIM I X A T O ~ ~ effects of the structure of R on the rate of chromic acid CHEMISTRY DEPARTMENT GSIVERSITYOF CALIFORNIA, SAS DIECO JVDITHC WARE cleavage of boronic acids, we might distinguish between LA JOLLA,CALIFORNIA T G TRAYLOR I1 and 111. This is possible only if one mechanism is RECEIVED AUGCST 16, 1963 operating. However, several species of chromic acid in aqueous solution could bring about oxidation. Among the species H,Cr04+, H2Cr04,HCrO4-, Cr04'-, Electrophilic Substitution. Chromic Acid Cleavage of H2Cr207,HCr207-,and Crz072-,only the first two have Carbon-Boron Bonds been definitely associated with oxidation in dilute Sir: solutions and these two usually are in competition. I n the previous communication' we indicated the This competition of mechanisms greatly complicates effect of structure in R on the stability of the threestudies of structural effects.? center transition state I. Lye wish to report an oxidation by HCrO4- in a reaction which promises important application both in synthesis and in mechanism studies. This finding also points the way to studies of reactions of the remaining CrV1 species. The reaction of chromic acid with t-butylboronic acid
H20
Cr03
I11 The general behavior of chromic acid oxidation of caused us to believe that chromic acid
+ ~-BUB(OH)~
_ -d[Crv'l =
dt
-2
-
r" 0
9
[CrVr]X k
=
0.054(HCrOa-)
+ 0.36(HzCr04) +
0.32(HzCr04)ho ( 3 )
-3 -
using the recordedgvalues of K H 2 C r O 4 = 1.21 andKHcro4= 3.2 X (Dimeric species are precluded by the low CrV1concentrations.) Therefore the oxidants in this reaction are H C r 0 4 - and H 3 C r 0 4 + ; Cr0.1'and H2Cr04 are relatively ineffectivelo (ie., H2Cr04
I I
I -4I
1
k2[RB(OH)2][Cr"]
The pH-rate profile is shown in Fig. l.5,6 The shape of this curve has three important implications. First, the oxidation of boronic acids a t pH 3-7 where alcohols are relatively stable to CrVr makes possible the synthesis of alcohols using chromic acid. Thus, oxidation with chromic acid a t pH ca. 5 will produce alcohols7 and a t higher acidities, e.g., 2 N acid, will produce ketoness Further, this reagent makes possible selective alkylborane cleavage because its reaction rate is much more sensitive to structure than that of hydrogen peroxide.' Thus, the k2 values for variE t , 6.6 X 10V4; hie, 2.4 X ous R aret-Bu, 7.5 X l.imole sec. in 0.114 ;M perchloric acid a t 30.0'. Secondly, the pH-rate profile is unusually informative about the mechanism of the reaction. This curve is accurately described by the equation
0-
-
+ Cr3+ (1)
is first order in each reagent over a wide range of pH ( - 1 to 9) and has a very small salt effect (5% increase a t 1.0 -IfNaC104) on rate.
I
-I
t-BuOH f HaBOa
0
9
I
8
I
I
7
6
I
5
I
4
3
I
I
2
1
I
0
.
HO
Fig. 1.-Plot of log kz a t 30.0" us. H0 (or p H ) for the rate expression d[Cr\"] /dt = kz[t-BuB(OH)2] [CrV1]. Circles are experimental points. The solid line represents eq. 3 and the dashed lines indicate unit slope. The Ho values for the perchloric acid solutions were taken from Long and Paul.6 (1) H M i n a t o , J . C W a r e , and T . G . T r a y l o r , J . A m . Chem. Soc , 86, 3024 (1963). (2) (a) F H . Westheimer, Chem. Rev., 46, 419 (1949); (b) F. Holloway, hi. Cohen, and F. H . Westheimer, J A m . Chem. Soc., 73, 05 (19511, (c) G . T E . G r a h a m and F H . Westheimer, ibid , 80, 3030 (1958); (d) R Brownell, A . Leo, Y W . C h a n g , and F H. Westheimer, i b i d . , 82, 406 (1900); (e) J Rofek and F. H . Westheimer. i b i d , 84, 2241 (1962), ( f ) Y. W . Chang a n d F. H . Westheimer, i b i d . . 82, 1401 (1900); (9) H. G. Kuivila and W J . Becker, i b i d , 1 4 , ,5329 (1952); (h) J . Rofek and J . Krupicka, Chem. Listy, 62, 1736 (1958); (i) F. H . Westheimer and A. S o v i c k , J . Chem. Phys , 11, 506 (1943).
(3) H G Kuivila, J . A m . Chem. Soc., 77, 4014 (19.50). (4) H. K w a r t and P . S. Francis, i b i d . , 77, 4907 (1955) (5) All rate d a t a were obtained b y t h e spectrophotometric method of Westheimer and S o v i c k 2 ' Borate buffer was used a t p H 9 , phosphate a t p H G 3, and acetate a t p H 4-5. At higher acidities perchloric acid was used. At p H D .5, t h e reaction is pseudo first order (KB(OH)2 in excess) for about one-half life and then acclerates rather sharply. Similar acceleration was noted with other buffers (except burate) above p H 8 . (6) M. A . Paul and F . A. Long, C h e m Rev., 67, 1 (19,57) (7) T h e chromic acid cleavage appears t o proceed with t h e s a m e stereo^ chemistry a n d yield as t h e hydrogen peroxide cleavage (unpublished results) (8) H. C . Brown aud C P . G a r g , J . A m . Chem S o c . , 83,2952 (19(il) (9) J . T Long and E L . King, i b i d . , 7 6 , 6180 (1953). (10) This preference of boronic acids for t h e anions is seen in other cases T h e "B n . m r. of ethylboronic acid" indicates i t t o be entirely trigonallZ in water (i.e,, not complexed with water) and tetrahedral in 0.8 .V sodium hydroxide. (11) W e are indebted to H Landesman, J . D i t t e r , and T. Burns, of t h e National Engineering Science C o , Pasadena, California, who very generously determined t h e IlB chemical shifts. (12) T . P Onak, H Landesman, R . E . Williams, and I. Shapiro, J . P h y s . Ciiem., 63, 1633 (1959).
