Stabilization of ketone and aldehyde enols by formation of hydrogen bonds to phosphazene enolates and their aldol products

Article abstract of DOI:10.1021/ja804221x

Solution properties of enolates generated using the phosphazene (Schwesinger) base P4-^tBu were investigated by NMR spectroscopy. With a full equivalent of base the benzyl ketones 1a and 1b, the acetophenone 2, the arylacetaldehyde 1c, and the methyl arylacetate 1d formed the expected "naked" (P4H+) enolates 3 and 7. However, at a half-equivalent of base the ketones 1a and 1b as well as the aldehyde 1c formed solutions of stable hydrogen-bonded dimeric (enol-enolate) structures (4). The acetophenone 2, on the other hand, forms only traces of the H-bonded dimer 8 during deprotonation of 2. The thermodynamic product was the isomeric self-aldol condensation product 12. The mechanism of this condensation was elucidated by low temperature rapid-injection (RI) NMR spectroscopy. Solutions of 8 stable enough for NMR characterization could be transiently generated by semiprotonation of the enolate 7 with HCl·OEt₂ at -130 °C using RINMR. The ester enolate 1d gave no trace of 4d even on a time scale as short as a few seconds at -130 °C either during the semideprotonation of 1d, or during semiprotonation of the enolate 3d. Long-lived solutions of the enols derived from 1a, 1b, 1c, and 2 (but not 1d) could be produced by full protonation of the phosphazene enolates with HCl·OEt₂ at low temperature. Copyright

Full text of DOI:10.1021/ja804221x

Published on Web 07/01/2008
Stabilization of Ketone and Aldehyde Enols by Formation of Hydrogen Bonds
to Phosphazene Enolates and Their Aldol Products
Kristopher J. Kolonko and Hans J. Reich*
Department of Chemistry, UniVersity of Wisconsin, 1101 UniVersity AVenue, Madison, Wisconsin 53706
Received June 4, 2008; E-mail: reich@chem.wisc.edu
The strong nonmetallic phosphazene or “Schwesinger” bases1
have found a variety of uses, from enolate and peptide alkylations2a
to catalytic aldol reactions,2b desilylations2c and even aromatic2d
and alkynyl2e deprotonations. The anionic species produced by
deprotonation with Schwesinger bases have been termed “naked”
due to the noncoordinating phosphazene counterion. The few
reported studies of the solution structure of these species have
focused on ester enolates whose behavior is complicated by
decomposition reactions.3 Here we report NMR solution studies
of phosphazene enolates formed by deprotonation of p,p′-difluo-
rodibenzyl ketone (1a), 1-phenyl-2-p-fluorophenylethanone (1b),
p-fluorophenylacetaldehyde (1c), methyl p-fluorophenylacetate (1d)
Figure 1. Titration of ketone 1a with P4 at -125 °C in 3:2 THF/ether.
and p-fluoroacetophenone (2) with tBu-P4 (P4, pKBH+ (MeCN) )
42.61).
A full equivalent of P4 base quantitatively deprotonated 1a to
generate a single species. The much larger 13C chemical shift
difference of the vinyl carbons compared to the lithium enolate
dimer4,5 suggest the formation of a “naked” P4H+ enolate 3a.
Figure 2. Mixing experiment with P4 and ketones 1a and 1b at -125 °C
in 3:2 THF/ether.
of 4a to its components, and the barrier of 11.4 kcal/mol represents
Addition of smaller increments of P4, however, unexpectedly
an upper limit on the strength of the hydrogen bond.
formed a different intermediate enolate species (Figure 1). At 0.25
Injection of ethereal HCl or HBF4 into a solution of 3a using
equiv of P4 at -125 °C no 3a was present, and half of the ketone
RINMR9 gave solutions of the enol 5a, which were stable for hours
1a was converted to this new species, which becomes the major
at -118 °C, allowing full spectroscopic characterization.4 If less
component (g95%) at 0.5 equiv of base. Further addition of P4
converted the intermediate to 3a. The vinyl carbon 13C shifts of
than 1 equiv of acid was added, mixtures of 1a, 4a, and 5a were
formed. The enol, like the enolate, was in slow dynamic exchange
the new species were similar to those of the dimeric lithium
enolate,4,5 as were the two signals in the 19F NMR spectrum. On
the basis of this data, as well as the observation of a downfield
singlet at δ 18.4 in the 1H NMR spectra, which is characteristic of
low-barrier hydrogen bonds (LBHB),6 we have identified it as the
enol-enolate H-bonded dimer 4a.7,8 The H/D isotope shift was
+0.35 (D upfield), consistent with a strong unsymmetrical LBHB.6
Evidently the strength of this hydrogen bond is sufficient to
overcome the unfavorable keto-enol equilibrium.
All three components in these equilibria (eq 1) are in slow
exchange on the NMR time scale below -15 °C. With 0.5 equiv
of P4 at -35 °C 1a, 3a, and 4a are present in a 1:2.4:7 ratio, giving
an association Kaeq of 650 M-1. Free energies of activation for the
exchange of 3a and 4a at several ratios of 3a to 4a were measured
by 19F DNMR at -25 °C, and were found to be independent of
the concentration of 3a. Thus, this is a unimolecular dissociation
with the H-bonded dimer 4a below -120 °C.
We investigated other enolizable carbonyl compounds to see if
H-bonded dimers were commonly formed during deprotonations
with P4 base. Ketone 1b also gave a preponderance of the H-bonded
dimer 4b at substoichiometric amounts of P4, although the
equilibrium was less favorable than for 1a (Kbeq ) 13 M-1 at -37
°C).
Further evidence for the dimeric nature of the H-bonded
structures was obtained by titrating P4 into a 1:1 mixture of ketones
1a and 1b. This produced the two homo H-dimers 4a and 4b at
<0.5 equiv of P4, as well as a new mixed H-dimer 6 (Figure 2).
It is surprising that 6 forms in larger than statistical amounts, even
1
though the upfield H shift of the H-bonded proton (from δ 18.4
for 4a to 16.9 for 6) suggests a weaker H-bond interaction.10
Increasing the amount of P4 to 0.75 equiv produced a solution of
9
9668 J. AM. CHEM. SOC. 2008, 130, 9668–9669
10.1021/ja804221x CCC: $40.75
2008 American Chemical Society

C O M M U N I C A T I O N S
structures (4) when excess carbonyl compound is present, and these
are thermodynamically more stable than the isomeric aldolates. This
“self-enolization” clearly has relevance to the role that LBHBs play
in enzyme catalyzed enolizations.16 Carbonyl compounds with less
favorable keto-enol equilibria do not form such dimers (1d), or
form them only transiently (2). In the latter case we were able to
follow the self-aldol condensation and elucidate its mechanism.
Figure 3. Self-aldol reaction of 2 at -125 °C in 3:2 THF/ether. In the
first step, 2 was injected into a solution of P4.
the H-dimer 4a and the enolate 3b, with P4 preferentially
deprotonating the more acidic ketone 1b. At a full equivalent of
base, the solution consisted of the enolates 3a and 3b.
Like 1a, the aldehyde 1c is converted quantitatively to the enolate
3c with 1 equiv of P4. At lower equivalents the H-bonded dimer
4c (mixture of E/Z isomers) is the major species formed (60% at
0.5 equiv of P4), but there are several minor components that have
not been identified (possible aldol products). Because of their high
propensity for aldol reactions, aldehyde metal enolates have been
rarely studied spectroscopically.11
Acknowledgment. The authors wish to thank Dr. Bob Shanks
and Dr. Charlie Fry for NMR assistance. We thank NSF for
financial support (Grant CHE-0074657) and funding for instru-
mentation (Grants NSF CHE-9709065, CHE-9304546).
Supporting Information Available: Experimental details and NMR
spectra. This material is available free of charge via the Internet at
http://pubs.acs.org.
References
The interaction of ester 1d with P4 has been previously studied.3
We find that P4 stoichiometrically forms the enolate 3d at -120
°C, with no indication of any H-bonded dimer 4d or Claisen
products. Apparently the energetic cost of enolization of 1d to 5d
outweighs the stabilizing effect of H-bond formation. Even the
RINMR injection of HCl into the enolate at -135 °C failed to form
either 4d or the ester enol 5d, only 1d was detected.
The effect of P4 on p-fluoroacetophenone (2) was more complex.
Unlike the behavior of 1a, the addition of 0.5 equiv of P4 to
solutions of 2 at -78 °C did not produce the H-dimer 8, but instead
the ketone self-aldol condensation product 12. RINMR experiments
using 19F and 1H NMR spectroscopy in which 2 was injected into
(1) Schwesinger, R.; Hasenfratz, C.; Schlemper, H.; Walz, L.; Peters, E. M.;
Peters, K.; von Schnering, H. G. Angew. Chem., Int. Ed. Engl. 1993, 32,
1361–1363.
(2) (a) Seebach, D.; Pietzonka, T. Chem. Ber. 1991, 124, 1837–1843. Seebach,
D.; Bezencon, O.; Jaun, B.; Pietzonka, T.; Matthews, J. L.; Ku¨hnle, F. N.;
Schweizer, W. B. HelV. Chim. Acta 1996, 79, 588–608. (b) Solladie´-Cavallo,
A.; Crescenzi, B. Synlett 2000, 327–330. Kobayashi, K.; Ueno, M.; Kondo,
Y. Chem. Commun. 2006, 3128–3130. (c) Ueno, M.; Hori, C.; Suzawa,
K.; Ebisawa, M.; Kondo, Y. Eur. J. Org. Chem. 2005, 1965–1968. Suzawa,
K.; Ueno, M.; Wheatley, A. E.; Kondo, Y. Chem. Commun. 2006, 4850–
4852. (d) Imahori, T.; Kondo, Y. J. Am. Chem. Soc. 2003, 125, 8082–
8083. (e) Tanaka, Y.; Arakawa, M.; Yamaguchi, Y.; Hori, C.; Ueno, M.;
Tanaka, T.; Imahori, T.; Kondo, Y. Chem. Asian J. 2006, 1, 581–585.
(3) (a) Solladie´-Cavallo, A.; Liptaj, T.; Schmitt, M.; Solgadi, A. Tetrahedron
Lett. 2002, 43, 415–418. (b) Fruchart, J.-S.; Lippens, G.; Kuhn, C.; Gras-
Masse, H.; Melnyk, O. J. Org. Chem. 2002, 67, 526–532. Fruchart, J.-S.;
Gras-Masse, H.; Melnyk, O.; Fox, D. T.; Bergman, R. G. Tetrahedron Lett.
2003, 44, 2243.
a solution of P4 at -118 °C initially produced the enolate 7 (t1/2
,
(4) The CR (Cꢀ) 13C NMR shifts are (Z isomer): δ 173.2 (87.3) for 3a, 165.1
(94.1) for 4a, 168.5 (96.1) for the lithium enolate dimer, 152.5 (109.4) for
the enol silyl ether of 1a, and 155.4 (106.7) for the enol methyl ether of
1a. The shifts of 5a were sensitive to the counterion, presumably due to
hydrogen-bonding interactions: δ 158.2 (98.7) for 5a from HCl injection,
155.0 (100.9) for HBF4 injection. The enol proton shifts were also sensitive
to counterion, appearing at δ 11.5 (HCl) and 9.5 (HBF4).
(5) The lithium enolate of dibenzyl ketone has been previously studied. (a)
Gareyev, R.; Ciula, J. C.; Streitwieser, A. J. Org. Chem. 1996, 61, 4589–
4593. Its behavior with P4 is nearly identical to that of 1a.
(6) Hibbert, F.; Emsley, J. AdV. Phys. Org. Chem. 1990, 26, 255–379.
(7) H-bonded dimers of enolic ꢀ-dicarbonyl compounds7a and phenols7b have
been previously reported. (a) Aladzheva, I. M.; Petrovskii, P. V.; Mast-
ryukova, T. A.; Kabachnik, M. I. Zh. Obshch. Khim. 1980, 50, 1442–1445;
English translation, 1980, 50, 1161-1164. Arnett, E. D.; Harrelson, J. A.
Gazz. Chim. Ital. 1987, 117, 237–243. (b) Bordwell, F. G.; McCallum,
R. J.; Olmstead, W. N. J. Org. Chem. 1984, 49, 1424–1427. Shan, S.; Loh,
S.; Herschlag, D. Science 1996, 272, 97–101.
ca. 1 min), followed by a self-aldol process with a t1/2 ) 30 min
(Figure 3). During the first 10 min a small amount of H-bonded
dimer 8 (3.5% compared to 7) was detected.12 The first-formed
product was a new species which we have identified as a 1:1
aldolate enolate hydrogen bonded complex 11 (δOH 12.4).13 The
rate of formation of 11 was first order in 2 and 7 and showed a
deuterium isotope effect of 1.0 when R-trideuterated 2 was used.
Intermediate 11 was then converted over several hours to the final
stable product of this sequence, the hydrogen-bonded aldolate
enolate 12 (δOH 11.7). This conversion showed a H/D kinetic isotope
effect of 2.4.
From these and other observations we conclude that the H-dimer
8 is in an unfavorable equilibrium with the enolate 7 and ketone 2.
Formation of the aldolate 10 is the rate determining step, but none
was detected since it rapidly deprotonates another ketone producing
the observed 11. Finally, 11 is converted by proton transfer and
loss of ketone in an irreversible step to 12.
In comparison, the reaction of 7 with p-fluorobenzaldehyde was
too fast to measure at -127 °C by RINMR (kobs g 2 s-1).14 Based
on temperature extrapolation of the rate of reaction of 2 with 7,
the aldehyde is at least 50000 times as reactive as the ketone.
Interestingly, the first-formed aldolate 13 was again not detected.
The products were acetal 1:2 adducts (14),15 in which the aldolate
is stabilized by addition to a second equivalent of aldehyde.
We have identified and characterized several phosphazenium
enolates generated with P4. Some of these enolates (3a, 3b, and
3c) form unique kinetically stable hydrogen-bonded dimeric
(8) Pan, Y.; McAllister, M. A. J. Org. Chem. 1997, 62, 8171–8176.
(9) Jones, A. C.; Sanders, A. W.; Bevan, M. J.; Reich, H. J. J. Am. Chem.
Soc. 2007, 29, 3492–3493.
(10) Kumar, G. A.; McAllister, M. A. J. Org. Chem. 1998, 63, 6968–6972.
(11) Wen, J. Q; Grutzner, J. B. J. Org. Chem. 1986, 51, 4220–4224.
(12) Transient solutions of 8 (t1/2 ) ca. 5 s), and 9 (t1/2 ) ca. 30 min) could be
formed by injection of HCl-etherate into solutions of 7 in 3:2 THF/ether
at-118 °C. More stable solution of 8 (t1/2 ) ca. 15 min), and 9 (t1/2 > 2 h
were formed in Me2O at-138 °C, which allowed characterization by 19F,
1H, and 13C NMR spectroscopy.
(13) Compound 11 could be fully characterized by 13C, 1H, 31P, and 19F NMR
spectroscopy. Saturation transfer measurements showed the enolate portion
of 11 to be in exchange with the enolate 7 in solution.
(14) For a previous RINMR study of a crossed aldol reaction see: Palmer, C. A.;
Ogle, C. A.; Arnett, E. M. J. Am. Chem. Soc. 1992, 114, 5619–5625.
(15) AsimilaradducthasbeenobservedfortheTASenolateofcyclohexanone:Nakamura,
E.; Shimizu, M.; Kuwajima, I.; Sakata, J.; Yokoyama, K.; Noyori, R. J.
Org. Chem. 1983, 48, 932–945.
(16) Cleland, W. W.; Frey, P. A.; Gerlt, J. A. J. Biol. Chem. 1998, 273,
25529–25532.
JA804221X
9
J. AM. CHEM. SOC. VOL. 130, NO. 30, 2008 9669