Lithium and cesium ion-pair acidities of dibenzyl ketone. Aggregation of lithium and cesium ion pairs of the enolate ion and dianion

Article abstract of DOI:10.1021/jo960013o

Spectral study of the cesium and lithium enolates of dibenzyl ketone (DBK) showed that both salts exist as contact ion pairs in THF solutions. The spectral data for the dicesium salt of DBK indicate that it exists as triple ions in which both cations are in contact with the dianion. The dilithium salt of DBK forms triple ions of two types in THF: in one, both cations are in contact with the DBK dianion, in the other, one of the lithium cations is solvent-separated. Evidence for dimerization of the ion pairs was obtained for both lithium and cesium enolates of DBK from UV-vis spectral (blue shift of the absorbance band at higher concentrations) and acidity (the decrease of pK at higher concentrations) studies. The dimerization constant for the cesium enolate of DBK obtained from the acidity data (3.5 × 10³ M^-1) is considered to be more accurate than that from the spectral analysis (1.7 × 10³ M^-1). The lithium enolate is much less dimerized than its cesium counterpart with a dimerization constant from acidity data of 4.2 × 10² M^-1. The first and second cesium pK values of DBK are 18.07 and 33.70, respectively, compared to the first lithium pK of 11.62.

Full text of DOI:10.1021/jo960013o

J . Org. Chem. 1996, 61, 4589-4593
4589
Lith iu m a n d Cesiu m Ion -P a ir Acid ities of Diben zyl Keton e.
Aggr ega tion of Lith iu m a n d Cesiu m Ion P a ir s of th e En ola te Ion
a n d Dia n ion 1,†
Roustam Gareyev, J ames C. Ciula, and Andrew Streitwieser2,*
Department of Chemistry, University of California, Berkeley, California 94720-1460
Received J anuary 2, 1996X
Spectral study of the cesium and lithium enolates of dibenzyl ketone (DBK) showed that both salts
exist as contact ion pairs in THF solutions. The spectral data for the dicesium salt of DBK indicate
that it exists as triple ions in which both cations are in contact with the dianion. The dilithium
salt of DBK forms triple ions of two types in THF: in one, both cations are in contact with the
DBK dianion, in the other, one of the lithium cations is solvent-separated. Evidence for dimerization
of the ion pairs was obtained for both lithium and cesium enolates of DBK from UV-vis spectral
(blue shift of the absorbance band at higher concentrations) and acidity (the decrease of pK at
higher concentrations) studies. The dimerization constant for the cesium enolate of DBK obtained
from the acidity data (3.5 × 103 M-1) is considered to be more accurate than that from the spectral
analysis (1.7 × 103 M-1). The lithium enolate is much less dimerized than its cesium counterpart
with a dimerization constant from acidity data of 4.2 × 102 M-1. The first and second cesium pK
values of DBK are 18.07 and 33.70, respectively, compared to the first lithium pK of 11.62.
In tr od u ction
ketone and the corresponding dianion salts were recorded
in 10-5 - 5 × 10-4 M THF solutions at 25 °C. The λmax
and molar extinction coefficients for each salt are listed
in Table 1. Values of λmax for CsDBK were found to
undergo a progressive blue shift as the concentration of
the salt was increased as illustrated in Figure 1. As we
had shown earlier9 such behavior reflects the formation
of ion pair aggregates having different spectra. With the
increase in concentration the aggregation equilibrium in
eq 1 shifts in favor of dimers or higher clusters that
absorb at shorter wavelength.
Heathcock3 has reviewed the use of metal enolates in
synthetic organic chemistry. Dianion salts of carbonyl
compounds are also finding increasing use in the syn-
thetic laboratory,4,5 but less is known of their properties6
compared to enolates. The reactivities of both enolate
ions and dianions are undoubtedly closely related to their
basicities and aggregation states. We would anticipate
that knowledge of these properties would be useful in the
practice of organic synthesis. We reported previously7
that the cesium enolate of dibenzyl ketone (DBK) exists
in dilute THF solutions as a mixture of monomeric and
dimeric contact ion pairs. The dimerization constant was
found from the dependence of the relative ion-pair acidity
of DBK on the cesium enolate concentration to be 1870
M-1. The ion pair pK of the monomeric cesium salt was
found by extrapolation to be 17.95. This value compares
well with the value of 18.7 reported by Bordwell in
DMSO.8 We present here an extended study of the
cesium ion pair acidity of DBK covering a wider range
of enolate concentrations, as well as its second cesium
pK, the lithium acidity, and a spectral study of the cesium
and lithium enolates (CsDBK and LiDBK, respectively)
and the corresponding salts of the dianions (Cs2DBK and
Li2DBK).
Kn
\
n M+A- y
z (M+A-)n
(1)
For the cesium enolate λmax changes gradually from
343.5 nm to 338 nm as the concentration is varied from
5 × 10-5 M to 5 × 10-4 M. We subjected the series of
spectra recorded in this concentration range to singular
value decomposition (SVD).9 This mathematical treat-
ment can indicate the number and type of species in
equilibrium and can provide the spectra of the individual
components. The results obtained from the SVD analysis
show the presence of two major species and are consistent
with a monomer-dimer equilibrium (n ) 2 in eq 1) with
dimerization constant K2 ) (1.7 ( 0.2) × 103 M-1. The
spectra of the monomer and dimer obtained by deconvo-
lution are shown in Figure 2, and their spectral charac-
teristics are given in Table 1. The extinction coefficient
of CsDBK is independent of the concentration at 339.5
Resu lts a n d Discu ssion
Sp ectr a l Da ta . Ion P a ir in g a n d Aggr ega tion . UV-
vis spectra of the lithium and cesium enolates of dibenzyl
1
nm (isosbestic point) with ꢀ(monomer) ) /2ꢀ(dimer) )
† Dedicated to Clayton H. Heathcock on the occasion of his 60th
birthday.
22700 ( 300 M-1 cm-1. Below 5 × 10-5 M no noticeable
shift was observed (λmax ) 343.5) and indicates that at
these concentrations the cesium enolate of DBK exists
mostly as monomeric ion pairs. These ion pairs are
undoubtedly contact (CIP) by analogy with the cesium
salts of many different organic anions as shown earlier
by an array of physicochemical methods.10
X Abstract published in Advance ACS Abstracts, J une 1, 1996.
(1) Carbon Acidity. 93.
(2) e-mail: astreit@garnet.berkeley.edu.
(3) Heathcock, C. H. Comprehensive Carbanion Chemistry; Elsevi-
er: New York, 1981; Vol. II.
(4) Bates, R. B.; Taylor, S. R. J . Org. Chem. 1994, 59, 245-246.
(5) Bartoli, G.; Bosco, M.; Dalpozzo, R.; Denino, A.; Palmieri, G.
Tetrahedron 1994, 50, 9831-9836.
(6) Thompson, C. M.; Green, D. L. C. Tetrahedron 1991, 47, 4223-
4285.
(9) Krom, J . A.; Petty, J . T.; Streitwieser, A. J . Am. Chem. Soc. 1993,
115, 8024-8030.
(10) Streitwieser, A.; Ciula, J . C.; Krom, J . A.; Thiele, G. J . Org.
Chem. 1991, 56, 1074-1076.
(7) Ciula, J . C.; Streitwieser, A. J . Org. Chem. 1992, 57, 431-432.
Correction, Ibid. 6686.
(8) Bordwell, F. G. Acc. Chem. Res. 1988, 21, 456-463.
S0022-3263(96)00013-8 CCC: $12.00 © 1996 American Chemical Society

4590 J . Org. Chem., Vol. 61, No. 14, 1996
Gareyev et al.
F igu r e 3. Spectra of dilithium (A) and dicesium (B) salts of
dibenzyl ketone in THF at 25 °C.
F igu r e 1. Dependence of λmax of the cesium enolate of dibenzyl
ketone on concentration in THF at 25 °C.
are generally those of nonconjugated compounds. It is
possible that π-coordination could compete with O-
coordination of DBK enolates, but the above results
suggest O-coordination for both LiDBK and CsDBK. The
similarity of the spectra suggests that both metals
coordinate similarly. If lithium were π-coordinated some
tendency toward forming the solvent-separated ion pair
(SSIP) would be expected; such an equilibrium between
CIP and SSIP was found for the di-salt, Li2DBK as
discussed below. In addition, π-coordination is expected
to show a smaller tendency toward aggregation than
O-coordination; that is, the dimer most probably has the
type of structure shown in 1. Finally, the similarity of
the spectra of monomer and dimer suggests that they
have similar coordinations.
F igu r e 2. Spectra of dibenzyl ketone cesium enolate ion pairs
(A) and their dimers (B) in THF at 25 °C.
Ta ble 1. Sp ectr a l Da ta for th e Cesiu m a n d Lith iu m
En ola tes a n d Dicesiu m a n d Dilith iu m Sa lts of Diben zyl
Keton e (DBK) in THF a t 25 ˚C
Successive addition of aliquots of cumylcesium to a
solution of CsDBK in THF leads to the formation of the
dicesium salt of DBK. The peak in the region of 340 nm
corresponding to CsDBK decreases as the absorption
band of Cs2DBK at 498 nm grows. The spectrum of Cs2-
DBK (B in Figure 3) was found to be independent of
concentration, common salt, and temperature effects and
indicates that this salt exists primarily as the contact-
contact triple ion; that is, both cesium cations are in
contact with the dianion. Moreover, its spectrum does
not show a dependence on the amount of CsDBK present
in solution. Spectra of pure Cs2DBK and those in the
presence of up to a 70-fold excess of CsDBK are identical.
Acidity data (see below) indicate that Cs2DBK most likely
exists in the presence of CsDBK as a complex, Cs2DBK‚
CsDBK; thus, CsDBK appears to have a negligible effect
on the spectral properties of its neighbor. The THF
solutions of Cs2DBK could only be studied at low con-
centrations because of its high extinction coefficient, and
at these low concentrations there is no indication that
the compound forms dimers; the dimer in this case, of
course, could also have the same spectrum as the
monomer.
species
CsDBKa
CsDBK(monomer)
CsDBK(dimer)
LiDBK
λ
max, nm
ꢀ, M-1 cm-1
338-343.5b
343.5
22700c
23600
46000
20400d
92000
43000
332
319-322b
498
Cs2DBK
Li2DBK
492.5
a
b
Mixture of monomer and dimer. Depending on the concen-
tration. c At the isosbestic point, 339.5 nm. At the isosbestic
d
point, 317 nm.
A blue shift of λmax from 322 nm to 319 nm was also
observed for the lithium enolate of DBK at concentrations
of 2 × 10-4 to 5 × 10-4 M. However, application of the
SVD technique was unsuccessful in this case probably
because the amount of dimer present was too small. The
second singular value had a rather small coefficient and
was hardly distinguishable from noise. The occurrence
of a blue shift at higher concentrations indicates, how-
ever, that some aggregation does occur. An isosbestic
point was found at 317 nm with ꢀ ) 20400 ( 300 M-1
cm-1
.
The lower λmax values for LiDBK compared to CsDBK
shows that the ion pairs of both compounds are CIP. This
result is expected if both salts have the metal cation
coordinated directly to oxygen rather than to the conju-
gated π-system. Such coordination is common in X-ray
crystal structures of metal enolates,11 but the structures
An X-ray crystal structure of Li2DBK shows two
equivalent lithiums π-coordinated above and below the
(11) Setzer, W. N.; Schleyer, P. v. R. Adv. Organomet. Chem. 1985,
24, 353-451.

Lithium and Cesium Ion-Pair Acidities of Dibenzyl Ketone
J . Org. Chem., Vol. 61, No. 14, 1996 4591
establish the individual spectra of the CIP and SSIP of
Li2DBK because the usual approach17 of modeling these
two extremes taking spectra in different solvents or with
different counterions or upon addition of complexing
agents was not applicable. The state of Li2DBK in THF
solutions can be qualitatively described as a mixture of
contact-contact and contact-solvent separated triple
ions. An example of the formation of such dual-type
triple ions with lithium cations was shown previously for
the dilithium salt of 9,9′-bifluorenyl.18 It is also signifi-
cant that the spectrum of the dicesium salt is in between
that of the CIP-SSIP and CIP-CIP dilithium salt.
Ion P a ir Acid ity. Ion pair acidities are defined by
the transmetalation equilibrium of eq 2, in which HA is
the acid whose pK is to be measured and HIn is a suitable
indicator; the equilibrium constants are converted to pK
differences.
F igu r e 4. Spectra of the dilithium salt of dibenzyl ketone in
THF at different temperatures.
HA + M+In- y
\
Kz HIn + M+A-
(2)
dianion plane.12 A comparable structure seems reason-
able for Cs2DBK as symbolized in 2. Ab initio calcula-
tions on the dilithium13 and dicesium14 salts of acetone
give similar structures. In the complex, 3, coordination
of the exposed oxygen to the metal cation of the enolate
seems reasonable. The repulsion of the third cation to
the first two approximately compensates for the attrac-
tion of the π-cations to the additional oxide anion. The
stabilization is then approximately the same as in the
enolate dimer itself. Comparison with (diphenylallyl)-
-log K ) pKHA - pKHIn
The pK differences are converted to absolute pKs by
assigning a reference indicator, fluorene, its pK value of
22.9 in DMSO.8 In this way we have established ion pair
pKs for lithium19 and cesium10 ion pairs in tetrahydro-
furan. All pK values are statistically corrected to reflect
the acidity per hydrogen.
The indicators used for the cesium ion-pair acidity
measurements of DBK were 9-phenylfluorene (pK 18.15)
and 9-biphenylylfluorene (pK 17.72).10 Experimental pK
values at different concentrations of CsDBK are pre-
sented in Table S2. The decrease of the observed pK with
increasing formal concentration of corresponding salt was
shown previously for ketones7,20 and other compounds9,21
to result from aggregation of ion pairs to dimers or higher
aggregates. The aggregation shifts the equilibrium 2 to
the right, increasing the apparent K. A plot of the
observed pK of the cesium enolate vs the logarithm of
its total concentration is shown in Figure 5. In such plots
the slope at any point is related to the mean degree of
aggregation at that point.20 For benzylic ketones such
plots are curved and can be analyzed as monomer-dimer
equilibria.7 At concentrations above 5 × 10-4 M the plot
becomes virtually linear with a slope corresponding to a
dimer.
For a monomer-dimer equilibrium, a plot of the exper-
imental equilibrium constants K in eq 2 vs C/K (C is
overall concentration of salt) can be shown to be linear.
Figure 6 shows such a plot for the present case. The
intercept gives the equilibrium constant for the monomer,
and the slope is equal to 2K2K2, in which K2 is the
dimerization constant (n ) 2 in equilibrium 1).9 The
monomer cesium ion-pair pK of DBK is 18.07 ( 0.03, and
the dimerization constant is (3.5 ( 0.5) × 103 M-1. The
errors are those arising from standard deviations of
intercept and slope. The error is larger for the dimer-
ization constant because it includes errors in both the
intercept and slope. The theoretical line corresponding
cesium is also interesting. The cesium salt has λmax
)
538 nm in THF.15 Even a simple Hu¨ckel treatment
predicts that DBK dianion has a higher transition energy
than the allylic anion; thus the electrostatic effect of the
O- substituent is relatively small and, accordingly, the
replacement of this substituent by the more complex
substituent in 3 has no significant further effect.
Solutions of Li2DBK were obtained by titration of a
solution of LiDBK in THF by lithium diisopropylamide
(LDA). No changes in spectra were observed by a change
in the concentration of Li2DBK or by addition of LiBPh4.
These results suggest no significant dissociation of the
salt to free ions or formation of dimer. However, the
broader spectral band of Li2DBK (A in Figure 3) com-
pared to Cs2DBK and its lower extinction coefficient
(Table 1) suggest the presence of more than one species
in solution. The spectra of Li2DBK at temperatures from
-17 to +52 °C show a gradual shift of λmax from 502 to
491 nm with an isosbestic point at 492.5 nm (Figure 4).
This behavior is typical of an equilibrium between solvent
separated and contact ion pairs (SSIP and CIP) of
organolithium compounds.16,17 We were not able to
(12) Dietrich, H.; Mahdi, W.; Wilhelm, D.; Clark, T.; Schleyer, P. v.
R. Angew. Chem. 1984, 96, 623-5.
(17) Gronert, S.; Streitwieser, A., J r. J . Am. Chem. Soc. 1988, 110,
2836-2842 and references cited therein.
(13) Kos, A. J .; Clark, T.; Schleyer, P. v. R. Angew. Chem. 1984, 96,
622-3.
(18) Stratakis, M.; Streitwieser, A. J . Org. Chem. 1993, 58, 1989-
1990.
(14) Abu-Hasanayn, F.; Streitwieser, A. Unpublished results.
(15) Thiele, G.; Streitwieser, A. J . Am. Chem. Soc. 1994, 116, 446-
454.
(16) Smid, J . in Ions and Ion Pairs in Organic Reactions; Szwarc,
M., Ed., Wiley-Interscience: New York, 1972; Vol. 1, p 85.
(19) Kaufman, M. J .; Gronert, S.; Streitwieser, A., J r. J . Am. Chem.
Soc. 1988, 110, 2829-2835.
(20) Kaufman, M. J .; Streitwieser, A., J r. J . Am. Chem. Soc. 1987,
109, 6092-6097.
(21) Gareyev, R.; Streitwieser, A. J . Org. Chem., in press.

4592 J . Org. Chem., Vol. 61, No. 14, 1996
Gareyev et al.
means that 27% of the indicator lithium salts are
dissociated at a typical acidity measurement concentra-
tion of 10-4 M (total concentration of salt). If both acids
in eq 2 are indicator hydrocarbons and their salts have
the same dissociation constants, no change in the ex-
perimental K is observed with the variation of their
concentration, and that constant represents the true ion-
pair acidity, even though free ions are also present in
the solution. In the case of DBK, however, as well as
other compounds forming contact ion pairs with lithium,
the dissociation constants are much lower, and this
difference appears as a decrease of the acidity at lower
concentrations of the salt. The observed relative acidity
then depends both on the concentration of LiDBK (if it
is aggregated) and the salt of the indicator. The dis-
sociation constant for LiDBK may be estimated from the
difference between the ionic pK of DBK (18.7)8 and
lithium ion pair pK (11.62, see below). The difference of
about 7 pK units corresponds to a difference of seven
orders of magnitude between Kd of the lithium salt of the
indicator and that of LiDBK; i.e., Kd for LiDBK is on the
order of 10-12 M. This value implies an extent of
dissociation of 0.01% in a 10-4 M solution, an amount
that is entirely negligible for the present purpose. The
necessary corrections can be made to K knowing the
concentration of the lithium salt of the indicator and
assuming its Kd to be 10-5 M. Both experimental and
corrected values of K are presented in Table S3. Exami-
nation of the corrected lithium pK values for DBK shows
that it varies with the LiDBK concentration much less
than the cesium pK. A plot of K vs C/K similar to that
of Figure 6 gives the dimerization constant for LiDBK of
420 ( 70 M-1, which is almost 1 order of magnitude lower
than that for CsDBK. This result is in accord with the
generalization that in THF lithium salts are generally
less aggregated than the corresponding cesium salts21
and that in some cases aggregation is observed for cesium
under conditions for which the lithium salts are mono-
meric.9 The monomer lithium ion pair pK of DBK was
extrapolated as 11.62 ( 0.02.
The second cesium pK of DBK, representing the ease
of abstraction of an allylic proton from its enolate, was
measured against di-o-tolylmethane as indicator (pK
34.22).10 The results summarized in Table S4 show a
remarkable invariability of the equilibrium constant. The
value of pK does not depend on the concentration of
dianion salt or that of enolate and is 33.70 ( 0.02. The
concentration range covered is (0.35 - 3.5) × 10-3 M in
CsDBK, a range in which the relative amounts of its
monomer and dimer vary substantially. The correspond-
ing concentration of Cs2DBK ranges from (0.1 - 2) × 10-4
M, sufficiently dilute that aggregation is not expected to
be important. Yet, if Cs2DBK were just monomeric in
these experiments the actual acidity equilibrium constant
should vary by more than a factor of two. CsDBK is in
5-70 fold excess over Cs2DBK and the results are
consistent with formation of a mixed aggregate that was
formulated above as 2 in equilbrium with Cs2DBK
monomer (eq 3).
F igu r e 5. Plot of the experimental cesium pK values for
dibenzyl ketone in THF at 25 °C vs the logarithm of formal
concentration of DBK cesium enolate present in solution. The
line shown is the theoretical curve for true pK ) 18.07 and
ion pair dimerization constant K2 ) 3500 M-1. Data are from
Tables S2.
F igu r e 6. Plot of the equilibrium constant for the reaction:
DBK + Cs(9-BpFl) h 9-BpFl + CsDBK vs the total concentra-
tion of CsDBK divided by this constant. The line shown is the
least squares fit: intercept 1.78 ( 0.12, slope 21800 ( 500,
correlation coefficient 0.995. Data are from Table S2.
to these values is shown in Figure 5 and fits experimental
points well. The above dimerization constant is higher
than that reported earlier.7 The difference results from
including acidity data at higher concentrations of the
enolate, which significantly increases the accuracy of the
determination of K2. It is notable also that the K2
obtained from the SVD analysis is lower than that from
the acidity data. This is in accord with the observation
that SVD generally underestimates the aggregation
constant.9,22 Nevertheless, SVD provides invaluable
information about the number and type of species present
in solution, in addition to their individual spectra.
Lithium ion-pair acidity results in THF obtained using
1,3-diphenylindene as indicator (pK 12.32)23 are pre-
sented in Table S3. In the case of lithium as a counterion
the analysis of the acidity data is complicated by the
dissociation of the salt of the indicator acid to free ions.
It was shown earlier19 that lithium salts of typical
indicator acids (fluorenyl and triarylmethyl hydrocarbons
with delocalized carbanions) have dissociation constants
(Kd) in THF that are all close to 1 × 10-5 M. This value
Kc
CsDBK + Cs2DBK y\z Cs2DBK‚CsDBK
(3)
The simple electrostatic argument given above sug-
gests that the equilibrium constant for complex forma-
tion, Kc, should be of similar magnitude to the dimeriza-
tion constant of CsDBK monomer, K2. Indeed, if Kc )
(22) Abbotto, A.; Streitwieser, A. J . Am. Chem. Soc. 1995, 117,
6358-9.
(23) Streitwieser, A. et al. Results to be published.

Lithium and Cesium Ion-Pair Acidities of Dibenzyl Ketone
J . Org. Chem., Vol. 61, No. 14, 1996 4593
2K2 the relative amounts of complex and monomer of Cs2-
DBK and of dimer and monomer of CsDBK remain the
same and the resulting proton transfer equilibrium is
independent of concentration, as found experimentally!
Our results reveal little about the stereochemistry of
the phenyl groups except that they must be significantly
conjugated to the carbanion center. A previous NMR
study has reported on such stereochemistry for Li2DBK
in THF solution.26
It is instructive to compare the second pK of DBK with
the acidity of 1,3-diphenylpropene (DPP), a compound
having a close structural relationship to DBK enolate.
This compound was reported to have a cesium pK in THF
of 27.85,15 almost 6 pK units more acidic than CsDBK.
The difference between the cesium salt of 1,3-diphenyl-
propene and Cs2DBK is the presence of a second cesium
cation and an O- group at the central carbon of the DBK
enolate allylic anion system. It was shown previously
that the presence of a COOLi group has an acidi-
fying effect on an adjacent benzylic hydrogen almost as
great as that of a COOR group, leading to a relatively
low second lithium ion pair pK for 1-naphthylacetic
acid.24
Exp er im en ta l Section
Gen er a l. The instrumentation and techniques used for the
UV-vis spectral studies and acidity measurements were re-
cently described in detail.9
Ma ter ia ls. Commercial dibenzyl ketone (99% Aldrich) was
vacuum distilled three times from calcium hydride in
a
Kugelrohr apparatus. After the final distillation the receiving
bulb was filled with dry argon and taken into the glove box.
Material was then transferred to a vial; melting point 35-36
°C (lit.27 35 °C). Further evidence of purity was obtained from
the NMR spectrum. Purification of THF was described
earlier.9
Sp ectr a l a n d Acid ity Mea su r em en ts. General proce-
dures were published recently.9 Bases used for deprotonation
were (diphenylmethyl)cesium9 and cumylcesium28 in the study
of CsDBK and Cs2DBK, respectively, and (9,9,10-trimethyl-
9,10-dihydroanthracenyl)lithium21 and LDA for LiDBK and
Li2DBK. A solution of LDA in THF was prepared immediately
prior to use from material twice sublimed and stored at -20
°C in a freezer inside the glovebox.
We were not able to determine the second lithium pK
of DBK because the base typically used in lithium acidity
measurements, the lithium salt of 9,9,10-trimethyl-9,10-
dihydroanthracene, failed to abstract the proton from
LiDBK. This failure is not necessarily that it is not
strong enough as a base (the estimated pK of 9,9,10-
trimethyl-9,10-dihydroanthracene is well above 30), but
rather that the rate of deprotonation is extremely low.
It was found previously25 that the deprotonation and the
establishment of lithium acidity equilibria are very slow
for compounds with pKs above 25. Even the rate of
deprotonation of LiDBK by LDA is only moderate and
the reaction requires about 30 min for completion.
Ack n ow led gm en t. This research was supported in
part by NSF grant CHE92-21277.
Su p p or tin g In for m a tion Ava ila ble: Tables S2-S4 (4
pages). This material is available in libraries on microfiche,
immediately follows this article in the microfilm version of the
journal, and can be ordered from the ACS; see any current
masthead page for ordering information.
J O960013O
(24) Gronert, S.; Streitwieser, A. J . Am. Chem. Soc. 1988, 110,
4418-4419.
(25) Xie, L.; Streitwieser, A. J . Org. Chem. 1995, 60, 1339-1346.
(26) Wilhelm, D.; Clark, T.; Schleyer, P. v. R. J . Chem. Soc., Perkin
Trans. 2 1984, 915.
(27) CRC Handbook of Chemistry and Physics; CRC Press: Boca
Raton, FL, 1985.
(28) Bors, D. A.; Kaufman, M. J .; Streitwieser, A., J r. J . Am. Chem.
Soc. 1985, 107, 6975-6982.