The extremely low intrinsic reactivity of the P-(formylmethyl)triphenylphosphonium cation: An artefact due to strong covalent hydration of the carbonyl group

Article abstract of DOI:10.1039/a903667i

The title cation is found to exhibit the lowest Marcus intrinsic reactivity known so far for a carbon acid; this situation is shown to reflect a rate-limiting hydration/ dehydration reaction of the aldehyde functionality.

Full text of DOI:10.1039/a903667i

View Article Online / Journal Homepage / Table of Contents for this issue
The extremely low intrinsic reactivity of the P-(formylmethyl)-
triphenylphosphonium cation: an artefact due to strong covalent
hydration of the carbonyl group
G. Moutiers,* Y. X. Peng, A. Peignieux, M. J. Pouet and F. Terrier*
o
Laboratoire SIRCOB, UPRES-A CNRS n 8086, Université de Versailles, Bâtiment Lavoisier,
4
5 avenue des Etats Unis, 78035 Versailles cedex, France.
E-mail: moutiers@chimie.uvsq.fr; terrier@chimie.uvsq.fr
Received (in Cambridge) 7th May 1999, Accepted 3rd June 1999
The title cation is found to exhibit the lowest Marcus
intrinsic reactivity known so far for a carbon acid; this
situation is shown to reflect a rate-limiting hydration/
dehydration reaction of the aldehyde functionality.
It is well known that the intrinsic reactivity (k in the Marcus
0
sense) of a carbon acid is closely related to the degree of reson-
1,2
ance stabilization of the conjugate carbanion. The greater the
importance of this stabilization, the greater the lag with which
it develops relative to the extent of proton transfer along the
reaction coordinate and the lower the intrinsic reactivity of the
1–4
carbon acid. In recent years, the use of the log k scale has
0
been very helpful in clarifying the electronic mode of action of
some unusual activating structures, e.g. neutral or positively
charged organometallic residues, in governing the reactivity
5,6
of carbon acid sites. In this communication we report our
finding that the ionization of the P-(formylmethyl)triphenyl-
phosphonium cation 1 is associated with the lowest intrinsic
reactivity known to date for a carbon acid in aqueous solution.
As will be shown, this result rules out classical carbon acid
behaviour according to eqn. (1), being instead a spectacular
Fig. 1 Brønsted plot for ionization of 1 by primary and secondary
Ϫ3
amines in 70%H O–30%Me SO (v/v) at I = 0.5 mol dm KCl and
2
2
T = 25 ЊC.
B
OH
Ϫ
_
+
+
k
p
[B] ϩ k
p
[OH ]
was appreciable at the highest pH studied, one obtains:
Ph PCH COR
ϩ
Ph PCHCOR (1)
3
2
BH
ϩ
H
2
O
3
k Ϫ p
[BH ] ϩ k Ϫ p
OH
4
3
Ϫ1 Ϫ1
kp = 9.1 × 10 dm mol s .
1
2
3
R = H
R = CH3
C-1
C-2
C-3
The collected data call for five principal comments: (1) the
intrinsic reactivity of 1 is by far the lowest which has ever been
7
R = C H5
measured for a carbon acid; (2) perhaps more importantly, this
6
intrinsic reactivity is lower by more than 6 log units than those
of the closely related P-acetonyl and P-phenacyltriphenyl-
phosphonium cations 2 and 3 which behave as many carbonyl
manifestation of the high susceptibility of the carbonyl group
of this Wittig reagent precursor to covalent hydration.
10
11
acids, i.e. log k = 3.44 for 2 (3.92 in 50/50 at 20 ЊC); log
0
Based first on eqn. (1), all rate and equilibrium measure-
ments pertaining to the ionization of 1 were carried out in a
k = 3.55 for 3, in 70%H O–30%Me SO; (3) the low reactivity
0
2
2
OH
of 1 is also reflected by the kp value which is more than 1000-
2
OH
8
3
7
0%H O–30%Me SO (v/v) mixture at 25 ЊC and constant ionic
fold lower than those for 2 (pK = 7.40, kp 8 = 1.7 × 10 dm
2
2
a
Ϫ3
Ϫ1 Ϫ1
3
OH
3
Ϫ1 Ϫ1
strength of 0.5 mol dm KCl. The pK value of 1 was deter-
mol s ) and 3 (pK = 6.55, kp = 2.1 × 10 dm mol s ) in
a
a
mined both by potentiometry and analysis of the observed
absorbance variations at λmax of C-1 (266 nm) obtained at
equilibrium as a function of pH using a set of carboxylic acid
the same solvent; (4) the β value is surprisingly high, being out
of the range 0.5 ± 0.1 commonly found for carbon acids,
1
including the two analogs 2 and 3; (5) the points for the six
1
buffers: pKa = 5.67 ± 0.03. Pseudo-first-order rate constants,
primary amines and the two secondary amines employed lie on
the same Brønsted plot. This is in marked contrast with the
general observation that because of differences in solvation
requirements of the conjugated ammonium cations, secondary
amines are more reactive than primary amines in proton trans-
fers at carbon. Altogether, these results suggest that the ioniz-
ation of 1 cannot proceed through the simple equilibrium
approach of eqn. (1).
kobsd, for the interconversion of 1 and C-1 were measured by
monitoring the appearance of C-1 at 266 nm in various primary
and secondary amine buffers (B). In each buffer, the concen-
Ϫ3
tration of the amine reagent was varied in the range 5 × 10 –
Ϫ3
Ϫ5
Ϫ3
0
.1 mol dm at constant pH ([1] ≈ 5 × 10 mol dm ) and a
direct dependence of kobsd upon [B] according to the reduced
eqn. (2) was observed. As shown in Fig. 1, the results obtained
Although the characterization of the hydrated species 1,H
B
OH
Ϫ
9,12
kobsd = kp [B] ϩ kp [OH ]
(2)
does not seem to have been reported previously, a most rea-
sonable way to account for the observed behaviour of 1 is in
for the series of amines used describe a satisfactory Brønsted
terms of a high susceptibility of the strongly activated aldehyde
1
3–16
relationship (β = 0.89). From this plot, the log value of the
function to undergo covalent hydration.
Interestingly, the
B
1
13
31
Ϫ3
intrinsic rate constant of 1 was readily determined as log
NMR evidence ( H, C, P) is that a 0.1 mol dm solution of
B
1
B
k = log (k /q) at pK = pKa ϩ log (p/q); log k = Ϫ3.30. From
1 in 50%H O–50%Me SO mixture consists of only 20% of a
0
p
a
0
2
2
the contribution of the hydroxide ion pathway to eqn. (2) which
mixture of the related Z- and E-enol tautomers but 80% of 1,H
J. Chem. Soc., Perkin Trans. 2, 1999, 1287–1288
1287

View Article Online
3
1
1
3
18
(
P: δ = 20.4; H: δHα = 5.05, δHβ = 3.60, J(H H ) = 5.6,
which has ever been reported for a carbon acid. However, it is
clear that this behaviour is an artefact, being the reflection of
the strong susceptibility of the aldehyde function to covalent
hydration. Interestingly, the available evidence is that the
approach to the 1,H 1 equilibrium is the rate-limiting step of
the overall ionization process. In order to delineate all the
mechanistic facets of the behaviour of this compound, we are
currently focussing our efforts on the design of experimental
conditions allowing a direct kinetic study of this step.
P
α
β
2
3
13
1
J(PH ) = 13, J(PH ) = 4.1; C: δCα = 87.6, δC = 32, J(PC ) =
β
7
α
β
β
1
5
3.3) with no trace at all of the parent aldehyde form. This
makes it reasonable to assume that an essentially complete con-
version of 1 to 1,H occurs at the low substrate concentrations
Ϫ5
Ϫ3
(
≈5 × 10 mol dm ) involved in our kinetic measurements
conducted in the 70%H O–30%Me SO mixture. It follows that
2
2
the ionization of 1 is better depicted by eqn. (3) rather than
CH
+
+
K
a
Ph PCH CH(OH)
Ph PCH CHO
3
2
2
3
2
2 H
H O (K )
β
1
α
,H
1
References and notes
_
1 C. F. Bernasconi, Adv. Phys. Org. Chem., 1992, 27, 119 and refer-
ences therein.
+
ϩ
Ph PCHCHO ϩ H
(3)
3
C-1
2 (a) C. F. Bernasconi and R. L. Montanez, J. Org. Chem., 1997, 62,
8
162; (b) C. F. Bernasconi, J. A. Moreira, L. L. Huang and K. W.
eqn. (1), implying that both the above measured pK and log k0
values are apparent values that we will now refer to as pKa
a
obsd
Kittredge, J. Am. Chem. Soc., 1999, 121, 1674.
3 (a) D. M. Heathcote, J. H. Atherton, G. A. de Boos and M. A. Page,
J. Chem. Soc., Perkin Trans. 2, 1998, 535; (b) ibid., 1998, 541.
obsd
and log k0
.
Based on the above evidence two different arguments can be
4
(a) G. Moutiers, B. El Fahid, R. Goumont, A. P. Chatrousse and
F. Terrier, J. Org. Chem., 1996, 61, 1978; (b) F. Terrier, E. Kizilian,
R. Goumont, N. Faucher and C. Wakselman, J. Am. Chem. Soc.,
1998, 120, 9496.
CH
used to obtain an estimate of the hydration constant K = K
/
H
a
obsd
Ka
. The first is based on previous findings that the relative
acidifying effects of the CHO and CORЈ (RЈ = CH , C H )
groups upon the carbon acidity of compounds of general struc-
ture RCH CORЈ depend very little on the RЈ substituent, e.g.
3
6
5
5 G. Moutiers, A. Peignieux, D. Vichard and F. Terrier, Organo-
metallics, 1998, 63, 1944.
2
COC
6
H
5
CHO
14
CH
6
7
C. F. Bernasconi, Chem. Soc. Rev., 1997, 26, 299.
In accord with previous reports (refs. 8 and 9), C-1 exists as a
mixture of the two cis- and trans-enolate isomers in H O–Me SO
pKa
Ϫ pKa
≈ 3. Based on this difference, a pKa
value of ≈3.50 can be calculated for 1 in 70%H O–30%Me SO.
2
2
obsd
2
2
Combining with pKa
= 5.67 yields K ≈ 150 (pK = Ϫ2.17)
H H
mixtures.
in this solvent. For comparison, a value of K = 37 has been
H
8
9
C. J. Devlin and B. J. Walker, Tetrahedron, 1972, 28, 3501.
N. A. Nesmeyanov, S. T. Berman, P. V. Petrovsky, A. I. Lutsenko
13
reported for the hydration of chloroacetaldehyde.
Providing that the ionization of the aldehyde form remains
the rate-limiting step of the overall process of eqn. (3), another
estimate of K can be obtained from the relationship (4).
and O. A. Reutov, J. Organomet. Chem., 1977, 129, 41.
1
13
31
10 H, C and P NMR spectra show that 2 and 3 exist essentially as
H
the two keto tautomers in H O–Me SO mixtures. See also ref. 9.
2
2
1
1 C. F. Bernasconi, value quoted in ref. 1 (Table 3).
12 X. M. Zhang and F. G. Bordwell, J. Am. Chem. Soc., 1994, 116, 968.
obsd
log k0
= log k ϩ pKH
(4)
0
1
1
3 R. P. Bell, Adv. Phys. Org. Chem., 1966, 4, 1.
4 M. P. Harcourt and R. A. More O’Ferrall, J. Chem. Soc., Perkin
Trans. 2, 1995, 1415.
5 (a) T. J. Burkey and R. C. Fakey, J. Am. Chem. Soc., 1983, 105, 868;
(b) L. H. Funderburk, L. Aldwin and W. P. Jencks, J. Am. Chem.
Soc., 1978, 100, 5444.
Should 1 actually behave as 2 and 3 and exhibit normal
carbon acid behaviour, an intrinsic reactivity typical of a carb-
1
onyl compound should be found with log k ≈ 3–3.5. From eqn.
0
(
4), one thus obtains: pK ≈ Ϫ6.5, a value which combined with
H
obsd CH
pKa
will imply that 1 is a very strong acid (pKa ≈ Ϫ1). In
16 (a) H. J. Cristau and F. Plénat, in The Chemistry of Organo-
phosphorous compounds, ed. F. R. Hartley, J. Wiley and Sons, 1994,
Vol. 3, Chapter 2, p. 45; (b) H. J. Cristau, Chem. Rev., 1994, 94, 1299.
view of the aforementioned relationship governing the carbon
14
acidities of RCH CORЈ (RЈ = H, CH , C H ) compounds,
2
3
6
5
1
13
1
7 δ in ppm relative to Me Si as the internal reference ( H, C) and
4
31
such a conclusion is totally unreasonable. We therefore con-
CH
H PO ( P); J in Hz.
3 4
clude that the set of pK and pK
values obtained in the first
H
a
1
8 F. Terrier, D. Croisat, A. P. Chatrousse, M. J. Pouet, J. C. Hallé and
G. Jacob, J. Org. Chem., 1992, 57, 3684.
approach is more realistic and that the rate limiting step of
eqn. (3) must be the hydration/dehydration pathway and not the
ionization of 1.
In summary, we have found that the ionization of 1 is
formally associated with the lowest Marcus intrinsic reactivity
Communication 9/03667I
1
288
J. Chem. Soc., Perkin Trans. 2, 1999 1287–1288