Spectroscopy and absolute reactivity of ketenes in acetonitrile studied by laser flash photolysis with time-resolved infrared detection

Article abstract of DOI:10.1021/ja964155b

Laser flash photolysis with time-resolved infrared detection of transients (LFP-TRIR) has been used to study the IR spectroscopy and reactivity of a number of substituted ketenes, prepared by the 308-nm photolysis of α-diazocarbonyl precursors in acetonitrile solution at room temperature. The correlation of the experimental ketene asymmetric stretching frequency to the Swain-Lupton field (F) and resonance (R) effect substituent parameters was unsatisfactory, whereas the correlation to the inductive substituent parameter (σ₁) of Charton gave excellent results. This suggests that the asymmetric stretching frequency of substituted ketenes depends mainly on the inductive (i.e., field) effect of the substituents. The mechanism and kinetics of the reactions of these ketenes with various amines in acetonitrile were also studied. An intermediate species identified as either zwitterionic ylide or amide enol formed in the nucleophilic addition of the secondary amine to the C(α) of the ketene is observed by TRIR. The decay of this species is assisted by the amine and is concomitant with the formation of an amide, the final product of the reaction. Our kinetic data on ketene amine reactions show a general trend, indicating a much higher reactivity (ca. 3 orders of magnitude difference in the corresponding rate constants) of secondary amines compared with that of tertiary amines. Secondary diethylamine shows reactivity similar to those observed for primary amines, while secondary piperidine seems to be, in general, somewhat more reactive. The observed trend is rationalized in terms of the steric effects exerted by both amine and ketene substituents. Our data on para-substituted phenyl ketenes provide support for the negative charge development on the ketene moiety in the transition state, with electron-withdrawing substituents accelerating and electron-releasing substituents slowing down the addition reaction.

Full text of DOI:10.1021/ja964155b

J. Am. Chem. Soc. 1998, 120, 1827-1834
1827
Spectroscopy and Absolute Reactivity of Ketenes in Acetonitrile
Studied by Laser Flash Photolysis with Time-Resolved Infrared
Detection
Brian D. Wagner,*,†,‡ Bradley R. Arnold,† Gerald S. Brown,† and Janusz Lusztyk*,†
Contribution from the Steacie Institute for Molecular Sciences, National Research Council of Canada,
Ottawa, Ontario, Canada K1A 0R6, and the Department of Chemistry, UniVersity of Prince Edward
Island, Charlottetown, Prince Edward Island, Canada C1A 4P3
ReceiVed December 2, 1996
Abstract: Laser flash photolysis with time-resolved infrared detection of transients (LFP-TRIR) has been used
to study the IR spectroscopy and reactivity of a number of substituted ketenes, prepared by the 308-nm photolysis
of R-diazocarbonyl precursors in acetonitrile solution at room temperature. The correlation of the experimental
ketene asymmetric stretching frequency to the Swain-Lupton field (F) and resonance (R) effect substituent
parameters was unsatisfactory, whereas the correlation to the inductive substituent parameter (σI) of Charton
gave excellent results. This suggests that the asymmetric stretching frequency of substituted ketenes depends
mainly on the inductive (i.e., field) effect of the substituents. The mechanism and kinetics of the reactions of
these ketenes with various amines in acetonitrile were also studied. An intermediate species identified as
either zwitterionic ylide or amide enol formed in the nucleophilic addition of the secondary amine to the CR
of the ketene is observed by TRIR. The decay of this species is assisted by the amine and is concomitant with
the formation of an amide, the final product of the reaction. Our kinetic data on ketene amine reactions show
a general trend, indicating a much higher reactivity (ca. 3 orders of magnitude difference in the corresponding
rate constants) of secondary amines compared with that of tertiary amines. Secondary diethylamine shows
reactivity similar to those observed for primary amines, while secondary piperidine seems to be, in general,
somewhat more reactive. The observed trend is rationalized in terms of the steric effects exerted by both
amine and ketene substituents. Our data on para-substituted phenyl ketenes provide support for the negative
charge development on the ketene moiety in the transition state, with electron-withdrawing substituents
accelerating and electron-releasing substituents slowing down the addition reaction.
Introduction
CdCdO moiety.11 The IR spectroscopy of ketenes has been
extensively studied in cryogenic matrices12-14 and has been
applied recently to investigate ketene reactions toward nucleo-
philes in this medium.15,16 We17-21 and others22 have also
recently demonstrated that LFP with time-resolved infrared
Ketenes are extremely important reactive species, which occur
as transients in numerous thermal and photochemical reactions.1-5
In many cases, ketenes have relatively weak UV absorption
bands below 400 nm, making them difficult to resolve from
other species in UV laser flash photolysis (LFP), although a
number of studies of ketene reactivity employing this technique
have been reported.6-10 However, all ketenes exhibit a very
(11) Lin-Vien, D.; Cothup, N. B.; Fatetey, W. G.; Graselli, J. G. The
Handbook of Infrared and Raman Characteristic Frequencies of Organic
Molecules; Academic: San Diego, CA, 1991.
(12) Moore, C. B.; Pimentel, G. C. J. Chem. Phys. 1963, 38, 2816-
2829.
(13) Torres, M.; Ribo, J.; Clement, A.; Strausz, O. P. NouV. J. Chim.
1981, 5, 351-352.
strong and characteristic IR band in the region of 2100 cm-1
arising from the out-of-phase or antisymmetric stretch of the
,
(14) Maier, G.; Preiss, T.; Reisenauer, H. P.; Hess, B. A., Jr.; Schaad,
L. J. J. Am. Chem. Soc. 1994, 116, 2014-2018.
(15) Qiao, G. G.; Andraos, J.; Wentrup, C. J. Am. Chem. Soc. 1996,
118, 5634-5638.
† National Research Council of Canada. NRCC publication 40855.
‡ University of Prince Edward Island.
(1) (a) Patai, S., Ed. The Chemistry of Ketenes, Allenes, and Related
Compounds; John Wiley and Sons: New York, 1980. (b) Tidwell, T. T.
Ketenes; John Wiley and Sons: New York, 1995.
(16) Visser, P.; Zuhse, R.; Wong, M. W.; Wentrup, C. J. Am. Chem.
Soc. 1996, 118, 12598-12602.
(17) Arnold, B. R.; Brown, C. E.; Lusztyk, J. J. Am. Chem. Soc. 1993,
115, 1576-1577.
(2) Seikaty, H. R.; Tidwell, T. T. Tetrahedron 1986, 42, 2587-2613.
(3) Tidwell, T. T. Acc. Chem. Res 1990, 23, 273-279.
(4) Snider, B. B. Chem. ReV. 1988, 88, 793-811.
(18) Wagner, B. D.; Zgierski, M. Z.; Lusztyk, J. J. Am. Chem. Soc. 1994,
116, 6433-6434.
(5) Ye, T.; McKervey, M. A. Chem. ReV. 1994, 94, 1091-1160.
(6) Allen, A. D.; Kresge, A. J.; Schepp, N. P.; Tidwell, T. T. Can. J.
Chem. 1987, 65, 1719-1723.
(19) Allen, A. D.; Colomvakos, J. D.; Egle, I.; Lusztyk, J.; McAllister,
M. A.; Tidwell, T. T.; Wagner, B. D.; Zhao, D. J. Am. Chem. Soc. 1995,
117, 7552-7553.
(7) Andraos, J.; Kresge, A. J. J. Am. Chem. Soc. 1992, 114, 5643-5646.
(8) Andraos, J.; Chiang, Y.; Huang, C.-G.; Kresge, A. J.; Scaiano, J. C.
J. Am. Chem. Soc. 1993, 115, 10605-10610.
(20) Camara de Lucas, N.; Netto-Ferreira, J. C.; Andraos, J.; Lusztyk,
J.; Wagner, B. D.; Scaiano, J. C. Tetrahedron Lett. 1997, 38, 5147-5150.
(21) Allen, A. D.; Colomvakos, J. D.; Diederich, F.: Egle, I.; Hao, X.;
Liu, R.; Lusztyk, J.; Ma, J.; McAllister, M. A.; Rubin, Y.; Sung, K.; Tidwell,
T. T.; Wagner, B. D. J. Am. Chem. Soc. 1997, 119, 12125-12130.
(22) Oishi, S.; Watanabe, Y.; Kuriyama, Y. Chem. Lett. 1994, 2187-
2190.
(9) Andraos, J.; Kresge, A. J. J. Photochem. Photobiol. A 1991, 57, 165-
173.
(10) Allen, A. D.; Andraos, J.; Kresge, A. J.; McAllister, M. A.; Tidwell,
T. T. J. Am. Chem. Soc. 1992, 114, 1878-1879.
S0002-7863(96)04155-8 CCC: $15.00 © 1998 American Chemical Society
Published on Web 02/14/1998

1828 J. Am. Chem. Soc., Vol. 120, No. 8, 1998
Wagner et al.
detection (LFP-TRIR) is ideally suited for both mechanistic and
kinetic studies of ketenes in solution since the characteristic
spectroscopic region near 2100 cm-1 is transparent to most
solvents and void of absorptions due to other photogenerated
transients and products. In this paper, we employ LFP-TRIR
for systematic studies of the infrared spectroscopic properties
of ketenes and explore the mechanism and kinetics of the ketene
amine reactions.
Ketene stretching frequencies are very sensitive to substitu-
tion, with values of the asymmetric stretch reported from 2085
to 2197 cm-1 11
There have been a number of efforts to
.
Figure 1. TRIR spectrum of phenyl ketene observed 4 µs after 308-
nm photolysis of R-diazoacetophenone in N2-saturated CH3CN solution.
correlate these observed frequencies to substituent parameters.
Gano and Jacob23 used experimental frequencies obtained from
the literature for a set of symmetrically disubstituted ketenes to
test the correlation to the field or inductive (F) and resonance
(R) parameters model of Swain and Lupton.24,25 These param-
eters were chosen because of their generality and the useful
separation of the substituent effect into field and resonance
components. In this approach, the ketene stretching frequencies
were fit to a model equation containing F and R as two
independent parameters. A good correlation with the following
fit parameters was obtained:
Scheme 1
have reported kinetic studies of the reaction of diphenyl ketene
with various bases in aqueous solution.7 They concluded from
the observed reactivities that the reaction proceeds via direct
nucleophilic attack of the base on the carbonyl carbon of the
ketene (CR) and not via base-catalyzed hydration.
νjx (cm-1) ) 2124 + 24Fx + 30Rx
(1)
There have been relatively few reported studies of ketene
reactivities in nonaqueous media. Satchell and co-workers30-33
studied the reactions of ketene, dimethyl ketene, and diphenyl
ketene with anilines in ether and benzene. They reported the
rate law
However, this correlation was based on a rather limited data
set of five experimental frequencies, which were measured in
different media (gas phase, cryogenic matrix). McAllister and
Tidwell26 extended this approach by using ketene stretching
frequencies obtained by ab initio calculations for both sym-
metrically disubstituted and monosubstituted ketenes. Their
calculated frequencies agreed within ca. 10 cm-1 with the
available experimental values but had a poorer correlation (r )
0.87) than Jacob and Gano’s, with a near zero contribution from
the resonance effect:
rate ) (k1[ArNH2] + k2[ArNH2]2)[ketene]
(3)
for which they proposed the existence of two different transition
states, one involving one aniline molecule, the other involving
two. These workers also studied the addition of bulky alcohols
to ketene in isooctane and CCl4 solutions and again reported
nonlinear quenching plots.34 In this case, the nonlinearity was
explained by a mechanism involving the more reactive (relative
to the monomer) trimers at high alcohol concentrations.
νjx (cm-1) ) 2124 + 91Fx - 6Rx
(2)
They obtained a similar correlation (r ) 0.86) by fitting their
data just to the field parameter F. Obviously, there is a need
for a larger set of experimental ketene frequencies obtained
under standardized experimental conditions to allow for more
definitive delineation of the field and resonance effects.27
Nucleophilic addition reactions of ketenes are well-known
and have been extensively reviewed.1-5 The vast majority of
studies, however, have dealt with the hydration kinetics of
ketenes in aqueous or mixed aqueous/organic solvent.6,8-10,28,29
It has been shown that the acid-catalyzed reaction involves rate-
limiting proton transfer to Câ perpendicular to the plane of the
ketene, whereas reaction with OH- or H2O involves nucleophilic
attack on CR in the plane of the ketene.6 Andraos and Kresge
Results and Discussion
Substituent Effects on the Ketene Asymmetric Stretching
Frequency. In this paper, we report the results on the IR
spectroscopy of a number of ketenes, studied by LFP-TRIR in
acetonitrile at room temperature. All except for one of the
ketenes were generated by the 308-nm photolysis of the
corresponding R-diazocarbonyl compounds, via the Wolff
rearrangement of the initial carbene.35-37 For example, phenyl
ketene was generated via photolysis of R-diazoacetophenone,
as illustrated in Scheme 1. Ketene VIIIa was generated from
the 308-nm photolysis of 2H-pyran-2-one.17 In all cases,
extremely strong and well-defined absorptions were observed
(23) Gano, J. E.; Jacob, E. J. Spectrochim. Acta 1987, 43A, 1023-1025.
(24) Swain, C. G.; Lupton, E. C., Jr. J. Am. Chem. Soc. 1968, 90, 4328-
4337.
(25) Swain, C. G.; Unger, S. H.; Rosenquist, N. R.; Swain, M. S. J. Am.
Chem. Soc. 1983, 105, 492-502.
(30) Briody, J. M.; Satchell, D. P. N. Tetrahedron 1966, 22, 2649-
2653.
(31) Lillford, P. J.; Satchell, D. P. N. J. Chem. Soc. B 1967, 360-365.
(32) Lillford, P. J.; Satchell, D. P. N. J. Chem. Soc. B 1968, 54-57.
(33) Lillford, P. J.; Satchell, D. P. N. J. Chem. Soc. B 1970, 1016-
1019.
(26) McAllister, M. A.; Tidwell, T. T. Can. J. Chem. 1994, 72, 882-
887.
(27) The only previous experimental results that we are aware of on the
substituent effect on the ketene stretching frequencies in solution are that
by: Melzer, A.; Jenny, E. F. Tetrahedron Lett. 1968, 4503-4506. The values
in 1,2-dichloroethane for the relevant diphenyl ketene, 4,4′-dimethoxy-
diphenyl ketene, and 4,4′-dinitrodiphenyl ketene are 2100, 2096, and 2109
cm-1, respectively.
(34) Donohoe, G.; Satchell, D. P. N.; Satchell, R. S. J. Chem. Soc.,
Perkins Trans. 2 1990, 1671-1674.
(35) Meier, H.; Zeller, K.-P. Angew. Chem., Int. Ed. Engl. 1975, 14,
32-43.
(28) Allen, A. D.; Tidwell, T. T. J. Am. Chem. Soc. 1987, 109, 2774-
2780.
(36) Tomioka, H.; Okuno, H.; Izawa, Y. J. Org. Chem. 1980, 45, 5278-
5283.
(29) Allen, A. D.; Stevenson, A.; Tidwell, T. T. J. Org. Chem. 1989,
54, 2843-2848.
(37) Scott, A. P.; Nobes, R. H.; Schaefer, H. F., III; Radom, L. J. Am.
Chem. Soc. 1994, 116, 10159-10164.

ReactiVity of Ketenes in Acetonitrile
J. Am. Chem. Soc., Vol. 120, No. 8, 1998 1829
Rh2 substituent parameters
Table 1. Observed Ketene Stretching Frequencies in Acetonitrile at 23 °Ca
R1
C
O
C
R2
precursors
structure
no.
no.
Ia
R1
R2
νj/cm-1
F
R
σI
O
I
H
H
t-Bu
2108
-0.02
-0.18
-0.01
CHN2
O
II
IIa
Ph
2118
2115
0.12
0.13
-0.13
-0.21
0.12
0.11
CHN2
O
CHN2
III
IIIa
H
H
H
p-MeOPh
p-CNPh
MeO
O
CHN2
IV
V
IVa
Va
2120
2122
NC
O
CHN2
p-NO2Ph
PhCtC
0.26
0.0
0.22
0.33
O2N
O
VI
VIa
H
H
H
2130
2116
2127
0.15
0.10
0.29
0.01
-0.17
-0.16
CHN2
Ph
O
VII
VIII
VIIa
VIIIa
PhCHdCH
Ph
O
CHN2
O
O
O
CH
HCCH
Ph
IX
X
IXa
Xa
Ph
Ph
Ph
2100
2119
Ph
N2
N2
O
Ph
Ph
PhC
O
O
a All frequencies are ( 2 cm-1. The substituent parameters F, R, and σI are given for monosubstituted ketenes where available.
in the ketene stretching region. The TRIR spectrum of phenyl
ketene in acetonitrile is shown in Figure 1. A strong, sharp
band is observed with a maximum at 2118 ( 2 cm-1, as well
as a weak bleaching at 2107 cm-1 resulting from the depletion
of the diazo band of the R-diazoacetophenone precursor. Similar
spectra were obtained for the other ketenes studied; the observed
ketene stretching frequencies are given in Table 1. The observed
frequencies for the monosubstituted ketenes range from 2108
to 2130 cm-1, a total range of only 22 cm-1. However, since
the data were obtained under standardized experimental condi-
tions, the set of seven monosubstituted ketenes does provide a
useful data set for correlations to substituent parameters. Table
1 also includes the values of F and R for these substituents,
taken from ref 38. The fit of the experimental frequencies to F
and R yielded the following equation:
Figure 2. Plot of ketene stretching frequencies calculated from eq 4
Vs experimental frequencies.
was much poorer, r ) 0.858 (Figure 2). The fit is especially
bad in the case of p-nitrophenyl ketene. The correlation of the
observed frequencies to just the field parameter F was even
poorer (r ) 0.768).
There has been some criticism of the Swain-Lupton ap-
proach.39,40 Hoefnagel et al.39 applied the Swain-Lupton
equation to a number of experimental systems, including the
pKa of substituted phenols and anilinium ions. Gross deviations
νjx (cm-1) ) 2117 + 43Fx + 35Rx
(4)
This is in reasonable agreement with the data obtained by Gano
and Jacob23 (eq 1), with nearly equal contributions from the
field and resonance terms. However, our correlation coefficient
(38) Hansch, C.; Leo, A.; Taft, R. W. Chem. ReV. 1991, 91, 165-195.

1830 J. Am. Chem. Soc., Vol. 120, No. 8, 1998
Wagner et al.
Figure 3. Plot of experimental ketene stretching frequencies Vs
Charton’s substituent parameter σI.
Figure 4. Observed phenyl ketene kinetic traces measured at 2118
cm-1 in the absence of quencher (a) and in the presence of 0.3 mM
diethylamine (b). The fit of the latter trace to a single-exponential decay
gave an excellent fit with a first-order decay constant (kobs) of 1.6 ×
from the experimental values were observed, leading to the
conclusion that the Swain-Lupton approach is much more
limited than was originally claimed. Charton40 came to the same
conclusion and proposed an alternative parameter, σI. This is
purely an inductive effect parameter, with no contribution from
delocalization (resonance) effects. Charton showed that for a
number of systems which are free from resonance effects, σI
was a much better representation of the localized field effect
than was F. Unfortunately, the values for this parameter were
only available for five of the monosubstituted ketenes that we
studied.40 An excellent correlation (r ) 0.993) was obtained
for this admittedly limited data set, as shown in Figure 3. The
fit equation obtained was
105 s-1
.
Figure 5. Quenching plot of the observed first-order decay constant
(kobs) for the decay of phenyl ketene Vs the concentration of diethylamine
in CH3CN. Linear least-squares analysis yields a slope of 4.8 × 108
M-1 s-1 with a correlation coefficient of 0.994.
νj (cm-1) ) 2108.8 + 63.5σI
(5)
The excellent correlation observed with σI may suggest that only
the inductive, or localized field, effect of substituents is
important in determining the stretching frequency of the
monosubstituted ketenes. This is in agreement with the conclu-
sions of McAllister and Tidwell;26 however, we find that σI is
a much better measure of the inductive effect than is F.
Equation 5 should provide a good predictive tool for the
expected frequency of the monosubstituted ketenes; obviously
more experimental frequencies would be useful for further
testing the validity of this equation.
failed to observe the expected36 ethoxy ketene upon photolysis
of the commercially available ethyl diazoacetate. Similarly,
photolysis of N,N-dimethyldiazoacetamide (synthesized in our
lab) did not produce detectable amounts of the expected N,N-
dimethylamino ketene. Rando41 and Tomioka et al.42 reported
predominant formation of the cyclic lactam products (formed
by intramolecular C-H insertion of the initially formed carbene)
upon photolysis of this compound, indicating that the yield of
the ketene product is relatively low.
Nucleophilic Addition Reactions of Amines with Ketenes.
Kinetic studies of the reactions of a family of ketenes with a
wide variety of amines are reported herein. In the absence of
added quencher, all the ketenes studied except VIIIa were stable
on the longest available experimental time scale (128 µs),
indicating lifetimes of greater than 500 µs. However, the
lifetimes of all the ketenes studied were shortened by added
amines. Figure 4 shows the observed kinetic trace of phenyl
ketene in the absence of quencher (a) and in the presence of
0.3 mM diethylamine (b). A stable signal is observed in the
former case, while in the latter, a well-defined first-order decay
is recorded. Assuming pseudo-first-order kinetics, the observed
rate constant of ketene decay (kobs) as a function of added
quencher concentration ([Q]) is given by
The observed correlation of the ketene stretching frequency
with the inductive substituent effect parameter can be explained
in terms of the following two major resonance forms for
monosubstituted ketenes:
The presence of electron-accepting groups (large σI) will tend
to stabilize the carbanion in resonance form 2, increasing its
contribution to the overall structure. This increases the CO bond
order and, hence, increases the observed CO stretching fre-
quency. In the case of electron releasing groups (small or
negative σI), the carbanion form 2 is destabilized, and thus, the
CO bond order is reduced, and the CO stretching frequency is
lowered.
kobs ) ko + kq[Q]
(6)
Efforts were made to expand the observed range of ketene
frequencies. Our attempts to observe ketenes with heteroatom
substituents (expected to have larger shifts in the stretching
frequency) from the corresponding R-diazocarbonyl precursors
in LFP-TRIR experiments were unsuccessful. For example, we
where ko is the observed decay rate constant in the absence of
quencher and kq is the second-order quenching rate constant.
The values of kq were thus obtained from the slopes of the plots
of kobs Vs [Q]. Figure 5 shows a representative quenching plot
for the case of phenyl ketene reacting with diethylamine. Within
(39) Hoefnagel, A. J.; Oøsterbeek, W.; Wepster, B. M. J. Org. Chem.
1984, 49, 1993-1997.
(40) Charton, M. J. Org. Chem. 1984, 49, 1997-2001.
(41) Rando, R. R. J. Am. Chem. Soc. 1970, 92, 6706-6707.
(42) Tomioka, H.; Kitagawa, H.; Izawa, Y. J. Org. Chem. 1979, 44,
3072-3075.

ReactiVity of Ketenes in Acetonitrile
J. Am. Chem. Soc., Vol. 120, No. 8, 1998 1831
Scheme 2
terionic ylides (Scheme 2) which result from attack of an amine
at the CR position of the ketene. More recently the species
identified as such zwitterions were characterized by IR in matrix-
isolation studies15,16 of this reaction with the corresponding
bands assigned in the 1650-1730-cm-1 region.44 Similarly,
the low-temperature reaction between fluorenylidene ketene and
imidazole in a phenolic resin was reported to produce a
zwitterionic ylide characterized by an absorption at 1635 cm-1
in the IR.46 Following these leads, in our own recent LFP-
TRIR studies of the reaction of fluorenylidene ketene with
diethylamine, the band at 1674 cm-1 was assigned to the
corresponding initial zwitterionic product.20 In this reaction,
however, as in the reaction of phenyl ketene with diethylamine
(Scheme 3), it is possible to envisage that the first observable
intermediate is an amide enol formed either directly through
the four-membered-ring transition state47 or via a short-lived
zwitterionic intermediate, and not the zwitterionic ylide itself.
To our knowledge, there has been only one direct observation
of amide enol: by 1H NMR in the reaction of bis(2,4,6-
triisopropylphenyl)acetic acid with Me2NH.48 IR bands corre-
sponding to the CdC stretching frequency of the amide enols
should occur in a similar region to that of the corresponding
ketene O,N-acetals.49 For a family of dimethyl ketene O,N-
acetals, Me2CdC(OR1)NR2R3, these frequencies were measured
in the 1640-1680 cm-1 region of the spectrum.52 They
correspond very well to the bands we observed at 1680 and
1674 cm-1, respectively, for the initial products in the reactions
of phenyl ketene and fluorenylidene ketene20 with diethylamine.
Thus, these bands may well correspond to the appropriate amide
enols.
Figure 6. (a) TRIR spectrum in the carbonyl region observed 2 µs
(b) and 60 µs (O) after 308-nm photolysis of R-diazoacetophenone in
the presence of 2.5 mM diethylamine in N2-saturated CH3CN solution.
(b) Observed kinetic traces at 1680 and 1640 cm-1
.
the explored concentration range of the quencher, good linearity
of the plot is obtained, yielding a rate constant of 4.8 × 108
M-1 s-1
.
Upon photolysis of R-diazoacetophenone in the presence of
diethylamine, a transient species with a band maximum at 1680
cm-1 was observed to grow in with kinetics identical with that
observed for the decay of the phenyl ketene band at 2116 cm-1
.
A plot of the observed first-order rate constant for the growth
of this transient Vs diethylamine concentration yielded a second-
order rate constant of 4.5 × 108 M-1 s-1, in excellent agreement
with that obtained for the decay of phenyl ketene, indicating
that this species is the initial product of the reaction of phenyl
ketene with diethylamine. This species was not stable, and its
decay was also dependent on the diethylamine concentration.
It leads to the formation of a stable species observed at 1640
cm-1, growing in with kinetics identical with that observed for
the decay of the 1680-cm-1 band. Figure 6 shows the TRIR
spectrum and kinetic traces observed in this region for the LFP
of a solution of R-diazoacetophenone containing 2.5 mM
diethylamine. At this concentration of diethylamine, growth
of the 1680-cm-1 band is instantaneous on the applied time
scale, while the decay of this band and the growth of the stable
product at 1640 cm-1 are clearly concomitant. Plots of the
observed first-order decay or growth rate constants Vs diethyl-
amine concentration yielded values for the second-order rate
constants of 2.1 × 107 M-1 s-1 for the decay of the 1680-cm-1
band and 2.0 × 107 M-1 s-1 for the growth of the stable 1640-
cm-1 band.
Our current data do not allow for unequivocal identification
of the species that are observed in the reaction of phenyl ketene
with diethylamine. The existing evidence indicates that the
1680-cm-1 band could correspond either to the zwitterion 3 or
the amide enol 4.
Recent theoretical treatment of a prototypical reaction of
ketene with ammonia53 supports the notion of the initial addition
(44) Recently a relevant zwitterionic complex
where Ar ) C6H5Cr(CO)3 and R2 ) (CH2)5, was isolated and its X-ray
structure established, suggesting a similarity of the CdO bond length in
the complex and in an amide moiety.45
(45) Chelain, E.; Goumont, R.; Hamon, L.; Parlier, A.; Rudler, M.;
Rudler, H.; Daran, J.-C.; Vaissermann, J. J. Am. Chem. Soc. 1992, 114,
8088-8098.
(46) Pacansky, J.; Chang, J. S.; Brown, D. W.; Schwarz, W. J. Org.
Chem. 1982, 47, 2233-2234.
The stable product at 1640 cm-1 can be readily identified as
N,N-diethylphenylacetamide. The assignment of the 1680-cm-1
band is more difficult since at least two intermediates can be
considered as initial products of the reaction of ketene with
amine. Transient species that have been previously observed
in a number of ketene reactions with pyridine in solution by
time-resolved UV-vis spectroscopy8,43 were assigned to zwit-
(47) Such a transition state may be extended by inclusion of the solvent
molecule(s).
(48) Frey, J.; Rappoport, Z. J. Am. Chem. Soc. 1996, 118, 3994-3995.
(49) For related systems, the CdC stretch was observed at 1661.9 cm-1
in matrix-isolation studies of syn-vinyl alcohol50 and in the range of 1610-
1680 cm-1 for a family of vinyl ethers.51
(50) Rodler, M.; Blom, C. E.; Bauder, A. J. Am. Chem. Soc. 1984, 106,
4029-4035.
(51) The Aldrich Library of FT-IR Spectra, 1st ed.; Pouchert, C. J., Ed.;
Aldrich: New York, 1985; Vol. I.
(43) Wang, J.-L.; Toscano, J. P.; Platz, M. S.; Nikolaev, V.; Popic, V.
J. Am. Chem. Soc. 1995, 117, 5477-5483. Boate, D. R.; Johnston, L. J.;
Kwong, P. C.; Lee-Ruff, E.; Scaiano, J. C. J. Am. Chem. Soc. 1990, 112,
8858-8863. Barra, M.; Fisher, T. A.; Cernigliaro, G. J.; Sinta, R.; Scaiano,
J. C. J. Am. Chem. Soc. 1992, 114, 2630-2634.
(52) Schaumann, E.; Sieveking, S.; Walter, W. Chem. Ber. 1974, 107,
3589-3601.
(53) Tidwell, T. T.; Sung, S. J. Am. Chem. Soc., in press.

1832 J. Am. Chem. Soc., Vol. 120, No. 8, 1998
Wagner et al.
Scheme 3
Table 2. Bimolecular Rate Constants for the Reaction of Ketenes with Amines in Acetonitrile Solution at 23 °C (Estimated Error (20%)
kq/107 M-1 s-1
ketene
Na
NH b
b
b
b
no.
Ia
IIa
IIIa
IVa
Va
VIa
VIIa
IXa
Xa
R1
R2
Et3Na
Et2NHb
i-PrNH2
n-BuNH2
PhCH2NH2
H
H
H
H
H
H
H
Ph
Ph
t-Bu
Ph
<0.01
<0.01
0.023
0.028
c
0.046
48
33
95
95
110
49
1.2
2.1
0.67
50
51
140
160
nmd
80
1.5
3.8
0.077
12
9.5
0.16
33
22
0.075
14
9.2
58
<0.01
<0.01
c
p-MeOPh
p-CNPh
p-NO2Ph
PhCtC
PhCHdCH
Ph
58
87
c
c
81
110
nmd
51
1.1
2.3
70
nmd
nmd
nmd
16
nmd
17
<0.01
<0.01
<0.01
0.011
<0.01
<0.01
0.44
0.94
0.54
1.1
O
PhC
a Amine concentrations in the range of 0.14-1.0 M were used for the quenching experiments. b Amine concentrations in the range of 0.05-1.2
M were used for the quenching experiments with Ia, 0.05-3.0 mM with IIa-VIa, and 1-30 mM with IXa and Xa. c Decays were not first-order.
d Not measured.
to the carbonyl bond of the ketene with the barrier for this
process being smaller by 9-13 kcal/mol than that for the
addition to the CdC bond. Theory also predicts53 that the
activation energy for the addition process is much lower for
the ammonia dimer compared with that calculated for a
monomer. Our observation of the linear dependence of the rate
constant for the addition on an amine indicates that in solution,
within the concentration range studied, the transition state for
the monomeric amine addition is solvent stabilized and the
reaction does not require co-association of another molecule of
amine to proceed. Theory also predicts53 a much higher stability
(by ca. 30 kcal/mol) of the amide compared with the amide
enol. If the 1680-cm-1 band observed in the reaction of phenyl
ketene with diethylamine corresponds to the amide enol 4, our
experimental data on its conversion to amide 5 corroborate
nicely with this calculation. The observed linear dependence
of the rate constant of the conversion with diethylamine
concentration indicates that this step is base-catalyzed.
to be in general somewhat more reactive. Recent detailed
studies of the kinetics of amine addition reactions to diphenyl
ketene in aqueous solutions showed a very similar reactivity
pattern.7 For example, for structurally relevant amines, the
values of 3.52 × 105 M-1 s-1 for n-butylamine, 1.94 × 105
M-1 s-1 for methylamine, 1.03 × 105 M-1 s-1 for piperidine,
1.91 × 105 M-1 s-1 for pyrrolidine, and 4.01 × 102 M-1 s-1
for N-methylpiperidine were reported.7 These rate constants
are ca. 2 orders of magnitude lower than those observed in our
studies. The decrease in reactivity of amine nucleophiles in
aqueous solutions has been reported before, for example, in the
reaction of amines with diarylmethyl cations.54 The second-
order rate constant for quenching by n-propylamine was found
to be 2 orders of magnitude lower in water than in acetonitrile.54
This was explained by a reaction mechanism in which the
hydrated amine complex, RNH2‚‚‚HOH, is unreactive and,
hence, an equilibrium desolvation must precede the quenching
reaction.
A compilation of the kinetic data obtained in this study for
ketene quenching by amines is given in Table 2. Our data show
a general trend, indicating a much higher reactivity (ca. 3 orders
of magnitude difference in the corresponding rate constants) of
secondary amines compared with that of tertiary amines.
Secondary diethylamine shows reactivity similar to those
observed for primary amines, while secondary piperidine seems
The observed trend of reactivities suggests7 that amine
addition to ketenes proceeds through direct nucleophilic attack
and that the basic strength of the amines that in general follows
the order of tertiary > secondary > primary is not a dominant
factor controlling the reactivity. Other considerations such as
(54) McClelland, R. A.; Kanagasabapathy, V. M.; Banait, N. S.; Steenken,
S. J. Am. Chem. Soc. 1992, 114, 1816-1823.

ReactiVity of Ketenes in Acetonitrile
J. Am. Chem. Soc., Vol. 120, No. 8, 1998 1833
steric hindrance have to be taken into account. The nucleophilic
attack occurs on the LUMO in the ketene plane.1 Thus, ketene
substituents, which also lie in the plane, could significantly
hinder a nucleophile approach as would bulky groups on the
attacking amine. Such steric effects are likely contributing to
the much lower reactivity of diphenyl ketene compared with
monosubstituted ketenes. They are also likely dominant in
lowering the reactivity of the tertiary amines. A small decrease
of the reactivities of isopropyl- and benzylamines compared to
n-butylamine as well as a gentle increase of the reactivity of
piperidine compared with diethylamine can possibly be at-
tributed to steric effects. Nucleophilic addition to ketenes is
likely to proceed via a highly polar transition state. The ability
of the system to stabilize developing charges would be important
in determining the energetics and kinetics of the reaction. Such
a property of the phenyl group is most likely responsible for
the much greater reactivity of phenyl ketene compared with that
of tert-butyl ketene. Although limited, our data on para-
substituted phenyl ketenes provide support for the negative
charge development on the ketene moiety in the transition state,
with electron-withdrawing substituents accelerating and electron-
releasing substituents slowing down the process. Hammett
treatment with the σp parameters yields F values of 0.43
(correlation coefficient, cc ) 0.994), 0.50 (cc ) 0.973), 0.88
(cc ) 0.991), 0.63 (cc ) 0.999), and 0.83 (cc ) 0.999) for
diethylamine, piperidine, and isopropyl-, n-butyl-, and benzyl-
amines, respectively. It is worth noting that similar treatment
of the kinetic data for the hydration of para-substituted phenyl
ketenes in aqueous solution yielded a F value of 1.19,55
indicating a possibly higher polarity of the transition state for
hydration than that for amine attack. With our set of data on
amine additions and those known for the hydration reactions,56
it is possible to correlate the trends of reactivities for these two
processes for the entire group of ketenes included in our studies.
Across the group, for all ketenes, the rectivities with amine and
water seem to follow the same general trend, although a larger
spread of the relative rate constants is observed for the hydration
reaction. The latter may reflect the higher polarity of the
aqueous medium and/or the transition state of the reaction.
Figure 7. Schematic diagram of the laser flash photolysis system with
time-resolved infrared detection (LFP-TRIR).
solution was extracted with diethyl ether and dried over magnesium
sulfate. Removal of the solvent yielded a solid, which was recrystallized
3 times from hexane (mp 54-56 °C). This was identified by GC-
MS as R-formyl-p-methoxyacetophenone (yield 56%). A solution of
sodium ethoxide was prepared by adding sodium metal (0.15 g, 6.52
mmol) to ethanol (20 mL) in a round-bottom flask fitted with a reflux
condenser. To this solution, R-formyl-p-methoxyacetophenone (1.0 g,
5.62 mmol) was added, and the solution was stirred for 30 min, resulting
in the precipitation of a white solid. Mesyl azide60 (0.75 g, 6.2 mmol)
was then added and the mixture stirred at room temperature in the dark
for 2 h. The completion of the reaction was noted by complete
dissolution of the precipitate and the formation of a deep red solution.
To the reaction mixture was added 10% aqueous NaOH (50 mL),
followed by extraction with diethyl ether (3 × 20 mL). The combined
ethereal extracts were dried, and the ether was removed, yielding a
brown oil. This was dissolved in a mixture of hexane and diethyl ether
to yield a yellow solution, which was filtered and rotary evaporated.
The resulting solid was recrystallized 3 times from 50/50 benzene/
hexane, yielding yellow crystalline plates (yield 20%). HRMS Calcd:
176.1128. Found: 176.0501. Mp: 85-86 °C. IR (CH3CN solu-
tion): 2108 (CN2), 1616 cm-1 (CdO). 1H NMR (CDCl3): δ (ppm)
3.86 (s, 3 H, OCH3), 5.85 (s, 1 H, N2CH), 6.93 (d, 2 H, aromatic),
7.74 (d, 2 H, aromatic). 13C NMR (CDCl3): 185.38, 163.48, 129.72,
114.01, 55.68, 53.68.
Time-Resolved Infrared (TRIR) Measurements. Figure 7 shows
a schematic diagram of the TRIR system. Previously, only details of
a much earlier version of the system61,62 or brief descriptions of the
current system18 have been given; full details of the current system are
therefore presented herein. Solutions (prepared to give an absorbance
of 0.3 at 308 nm) were flowed through a CaF2 cell, the path length of
which was adjustable from 0.5 to 6 mm. The output of a Lumonics
Excimer-500 laser (XeCl, 308 nm; 10-ns pulse width; 10-Hz repetition
rate) was focused onto the sample cell from one side. The absorption
by the solution of these excitation pulses generated the transients of
interest. The output of a Mutek Model MPS-1000 diode laser was
passed through the solution cell from the opposite side as the excitation
pulses and focused onto a CdHgTe semiconductor IR detector, mounted
on a liquid N2 Dewar. This detector output a voltage directly
proportional to the incident IR intensity and had a response time of
ca. 250 ns. The diode laser contained ports for four different diodes,
the temperature of which were controlled in the range of 20-90 K by
a liquid He cryostat. By varying the temperature of and current passing
through the diode, a tuning range of approximately 200 cm-1 was
obtained for each diode, yielding a total range of the system of 800
cm-1. The diodes currently installed gave a nearly continuous tuning
range from 1500 to 2300 cm-1. The spectral width of the diode laser
output was less than 1 cm-1. The temperature and current were set
manually; the frequency of the resulting output was determined using
a Mutek Model MDS-1200 monochromator. The cw diode laser output
was chopped to give a square-wave IR probe of 10-Hz frequency and
500-ms duration. This was necessary both to provide the timing trigger
for the excimer laser and the electronics and to prevent saturation of
Experimental Section
Materials. Acetonitrile (BDH Omnisolve), piperidine (BDH),
diethylamine (BDH), isopropylamine (Aldrich), n-butylamine (Aldrich),
benzylamine (Aldrich), 2H-pyran-2-one (VIII, Aldrich), triethylamine
(Fisher), and pyridine (Anachemia) were used as received. Precursors
I,9 II,6,57 IV,3,57 VII,10,57 IX,6 and X58 were kindly supplied by Prof. A.
J. Kresge and used without further purification. Precursor VI10 was
kindly supplied by Prof. T. T. Tidwell and used without further
purification. Precursor V59 was kindly supplied by Prof. M. H. Liu
and used without further purification.
r-Diazo-p-methoxyacetophenone (III). To a stirred suspension of
sodium hydroxide (1.0 g, 41.7 mmol) in anhydrous ether (100 mL)
was added p-methoxyacetophenone (5.0 g, 33.3 mmol) and an excess
of ethyl formate (12.0 g, 162 mmol). After GC-MS analysis revealed
complete consumption of the p-methoxyacetophenone (ca. 2 h), ethanol
(5 mL) was added, followed by 0.1 N aqueous HCl (100 mL). The
(55) Bothe, E.; Meier, H.; Schulte-Frohlinde, D.; von Sonntag, C. Angew.
Chem., Int. Ed. Engl. 1976, 15, 380-381.
(56) The corresponding rate constants in water are for Ia, 14.6 s-1; for
IIa, 4.77 × 103 s-1; for IIIa, 4.5 × 103 s-1; for IVa, 25.6 × 103 s-1; for Va,
49.5 × 103 s-1; for VIa, 71.6 × 103 s-1; for VIIa, 5.76 × 103 s-1; for IXa,
0.275 × 103 s-1; and for Xa, 6.29 × 103 s-1. See ref 1b, pp 578-580.
(57) Danheiser, R. L.; Miller, R. F.; Brisbois, R. G.; Park, S. Z. J. Org.
Chem. 1990, 55, 1959-1964.
(60) Boyer, J. H.; Mack, C. H.; Goebel, W.; Morgan, L. R., Jr. J. Org.
Chem. 1958, 23, 1051-1053. Taber, D. F.; Ruckle, R. E., Jr.; Hennessy,
M. J. J. Org. Chem. 1986, 51, 4077-4078.
(61) Rayner, D. M.; Nazran, A. S.; Drouin, M.; Hackett, P. A. J. Phys.
Chem. 1986, 90, 2882-2888.
(58) Popic, V. V.; Korneev, S. M.; Nikolaev, V. A.; Korobitsyna, I. K.
Synthesis 1991, 195-198.
(59) Bradley, W.; Schwarzenbach, G. J. Chem. Soc. 1928, 2904-2912.
(62) Ishikawa, Y.; Hackett, P. A.; Rayner, D. M. J. Phys. Chem. 1988,
92, 3863-3869.

1834 J. Am. Chem. Soc., Vol. 120, No. 8, 1998
Wagner et al.
the IR detector. The synchronous relative timing of the excimer laser
and electronics was controlled using the chopper signal by a Stanford
Model DG535 delay generator.
If no shock wave was observed, then the kinetic traces were corrected
by subtracting the trace measured with the excimer beam blocked from
hitting the sample. This is preferable, as the detector signal was not
flat, due to detector saturation, and the slope of this signal was
dependent on the signal strength. Spectra were obtained by measuring
the individual kinetic traces at 2-10-cm-1 increments throughout the
region of interest and plotting ∆ absorbance Vs wavenumber for a fixed
time after the UV pulse.
The kinetic traces at a particular IR frequency were obtained by
measuring the IR intensity (as a voltage at the detector) passing through
the sample before, during, and after absorption of the UV laser pulse.
The detector signal was observed on a Tektronix Model 7603
oscilloscope as voltage Vs time. This oscilloscope trace was then
digitized using a DSP Technology Model 6001 transient digitizer,
controlled by a OGIVAR 386 computer. Absorption of the excitation
pulse sometimes caused a shock wave in the kinetic traces, which is
dependent on the solution and path length. In such cases, all kinetic
traces were corrected after data acquisition by subtracting the shock
wave measured at an IR frequency at which no transient was observed.
Acknowledgment. We are grateful to Prof. T. T. Tidwell
for valuable discussions and sharing the results prior to
publication and to Prof. A. J. Kresge for valuable comments.
JA964155B