Time-resolved IR studies of & α-lactones

Article abstract of DOI:10.1002/poc.789

A series of & α-lactones were generated from the reaction of phenylchlorocarbene, 4-nitrophenylchlorocarbene, diphenylcarbene, bis(4-nitrophenyl)carbene and bis(4-methoxyphenyl)carbene with carbon dioxide and examined by nanosecond time-resolved infrared (TRIR) spectroscopy. Estimated second-order rate constants for the reaction of these carbenes with carbon dioxide indicate that more nucleophilic carbenes react at faster rates, in agreement with previous low-temperature matrix experiments. Spectral TRIR data confirms that the structure of & αlactones is dependent both on substituents at the & α-carbon and on solvent polarity, with electron-donating substituents and polar solvents favoring a zwitterionic ring-opened structure as opposed to the three-membered ring oxiranone form. B3LYP calculations using self-consistent reaction field (SCRF) methods also provide support for these experimental investigations. Copyright

Full text of DOI:10.1002/poc.789

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY
J. Phys. Org. Chem. 2004; 17: 743–748
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/poc.789
Time-resolved IR studies of a-lactonesy
Brett M. Showalter and John P. Toscano*
Department of Chemistry, Johns Hopkins University, 3400 North Charles Street, Baltimore, Maryland 21218, USA
ABSTRACT: A series of ꢀ-lactones were generated from the reaction of phenylchlorocarbene, 4-nitrophenylchlor-
epoc
ocarbene, diphenylcarbene, bis(4-nitrophenyl)carbene and bis(4-methoxyphenyl)carbene with carbon dioxide and
examined by nanosecond time-resolved infrared (TRIR) spectroscopy. Estimated second-order rate constants for the
reaction of these carbenes with carbon dioxide indicate that more nucleophilic carbenes react at faster rates, in
agreement with previous low-temperature matrix experiments. Spectral TRIR data confirms that the structure of ꢀ-
lactones is dependent both on substituents at the ꢀ-carbon and on solvent polarity, with electron-donating substituents
and polar solvents favoring a zwitterionic ring-opened structure as opposed to the three-membered ring oxiranone
form. B3LYP calculations using self-consistent reaction field (SCRF) methods also provide support for these
experimental investigations. Copyright # 2004 John Wiley & Sons, Ltd.
Additional material for this paper is available in Wiley Interscience
KEYWORDS: ꢀ-lactones; oxiranones; time-resolved infrared spectroscopy; phenylchlorocarbene; 4-nitrophenylchloro-
carbene; diphenylcarbene; bis(4-nitrophenyl)carbene; bis(4-methoxyphenyl)carbene
INTRODUCTION
substituent R. In cases where R is alkyl or phenyl, ethanol
reacts with ꢀ-lactones to produce ꢀ-ethoxy acids (3),
presumably via zwitterionic form 2.10–12 When R is the
strongly electron-withdrawing trifluoromethyl group
(such as in 4), however, ethanol reacts to produce an ꢀ-
hydroxy ester (5), which presumably is formed by attack
on ring-closed form 1.5 The trifluoromethyl groups ob-
viously destabilize the dipolar ring-opened structure.
ꢀ-Lactones (oxiranones) (1) are three-membered hetero-
cyclic rings that have been invoked as intermediates in a
variety of organic transformations1 and in enzymatic
glycosyltransferase reactions.2 Most ꢀ-lactones are un-
stable at room temperature;3 those that are stable require
bulky4 or strongly electron-withdrawing substitution5 at
the ꢀ-carbon. Much of the chemistry of ꢀ-lactones has
been explained by invoking a higher energy ring-opened
zwitterionic form (2). The relative stability of ring-closed
form 1 vs ring-opened form 2 is dependent on the
substituents R. Calculations have predicted that for the
parent ꢀ-lactone (R ¼ H), ring-closed form 1 is signifi-
cantly lower in energy than ring-opened form 2.6–8
Calculations have also shown that, as expected, substitu-
tion with electron-donating R groups, and also polar
solvation, preferentially stabilize 2.9
ꢀ-Lactones readily undergo decarbonylation and also
polymerization reactions. Loss of carbon monoxide most
likely occurs through ring-closed form 1, whereas poly-
merization occurs through ring-opened form 2. Hence
structures 1 and 2 both play important roles in ꢀ-lactone
chemistry; however, in essentially all cases where ring-
opened form 2 dominates the observed reactivity, it has
not been detected spectroscopically. (We are aware of
only one instance in which zwitterion 2 has been directly
observed;13 this study will be discussed in more detail
below.) For example, the reactivity of ꢀ-lactones 6 and 7,
even at temperatures as low as ꢀ100 ꢁC, is consistent
with ring-opened form 2.10–12 However, low-temperature
—
Nucleophilic attack of alcohols (e.g. ethanol) on ꢀ-
lactones gives different regiochemistry depending on the
*Correspondence to: J. P. Toscano, Department of Chemistry, Johns
Hopkins University, 3400 North Charles Street, Baltimore, Maryland
21218, USA.
E-mail: jtoscano@jhu.edu
Contract/grant sponsor: National Science Foundation; Contract/grant
number: CHE-0209350.
Contract/grant sponsor: Petroleum Research Fund.
yDedicated to Professor Kurt Mislow on the occasion of his 80th
birthday.
(77 or 10 K) matrix IR spectroscopy of 6 (ꢁC
¼
O
—
1900 cmꢀ1 11
)
and 7 (ꢁC
¼ 1890 cmꢀ1 14,15
)
has re-
—
O
—
vealed carbonyl stretching modes at considerably high
frequencies, indicative of ring-closed form 1. These
observations have been rationalized in terms of a small
energy gap between structure 1 and its more reactive, but
Copyright # 2004 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2004; 17: 743–748

744
B. M. SHOWALTER AND J. P. TOSCANO
higher energy counterpart 2, and are consistent with
previous computational work.
Since the IR signatures of structures 1 and 2 are
expected to be different, we generated a series of
ꢀ-lactones and report here their characterization by
nanosecond time-resolved infrared (TRIR) spectroscopy.
We examined the influence of substituents (R) and
solvent on the relative stability of 1 and 2. These
experimental investigations were supported by B3LYP
calculations using self-consistent reaction field (SCRF)
methods.
RESULTS AND DISCUSSION
ꢀ-Lactones were generated from the reaction of carbenes
with carbon dioxide and examined by TRIR spectro-
scopy. (This method of ꢀ-lactone generation has been
used by Kistiakowsky and Sauer in the gas phase,16 by
Wheland and Bartlett in solution12 and by both Milligan
and Jacox17 and Sander and co-workers14,15 in low-
temperature matrices; Kovacs and Jackson reported a
comprehensive computational investigation of the reac-
tion of methylene with CO2.18) Representative TRIR data
for the reaction of phenylchlorocarbene (9), produced by
laser photolysis of phenylchlorodiazirine (8), with CO2 in
dichloromethane are shown in Figs 1 and 2. Depletion of
reactants gives rise to negative signals and the formation
of transient intermediates or products leads to positive
bands. The negative signal at 1566 cmꢀ1 is assigned to
depletion of diazirine 8, whereas the positive bands at
1586, 1910 and 1776 cmꢀ1 are assigned to carbene 9,15
ꢀ-lactone 1015 and acid chloride 11,19 respectively, based
on previous studies (Scheme 1).15,20
Figure 2. Representative kinetic traces observed following
355 nm laser photolysis of diazirine 8 (15.7 mM) in CO2-
saturated dichloromethane showing (a) decay of carbene 9,
(b) growth of ꢀ-lactone 10, (c) decay of ꢀ-lactone 10 and (d)
growth of acid chloride 11. The dotted curves are experi-
mental data; the solid curves are the calculated best fit to a
single exponential function
Figure 1. TRIR difference spectra averaged over the time-
scales indicated following 355 nm laser photolysis of diazir-
ine 8 (15.7 mM) in CO2-saturated dichloromethane
Scheme 1
Copyright # 2004 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2004; 17: 743–748

TIME-RESOLVED IR STUDIES OF ꢀ-LACTONES
745
Table 1. Estimated second-order rate constants for carbene
reactions with CO2 in dichloromethane and IR frequencies of
the corresponding ꢀ-lactones
Consistent with Scheme 1, we observe (1) depletion of
the diazirine and formation of the carbene within the time
resolution (50 ns) of our experiment, (2) subsequent
decay of the carbene (kosbd ¼ 3.0 ꢂ 105 sꢀ1) at the same
rate within experimental error ( ꢃ10%) that the ꢀ-lactone
is produced (kosbd ¼ 3.2 ꢂ 105 sꢀ1) and (3) final decay of
the ꢀ-lactone (k ¼ 2.0 ꢂ 104 sꢀ1) at the same rate as the
acid chloride product is formed (kosbd ¼ 1.8 ꢂ 104 sꢀ1).
The position of the ꢀ-lactone band at 1910 cmꢀ1 is
clearly indicative of ring-closed form 1 and in very
good agreement with the signal observed at 10 K
(1920 cmꢀ1) by Sander and co-workers.15
Carbene
kCO (l molꢀ1 sꢀ1
)
ꢀ-Lactone (cmꢀ1
)
2
9
1 ꢂ 106
1910
1920
1880
12
13
14
15
2 ꢂ 105
a
—
2 ꢂ 107
1620, 1576
—
<104
b
a A carbene band is not observed in dichloromethane owing to overlap with
a diazo precursor depletion band.
b
Carbene 15 is unreactive with CO2; no ꢀ-lactone signals are observed.
Sander and co-workers also previously examined the
effects of carbene spin state and philicity on the carbox-
ylation reaction in low-temperature matrices.15 They
concluded that the reactivity of carbenes towards CO2
is determined by their philicity (more nucleophilic car-
benes are more reactive) and that carbene spin state
interestingly has little effect. Kovacs and Jackson sug-
gested that this reactivity pattern may be explained by a
non-equilibrium surface crossing mechanism.18
Using the pseudo-first-order equation kobsd ¼ k0þ
kCO ½CO2ꢄ (where kCO is the second-order rate constant
2
2
for the reaction of carbene with CO2 and k0 is the rate of
carbene decay in the absence of CO2), we estimated
solution-phase values of kCO for phenylchlorocarbenes
2
9 and 12 and diphenylcarbenes 14 and 15 in dichloro-
methane (Table 1). (The concentration of CO2 in satu-
rated dichloromethane solution at 25 ꢁC and 1 atm is
196 mmol lꢀ1.21) The trend of these estimated second-
order order rate constants agrees with that observed in
low-temperature matrices by Sander and co-workers.
Although this series of carbenes contains both ground-
state singlets (9 and 12) and ground-state triplets (13–15),
unlike in Sander and co-workers’ low-temperature ex-
periments,14,15 we cannot rule out the thermal population
of and reaction from the higher spin state in our room
temperature studies.
Figure 3. TRIR difference spectra averaged over the time-
scales indicated following 266 nm photolysis of (a) diphe-
nyldiazomethane (3.6 mM) and (b) bis(4-methoxyphenyl)dia-
zomethane (1.1 mM) in CO2-saturated dichloromethane
zwitterion 17. In this latter case [Fig. 3(b)], the spectral
region between 1800 and 2000 cmꢀ1 is devoid of any
signal that may be attributed to the ring-closed form.
Instead, intense IR bands are detected at 1620 and
1576 cmꢀ1, consistent with B3LYP/6–31G(d) optimized
geometries (all calculations were camed out with the
Guassian suite of programs22) (Fig. 4 and Supplementary
material) and calculated frequencies (scaled by 0.96)23
(Table 2). Indeed, even geometry minimizations starting
from the ring-closed form of 17 resulted in a ring-opened
minimum. In addition, B3LYP/6–31G(d) calculations in-
dicate that the corresponding triplet-biradical 18 is
16.1 kcal molꢀ1 higher in energy than zwitterion 17.
Although carbene 15 is unreactive with CO2, TRIR
data were obtained for the reaction of carbenes 9, 12–14
with CO2; the experimentally observed C—O stretching
frequencies are given in Table 1. Spectral data for the ꢀ-
lactone derived from carbene 12 (see Supplementary
material), although significantly weaker in intensity, are
analogous to those derived from carbene 9. The TRIR
spectra for the ꢀ-lactones derived from carbenes 13 and
14, however, are dramatically different from each other
(Fig. 3). Whereas the product of CO2 reaction with 13 in
dichloromethane is clearly the ring-closed ꢀ-lactone 16
—
(ꢁC
¼ 1880 cmꢀ1, again in good agreement with that
O
—
observed by low-temperature matrix IR spectroscopy14 at
1890 cmꢀ1), the carboxylation of 14 in the same solvent
leads to a structure that appears to be best described by
Copyright # 2004 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2004; 17: 743–748

746
B. M. SHOWALTER AND J. P. TOSCANO
zwitterionic form 2. TRIR results observed following
photolysis of diphenyldiazomethane (22) in acetonitrile-
d3 (Fig. 5) contain signals that can be attributed to both
ring-closed form 16 and ring-opened form 23. SCRF
B3LYP/6–31G(d) calculations using the Onsager
model24,25 for acetonitrile (" ¼ 35.9) predict that ring-
closed form 16 is 2.6 kcal molꢀ1 (1 kcal ¼ 4.184 kJ) more
stable than ring-opened form 23. Strong IR bands are
predicted at 1886 cmꢀ1 for 16 (in good agreement with
the experimentally observed 1880 cmꢀ1 band) and at
1314 cmꢀ1 for 23 (in reasonable agreement with the
experimentally observed 1356 cmꢀ1 band). Less intense
IR bands are also predicted at 1638 and 1577 cmꢀ1 for 23.
(See Supplementary material for a complete tabulation of
computational results obtained for 16 and 23.) These
calculated bands are probably obscured by an overlap-
ping diazo depletion band at 1592 cmꢀ1 in the experi-
mental data.
Figure 4. B3LYP/6ꢀ31G(d)-calculated geometries of diphe-
nyloxiranone (16) and bis(4-methoxyphenyl)oxiranone (17).
The aryl groups have been removed for clarity
Table 2. Selected experimental and B3LYP/6ꢀ31G(d)-calcu-
lated IR frequencies (scaled by 0.96) for diphenyloxiranone
(16) and bis(4-methoxyphenyl)oxiranone (17)
Experimental
)
Calcuꢀla1ted
(cm )
ꢀ-Lactone
(cmꢀ1
Assignment
The kinetics observed for the 1880 cmꢀ1 band of 16
and the 1356 cmꢀ1 band of 23 are in good agreement with
each other, indicating that these two species are in
equilibrium (16, kgrowth ¼ 1.5 ꢂ 106 sꢀ1, kdecay ¼ 1.1 ꢂ
105 sꢀ1; 23, kgrowth ¼ 1.2 ꢂ 106 sꢀ1, kdecay ¼ 1.2 ꢂ 105 sꢀ1).
Given that we can observe an equilibrium mixture of 16
and 23, the calculated energy difference of 2.6 kcal molꢀ1
between the two is almost certainly overestimated.
TRIR data observed following photolysis of 22 in
cyclohexane (Fig. 6) contains a band easily attributed to
ring-closed form 16, but in contrast to data obtained in
acetonitrile, no signals indicating the presence of ring-
opened form 23. SCRF B3LYP/6–31G(d) calculations
using the Onsager model24,25 for cyclohexane (" ¼ 2.02)
predict a strong IR band at 1900 cmꢀ1, in excellent
agreement with the experimentally observed band. These
calculations, however, indicate that ring-opened form 23
is not a minimum on the potential energy surface, but
16
17
1880
1576
1907
1587
1597
1658
Carbonyl str.
Phenyl C–C str.
Phenyl C–C str.
Asym. C–O str.
1620
Lew et al.13 photolyzed 9-hydroxy-9-fluorenecar-
boxylic acid in hexfluoro-2-propanol (HFIP) and detected
a transient (ꢂ ꢅ 20 ms) with ꢃmax ¼ 495 nm and strong IR
bands at 1575, 1600, and 1620 cmꢀ1, which was assigned
to zwitterion 20, in excellent agreement with our mea-
sured IR bands for ring-opened form 17 (Scheme 2).
They examined the region 1988–1855 cmꢀ1 and reported
that no signals were observed for ring-closed form 21.
The difference in structure between 16 and 17 is
readily understood in terms of the addition of strongly
electron-donating substituents, but the contrast between
16 and 20 is less easily rationalized. Photolysis of 19 was
carried out in HFIP [dielectric constant (") ¼ 16.75]
whereas the TRIR experiments were carried out in
dichloromethane (" ¼ 9.08), suggesting that the ꢀ-lactone
structure is dependent on solvent polarity.
rather
a transition-state structure (one imaginary
frequency ¼ 72 cmꢀ1). The calculated frequencies for
this structure (Supplementary material) are, however,
similar to those obtained in the acetonitrile calculations.
We examined, both experimentally and computation-
ally, the role of solvent polarity on the stabilization of
Figure 5. TRIR difference spectra averaged over the
time-scales indicated following 266 nm laser photolysis
of diphenyldiazomethane (22) (9.8 mM) in CO2-saturated
acetonitrile-d3
Scheme 2
Copyright # 2004 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2004; 17: 743–748

TIME-RESOLVED IR STUDIES OF ꢀ-LACTONES
747
Figure 6. TRIR difference spectra averaged over the time-scales indicated following 266 nm laser photolysis of diphenyldia-
zomethane (22) (6.3 mM) in CO2-saturated cyclohexane and cyclohexane-d12. Since the detection of transient species is more
problematic in regions with strong solvent bands owing to the low transmission of IR light, cyclohexane-d12 was required for the
spectral region below 1600 cmꢀ1
CONCLUSIONS
at 20 Hz or the Quantronix Q-switched Nd:YAG laser
operating at 200 Hz.
TRIR spectroscopy was used to characterize a series of ꢀ-
lactone intermediates and provided direct information
concerning their structure in solution. The structure of
ꢀ-lactones, as reflected in the observed TRIR spectra, is
dependent both on the substituents at the ꢀ carbon and on
solvent polarity, with electron-donating substituents and
polar solvents favoring a zwitterionic ring-opened struc-
ture. B3LYP calculations using SCRF methods to account
for solvent polarity are consistent with these experimen-
tal conclusions. We are currently extending our TRIR
studies to the investigation of ꢀ-lactam intermediates.
Supplementary material
TRIR data for 4-nitrophenylcarbene and tabulations of
optimized Cartesian coordinates, total energies, and cal-
culated frequencies and intensities are available in Wiley
Interscience.
Acknowledgments
J.P.T. thanks the National Science Foundation (CHE-
0209350) and the donors of the Petroleum Research
Fund, administered by the American Chemical Society,
for generous support of this research. Christopher M.
Hadad is thanked for helpful discussions.
EXPERIMENTAL
Phenylchlorodiazirine (1),26 4-nitrophenylchlorodiazir-
ine,27 diphenyldiazomethane,28 bis(4-nitrophenyl)diazo-
methane,29 and bis(4-methoxyphenyl)diazomethane30
were prepared according to literature procedures.
We conducted TRIR experiments following the method
of Hamaguchi and co-workers31 as described pre-
viously.32 Briefly, the broadband output of an MoSi2 IR
source (JASCO) is crossed with excitation pulses from an
Nd:YAG laser. Changes in IR intensity are monitored by
an MCT photovoltaic IR detector (Kolmar Technologies,
KMPV11-1-J1), amplified and digitized with a Tektronix
TDS520A oscilloscope. The experiment was conducted
in the dispersive mode with a JASCO TRIR-1000 spec-
trometer. TRIR difference spectra were collected at
16 cmꢀ1 resolution using either a Continuum HPO-300
diode-pumped Nd:YAG laser (266 nm, 10 ns, 0.4 mJ;
355 nm, 12 ns, 0.6 mJ) or a Quantronix Q-switched
Nd:YAG laser (266 nm, 90 ns, 0.4 mJ; 355 nm, 90 ns,
1.5 mJ), both operating at 200 Hz. Kinetic traces were
collected using a Continuum Minilite II Nd:YAG laser
(266 nm, 5 ns, 1–4 mJ; 355 nm, 5 ns, 2–8 mJ) operating
REFERENCES
1. (a) Niwayama S, Noguchi H, Ohno M, Kobayashi S. Tetrahedron
Lett. 1992; 34: 665–668; (b) Schmittel M, von Seggern H.
Liebigs Ann. Chem. 1995; 1815–1821; (c) Strijtveen B, Kellogg
RM. Recl. Trav. Chim. Pays-Bas 1987; 106: 539–542; (d)
Adam W, Blancafort L. J. Am. Chem. Soc. 1996; 118:
4778–4787.
2. (a) Ashwell M, Guo XM, Sinnott ML. J. Am. Chem. Soc. 1992;
114: 10158–10166; (b) Chong AKJ, Pegg MS, Taylor NR,
Vonitzstein M. Eur. J. Biochem. 1992; 207: 335–343; (c) Guo
XM, Sinnott ML. Biochem. J. 1993; 296: 291–292; (d) Guo XM,
Laver WG, Vimr E, Sinnott ML. J. Am. Chem. Soc. 1994; 116:
5572–5578; (e) Horenstein BA, Bruner M. J. Am. Chem. Soc.
1996; 118: 10371–10379; (e) Horenstein BA. J. Am. Chem. Soc.
1997; 119: 1101–1107.
´
3. L’abbe G. Angew. Chem., Int. Ed. Engl. 1980; 19: 276–289.
4. Coe PL, Sellers A, Tatlow JC, Whittaker G, Fielding HC. J. Chem.
Soc., Chem. Commun. 1982; 362–363.
5. Adam W, Liu J, Rodriguez O. J. Org. Chem. 1973; 38: 2269–
2270.
Copyright # 2004 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2004; 17: 743–748

748
B. M. SHOWALTER AND J. P. TOSCANO
6. Liebman JF, Greenberg A. J. Org. Chem. 1974; 39: 123–130.
22. Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA,
Cheeseman JR, Zakrzewski VG, Montgomery JA Jr, Stratmann
RE, Burant JC, Dapprich S, Millam JM, Daniels AD, Kudin KN,
Strain MC, Farkas O, Tomasi J, Barone V, Cossi M, Cammi R,
Mennucci B, Pomelli C, Adamo C, Clifford S, Ochterski J,
Petersson GA, Ayala PY, Cui Q, Morokuma K, Malick DK,
Rabuck AD, Raghavachari K, Foresman JB, Cioslowski J,
Ortiz JV, Stefanov BB, Liu G, Liashenko A, Piskorz P,
Komaromi I, Gomperts R, Martin RL, Fox DJ, Keith T, Al-Laham
MA, Peng CY, Nanayakkara A, Gonzalez C, Challacombe M, Gill
PMW, Johnson B, Chen W, Wong MW, Andres JL, Gonzalez C,
Head-Gordon M, Replogle ES, Pople JA. Gaussian 98, Revision
A.7. Gaussian: Pittsburgh, PA, 1998.
7. Chung CSC. J. Mol. Struct. 1976; 30: 189–191.
8. Antolovic D, Shiner VJ, Davidson ER. J. Am. Chem. Soc. 1988;
110: 1375–1381.
9. (a) Firth-Clark S, Rodriquez CF, Williams IH. J. Chem. Soc.,
Perkin Trans. 2 1997; 1943–1948; (b) Ruggiero GD, Williams IH.
J. Chem. Soc., Perkin Trans. 2 2001; 733–737; (c) Buchanan JG,
Charlton MH, Mahon MF, Robinson JJ, Ruggiero GD, Williams
IH. J. Phys. Org. Chem. 2002; 15: 642–646.
¨
10. Adam W, Rucktaschel R. J. Am. Chem. Soc. 1971; 93: 557–559.
11. Chapman OL, Wojtkowski PW, Adam W, Rucktaschel RR.
¨
J. Am. Chem. Soc. 1972; 94: 1365–1367.
12. Wheland R, Bartlett PD. J. Am. Chem. Soc. 1970; 92: 6057–6058.
13. Lew CSQ, Wagner BD, Angelini MP, Lee-Ruff E, Lusztyk J,
Johnston LJ. J. Am. Chem. Soc. 1996; 118: 12066–12073.
14. Sander WW. J. Org. Chem. 1989; 54: 426–427.
15. Wierlacher S, Sander W, Liu MTH. J. Org. Chem. 1992; 57:
1051–1053.
23. Scott AP, Radom L. J. Phys. Chem. 1996; 100: 16502–16513.
24. Wong MW, Frisch MJ, Wiberg KB. J. Am. Chem. Soc. 1991; 113:
4776–4782.
25. Wong MW, Wiberg KB, Frisch M. J. Chem. Phys. 1991; 95:
8991–8998.
16. Kistiakowsky GB, Sauer K. J. Am. Chem. Soc. 1958; 80: 1066–
1071.
26. Graham WH. J. Am. Chem. Soc. 1965; 87: 4396–4397.
27. Liu MTH, Toriyama K. Can. J. Chem. 1972; 50: 3009–3016.
28. Smith LI, Howard KL. Org. Synth., Coll. Vol. 1955; 3: 351–352.
29. Just G, Wang ZY, Chan L. J. Org. Chem. 1988; 53: 1030–1033.
30. Kimse W, Kilian J, Steen S. J. Am. Chem. Soc. 1990; 112:
6399–6400.
31. (a) Iwata K, Hamaguchi H. Appl. Spectrosc. 1990; 44:
1431–1437; (b) Yuzawa T, Kato C, George MW, Hamaguchi H.
Appl. Spectrosc. 1994; 48: 684–690.
17. Milligan DE, Jacox ME. J. Chem. Phys. 1962; 36: 2911–2917.
18. Kovacs D, Jackson JE. J. Phys. Chem. A 2001; 105: 7579–7587.
19. Pouchert CJ (ed). The Aldrich Library of Infrared Spectra
(3rd edn). Aldrich Chemical: Milwaukee, WI, 1981; 1057.
20. Ganzer GA, Sheridan RS, Liu MTH. J. Am. Chem. Soc. 1986;
108: 1517–1520.
21. Fogg PGT (ed). Carbon Dioxide in Non-aqueous Solvents at
Pressures Less Than 200 kPa. Pergamon Press: Oxford, 1992;
259.
32. Wang Y, Yuzawa T, Hamaguchi H, Toscano JP. J. Am. Chem. Soc.
1999; 121: 2875–2882.
Copyright # 2004 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2004; 17: 743–748