Toluene carbonylation catalyzed by mixtures of water and trifluoromethanesulfonic acid

Article abstract of DOI:10.1006/jcat.1999.2592

The effect of water in trifluoromethanesulfonic acid (CF₃SO₃H or triflic acid) on the carbonylation of toluene is reported at room temperature (297 K) and constant CO partial pressure (7.68 MPa = 1100 psig). A method has been developed to measure low concentrations of water in triflic acid by ¹H NMR. The NMR signal increased in value linearly with increasing mole percent of water in the superacid. The first-order rate constant for toluene conversion decreases with increasing water content of the triflic acid when all other experimental variables are held constant. The triflic acid hemihydrate [(CF₃SO₃H)₂:H₂O] is inactive for the carbonylation reaction of toluene.

Full text of DOI:10.1006/jcat.1999.2592

Journal of Catalysis 186, 345–352 (1999)
Article ID jcat.1999.2592, available online at http://www.idealibrary.com on
Toluene Carbonylation Catalyzed by Mixtures of Water
and Trifluoromethanesulfonic Acid
Bo-Qing Xu, ,1 Dhiraj S. Sood, ,2 Leslie T. Gelbaum,† and Mark G. White
School of Chemical Engineering and †School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30332
Received January 27, 1999; revised May 27, 1999; accepted June 7, 1999
the effect of small amounts of water on the carbonylation
chemistry since it is known that superacids are hygroscopic.
Indeed, Olah and co-workers found that the addition of
only small amounts of water to triflic acid (<10 mol% or
1.3 wt% ) changed the acidity and product selectivity for the
isobutane–isobutylene alkylation (4). Thus, one purpose of
this study was to document the effects on toluene carbony-
lation of adding water to triflic acid.
Measuring the water content in strong acids is a daunting
task since triflic acid is hygroscopic and the density of the
acid is much higher than that of water; therefore accurate
measurements of a small amount of water in the acid could
be difficult to obtain using traditional methods. For exam-
ple, the conventional titration method with a base requires
a large amount of sample for a reliable measurement when
The effect of water in trifluoromethanesulfonic acid (CF3SO3H
or triflic acid) on the carbonylation of toluene is reported at room
temperature (297 K) and constant CO partial pressure (7.68 MPa =
1100 psig). A method has been developed to measure low concen-
1
trations of water in triflic acid by H NMR. The NMR signal in-
creased in value linearly with increasing mole percent of water in
the superacid. The first-order rate constant for toluene conversion
decreases with increasing water content of the triflic acid when
all other experimental variables are held constant. The triflic acid
hemihydrate [(CF3SO3H)2 :H2O] is inactive for the carbonylation
c
reaction of toluene.
1999 Academic Press
INTRODUCTION
The carbonylation of aryl substrates is an important syn- the water content is very low ( 0.1 wt% ), and the sam-
thesis reaction for aryl aldehydes. Earlier work by Olah ple is destroyed by the analysis. A direct approach such as
et al. shows that the carbonylation of aryl substrates pro- GC/MS may not be possible since wet triflic acid will readily
ceedssmoothlyin strongacidsat room temperature (1a;1c). corrode stainless steel and will react with most stationary
One reaction of particular interest is the carbonylation of phases. Therefore, an additional purpose of this study was
toluene by the action of trifluoromethanesulfonic acid, also to develop an analytical technique to measure accurately
known as triflic acid (CF3SO3H). This strong acid catalyzes the water content of the acid well below the 1 wt% level. It
the production of p-tolualdehyde in high yields (>95% is desired that this technique be nondestructive and require
isomer selectivity) at high conversions of toluene ( 65% ) only very small amounts of the sample.
when the acid/substrate ratio is greater than 2 (2). After the
Nuclear magnetic resonance of protons is a well-
reaction, the product mixture must be treated with ice/water established technique for determining quantitatively the
to release the products from the acid catalyst (1, 2). The an- number of protons in a specific environment. The NMR
hydroustriflicacid can be recovered from aqueoussolutions technique is a nondestructive method for which the sample
by distilling in a Vigreux column under vacuum (15 Torr need not come in contact with the metal parts of the spec-
and 85 C) (3a) or by an alternative method patented by trometer; thus, corrosion of the instrument by the “wet”
Halder-Topsoe. By this technology, Hommeltoft reports a acid is avoided. In the present report, we will show that
process for drying triflic acid using a base such as a tri- the sensitivity of the NMR technique permits an accurate
alkyl amine (3b). The active agent for the dehydration of quantitation of water in wet triflic acid with only a small
the acid is either the trialkyl amine or its salt with triflic amount of sample.
acid. Recycling of the acid catalyst may permit the practi-
cal use of carbonylation chemistry. It is surprising, however,
that little information appears in the literature to describe
EXPERIMENTAL PROCEDURES
Chemicals
1 Present address: Department of Chemistry, Tsinghua University,
Trifluoromethanesulfonic acid was obtained from Alfa
Aesar with a water content claimed to be less than 1 mol% .
Beijing 100084 China.
2 Present address: Albemarle Corp., Baton Rouge, LA 70805.
345
0021-9517/99 $30.00
c
Copyright
1999 by Academic Press
All rights of reproduction in any form reserved.

346
XU ET AL.
We selected a portion of this acid to be purified by distil- NMR signal of each triflic acid sample was divided by the
lation under a partial vacuum (14–14.5 Torr; 1.94 kPa) for signal of the internal water standard. Acid/water standards
which the distillate was condensed at 63.5 C (336.6 K). The were prepared for 10 mixtures showing water concentra-
distilled acid was a colorless, clear liquid and this material tions between 0.1 and 5 wt% (0.8–30.5 mol% ). Five of these
was used in preparing the water/acid standards (vide infra). mixtures were prepared with water concentrations less than
Deionized water was used in preparing the standards.
1 wt% (7.7 mol% ). Two samples of triflic acid were pre-
Reagent-grade toluene, obtained from Aldrich, was used pared by distillation of the triflic acid obtained from the
˚
after drying over 3-A molecular sieves. High-purity car- Alfa Aesar (vide supra).
bon monoxide, obtained from Matheson Chemicals, was
Reactions. The reactionswere completed in a stirred au-
used without further purification. Standards of o-, m-, and
p-tolualdehyde were obtained from Aldrich for calibrating
the GC/MS.
toclave, 50 cm3, obtained from Autoclave Engineers (Erie,
PA). The autoclave was rated to 7000 psig (48.43 MPa)
at 400 C (673 K) and it was fabricated from Hastelloy-C
to resist the corrosive effects of the water/acid mixtures.
The organic products were analyzed in a HP-5890 Series II
Plus GC coupled to a HP-5972 mass spectrometer. High-
purity helium (99.995% , Air Products and Chemicals) was
the carrier gas and the partitioning column was a cross-
linked polyethylene glycol column obtained from Hewlett
Packard (Part No. 19019 N-233, 30 m long, 0.25 mm i.d.,
0.5-mm film thickness).
Apparatus and Standards
NMR calibration. We used a Bruker DRX 500, 500-
MHz instrument housed in the Georgia Tech NMR Center.
An internal standard was prepared by placing deionized
water in a sealed capillary tube. The capillary tubes were
obtained from Kimax (Catalog No. 34507) and had dimen-
sions of 0.8–1.1 mm by 90 mm. A number of these capillary
tubes containing the water were examined in separate
experiments to determine the response of the instrument
for the 1H NMR. Ten of these samples that showed relative,
integrated intensities between 0.9990 and 1.0004 for the
proton resonance were selected. These results indicate
that each internal standard presented the same amount of
water to the spectrometer. One internal standard was put
in each NMR tube containing the water/triflic acid mixture
(Fig. 1). The NMR tubes were obtained from Wilmad
(5 mm 177.8 mm) and were sealed with a Teflon cap. The
The reaction conditions for all of the studies were as fol-
lows: CO pressure 7.68 MPa; acid/toluene ratio 2 mol/1mol;
temperature 297 K; duration of reaction 30 min. At the
completion of the reaction, the CO pressure was reduced to
1atm, the vesselwaspurged with He, and then ice wasadded
to the reaction mixture to dilute the acid. The organic and
aqueous layers were separated and the organic layer was
neutralized with sodium bicarbonate until no more CO2
was observed. The liquid was separated from the solid by
filtration and this liquid was diluted in diethyl ether.
To examine the effects of water upon reactivity, we mixed
0.15 mol of freshly distilled triflic acid with known amounts
of water varying between 0 and 0.15 mol to produce the
desired ratio of acid/water. A small amount of each mixture
was examined by 1H NMR to determine the water content.
Each mixture was added to 0.077 mol of toluene in the
autoclave, the contents were stirred for several minutes,
and the reactor was closed and purged with He. Then, CO
was added to the desired pressure. To examine the effect of
acid/substrate ratio in anhydrous conditions, we added to
0.077 mol of substrate varying amounts of triflic acid (0.038
to 0.15 mol). The reaction conditions for these tests were
the same as those for the mixtures of acid and water.
RESULTS AND DISCUSSION
NMR Determination of Water Content in Wet Triflic Acid
Figure 2 shows 1H NMR spectra of the distilled CF3SO3H
(Fig. 2A), a CF3SO3H/H2O mixture with 7.60 mol% water
(Fig. 2B), and the hemihydrate of the acid [(CF3SO3H)2 :
H2O] (Fig. 2C). The signal at 4.670 ppm in each spectrum
is the internal water standard while the signal at >10 ppm
is the proton resonance from the mixture CF3SO3H/H2O.
FIG. 1. Schematic diagram of NMR tube with capillary containing
standard.

FIG. 2. (A) 1H NMR of distilled triflic acid and standard. Peak at 4.669 ppm is the water standard; peak at 10.5142 ppm is the triflic acid. Numbers
1
above the peaks are the integrated peak areas. (B) H NMR of triflic acid/water (7.6 mol% ) and standard. Peak at 4.670 ppm is the water standard;
1
peak at 11.28 ppm is the triflic acid/water mix. Numbers above the peaks are the integrated peak areas. (C) H NMR of triflic acid hemihydrate and
standard. Peak at 4.670 ppm is the water standard; peak at 11.57 ppm is the triflic acid/water mix. Numbers below the peaks are the integrated peak
1
areas. (D) H NMR of triflic acid in deuterated chloroform. Peak at 10.5246 ppm is the distilled triflic acid.

348
XU ET AL.
FIG. 2—Continued

TOLUENE CARBONYLATION
349
be unwise to use these correlations as a means to inter-
pret NMR data from other instruments without calibrating
these instruments using the techniques we described above.
Equation [1] was correlated on the range of the ratio of
NMR signals between 1.80 and 3.40 and thus should not be
extrapolated beyond this range. The present method seems
not to be applicable to samples of the CF3SO3H/H2O sys-
tem with water content >33.3 mol% .
The phase diagram for triflic acid and water shows that
triflic acid hydrates (CF3SO3HnH2O, n = 0.5, 1, 2, 4, and 5)
have been identified (5, 6, 7). One of these hydrates is the
hemihydrate, which shows a water content of 33.3 mol% .
The phase diagram does not ensure that these hydrates exist
as discrete molecules under the conditions of the analysis
or the reaction. However, we can imagine that as water is
added to triflic acid, hydronium ions are created in a one-
to-one ratio to the moles of water added. The integrated
NMR signal will double when the number of hydrogens is
doubled. This result is observed when one considers the
absolute values of the NMR signal for the anhydrous acid
(signal 1.70; 1 equivalent of protons: CF3SO3H) and for
the mixture containing 33.3 mol% water {signal 3.50, two
equivalents of protons: 1/2[(CF3SO3H)2 :H2O]}. Thus, the
ratio of the 1H NMR peaks should increase in a linear fash-
ion as shown in Fig. 3.
FIG. 3. 1H-NMR signal as a function of mole fraction water.
These assignments are confirmed with separate measure-
ments of the pure acid (10.53 ppm, Fig. 2D) diluted in
deuterated chloroform (1% sample in chloroform) to cal-
ibrate the chemical shift. A total of 12 samples were mea-
sured in this manner and we recorded for each the ratio of
the integrated NMR signals for the peaks at >10 ppm to
that for the internal standard at 4.67 ppm.
When the integrated intensities of the samples are di-
vided by the signals of the internal standard in each capil-
lary tube, we found that this ratio of integrated NMR signals
increased linearly as a function of mole percent of water for
the 12samplesin the range of0–33mol% water (Fig. 3). This
correlation may be inverted to give the desired target vari-
able (mole percent of water) as a function of the measured
variable (ratio of NMR integrated peak areas : unknown to
internal standard).
Toluene Carbonylation Studies
The conversion of toluene after 30 min of reaction at
room temperature decreases as the mol percent of water in
the original acid increases (Fig. 4). The conversion for the
anhydrous acid was 20.4% under standard conditions and
the yield to p-tolualdyde was 18.5% in this base case. When
just 1.5 mol% water was added to the acid, the conversion of
toluene decreased to 14% and the conversion was zero for
a water concentration of 33 mol% (i.e., the hemihydrate is
not active). The response of the data suggested a nonlinear
behavior in the conversion with water content of the acid;
the fastest decrease in the conversion appeared with the
addition of the first 2 mol% water.
mol% H2O = 21.0 (NMR ratio) 39.2
[1]
As an example of the application of this technique, we ex-
amined the sample as received from Alfa Aesar and found
that it contained 0.39 mol% water (0.05 wt% ). This de-
termination agrees well with the water content claimed
by the vendor (<1 mol% ). Thus, the present technique
provides a reliable estimate of the water concentration in
aqueous mixtures of triflic acid when the water content is
33 mol% .
Several comments are appropriate to the use of these
equations. First, the ratio of integrated NMR signals of
CF3SO3H/H2O depends on the size (diameter) of both the
NMR tube and the capillary tube since the relative amount
of the sample to the “internal” water standard varies with
the size of the tubes. Moreover, both the sample tube
and the capillary containing the internal standard must be
longer than the coil length of the NMR instrument. Finally,
the capillary tube should be placed in the NMR tube so that
the capillary is completely in the coil area. Thus, it would
FIG. 4. Conversion of toluene versus mole percent water.

350
XU ET AL.
ter or (ii) under anhydrous conditions when xHOTf < 0.67.
The rate constants appear to describe a single curve when
HOTf > 0.67. For toluene carbonylation, very high acid
x
strength is required to catalyze the desired reaction to the
desired p-tolualdehyde (1).
This result can be understood in terms of the reaction
mechanism given below:
CO(gas) ⇔ CO(dissolved in acid/hydrocarbon):
x[CO] = hPCO [3]
CO(dissolved) + HOTf ⇔ COH+OTf : x[COH
+OTf
]
CO HOTf COH+OTf
x[CO]x[HOTf] [4]
= K2
/
Toluene + (COH+OTf ) → tolualdehyde–H+ + OTf :
R = k3aTOLaCOH+OTf [5]
FIG. 5. Yields versus toluene conversion.
Tolualdehyde–H+OTf + H2O
→ tolualdehyde + H3O+ + OTf . [6]
The yields of o- and p-tolualdehyde and heavy hydrocar-
bons (i.e., di- and tritolymethane) also change with water
content of the acid. The o- and p-tolualdehyde decreased
in a manner similar to the toluene conversion with increas-
ing water content; however, these yields were very small
as were the yields of di- and tritolymethane. We plotted
the yields versus toluene conversion (Fig. 5) to show the
correlation of yields of tolualdehyde with conversion.
To understand better the effect of water upon the chem-
istry, data of toluene conversion, f, were converted into
rate constants, k, using the integrated rate equation for
an isothermal batch reactor under first-order kinetics for
toluene consumption
Carbon monoxide must be dissolved into the reaction
mixture for the reaction to occur. We model this step using
Henry’s law for which the mole fraction of dissolved CO
is linearly related to the partial pressure of CO gas by the
constant, h, in Eq. [3]. The constant, h, is a function of tem-
perature and the acidity of the liquid (9). The protonation
of CO by a strong acid to form the formyl ion [HCO+] is
necessary for the reaction to proceed (1, 4). We show this
step to be a reversible reaction characterized by the equilib-
rium constant K2 in Eq. [4]. Here we use the solution model
to express activities in terms of mole fractions and activ-
ity coefficients. We assumed that
were functions of temperature and acidity of the reaction
, HOTf, and
CO
COHOTf
ln(1 f )/t = k,
[2]
where k isthe observed first-order rate constant. The useful-
ness of this kinetic rate law is supported by the recent work
of Sood (2), who showed that the carbonylation reaction
followed first-order kinetics in toluene when the conver-
sion was less than 25% . Nonlinear kinetics in toluene was
observed for higher conversion as a result of secondary re-
actions between the product aldehydes and toluene to form
di- and tritolymethane.
This rate constant increased with increasing values of the
triflic acid mole fraction: xHOTf. Under anhydrous condi-
tions, this effect may be exploited to achieve nearly total
conversion of substrate to p-tolualdehyde at room temper-
ature and constant CO partial pressure (7.68 MPa) (8). In
order to understand better the effect of water upon the
carbonylation chemistry, Fig. 6 shows data for two sep-
arate tests in which the mole fraction of triflic acid was
changed by (i) varying the amount of the anhydrous acid
(squares) or (ii) adding varying amounts of water to con-
stant amounts of the anhydrous acid (circles). Apparently,
the response of the rate constants was different for chang-
FIG. 6. Observed rate constant as a function of mole fraction of triflic
acid. (a) Squares are results under anhydrous conditions. (b) Circles are
results for mixtures of acid/water. Reaction conditions: T = 297 K; p =
7.68 MPa; 0.077 mol of toluene; variable amounts of water added to reac-
ing the acid/substrate ratio either (i) by the addition of wa- tion mixture, initially.

TOLUENE CARBONYLATION
351
mixture; whereas K2 was only a function of temperature. against the mole fraction of triflic acid remaining unreacted
The rate-determining step was assumed to be the elec- with water.
trophilicsubstitution reaction described byEq. [5]and char-
The rate constant data for the acid/water mixtures (cir-
acterized by rate constant k3. For this rate equation, we use cles, Fig. 6) fall below the curve for the effect of acid/
the activities of toluene and COH+OTf to express the ef- substrate ratio under anhydrous conditions (squares, Fig. 6)
fect of concentration upon rate. Water added to the reac- when xHOTf < 0.67. Thus, we conclude that the effect of
tion mixture released the tolualdehyde by Eq. [6], which adding water is more than just converting triflic acid to
has been practiced in the literature (1) and was recently inert triflic acid monohydrate. These data for the anhy-
confirmed by Sood (2).
Considering the fact that the reaction rate of toluene is described by two straight lines, with a higher slope for
carbonylation increases with increasing values of the acid/ HOTf > 0.67. This change in slope suggests that the factor
substrate ratio, which is well documented in the literature (hPCO K2 COHOTf) increases with increas-
drous acid (squares, Fig. 6) show that the rate constant
x
/
toluene CO HOTf
(1, 2), we use an empirical relationship for the reaction rate: ing triflic acid mole fraction. The partial pressure of CO,
PCO, was constant in all of these tests. We expect that the
R = k3xtoluene g(mol HOTf/mol toluene).
[7]
Henry’s law constant, h, and the chemical equilibrium con-
stant for Eq. [2], K2, should not change much as a function
of xHOTf, but the activity coefficients for toluene, dissolved
This rate expression accounts for the reaction order (first
order in the toluene mole fraction, x[toluene],) and varies acc-
ording to some function, g, of the ratio moles of HOTf/
moles of toluene. Here x[toluene] is defined as [mole toluene/
(mole toluene + mole acid + mole of water + mole of
hemihydrate)].
CO, protonated CO, and HOTf may change with xHOTF
The data for the acid/water mixtures show that the factor
(hPCO K2 COHOTf) is a different function
.
/
toluene CO HOTf
of xHOTF when xHOTf is less 0.67. This result may be a con-
sequence of the lower acid strength of the mixture when
water is added to triflic acid to form the hemihydrate, a
much weaker acid. Olah et al. showed that the addition of
even a small amount of water (ca. 1 mol% ) to triflic acid re-
duced the acid strength by as much as 1 order of magnitude
when measured by the Hammett (see Ref. 10) acidity func-
tion (Ho = 14 to 13). Further addition of water to triflic
acid (ca. 15 mol% water) reduced the Ho to 11 (4), which
is an order of magnitude lower than that ascribed to 100%
H2SO4. The observation that the hemihydrate is inactive
for the carbonylation reaction suggests that the reaction is
promoted only by a superacid. The decrease in acid strength
should reduce the Henry’s law constant since Booth and El-
Fekky (8) showed that CO is much more soluble in triflic
acid (155 cm3/liter acid) than in 95% sulfuric acid (21 cm3/
liter acid) at 300 K and a CO partial pressure of 205 kPa.
Our expectation is that the CO solubility will decrease as
the strength of the mixed acid (HOTf + acid hemihydrate)
decreases. The observed reaction rate, k, should decrease
as xHOTf decreases even though the CO partial pressure re-
mains constant. We observed this result.
The mechanism given in Eqs. [3]–[5] may be used to de-
velop a rate equation assuming that Eq. [5] is the rate-
determining step. The mole fraction of COH+OTf may
be eliminated in Eq. [5] considering the equilbria (Eqs. [3]
and [4]) to give
CO HOTf COH+OTf
R = k3 toluenehPCO K2(
/
)x[HOTf]x[toluene]
= kx[toluene]
,
[8]
where k is the observed first-order rate constant shown in
Eq. [2]. Upon comparing Eqs. [7] and [8], we observe that
the function g in Eq. [7] is
toluene CO HOTf COH+OTf
k = k3g; g = hPCO K2(
Thus, we suggest that the product (hPCO K2
/
)xHOTf.
(
toluene CO
COH+OTf
/
)x[HOTf]) increases with increasing mole
HOTf
fraction HOTf. In the present case we assume that h,
,
toluene
,
HOTf, and COHOTf may be functions of xHOTf.
CO
Equations [7] and [8] will be used to interpret the data of
Fig. 6. For one set of tests, we changed the acid/substrate
ratio by mixing known amounts of water (0–0.077 mol) into
fixed amountsofacid and substrate (0.15molacid;0.077mol
substrate). Here, the water was added in just such amounts
so as to make a mixture of triflic acid and its hemihydrate.
One result of adding water to the triflic acid was to re-
duce the effective amount of the acid catalyst by a reaction
to form the triflic acid hemihydrate since the hemihydrate
was inactive for the toluene carbonylation reaction (Fig. 3).
Thus the amount of the effective acid catalyst, N[HOTf], is
SUMMARY
The present work suggests a method to measure low con-
centrations of water in triflic acid. This technique is based
on the observation that the intensity of the 1H NMR signal
for wet triflic acid is proportional to the amount of water
in the sample. This analytical tool was used to characterize
the water content of several acid/water mixtures to catalyze
N[HOTf] = N[HOTf]o 2N[H2O]
,
where N[HOTf]o is the number of moles of triflic acid prior the carbonylation of toluene at room temperature. Data of
to reacting with water and N[H2O] is the number of moles first-order rate constants showed that the water inhibited
of water added to the triflic acid. The data were correlated the carbonylation reaction (i) directly by converting HOTf

352
XU ET AL.
to an inert species, such as the inactive acid hemihydrate,
and (ii) indirectly, perhaps by lowering the CO solubility
in the toluene/acid mixture at fixed temperature and CO
partial pressure.
2. Sood, D. S., Ph.D. Thesis, Georgia Institute of Technology, Atlanta,
GA, 1998.
3. (a) Goeppart, A., Louis, B., and Sommer, J., Catal. Lett. 56, 43–48
(1998); (b) Hommeltoft, S. I., European Patent 687658A1 (29.05.95),
and U.S. Patent 5,759,357; “Process for the Recovery of a Strong Acid
from an Aqueous Solution,” 1998.
4. Olah, G. A., Batamack, P., Deffieux, D., Torok, B., Wang, Q., Molanar,
A., and Surya Prakash, G. K., Appl. Catal. A Gen. 146, 107 (1996).
5. DeLaplane, R. G., Lundgren, J. O., and Olovsson, I., Acta Crystallogr.
Sect. B 31, 2202 (1975).
6. Spencer, J. B., and Lundgren, J. O., Acta Crystallogr. Sect. B 29, 1923,
(1973).
ACKNOWLEDGMENTS
We acknowledge the Hoechst-Celanese Corp. for financial support for
this project. We thank Dr. Jeffrey C. Kenvin (Hoechst-Celanese Corp.)
for his insightful contributions to the writing of this article. We thank the
National Science Foundation [Grant BIR-9306392] for partial funding of
the NMR spectrometer.
7. DeLaplane, R. G., Lundgren, J. O., and Olovsson, I., Acta Crystallogr.
Sect. B 31, 2208 (1975).
8. Booth, B. L., and El-Fekky, T. A., J. Chem. Soc. Perkins I, 2441–2446
(1979).
REFERENCES
9. Kenvin, J. C., Schiraldi, D. A., Sood, D. S., and White, M. G., “Pro-
ceedings, 11th International Symposium on Homogeneous Catalysis,
Edinburgh, Scotland, July 1998.”
10. Hammett, L. P., and Paul, M. A., J. Am. Chem. Soc. 56, 830
(1934).
1. (a) Olah, G. A., Ohannesian, L., and Arvanaghi, M., Chem. Rev. 87,
671 (1980); (b) Booth, B. L., El-Fekky, T. A., and Noori, G. F. M.,
J. Chem. Soc. Perkins I, 181 (1980); (c) Olah, G. A., Laali, K., and
Farooq, O., J. Org. Chem. 50, 1483–1486 (1985).