Structure Insensitivity and Effect of Sulfur in the Reaction of Hydrodechlorination of 1,1-Dichlorotetrafluoroethane (CF3-CFCl2 ) over Pd Catalysts

Article abstract of DOI:10.1006/jcat.1998.2059

The kinetics of the hydrodechlorination of 1,1-dichlorotetrafluoroethane (CFC 114a) was studied on Pd(111), Pd(100), and a Pd foil at atmospheric pressure. The three products formed were CF₃-CFH₂ (HFC 134a), CF₃-CFClH (HCFC 124), and CF₃-CH₃ (HFC 143a) with selectivities independent of conversion. The single crystals and foil (model catalysts) were studied in an apparatus that permitted the direct transfer of samples between a high pressure cell (1 atm) and an ultrahigh vacuum chamber. The reaction rates were measured in the temperature range of 350 to 470 K. The reaction is not sensitive to the structure of the catalyst, as indicated by the similar turnover rates for all catalysts tested. The reaction is inverse first order in the reaction product HCl on all samples. Sulfur adsorbed on the Pd surface depressed the rates of formation of 134a more strongly than the rates of 124 and 143a.

Full text of DOI:10.1006/jcat.1998.2059

JOURNAL OF CATALYSIS 176, 352–357 (1998)
ARTICLE NO. CA982059
Structure Insensitivity and Effect of Sulfur in the Reaction of
Hydrodechlorination of 1,1-Dichlorotetrafluoroethane
–
(CF3 CFCl2) over Pd Catalysts
,1
ꢀ
ꢀ
ꢀ
ꢀ
F. H. Ribeiro, C. A. Gerken, G. Rupprechter, G. A. Somorjai, C. S. Kellner,† G. W. Coulston,†
L. E. Manzer,† and L. Abrams†
ꢀ
Lawrence Berkeley National Laboratory, 1 Cyclotron Rd., Berkeley, California 94720 and Department of Chemistry, University of California
at Berkeley, Berkeley, California 94720; and †E. I. DuPont Co., Experimental Station, Central Research and Development,
Corporate Catalysis Center, Wilmington, Delaware 19880
Received September 11, 1997; revised February 19, 1998; accepted February 23, 1998
and single crystals (of (111) and (100) orientation) at low
conversion (<1% ). Atmospheric pressures were utilized,
and the temperature range was 350 to 470 K. We found the
turnover rates to be the same within error for all of the
different types of palladium catalysts, indicating that
the reaction is insensitive to the catalyst surface structure.
The kinetics of the hydrodechlorination of 1,1-dichlorotetra-
fluoroethane (CFC 114a) was studied on Pd(111), Pd(100), and a
Pd foil at atmospheric pressure. The three products formed were
–
–
–
CF3 CFH2 (HFC 134a), CF3 CFClH (HCFC 124), and CF3 CH3
(HFC 143a) with selectivities independent of conversion. The sin-
gle crystals and foil (model catalysts) were studied in an apparatus
that permitted the direct transfer of samples between a high pres-
sure cell (1 atm) and an ultrahigh vacuum chamber. The reaction
rates were measured in the temperature range of 350 to 470 K. The
reaction is not sensitive to the structure of the catalyst, as indicated
by the similar turnover rates for all catalysts tested. The reaction is
inverse first order in the reaction product HCl on all samples. Sulfur
adsorbed on the Pd surface depressed the rates of formation of 134a
c
2. EXPERIMENTAL METHODS
The experiments on the foil and single crystals were car-
ried out in a high pressure (1 atm) batch reactor attached
to an ultrahigh vacuum (UHV) chamber with a base pres-
sure of 8 ꢃ 10ꢄ10 Torr (1 Torr = 133 Pa). The chamber was
equipped with low energy electron diffraction (LEED),
Auger electron spectroscopy (AES), and a mass spectrome-
ter for temperature programmed desorption (TPD) of CO.
The sample was transferred between the UHV chamber
and the high pressure reactor with a welded bellows assem-
bly capable of maintaining an UHV environment during
sample transfer. The foil was attached to a sample cart con-
taining two pins for resistive heating and two pins (chromel
and alumel) for thermocouple temperature readings so the
sample could be heated and the temperature measured
both in the UHV chamber and in the reactor. The foil was
spotwelded directly to the pins, and for the single crystal
samples, stainless steel tabs were used to spotweld the side
of the crystal to the pins. The type K thermocouple was
spotwelded to the center of the rear of the foil and to the
side of the single crystals. After reaction, due to desorption
of reaction gases, the pressure could not be maintained be-
low 10ꢄ7 Torr even though a 60 liter sꢄ1 turbomolecular
pump was connected to the transfer line. After the sample
was transferred, the pressure in the main chamber returned
to the base pressure in a few minutes.
more strongly than the rates of 124 and 143a.
ꢁ 1998 Academic Press
1. INTRODUCTION
Hydrodechlorination is a gentle chemical reaction that
removes the chlorine from a carbon-chlorine group and
replaces it with a hydrogen. The process is catalyzed by
several transition metals, and palladium is one of the best
catalysts among them. This reaction plays a key role in
the removal of chlorine from chlorofluorocarbons (CFCs)
to produce hydrofluorocarbons (HFCs) with similar phys-
ical properties that are more environmentally benign (1).
–
1,1-dichlorotetrafluoroethane, CF3 CFCl2 (also called CFC
114a) is one of the most important molecules in this family
of compounds. In this paper, we report its hydrodechlorina-
–
tion to CF3 CFH2 (also called HFC-134a) over palladium
catalysts. Its utility derives from the use of HFC-134a as a
substitute for the refrigerant CF2Cl2.
We explored the kinetics and mechanism of this reac-
tion in studies using small area (ꢂ1 cm2) palladium foils
The batch reactor had a volume of about 700 cm3. The re-
actants were circulated with a metal bellows pump (Parker-
Model MB 21) with a flow rate of about 1 liter minꢄ1. This
1 Present address: Department of Chemical Engineering, Worcester
Polytechnic Institute, Worcester, MA 01609-2280. E-mail: Fabio@wpi.edu.
352
0021-9517/98 $25.00
c
Copyright ꢁ 1998 by Academic Press
All rights of reproduction in any form reserved.

HYDRODECHLORINATION OF CFC 114a OVER Pd
353
flow rate permitted a small conversion per pass. Heat and
mass transfer limitations are of no concern in this type of background catalytic activity was not significant. Those ex-
experiment since all the active surface area is exposed. periments were run on a stainless steel foil with a feed of
Blank experiments were carried out to certify that the
The Pd polycrystalline foil used in this study was 0.1 mm 150 Torr of CFC 114a and 630 Torr of H2. At 423 K, where
thick, had a total surface area of about 0.5 cm2 and a pu- most of the experiments were run, background reactions
rity of 99.99% (Johnson Matthey). The single crystals were could not be detected, to a level at least 1000 times lower
about 1-mm thick, 0.5to 0.8cm2, and were cut from bulk ma- than the level obtained with a Pd foil.
terial (single crystal rods, Goodfellow) with 0.5ꢅ precision
and polished with standard techniques. Due to the sample
cart design, only one side of the sample could be cleaned
by Ar sputtering. The sample was cleaned by cycles of Ar
sputtering at 1000 K and 5 ꢃ 10ꢄ5 Torr, and annealing at
1000 K until no foreign peaks were found by Auger elec-
tron spectroscopy (AES). On the single crystals, the LEED
pattern was also examined before reaction to verify the sin-
gle crystal structure. After each reaction, Ar sputtering for
15 min followed by annealing at 1000 K was sufficient, as
shown by AES, to clean the surface.
The temperature-programmed desorption of CO was
used to characterize the Pd foil before and after reaction.
Carbon monoxide was dosed on the surface in various
amounts up to saturation. The sample was then heated lin-
early at a rate of 55 K sꢄ1 while the temperature and mass
28 signal were recorded with the help of a computer.
After the sample was transferred into the reactor, it was
heated to 373 K and then H2 was introduced, followed by
CFC 114a and Ar (make-up gas) to achieve 770 Torr to-
tal pressure. The sample was heated to 373 K before the
introduction of reactants because previous experience (2)
demonstrated that a combination of high hydrogen pres-
sures (e.g., 500 Torr) and low temperatures (RT) would
warp the Pd foil, probably due to the formation of palla-
dium hydride.
3. RESULTS
3.1. Measurement of Rates
We reported before that on the Pd foil the rates of forma-
tion of HFC 143a, HFC 134a, and HCFC 124 are inhibited
by the reaction product HCl (2). For all the catalysts used
in this study, the reaction order in HCl was also ꢄ1 as cal-
culated from the accumulation plots of the batch reactor
runs (details below). The immediate consequence of this
inhibition is that the turnover rate will be a strong function
of conversion (Fig. 1.A). An explanation for the procedure
used to calculate the turnover rate follows. For a batch re-
actor operating at the low conversion levels of this study
(<1% ) the rate of reaction should have been constant with
time (pseudo-zero order reaction) and straight lines should
have been obtained in Fig. 1.A. However, because the reac-
tion is inverse first order in the product HCl the rate will be
–
r = A/ꢀ, where ꢀ is the conversion and A depends on the
concentration of the reactants and is essentially constant
(low conversion). Using the rate above, the equation for
the batch reactor will then be
dꢀ/dt = B/ꢀ,
[1]
where t is time and B is a constant. Integrating this equation
will produce ꢀ2 = 2Bt. This integrated equation has a lin-
ear dependence between the square of conversion (or the
number of turnovers) and time (Fig. 1.B). The slope of the
line (2 B) can then be used in Eq. [1] to calculate the rate
at any conversion.
The chlorofluorocarbon (CFC 114a) was obtained from
DuPont Co. with only one impurity detected, possibly
–
CF3 CH2Cl. This impurity did not change its concentration
during the experiments. The CFC was degassed by freeze-
pump-thaw cycles and then stored in a glass vial filled
with a previously reduced 0.5% Pd on carbon supported
catalyst. This was done to scavenge any sulfur-containing
compounds from the reactant feed. Hydrogen (Matheson,
Prepurified) was passed through a liquid nitrogen cold trap
to remove any condensable impurities. The Ar (Liquid
Carbonic, 99.995% ) was used without further purification.
The reaction rates were measured in the temperature
range of 350 to 470 K, pressure range of 23 to 670 Torr in
The measurement of rates on the model catalysts were
made difficult by the deposition of sulfur on the sample
which caused the catalyst to deactivate. Sulfur was de-
posited whenever HCl was formed during reaction. It is
necessary to know if sulfur deposition affects the measure-
ment of rates. The first step was to find a calibration for
the amount of sulfur, determined by the ratio of Pd to sul-
fur AES peaks. This calibration was possible because sulfur
forms ordered structures on Pd single crystal surfaces.
At the end of the reaction (after 2 to 3 h), a ratio of Auger
intensities S152/Pd330 of about 0.25 was determined for the
foil, 0.24 for Pd(111) and 0.20 for Pd(100). Table 1 shows
the literature data for the structures of sulfur formed on Pd
single crystals (as determined by LEED) and correspond-
ing sulfur coverages (as determined by AES). All measure-
ments reported here are for a reaction temperature of 423 K
–
H2, 21 to 511 Torr in CF3 CFCl2 and balance Ar at a total
pressure of 770 Torr. The reaction products were analyzed
with a gas chromatograph (Hewlett-Packard 5790A) with a
flame ionization detector. The products were separated in
a 5% KrytoxTM 143AC, 60/80 Carbopack B HT 200 ꢃ 1/800
column (Supelco). The first analysis was made after 3 min
of reaction and subsequent analyses were carried out at
intervals of 18 min.

354
RIBEIRO ET AL.
patterns which were consistent with the highest-coverage
structures of sulfur known on the respective crystal faces, at
least 0.4 monolayers in each case. Specifically, the 0.5 mono-
layer c(2 ꢃ 2) structure was formed on Pd(100) (3) and the
√
√
√
√
(
7ꢃ 7)R19ꢅ structure on Pd(111). The ( 7ꢃ 7)R19ꢅ S
structure on Pd(111) has been studied by several groups us-
ing LEED, AES, and STM (4–7). General agreement exists
that this structure is formed above 400 K (which is close
to our reaction temperature) and that the geometry of this
structure is rather complex. Presumably this phase com-
prises a mixed palladium–sulfur overlayer, where some of
the sulfur atoms have been incorporated into the subsur-
face region. The agreement of the AES values measured
here approximates the values in the literature. The higher
coverage c(2 ꢃ 2) structure exhibits a lower coverage in our
AES data. One possible explanation is that the AES spectra
√
√
of the ( 7 ꢃ 7)R19ꢅ structure contains additional disor-
dered sulfur which does not appear in the LEED pattern.
Bo¨mermann et al. (6) detected disordered layers of sulfur
on Pd(111) which could account for the higher AES inten-
sities we have observed on this face. AES spectra obtained
on both crystal faces were utilized to calibrate the sulfur
coverage of the foil. On the foil, we find an AES peak ratio
S/Pd equal to 0.25, which corresponds, based on the calibra-
tion of the single crystals, to a saturation surface coverage
of 0.4–0.5 monolayers.
The sulfur contamination did not come from impurities
in the H2 or CFC 114a, as the amount of sulfur after heating
the sample in pure H2 or pure CFC 114a at reaction temper-
ature was close to background level. The amount of sulfur
on the sample correlated with the amount of HCl (amount
of conversion) produced in the reaction. We believe that
the source of sulfur is the following: H2/HCl mixtures are
good hydrodesulfurization agents (8) and any residual sul-
fur compound can be transformed into the corresponding
chloride and H2S. This last compound will eventually con-
FIG. 1. (A) Accumulation plot (number of turnovers versus time)
and (B) integrated equation plotted in linear form (accounting for
–
effect of HCl) for the reaction of CFC 114a (CF3 CFCl2) (24.9 Torr),
H
2 (50.2 Torr), Ar (697.7 Torr) at 423 K on Pd (111). The reaction prod- taminate the catalyst. The most significant breakthrough in
–
–
ucts were: HFC 134a (CF3 CFH2) (4); HCFC 124 (CF3 CFClH) (᭺); and
controlling the sulfur poisoning was to remove the rough
vacuum oil pump, which was used to pump down the reac-
tion manifold to 10ꢄ1 Torr before the turbopump could be
used. The use of the rough pump constantly added sulfur
compounds to the reactor by backstreaming. The replace-
ment of the rough pump with a sorption pump decreased
the amount of contamination to a level where initial rates
could be measured without the interference of sulfur. To
prove this point, the amount of sulfur as a function of re-
action time was studied in a series of experiments at 423 K,
260 Torr of CFC 114a and 510 Torr H2 on the Pd foil (Fig. 2).
As explained above, when the square of concentration is
plotted as a function of time (Fig. 2.B), the dependence
on HCl is removed and the new plot should be linear with
time. Any curvature in the graph will be due to sulfur poi-
soning. It is concluded then that at the particular conditions
used, sulfur starts to affect the rates significantly after about
–
HFC 143a (CF3 CH3) (ᮀ).
and are presumably saturation coverage. No higher cover-
age was detected by AES at higher reaction temperature,
but at low temperature (e.g., at 400 K) the values could be
50 to 80% of the saturation values which also corresponded
to a lower deactivation. The single crystals gave LEED
TABLE 1
Ordered Sulfur Structures on Pd Single Crystals
Sulfur
Nominal
coverage
Surface
structure
S
152/Pd330 eV
References
Pd(100)
Pd(111)
c(2 ꢃ 2)
0.50
0.43
0.195 ꢆ 0.016
0.239 ꢆ 0.021
(3)
(4–7)
√
√
(
7 ꢃ 7)R19ꢅ

HYDRODECHLORINATION OF CFC 114a OVER Pd
355
planes with defects. The CO TPD from the foil, however,
is very similar to the TPD of CO from Pd(111) (9, 10) and
thus the defect sites do not produce a different CO desorp-
tion peak. Thus, we were not able to quantify the defects
on the surface. Table 2 reveals a small variation for the
rates. For the product formed with the highest selectivity
(HFC 134a) the ratio between the maximum and the min-
imum rate among the various catalysts is less than three.
The widest variation is on the rate of the product formed in
the least amount (HFC 143a). The selectivities in Table 2
vary even less than the rates. Note that the comparison of
reaction selectivities is simplified in this reaction because
the selectivities are independent of conversion. It seems
that the reaction is insensitive to the structure of the Pd
catalyst.
The foil and single crystals were characterized after re-
action. The AES spectra after reaction revealed only sulfur
and Pd on the surface. The LEED revealed that the crys-
tallinity of the single crystals was not affected by the reac-
tion and that additionalspots, due to a sulfur superstructure,
were detected. As discussed before, the Pd(100) formed a
√
√
c(2 ꢃ 2) structure and Pd(111) exhibited a ( 7 ꢃ 7)R19ꢅ
structure.
It is surprising that no chlorine and fluorine are detected
on the surface of the catalyst after reaction when we know
from the reaction kinetics on the foil (2) that chlorine is the
most abundant reaction intermediate. To test why chlorine
is not observed after reaction, we first dosed Cl2 to satura-
tion (60 Langmuir) on the Pd foil through a leak valve and
verified that the excitation electron beam used in AES did
not desorb it from the foil. Next, we verified that heating
the chlorine saturated Pd surface at 373 K in 1.0 Torr of H2
for 1 min in the high pressure cell was sufficient to remove
the chlorine. Thus, when the reactor is pumped down after
reaction there is enough hydrogen to react away chlorine
adsorbed on the surface. For surface fluorine, there may be
a problem in detecting it by AES. It seems that fluorine is
easily desorbed by an electron beam, as used for excitation
in AES (11).
–
FIG. 2. (A) Amount ofHFC 134a (CF3 CFH2) (number ofturnovers)
and amount of sulfur versus time. (B) Integrated equation plotted in linear
form (accounting for effect of HCl). Reaction conditions were: CFC 114a
(CF3 CFCl2) (260 Torr); H2 (510 Torr), 423 K on a Pd foil.
–
4000 s. The curvature on the linearized graphs depended on
the amount of HCl formed; at low temperature (e.g., 350 K)
the rate of formation of HCl is slow, as is the rate of S de-
position, and no curvature due to sulfur is found. At the
highest temperature tried (470 K), the curvature appears
after 1000 s. In conclusion, the initial rates could be mea-
sured without interference from sulfur.
TABLE 2
Turnover Rate for Model Pd Catalysts
Turnover ratea/sꢄ1 (selectivityb/% )
reaction products
–
–
–
CF3 CH3
CF3 CFH2
CF3 CFClH
3.2. Influence of Catalyst Structure on Rates and Selectivity
Catalyst
(HFC-143a)
(HFC-134a)
(HCFC-124)
To study the dependence of turnover rates with the cata-
lyst structure, the turnover rates (based on the total num-
ber of Pd surface atoms) on two single crystals and a foil
were compared (Table 2). The three catalysts have dif-
ferent structures. The single crystals are the close packed
(pseudo-hexagonal) Pd(111) and square Pd(100). The Pd
foil is polycrystalline and it is probably composed of (111)
Pd foil
Pd(100)
Pd(111)
6.4 ꢃ 10ꢄ2 (3)
8.0 ꢃ 10ꢄ2 (6)
18 ꢃ 10ꢄ2 (10)
2.1 (85)
1.2 (84)
1.3 (74)
3.0 ꢃ 10ꢄ1 (12)
1.4 ꢃ 10ꢄ1 (10)
2.8 ꢃ 10ꢄ1 (16)
a
–
Rates corrected for 50 Torr CF3 CFCl2, 100 Torr H2, 0.1 Torr HCl,
423 K.
b
Selectivity is constant for conversion lower than 50% . Selectivity =
134a/(143a + 134a + 124).

356
RIBEIRO ET AL.
The activation energies are shown in Table 3 and are
around 100 kJ molꢄ1. There may be a trend of higher ac-
tivation energies as more halogens are substituted by hy-
drogen, except for the foil where the activation energy for
HFC 143a is lower than for HFC 134a.
3.3. Effect of Sulfur
Sulfur was detected by AES at the end of reaction on the
Pd foil and single crystals. It was already discussed that the
time scale of the sulfur deposition was slow enough to allow
for the measurement of initial rates and thus it did not dis-
turb our rate measurements. Sulfur deposition eventually
decreased the rates and selectivity. To measure the effect
of sulfur, the rates were measured on a sample that had its
surface covered with sulfur to saturation. This sample was
prepared by carrying out the hydrodechlorination reaction
on a clean foil, taking the sample to the UHV chamber
for characterization, and not removing the accumulated
sulfur before reaction. The final rates on the Pd foil with
sulfur for 143a, 134a, and 124 at the standard conditions
–
given in Table 2 (50 Torr CF3 CFCl2, 100 Torr H2, 0.1 Torr
HCl, 423 K) were respectively 3.7 ꢃ 10ꢄ2, 8.0 ꢃ 10ꢄ2, and
4.4 ꢃ 10ꢄ2 sꢄ1. The rates were decreased by a factor of 2 for
143a, 25 for 134a, and 7 for 124. Thus, the selectivity to 134a
(the desired product) decreased as sulfur was deposited on
the surface. Figure 3 shows the effect of sulfur on the TPD of
CO. The addition of sulfur eliminates the higher adsorption
energy CO adsorption sites on the foil and decreases the
total uptake by a factor of 13. The same decrease in the CO
adsorption energy has been observed for sulfur adsorbed
on Pd(100) (12). On Pd(111) (13), the sample prepared
by adding sulfur to saturation did not adsorb CO at room
temperature until it was annealed to create adsorption sites.
For the samples that were exposed to sulfur short of sulfur
saturation, the CO uptake was proportional to the number
of sites that could adsorb acetylene (13). Since the reaction
appears to be structure insensitive, the decrease in rate
with the addition of sulfur should be directly proportional
to the decrease in surface area. The reaction selectivities
should not be affected by the addition of sulfur either. The
rates and CO uptake decrease by an amount higher than
FIG. 3. (A) TPD of CO on Pd foil. CO exposures were 0.1, 0.2, 0.5,
0.8, 1.0, 2.0, 5.0, and 50.0 Langmuir. (B) TPD of CO on Pd foil with 0.5
monolayer of sulfur. CO exposures were 0.1, 0.5, 1.0, 5.0, 10.0, and 50.0
Langmuir. Heating rate 55 K sꢄ1
.
predicted by a blocking mechanism assuming one sulfur
atom blocking one Pd atom (about 50% of the surface
is covered with sulfur). Changes in the CO uptake, CO
desorption temperature, rate, and selectivity suggests that
sulfur affects the binding of reaction intermediates on the
surface.
We have also found that the deposition of sulfur does
not affect the reaction order for HCl, CFC 114a, and H2 on
the Pd foil. The same chemistry is operative when sulfur is
deposited.
TABLE 3
Apparent Activation Energy for Model Pd Catalysts
Apparent activation energy/kJ molꢄ1
reaction products
4. DISCUSSION
–
–
–
CF3 CH3
CF3 CFH2
CF3 CFClH
The main result of this paper is the small variation of rates
and selectivities as the structure of the catalyst is changed.
These results indicate that the reaction is structure insensi-
tive. Fung and Sinfelt (14), based on the small variation of
turnover rates (two orders of magnitude) across a period
in the periodic table for the hydrodechlorination of methyl
Catalyst
(HFC-143a)
(HFC-134a)
(HCFC-124)
Pd foil
Pd(100)
Pd(111)
95
125
135
110
85
100
95
80
85
Note. The error bars were about 10 kJ molꢄ1
.

HYDRODECHLORINATION OF CFC 114a OVER Pd
357
chloride, suggested that hydrodechlorination reactions may
be structure insensitive on group VIII metals.
It is surprising that sulfur modifies the reaction selectivity
when changes in the catalyst morphology will not affect the
A comparison of our rates with reaction rates reported rates. The fact that it affects the selectivities of a structure
in the literature was difficult because there are not many insensitive reaction is an indication that sulfur is changing
studies reported for CFC 114a on Pd. Gervasutti et al. (15) the binding energies of the reaction intermediates.
also studied hydrodechlorination of CFC 114a, but these
5. CONCLUSION
authors did not report a turnover rate, making a compari-
son with our rates very difficult. Karpinski et al. (16) studied
The reaction of hydrodechlorination of 1,1-dichloro-
this reaction on Pd supported on alumina, but their reac-
tetrafluoroethane is insensitive to the structure of the Pd
catalyst. This result implies that the rates of hydrodechlo-
rination will be proportional to the total Pd surface area.
Sulfur decreases the rates of all products, but it decreases
tion order in HCl appears to be zero, as it is apparent from
the contact time versus conversion plot, and not ꢄ1 order
in HCl as in this case. Thus a comparison cannot be made
in this case since the reaction chemistry seems to be very
the rate of HFC 134a more strongly than the other two
different on alumina-supported samples. Another possibil-
products, thus decreasing the reaction selectivity for this
ity for comparison is to use the rate data for similar CFCs.
most desirable product.
However, this is difficult because the rate is highly depen-
dent on the distribution of halogens in the molecule, and
there is no correlation available to account for these differ-
ACKNOWLEDGMENTS
ences.
This research was supported by the Laboratory Technology Research
Comparison of activation energies is also difficult for the
same reasons discussed above for the turnover rates. The
only additional comment is that if the inhibition by HCl is
not accounted for in the calculation of the rate constant,
the resulting Arrhenius plot (usually obtained by varying
the temperature and keeping the other variables constant)
will produce an “apparent activation energy” that will be
two times lower than the one with the HCl effect accounted
for. In other words, the rates will appear not to increase as
much as expected with the temperature due to the effect of
HCl. The activation energies here are higher than the ones
observed by Karpinski et al. (16) by a factor of almost two
but the difference may be again related to the alumina sup-
port. It appears that the activation energy increases with the
number of halogen atoms that are removed from the reac-
tant (Table 3). The mechanism advanced from the foil data
for the two most abundant products (134a and 124) (2) has
a common rate-determining step for these two compounds
and thus the activation energies should be the same. More
experiments are necessary to decide if there is a trend of
Program (ER-LTR), Office of Computational and Technology Research,
U.S. Department of Energy under a CRADA (Cooperative Research and
Development Agreement) between Ernest Orlando Lawrence Berkeley
National Laboratory (LBNL) and E.I. DuPont de Nemours and Company,
Wilmington, Delaware, under U.S. D.O.E. Contract DE-AC03-76SF00098.
REFERENCES
1. Manzer, L. E., and Rao, V. N. M., Adv. Catal. 39, 329 (1993).
2. Ribeiro, F. H., Gerken, C. A., Somorjai, G. A., Kellner, C. S., Coulston,
G. W., Abrams, L., and Manzer, L. E., Catal. Lett. 45, 149 (1997).
3. Berndt, W., Hora, R., and Scheffler, M., Surf. Sci. 117, 188 (1982).
4. Forbes, J. G., Gellman, A. J., Dunphy, J. C., and Salmeron, M., Surf.
Sci. 279, 68 (1992).
5. Grillo, M. E., Stampfl, C., and Berndt, W., Surf. Sci. 317, 84 (1994).
6. Bo¨mermann, J., Huck, M., Kuntze, J., Rauch, T., Speller, S., and Hel-
land, W., Surf. Sci. 358, 849 (1996).
7. Patterson, C. H., and Lambert, R. M., Surf. Sci. 187, 339 (1987).
8. Matheson Gas Products, Technical Brief TB-206.
9. Kiskinova, M. P., and Bliznakov, G. M., Surf. Sci. 123, 61 (1982).
10. Noordermeer, A., Kok, G. A., and Nieuwenhuys, B. E., Surf. Sci. 165,
375 (1986).
11. Wheeler, D. R., and Pepper, S. V., J. Vac. Sci. Technol. A 8, 4046 (1990).
higher activation energies as more halogen bonds are bro- 12. Yamada, T., Zhai, R., Iwasawa, Y., and Tamaru, Surf. Sci. 205, 82
(1988).
ken. However, it is reasonable to assume that for 143a the
higher activation energy indicates that the reaction mech-
anism is different than for the other two products in that a
13. Gellman, A. J., J. Phys. Chem. 96, 790 (1992).
14. Fung, S. C., and Sinfelt, J. H., J. Catal. 103, 220 (1987).
15. Gervasutti, C., Marangoni, L., and Marra, W., J. Fluorine Chem. 19,
higher energy carbon-fluorine bond may be broken in the
rate-determining step.
1 (1981/82).
16. Karpinski, Z., Early, K., and d’Itri, J. L., J. Catal. 164, 378 (1996).