n-Butane isomerization on sulfated zirconia: How olefins affect surface intermediate concentration

Article abstract of DOI:10.1016/j.jcat.2006.10.007

Isotopic transient kinetic analysis (ITKA) was used to study the effect of various olefin additions (ethylene, propylene, 1-butene, and 1-pentene) on n-butane isomerization on sulfated zirconia (SZ) at 100 °C. Earlier work showed that the activity of this catalyst is enhanced not only by the addition of butene, but also by the addition of ethylene, propylene, and 1-pentene. This reaction activity enhancement observed for olefins other than butene was suggested to be due to the isomerization of n-butane via different oligomeric species. In this work, we show that at the maximum reaction rate after the initial induction period, the activity of the reaction sites appeared to be similar, regardless of whether they were activated by the various added olefins or by butene produced during reaction. The addition of butene was able to lower the average surface residence time of the active intermediates leading to isobutane (τ_isoC4^*) at the very beginning of the reaction induction period, reinforcing an early suggestion that butene formation and accumulation of olefinic intermediates are required for the reaction. The improved isobutane formation rate caused by the addition of olefins was due to an increased concentration of surface intermediates leading to isobutane (N_isoC4^*). Any type of added olefin led to the formation/utilization of additional active olefin-modified sites (i.e., adsorbed carbenium ions on acid sites). A larger N_isoC4^* was obtained compared with the reaction without olefin addition when the olefins were added only during the initial 2 min of reaction and then terminated. This suggests that most of the olefin-modified sites were probably formed by the addition of olefin at the early stages of the reaction. Those sites were able to integrate themselves into the reaction cycles and to persist for multiple turnovers.

Full text of DOI:10.1016/j.jcat.2006.10.007

Journal of Catalysis 245 (2007) 198–204
www.elsevier.com/locate/jcat
n-Butane isomerization on sulfated zirconia: How olefins affect surface
intermediate concentration
Nattaporn Lohitharn, James G. Goodwin Jr. ∗
Department of Chemical and Biomolecular Engineering, Clemson University, Clemson, SC 29634, USA
Received 11 August 2006; revised 3 October 2006; accepted 8 October 2006
Abstract
Isotopic transient kinetic analysis (ITKA) was used to study the effect of various olefin additions (ethylene, propylene, 1-butene, and 1-pentene)
◦
on n-butane isomerization on sulfated zirconia (SZ) at 100 C. Earlier work showed that the activity of this catalyst is enhanced not only by the
addition of butene, but also by the addition of ethylene, propylene, and 1-pentene. This reaction activity enhancement observed for olefins other
than butene was suggested to be due to the isomerization of n-butane via different oligomeric species. In this work, we show that at the maximum
reaction rate after the initial induction period, the activity of the reaction sites appeared to be similar, regardless of whether they were activated
by the various added olefins or by butene produced during reaction. The addition of butene was able to lower the average surface residence time
∗
of the active intermediates leading to isobutane (τ
) at the very beginning of the reaction induction period, reinforcing an early suggestion
isoC4
that butene formation and accumulation of olefinic intermediates are required for the reaction. The improved isobutane formation rate caused by
the addition of olefins was due to an increased concentration of surface intermediates leading to isobutane (N
∗
). Any type of added olefin led
isoC4
∗
to the formation/utilization of additional active olefin-modified sites (i.e., adsorbed carbenium ions on acid sites). A larger N
was obtained
isoC4
compared with the reaction without olefin addition when the olefins were added only during the initial 2 min of reaction and then terminated. This
suggests that most of the olefin-modified sites were probably formed by the addition of olefin at the early stages of the reaction. Those sites were
able to integrate themselves into the reaction cycles and to persist for multiple turnovers.
© 2006 Elsevier Inc. All rights reserved.
Keywords: Sulfated zirconia; n-Butane isomerization; Olefin addition; Isotopic transient kinetic analysis; Active sites
1. Introduction
though SZ has high activity for alkane isomerization even at
low temperatures [1], it also deactivates rapidly.
The structural isomerization of linear paraffins to branched
ones is an important process in raising the octane number of
gasoline. Most current commercial processes are based on the
use of chlorinated Pt/Al2O3 that requires the continuous addi-
tion of chlorine. To make this process more environmentally
friendly, heterogeneous strong acid catalysts that do not require
the continuous addition of an undesirable chemical, such as
chlorine, have been sought.
Modified zirconias have been shown to be very strong solid
acids and much research has been conducted on these catalysts.
One of the most active modified zirconias is sulfated zirconia
(SZ), on which one current UOP-licensed process is based. Al-
The mechanisms of alkane isomerization, particularly for
n-butane isomerization, have been widely studied. Numerous
studies have shown that an increase in SZ activity is manifested
when butene is added [2,3] or is present as an impurity [4–7].
This result has led to a suggestion of a bimolecular mechanism
that requires the formation of C+8 species via the oligomeriza-
tion of butene and n-butane before isomerization and β-scission
to form isobutane [2,3,5,8–12]. However, Li et al. [13] sug-
gested that n-butane isomerization proceeds via a monomole-
cular pathway in which isobutane is formed by the skeletal iso-
merization of sec-butyl ion produced by a protonation of butene
to tert-C+4 intermediates. An increase in the surface concentra-
tion of sec-butyl ions due to the presence of butene led to an
enhanced reaction rate. Recently, we have put forth a new hy-
pothesized mechanism based on the results that any olefin can
*
Corresponding author. Fax: +1 864 656 0784.
E-mail address: james.goodwin@ces.clemson.edu (J.G. Goodwin).
0021-9517/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2006.10.007

N. Lohitharn, J.G. Goodwin Jr. / Journal of Catalysis 245 (2007) 198–204
199
enhance n-butane isomerization on SZ [14]. Not only butene,
2.2. Reaction rate analysis
but also ethylene, propylene, and 1-pentene, would appear to
be able to form active “olefin-modified sites,” which have been
suggested to be the main centers of reaction [2]. In this pro-
posed mechanism, isomerization takes place by a monomolec-
ular skeletal rearrangement of oligomeric species through pro-
tonated cyclopropane states [14]. These oligomeric intermedi-
ates are formed via oligomerization of olefin and an adsorbed
carbenium ion, the so-called “olefin-modified site” [2]. The
comprehensive mechanism, having attributes of both the pre-
viously hypothesized monomolecular and bimolecular path-
ways, is able to resolve all the disparate results in the litera-
ture.
The study of n-butane isomerization was performed in a
quartz microreactor (8 mm i.d.). The reaction was carried out
under differential reaction conditions with a maximum conver-
sion <5% to minimize temperature and concentration gradi-
ents. The SZ catalyst (0.2 g) was pretreated in situ at 315 ◦C
under 30 cc/min of dry air (National Specialty Gases, Zero
Grade) for 4 h, where the temperature was ramped at 2 ◦C/min
and held constant at 100 ◦C for 1 h before heating to 315 ◦C.
Then the reactor was cooled down to 100 ◦C under air and
flushed with He (National Specialty Gases, UHP) for 30 min
before the reaction. Reactions were carried out at 100 ◦C and
at a constant pressure of 1.5 atm, where reaction rates were af-
fected by neither mass nor heat transfer limitations.
A trap containing 10 g of H-mordenite held at room tem-
perature was used to remove any olefin impurities present in
the n-butane feed. The impurities remaining in the n-butane
feed after the trap were 1 ppm ethane, 17 ppm propane, and
90 ppm isobutane. The total flow rate of the reaction mix-
ture was 30 cc/min (STP), containing 15 cc/min of 5% n-
12C4H10 + 1% Ar (balance in He) (National Specialty Gases)
and 15 cc/min of He. The flow of pure He was reduced to main-
tain a constant concentration of n-butane when olefins (1% of
C=2 , C=3 , 1-C4=, or 1-C=5 in He [National Specialty Gases, UHP])
were added to the feed stream. Ethylene was added to the reac-
tion at the lowest olefin-to-paraffin (O/P) ratio of 0.003, due
to its strong catalyst activation–deactivation response, whereas
propylene, 1-butene, and 1-pentene were added at the op-
timum O/P ratio for maximum rate enhancement, as deter-
mined previously (0.009, 0.012, and 0.009, respectively) [3].
Reaction samples were analyzed using a Varian 3700 gas
chromatograph equipped with a 12-ft 15% Squalane CP-AW-
DMCS/Chromosorb 80/100-mesh column (Alltech) and a flame
ionization detector. All experiments were reproducible within a
maximum error rate of 5%.
It has been shown that the addition of butene does not have
an impact on the average site activity of the SZ catalyst, because
the average surface residence time of isobutane intermediates
(τi∗soC ) remained essentially unchanged for n-butane isomer-
izatio4n at 150 ◦C as determined by isotopic transient kinetic
analysis (ITKA) [2]. The increase in the isobutane formation
rate caused by the addition of butene at 150 ◦C was shown to be
due to an increase in the concentration of surface intermediates
leading to isobutane (Ni∗soC ). However, it is important to know
4
the effects of adding ethylene, propylene, or 1-pentene on the
surface kinetic parameters of n-butane isomerization. By deter-
mining the surface kinetic parameters using ITKA, a better un-
derstanding can be developed of how the addition of olefins in-
fluence reaction rate with TOS, since understanding what hap-
pens on the catalyst surface during the induction period is key to
understanding how the catalyst works and how olefins influence
activity and reaction. Although isotopic tracing measurements
were carried out previously for n-butane isomerization with the
addition of 1-butene [2], the surface reaction parameters during
the reaction induction period or for reaction at 100 ◦C have not
been reported. Therefore, in this work, ITKA was carried out
for n-butane isomerization at 100 ◦C in the presence and ab-
sence of various added olefins (ethylene, propylene, 1-butene,
and 1-pentene). The concentrations of surface intermediates
and other surface kinetic parameters during the induction pe-
riod of n-butane isomerization on SZ were measured and are
reported herein.
2.3. Isotopic transient kinetic analysis
For ITKA during the reaction, a Valco two-position valve
with an electric actuator was used to switch between 5% n-
12C4H10 + 1% Ar in He and 5% 13CH3(12CH2)213CH3 in He
(Isotec, 99%) without disturbing the other reaction conditions.
A trace of Ar was present in the n-12C4H10 to measure the
gas-phase holdup for the reaction system. A 34-port VICI au-
tosampling valve was used to collect 16 effluent samples during
the 2-min isotopic transient periods. Then the collected efflu-
ent samples were separated by a gas chromatograph with 24-ft
15% Squalane CP-AW-DMCS/Chromosorb 80/100-mesh col-
umn (Alltech) held at 27 ◦C. A thermal conductivity detector
(TCD) was used for product analysis to prevent product de-
struction. H2/He (8.5%) at a flow rate of 20 cc/min was used as
the carrier gas in the GC column and as a source of H2 for a hy-
drogenolysis unit in which the separated effluent was converted
to methane (CH4) after GC separation. The hydrogenolysis re-
actor containing 5 g of 5% Pt on Al2O3 was held at 250 ◦C.
The product CH4 was subsequently introduced into a Balzers–
2. Experimental
2.1. Catalyst preparation and characterization
The SZ catalyst was prepared by calcining a sulfate-doped
zirconium hydroxide [Zr(OH)4] precursor (XZ0 1249/01, ob-
tained from MEI, Flemington, NJ) at 600 ◦C under static air
for 2 h. It was then cooled to room temperature over a 4-h pe-
riod. BET surface area of the calcined SZ catalyst using N2
adsorption was determined with a Micromeritics ASAP 2010.
The sulfur content was measured by Galbraith Laboratories,
Inc. (Knoxville, TN). The crystallinity of the calcined catalyst
was analyzed using a Philips X’Pert X-ray diffractometer with
monochromatized CuKα radiation and a Ni filter, operated at
40 kV and 30 mA.

200
N. Lohitharn, J.G. Goodwin Jr. / Journal of Catalysis 245 (2007) 198–204
Pfeiffer Prisma 200-amu quadrupole mass spectrometer (Pfeif-
fer Vacuum) via a 1/16-inch capillary tube with differential
pumping. The MS data were collected by a personal computer
using Balzers Quadstar 422 v 6.0 software. Surface kinetic pa-
rameters such as the average surface residence time (τisoC and
4
τn-C ) were determined from the isotopic transients using ITKA
4
data analysis software. The analysis method and system setup
have been described previously [15,16].
We previously reported that, if not taken into account, the
impact of isobutane readsorption on ITKA can cause an over-
estimation of the average residence time of surface intermedi-
ates leading to isobutane (τisoC ), resulting in an overestima-
4
tion in the concentration of surface intermediates for isobutane
(NisoC ) [16,17]. Therefore, to correct τisoC for readsorption,
4
4
experiments were performed at different space times by vary-
ing catalyst weight from 0.05 to 0.25 g, and the values of τisoC
4
were measured. These values were then used to determine a cor-
rected value of τisoC . The impact of reversible isobutane read-
4
sorption on τisoC was significant for every situation studied
4
(not shown). By extrapolating to zero space-time, a better value
for τisoC in the absence of readsorption effects was obtained,
4
designated as τi∗soC . Thus, the τi∗soC and Ni∗soC values reported
4
4
4
in this paper are essentially free from isobutane readsorption
effects. Ni∗soC was det∗ermined by multiplying the formation
4
rate of isobutane by τisoC [16]. Unlike isobutane, the aver-
4
age residence time of n-butane (τn-C ) was essentially constant
4
independent of space-time within the limits of experimental er-
ror (not shown). This indicates that reversible readsorption of
n-butane was not an issue, possibly due in part to the relatively
high partial pressure of n-butane, as was also suggested by Kim
et al. [16]. Bajusz et al. [18] found that at relatively high partial
pressures, the competition for adsorption increases. This in turn
minimizes readsorption of the n-butane reactant molecules. Al-
though clearly some n-butane had to adsorb on the reaction sites
to form the isobutane product, much more (1–2 orders more)
was able to adsorb and desorb reversibly from other sites with-
out reaction.
Fig. 1. The effect of continuous 1-butene addition (O/P ratio = 0.012) on the
◦
surface kinetic parameters with TOS at 100 C.
in an isobutane formation rate [3]. The variations in isobutane
formation rate, the average surface residence time of the in-
termediates leading to isobutane (τi∗soC ), and the concentration
4
3. Results and discussion
of active surface intermediates leading to isobutane (Ni∗soC ) at
4
100 ◦C with TOS (corrected for readsorption) in the presence
and the absence of 1-butene addition are shown in Fig. 1.
In the absence of 1-butene addition, τi∗soC appeared to de-
crease somewhat with TOS before reaching a 4steady-state value
of 8.5 s at 20 min TOS, where the maximum activity of the
catalyst was exhibited (Fig. 1). This suggests that the aver-
3.1. Catalyst characterization
The BET surface area and sulfur content of the fresh cal-
cined SZ were measured as 137 m2/g and 1.89 wt% (590 µmol
sulfur/g), respectively. XRD analysis showed that ZrO2 was
present only in the tetragonal phase. No sulfur compounds were
detected due to the low concentration of sulfur present.
age site activity of the catalyst (TOF∗ITK, where TOFI∗TK
=
rateisoC /N∗
≈ 1/τ∗isoC ) increased at the beginning of re-
4
isoC4
action (the first 20 min T4OS). The initial decrease in τ∗isoC
3.2. The continuous addition of various olefins
4
obtained when the catalyst was not exposed to added olefin
is probably because of the greater amount of butene produced
with TOS. Hammache and Goodwin [2] suggested that butene
(added or formed) is able to adsorb on catalytic sites and form
olefin-modified sites that contribute to isobutane formation. As
was also suggested by others [2,3,5,7,13,19], n-butane isomer-
ization on SZ requires some amount of butene to initiate the
reaction.
It has been shown that the presence of 1-butene in the feed
stream affects the behavior of n-butane isomerization on SZ
by shortening the induction period and increasing the rate of
catalyst deactivation [2,3,5]. Thus, ITKA was carried out for n-
butane isomerization at 100 ◦C as a function of TOS in the pres-
ence of 1-butene at an O/P ratio of 0.012, which has been shown
to be the optimum ratio to ensure a maximum enhancement

N. Lohitharn, J.G. Goodwin Jr. / Journal of Catalysis 245 (2007) 198–204
201
Unlike without olefin addition, smaller τ∗isoC values at the
4
very beginning of the induction period were observed when
1-butene was added to the reaction. τ∗isoC was 5.5 s at 1.5 min
4
but increased rapidly up to 8 s at 6 min TOS. Adding 1-butene
was able to lower τ∗isoC , whereas deactivation of catalyst was
4
not significant. This suggests that the presence of butene in
the feed stream activated catalytic sites to promptly react with
n-butane (or, rather, butene formed from n-butane). Butene for-
mation has been suggested to occur via the oxidative dehydro-
genation of n-butane by sulfate groups where the oxidizing and
reducing species are SO3 and SO2, respectively [7,20]. How-
ever, τ∗isoC was observed to increase after the first 3 min of
4
reaction. This was probably due to a higher participation of
weaker acid sites in overall reaction as a result of rapid cata-
lyst deactivation of the most active sites caused by the addition
of 1-butene. Yaluris et al. [21] and Kim et al. [22] found that
strong acid sites (i.e., sites with a higher activity contributing
◦
Fig. 2. The reaction rates at 100 C when ethylene, propylene, 1-butene and
1-pentene were added continuously. (Reproduced from Ref. [14].)
to a smaller τ∗isoC ) tend to deactivate at a faster rate compared
4
alyst activity and significantly shortened the induction period,
whereas adding propylene, 1-butene, and 1-pentene produced
relatively comparable reaction rate profiles with a 6-min induc-
tion period at these reaction conditions before catalyst deacti-
vation became dominant.
with the less active sites. It has also been reported that the deac-
tivation rate constant for 1-butene addition at 100 ◦C is 4 times
higher than that for the reaction without 1-butene addition [3].
Fig. 1 clearly shows that the activity and deactivation be-
haviors of the catalyst (i.e., increase and decrease in the rate
of∗ isobutane formation) were due primarily to changes in
NisoC . N∗isoC was significantly larger with the addition of 1-
As shown in Fig. 2, olefins influence the maximum catalyst
activity to varying degrees. Thus, the ITKA study was per-
formed at the maximum isobutane formation rate (i.e., 20, 1.5,
and 6 min TOS for no olefin, ethylene, and propylene, 1-butene
and 1-pentene addition, respectively) and the results are shown
4
4
butene than in the absence of the added olefin. The larger val-
ues for N∗isoC observed from continuous addition of 1-butene
4
(1.2 µmol/g at the maximum activity) compared with those in
its absence (0.6 µmol/g) confirms the concept that the addi-
tion of butene increases the number of olefin-modified sites by
in Table 1. At the maximum reaction rate, τ∗isoC values were
insignificant different among the reactions with and without
4
olefin∗addition, showing relatively the same average site activity
forming or activating additional active sites. However, Ni∗soC
4
(TOFITK ≈ 1/τ∗ ) regardless of whether the olefin-modified
isoC4
went through a maximum value of 1.2 µmol/g at 6 min TOS
and then decreased, just like the overall rate of isobutane for-
mation. A loss of active sites by coke/oligomer formation has
been reported to be the major cause of catalyst deactivation for
n-butane isomerization on SZ [9,16,23,24]. In contrast, little
sites were formed by different added olefins or by butene pro-
duced during reaction. This reinforces the previous suggestion
that any type of olefin (ethylene, propylene, and 1-pentene) be-
sides butene is able to adsorb on the catalyst surface and form
olefin-modified sites [14]. Such sites have been proposed to be
main centers of reaction that could lead to the formation of
oligomer when reacted with a butene (from n-butane) molecule.
Those oligomeric species were hypothesized to be C+6 , C7+, C8+,
and C+9 due to the dissimilarity among the olefin-modified sites
(i.e., C+2 , C3+, C4+, and C5+ from addition of ethylene, propylene,
1-butene, and 1-pentene, respectively). However, it is hard to
imagine that those sites were so similar when they were formed
by different hypothesized oligomeric species. Considering the
significant change in N∗isoC was observed after 10 min TOS
4
for the reaction without added olefin, due to the lower catalyst
deactivation rate at this low reaction temperature.
ITKA measurements were also made for reaction with the
addition of ethylene, 1-propylene, and 1-pentene to study the
effects of olefin type on the surface reaction parameters of
the catalyst. Propylene, 1-butene, and 1-pentene were added at
the optimum O/P ratios (0.009, 0.012, and 0.009, respectively)
known to maximally enhance catalyst activity [3]. (Increasing
the concentration of added olefins beyond these values only
accelerates catalyst deactivation and lowers isobutane selec-
tivity.) Ethylene was added at an O/P ratio of 0.003, because
ethylene strongly promotes catalyst activation and deactivation.
Fig. 2 shows that the activity of SZ for n-butane isomerization
is increased by the addition not only of 1-butene but also of
ethylene, propylene, and 1-pentene. The enhanced maximum
activity of the catalyst was observed in varying degrees, de-
pending on the type of olefin added. The increased activity for
isobutane observed during the addition of various olefins was
much greater than the simple conversion of the added olefin
to isobutane. Adding ethylene significantly promoted the cat-
small change in τ∗isoC values, it appears that the activity of ac-
4
tive sites modified by added ethylene was higher (i.e., τi∗soC
was lower) than in the absence of olefin addition. Only ∗a rela4-
tively small difference in τ∗isoC was observed, because τisoC is
the average value for all of the active intermediates, includ4ing
more than a third (0.7 vs 1.9 µmol/g) generated by butene from
the n-butane reactant.
4
Obviously, the increase in N∗isoC observed during reaction
4
with added olefins appeared to be the main cause of the higher
catalyst activities observed. As would be expected, N∗isoC was
most enhanced by the∗addition of ethylene. Adding eth4ylene
was able to increase NisoC to almost 3 times that extant in the
4

202
N. Lohitharn, J.G. Goodwin Jr. / Journal of Catalysis 245 (2007) 198–204
Table 1
◦
The effect of the continuous addition of various olefins on surface kinetic parameters at 100 C at maximum reaction rate
a
∗
c
∗
e
∗
g
b
∗
d
f
h
Olefin
O/P
ratio
TOS
(min)
Rate
N
TOF
θ
/
TOF
τ
N
θ
/
n-C4
isoC4
sulfur
−4 −1
n-C4
ITK
−1 −1
)
isoC4
isoC4
isoC4
−2
)
−3
)
(µmol/(g s))
(µmol/g)
(10
s
(10
s
)
(µmol/g)
0.7
sulfur (10
2.5
(s)
sulfur (10
1.2
None
–
20
1.5
6
0.083
0.235
0.152
0.155
0.159
1.41
3.98
2.58
2.63
2.69
8.5
7.9
8.0
8.0
8.3
1.18
1.27
1.25
1.25
1.20
14.5
14.0
14.0
15.0
14.8
=
C
0.003
0.009
0.012
0.009
1.9
3.2
2.4
2
=
=
1-C
4
=
1-C
5
i
i
i
C
1.2
2.1
2.4
3
6
1.2
2.1
2.6
6
1.3
2.2
2.5
a
Max error = 5%.
b
c
TOF
= Rate
/(590 µmol sulfur/g); Max error = 5%.
sulfur
isoC4
×τ
∗
∗
N
= Rate
; Max error = 0.06 µmol/g.
isoC4
isoC4
isoC4
d
e
Corrected τ
due to isobutane readsorption by extrapolating to zero space time; Max error = 0.5 s.
isoC4
isoC4
∗
∗
∗
−1
.
TOF
= 1/τ
= Rate
/N
; Max error = 0.007 s
isoC4
ITK
isoC4
f
N
= flow rate n-C (2.55 µmol/(g s)) ×τ
(no significant readsorption of n-C in the catalyst bed determined); Max error = 0.6 µmol/g.
n-C4
4
n-C4
4
g
∗
∗
θ
= N
= N
/(590 µmol sulfur/g); Max error = 0.0009.
isoC4
isoC4
h
i
θ
/(590 µmol sulfur/g); Max error = 0.0012.
n-C4
n-C4
The optimum olefin-to-paraffin ratio for the maximum isobutane formation rate.
absence of added olefin and 2 times that produced by the addi-
tion of any other olefin. Ethylene probably accomplished this by
rapidly populating and forming a greater number of active sites
on the catalyst surface as a result of a very high reactivity of the
primary carbenium ion-modified site produced by the adsorp-
tion of ethylene. Table 1 also shows that the similarity in the
isobutane formation rate enhancement caused by the addition
of propylene, 1-butene, and 1-pentene was due to similar val-
ues of N∗isoC . N∗isoC was observed to be similar probably be-
4
4
cause of a similar reactivity of these 3 olefins to form secondary
carbocations on the acid sites. Although experiments with dif-
ferent olefin concentrations were not performed in this study,
we believe, based on these results and those previously [3], that
N∗isoC would decrease with higher olefin concentrations due to
4
enhanced deactivation.
◦
Fig. 3. The reaction rates at 100 C when propylene, 1-butene and 1-pentene
were added during the first 2 min of reaction at the optimum O/P ratio (O/P ratio
∗
The N
value was observed to be very low compared
isoC4
=
=
=
with Nn-C (the concentration of n-butane reversibly adsorbed),
= 0.009 for C and 1-C , and 0.012 for C ). (Reproduced from Ref. [14].)
4
3
5
4
which was approximately 14.5 µmol/g of catalyst. Only a frac-
tion of the total adsorption sites (1.2 µmol/g, or <10%) con-
ethylene considerably increased catalyst activity and signifi-
cantly promoted catalyst deactivation in less than 2 min (see
Fig. 2). The effect of an initial 2-min addition of the other vari-
ous olefins on the formation rate of isobutane is shown in Fig. 3.
Although the flow of olefin was introduced to the feed stream
for only 2 min at the beginning of the reaction and then termi-
nated, isobutane formation was still higher compared with that
when no olefin was added. The initial 2-min addition of these
three olefins also produced similar effects on catalyst activity as
seen from continuous addition. To gain more insight into how
olefins promote catalyst activity after their addition is termi-
nated, ITKA measurements were made for the reaction at two
different TOS (i.e., 5 min and at maximum catalyst activity) be-
fore significant catalyst deactivation.
tributed to the formation of isobutane. Nn-C was found to be
4
relatively constant within the limits of experimental error and
was not affected by the addition of any olefin. In addition, sur-
face coverages of sulfur (assuming that all sulfur entities are
accessible) by n-butane (θn-C ) and isobutane (θ∗ ) are given
4
in Table 1. It was found that probably only ca. i2so.5C%4 of sulfur
loading was used in the adsorption of n-butane (θn-C ), which
4
was an order of magnitude higher than θi∗soC
.
4
3.3. Initial 2-min addition of olefins
The positive effect on the formation rate of isobutane from
continuous feeding of olefins to the reaction was overwhelmed
long term by more rapid catalyst deactivation. To minimize this
negative impact of olefin addition, propylene, 1-butene, and 1-
pentene were added to the reaction at 100 ◦C for only the first
2 min of the reaction. The addition of ethylene during the ini-
tial 2 min of the reaction was not done in this study because
A comparison of ITKA results at 5 min TOS and at maxi-
mum isobutane formation between the reaction with and with-
out the initial 2 min addition of olefins (propylene, 1-butene and
1-pentene) is shown in Table 2. Adding olefin only during the
initial 2 min of reaction resulted in the same effect on τi∗soC
4

N. Lohitharn, J.G. Goodwin Jr. / Journal of Catalysis 245 (2007) 198–204
203
Table 2
4. Conclusions
The effect of initial 2 min addition of various olefins on surface kinetic parame-
ters at 100 C
◦
Our previous proposal [14] that the active sites for n-butane
isomerization on SZ can be activated by any olefin has been
strongly confirmed by a similar effect of adding olefins on sur-
∗
c
e
b
∗
d
TOS
(min)
Olefin O/P
ratio
N
N
n-C4
Rate
τ
isoC4
isoC4
isoC4
a
(µmol/g)
(µmol/(g s))
(µmol/g)
0.2
(s)
face reaction parameters, such as TOF∗ITK (1/τi∗soC ) and the
5
5
5
5
None
–
0.016
0.092
0.091
0.088
0.083
0.123
0.130
0.117
13.2
8.2
9.1
8.3
8.5
9.1
8.8
8.2
14.3
16.8
14.5
15.0
14.5
15.3
13.5
14.4
4
=
=
1-C
4
=
1-C
5
concentration of active surface intermediates (Ni∗soC ). At the
C
0.009
0.012
0.009
–
0.8
3
4
maximum reaction rate, τ∗isoC was observed to be essentially
0.8
4
0.7
equal for all continuously added olefins (ethylene, propylene,
1-butene, and 1-pentene) within the limits of experimental er-
ror. An increased concentration of active surface intermediates
f
20
None
0.7
f
f
f
=
11
11
11
a
C
0.009
0.012
0.009
1.1
3
leading to isobutane (Ni∗soC ) with the addition of various olefins
=
4
1-C
1-C
1.1
4
=
was shown to be the cause of the enhanced catalyst activity.
1.0
5
N∗isoC was most increased by the addition of ethylene, result-
The optimum olefin-to-paraffin ratio for the maximum isobutane formation
4
ing in the highest maximum formation rate of isobutane. The
rate.
value of N∗isoC at the maximum rate was similar for the addi-
b
Max error = 5%.
4
c
∗
∗
N
= Rate
×τ
; Max error = 0.04 µmol/g.
due to isobutane readsorption by extrapolating to zero
isoC4
tion of propylene, 1-butene, and 1-pentene, resulting in similar
isoC4
isoC4
d
rate maxima. As expected, the increase in N∗isoC caused by the
Corrected τ
isoC4
4
space time; Max error = 0.3 s.
initial 2-min olefin additions was lower than that with contin-
uous olefin additions. However, although the olefin feed was
terminated at least 3 min ∗before the isotopic tracing measure-
e
N
= flow rate n-C (2.55 µmol/(g s)) ×τ
(no significant readsorp-
n-C4
n-C4
4
tion of n-C in the catalyst bed determined); Max error = 1 µmol/g.
4
f
TOS at the maximum isobutane formation rate.
ments were made, the NisoC was still larger compared with
4
that for the case without added olefin, regardless of the type
of olefin added. This reinforces our previous suggestion that
olefin-modified sites can be formed by adding any olefin at
the beginning of the reaction induction period and persist for
multiple reaction cycles. Certainly, the fact that the addition of
1-butene was able to lower the τ∗isoC value at early stages of
the induction period lends credence 4to the widely held belief
that in the absence of added olefins, n-butane isomerization re-
quires the formation of butene and an accumulation of surface
intermediates to carry out the reaction.
just like the continuous addition of olefins, regardless of the
type of olefin. Taking into account experimental error, τi∗soC
remained unchanged with TOS during the induction period fo4r
2-min olefin addition. No differences in intrinsic site activity
were seen among the reactions with the addition of these three
olefins. These values (τi∗soC ) were similar to those seen when
4
the catalyst was continuously exposed to added olefins within
experimental error. This indicates that the average site activ-
ity was not influenced by either olefin type or the addition of
olefin.
Similar values of N∗isoC appeared to be associated with the
4
Acknowledgments
similar catalyst activities observed for reaction with an initial
2-min addition of propylene, 1-butene, and 1-pentene, but these
values were slightly lower than those seen when olefins were
added continuously. Similar to continuous 1-butene addition,
This material is based on work supported by the National
Science Foundation (NSF) under grant CTS-0211495. Any
opinions, findings, and conclusions or recommendations ex-
pressed in this material are those of the authors and do not
necessarily reflect the views of the NSF.
ITKA done before the point at which catalyst deactivation be-
∗
came dominant showed that N
increased with TOS dur-
isoC4
ing the induction period for every olefin added. Although the
isotopic tracing measurements were performed after olefin ad-
References
dition was stopped, a higher N∗isoC compared with that for
4
the reaction without added olefin was obtained. This suggests
that these additional active sites likely were activated by butene
formed from n-butane or perhaps by olefins fed slowly dur-
ing the first 2 min of reaction. Olefin-modified sites are known
to be able to persist for multiple reaction cycles, producing
a long-lasting impact of olefin addition on the catalyst. Each
added butene molecule has been reported to result in more
than 10–700 additional isobutane molecules, depending on re-
action conditions [2,3,8]. Tabora and Davis [8] reported that at
least 10 excess isobutane molecules were produced per mole-
cule of added trans-2-butene on iron- and manganese-promoted
SZ during n-butane isomerization at 100 ◦C; Hammache and
Goodwin [2] found this value to be 700 when the reaction was
carried out at 150 ◦C with the addition of 1-butene.
[1] C.Y. Hsu, C.R. Heimbuch, C.T. Armes, B.C. Gates, J. Chem. Soc. Chem.
Commun. (1992) 1645.
[2] S. Hammache, J.G. Goodwin Jr., J. Catal. 211 (2002) 316.
[3] N. Lohitharn, J.G. Goodwin Jr., E. Lotero, J. Catal. 234 (2005) 199.
[4] V. Adeeva, G.D. Lei, W.M.H. Sachtler, Catal. Lett. 33 (1995) 135.
[5] H. Liu, V. Adeeva, G.D. Lei, W.M.H. Sachtler, J. Mol. Catal. A Chem. 100
(1995) 35.
[6] M.A. Coelho, W.E. Alvarez, E.C. Sikabwe, R.L. White, D.E. Resasco,
Catal. Today 28 (1996) 415.
[7] X.B. Li, K. Nagaoka, L.J. Simon, R. Olindo, J.A. Lercher, A. Hofmann,
J. Sauer, J. Am. Chem. Soc. 127 (2005) 16159.
[8] J.E. Tabora, R.J. Davis, J. Catal. 162 (1996) 125.
[9] T.K. Cheung, J.L. Ditri, B.C. Gates, J. Catal. 151 (1995) 464.
[10] S.Y. Kim, J.G. Goodwin Jr., S. Hammache, A. Auroux, D. Galloway,
J. Catal. 201 (2001) 1.
[11] V. Adeeva, G.D. Lei, W.M.H. Sachtler, Appl. Catal. A Gen. 118 (1994)
L11.

204
N. Lohitharn, J.G. Goodwin Jr. / Journal of Catalysis 245 (2007) 198–204
[12] S. Hammache, J.G. Goodwin Jr., J. Catal. 218 (2003) 258.
[13] X.B. Li, K. Nagaoka, L.J. Simon, R. Olindo, J.A. Lercher, J. Catal. 232
(2005) 456.
[14] N. Lohitharn, E. Lotero, J.G. Goodwin Jr., J. Catal. 241 (2006) 328.
[15] S.L. Shannon, J.G. Goodwin Jr., Chem. Rev. 95 (1995) 677.
[16] S.Y. Kim, J.G. Goodwin Jr., D. Galloway, Catal. Lett. 64 (2000) 1.
[17] S.H. Ali, J.G. Goodwin Jr., J. Catal. 171 (1997) 339.
[18] I.G. Bajusz, J.G. Goodwin Jr., D. Galloway, N. Greenlay, Langmuir 14
(1998) 1846.
[19] J.E. Tabora, R.J. Davis, J. Am. Chem. Soc. 118 (1996) 12240.
[20] D. Farcasiu, A. Ghenciu, J.Q. Li, J. Catal. 158 (1996) 116.
[21] G. Yaluris, R.B. Larson, J.M. Kobe, M.R. Gonzalez, K.B. Fogash, J.A.
Dumesic, J. Catal. 158 (1996) 336.
[22] S.Y. Kim, J.G. Goodwin Jr., D. Galloway, Catal. Today 63 (2000) 21.
[23] B. Subramaniam, V. Arunajatesan, C.J. Lyon, Catal. Deact. 126 (1999)
63.
[24] S.Y. Kim, N. Lohitharn, J.G. Goodwin Jr., R. Olindo, F. Pinna, P. Canton,
Catal. Commun. 7 (2006) 209.