Ir/KLTL zeolites: Structural characterization and catalysis of n-hexane reforming

Article abstract of DOI:10.1006/jcat.1996.0059

Ir/KLTL zeolite catalysts were prepared by incipient wetness impregnation of LTL zeolites with [Ir(NH₃)₅Cl]Cl₂. The catalysts were characterized by extended X-ray absorption fine structure (EXAFS) spectroscopy, infrared spectroscopy, and H₂ chemisorption. EXAFS data show that the average Ir cluster size (after treatment at 300°C in H₂) increased from about 7 to 20 A as the zeolite K : Al atomic ratio increased from 0.34 to 1.56. Infrared spectra of adsorbed CO show that the electron donation to the Ir increased as the K : Al ratio increased. In contrast to the performance reported for Pt/KLTL zeolites with metal clusters as small as those observed in the present experiments, the Ir/KLTL catalyst has a low selectivity for dehydrocyclization of n-hexane at 440-480°C and 1 atm with a H₂ : n-hexane feed molar ratio of 6. Instead, the catalysts are selective for hydrogenolysis. The selectivity is insensitive to the K : Al ratio, but the activity for dehydrocyclization is a maximum at a K : Al atomic ratio of about 1. The results show that even the smallest Ir clusters to which electron donation is significant still behave essentially like metallic Ir in the catalytic reactions.

Full text of DOI:10.1006/jcat.1996.0059

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
JOURNAL OF CATALYSIS 159, 14–22 (1996)
ARTICLE NO. 0059
Ir/KLTL Zeolites: Structural Characterization and
Catalysis of n-Hexane Reforming
N. D. Triantafillou,ꢀ S. E. Deutsch,† O. Alexeev,† J. T. Miller,‡ and B. C. Gatesꢀ, †, 1
ꢀ
Center for Catalytic Science and Technology, Department of Chemical Engineering, University of Delaware, Newark, Delaware 19716;
†
Department of Chemical Engineering and Materials Science, University of California, Davis, California 95616; and
‡
Amoco Oil Company, Research and Development Department, P.O. Box 3011, Naperville, Illinois 60566
Received January 3, 1995; revised October 10, 1995
are so selective for dehydrocyclization. The various sug-
gestions for why the catalysts are so selective are summa-
rized elsewhere (11); in short, high dehydrocyclization se-
lectivites are attributed to some combination of the effects
of Pt cluster size (8), the constraints imposed by the zeolite
pores, and electronic effects associated with the basicity of
the support.
Ir/KLTL catalysts have been prepared with clusters of
about five to six atoms each, which are about the same size
as those in well-prepared Pt/KLTL zeolites (11). In con-
trast to Pt, Ir is quite resistant to sintering on a number
of metal oxide and zeolite supports in the presence of H2
at relatively high temperatures (12–17). Ir on various sup-
ports (MgO, Al2O3, NaY zeolite, and NaX zeolite) is both
a hydrogenolysis and a dehydrogenation catalyst (18–20).
Because Ir is highly active for hydrogenolysis (18–20), it is
not applied as a hydrocarbon-reforming catalyst. However,
it has been used with Pt in bimetallic-reforming catalysts,
in which it seems to minimize deactivation of the catalyst
resulting from coke formation (18, 21, 22).
Ir/KLTL zeolite catalysts were prepared by incipient wetness im-
pregnation of LTL zeolites with [Ir(NH3)5Cl]Cl2. The catalysts were
characterized by extended X-ray absorption fine structure (EX-
AFS) spectroscopy, infrared spectroscopy, and H2 chemisorption.
EXAFS data show that the average Ir cluster size (after treatment
at 300 C in H2) increased from about 7 to 20 A as the zeolite K : Al
atomic ratio increased from 0.34 to 1.56. Infrared spectra of ad-
sorbed CO show that the electron donation to the Ir increased as
the K : Al ratio increased. In contrast to the performance reported
for Pt/KLTL zeolites with metal clusters as small as those observed
in the present experiments, the Ir/KLTL catalyst has a low selec-
tivity for dehydrocyclization of n-hexane at 440–480ꢀC and 1 atm
with a H2 : n-hexane feed molar ratio of 6. Instead, the catalysts
are selective for hydrogenolysis. The selectivity is insensitive to the
K : Al ratio, but the activity for dehydrocyclization is a maximum
at a K : Al atomic ratio of about 1. The results show that even the
smallest Ir clusters to which electron donation is significant still
ꢀ
˚
c
behave essentially like metallic Ir in the catalytic reactions. ꢁ 1996
Academic Press, Inc.
The goal of the research summarized here was to pre-
pare and characterize a family of KLTL zeolite-supported
Ir catalysts with various K : Al ratios and to determine how
the Ir cluster size and the support properties influence the
catalytic properties for n-hexane conversion. Ir was chosen
because well-defined Ir clusters, including Ir4 and Ir6, have
been prepared on metal oxide and zeolite supports, and one
of the goals was to prepare and investigate such small clus-
ters on supports similar to those used for Pt clusters, namely,
zeolite LTL. The catalysts were characterized by extended
X-ray absorption fine structure (EXAFS) spectroscopy and
infrared spectroscopy and used for dehydrocyclization and
hydrogenolysis of n-hexane.
INTRODUCTION
An important reaction in catalytic reforming of naphtha
is dehydrocyclization of C6 and C7 paraffins. These reac-
tions are catalyzed selectively when the catalyst consists
of Pt clusters supported in zeolite LTL containing K+ or
K+ and Ba2 + exchange ions (1–3). The zeolite LTL cata-
lysts are nonacidic, and evidently the Pt clusters alone are
the catalytically active species. LTL zeolite is evidently the
best support for selective Pt-catalyzed dehydrocyclization
reactions; this zeolite incorporates straight parallel pores
that incorporate cages that are approximately ellipsoidal in
˚
shape with approximate dimensions of 4.8 ꢁ 12.4 ꢁ 10.7 A;
the apertures are 12-membered oxygen rings with diame-
˚
ters of about 7.5 A.
EXPERIMENTAL
Notwithstanding many data in the literature (4–10), there
is still no consensus about why Pt-containing LTL zeolites
Sample Preparation
1 To whom correspondence should be addressed at the University of
California at Davis.
KLTL zeolites (Union Carbide) were prepared with var-
ious K : Al ratios. They were converted into the NH+4 form
14
0021-9517/96 $18.00
c
Copyright ꢂ 1996 by Academic Press, Inc.
All rights of reproduction in any form reserved.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
n-HEXANE CONVERSION IN Ir/KLTL ZEOLITE
15
by aqueous ion exchange or simply washed with water to aged. A Si(111) double-crystal monochromator was used,
lower the K+ content; alternatively, some samples were with a resolution of 1E/E = 2 ꢁ 10ꢄ4. Higher harmonics
impregnated with KNO3 to increase the K+ content. The were rejected by detuning the monochromator by 20%.
several procedures were as follows:
EXAFS Data Analysis (11)
(A) KLTL zeolite (130 g) was ion exchanged with 1 liter
of 1.5 M NH4NO3 at 60ꢃC for 2 h. The zeolite was filtered,
Phase shift and backscattering amplitude data were
obtained from EXAFS spectra of suitable reference
compounds. Pt foil was used as a reference for Ir–Ir
absorber–backscatterer pairs. Na2Pt(OH)6 was used for Ir–
O absorber–backscatterer pairs. The crystallographic data
and details of the preparation of the reference files are
summarized elsewhere (11). The transferability of phase
shifts and backscattering amplitudes has been shown theo-
retically (23, 24) and experimentally (12, 25–27). Standard
procedures (27) were used to extract the EXAFS (chi) func-
tion from the raw data. Normalization of the EXAFS data
was performed by dividing the absorption intensity by the
height of the background absorption at 50 eV beyond the
absorption edge. The main EXAFS contributions were iso-
lated by Fourier filtering of the final EXAFS function (ex-
tracted EXAFS data), with k3 weighting (k is the wave vec-
tor) and phase and amplitude correction by using the Pt foil
reference (Table 1). Data analysis was performed on the fil-
tered data. Best-fit parameters were obtained by fitting with
the Koningsberger difference file technique (28). Low-Z
and high-Z contributions were fitted on both r space (r is
the radial distance from absorbing atom to backscattering
atom) and k space, with both k1 and k3 weighting so as not to
overemphasize either the low-Z or high-Z backscatterers.
Details of the fitting procedures are given elsewhere (27).
washed, and calcined at 540ꢃC for 3 h.
(B) KLTL zeolite (50 g) was washed six times with 500 ml
of hot water (pH 8.0); the resultant sample was calcined at
400ꢃC for 3 h.
(C) KLTL zeolite (100 g) was washed three times with
500 ml of hot water (pH 9.0) and calcined at 400ꢃC for 3 h.
(D) KLTL zeolite (15 g) was added to a solution of 0.78 g
KNO3 in 11 ml of water, dried, and calcined at 400ꢃC for 3 h.
(E) KNO3 (1.5 g) was dissolved in 11 ml of water and
added to KLTL zeolite (15 g), which was then dried and
calcined at 400ꢃC for 3 h.
(F) KNO3 (2.5 g) was dissolved in 11 ml of water and
added to 15 g of zeolite, which was dried and calcined at
400ꢃC for 3 h.
Each of these samples was dried, crushed into a fine pow-
der, and brought in contact with [Ir(NH3)5Cl]Cl2 (Johnson
Matthey) in water by the incipient wetness method. For ex-
ample, 2.0 g of [Ir(NH3)5Cl]Cl2 dissolved in 40 ml of water
at 70ꢃC was added to 50.0 g of zeolite. After the addition
of a few drops of HNO3, the pH of the [Ir(NH3)5Cl]Cl2
solution was adjusted to 7 by addition of NH4OH. The so-
lution was filtered and added dropwise to the zeolite. The
resultant slurry was allowed to stand at room temperature
for 3 h, and the sample was dried at 120ꢃC overnight. The
crystallites of zeolite were held together without a binder.
Chemisorption Measurements
Elemental Analysis
H2 chemisorption was carried out with an RXM-100 cata-
lyst testing and characterization apparatus (Advanced Sci-
entific Design, Inc.). The Ir/KLTL zeolites were reduced
at either 300 or 500ꢃC for 1 h and then evacuated at pres-
sures <10ꢄ6 Torr for 2 h at the reduction temperature. The
The zeolites were analyzed for Ir by Schwartzkopf Micro-
analytical Laboratories, Woodside, NY, and for Al and K
by the analytical division of Amoco Oil Co. by inductively
coupled plasma analysis.
TABLE 1
Extended X-Ray Absorption Fine Structure Spectroscopy
Parameters Used for Isolation of Main EXAFS Data
in Fourier Filteringa
In preparation for EXAFS spectroscopy, the catalysts
were treated for 1 h at 300ꢃC in flowing H2 (99.999 + %,
formed by electrolysis of water in a Balston generator).
Traps containing Cu2O and 4A zeolite were used to re-
move traces of oxygen and water, respectively. The cata-
lysts were pressed into self-supporting wafers and loaded
into a cell in the absence of air. EXAFS measurements were
done at Beamline X-11A at the National Synchrotron Light
Source, Brookhaven National Laboratory, Upton, NY. The
synchrotron energy was maintained at 2.5 GeV, and the ring
current was at least 110 mA. Two scans were done for each
sample at the Ir LIII edge (11215 eV), with the sample at
near liquid nitrogen temperature, and the scans were aver-
K : Al atomic ratio
ꢄ1
˚
˚
in zeolite
1k (A
)
1r (A)
p
0.34
0.95
1.34
1.56
3.59–14.55
3.48–13.96
3.62–13.60
3.70–13.60
1.74–3.47
1.42–3.52
1.70–3.38
1.67–3.40
13
15
12
12
a
Notation: 1k, range used for forward Fourier transform; 1r,
range used for isolation of main data; p, number of free parame-
ters from Nyquist theorem, p = (21k1r)/ꢀ + 1. Spectral isolation
was carried out on k3-weighted Fourier transform, with phase and
amplitude correction using Pt foil reference.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
TRIANTAFILLOU ET AL.
16
catalysts were cooled to room temperature under vacuum
before the chemisorption experiments were carried out.
Infrared Spectroscopy of Chemisorbed CO
Catalysts reduced at either 300 or 500ꢃC were loaded
into a gas-tight infrared cell and treated with flowing CO
for 30 min at room temperature and then evacuated at pres-
sures <10ꢄ3 Torr to remove any gas-phase or physisorbed
CO. Infrared spectra were recorded in the transmission
mode with a Nicolet 510 M FTIR or a Bruker IFS-66 v
FTIR instrument.
Catalytic Reaction Experiments
Details of the apparatus used for catalytic reforming of
n-hexane are given elsewhere (11). Catalysts were reduced
at 500ꢃC for 1 h and then cooled under flowing H2 to the
operating temperature of 400, 440, or 480ꢃC prior to the
start of reactant flow. n-Hexane, fed at rates of 1.33 to 2.7
ml/h, was vaporized before mixing with H2 and He. The
H2, He, and n-hexane flow rates were chosen so that the
n-hexane partial pressure was 76 Torr and the molar H2 : n-
hexane ratio 6. The mass of catalyst varied from 15 to 600
mg. Productswereanalyzedbyon-linegaschromatography.
FIG. 1. Raw EXAFS data characterizing the sample with K : Al = 0.34
after treatment in H2 at 300ꢃC.
to Ir–Ir interactions, was first estimated for k > 8 Aꢄ1, be-
˚
cause this part of the spectrum is dominated by Ir–Ir con-
tributions. The EXAFS function representing the best-fit
parameters was then subtracted from the filtered data over
the entire k range. The residual spectrum was fitted with
two Ir–O contributions. The EXAFS function representing
these two contributions was subtracted from the full data
to give a better estimate of the Ir–Ir contribution. The fit-
ting of the Ir–Ir contribution and the Ir–O contributions
was then continued iteratively. Figures 2 and 3 show the
difference files representing the sample with K : Al = 0.34;
these data are typical. The best-fit parameters from all
three contributions were added and compared with the fil-
tered data, both in r space and in k space, until the fit was
satisfactory.
RESULTS
Catalyst Compositions
The catalyst compositions are summarized in Table 2. In
what follows, the KLTL zeolites are referred to by the K : Al
atomic ratio.
EXAFS Data
The fits obtained for each sample are very good. The
The data show oscillations in the raw EXAFS up to EXAFS parameters are summarized in Table 3. The Ir–Ir
k = 14 Aꢄ1 or greater. Typical data are shown in Fig. 1. The first-shell coordination number increased from 3.7 to 7.9
˚
largest contribution in each EXAFS spectrum, attributed as the K : Al ratio increased from 0.34 to 1.34. However,
TABLE 2
Elemental Analysis Results for Ir in KLTL Zeolites with Different K : Al Ratios and Results
of H2 Chemisorption Experiments
K : Al atomic ratio
in zeolite
H/Ir ratio after
H/Ir ratio after
Estimated
particle size (A)
a
b
treatment at 300ꢃC
treatment at 500ꢃC
NIr–Ir
Reference
˚
Ir (wt%)
0.34
0.95
0.99
1.08
1.08
1.14
1.34
1.56
1.16
1.87
0.95
1.03
1.03
1.66
1.29
1.56
0.30
0.25
1.24
0.63
0.70
1.44
3.7
6.3
7
11
3.2
6
8
(11)
(11)
4.2c
1.69
0.80
0.49
1.99
1.04
0.73
7.9
7.8
18
18
a NIr–Ir: Ir–Ir first shell coordination number from EXAFS analysis, after treatment in H2 at 300ꢃC, except where noted.
b Particle size estimated for spherical particle [Ref. (13)].
c After treatment in H2 at 500ꢃC.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
n-HEXANE CONVERSION IN Ir/KLTL ZEOLITE
17
TABLE 3
Best-Fit EXAFS Parameters for Samples after Treatment in H2 at 300ꢃC for 1 ha
K : Al atomic
ratio in zeolite
2
1ꢁ2 (A )
1E0 (eV)
Reference
˚
˚
Shell
N
R (A)
0.34
Ir–Ir
Ir–O
Ir–O
Ir–Ir
Ir–O
Ir–O
Ir–Ir
Ir–O
Ir–O
Ir–Ir
Ir–O
Ir–O
Ir–Ir
Ir–O
Ir–O
Ir–Ir
Ir–O
Ir–O
3.7
0.9
0.9
6.3
1.6
0.6
3.2
1.2
2.5
4.2
0.7
1.2
7.9
0.9
0.5
7.8
1.2
0.3
2.71
2.04
2.54
2.70
2.02
2.53
2.70
2.18
2.60
2.70
2.09
2.67
2.71
2.01
2.50
2.71
2.03
2.48
0.00119
0.00560
ꢄ0.00518
0.00209
0.00990
ꢄ0.00428
0.0017
0.0034
0.0088
0.0022
0.0085
ꢄ2.72
2.82
9.81
This work
This work
This work
This work
This work
This work
(11)
(11)
(11)
(11)
(11)
0.95
1.08b
1.08c
1.34
1.56
ꢄ1.20
6.34
7.15
ꢄ0.45
ꢄ4.68
ꢄ0.47
0.68
7.05
0.0085
ꢄ6.50
ꢄ0.81
2.03
(11)
0.00169
0.00990
ꢄ0.00424
0.00161
0.00990
ꢄ0.00597
This work
This work
This work
This work
This work
This work
0.03
ꢄ1.69
1.69
2.12
a Notation: N, coordination number for absorber–backscatterer pair; R, radial distance from
absorber to backscatterer; 1ꢁ2, Debye-Waller factor; 1E0, inner potential correction. Estimated
errors: N (Ir–Ir), ꢅ15%; N (Ir–O), ꢅ20%; R, ꢅ1%, 1ꢁ2, ꢅ10%, 1E0, ꢅ10%.
b Sample reduced at 300ꢃC.
c Sample reduced at 500ꢃC.
˚
there was no significant change in the Ir–Ir coordination Ir–O distances, 2.02 ꢅ 0.02 A, which are bonding distances
number as the K : Al ratio increased further to 1.56. The Ir– (35), are slightly less than those observed for other sup-
Irdistanceineachsamplewasthesamewithinexperimental ported Ir samples (12, 33, 34, 36). The longer Ir–O distances,
˚
˚
error, 2.70–2.71 A.
2.48–2.54 A, are about the same as Pt–O (36) and Ir–O
Two Ir–O contributions were needed to fit the data sat- (37) distances reported for supported metals, but a value
˚
isfactorily (Table 3). Similar metal–oxygen contributions of about 2.7 A is more typical of metal oxide-supported Pt
have been reported for highly dispersed Ir and other metals and Ir (12, 28, 30, 33–35), perhaps being associated with
on metal oxide supports (9, 10, 13–15, 28–34). The shorter hydrogen at the metal–support interface, such as in Pt–H–
FIG. 2. EXAFS data characterizing the sample with K : Al = 0.34: k1-
weighted Fourier transform of the filtered EXAFS data minus the cal-
culated Ir–Ir contribution (solid line) and the best-fit Ir–O contributions
(dashed line), phase corrected using Na2Pt(OH)6 reference file.
FIG. 3. EXAFS data characterizing the sample with K : Al = 0.34; k3-
weighted Fourier transform of the filtered EXAFS data minus the sum of
the calculated Ir–O contributions (solid line) and best-fit Ir–Ir contribution
(dashed line), phase and amplitude corrected using Pt foil reference file.