Impact of zeolite aging in hot liquid water on activity for acid-catalyzed dehydration of alcohols

Article abstract of DOI:10.1021/jacs.5b06169

The location and stability of Br?nsted acid sites catalytically active in zeolites during aqueous phase dehydration of alcohols were studied on the example of cyclohexanol. The catalytically active hydronium ions originate from Br?nsted acid sites (BAS) of the zeolite that are formed by framework tetrahedral Si atom substitution by Al. Al K-edge extended X-ray absorption fine structure (EXAFS) and ²⁷Al magic angle spinning (MAS) nuclear magnetic resonance (NMR) spectroscopies in combination with density functional theory (DFT) calculations are used to determine the distribution of tetrahedral Al sites (Al T-sites) both qualitatively and quantitatively for both parent and HBEA catalysts aged in water prior to catalytic testing. The aging procedure leads to partial degradation of the zeolite framework evidenced from the decrease of material crystallinity (XRD) as well as sorption capacity (BET). With the exception of one commercial zeolite sample, which had the highest concentration of framework silanol-defects, there is no evidence of Al coordination modification after aging in water. The catalyst weight-normalized dehydration rate correlated best with the sum of strong and weak Br?nsted acidic protons both able to generate the hydrated hydronium ions. All hydronium ions were equally active for the acid-catalyzed reactions in water. Zeolite aging in hot water prior to catalysis decreased the weight normalized dehydration reaction rate compared to that of the parent HBEA, which is attributed to the reduced concentration of accessible Br?nsted acid sites. Sites are hypothesized to be blocked due to reprecipitation of silica dissolved during framework hydrolysis in the aging procedure.

Full text of DOI:10.1021/jacs.5b06169

HBEA30
HBEA30w
HBEA37
HBEA37w
HBEA52
HBEA52w
HBEA80
HBEA80w
HBEA150
HBEA150w
y = (0.043 ± 0.002)x – (1.7 ± 0.5)
R2 = 0.989

Journal of the American Chemical Society
Page 2 of 11
1
2
3
4
5
6
7
8
9
Impact of zeolite aging in hot liquid water on activity for
acid-catalyzed dehydration of alcohols
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Aleksei Vjunov,a Miroslaw A. Derewinski,a John L. Fulton,a Donald M. Camaioni,a and Johannes A.
Lerchera,b
a
Institute for Integrated Catalysis, Pacific Northwest National Laboratory, P.O. Box 999, Richland, WA 99352 (United
States)
b Department of Chemistry and Catalysis Research Institute, TU München, Lichtenbergstrasse 4, 85748 Garching (Germany)
KEYWORDS: alcohol dehydration, catalysis, aqueous medium, zeolite stability
ABSTRACT: The location and stability of Brønsted acid sites catalytically active in zeolites during aqueous phase dehydration of
alcohols were studied on the example of cyclohexanol. The catalytically active hydronium ions originate from Brønsted acid sites
(BAS) of the zeolite that are formed by framework tetrahedral Si atom substitution by Al. Al K-edge extended X-ray absorption
fine structure (EXAFS) and 27Al magic angle spinning (MAS) nuclear magnetic resonance (NMR) spectroscopies in combination
with density functional theory (DFT) calculations are used to determine the distribution of tetrahedral Al sites (Al T-sites) both
qualitatively and quantitatively for both parent and HBEA catalysts aged in water prior to catalytic testing. The aging procedure
leads to partial degradation of the zeolite framework evidenced from the decrease of material crystallinity (XRD) as well as sorp-
tion capacity (BET). With the exception of one commercial zeolite sample, which had the highest concentration of framework si-
lanol-defects, there is no evidence of Al coordination modification after aging in water. The catalyst weight normalized dehydration
rate correlated best with the sum of strong and weak Brønsted acidic protons both able to generate the hydrated hydronium ions. All
hydronium ions were equally active for the acid-catalyzed reactions in water. Zeolite aging in hot water prior to catalysis decreased
the weight normalized dehydration reaction rate compared to that of the parent HBEA, which is attributed to the reduced concentra-
tion of accessible Brønsted acid sites. Sites are hypothesized to be blocked due to re-precipitation of silica dissolved during frame-
work hydrolysis in the aging procedure.
INTRODUCTION
To address the abovementioned questions, we report a sys-
tematic study of the impact of zeolite aging in water on the
catalytic activity for acid-catalyzed dehydration of alcohols.
We have chosen the aqueous dehydration of cyclohexanol as
the test reaction, because it is the rate-determining step in phe-
nol hydrodeoxygenation (HDO).7 A combination of Al ex-
tended X-ray absorption fine structure (EXAFS) and magic
angle spinning (MAS) nuclear magnetic resonance (NMR)
spectroscopies8 are used to probe the distribution of Al T-sites
and to monitor the possible structural changes resulting from
the aging of the zeolite in water. Two main points addressed
by this work are (1) the effect of hot liquid water on the active
site integrity and catalytic reaction TOF normalized to the
number of Brønsted acid sites (BAS), and (2) the influence of
Al-concentration in the zeolite framework on the TOF in
aqueous phase. In this work we demonstrate that all accessible
Brønsted acid sites irrespective of their concentration in the
zeolite framework are equally active for acid-catalyzed dehy-
dration of cyclohexanol in water. Our results indicate that fur-
ther research is needed in the field of zeolite post-synthetic
modification leading to materials with very few defects and
highly accessible BAS as well as provide a conceptual path-
way for the successful use of zeolites in water.
Microporous materials such as zeolites are prospective solid
acid catalysts for tailored biomass conversion. The exceptional
catalytic performance of these materials is often attributed to
the pore environment of the acid sites.1 The transformation of
biomass-derived phenolic molecules (lignin) to diesel range
hydrocarbon fuels, for example, requires a complex cascade of
chemical reactions, i.e., hydrogenation of phenols to cycloal-
kanols, dehydration to cyclic olefins, and hydrogenation to
cycloalkanes requiring acid catalysis in several steps.2 Per-
forming these reactions in aqueous media3 is necessitated in
practice by the high water content of the feedstock materials.
Achieving practical reaction rates currently requires the reac-
tions be performed at temperatures above 150 °C, conditions
under which partial dissolution of zeolites may occur. Previ-
ous work suggested that lattice defects, e.g., silanol-nests, are
the reactive centers at which the zeolite framework decompo-
sition occurs in hot water,4 and that the degradation of zeolite
HBEA proceeds via selective T–O–T bond hydrolysis.5 Re-
sasco et al. have shown that curing the Si–OH species via
post-synthetic modification of the zeolite with organosilanes
allows forming a hydrophobic particle surface6 that leads to
stabilization of the zeolite in aqueous medium. However, an
understanding of the effects of hot liquid water on stability of
zeolite active sites, the Brønsted acid sites created by Al in-
corporation in the framework, and the catalyst activity is still
critically needed.
RESULTS AND DISCUSSION
Catalyst properties. In this work we used four lab-
synthesized (HBEA30, HBEA37, HBEA52, HBEA80) cata-
ACS Paragon Plus Environment

Page 3 of 11
Journal of the American Chemical Society
lysts and one commercial (HBEA150) catalyst. The properties
1
2
3
4
of these catalysts, as analyzed by elemental, Brunauer-
Emmett-Teller (BET)9 and temperature programed desorption
of bases, are compiled in Table 1. The 48 h 160 ºC water treat-
ed samples are also listed in Table 1 and are marked with a
suffix “w”.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. HIM images of the untreated (a & b) and the 48 h
160 °C (c & d) water treated HBEA52 samples. The magnifica-
tion is reported in the plots.
Figure 1. HIM images of the untreated (a & b) and the 48 h
160 °C (c & d) water treated HBEA30 samples. The magnifica-
tion is reported in the plots.
Figure 4. HIM images of the untreated (a & b) and the 48 h
160 °C (c & d) water treated HBEA80 samples. The magnifica-
tion is reported in the plots.
To determine the surface morphologies and particle size dis-
tributions of the studied catalyst we have imaged all samples
by He ion microscopy (HIM). Let us first turn to the parent
HBEA catalysts, which all have similar morphologies (panels
a and b in Figures 1-5). The particle sizes for these zeolites
vary from ~ 220 to 450 nm and increasing in size with the
Si/Al ratio. Note that the particles are agglomerates of small
crystals, which leads to the formation of inter-crystal meso-
pores. The HIM images obtained for the water treated samples
(Figures 1-5, panels c and d) do not show substantial changes
in particle sizes, however, particle surfaces appear smoothened
compared to the parent samples. We attribute this to the re-
moval and re-precipitation of silica, both dissolved from the
framework and amorphous synthesis remnants.
Figure 2. HIM images of the untreated (a & b) and the 48 h
160 °C (c & d) water treated HBEA37 samples. The magnifica-
tion is reported in the plots.
2
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 11
tional mesoporosity12 as well as to an increased concentration
of Si–OH defect sites, which greatly influence the hydrother-
mal stability of such zeolites.5
a)
b)
1
2
3
4
5
6
7
8
9
In the case of the HBEA samples aged in water prior to cat-
alytic testing, the zeolite pore volumes have decreased by ~ 30
- 40 % compared to the parent material (Table 1) for all cata-
lysts. This pore volume decrease is due to the destruction of
both micro- and mesopores (Table 1). We attribute the ob-
served changes to hydrolysis of framework Si–O–Si groups
and Si–O–Si–OH groups at the surfaces of crystallites in the
agglomerate particles.
200 nm
50 nm
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
c)
d)
Table 1 also compiles the concentrations of strong and total
Brønsted acid sites determined from the IR-spectra of ad-
sorbed pyridine. The concentrations of BAS are in all cases
lower than the concentrations of Al and the difference increas-
es with the Al concentration. HBEA52 has a BAS concentra-
tion similar to HBEA37, even though HBEA37 has 37 % more
Al than HBEA52. We suggest the discrepancy between Al and
BAS is largely due to a fraction of the BAS being inaccessible
to sorbates. Compared to the parent zeolites, the water treated
zeolites showed an approximately 15 - 30 % decrease in both
the total and strong BAS concentration. We suggest this de-
crease is due pore blocking via dissolution and re-precipitation
of silica causing more Al T-sites to become inaccessible to
pyridine molecules.
X-ray diffraction. The XRD patterns for the parent and wa-
ter treated samples are shown in the Supporting Information
(SI) and are used to estimate the degree of crystallinity of the
HBEA13 samples. The XRD signal intensity at 2θ = 7.7º varied
without a clear trend for the parent samples. For the water
treated samples, the XRD intensity of this peak was reduced to
~ 25-40 % of that observed for the respective parent material.
In contrast, the intensity of the reflection for HBEA37w was
little changed by water.
Al XAFS analysis. The variation of the X-ray absorption
(xµE) in the energy region of the Al XANES is shown in Fig-
ure 6a. The strong signal intensity observed at 1566 eV for all
samples is assigned14 to the tetrahedral Al in the zeolite sam-
ples. The prominent tetrahedral Al feature is attributed primar-
ily to excitations from Al 1s to a mixture of O 3p and Al 3p
states.8 In the case of lab-synthesized HBEA, both parent and
water-treated, there is no indication of 5- or 6-coordinated Al
species, which are typically observed at 1567.5 eV and 1568
eV, respectively.15 In the case of the commercial HBEA150
and HBEA150w, the XANES suggest 7 and 8 % octahedral
Al3+, respectively. The Al signal intensity is slightly enhanced
for all water treated samples, suggesting that the overall Al
200 nm
50 nm
Figure 5. HIM images of the untreated (a & b) and the 48 h
160 °C (c & d) water treated HBEA150 samples. The magnifica-
tion is reported in the plots.
Table 1 also reports the micro- and mesoporosity of the zeo-
lites determined by the BET analysis of N2 adsorption. The
parent samples have pore volumes ranging from about 0.30
(HBEA52 and HBEA80) to 0.42 (HBEA30) cm3g-1. In gen-
eral, the microporosity varies inversely with the concentration
of Al. In the case of the lab-synthesized zeolites, the mi-
cropore volume increases with decreasing Al concentration.
An opposite trend is observed for the mesopore volumes orig-
inating from the inter crystal voids in the agglomerates of
small crystals. This is particularly evident in the samples with
high Al content (HBEA30 and HBEA37). This tendency to
smaller crystallites leads to higher concentrations of silanols
terminating the crystals and in consequence to a higher con-
centration of mesopores.10 The commercial HBEA150 zeolite,
which has the lowest Al concentration (Si/Al = 75), exhibits
micro- and mesoporosity comparable to those observed for the
lab-synthesized HBEA30 and HBEA37 with Si/Al ratios of 15
and 19, respectively. We hypothesize that the porosity of the
commercial HBEA150 sample indicates that it had undergone
post-synthetic modification. Typically, commercial zeolites11
are synthesized with low Si/Al ratios and are then modified via
a sequence of steaming and/or mineral acid leaching steps.10
The removal of Al from the zeolite framework leads to addi-
Table 1. Studied HBEA zeolite samples with corresponding elemental analysis, BET and pyridine adsorption characterization data.
Si/Al
ratio b
15
Al concentration
Mesopore volume
Micropore volume
Total Brønsted acid site
Strong Brønsted acid site
Catalyst a
(µmol/g) b
890
(cm3/g)
0.27
0.18
0.09
0.07
0.15
0.16
0.12
0.07
0.11
0.15
(cm3/g)
0.15
0.18
0.20
0.24
0.18
0.10
0.10
0.09
0.12
0.08
concentration (µmol/g) c
concentration (µmol/g) c
HBEA30
HBEA37
400
360
360
220
150
290
320
300
180
115
330
260
280
210
140
230
290
250
170
110
19
740
HBEA52
26
540
HBEA80
40
360
HBEA150
HBEA30w
HBEA37w
HBEA52w
HBEA80w
HBEA150w
75
190
15
910
18
780
24
570
41
360
66
220
a Zeolite samples that were aged in water prior to catalytic testing are labeled with “w”; b Determined from elemental analysis; c Determined from pyridine adsorption.
3
ACS Paragon Plus Environment

Page 5 of 11
Journal of the American Chemical Society
concentration in the zeolites was slightly increased after the
HBEA30
HBEA30w
HBEA37
HBEA37w
HBEA52
HBEA52w
HBEA80
HBEA80w
HBEA150
HBEA150w
1
2
3
4
5
6
7
8
a)
water treatment due to Si dissolution from the zeolite frame-
work. This observation is in agreement with the changes in the
HBEA Si/Al ratios determined from element analysis (Table
1).
200
a)
HBEA30
HBEA30w
HBEA37
HBEA37w
HBEA52
150
HBEA52w
9
HBEA80
HBEA80w
HBEA150
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
HBEA150w
100
60
40
20
0
Chemical shift (ppm)
50
0
T1
T2
T3
T4
T5
T6
T7
T8
T9
b)
1560
1580
1600
1620
E (eV)
8
6
4
2
0
HBEA30
HBEA30w
HBEA30
HBEA30w
HBEA37
HBEA37w
HBEA52
HBEA52w
HBEA80
b)
HBEA80w
HBEA150
HBEA150w
HBEA37
HBEA37w
HBEA52
HBEA52w
HBEA80
HBEA80w
HBEA150
HBEA150w
χ
70
65
60
55
50
45
40
Chemical shift (ppm)
Figure 7. 27Al MAS NMR spectra of the parent and 48 h 160
ºC water treated HBEA samples are shown in a). The region for
the tetrahedral Al signal is shown in b). The chemical shift values
for the nine T-sites of HBEA calculated using DFT are shown for
reference. The color-coding is reported in the legend.
1
2
3
4
5
6
R, Å
Figure 6. a) XANES xµE spectra for the parent and water
treated HBEA samples. Note the increase in Al intensity for the
water treated samples suggesting higher Al concentration in these
samples. The observation is attributed to selective Si-removal in
water. b) EXAFS Img[ꢀ ꢀ ] spectra for the parent and water
treated HBEA samples. The spectra are offset (y-axis) for better
visualization. The color-coding is reported in the legend.
While the lab-synthesized samples exhibit almost no Al in
octahedral coordination, (0 ppm region18) a small fraction of
octahedral Al species was observed for the commercial
(HBEA150 and HBEA150w) samples. The NMR region as-
signed to the 4-coordinated framework Al of the HBEA spans
from 52 - 62 ppm and has two distinct peaks at 54 and 57
ppm. There is no indication of 5-coordinated and extra-
framework tetrahedral Al, which are typically observed in
NMR in the regions of ~ 30 - 40 ppm19,20 and ~ 40 - 45 ppm,21
respectively. The HBEA30, HBEA37, HBEA52 and HBEA80
as well as the respective water treated samples show a similar
NMR peak shape and intensity. The main tetrahedral Al peak
accounting for 3/5 of the spectral intensity was observed at
57.6 ppm, the shoulder accounting for ~ 2/5 at 54.3 ppm. A
significant change was observed for the HBEA150 and
HBEA150w samples. The water treatment not only resulted in
a downfield shift of the tetrahedral Al NMR peak, but also in a
change of the peak intensity. Specifically, both tetrahedral Al
NMR peaks of HBEA150 are shifted by –0.3 ppm and a frac-
tion of tetrahedral Al observed at 54.7 ppm (HBEA150) is
hypothesized to be selectively converted to octahedral Al.
Overall, the NMR spectral intensities are consistent with the
Al concentrations measured by elemental analysis (Table 1).
The water treatment resulted in two changes in the NMR spec-
tra, (1) a minor downfield shift in peak positions and (2) a
slight broadening of the peaks, which is attributed to distor-
tions in the Si-structure induced by hydrolysis of Si–O–Si
The EXAFS Img[ꢀ ꢀ ] spectra for the parent and water
treated HBEA samples are shown in Figure 6b. With the ex-
ception of HBEA150, the Img[ꢀ ꢀ ] of the parent HBEA
samples appear quite similar in representation. This observa-
tion suggests that in the case of HBEA30, HBEA37, HBEA52,
and HBEA80 Al is distributed similarly among the same sets
of T-sites. Since the EXAFS spectra indicate that the Al–O
bond and Al–Si atom distances of the 48 h 160 ºC water treat-
ed samples are nearly identical to those of the parent samples,
the treatment is suggested not to lead to selective Al removal
from the framework. We note that peak positions and intensity
in the region from 2 - 5 Å in the EXAFS spectra of HBEA150
and HBEA150w differ only in minor aspects from those ob-
served for the lab-synthesized samples.
Quantitative analysis from 27Al NMR. The NMR spectra of
the parent and water treated samples are shown in Figure 7a
and b. The NMR chemical shifts calculated8 for the 9 different
T-sites of HBEA are also shown for reference. The measured
chemical shifts are in agreement with chemical shift values
reported for HBEA previously.16,17
4
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 11
linkages in the vicinity of the Al T-sites. Both changes lead to distributions determined for both the parent and water treated
1
2
3
4
5
6
7
8
inhomogeneously broadened NMR peak shapes consisting of a
continuum of independent lines,22 due to the introduction of
defects into the symmetric structure of the HBEA framework.
The broadening also affects the quadrupolar coupling and,
therefore, the chemical shift.
samples by EXAFS and NMR are in very good quantitative
agreement and suggest a high concentration of Al in T2 sites
in the lab-synthesized samples.
In the case of HBEA150 a fraction of tetrahedral Al was
concluded to be converted to octahedral coordination. We
speculate that the more prominent transformation (compared
to the lab-synthesized samples) is due to a higher concentra-
tion of framework defects.
We suggest that the population of T-sites is due to the kinet-
ic control during zeolite synthesis, tentatively attributed to the
directing role of tetraethyl ammonium hydroxide (TEAOH)
used as template for the lab-synthesized HBEA. Previously the
key role of the template was shown for ferrierite, indicating
that the organic template preferentially occupied well-defined
positions within the void volume of the zeolite framework.23
The synthesis conditions used in this work have resulted in the
formation of isolated (determined from 29Si NMR) Al T-sites.
This finding simplifies the analysis by eliminating to consider
effects of Al-pairing. Dedecek et al. have shown that the
sources of Si and Al as well as synthesis temperature can af-
fect Al distribution and alter the pattern of the Al–O–(Si–O)n–
Al sequences.24
Al-distribution in HBEA. The distribution of Al in HBEA
was determined from the combination of EXAFS and NMR
supported by theory as described in the experimental section
and prior work.5,8 For the EXAFS analysis, the nine T-sites of
HBEA are grouped into three sets, A (T1, T2, T5, T6), Set B
(T3, T4) and Set C (T7, T8, T9), based on the similarities of
the EXAFS spectra of the T-sites of HBEA predicted from
DFT calculated structures. Molecular Dynamics (MD) EXAFS
spectra are calculated for one representative Al T-site of each
set. Linear combinations of these MD-EXAFS calculated
spectra along with the EXAFS spectrum of Al(H2O)63+, as a
model of the octahedral Al sites, were then used to fit the
measured EXAFS spectra. The measured NMR spectra are
fitted using chemical shift values for the nine T-sites of HBEA
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
–
calculated for an Al–(OSi)4(OH)12 cluster derived from the
DFT optimized HBEA unit cell.5,8 The Al distributions deter-
mined for the parent and water treated HBEA samples are
summarized in Table 2S in the SI.
29Si MAS NMR. The acquired single pulse (SP) and cross-
polarization (CP) 29Si NMR spectra for the parent and water
treated HBEA zeolites are shown in Figures 8 and 9, respec-
tively. The SP NMR spectra for the parent and water treated
HBEA samples are shown in Figure 8a and 8b, respectively. In
the SP NMR the main peak centered at ~ –111.5 ppm is at-
tributed to the Si Q4 atoms.25 Shoulder peaks left and right of
the main peak are due to the distribution of NMR chemical
shifts of the nine T-sites of HBEA. The signal at ~ –103 ppm26
is assigned to the Si Q3 [Si(OSi)3OAl]. The variation of inten-
sity for this peak is explained by the differences in the Si/Al
ratios of the studied samples. The SP NMR spectra show little
intensity at –98 ppm indicating that the fractions of Si in Q2
sites27 as well as in paired Al sites (Al–O–Si–O–Al) are negli-
gible. The SP NMR spectra of the treated samples reflect the
changes in the Si–O backbone of the zeolites in water. A slight
decrease in signal intensity as well as broadening of the NMR
spectra is observed for all samples and is attributed to an in-
crease in structural defects, as the zeolite framework is hydro-
lytically degraded.
The Al EXAFS results indicate that the Al preferentially
populates T-sites of Set A (T1, T2, T5, T6) in the parent zeo-
lite samples. Because the fit quality did not improve signifi-
cantly when we included contributions of either Set B or Set C
sites,8 an accurate determination of the Al fraction populating
set B and/or C from the EXAFS analysis alone remains am-
biguous. However, the tetrahedral Al contributions can be
further resolved by fitting the NMR spectra using the chemical
shift values calculated from DFT. The measured NMR spectra
are fitted such that the peak intensities, the offset relative to
the aqueous Al3+ reference, the peak width and peak shape
were optimized with constraints that the peaks have the same
widths, Guassian-Lorentzian ratios and absolute chemical
shifts of the Al T-sites.5,8 By allowing different values for the
aqueous Al3+ reference several unique fits were obtained for
each of the samples. The best fit was selected as the one that is
in agreement with the distribution determined from EXAFS
analysis, which indicated ~ 75 % of Al T-atoms occupied Set
A sites. The analysis of the NMR spectra then shows that this
Al is distributed in sites, T1 and T2. The remaining ~ 25 % Al
is distributed in sites T7 and T9. In the synthesized HBEA
samples, 55 % of framework Al populates the T2 sites. Al also
populates the T1 and T7 sites, each to a level of ~ 15 - 25 %.
The T9 sites are populated to a level of up to 14 %. On the
contrary, in the case of the commercial HBEA150 zeolite,
approximately half of the tetrahedral Al populates the T1 site.
The T2 site is populated at ~ 20 %. The remaining tetrahedral
Al is distributed statistically between sites T7 and T9. The
NMR spectral fitting results indicate that Al does not populate
to a significant extent the T3, T4, T5, T6, and T8 sites in any
of the samples. Finally, the fraction of Al in T-sites T1 and T2
increases with the increase in the Si/Al ratio at the expense of
sites T7 and T9.
15
10
5
15
10
5
b)
a)
29Si SP NMR
29Si SP NMR
HBEA30w
HBEA30
HBEA37
HBEA52
HBEA80
HBEA150
HBEA37w
HBEA52w
HBEA80w
HBEA150w
0
0
-80 -90 -100 -110 -120
Chemical shift (ppm)
-80 -90 -100 -110 -120
Chemical shift (ppm)
In the case of the water treated samples, HBEA30w,
HBEA37w, HBEA52w, and HBEA80w, significant changes
in the Al distribution compared to the respective parent sam-
ples were not observed. The analysis shows that the water
treatment causes neither selective Al removal nor substantial
re-distribution of Al among the zeolite T-sites. The Al–
Figure 8. 29Si MAS NMR single pulse (SP) spectra of the par-
ent (a) and water treated (b) HBEA samples. The color-coding is
shown in the legend.
5
ACS Paragon Plus Environment

Page 7 of 11
Journal of the American Chemical Society
tive agreement with the changes observed in the XRD pat-
29Si CP NMR
29Si CP NMR
b)
1
2
3
4
5
6
7
8
a)
terns: as the intensity of the reflection at 2θ = 7.7º decreased,
the concentration of Si–OH (CP NMR) increased. For exam-
ple for the lab-synthesized samples, both XRD and CP NMR
signal intensities are comparable between HBEA37w and
HBEA80w as well as between HBEA30 and HBEA52w. In-
terestingly, the degree of crystallinity (based on the peak at 2θ
= 7.7º) is lower for the parent HBEA30 and HBEA52. It ap-
pears that while the rates of hydrolysis are similar for these
HBEA zeolites, the samples with lower concentrations of lat-
tice defects exhibit greater stabilities in water.
5
4
3
2
1
0
5
4
3
2
1
0
HBEA30w
HBEA30
HBEA37
HBEA52
HBEA80
HBEA150
HBEA37w
HBEA52w
HBEA80w
HBEA150w
9
10
11
12
13
14
15
The formation of surface OH groups (typically a signal at –
100 ppm)30 cannot be unambiguously determined from the CP
NMR due to substantial peak broadening. Lippmaa et al.
showed that the higher chemical shift value can be related to a
rearrangement in the neighboring Si–OH groups in geminal
[Si(OH)2(OSi)2] or vicinal [(OSi)3Si–OH HO–Si(OSi)3] con-
formation.31 As only a minor signal (by comparison) is ob-
served at –90 ppm,32,33 which is attributed to the Si Q2
[Si(OH)2(OSi)2], most Si–OH groups are suggested to be in a
vicinal conformation in both the parent and water treated sam-
ples.
-80 -90 -100 -110 -120
-80 -90 -100 -110 -120
Chemical shift (ppm)
Chemical shift (ppm)
Figure 9. 29Si MAS NMR cross-polarization (CP) spectra of
the parent (a) and water treated (b) HBEA samples. The color-
coding is shown in the legend.
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The CP NMR spectra for the parent and water treated
HBEA samples are shown in Figure 9a and 9b, respectively.
The large peak at ~ –101 ppm is assigned to the Si(OH)(OSi)3
species.28 The Si Q3, Si(OSi)3OAl, atoms that neighbor Al T-
sites are assigned to NMR signals around ~ –106 ppm.26 The
CP NMR signals from –110 to –115 ppm are assigned to the
Si Q4 atom of the zeolite.26 In the case of the HBEA30,
HBEA37, HBEA52 and HBEA80 zeolites, the CP NMR spec-
tra of the parent samples exhibit a general trend of an increase
of signal intensity in the Si Q3 region with decreasing Si/Al
ratios in the sample. The higher Al concentration in the syn-
thesized HBEA leads to the formation of smaller crystallites
(kinetic control during synthesis), which in turn leads to a high
concentration of inter-crystal growth faults. These faulting
regions contain a high concentration of Si–OH defect sites and
Si–OH nests. The signal intensity at –101 ppm (Si Q3 region)
is markedly increased for the water treated samples, suggest-
ing an increased concentration of Si–OH groups within the
lattice structure (Si–OH nests).29
Catalytic testing. The activities of the parent and water-aged
HBEA zeolites were assessed on the aqueous phase cyclohex-
anol dehydration reaction. Cyclohexene was observed to be
the primary product in this reaction (see Table 2 for details).
Dicyclohexyl ether was also formed in the initial reaction
phase with a maximum yield well below 1 %, and is hence not
discussed further. Table 2 also reports two turnover frequen-
cies, one normalized only to the concentration of strong BAS
–1
(mol molStrong BAS s–1), and the other normalized to the con-
centration of all (weak and strong) BAS (molcyclohexene molTotal
–1 s–1). In the first case, the TOF based on strong BAS, we
BAS
note that the TOF varies by nearly a factor of 2. In the second
case, the TOF based on the total BAS ranges over values that
differ only by a factor of 1.4.
–1
The initial rates (molcyclohexene gHBEA s–1) of cyclohexene
formation catalyzed by the water treated samples are typically
~ 25 % lower than the rates catalyzed by the corresponding
parent samples. However, as we have observed (Table 1) that
the water treatment reduces the BAS concentrations by ~ 25
%, the TOF for treated and parent catalysts are considered to
be identical.
While the CP NMR signal is enhanced for all water treated
samples, there is no correlation with the Al concentration in
the sample. Based on this observation, we hypothesize that the
concentration of Si–OH defects, not the Al concentration, de-
termines the integrity of the HBEA framework. Note that the
CP NMR intensity of the water treated samples is in qualita-
Table 2. The catalytic performance of the parent and water zeolite samples is tested on the cyclohexanol dehydration reaction. The
rates of formation as well as TOF are reported for cyclohexene, the main dehydration product. Reaction conditions: 80 mL 0.33M cy-
clohexanol + 58 mg HBEA reacting at 160 ºC.
Rate of cyclohexene
formation
TOF normalized to strong BAS c
concentration
(mol × molStrong BAS–1 × s–1) ×103
TOF normalized to total BAS c
concentration
(mol × molTotal BAS–1 × s–1) ×103
Si/Al
ratio a
Cyclohexene
yield (%) b
Catalyst
(µmolcyclohexene × gHBEA-1 × s–1)
14.6 ± 1.0
HBEA30
HBEA37
15
19
26
40
75
15
18
24
41
66
11.4
11.1
11.0
6.0
45 ± 3
54 ± 4
50 ± 4
35 ± 3
28 ± 2
47 ± 3
43 ± 3
47 ± 3
36 ± 3
30 ± 2
36 ± 3
39 ± 3
38 ± 3
34 ± 2
28 ± 2
37 ± 3
39 ± 3
39 ± 3
34 ± 2
30 ± 2
14.1 ± 1.0
HBEA52
13.8 ± 1.0
HBEA80
7.5 ± 0.5
HBEA150
HBEA30w
HBEA37w
HBEA52w
HBEA80w
HBEA150w
3.2
4.1 ± 0.3
8.5
10.7 ± 0.8
9.9
12.4 ± 0.9
9.2
11.6 ± 0.8
4.9
6.2 ± 0.4
2.8
3.5 ± 0.3
a Determined from elemental analysis; b Determined yield for the 1h reaction; c BAS concentration accuracy estimated at ±10 %
6
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 11
Figure 10 shows the rate of cyclohexene formation plotted tration of surface Si-OH groups as well as silanol-nests, which
1
2
3
4
5
6
7
8
against the concentrations of (a) total and (b) strong BAS. It is
seen that the rates are linearly correlated with the total BAS
giving an R2 = 0.989 compared to R2 = 0.920 for the strong
BAS. Note that in either case some BAS determined by pyri-
dine adsorption are not active in the dehydration of cyclohex-
anol. This could be due to the inherent differences for the ac-
cessibility of pyridine and cyclohexanol to the hydronium
ions, or it could be that a constant fraction of sites is destroyed
or converted by the immersion in hot water during catalysis.
have been identified as sites initiating the framework decom-
position.35 Conceptually, zeolite framework stability, there-
fore, can be improved by minimizing the Si-OH concentration,
either via synthesis or by post-synthesis chemistry (e.g., by
converting two Si-OH groups to one Si-O-Si group).
CONCLUSION
The impact of the aqueous reaction medium on the activity
of zeolites for acid-catalyzed reactions is studied on the exam-
ple of cyclohexanol dehydration. The turnover frequency in
water is proportional to the sum of the concentration of strong
and weak Brønsted acid sites. This is attributed to the fact that
the Brønsted acid sites of the solid formed hydrated hydroni-
um ions. The location of Al T-sites in the zeolite did not affect
the turnover frequency in aqueous medium, indicating that all
hydrated hydronium ions have the same activity for alcohol
dehydration. Exposure of the zeolite to hot water decreased the
reaction rate. This was caused by framework hydrolysis of Si
reducing the accessibility of tetrahedrally coordinated Al
through, charge compensation by cations as well as by pore or
site blocking caused by dissolved and re-precipitated silica.
The induced presence of oxidic species in the pores is shown
by the reduction of the micropore volume. Only in the case of
HBEA150, the decrease in activity was also associated with
the conversion of a fraction of tetrahedral Al to octahedral
coordination. Key to a new generation of active and stable
catalysts is the reduction of the rate of hydrolysis of Si–O–Si
bonds by lowering the concentration of the Si–OH groups
either by design or by post-synthetic treatment. Detailed kinet-
ic and synthetic experiments are underway to test the hypothe-
ses and to enhance the stability of zeolites in water.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
20
a)
HBEA30
HBEA30w
HBEA37
HBEA37w
HBEA52
HBEA52w
HBEA80
HBEA80w
HBEA150
HBEA150w
15
10
5
y = (0.043 ± 0.002)x – (1.7 ± 0.5)
R2 = 0.989
0
0
100
200
300
400
-1
Concentration of Brønsted acid sites (µmolBAS gHBEA
)
20
15
10
5
b)
HBEA30
HBEA30w
HBEA37
HBEA37w
HBEA52
HBEA52w
HBEA80
HBEA80w
HBEA150
HBEA150w
EXPERIMENTAL SECTION
Reagents. HBEA150 (Si/Al = 75) was obtained from Clari-
ant in H-form. Cyclohexanol (Sigma-Aldrich, 99 %), cyclo-
hexene (Sigma-Aldrich, 99 %, GC-grade), 1,3-dimethoxy-
benzene (Sigma-Aldrich, 99 %), dichloromethane (Sigma-
Aldrich, HPLC grade), sodium sulfate (Acros Organics, 99 %,
anhydrous) are used as-received without further purification.
Catalyst preparation. HBEA zeolites are synthesized ac-
cording to the procedure reported by M. Derewinski and F.
Fajula.36 The method is briefly reviewed on the example of
HBEA30 (Si/Al=15): 260.2 g tetraethyl ammonium hydroxide
(TEAOH) and 39.6 g water are mixed in a polypropylene
beaker. 4.57 g sodium aluminate is then added under vigorous
stirring at RT until a clear solution was obtained. 50.0 g Zeosil
175mp (silica source, Rhone-Poulenc) is slowly added to the
solution. The system is aged for 24 h at RT under vigorous
stirring and then the aged gel is charged into an autoclave with
a Teflon liner. The synthesis is performed at 150 °C for 40 h
under stirring. The synthesis time (40 h) was identical for all
synthesized zeolites. Choosing the appropriate concentration
of sodium aluminate in the gel controls the Si/Al ratio in the
zeolite. The white product (powder) is filtered and washed
with deionized water until pH = 7.5. The material is dried
overnight at 80 °C and subsequently calcined at 550 °C in the
flow of dry air. The calcined material is ion-exchanged using a
0.1 M NH4NO3 solution at 80 °C for 2 h, the procedure is re-
peated 2 times. The zeolite in NH3-form is treated for 8 h at
450 °C in 100 mL/min nitrogen flow to obtain the H-form
(HBEA).
y = (0.059 ± 0.006)x – (3.5 ± 1.4)
R2 = 0.920
0
0
100
200
300
400
-1
Concentration of Strong Brønsted acid sites (µmolBAS gHBEA
)
Figure 10. The rate of cyclohexene formation as a function of
the concentration of a) total (solid markers) and b) strong (hollow
markers) BAS determined by gas-phase pyridine adsorption. The
color-coding for the ten HBEA samples is reported in the legend.
The correlations between measured rates and BAS concentrations
as well as the respective slopes are also shown.
A correlation between the Al concentrations and the weight-
normalized rates was not observed. Since the concentration of
BAS was substantially lower (~ 50 %) than the concentration
of tetrahedrally coordinated Al (Table 1) we hypothesize that
not all BAS are accessible to pyridine (blocked or being
charge compensated by cations) or cyclohexanol. We conclude
that the accessible fraction of strong and weak BAS have a
similar activity because in the presence of water zeolite pro-
tons are present as hydrated H3O+ ions.34 This in turn implies
that the intrinsic acid site strength and catalytic performance is
independent of acid strength of the parent dry zeolite.
The decrease in the BAS concentration in the presence of
hot water is proposed to be retarded by decreasing the concen-
7
ACS Paragon Plus Environment

Page 9 of 11
Journal of the American Chemical Society
An example of typical sample preparation via hot liquid wa-
used. The XAFS data were weighted by k2, and truncated us-
ing a Hanning window with dk = 1.0 Å–1 in the range of 1.5 <
k < 8.0 Å–1. Molecular dynamics (MD) EXAFS38 was used to
analyze the experimental EXAFS data for the series of HBEA
samples. The analytical approach for determining the Al-
distribution in HBEA from the EXAFS analysis has been de-
scribed previously.8 In short, based on the BEA single crystal39
a DFT structure optimization of the zeolite T-sites (populating
a single T-site of the unit cell with Al) are arranged into
groups based on the similarities of the calculated EXAFS
spectra. These groups are named Set A (T1, T2, T5, T6), Set B
(T3, T4) and Set C (T7, T8, T9). In the next step MD trajecto-
ry calculations are performed for a representative T-site from
each set (T1, T3, T7 for A, B, C, respectively). Structures
from these trajectories are used to calculate an EXAFS spec-
trum for each set. Finally, a linear combination of these MD-
EXAFS calculated spectra along with the EXAFS spectrum of
Al(H2O)63+, as a model of the octahedral Al sites, is used to fit
the measured EXAFS spectra.8
1
2
3
4
5
6
7
8
ter treatment: 0.5 g zeolite and 20 mL deionized water are
sealed in a Teflon lined autoclave and heated for 48 h at 160
°C. The autoclave is cooled using ice/water and the samples
are centrifuged and dried in vacuum for 24 h prior to storing.
Prior to catalyst characterization and catalytic testing the
HBEA samples are stored for 48 h in a desiccator over a satu-
rated Ca(NO3)2 aqueous solution in order to achieve full hy-
dration of the zeolite pores.
Catalyst characterization. The BET surface area and pore
volume (Table 1), as well as pore size distributions were de-
termined by N2 adsorption-desorption isotherms measured at
77.3 K using a ASAP2020 automatic BET-Sorptometer. The
concentrations of Brønsted acid sites were determined using
infrared spectroscopy (IR) with pyridine as probe molecule.
The catalyst samples were activated in vacuum at 723 K for 1
h with at a heating rate of 10 K min–1 from 423 to 723 K. Pyr-
idine was adsorbed and equilibrated (1 mbar, 423 K) for 1 h.
Subsequently, the sample was outgassed for 1 h in order to
remove physisorbed molecules. The sample was then heated at
a rate of 10 K min–1 from 423 to 723 K to allow partial pyri-
dine desorption (removing molecules adsorbed on weak acid
sites). The IR spectra are acquired after each step using a
ThermoScientific Nicolet IR spectrometer. For calibration, a
standard (Zeolite HZSM-5 with Si/Al = 45, acid site concen-
tration = 360 µmol g–1) was used.
Catalytic testing. The catalytic testing was performed using
the aqueous 0.33 M cyclohexanol dehydration at 160 °C as a
test reaction. An example of a typical experiment using a zeo-
lite catalyst: 58 mg HBEA and 80 mL 0.33 M cyclohexanol
are sealed in a Hastelloy PARR reactor, pressurized to 50 bar
using H2 and stirred vigorously while heated to 160 °C. The
reaction time is reported counting from the point when the set
temperature is reached (~12 min). Upon completion the reac-
tor is cooled using ice/water and the contents are extracted
using dichloromethane. The organic phase is analyzed on an
Agilent 7890A GC equipped with HP-5MS 25-m 0.25-µm i.d.
column, coupled with Agilent 5975C MS. 1,3-dimethoxy-
benzene was used as internal standard for compound quantifi-
cation.
Al K-edge XAFS. The Al K-edge XAFS measurements were
performed at the Phoenix I, elliptical undulator beamline at the
Swiss Light Source (SLS) at the Paul Scherrer Institute. Ener-
gy calibration was achieved by setting inflection point of an Al
foil spectrum to 1559.6 eV. The double-crystal monochroma-
tor employed a set of KTiOPO4 (011) crystals to provide an
energy resolution of about 0.6 eV over a scan range for the Al
K-edge from 1500 to 2150 eV. Two Ni-coated mirrors were
set at an angle of 1.45° to provide cutoff of higher harmonics.
An unfocused 1.0 × 1.0-mm beam having a flux of approxi-
mately 109 photons/sec was used. The sample chamber pres-
sure was maintained at approximately 2.5 × 10–4 mbar. Meas-
urements were performed in fluorescence mode. I0 was meas-
ured as total electron yield signal taken from a 0.5 µm thin
polyester foil, which was coated with 50 nm of Ni. This I0
detector was held in a miniaturized vacuum chamber (2.9×10–6
mbar), which is separated by a thin Kapton foil from the
measurement chamber itself. The X-ray fluorescence was de-
tected using a 4-element Vortex Si-drift diode detector.
ATHENA37 software was used during the background pro-
cessing necessary to extract the χ(k) data from the background
function. A Fourier filter cutoff distance, Rbkg, of 1.0 Å was
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
X-ray diffraction (XRD). XRD patterns were collected on a
Rigaku Mini Flex II bench top X-ray diffractometer using a
Cu-Kα radiation of 0.154056 nm (30 kV and 15 mA). Experi-
ments were performed on a rotating powder sample holder in a
2θ range of 5º to 60º with a step size of 0.02 °/s. All measure-
ments were performed under ambient conditions.
27Al MAS NMR. The Ultra-high field 27Al MAS NMR ex-
periments were performed on a Varian-Agilent Inova 63-mm
wide-bore 850 MHz NMR spectrometer. Experiments were
performed using a commercial 3.2 mm pencil type MAS
probe. In a typical experiment about 15 mg of sample powder
were loaded in the rotor and measured at ambient temperature.
The HBEA samples were stored for 48 h in a desiccator over a
saturated Ca(NO3)2 aqueous solution to reproducibly hydrate
the zeolites. This procedure leads to Al tetrahedra that have
minimal distortions and that have the maximum 27Al MAS
NMR spectral resolution.40 A single pulse sequence with a
pulse length of 2.0 ms, corresponding to a pulse angle of 45°,
was selected for acquiring each 27Al MAS NMR spectrum
with a recycle time of 1 s and total accumulation of 5000
scans. The spectra were acquired at a sample spinning rate of
20 kHz ± 2 Hz and were referenced to 1.5 M Al(NO3)3 in H2O
(0 ppm) using the center of the octahedral peak of solid γ-
Al2O3 (at 13.8 ppm) as a secondary reference. For quantitative
measurements the weights of samples loaded into the MAS
rotor were recorded and four spectra were acquired to check
the stability of the spectrometer. The matching and tuning
conditions of the RF circuit of the NMR probe were set using
a network analyzer. All other experimental conditions were
kept identical for all analyzed samples. In this way, the abso-
lute peak areas normalized to the spectrometer standard were
proportional to the Al in the sample. The spectra were ana-
lyzed using the MestreNova 8.1 software package.
29Si MAS NMR. The single pulse (SP) and cross-polarization
(CP) 29Si MAS NMR experiments were performed using a
Varian Inova 89-mm wide-bore 300 MHz NMR spectrometer
and a 5 mm HXY MAS Chemagnetics style probe. The fol-
lowing parameters for the cross-polarization pulse sequence
were used: the H90 was set to 3.5 µsec, the contact time was 1
ms and the decoupling field of 62.5 KHz was applied for 10
ms during the acquisition time. The spinning speed was set to
5 KHz.
DFT Calculations. The NMR chemical shifts for the HBEA
Al T-sites were calculated using the gauge invariant atomic
8
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 11
orbital (GIAO) approach41 in the NWChem software pack-
age.42 The modeled clusters as well as further computational
1
2
details are reported elsewhere.8
[3] Pinkert, A.; Marsh, K. N.; Pang, S. S.; Staiger, M. P., Chem.
3
4
5
6
7
8
9
Rev. 2009, 109, 6712–6728.
Helium ion microscopy (HIM). HIM images were obtained
using 35 keV He ions with 0.1 pA beam current at normal
incidence. Secondary electrons were detected using an Ever-
hart-Thornley detector. Because the samples were completely
insulating a thin layer of carbon (<1 nm) was coated using a
carbon sputter deposition system. The instrument resolution
was 0.35 nm.
[4] Ravenelle, R. M.; Schüssler, F.; D’Amico, A.; Danilina, N.; van
Bokhoven, J. A.; Lercher, J. A.; Jones, C. W.; Sievers, C., J. Phys.
Chem. C 2010, 114, 19582–19595.
[5] Vjunov, A.; Fulton, J. L.; Camaioni, D. M.; Hu, J. Z.; Burton,
S. D.; Arslan, I.; Lercher, J. A., Chem. Mater. 2015, 27, 3533–3545.
[6] Zapata, P. A.; Faria, J.; Ruiz, M. P.; Jentoft, R. E.; Resasco, D.
E., J. Am. Chem. Soc. 2012, 134, 8570–8578.
[7] Zhao, C.; Kou, Y.; Lemonidou, A. A.; Li, X.; Lercher, J. A.,
Chem. Commun. 2010, 46, 412–414.
[8] Vjunov, A.; Fulton, J. L.; Huthwelker, T.; Pin, S.; Mei, D.;
Schenter, G. K.; Govind, N.; Camaioni, D. M.; Hu, J. Z.; Lercher, J.
A., J. Am. Chem. Soc. 2014, 136, 8296–8306.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ASSOCIATED CONTENT
Supporting Information. XRD patterns, BET measurement data
as well as NMR and EXAFS fit can be found in Supporting In-
formation. This material is available free of charge via the Internet
at http://pubs.acs.org.
[9] Brunauer, S.; Emmett, P. H.; Teller, E., J. Am. Chem. Soc.
1938, 60, 309–319.
[10] Barrer, R. M., Zeolites and Clay Minerals as Sorbents and
Molecular Sieves, Academic Press, London, UK, 1978.
[11] Rüscher, C. H.; Salman, N.; Buhl, J.–C.; Lutz, W., Micropor.
Mesopor. Mater. 2006, 92, 309–311.
[12] Zhdanov, S. P.; Khvoshchev, S. S.; Feoktistova, N. N., Syn-
thetic Zeolites, vol. 1, Gordon & Breach Science Publishers, New
York, NY, USA, 1990.
[13] Wright, P. A.; Zhou, W.; Perez–Pariente, J.; Arranz, M.; J.
Am. Chem. Soc. 2005, 127, 494–495.
[14] van Bokhoven, J. A.; Sambe, H.; Ramaker, D. E.; Konings-
berger, D. C., J. Phys. Chem. B 1999, 103, 7557–7564.
[15] Li, D.; Bancroft, G. M.; Fleet, M. E.; Feng, X. H.; Pan, Y.,
Am. Miner. 1995, 80, 432–440.
[16] van Bokhoven, J. A.; van der Berden, A. M. J.; Koningsber-
ger, D. C.; J. Am. Chem. Soc. 2003, 125, 7435–7442.
[17] Maier, S. M.; Jentys, A.; Lercher, J. A., J. Phys. Chem. C
2011, 115, 8005–8011.
[18] van Bokhoven, J. A.; Koningsberger, D. C.; Kunkeler, P.; van
Bekkum, H.; Kentgens, A. P. M., J. Am. Chem. Soc. 2000, 122,
12842–12847.
AUTHOR INFORMATION
Corresponding Author
* Johannes A. Lercher; Johannes.Lercher@pnnl.gov
Present Addresses
Institute for Integrated Catalysis, Pacific Northwest National La-
boratory, P.O. Box 999, Richland, WA 99352 (United States).
Author Contributions
The manuscript was written through contributions of all authors.
ACKNOWLEDGMENT
Authors thank B. W. Arey (PNNL) for HIM measurements, T.
Huthwelker for support during Al XAFS measurements at the
Swiss Light Source (PSI, Switzerland), J. Z. Hu and S. D. Burton
(PNNL) for support during NMR experiments. This work was
supported by the U. S. Department of Energy (DOE), Office of
Science, Office of Basic Energy Sciences, Division of Chemical
Sciences, Geosciences & Biosciences. MD acknowledges support
by the Materials Synthesis and Simulation Across Scales (MS3
Initiative) conducted under Laboratory Directed Research & De-
velopment Program at PNNL. NMR experiments were performed
at the Environmental Molecular Science Laboratory, a national
scientific user facility sponsored by the DOE Office of Science,
Office of Biological and Environmental Research, and Physical
Science Laboratory both located at Pacific Northwest National
Laboratory (PNNL).
[19] Deng, F.; Yue, Y.; Ye, C., J. Phys. Chem. B 1998, 102, 5252–
5256.
[20] Hunger, M.; Engelhardt, G.; Weitkamp, J., Microporous Ma-
ter. 1995, 3, 497.
[21] Perez–Pariente, J.; Sanz, J.; Fornes, V.; Corma, A., J. Catal.
1990, 124, 217–223.
[22] Schmidt–Rohr, K.; Spiess, H. W., Multidimensional Solid–
State NMR and Polymers 1994, Academic Press, London, ISBN 0–
12–626630–1.
[23] Pinar, A. B.; Gomez–Hortguela, L.; McCusker, L. B.; Perez–
Pariente, J.; J. Chem. Mat. 2013, 25, 3654–3661.
[24] Gabova, V.; Dedecek, J.; Cejka, J., Chem. Commun. 2003,
1196.
[25] Lippmaa, E.; Mägi, M.; Samoson, A.; Tarmak, M.; Engel-
hardt, G., J. Am. Chem. Soc. 1981, 103, 4992–4996.
[26] Čapek, L.; Dědeček, J.; Wichterlová, B., J. Catal. 2004, 227,
352–366.
ABBREVIATIONS
DFT, density functional calculations; EXAFS, extended X-ray
absorption fine structure spectroscopy; XANES, X-ray absorption
near edge spectroscopy; MAS, magic angle spinning; NMR, nu-
clear magnetic resonance; MD, molecular dynamics; HDO, hy-
drodeoxygenation; XAFS, X-ray absorption spectroscopy; XSW,
X-ray standing wave; XRD, X-ray diffraction; TPD, Temperature
programed desorption.
[27] Fyfe, C. A.; Feng, Y.; Grondey, H.; Kokotailo, G. T.; Gies, G.,
Chem. Rev. 1991, 91, 1525.
[28] Hunger, M.; Kärger, J.; Pfeifer, H.; Caro, J.; Zibrowius, B.;
Bülow, M.; Mostowicz, R., J. Chem. Soc., Faraday Trans.1 1987, 83,
3459–3468.
[29] Zhuravlev, L. T., Colloids Surf. A 1993, 74, 71–90.
[30] Engelhardt, G.; Lohse, U.; Samoson, S.; Mägi, M.; Tarmak,
M.; Lippmaa, E., Zeolites 1982, 2, 59–62.
[31] Lippmaa, E.; Samoson, A.; Brei, V. V.; Gorlov, Y. I., Fiz.
Chim. 1981, 259, 403.
[32] Fyfe, C. A.; Gobbi, G. C.; Kennedy, G. J., J. Am. Chem. Soc.
1985, 89, 277.
REFERENCES
[1] Antal, M. J.; Carlsson, M.; Xu, X.; Anderson, D. G. M., Ind.
Eng. Chem. Res. 1998, 37, 3820–3829.
[2] Zhao, C.; Lercher, J.A., Angew. Chem. Int. Ed. 2012, 51, 5935–
5940.
9
ACS Paragon Plus Environment

Page 11 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
[33] Pines, A.; Gibby, M. G.; Waugh, J. S., J. Chem. Phys. 1973,
59, 569.
[34] Corma, A.; Lopez Agudo, A.; Fornes, V., J. Chem. Soc.,
Chem. Commun. 1983, 942–944.
[35] Ravenelle, R. M.; Schüssler, F.; D’Amico, A.; Danilina, N.;
van Bokhoven, J. A.; Lercher, J. A.; Jones, C. W.; Sievers, C., J.
Phys. Chem. C 2010, 114, 19582–19595.
[36] Derewinski, M. A.; Fajula, F., Appl. Catal. A:General 1994,
108, 53–61.
[39] Baerlocher, C.; McCusker, L. B., Database of Zeolite Struc-
tures, accessed Jan. 2015; http://www.iza–structure.org/databases/
[40] Jiao, J.; Kanellopoulos, J.; Wang, W.; Ray, S. S.; Foerster, H.;
Freude, D.; Hunger, M., Phys. Chem. Chem. Phys. 2005, 7, 3221 –
3226.
[41] Wolinski, K.; Hinton, J. F.; Pulay, P. J. Am. Chem. Soc. 1990,
112, 8251-8260.
[42] Valiev, M.; Bylaska, E. J.; Govind, N.; Kowalski, K.; Straats-
ma, T. P.; Van Dam, H. J. J.; Wang, W.; Nieplocha, J.; Apra, E.;
Windus, T. L.; de Jong, W. A., Computer Physics Communications
2010, 181, 1477.
[37] a) Ravel, B.; Newville, M., J. Synchrotron Radiat. 2005, 12,
537; b) Newville, M., Ibid. 2001, 8, 322.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
[38] Palmer, B. J.; Pfund, D. M.; Fulton, J. L., J. Phys. Chem.
1996, 100, 13393–13398.
10
ACS Paragon Plus Environment