Insight into the Selectivity of Isopropanol Conversion at Strontium Titanate (100) Surfaces: A Combination Kinetic and Spectroscopic Study

Article abstract of DOI:10.1021/acscatal.7b02417

This work aims to better understand the role of interfacial molecular structure that governs selectivity and activity in heterogeneous catalytic reactions. To address this, a comprehensive study of isopropyl alcohol conversion over an archetypal perovskite material, strontium titanate (SrTiO₃ or STO), was performed with an array of techniques sensitive to orthogonal aspects of the ensuing chemistry. Cubic-shaped STO nanoparticles with only the (100) facet exposed were synthesized and used to study the ensemble kinetic conversion of isopropyl alcohol over the surfaces, which showed selectivity to form acetone, with minor propylene products appearing at elevated temperatures. These results in combination with inelastic neutron scattering measurements provide not only insight into the selectivity and overall activity of the catalysts but also low-frequency vibrational signatures of the adsorbed and reacted species. To complement these measurements, pristine thin films of STO (100) were synthesized and used in combination with vibrational sum frequency generation spectroscopy to extract the absolute molecular orientation of the adsorbed molecules at the interface. It was found that the isopropyl alcohol assumes an orientation where the -CH group points toward the STO surface; this prereaction geometry offers an obvious pathway to produce acetone by abstracting the α proton and, thus, provides a mechanistic explanation of selectivity at STO (100) surfaces. This insight opens up pathways to explore and modify surfaces to tune the activity/selectivity through a molecular level understanding of the reactions at the surface.

Full text of DOI:10.1021/acscatal.7b02417

Subscriber access provided by LAURENTIAN UNIV
Article
Insight into the Selectivity of Isopropanol Conversion at Strontium
Titanate (100) Surfaces: A Combination Kinetic and Spectroscopic Study
Shuai Tan, Matthew B. Gray, Michelle K. Kidder, Yongqiang Cheng, Luke L. Daemen,
Dongkyu Lee, Ho Nyung Lee, Ying-Zhong Ma, Benjamin Doughty, and Daniel A. Lutterman
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.7b02417 • Publication Date (Web): 13 Oct 2017
Downloaded from http://pubs.acs.org on October 13, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street
N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
Insight into the Selectivity of Isopropanol Conversion at
Strontium Titanate (100) Surfaces: A Combination Kinetic
and Spectroscopic Study
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
1
2
1
3
3
Shuai Tan , Matthew B. Gray , Michelle K. Kidder , Yongqiang Cheng , Luke L. Daemen ,
4
4
1
1,
Dongkyu Lee , Ho Nyung Lee , YingꢀZhong Ma , Benjamin Doughty * and Daniel A.
1,
Lutterman *
AUTHOR ADDRESS:
1
2
. Chemical Sciences Division, Oak Ridge National Laboratory, Oak Ridge, TN, 37831, USA
. Department of Chemistry and Biochemistry, Ohio State University, Columbus, Ohio 43210,
United States
. Chemical and Engineering Materials Division, Neutron Sciences Directorate, Oak Ridge
National Laboratory, Oak Ridge, TN, 37831, USA
3
4
. Materials Science and Technology Division, Oak Ridge National Laboratory, Oak Ridge, TN,
3
7831, USA
Notice: This manuscript has been authored by UTꢀBattelle, LLC under Contract No. DEꢀAC05ꢀ
0OR22725 with the U.S. Department of Energy. The United States Government retains and the
0
publisher, by accepting the article for publication, acknowledges that the United States
Government retains a nonꢀexclusive, paidꢀup, irrevocable, worldꢀwide license to publish or
reproduce the published form of this manuscript, or allow others to do so, for United States
Government purposes. The Department of Energy will provide public access to these results of
federally sponsored research in accordance with the DOE Public Access Plan
(
http://energy.gov/downloads/doeꢀpublicꢀaccessꢀplan).
1
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Abstract:
This work aims to better understand the role of interfacial molecular structure that governs
selectivity and activity in heterogeneous catalytic reactions. To address this, a comprehensive
study of isopropanol conversion over an archetypal perovskite material, strontium titanate
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(SrTiO or STO), was performed with an array of techniques sensitive to orthogonal aspects of
3
the ensuing chemistry. Cubicꢀshape STO nanoparticles with only the (100) facet exposed were
synthesized and used to study the ensemble kinetic conversion of isopropanol over the surfaces,
which showed selectivity to form acetone, with minor propylene products appearing at elevated
temperatures. These results in combination with inelastic neutron scattering measurements
provide not only insight into the selectivity and overall activity of the catalysts but also low
frequency vibrational signatures of the adsorbed and reacted species. To compliment these
measurements, pristine thin films of STO (100) were synthesized and used in combination with
vibrational sum frequency generation spectroscopy to extract the absolute molecular orientation
of the adsorbed molecules at the interface. It was found that the isopropanol assumes an
orientation where the ꢀCH group points towards the STO surface; this preꢀreaction geometry
offers an obvious pathway to produce acetone by abstracting the alphaꢀproton and, thus, provides
a mechanistic explanation of selectivity at STO (100) surfaces. This insight opens up pathways
to explore and modify surfaces to tune the activity/selectivity though a molecular level
understanding of the reactions at the surface.
Keywords:
Perovskite, Strontium Titanate, Isopropanol, Neutron Spectroscopy, SFG
2
ACS Paragon Plus Environment

Page 3 of 34
ACS Catalysis
1
2
3
4
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
Introduction:
Understanding how catalyst structure and composition affect product distributions is a long
standing goal in the field of heterogeneous catalysis.[1ꢀ2] It is thought that with an understanding
of how structural organization promotes specific reactivity and selectivity, one could tune the
catalytic properties to yield the desired products.[3ꢀ4] In a heterogeneous catalytic system, the
initial microscopic mechanistic step involves the adsorption of a molecule onto a reactive
catalytic surface. A description of the composition, structure, and orientation of the adsorbed
molecules on the surface from this initial interfacial interaction can aid in the understanding of
how catalysts operate and differentiate selectivity at various surfaces.
In this work, we make use of the reactions of alcohols, which are known to generally proceed
via dehydration pathways to produce alkenes and water on acidic oxides or dehydrogenation
pathways on basic oxides producing aldehydes and hydrogen (see Scheme 1).[5ꢀ7] Recently, we
examined isopropanol adsorption on the surface of a model CeO catalyst that was synthesized
2
with only the (100) facets exposed. We found that isopropanol readily deprotonated at the CeO2
(100) surface, and the molecules appears to arrange to a ‘preꢀreaction’ geometry to allow for Hꢀ
abstraction from a methyl group at elevated temperatures.[8] While CeO drives isopropanol to
2
produce propylene upon heating through a dehydration mechanism, other metal oxides and
bimetallic oxide systems, such as strontium titanate (SrTiO , referred to here as STO), are known
3
to promote the production of acetone via dehydrogenation.[9] Despite this knowledge, the
mechanism underlying the conversion is not well understood, nor is there an obvious reason why
3
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the two oxide surfaces would uniquely drive chemistry in the way that they do. The question
is then: How do the structures and orientations of adsorbed molecules differ in these two
systems and what can that tell us about selectivity? To this end, this work aims to elucidate
what governs the specificity of catalysts and what chemistry is uniquely driven using a
combination of neutron scattering, kinetic measurements, and nonlinear optical spectroscopy.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 1. Reaction pathways of isoꢀpropanol over different catalysts.
Neutron spectroscopy is particularly a powerful approach to study heterogeneous
catalysis due to the penetrating power of neutrons and their unparalleled sensitivity to light
elements and hydrogen,[10] which is important to the reactions studied here. Neutron
scattering is an overall underutilized technique in the field of catalysis, due to its infamous
reputation for requiring large sample sizes and long collection times for measurements.
However, the neutron flux at the Spallation Neutron Source (SNS) at Oak Ridge National
Laboratory is 4000 times greater than any of its predecessors, which enables us to attain
higher resolution spectra in shorter amounts of time with smaller sample sizes than ever
before.[11] As we will show below, applying inelastic neutron scattering spectroscopy to
catalytic questions provides a wealth of information regarding the structure and dynamic of
adsorbates at the catalyst surface during the reaction. This information together with reaction
kinetics provide the macroscopic performance and selectively of STO nanocube catalysts,
which offer new insight into the origin of the activity and selectivity of the surfaces. To
4
ACS Paragon Plus Environment

Page 5 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
evaluate the molecular orientation at the interfaces, we employ vibrational sumꢀfrequency
generation (SFG), a surface specific nonlinear optical probe capable of providing absolute
molecular orientation at surfaces.[12ꢀ20] Through this comprehensive approach, we can directly
correlate the preꢀreaction orientations of isopropanol extracted via SFG and the
activity/selectivity of these nanoshape surfaces with the same exposed facets using neutron
scattering and kinetic measurements to understand the molecular scale events that drive
selectivity at catalytic interfaces.
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
Experimental Details:
Materials Synthesis:
The following chemicals were used as received: Strontium oxalate dihydrate (SrC O ꢁ2H O,
2
4
2
9
9%, Alfa Aesar), Titanium Oxideꢀanatase (TiO , 99.9%, Aldrich Chemical), Sodium Chloride
2
(NaCl, 99%, Fisher Chemical)
SrTiO cuboidal nanoparticles were synthesized via a modified molten salt process described
3
in Stair et al..[21] Briefly, a mixture of SrC O ꢁ2H O: TiO : NaCl in a 1:1:20 molar ratio was
2
4
2
2
rigorously ground in a mortar and a pestle and loaded into a crucible. The sample was heated to
50 °C in air for 4 hours followed by cooling down to room temperature. The resulting powder
8
was then washed with excess deionized H O in order to remove NaCl, and subsequently dried in
2
an oven at 110 °C overnight. Commercial STO powders (99%, Sigma Aldrich) were purchased
and used after calcination at 550 °C for 4 hours.
In order to obtain a wellꢀdefined and flat surface for nonlinear optical measurements, SrTiO3
(100) singleꢀcrystalline substrates were chemically and thermally treated. First the samples were
5
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
chemically etched in buffered hydrofluoric acid for 30 seconds to obtain a TiO ꢀterminated
2
surface, followed by thermal annealing at 1000 °C in air for 90 minutes to develop an atomically
smooth surface (see Supporting Information, Figure S1) characterized by atomic force
microscopy (AFM) measurement.[22]
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Characterization Methods:
XRD patterns were measured with an X’pert Pro Xꢀray diffractometer from PANalytical
with nickelꢀfiltered Cu Kα radiation (λ = 1.5418 Å), with generator setting of 40 mA and 45
kV. The scanning region was set to be 10 ꢀ 90° with a step size of 0.025°.
Specific surface area was determined by N adsorptionꢀdesorption isotherms carried out
2
with a Quantachrome 1C Autosorb. About 0.1 g of sample was outgassed with helium at
150 °C for 5 h.
Scanning electron microscope (SEM) was performed using a Hitachi HDꢀ
000 equipment operating at 200 kV. Samples were loaded onto a 300 mesh copper grid with
2
lacey carbon backing.
Atomic force microscopy (AFM) was used to examine the surface morphology of (100)
SrTiO singleꢀcrystalline substrates. An atomically flat surface was observed with a rootꢀ
3
meanꢀsquare (RMS) roughness value below 0.1 nm.
Inelastic neutron scattering (INS) measurements were carried out on beamline 16B, the
Vibrational Spectrometer (VISION) at the Spallation Neutron Source, at Oak Ridge National
Laboratory. VISION is a highꢀflux highꢀthroughput INS instrument covering a broad energy
range from ꢀ2 to 1000 meV.[23] In these experiments, about 3 g of sample was loaded into a
homeꢀbuilt cylindrical quartz reactor. The reactor was sealed and heated under vacuum at
6
ACS Paragon Plus Environment

Page 7 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
ꢀ6
1
20 °C overnight until the measured pressure was below 1×10 mbar to remove air and loosely
bound surface H O prior the measurement. The INS spectra of pure materials, including the
2
nanocatalyst, bulk phase of Isopropanol (IPA)/acetone/propylene were measured by adding
sample into the empty reactor. Adsorbates were added to the catalysts held at 30 K within a time
span of 5 ꢀ 10 min. The adsorbate was dosed via a capillary into the sample can using a homeꢀ
built gasꢀmanifold. Our INS measurements were carried out at 5 K and started roughly 20 min
after loading. Additional experimental conditions are provided specifically in the Results section.
To isolate the adsorption and subsequent reactions of molecules at the nanocube surfaces, the
background, which is the signal of the pure materials, were subtracted. INS spectra were
collected with neutron energy loss up to energy transfer values of 1000 meV, corresponding to
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
ꢀ1
about 8000 cm .
Sumꢀfrequency generation (SFG) spectroscopy was performed using a homeꢀbuilt
spectrometer described in detail elsewhere.[8, 24ꢀ25] Briefly, isopropanol was drop cast on
pristine STO (100) surface allowed to dry in atmosphere before SFG measurements commenced.
Adsorption of species from the atmosphere was previously found to not impact the spectra or
orientational results in a meaningful way.[8] The SFG spectra were collected in three
polarization combinations denoted as SSP, PPP, and SPS. Following convention, the
polarization combinations describe the radiated SFG (Sꢀ), incident nearꢀinfrared (Sꢀ), and midꢀIR
pulses (Pꢀ), light fields, respectively. Transition frequencies and amplitudes that characterize the
molecules at the surface are extracted by fitting the data to equation 1:[26]
ꢓ
ꢈꢉ
= ꢄꢅꢆꢇ + ∑ꢑꢒ
ꢀꢁꢂꢃ
ꢄ
(eq. 1)
ꢊꢋꢌꢍꢊ ꢎꢏꢐ
ꢉ
ꢉ
where ISFG represents the measured SFG signal, ANR is a nonꢀresonant contribution to the SFG
th
signal, A is the amplitude of the q ꢀresonance, ω and ω are the frequencies of the incident
q
IR
q
7
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
light and the vibrational resonance, respectively, and Γ is the linewidth. Fitting to more
q
complex functions, such as a Voigt profile, was not attempted due to the increased number of
freely adjustable parameters which would result in over fitting the data. [27]
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Activity Tests:
Isopropanol (IPA) conversion reactions were carried out in a 1/4ꢀinch diameter tubular
quartz reactor. In a typical run, ca. 30 mg catalysts was mixed with 370 mg 60ꢀ80 mesh
quartz sand and loaded in the reactor. The catalyst bed was held by quartz wool at both ends.
The reactor was inserted to CarboliteꢀGero furnace (MTF10ꢀ15ꢀ130), and purged with 50
sccm Ar flow for 0.5 h, followed by heating to desired temperature (i.e., 300 °C). For the
activation treatment procedure, a 50 sccm, 10 % O /Ar flow was applied for 1 h at 550 °C,
2
followed by 50 sccm Ar flow purge for 5 min. The IPA was injected to the heated gas line
with syringe pump (The New Era Pump Systems Inc.) at various rates. The gas line was
wrapped with heating cord, which is set to 95 °C in order to ensure the IPA stays fully
vaporized throughout the path to the reactor.
The samples were analyzed on line by a Buck Scientific 910 gas chromatograph equipped
with a Restek Packed Column (15% CWꢀ20M, 6 ft, 2 mm I.D., 1/8 in O.D.) and a flame
ionization detector (FID), respectively. After the reaction, the catalyst was purged with Ar,
cooled to room temperature, removed from the reactor and saved in a glass vial for additional
characterization.
The conversion and selectivity were calculated by using the following equations:
∑
ꢠꢡ,ꢢꢣꢤ
ꢀ
ꢔꢅꢕꢖꢗꢘꢙꢚꢛꢜꢝꢗꢘꢕꢞ%ꢟ =
× 100%
(eq.2)
∑ꢠꢡ,ꢢꢣꢤ ꢎꢥꢦꢈꢢꢣꢤ
ꢠꢡ,ꢢꢣꢤ
ꢧꢚꢨꢚꢩꢪꢝꢙꢝꢪꢫꢕꢞ%ꢟ = ∑ꢠꢡ,ꢢꢣꢤ × 100%
(eq. 3)
8
ACS Paragon Plus Environment

Page 9 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
th
whereas Xi,out refers to concentration of the i product (i.e., CO , propylene, acetone), and IPA
2
out
is the concentration of IPA measured in the exit port.
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
Results and Discussion
Figure 1 shows the XRD patterns of the synthesized and commercial STO catalysts. It can be
seen that for the cubic STO sample, peaks were apparent at 32.6°, 39.9°, 46.5°, 57.9°, 67.9°, and
77.2°, which are assigned to SrTiO (JCPDS: 35ꢀ0734). It should be mentioned that a
3
commercial STO sample shows almost identical XRD peaks, except one shoulder peak at ca.
31.6°. This is possibly due to presence of other phase, such as Sr TiO (JCPDS: 39ꢀ1471) and
2
4
Sr Ti O (JSCPD: 11ꢀ0663).
3
2
7
cubic STO Pre
cubic STO Post
commercial STO Pre
commercial STO Post
(
): SrTiO , JCPDS: 35-0734
3
1
5000
[
]: Carbon, JCPDS: 26-1080
(
110)
(
200)
(
211)
(111)
(
220)
(
100)
(310)
(
210)
[
004]
20
30
40
50
2
60
70
80
ο
θ τ τ
Figure 1. XRD patterns of synthesized cubic STO and commercial STO, before and after
reactions.
9
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The morphology of STO samples was determined by SEM. As can be seen in Figure 2,
the as synthesized STO sample has a roughly cubic shape, with ca. 100 ꢀ 200 nm diameter
(Figure 2a). It should be mentioned that cubic STO samples have primarily (100) face
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
exposed due to the low surface energy with only a small fraction of (110) sites at the
corner.[21, 28] In contrast, the commercial STO powder has inhomogeneity particle sizes
with irregular shapes. These particles are without wellꢀdefined surface facets exposed (Figure
2c).
(
a)
(b)
(d)
(c)
Figure 2. SEM images of synthesized cubic STO sample (a) before and (b) after 6 h reaction,
and commercial STO sample (c) before and (d) after 6 h reaction.
The textural properties (i.e., surface area etc.) of the commercial and synthesized cubic
STO, are determined from N ꢀphysisorption isotherms (Figure 3). In short, both the
2
10
ACS Paragon Plus Environment

Page 11 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
commercial and synthesized STO are nonporous material. However, the surface area of the
2
commercial sample is over 3ꢀfold larger than the cubic STO (24 vs. 7 m /g). As the surface area
of STO is strongly structureꢀdependent, the low value for cubic STO is likely due to the flat
exposed facets, whereas the irregular and rough surfaces of commercial STO lead to a high value.
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
As depicted in Figure 3, both these materials show very low adsorption over a wide relative
pressure regime. Such phenomenon indicates that there is no identifiable monolayer formation
and show stronger adsorbateꢀadsorbate interaction rather than adsorbateꢀadsorbent interaction.
Hence, the isotherms of the STO samples can be identified as type III, which is characteristic of
nonporous materials.[29] In addition, type H3 hysteresis loops, according to IUPAC, were
detected in all catalysts. This type of loop is characteristic of metastability of the adsorbed
multilayer and delayed capillary condensation.[30]
180
STO-Commercial
STO-Cubic
1
1
1
1
60
40
20
00
8
6
4
2
0
0
0
0
0
0.0
0.2
0.4
0.6
0.8
1.0
P/P0
Figure 3. N ꢀphysisorption isotherms of the commercial and cubic SrTiO samples.
2
3
11
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Prior to the catalytic tests, control experiments were performed in an empty tube, and
over quartz sand under the same reaction conditions. All experiments had ca. 1.2 s residence
times. The background IPA conversion was negligible.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The catalytic performance of cubic and commercial STO samples with/without activation
of catalyst is shown in Figure 4. To start, we consider the conversion of IPA over the
synthesized cubic nanocatalysts at 280 °C without activation, as shown in Figure 4a (black
solid symbol). Here, we find that the conversion is ca. 2 %, with the acetone being the only
product (Figure 4b, black solid symbol being overlapped by black open symbol). As the
temperature increased to 300 °C and 320 °C, the conversion rises to 6 % and 8 %,
respectively (solid blue and green symbols in Figure 4a), for the cubic nanocatalysts. In all
cases, acetone was observed as the major product (Figure 4b) with up to 10 % propylene
selectivity (Figure 4c) at the highest temperatures (320 °C). The change in catalytic
selectivity at elevated temperatures indicates an increase of the dehydration pathway at
elevated temperatures, where the intermediate formed at the STO surface can be transformed
to propylene. In contrast, the catalytic performance of commercial STO was found to yield
roughly the same conversion level as compared to the inactivated cubic STO, but remarkably
different selectivities at 300 °C. To be specific, the commercial STO sample shows ca. 63 %
acetone and 37 % propylene, while cubic STO shows 94 % acetone, and 6 % propylene.
Taking into the account that the surface area of commercial STO is ca. 3 times larger than
that of cubic STO, it indicates that the cubic sample exhibits superior activity in at least 3ꢀ
fold and is highly acetoneꢀselective. Taking the morphology into account, the differences in
selectivity also point to the role of different facets in driving the chemistry to form acetone at
12
ACS Paragon Plus Environment

Page 13 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
the (100) facets in the nanocube samples, whereas the presence of different facets on the more
rounded edges of the commercial sample likely results in different chemical environments that
uniquely drive the chemistry and thus explains the breadth of the product distribution observed.
As a comparison, in a separate experiment, an O ꢀactivtion step (50 sccm 10% O /Ar flow
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
2
2
for 1 h at 550 °C, followed by purged with 50 sccm Ar flow for 10 min) was applied to the
catalysts prior the reaction, and the data was also shown in Figure 4 (legends with “ꢀAct”). It
should be also mentioned that, for cubic STO samples, such activation could enhance the
conversion and acetone selectivity. However, the IPA conversion rapidly dropped to steady level
within the first 60 ꢀ 120 min. This could be explained by the quick consumption of surface
oxygen species as well as residual O in system. The higher level of acetone selectivity is
2
possibly due to the enhancement of dehydrogenation pathway in the “oxidative environment” as
proposed in Vedrine’s work.[31] However, a detail study on the effect of such an “activation step”
is still needed.
3
0
8
(
a)
2
o
2
3
3
80 C (Cubic-Inact)
26
24
o
00 C (Cubic-Inact)
o
20 C (Cubic-Inact)
o
22
20
18
16
280 C (Cubic-Act)
o
300 C (Cubic-Act)
o
320 C (Cubic-Act)
o
300 C (Commercial-Inact)
14
12
1
0
8
6
4
2
0
50
100
150
200
250
300
350
400
Time (min)
13
ACS Paragon Plus Environment

ACS Catalysis
Page 14 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
b)
1
00
90
o
2
80 C (Cubic-Inact)
o
300 C (Cubic-Inact)
8
7
6
0
0
0
o
3
20 C (Cubic-Inact)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
o
2
80 C (Cubic-Act)
o
300 C (Cubic-Act)
o
320 C (Cubic-Act)
o
3
00 C (Commercial-Inact)
0
50
100
150
200
Time (min)
250
300
350
400
(
c)
40
o
280 C (Cubic-Inact)
3
2
1
0
0
0
0
o
3
3
2
00 C (Cubic-Inact)
o
20 C (Cubic-Inact)
o
80 C (Cubic-Act)
o
300 C (Cubic-Act)
o
3
20 C (Cubic-Act)
o
3
00 C (Commercial-Inact)
0
50
100
150
200
Time (min)
250
300
350
400
Figure 4. (a) IPA conversion, (b) Acetone selectivity and (c) Propylene selectivity over the
synthesized cubic and commercial STO catalysts. Reaction conditions: 30 mg catalysts mixed
with 370 mg quartz sand, 1µL/min IPA injection rate, 50 sccm Ar flow, and 1 atm total pressure.
The reacted samples were examined after experiments to see how the chemistry impacted
the structure and composition of the materials. Accordingly, both cubic and commercial
STO sample were reꢀcharacterized (data given in Figure 1). The results indicate that the
synthesized nanocube STO samples remain unchanged based on observation of the peaks at
the same positions, whereas the commercial samples show new XRD peaks, indicative of the
1
4
ACS Paragon Plus Environment

Page 15 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
possible deposition of carbon (JCPDS: 26ꢀ1080) on the surface (see arrow in Figure 1), which is
likely formed during the reaction. The samples were also reꢀexamined by SEM for structural
comparison. As shown in Figure 2b and 2d, these results reveal that the synthesized STO
remained nearly ideally cubic with slightly truncated corners, whereas the commercial STO
powder shows inhomogeneous particle sizes and shapes.
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
The temperature effect on the catalytic performance was also studied and the results were
shown in Figure 5. As can be seen, regardless of pretreatment, increasing temperature from
280 °C to 320 °C, leads to an increase of both conversion and propylene selectivity, while
acetone selectivity suffers a slight drop for the same temperature change. It should be mentioned
that, each point is the average of nine measurements. There is generally less than 5% difference
of the original value due to systematic error. Again, the activated samples exhibit a higher
conversion and acetone selectivity but a lower propylene selectivity than that of inactivated
sample, which is consistent with the observation depicted in Figure 4.
110
100
90
Avg. Conv. (inactivated)
Avg. Conv. (activated)
Avg. Prop. Sel. (inactivated)
Avg. Prop. Sel. (activated)
Avg. Ace. Sel. (inactivated)
Avg. Ace. Sel. (activated)
8
1
0
2
9
6
3
0
2
70
280
290
300
310
320
330
o
Temperature ( C)
15
ACS Paragon Plus Environment

ACS Catalysis
Page 16 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 5. Temperature effect on IPA (squares) conversion, Propylene (triangles) and Acetone
(diamonds) selectivity over inactivated (solid symbols) and activated (open symbols) cubic STO
catalyst. Reaction conditions: 30 mg catalysts mixed with 370 mg quartz sand, 1 µL/min IPA
injection rate, 50 sccm Ar flow, and 1 atm total pressure.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In the kinetic analysis of a gas phaseꢀinvolved reaction, partial pressure could
conveniently substitute for concentration at constant temperature. Accordingly, the effect of
IPA partial pressure (due to different IPA injection rate) over commercial/synthesized STO
was also investigated and the results are summarized in Figure 6. It can be seen that an
increase of partial pressure leads to a drop in conversion in both cases. This might be due to
the increase of WHSV (weight hour space velocity). Once the partial pressure is above 1000
Pa, the conversion reaches a steady level. Meanwhile, the acetone selectivity significantly
increases from 48 % to 68 % within the partial pressure range for the commercial STO
sample, while propylene selectivity decreases. The observed effect is not so obvious with
respect to selectivity over cubic STO, although a slight change is seen. Such a phenomenon
suggests that high population of IPA at the catalyst surface (high partial pressure) leads to an
increase of dehydrogenation pathway (to produce acetone). With the current data alone, one
cannot conclude a simple reason that causes such behavior. For one reason, it is difficult to
quantitatively measure the coverage of IPA over STO surface. In addition, the generated
interꢀmolecular interaction (e.g., Hꢀbonding) may also change the surface orientation.
However, it is commonly accepted that the selectivity is somehow coverage dependent.[32ꢀ
34] In this case, an increasing of injection rate leads to IPA to be more packing at the surface
and overall different possible orientations that the functional groups can take at the surface.
16
ACS Paragon Plus Environment

Page 17 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
Consequently, as such, the branching between dehydration and dehydrogenation pathways could
potentially change as a function of IPA coverage. This remains an open topic for continued
investigation at controlled pressures.
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
Wachs and coꢀworkers have investigated methanol and isopropanol conversion over a variety
of metal oxides, and found that the average active site density of SrO and TiO (the two possible
2
2
termination surfaces of STO) to be 4.3 and 3.7 µmol/m , respectively.[35ꢀ36] Taking the surface
area into account, the turnover frequency (TOF) can be determined by normalizing the activity
under differential conversion regime (< 15%). Hence, the TOF at each IPA partial pressure under
3
00 °C was calculated from Figure 6. The best TOFs for IPA conversion for cubic and
ꢀ1
commercial STO are 0.0143 and 0.0059 s , respectively (IPA partial pressure is 1934 Pa).
(
a)
Cubic-Inact (Conv.)
12
Commercial-Inact (Conv.)
1
0
8
6
4
2
0
500
1000
1500
2000
Partial Pressure (Pa)
17
ACS Paragon Plus Environment

ACS Catalysis
Page 18 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
00
(
b)
9
0
80
7
6
5
0
0
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Cubic-Inact (Prop. sel.)
Cubic-Inact (Ace. sel.)
Commercial-Inact (Prop. sel.)
Commercial-Inact (Ace. sel.)
40
3
2
0
0
10
0
0
500
1000
1500
2000
Partial Pressure (Pa)
Figure 6. Partial pressure effect on (a) IPA conversion, and (b) Propylene (circles) and Acetone
triangles) selectivity over inactivated cubic STO (solid symbols) and commercial STO (open
(
symbols). Reaction conditions: 30 mg catalyst mixed with 370 mg quartz sand, 0.25 ~ 3 µL/min
IPA injection rate, 50 sccm Ar flow, 300 °C, and 1 atm total pressure.
A plot of logarithm pressure versus logarithm rate allows us to determine the overall
order of the reaction (the slope of linear fitting equation), as well as to identify individual
production formation channel. As can be seen from Figure 7, the apparent reaction order of
IPA, acetone, and propylene over the cubic STO at 300 °C is 0.49, 0.52, and 0.3, respectively.
While the corresponding values for the commercial STO case are 0.36, 0.49, and 0.17,
respectively. Although the observed reaction orders are slightly different for both type of
catalysts, the reaction orders for IPA consumption, as well as acetone formation order
appears to be approximately 0.5. There is certain debate regarding the reaction mechanism
(
derived from the reaction order) of isopropanol decomposition. For example, Vedrine
18
ACS Paragon Plus Environment

Page 19 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
proposed a LangmuirꢀHinshelwood kinetics for dehydration of IPA over Nb O , based on the
2
5
observation of propylene reaction order being 1 at low partial pressure, and 0 at high partial
pressures (> 2 kPa).[31] In contrast, Lorenzelli et al. proposed a Marsꢀvan Krevelen mechanism
of IPA oxidation over Co O and MgCr O , with evidence of nucleophilic lattice oxygen being
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
3
4
2
4
involved in the reaction.[37] Moreover, Barteau and coꢀworkers found a 0.5 order with respect to
IPA up to 16 kPa partial pressure over anatase TiO .[38] Notice that most studies are over single
2
metal oxides, and only few reports have been focused on decomposition of vapor phase IPA over
perovskite materials.[39]
-14.5
(a)
-15.0
2
IPA, y=0.49x-18.46, R =0.99
-
-
-
-
-
-
15.5
16.0
16.5
17.0
17.5
18.0
2
Acetone, y=0.52x-18.87, R =0.99
2
Propylene, y=0.3x-19.3, R =0.945
4
.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
ln (IPA partial pressure, Pa)
19
ACS Paragon Plus Environment

ACS Catalysis
Page 20 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
-
14.0
(
b)
-14.5
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
IPA, y=0.36x-17.3, R =0.93
-15.0
2
-15.5 Acetone, y=0.49x-18.67, R =0.96
-16.0
2
Propylene, y=0.17x-16.9, R =0.79
-16.5
4
.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
ln (IPA partial pressure, Pa)
Figure 7. Reaction order of IPA (squares), Propylene (circles) and Acetone (triangles) over (a)
cubic STO and (b) commercial STO. Reaction conditions: 30 mg catalyst mixed with 370 mg
quartz sand, 0.25 ~ 3 µL/min IPA injection rate, 50 sccm Ar flow, 300 °C, and 1 atm total
pressure.
Since neutron scattering methods are exquisitely sensitive to hydrogen containing
molecules and vibrations, INS serves as an ideal platform for the chemistry study of
hydrocarbons over a reactive nanocatalysts. Here, we employ INS to measure the surface
chemistry during IPA conversion at cubic STO surfaces. Figure 8 shows the spectrum of IPA
adsorbed on cubic STO surface. Before the adsorption experiment, the overall signals from
bulk phase of STO and IPA, including the sample container and equipment (Aluminum)
background were measured separately as references and were subtracted from the IPA
adsorption spectra. It can be seen that the bulk IPA gives rise to several characteristic peaks
ꢀ1
at positions of ca. 239, 280, 373, 492, and 934 cm . The detail vibrational modes are
20
ACS Paragon Plus Environment

Page 21 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
described in Supporting Information (Figure S2). In contrast pure cubic STO does not show
peaks at these positions. For the spectrum acquired with 10 µL IPA adsorbed on STO at 120 °C
(already subtracted the bulk STO signal), these characteristic bands due to IPA were observed,
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
indicating surface adsorption of isopropanol.
2.0
Bulk IPA
Bulk STO
IPA ads. on STO
1
1
1
.8
.6
.4
1.2
1
0
.0
.8
0.6
0
0
0
.4
.2
.0
0
200
400
600
800
1000
1200
1400
1600
-
1
Wavenumber (cm )
Figure 8. Neutron Vibrational Spectroscopy (NVS) of IPA adsorption on cubic STO.
Next, to study the conversion of IPA over STO nanocatalysts with INS (see Figure 9) we first
recorded the INS spectrum from adsorption of possible products (i.e., Propylene (Prop), Acetone
(Ace), and Isopropyl Ether (IPE)) to the STO nanocatalysts. Notably, the peak observed at 373
ꢀ1
cm is quite unique for IPA, which is absent in product spectra. As such, this band can be used
to verify reaction occurrence and IPA consumption. For a reaction taking place at 200 °C for 5
min, the featured bands of IPA remain at a comparable level with respect to the standard
ꢀ1
spectrum. However, more notably, 3 new bands appear at ca. 116, 180, and 330 cm . They are
attributed to propylene/acetone, acetone, and propylene/acetone, respectively. Such observation
2
1
ACS Paragon Plus Environment

ACS Catalysis
Page 22 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
indicates the occurrence of a reaction. Further heating the sample to 300 °C for an additional
ꢀ1
5
min leads to a notable intensity decrease for the peak at 373 cm , suggesting a further
conversion of IPA. This peak continues to decrease as the reaction proceeds (see spectrum
with another 10 min treatment at 300 °C). Meanwhile, the intensity of the bands at 116, 180,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ꢀ1
and 330 cm increase. Taking into the account the activity results shown above, which
suggests occurrence of certain conversion between 280ꢀ320 °C, the new bands measured in
the INS spectra are likely due to the formation of acetone that is adsorbed to the surfaces of
the nanocatalysts. While INS is able to probe the chemical species at the catalyst interface,
molecular level information as to the structure and orientation of the adsorbed molecules at
ambient conditions is difficult to extract from these measurements alone.
2
.5
2
.0
2
0 min
0 min
min
1
1.5
5
IPA ads.
IPE ads.
1
.0
0.5
Prop ads.
Ace ads.
0
.0
0
200
400
600-1
800
1000
Wavenumber (cm )
Figure 9. Neutron Vibrational Spectroscopy (NVS) of IPA reaction over STO. The bulk STO
background has been subtracted from all spectra. Reaction temperature: 200 ꢀ 300 °C,
Spectroscopic measurement temperature: ꢀ268 °C.
22
ACS Paragon Plus Environment

Page 23 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
In order to gain insights into catalytic activity/selectivity over cubic STO surface, the
adsorption behavior was investigated by means of SFG in Figure 10. Flat interfaces were used in
SFG measurements to allow for conventional orientational analyses to be applied to the
measured data to extract the absolute molecular orientation of IPA at STO (100) surfaces.
Studies of powdered materials would not allow such a direct analysis due to the symmetry of the
particles and associated complex nonlinear scattering phenomena.[40] However, since the
nanoshapes are predominantly (100) faceted, and the flat STO surface is also (100), the
conclusions and orientational information drawn from the pristine flat interface can be directly
applied to the nanoparticles. Building on recent Raman measurements, SFG experiments with
theoretical support and polarization selection rules, the resolved bands for isopropanol adsorbed
to STO (100) surfaces are assigned. Notably, the spectra observed on STO (100) surfaces are
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
substantially different from isopropanol on CeO (100) interfaces in both the number of observed
2
bands as well as the appearance of the spectra. More specifically, in the SSP spectrum many of
the features appear as dips, which are common in SFG measurements of molecules adsorbed
onto the metallic interfaces and are indicative of a different orientation of isopropanol at the
surface of STO vs. CeO . With respect to the SFG spectrum of bare STO, it is well known that in
2
broadband SFG measurements the response of any nonꢀresonant interface matches that of the
driving IR spectrum. As such, the response of the bare STO surface is simply a Gaussian like
spectrum that is scaled out during analysis. Hence, we use a gold film as a reference (see
Supporting Information, Figure S3) simply because it gives a larger signal with better signalꢀtoꢀ
noise than the bare STO itself. By scaling the spectrum by a reference, the wavelength
dependent intensity is intrinsically removed.
23
ACS Paragon Plus Environment

ACS Catalysis
Page 24 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Similar to our previous work, bands a and f are assigned to the –CH stretch of isopropanol in
agreement with previous DFT calculations[8] and high resolution Raman measurements.[41]
ꢀ1
We note that the prominent bands at 2943 cm observed in all three polarization combinations
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ꢀ
(
bands d, j, and k) are readily attributed to the CH ꢀasymmetric stretching (CH ꢀas, r ) modes of
3
3
deprotonated isopropanol, in agreement with work on CeO (100) surfaces.[8] However, a small
2
ꢀ1
feature at ~2960 cm (bands e and l) is also observed and is in good agreement with the location
of the CH ꢀas stretch of protonated isopropanol.[13, 42ꢀ43] This observation points towards the
3
coexistence of both protonated and deprotonated isopropanol species at STO (100) surfaces at
room temperature. This also means that the STO (100) surface is less basic than the CeO (100)
2
surface, where isopropanol was completely deprotonated. These assignments follow selection
rules[12, 42] for asymmetric stretching modes and are in excellent agreement with our previous
experiments, analysis, and DFT calculations[8] as well as other literature probing neat
isopropanol interfaces.[13, 42ꢀ43] Noting that two species are in coexistence at the surface
greatly complicates the assignment of bands in the middle of the spectra since features
corresponding to the protonated and deprotonated species will overlap. As such, we can only
attribute bands b, c, g, h, and i to some combination of Fermi resonances, CH ꢀsymmetric
3
stretching, and –OH stretches in a shared Hꢀbond geometry.[8] It is important to note that in
comparing these results to CeO , we achieve two important goals: 1) we show agreement with
2
previous result in terms of band positions for the deprotonated species, while showing the
existence of a protonated species, and 2) we show how the two catalysts display different
selectivity and how this difference corresponds with the absolute molecular orientation.
24
ACS Paragon Plus Environment

Page 25 of 34
ACS Catalysis
1
2
3
4
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 10. SFG spectra of isopropanol at STO (100) interfaces collected in the SSP (a), PPP (b),
and SPS (c) polarization combinations. Bands are labeled as letters from aꢀl.
Having assigned and resolved vibrations from two well defined functional groups on the
deprotonated isopropanol structure, we can use the extracted amplitudes to determine the
orientations of –CH and –C(CH ) functional groups can be determined using the frame work
3
2
described in our previous work. Briefly, Fresnel factors for the incoming and outgoing light
fields were calculated[12, 14, 44] assuming an interfacial index of refraction of 1.18. The choice
of the interfacial index to be 1.18 is supported by excellent agreement our previous results on
CeO with theory, along with other reports where 1.18 is a commonly used interfacial index of
2
refraction value for organic species far from electronic resonance.[12, 45ꢀ46] The indices of
refraction for STO were taken to be 2.38, 2.34, and 2.21 for the SFG, NIR and IR light,
respectively.[47] Using the intensity ratio of –CH amplitudes obtained in the SSP and PPP
spectra the –CH bond angle with respect to the surface normal was determined to be ꢬꢭꢮ = 51 ±
8° angle with respect to the surface normal. The methyl groups are treated in by the unified atom
model[8, 13] effectively reducing the tetrahedral geometry surrounding the central carbon of
isopropanol to trigonal planer structure having −CH, −CO, and −C(CH ) groups. Inserting
3
2
measured intensity ratios for the SSP, PPP, and SPS polarization combinations as extracted for
25
ACS Paragon Plus Environment

ACS Catalysis
Page 26 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the methyl asymmetric stretch, we can obtain a family of solutions describing possible twistꢀtilt
angle combinations that the methyl groups could possibly take given the measured data. The
correct twist/tilt angle pair is determined using the constraint that the angle between the –CH and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
–
C(CH ) groups is 120° and that the measured orientational angle of the –CH group be satisfied.
3 2
Here we assume the –CH group is pointed downwards towards the surface such that the angle the
CH takes is 90° + 51° = 141° and is supported by the dramatically different SFG signals,
–
particularly in the SSP spectrum, observed here vs. at CeO (100) surfaces, which is indicative of
2
differing upꢀdown orientations of the functional groups in isopropanol.[46] Applying these
constraints, we arrive at the dashed red lines in Figure 11. The intersection area of the dashed
red lines and solid blue lines represent the allowed angles the methyl groups can take that
simultaneously satisfy the measured –CH bond angle with respect to the surface normal. From
this analysis, the average orientational twist and tilt angles for the –C(CH ) group are found to
3
2
be ψ = 35 ± 15° and θ = 29 ± 7°, respectively. The somewhat large uncertainties in these angles
are a result of challenges in fitting the SFG spectra in the presence of strong nonꢀresonant
contributions with multiple overlapping bands from different chemical species.
26
ACS Paragon Plus Environment

Page 27 of 34
ACS Catalysis
1
2
3
4
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 11. Intersection of possible methyl twist/tilt angles and constraints as defined in the text is
used to uniquely determine the orientation of methyl groups. The intersection of these two
regions represents the unique simultaneous solution for the measured intensities in SFG
experiments.
The absolute molecular orientation is obtained by first defining the molecular orientation
such that the –C(CH ) group is pointed in the +Z direction as shown in Figure. 12a.[8, 48] Blue
3
2
arrows are unit vectors pointing along the various bond axes/bisector that in turn define the
molecular plane. The coordinates of the −C(CH ) group are represented by the direction cosine
3 2
ꢱ
ꢱ
ꢱ
) = (0, 0, 1) whereas the –CH group is given as (ꢨꢱ , ꢨꢱ , ꢕꢨ ꢭꢱ ꢮ,ꢳ) = (cos(ꢀ
(ꢨ
, ꢨ
, ꢕꢨ
ꢭꢮꢯ,ꢰ ꢭꢮꢯ,ꢲ
ꢭꢮꢯ,ꢳ
ꢭꢮ,ꢰ ꢭꢮ,ꢲ
30°), 0, cos(120°)). Next, two consecutive rotations are applied, the first being ψ = 35° twist
about the Zꢀaxis lied followed by tilting by θ = 29°, about the Yꢀaxis (sketched in Figure 12a).
The laboratory frame direction cosines for the –C(CH ) and –CH group are (cos(61°), cos(90°),
3
2
cos(29°)) and (cos(68°), cos(60°), cos(141°)), respectively. As necessary, the projection of the –
CH bond angle with respect to the normal is in agreement with independent orientational
27
ACS Paragon Plus Environment

ACS Catalysis
Page 28 of 34
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
analysis. The absolute molecular orientation is defined as the angle of the molecular normal
(green arrow in Figure 12b) with respect to the surface normal, which is the cross product of
the rotated –C(CH ) and –CH direction cosines to obtain a new vector normal to the
3
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
molecular plane. The projection of this vector onto the laboratory Zꢀaxis then provides the
absolute orientational angle,[8, 48] which is determined to be φ = 74°, as shown in Figure
12b for isopropanol at STO (100) interfaces.
Figure 12. Rotation of isopropanol to the laboratory frame. Angles between the surface normal
and various unit vectors are shown as described in the text.
Insight into Catalytic Selectivity
Since there is a 1:1 correspondence with the exposed (100) facets on the nanoparticles
and the (100) flat surface, we can rationalize the measured selectivity from kinetic
measurements though a molecular level picture measured with SFG. Our kinetic results in
Figure 4 clearly show that STO nanocubes with exposed (100) facets convert isopropanol to
acetone as the dominant product at elevated temperatures. The question is then why does
STO (100) preferentially form this product? This is addressed by turning to SFG data and
28
ACS Paragon Plus Environment

Page 29 of 34
ACS Catalysis
1
2
3
4
5
6
7
8
9
the orientational information extracted from it. Hereby, we propose a reaction mechanism of
isopropanol over STO (100) surface as illustrated in Scheme 2. Mechanistically, we find that the
first step is that on adsorption, isopropanol exists in an equilibrium between a protonated and a
deprotonated species illustrated in step a) of Scheme 2. This is evidenced by INS data (Figure 8)
and the existence of two species in the SFG spectrum (Figure 10). The deprotonated species
assumes an orientation at the interface such that the methine (ꢀCH) group is pointed towards the
STO surface as sketched in step Scheme 2b). Since we know from kinetic measurements that
acetone is produced as the dominant product and that abstraction of the hydrogen atom from the
central carbon in isopropanol has to occur to form acetone, it is reasonable to consider that at
elevated temperatures, the energetic barrier for abstraction of this proton is overcome at elevated
temperatures (step c) and thus producing acetone at the interface, which can subsequently desorb
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
(step d). This scheme is supported by INS measurements showing the existence of acetone at
STO (100) nanocube surfaces post reaction. This also agrees with our recent work on CeO (100)
2
surfaces, where the dominant product is known to be propylene and the adsorbed isopropanol
molecules are assumed a different orientation, in which methyl groups are in closer proximity to
the surface.[8] For the case of CeO , we showed previously that the molecule is arranged such
2
that a proton from a methyl group is pointed towards the CeO (100). Notably, isopropanol
2
assumes a preꢀreaction geometry where this critical proton is pointed towards the STO (100)
interface in an analogous way to the results previously reported by our group for CeO (100).
2
Through the combination of these complimentary techniques, we have shown how simple
structural arguments can explain the selectivity in heterogeneous catalysis of wellꢀdefined
systems. Thus, these results show clearly that the selectivity of the catalyst is at least in part
dictated by the adsorption geometry of the molecule to the interface of interest. This is a
29
ACS Paragon Plus Environment