ACS Catalysis
Research Article
solutions. Raman spectra of the dehydrated catalysts were
obtained from catalyst pellets placed in an environmentally
controlled high-temperature cell reactor (Linkam, CCR1000).
Before analysis, the catalysts were dehydrated at 823 K for 1 h
aliquots were analyzed using a gas chromatograph (Agilent
7890A) equipped with a flame ionization detector (FID).
Reactants and products were separated in a column (HP-
INNOWax, 30 m × 0.53 mm i.d., 1.0 μm film thickness). Peaks
within the gas chromatograms were identified by comparisons
to mixtures of known standards and with a mass-selective
detector (MSD, Agilent, 5975B) operating with electron
ionization. The alkene conversion and product selectivity
were calculated as follows
3
−1
in flowing O (Airgas, ultrahigh purity, 20 cm min ) to
2
desorb the adsorbed moisture. The spectra were collected after
cooling the catalyst to room temperature in flowing O gas (20
2
3
−1
cm ·min ). Spectra were acquired with a Raman spectrometer
Renishaw, inVia) equipped with a 532 nm laser that was less
than 41 mW·μm at the catalyst surface. The power density
was measured directly using a portable power meter (Gentec-
EO, PRONTO-Sl). Experimental spectra represent the average
of 1000 accumulations with 0.1 s per accumulation.
(
−
2
moles of alkene reacted
alkene conversion (%) =
× 100
moles of alkene fed
The Raman spectra of the catalysts under the reaction
conditions were taken within a custom liquid flow cell. The
pelletized catalysts were placed in this cell, immersed in
moles of C in product
moles of C in all products
selectivity (%) =
× 100
flowing acetonitrile solutions (CH CN, Fischer Chemicals,
HPLC grade), and heated to 313 K. Steady-state spectra of the
reactive intermediates were collected (0.1 s per accumulation,
3
Analysis of pseudosteady-state turnover rates for oxidative
cleavage as a function of reactant concentration is reported on
the basis of moles of carbon within the products normalized by
the number of active sites (Section 2.5) per unit time (i.e.,
1
000 accumulations) in flowing solutions of H O and
2 2
−
1
CH CN (0.5 M H O , 1.98 M H O, 313 K) delivered by an
3
2
2
2
(mol C)(mol W·s) ). Reactions of 4-octene with H O give
2 2
3
−1
HPLC pump (SSI, LS Class; 1 cm min ). Spectra were
acquired using a 532 nm laser with areal power density (5.2
mW·μm for X-WO −Al O and 0.2 mW·μm for WO3).
selectivity to butanal of ∼95% over both WO and WO −
3
x
Al O catalysts at the reaction conditions and low conversions
2
3
−
2
−2
x
2
3
examined here. Therefore, moles of carbon within produced
butanal are used to determine turnover rates for oxidative
cleavage. Carbon balances for reactions with 4-octene are
typically 85−95%. Reaction with oleic acid, however, gave
different selectivities to epoxide (i.e., 8-(3-octyloxiran-2-
yl)octanoic acid) and oxidative cleavage products (i.e.,
nonanal, nonanoic acid, 9-oxo-nonanoic acid, azelaic acid)
over WO and WO −Al O catalysts. Over WO , selectivity for
In situ UV−vis spectra were collected using a 45° diffuse
reflection probe (Avantes, solarization-resistant fibers) coupled
to a fiber-optic spectrometer (Avantes, AvaFast 2048) with a
compact deuterium−halogen light source (Avantes, AvaLight-
DHc). The pelletized catalyst was loaded into a similar liquid
flow cell as used for Raman spectra but equipped with a sealed
port to admit the diffuse reflectance probe. The sample was
3
x
2
3
3
contacted with flowing aqueous CH CN (1.98 M H O in
3
2
epoxide is greater than 98% and selectivity for oxidative
cleavage products is less than 2%. On the other hand,
selectivity for oxidative cleavage products is greater than 80%
over WO −Al O catalyst. Turnover rates for oleic acid
3
−1
CH CN, 1 cm ·min ) and heated to 313 K while
3
continuously scanning. Background spectra were obtained
when the system reached steady state. The spectra of the
reactive species were collected when samples contacted flowing
x
2
3
consumption were determined from the sum of the formation
rates for all detected products (i.e., 8-(3-octyloxiran-2-yl)-
octanoic acid, nonanal, nonanoic acid, 9-oxo-nonanoic acid,
and azelaic acid) over both catalysts. Carbon balances for
reactions with oleic acid are 60−80% and are systematically
lower than for reactions with 4-octene.
3
H O solutions (0.5 M H O , 1.98 M H O in CH CN, 1 cm ·
2
2
2
2
2
3
−
1
min ) at 313 K, and the system was allowed to reach steady
state once more. Subsequently, the solution was switched to
3
−1
pure CH CN (1 cm ·min ) to determine if the H O -
3
2
2
activated species on these catalysts form reversibly. All UV−vis
spectra of H O -derived intermediates represent the difference
The concentration of H O was measured by colorimetric
2
2
2
2
between the experimental spectrum (i.e., sample within
aqueous H O solutions) and the background spectrum
titration using a titrant solution which is an aqueous solution of
2
2
CuSO (8.3 mM, Fisher Chemicals, >98.6%), neocuproine (12
4
obtained within aqueous CH CN in the absence of H O .
mM, Sigma-Aldrich, >98%), and ethanol (25% v/v, Decon
3
2
2
2
.4. Catalytic Reaction Rate Measurements. Rates for
Laboratories, 100%). The reaction solution was diluted to 1−
oxidative cleavage of oleic acid (C H O , TCI Chemical,
10% v/v reaction solution with CH CN. The diluted reaction
18
34
2
3
3
>
99%) and 4-octene (trans-4-C H , Sigma-Aldrich, 98%)
solution (10 μL) was titrated with the titrant (0.2 cm ), and
8
16
3
were measured within batch reactors (100 cm , three-neck
round-bottom flasks) equipped with reflux condensers. The
organic reactant was combined with hexane (an internal
standard for GC analysis, C H , Sigma-Aldrich, ≥95%) and
the absorbance at 454 nm was determined using a multi-
detection microplate reader (BioTek, Synergy 2). The
measured H O concentration values were used to calculate
2
2
6
14
H O decomposition rates.
2
2
the catalyst (30−100 mg) and added to CH CN. The mixture
We carried out a hot filtration test for WO −Al O catalyst
3
x
2
3
was heated to the desired temperature (313−343 K) and
stirred at 700 rpm for 0.5 h. The reaction was initiated by
adding the necessary amount of aqueous H O (Fischer
to determine if active WO species are leaching from the
alumina support (Section S1.1). During a reaction, an aliquot
3
of the reaction mixture (8 cm ) was taken by syringe at 30 min,
2
2
3
Chemicals, 30% in H O). Aliquots (∼0.8 cm ) of the reaction
solution were extracted as a function of reaction time using a
the solid catalyst was filtered, and the liquid solution was
2
3
transferred into a 20 cm scintillation vial (700 rpm, 333 K).
syringe equipped with a filter (0.05 μm, polystyrene for WO3,
Aliquots were taken from the batch reactor and scintillation
vial as a function of time, and concentrations of reactant and
products were measured by gas chromatography. The
concentrations of butanal do not change following filtration
of the catalyst (Figure S1), which indicates that the active WOx
0
.22 μm, polypropylene for X-WO −Al O ) to separate the
x
2
3
catalyst particles from the reaction solution. Analogous
approaches were used to determine the reaction rates for
epoxide ring opening and the oxidative cleavage of diols. The
3
140
ACS Catal. 2021, 11, 3137−3152