Development of NiCu catalysts for aqueous-phase hydrodeoxygenation

Article abstract of DOI:10.1021/cs500562u

Herein, we report on the development of supported Ni and Raney Ni catalysts doped with Cu for the hydrothermal hydrodeoxygenation (HDO) of o-cresol. Raney Ni catalysts doped with >10 wt % Cu show a significant reduction in gasification activity and produc

Full text of DOI:10.1021/cs500562u

Subscriber access provided by UNIV OF UTAH
Article
Development of NiCu Catalysts for Aqueous-Phase Hydrodeoxygenation
Jacob Gerald Dickinson, and Phillip E. Savage
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/cs500562u • Publication Date (Web): 27 Jun 2014
Downloaded from http://pubs.acs.org on July 4, 2014
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 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
Development of NiCu Catalysts for Aqueous-Phase Hydrodeoxygenation
Jacob G. Dickinson, Phillip E. Savage*
Department of Chemical Engineering, University of Michigan, Ann Arbor, MI 48109, United
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
States
E-mail: psavage@umich.edu
Keywords
Hydrodeoxygenation
Hydrothermal
Raney Ni
o-cresol
Aqueous Phase
NiCu Catalysts
Alumina
Biomass
Abstract
Herein, we report on the development of supported Ni and Raney Ni catalysts doped
with Cu for the hydrothermal hydrodeoxygenation (HDO) of o-cresol. Raney Ni catalysts
doped with greater than 10 wt% Cu show a significant reduction in gasification activity,
and produce a higher yield of liquid products than the unmodified Raney Ni catalyst.
Adding Cu did not, however, increase the yield of the desired HDO products, liquid
hydrocarbons. The addition of acid sites to the catalysts, by supporting Ni and NiCu on
Al O
2 3
and by calcining the Raney Ni to produce Al
2
O within the catalyst, however, did
3
significantly increase the HDO activity of the catalysts such that yields of liquid
Dickinson 1
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 37
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
hydrocarbons exceeded 60%. Two catalysts, a novel calcined 5% Raney NiCu catalyst and a
NiCu/Al
2
O catalyst, produced the highest liquid hydrocarbon yields to date (~70%) for
3
stable, non-noble-metal, hydrothermal HDO catalysts.
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. Introduction
Hydrodeoxygenation (HDO) is an enabling technology for the conversion of biomass
into liquid fuels and chemicals, because it decreases the viscosity and increases the energy
density of bio-oils. HDO has been the focus of several extensive reviews 1-3. The present
work focuses on HDO occurring in a high-temperature, aqueous (i.e., hydrothermal)
reaction environment. Hydrothermal HDO is of interest for several reasons. First, most
biomass liquefaction technologies (e.g. fast pyrolysis, pyrolysis, hydrothermal liquefaction)
result in a mixture or solution of bio-oil and water that can range from 20-30 wt% water,
4
for bio-oils obtained from the fast pyrolysis of lignocellulosic biomass , to 80-90 wt%
water, for the reactor effluent from the hydrothermal liquefaction (HTL) of microalgae 5.
The water content in the reactor effluent is a function of the type of biomass, the water
content of the biomass, and the processing conditions. Separation of the bio-oil and water
after the biomass treatment step is sometimes possible, but can be difficult when the
6
material is a single phase or an oil/water emulsion . Second, the oxygen content of bio-oil
ranges from 5 wt% for bio-oils obtained from the HTL of microalgae 5, 7 to 50 wt% for bio-
4
oils obtained from the fast pyrolysis of lignocellulosic biomass . When this oxygen is
removed from the bio-oil with H , water is formed. Stoichiometry indicates that a bio-oil
2
with 15 wt% oxygen will form an oxygen-free bio-oil and water mixture containing 17 wt%
Dickinson 2
ACS Paragon Plus Environment

Page 3 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
water upon complete HDO. Therefore, water will be present, likely in high quantities,
during HDO.
Hydrothermal HDO presents significant challenges to catalyst stability and activity.
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
Many common hydrotreating catalysts such as NiMo and CoMo oxidize under
8
hydrothermal conditions , and common catalyst supports such as γ-Al
O
2 3
and SiO may be
2
unstable under some hydrothermal conditions 8-10. This realization has led most
1
1
researchers to use noble metal catalysts (Pt, Pd, Ru, and Rh) for hydrothermal HDO .
Dumesic and co-workers have pioneered the conversion of sugars and biofuel byproduct
molecules, such as ethylene glycol, to H
focused on aqueous-phase reforming of sugars with the goal of producing H
authors found that some metals, such as Ni, Rh, and Ru were more selective for alkane
2
and alkanes 12-16. Many of these studies have
2
, but the
1
4
production than gas formation . Huber et al. examined the conversion of sorbitol over Pt
and achieved a high gas phase hydrocarbon yield and found the catalyst to be stable over a
1
5
six-day period . Furthermore, they found that the liquid alkane yield was greater when
they used a mixture of Pt/Al and SiO -Al . This increase in liquid alkane yield was
attributed to the presence of the solid acid (SiO -Al ) catalyst.
O
2 3
2
O
2 3
2
O
2 3
Lercher and co-workers have extensively examined the HDO of phenolic molecules
or HZSM-5 17-19 and achieved complete
between 150 and 200 °C using Pd/C with H
3
PO
4
conversion and high selectivity (> 80%) to cycloalkanes. Interestingly, without the acid
catalyst (i.e., H PO or HZSM-5), no cycloalkanes were observed. In later studies, they
3
4
combined the active metal, Pd, and acid catalysts into a single, bifuntional catalyst,
Pd/HBEA, that produced primarily larger, oxygen free hydroalkylation products from
2
0
phenolic molecules .
Dickinson 3
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 37
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
Savage and coworkers have also made significant progress in catalytically
converting bio-oils and bio-derived molecules into liquid hydrocarbons under
hydrothermal conditions 21-30. In general, these studies have taken place at higher
temperatures (350 – 500 °C) and pressures (150 to 305 bar) than those examined by
Dumesic and Lercher (~200 °C, ~25 bar). These studies revealed that Pt/C is an active and
stable catalyst for the decarboxylation of palmitic acid and the hydrothermal HDO of o-
cresol and benzofuran 21, 22, 29. It is important to note, that unlike the work of Lercher et al.,
these reactions did not require an additional acid catalyst to perform HDO. Nonetheless,
these studies have focused on using noble metal catalysts whose high cost make them
undesirable.
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
Reports of active non-noble metal catalysts for hydrothermal HDO are rare. Lercher
and associates examined the use of Raney Ni with homogeneous (H
3
PO and acetic acid) or
4
solid acid catalysts (Nafion/SiO ) and determined that only the Raney Ni and Nafion/SiO
2
2
catalyst combination was effective at producing cycloalkanes from 4-n-propylphenol at
31
2
00 °C . Again, the researchers determined that the solid acid catalyst was essential for
producing cycloalkanes, as conversions and selectivities were negligible without it. These
researchers have also examined the use of Ni/HZSM-5 and Ni/γ-Al -HZSM-5 catalysts for
O
2 3
phenol and phenolic monomer hydrothermal HDO and obtained a 100% yield of
hydrocarbons between 200 and 250 °C 32, 33. Unfortunately though, these catalysts showed
substantial deactivation in catalyst recycle experiments, even with catalyst regeneration
occurring between cycles, due to catalyst particle sintering, Ni leaching, and structural
changes to the catalyst support 32.
Dickinson 4
ACS Paragon Plus Environment

Page 5 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
To address this need to develop active and stable non-noble metal catalysts for
hydrothermal HDO, we previously examined a Raney Ni catalyst doped with 10 wt% Cu
and found the catalyst to be active, selective, and stable for hydrothermal HDO at 380 °C 21.
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
Cu was chosen as a dopant because the base Raney Ni catalyst was active for C-C bond
hydrogenolysis, producing primarily methane from the o-cresol fed to the reactor, and
8
because Elliot et al. showed that Cu was inactive for gasification . We associated this
increase in selectivity to liquid phase products to the fact that, in a different catalyst system,
Cu reduces the adsorption energy of aromatic molecules when alloyed with Pd to make a
3
4
PdCu catalyst . The present study expands on this previous work to examine the effect of
Cu content on the products, selectivities, and yields from reacting o-cresol, a model oxygen-
containing compound found in most bio-oils 35, 36, with H
over various catalysts. This work
2
shows that increasing Cu content drastically reduces the gasification activity of the Raney
NiCu catalysts, but is ineffective at increasing the yield of the desired liquid hydrocarbons.
To increase the liquid hydrocarbon yield, we added Al
parallel approaches. One approach was to synthesize a NiCu/Al
was to calcine the Raney Ni catalyst thereby oxidizing the Al in the Raney Ni catalyst to
Al . With these catalyst modifications, we provide, to the best of our knowledge, the first
O
2 3
to the catalysts through two
O
2 3
catalyst and the second
O
2 3
report of a high (≥ 60 %) and stable liquid hydrocarbon yield by using only a non-noble
metal catalyst for HDO in hydrothermal conditions.
2. Materials and Methods
This section provides the details of the materials and methods used to complete the
experimental work and analyses.
Dickinson 5
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 37
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
2
.1 Materials
We procured all solvents, reagents, and catalyst precursors from Fischer Scientific
≥ 99% purity) and used them as received. Raney Ni 2800 was obtained from Sigma Aldrich,
and the γ-AlOOH that was used as a catalyst support and a catalyst bed diluent was
obtained from Morton Thiokol. γ-AlOOH was converted to γ-Al during the calcination
and reduction of the catalysts (shown later), therefore, we refer to all such catalysts as
Al -supported. The ZrO was obtained from Alpha Aesar and deionized water was
prepared in house.
.2 Catalyst Synthesis
We synthesized the Raney NiCu catalysts by dissolving Cu(NO
ethanol, adding this solution to reduced Raney Ni, heating the resulting mixture to 100 °C
in a sealed vial for 1 hr, and then reducing the catalyst in flowing H at 400 °C for 3 hr with
a 5°C/min ramp rate. For example, to synthesize the 5% Raney NiCu catalyst we added, by
puncturing the parafilm covering the vial, a solution containing 0.137 g of Cu(NO ꢀ2.5H
(
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
2 3
O
O
2 3
2
2
3
)
2
ꢀ2.5H O in 5 mL of
2
2
3
)
2
2
O
and 5 mL of ethanol to an Ar filled vial containing 0.713 g of reduced Raney Ni before
carrying out the heating and reduction procedures above.
The Al
support with a Ni(NO
In general, to achieve the desired metal loadings we performed two impregnations of the γ-
AlOOH for the Al -supported catalysts, and three impregnations of the ZrO for the ZrO
supported catalyst. For example, to synthesize the 10% 0.5% NiCu/Al catalyst we
performed two impregnations of 4.5 g of γ-AlOOH with a solution that contained 6.6 g of
Ni(NO ꢀ6H O and 0.25 g of Cu(NO ꢀ2.5H O in 13.2 mL of DI H O. Following each
2
O
3
- and ZrO
2
-supported catalysts were synthesized by impregnating the
3
)
2
ꢀ6H
2
O or a Ni(NO ꢀ6H O and Cu(NO ꢀ2.5H O solution in DI H O.
3
)
2
2
3
)
2
2
2
2
O
3
2
2
-
O
2 3
)
3 2
2
)
3 2
2
2
Dickinson 6
ACS Paragon Plus Environment

Page 7 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
impregnation, we dried the catalysts at 110 °C for 12 hr and then calcined them by
increasing the temperature at 10 °C/min to 400 °C where the temperature was held for 4
hr. The calcined catalysts were crushed and sieved (150 µm), and then reduced in flowing
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
H
2
by increasing the temperature at 5 °C/min to 400 °C where the temperature was held
for 3 hr.
Last, to produce the calcined Raney Ni and calcined Raney NiCu catalysts, we heated
the Raney Ni in DI water at 80 °C for 3 hr, followed by drying and calcining the catalysts
using the procedure outlined above. The calcined 5% Raney NiCu catalyst was synthesized
by adding 0.590 g of Cu(NO
3 2
)
ꢀ2.5H O in 5 mL of ethanol to 3.047 g of calcined Raney Ni.
2
The resulting mixture was heated and reduced using the procedure discussed for
producing the Raney NiCu catalysts.
2.3 Catalyst Characterization
We performed X-ray Diffraction (XRD), transmission electron microscopy (TEM),
and CO temperature programmed desorption (TPD) using catalysts passivated overnight at
70 °C in 1% O in He. XRD characterization was performed on a Rigaku Miniflex 600,
2
TEM/STEM was performed on a Jeol 2010f, and CO TPD was performed on a Micromeritics
Autochem 2910. The passivated catalysts were reduced in situ at 430 °C for 180 min, prior
to the TPD experiments. CO was added to the catalyst surface at 25 °C and the catalyst was
heated at 10 °C/min to 550 °C.
2.4 Batch Reactor Procedure
For each hydrothermal HDO experiment, the reduced catalysts were transferred to
an Ar glove box in the reduction tube to minimize oxygen exposure. The reduced catalysts
were loaded into 4.1 mL Swagelok batch reactors in the Ar glove box, and the open end of
Dickinson 7
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 37
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
each reactor was sealed with parafilm. The batch reactors were filled outside of the glove
box with 0.67 mL of DI H
parafilm, thereby minimizing exposure of the catalyst to air. The batch reactors were then
capped, purged with He, and pressurized with He and H . Details on the construction,
pressurization, extraction, gas analysis, and liquid product analysis are available in our
2
O and either 20 or 100 mg of o-cresol by poking holes through the
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
2
2
2
previous work . In short, we heated the reactors to 380 °C in a fluidized sand bath,
quenched the reactors in water, analyzed the gas content of the reactors with GC/TCD,
extracted the reactors with 10 mL of acetone, and analyzed the liquid contents of the
reactors with GC/FID. In some instances, multiple reactions were run at a single reaction
condition to obtain estimates of the experimental uncertainty, represented by one standard
deviation. We report data only from reactors with carbon balances greater than 80%, but in
general, the carbon balances were greater than 90%.
2.5 Flow Reactor Description
We used two different flow reactor configurations. The first configuration used a
sandbath as the heating source and had a feed that contained DI water, formic acid, and o-
2
1
cresol. This flow reactor configuration was detailed in our previous work , and in the
present work was used only with the 2% Raney NiCu catalyst. A second flow reactor
configuration, shown in fig. 1, is similar to the previous setup, but with a few major
improvements. First, H
2
was delivered to the flow reactor with a mass flow controller,
source. Second, separate
instead of using the in situ decomposition of formic acid as the H
2
pumps for DI water and o-cresol allowed for varying concentrations of o-cresol to be fed to
the reactor. Last, we used a tube furnace as the heating source of the reactor to allow for
faster heat up and cool down times.
Dickinson 8
ACS Paragon Plus Environment

Page 9 of 37
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
Fig. 1. Flow reactor process flow diagram.
The modified flow reactor, shown in fig. 1, has three 316 stainless steel inlet lines.
Chrom Tech Series III pumps fed DI water and the internal standard solution to the reactor
and cooled reactor effluent, respectively. o-Cresol, heated to 35 to 40 °C by heat tape, was
fed by an ISCO 260D syringe pump, and H
controller from a 6000 psig H cylinder. The preheating tubing within the Applied Test
Systems furnace, which was controlled by an Omega PID controller, was 1/16 in. o.d. tubing.
The water, H , and o-cresol preheating lines were 80, 60, and 15 in. in length, respectively,
2
was fed through a Brooks 5850 TR mass flow
2
2
and mixed in a cross fitting prior to entering the catalyst bed. The temperature within this
cross fitting was monitored by a thermocouple and data logger. The 3.5 in. catalyst bed was
constructed from 1/4 in. o.d. tubing with 5 µm Hastelloy frits placed at both ends. The
catalyst bed was loaded with catalyst and γ-AlOOH, a diluent, in the Ar glove box. We
covered the ends of the catalyst bed with parafilm in the glove box to minimize the
exposure of the catalyst to air when the catalyst bed was transferred to the flow reactor. A
second thermocouple, positioned in a T fitting, measured the temperature of the product
stream exiting the catalyst bed. A tube-in-tube heat exchanger cooled the reactor effluent to
room temperature, and a backpressure regulator maintained the desired reaction pressure.
Dickinson 9
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 37
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
We added an internal standard solution consisting of 4-isopropylphenol (4 g/L) in
2-propanol to the cooled reactor effluent prior the backpressure regulator to form a single
liquid phase and to provide an internal standard for use in product quantification. 15 to 20
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
mL/min of N was added to the reaction stream through an Omega 5400/5500 mass flow
2
controller after the backpressure regulator to provide a reference gas. A Gilson 223 fraction
collector with an automated switching valve collected liquid samples in test tubes
containing ~3 mL of 2-propanol. When the fraction collector was not collecting liquid
samples, the reactor effluent was diverted to a 250 mL flash column that separated the
liquid and gas products. The liquid products were sent to a waste container while the gas
products were sent to the GC/TCD. The liquid samples were analyzed on an Agilent 6890N
gas chromatograph with a flame ionization detector using a 50m HP-5MS capillary column.
The injected sample, 1 µL, was heated by an inlet at 310 °C and separated by the capillary
column by holding the column at 35 °C for 10 min, and then ramping the oven temperature
at 2 °C/min to 50 °C, 10 °C/min to 160 °C, 30 °C/min to 300 °C, and holding for 2 min. An
Agilent 6890N gas chromatograph with a mass spectrometric detector used this same
method and column to determine the identities of some molecules.
2.6 Flow Reactor Procedure
To start up the flow reactor shown in fig. 1, we first attached the catalyst bed to the
reactor, then pressurized the system with 70 bar H
2
to check for leaks. After confirming the
at 50 – 100
absence of leaks, we released the reactor pressure, and began flowing H
2
mL/min while heating the reactor to 550 °C for at least one hour for the in situ catalyst
reduction. We then cooled the reactor to approximately 400 °C and started flowing water,
H
2
, and o-cresol. For the calcined Raney Ni and calcined 5% Raney NiCu catalysts we set the
Dickinson 10
ACS Paragon Plus Environment

Page 11 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
H
2
O, o-cresol, and H
2
flow rates to 1 mL/min, 0.030 mL/min, and 0.006 mol/min,
respectively, to build the reaction pressure. Upon the reactor reaching ~240 bar, we
lowered the H O flow rate to the desired steady state value of 0.270 mL/min, while
maintaining the other flow rates. At this point, we also started the internal standard
solution flowing at 0.400 mL/min and the N reference gas flowing at 20 mL/min. The
startup procedure for the 10%, 0.5% NiCu/Al catalyst differed only in the H O flow rate,
2
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
O
3
2
which was set at 0.270 mL/min during the entire startup period. Upon reaching the desired
temperature of 365 °C at the mixing point thermocouple and the desired reactor pressure
of 280 bar, we began collecting samples. We found that the reactor took approximately 225
min to achieve steady carbon balances and we report this time as zero min on stream.
2.7 Data Analysis
We calculated conversion, yield, selectivity, and carbon recovery as follows:
ꢀꢈꢉ,ꢊ
ꢀꢈꢉ,ꢋ
ꢀꢁꢂꢃꢄꢅꢆꢇꢁꢂ = 1 −
ꢀꢐ
ꢌꢌ(1)
ꢍꢇꢄꢎꢏ =ꢌ
ꢌꢌꢌ(2)
ꢀꢈꢉ,ꢋ
ꢀꢐ
ꢑꢄꢎꢄꢒꢓꢇꢃꢇꢓꢔ =
ꢌꢌ(3)
∑
∑
ꢀꢐ
ꢀꢐ
ꢀ
ꢕꢅꢖꢁꢂꢌꢅꢄꢒꢁꢃꢄꢅꢔ =ꢌ
ꢌꢌꢌ(4)ꢌ
ꢀꢈꢉ,ꢋ
where Coc,e, Coc,o, and C refer to the concentrations, in mol C/L, of o-cresol exiting the
i
reactor, of o-cresol entering the reactor, and of any product, respectively. We also will refer
to the liquid hydrocarbon yield, which is the sum of the yields of all hydrocarbons in the
liquid phase reactor effluent. On occasion, the flow reactor would release a large volume of
Dickinson 11
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 37
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
gas, resulting in a slight pressure drop in the reactor. This increased gas flow resulted in
high yields of the carbon containing gasses (i.e., methane and CO ) and high carbon
2
recoveries (≥ 120%) if sampling occurred simultaneously with a gas release. When such
anomalous data were collected, we removed them from the data set.
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
3. Results and Discussion
This section is divided into two major parts. The first section examines the effect of
Cu loading on product yields and selectivities for the hydrothermal HDO of o-cresol. The
second section discusses improvements to the NiCu catalysts made through the addition of
acid sites. Within each major section, we first discuss reaction results and then examine the
findings of the catalyst characterization.
3.1 Raney NiCu catalysts and the effect of Cu loading
Previous work 21 indicates that Raney Ni 2800 promoted with 10 wt% Cu is an
active and stable catalyst for hydrothermal HDO. The goal of this section is to examine the
effect of Cu loading and to develop a catalyst and reaction scheme for hydrothermal HDO
that produces high liquid hydrocarbon yield. To achieve this goal, the catalyst must not
only be active, it must also be selective for C-O bond hydrogenolysis.
3.1.1 2% Raney NiCu
Previous work 21 showed that a 10% Raney NiCu catalyst had a lower HDO turnover
frequency (TOF) than did unmodified Raney Ni. Further, the total activity (i.e., TOF for
conversion) of the 10% Raney NiCu catalyst is even more suppressed when compared with
the unmodified Raney Ni catalyst. We desired to learn whether reducing the Cu content
would improve overall and HDO activity without diminishing the selectivity increases, and
Dickinson 12
ACS Paragon Plus Environment

Page 13 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
therefore the higher liquid hydrocarbon yield, observed with the 10% Raney NiCu catalyst.
Therefore, we synthesized a 2% Raney NiCu catalyst and tested this catalyst in the flow
system described previously 21 at 380 °C using formic acid decomposition as the H
source.
2
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
Fig. 2 shows the variation of conversion and liquid hydrocarbon yield with time on
stream (TOS). The most abundant liquid hydrocarbon products were toluene,
methylcyclohexane, benzene and cyclohexane. In addition to these products, oxygenated
products consisting primarily of methylcyclohexanone and phenol formed. Methane and
CO were the only gaseous products, but due to the presence of formic acid in the reactor
2
feed, it is unclear whether these carbon-containing molecules originate from the formic
acid or o-cresol.
Fig. 2. Conversion and liquid hydrocarbon yield from reacting o-cresol over 2% Raney NiCu
(
W/F = 56 min, 380 °C, 305 bar, feed solution (ambient conditions): 24.7 g/L o-cresol, 147
mL/L formic acid, balance DI H O, feed flow rate (ambient temperature) = 0.218 mL/min,
catalyst reduction).
2
4
30 °C in situ H
2
Fig. 2 shows that the maximum liquid hydrocarbon yield of 19% occurred around
00 min on stream. This maximum liquid hydrocarbon yield is similar to the 21 ± 4%
2
average liquid hydrocarbon yield obtained previously from a 10% Raney NiCu catalyst
Dickinson 13
ACS Paragon Plus Environment

ACS Catalysis
Page 14 of 37
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
2
1
using the same reaction conditions . After reaching this maximum, the liquid hydrocarbon
yield quickly declined to 0% at 470 min on stream and thereafter. Likewise, conversion
decreased rapidly after 200 min on stream. It is clear that the 2% Raney NiCu catalyst
deactivates rapidly after about 240 min on stream.
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 cause of the deactivation is unknown, and is beyond the primary focus of this
work. We did perform several simple experiments, however, that permit speculation
regarding the cause of deactivation. We previously reported that CO is in the reaction
stream that contacts the catalyst bed. Fig. 3 shows the results from CO TPD of the
2
1
unmodified Raney Ni catalyst and the 10% Raney NiCu catalyst used previously . These
results will be discussed more thoroughly in a later section. At present, we will simply
remark that CO remained adsorbed to both the Raney Ni and 10% Raney NiCu catalysts at
temperatures approaching 600 °C. This finding is important because it points to a possible
cause of catalyst deactivation, CO poisoning, and a possible means of regenerating the
catalyst, reduction of the catalyst at high temperature (e.g., 550 °C) in flowing H . Literature
2
37
also indicates that CO
2
, when reacted over a Ni catalyst, can form adsorbed CO . The
preferential pathway for the removal of adsorbed CO when H
2
is present is through a HCO
3
8
or COH pathway that converts bound CO to methane .
Dickinson 14
ACS Paragon Plus Environment

Page 15 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
Fig. 3. CO temperature programmed desorption spectra of Raney Ni and 10% Raney NiCu.
Fig. 4 shows the liquid hydrocarbon yield from three cycles of reacting o-cresol over
the same 2% Raney NiCu catalyst that was used in fig. 2 (table S1, supplementary
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
information). Each cycle consisted of reducing the catalyst at 550 °C in flowing H , then
2
passing the feed solution over the catalyst for 720 min. It is clear from fig. 4 that the liquid
hydrocarbon yields for each cycle overlap, indicating that the HDO activity of the catalyst
was regenerated after each cycle by simply re-reducing the catalyst. This result is
consistent with CO poisoning causing the loss in catalyst activity, but it does not preclude
other methods of deactivation. The catalyst could be losing activity from oxidation of the Ni
metal by supercritical water, or from some other unknown catalyst poison. Other common
deactivation causes such as sintering or coking are unlikely, because these conditions are
irreversible by reduction.
Fig. 4. Liquid hydrocarbon yield from reacting o-cresol over 2% Raney NiCu (W/F = 80 min,
380 °C, 305 bar, feed solution (ambient conditions): 24.7 g/L o-cresol, 147 mL/L formic
acid, balance DI H
2
O, feed flow rate (ambient temperature) = 0.200 mL/min, 550 °C in situ
H
2
catalyst reduction between runs).
Dickinson 15
ACS Paragon Plus Environment

ACS Catalysis
Page 16 of 37
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
3.1.2 Cu loading effects
This section elucidates the effect of Cu loading on the HDO selectivity of 1-40 wt%
Raney NiCu catalysts using H (rather than formic acid) as the reductant to avoid the
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
2
formation of CO. We performed these reactions in 4.1 mL batch mini-reactors to rapidly
screen these catalysts.
Fig. 5 and table S2 shows that at low Cu loading (e.g. 1 - 7%) the conversion of o-
cresol was nearly 100%, and the selectivity to gas products, such as methane (55 - 68%
yield), ethane (~1% yield), and CO (15 – 30% yield), was high. Selectivity to liquid
2
products was low. The liquid products were mainly oxygenated intermediates such as
phenol (1.4 – 8.3% yield), cyclohexanone (0-2.7% yield), methylcyclohexanone (0.4-2.4%
yield), and methylcyclohexanol (0 – 2.4% yield).
Fig. 5. Effect of Cu loading for Raney NiCu catalysts on selectivity, conversion, and liquid
hydrocarbon yield in 4.1 mL reactors (380 °C, 30 min, 20 mg catalyst, 20 mg of o-cresol,
0.67 mL DI H
2
O, 19 bar H
2
(STP), w×t/m = 30 min, 17:1 H to o-cresol).
2
Dickinson 16
ACS Paragon Plus Environment

Page 17 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
There is a dramatic shift in product selectivities and conversion when the Cu
content of the Raney NiCu catalyst increased beyond 7% Cu. Fig. 5 shows that the
conversion of o-cresol drops beyond 7% Cu loading and remains relatively steady at ~70%.
Also between 7 and 25% Cu, the product selectivities shift from forming gas to forming
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
liquid products. Again, methane and CO were the main gasses while the aforementioned
2
oxygenated intermediate molecules were present in high yields (table S2).
These results support the conclusion that Cu decreases the hydrocracking (C-C bond
hydrogenolysis with H ) activity of the catalyst. It is likely, on the low Cu loaded catalysts,
2
that cyclohexanol, after forming from the hydrogenation of cyclohexanone, continues to
react to form methane. This hypothesis is supported by literature showing that Ni is a more
8
active hydrogenation and hydrocracking metal than Cu . The lower overall activity of the
high Cu loaded catalysts manifests in the decreased conversion observed. In contrast to the
conversion and liquid and gas product selectivities in fig. 5, the liquid hydrocarbon yield
varied little as the Cu content of the catalysts is varied. The liquid hydrocarbon yield varied
only from a low of 2.5 ± 0.2% for the 1% Raney NiCu catalyst to 5.6 ± 1.4% for the 5%
Raney NiCu catalyst.
The goal of this work was to produce a non-noble metal catalyst that is active,
selective, and stable for hydrothermal HDO because such a catalyst would result in high
liquid hydrocarbon yield, while minimizing the formation of lower value gas products. The
results above indicate that Cu doping the Raney Ni catalyst is itself, insufficient to achieve
this goal because the addition of Cu results only in minimizing unwanted gas products and
not in increasing the yield of the desired liquid hydrocarbon products.
3.1.3 Raney NiCu catalyst characterization
Dickinson 17
ACS Paragon Plus Environment

ACS Catalysis
Page 18 of 37
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 reaction results presented in section 3.1.2 indicate that the addition of Cu to a
Raney Ni catalyst has a dramatic effect on the product selectivities when o-cresol is reacted
at 380 °C in supercritical water. This section elucidates the morphology of the Raney NiCu
catalysts by discussing the results from titration, diffraction, and microscopy experiments
used to characterize the catalysts.
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
Fig. 3 shows the results of CO TPD of both the unmodified Raney Ni and 10% Raney
NiCu catalysts. Both spectra contain two major peaks at 75 and 405 °C. The Raney Ni
spectrum contains one shoulder between 105 and 140 °C, while the Raney NiCu spectrum
contains two shoulders, one between 105 and 115 °C, and the other, broader shoulder,
between 145 and 210 °C. This second broad shoulder is an entirely new feature in the 10%
Raney NiCu spectrum that was not observed in the Raney Ni spectrum. Both spectra were
integrated, and the 10% Raney NiCu catalyst adsorbed 1.7 times the CO of the Raney Ni
catalyst. The new shoulder and the increased CO uptake suggest that a new phase is formed
when the Raney Ni catalyst is doped with Cu, but the results do not provide direct evidence
of the type of phase the Cu forms. We also examined a Cu standard (100 nm Cu particles)
under the same reduction and TPD conditions and found no significant CO uptake because
the Cu particles sintered. Therefore, it is likely that the Cu and Ni in the 10% Raney NiCu
catalyst form an alloy due to the relatively high CO uptake observed.
Fig. 6 shows the XRD patterns of the 1-40% Raney NiCu catalysts. The dominant
features of each diffraction pattern at low Cu loading (1-7%) are the Ni planes. As the Cu
loading increases, a peak forms at the base of the Ni(111) peak in the 10% Raney NiCu
catalyst. Further increases in Cu loading increase the size of this peak, identified as a
Ni0.5Cu0.5, indicating the formation of an alloy of Ni and Cu. The 40% Raney NiCu catalyst
Dickinson 18
ACS Paragon Plus Environment

Page 19 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
shows the formation of a third peak in the (111) plane, identified as a Cu rich phase,
(Cu19Ni)0.2. In addition to the (111) plane, one also observes the formation of the Ni0.5Cu0.5
and (Cu19Ni)0.2 phases in the (200) and (220) planes. These results indicate that Ni and Cu
form an alloy under the synthesis and reduction conditions used, and that at low to
moderate Cu loading, this alloy appears to be roughly equal parts Ni and Cu.
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
Fig. 6. X-ray diffraction patterns of the Raney NiCu catalysts (passivated in 1% O
overnight, Cu Kα source, 40 kV, 15 mA, 1.25°/min).
2
at 70 °C
Fig. 7 shows TEM and STEM images of the 10% Raney NiCu catalyst. Fig. 7a shows a
TEM image of the catalyst particle examined. Fig. 7d shows the STEM image used to
generate the elemental map. We calculated the x distance across Fig. 7d, using the scale bar
from Fig. 7a and common points in both images, as 870 nm. Fig. 7b, c, e, and f show the
elemental maps of Ni, Cu, Zn, and Al. One expects Ni, Cu, and Al to be present in the catalyst.
Zn, on the other hand, (Fig. 7e) served as a control element to determine the amount of
background signal one should expect. Examination of Fig. 7e indicates that the background
signal is low compared to the signals obtained from the other elements. Fig. 7f shows that
Dickinson 19
ACS Paragon Plus Environment

ACS Catalysis
Page 20 of 37
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 Al appears to be segregated to both ends of the catalyst particle examined, with
relatively little Al present in the middle of the image. This segregation is likely a result of
the leaching methods that are used to remove Al from the Ni matrix during the synthesis of
Raney Ni.
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
Fig. 7b and c indicate that Ni and Cu appear to spread evenly across the catalyst
surface. This finding likely indicates that the Cu added to the catalyst is not present in Cu
islands, but rather is present in a NiCu alloy that is relatively evenly dispersed on the
catalyst surface. This finding is also in agreement with the XRD and TPD results presented
earlier. All of the catalyst characterization findings are consistent with the results of Huber
and Shabacker who used a Sn doped Raney Ni catalyst for aqueous phase reforming
reactions and found that the Sn incorporated in the Ni catalyst 39-41
.
Fig. 7. TEM and STEM elemental maps of 10% Raney NiCu. Fig. 7a. shows the original TEM
image of the catalyst. Fig. 7d. shows the STEM image taken of the section of the catalyst
used for the elemental map. Fig. 7b, c, e, and f. show the result from elemental maps of Ni,
Cu, Zn, and Al respectively.
3.2 Improved Ni-based Hydrothermal HDO catalysts with the addition of acid sites
Dickinson 20
ACS Paragon Plus Environment

Page 21 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
Section 3.1 showed that the addition of Cu to Ni catalysts suppresses gasification
activity, but has little effect on the HDO activity of these catalysts. In short, the Raney NiCu
catalysts are active for the hydrogenation of o-cresol, but are not sufficiently active for the
C-O hydrogenolysis required for HDO. Recent research suggests that acid sites are
essential for hydrothermal HDO reactions over Ni catalysts 31, 42. Zhao et al. found that the
Raney Ni 4200 and 2400 are active for HDO only when an additional acid catalyst, in this
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
1
case Nafion supported on SiO
that the non-hydrothermal HDO of phenol occurred more rapidly on acidic supports such
as ZrO and Al , leading these researchers to speculate that the HDO reaction actually
2
, was added the reaction mixture . Mortensen et al. found
2
O
2 3
occurred at the metal/support interface 42.
This section elucidates the influence of added acidity on the product yields and
selectivities by modifying the reaction system in three ways. First, we used the 10% Raney
NiCu catalyst in an aqueous solution of HCl or H
catalysts on acidic supports (e.g., Al and ZrO
catalyst by oxidizing some of the Al in the catalyst to Al
00 °C (see section 2). This calcined Raney Ni catalyst is a novel catalyst, that to the best of
our knowledge, has never before been used for HDO. The only prior report we found of
intentionally producing Al through calcination on a Raney Ni catalyst used the Al
binder for an extruded catalyst .
.2.1 Batch experiments with acidic Ni catalysts
We examined a variety of catalysts in 4.1 mL batch mini-reactors to quickly assess
2
SO
). Third, we further modified the Raney Ni
by calcining the catalyst in air at
4
. Second, we synthesized Ni and NiCu
O
2 3
2
O
2 3
4
O
2 3
2
O as a
3
4
3
3
their activity for hydrothermal HDO under a variety of conditions. Table 1 summarizes
these results and conditions used. These experiments used less catalyst (10 vs. 20 mg),
Dickinson 21
ACS Paragon Plus Environment

ACS Catalysis
Page 22 of 37
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
more o-cresol (100 vs. 20 mg), and a lower H
2
to o-cresol molar ratio (6:1 vs. 17:1) than did
the batch reactions that generated fig. 5. These reaction conditions were chosen primarily
to prevent complete conversion. Table 1 indicates that the addition of HCl and H
2
SO had a
4
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
negative effect on liquid hydrocarbon yield and conversion. We suspect, although we did
not test this hypothesis, that the Cl and S poisoned the 10% Raney Ni catalyst and rendered
the catalyst almost completely inactive.
Table 1 also summarizes the results of several active catalysts. The temporal
variation of conversion and liquid hydrocarbon yield from these catalysts appear in fig. 8.
Table 1 and fig. 8 reveal that the 10% Ni/ZrO
2
catalyst had a low liquid hydrocarbon yield
catalyst showed
and conversion, but a high selectivity to liquid products. The 10% Ni/Al
O
2 3
the highest conversion, liquid hydrocarbon yield, and selectivity to liquid products. Last,
the calcined Raney Ni catalyst had a high conversion, a relatively low liquid hydrocarbon
yield, and a low selectivity to liquid products when compared to the 10% Ni/Al
2
O catalyst.
3
Nonetheless, the calcined Raney Ni catalyst generally had higher liquid hydrocarbon yields
than the 10% Raney NiCu catalyst. For the catalysts tested in fig. 8, the liquid hydrocarbons
were cyclohexane, benzene, toluene and methylcyclohexane. The major oxygenated
products were methylcyclohexanol, methylcyclohexanone, phenol, cyclohexanol, and
cyclohexanone. The gas product was primarily methane with smaller amounts of CO
2
. The
yields of individual products are shown in table S3.
Dickinson 22
ACS Paragon Plus Environment

Page 23 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
Table 1. Summary of catalysts tested in 4.1 mL batch reactors (380 °C,15-180 min, 10 mg
catalyst, 100 mg of o-cresol, 0.67 mL DI H
2
O or DI water acid solution, 41 bar H
2
(STP), 6:1
H
2
to o-cresol, X= conversion, LH = liquid hydrocarbon, Sliq = selectivity to liquid products).
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
Catalyst
X
LH Yield
0-7%
0%
Sliq
65-87%
----
Time (min) W×t/m (min)
1
0% Raney NiCu
37-80%
1-4%
15-45
15-60
1.5-4.5
1.5-6.0
1
10% Raney NiCu + HCl
2
1
0% Raney NiCu + H SO4
0-5 %
0-1%
----
30-60
3.0-6.0
2
1
0% Ni/Al O3
54-93% 12-35% 85-91%
14-67% 0-8% 64-98%
57-88% 3-12% 44-68%
30-150
0.15-1.5
2
10% Ni/ZrO2
15-180
15-60
0.3-1.8
1.5-6.0
Calcined Raney Ni
1
1
2
.04 wt% HCl in DI water. 0.2 wt% H SO in DI water
2 4
Fig. 8. Conversion (dashed lines) and liquid hydrocarbon yield (solid lines) from Ni and
NiCu catalysts in 4.1 mL batch reactors. Table 1 summarizes all of catalysts tested (380 °C,
0 mg catalyst, 100 mg of o-cresol, 0.67 mL DI H
1
2
O, 41 bar H
2
(STP), 6:1 H to o-cresol).
2
The high liquid hydrocarbon yield from the 10% Ni/Al
2
O catalyst and the increased
3
liquid hydrocarbon yield from the calcined Raney Ni catalyst support the hypotheses by
Mortensen et al. 42 and Zhao et al. 32 that acid sites, like those found in Al
O , assist in C-O
2
3
hydrogenolysis, and therefore increase HDO activity. The results obtained with the calcined
Dickinson 23
ACS Paragon Plus Environment

ACS Catalysis
Page 24 of 37
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
Raney Ni catalyst were especially encouraging because such a catalyst has not been studied
previously for HDO. Furthermore, it is suspected that the liquid hydrocarbon yield, shown
in fig. 8, was low due to the relatively high gasification activity of the catalyst (i.e., low
selectivity to liquid products in table 1), but in section 3.1 it was shown that Cu decreases
the gasification activity of Raney Ni. Therefore, with these two tools, the addition of Cu 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
the addition of Al
2
O , the activity and selectivity of Ni catalysts can be tuned.
3
3.2.2 Flow reactions
At this point, this work identified several catalysts that are active for hydrothermal
HDO, but several challenges remain. The first challenge is to obtain a high liquid
hydrocarbon yield (e.g., > 50%) with these catalysts, and the second challenge is to
determine the stability of these catalysts under the harsh hydrothermal conditions. These
challenges were addressed by performing a series of flow reactor experiments using the
reactor shown in fig. 1 and described in detail in the methods section.
3.2.2.1 Calcined Raney Ni
Calcined Raney Ni was tested in the flow reactor, and fig. 9 shows the results. The
conversion of o-cresol, at a W/F (mass of catalyst active metal/mass flow rate of o-cresol)
of 12.7 min, is generally between 75 and 90%, and appears to decrease slightly with TOS.
The liquid hydrocarbon yield was generally between 25 and 45% with the higher yields
appearing at later TOS. The linear trend line fit to the liquid hydrocarbon yield data
indicates an increasing liquid hydrocarbon yield with TOS. Fig. 9 also shows that the yields
of the two major liquid hydrocarbon products, methylcyclohexane and cyclohexane, appear
relatively steady with TOS, if not increasing slightly. A variety of other minor liquid
hydrocarbon products also formed, and in general had low individual yields (< 2%). The
Dickinson 24
ACS Paragon Plus Environment

Page 25 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
sum of these products, however, significantly increased the liquid hydrocarbon yield. These
products include butane, 2-methylbutane, hexane, methylcyclopentane, 1,3-
dimethylcyclopentane, ethylcyclopentane, toluene, and methylcyclohene (table s4). These
minor products indicate that this catalyst has a relatively high isomerization activity
compared with the other catalysts examined.
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
Fig. 9 shows that the methane yield was between 20 and 40% during the reaction,
indicating that this catalyst still retains significant gasification activity. This gasification
activity also appears to be maintained with TOS indicating that this catalyst is unlikely to
become more selective for liquid products, either oxygenated or hydrocarbons, without a
modification to the catalyst.
Fig. 9. Results from reacting o-cresol over calcined Raney Ni (W/F = 12.7 min, Tinlet = 365 °C,
T
outlet = 390 °C, 280 bar, feed flow rates (mL/min, ambient temperature): H
cresol = 0.030, internal standard solution =0.400, H = 0.006 mol/min, H to o-cresol molar
ratio = 20:1, 550 °C in situ H catalyst reduction).
2
O = 0.270, o-
2
2
2
Previous work revealed an increase in liquid hydrocarbon yield with TOS, similar to
that in fig. 9, when using a Pt/C catalyst for the hydrothermal HDO of o-cresol. A potential
Dickinson 25
ACS Paragon Plus Environment

ACS Catalysis
Page 26 of 37
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
cause of this increase in activity was reduction of the catalyst surface that had been
2
1
oxidized during reactor startup . Similarly, oxidation of the present Ni catalyst may have
occurred during the reactor startup, and gradual re-reduction during the reaction led to
increasing liquid hydrocarbon yields. Another possible cause of the increase in liquid
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
hydrocarbon yield with TOS is an increase in the Al
caused by the oxidizing hydrothermal environment. Suchanek developed an Al
phase diagram by subjecting γ-Al(OH) and γ-AlOOH to a hydrothermal environment and
showed that under conditions similar to those used in this reaction, α-Al
O
2 3
content of the Raney Ni catalyst
2 3
O -H
2
O
3
is formed 44.
O
2 3
Therefore, it is also possible that the increase in liquid hydrocarbon yield observed in fig. 9
is due to an increasing concentration of acid Al sites adjacent to the Ni active metal,
allowing for HDO reactions to occur more readily.
Previous work with Raney Ni, using formic acid as the H
reaction conditions, provided only very low yields of liquid hydrocarbons, 6.4%, at a very
O
2 3
2
source and similar
2
1
similar W/F of 12 min . Therefore, the development of calcined Raney Ni represents a
significant step forward in developing a low-cost, active, and stable catalyst for
hydrothermal HDO.
3.2.2.2 Supported NiCu catalyst
Fig. 10 shows the results from reacting o-cresol over a 10%, 0.5% NiCu/Al
2 3
O
catalyst at a W/F of 3.22 min using the same reaction conditions as above. We doped this
catalyst with a small amount of Cu (0.5%) to suppress the gasification pathway.
Fig. 10 shows the conversion remained steady at ~100% throughout the 24 hr
experiment. The liquid hydrocarbon yield was generally between 50 and 80%, with an
average of 71 ± 19%. During this experiment the reaction pressure would occasionally
Dickinson 26
ACS Paragon Plus Environment

Page 27 of 37
ACS Catalysis
1
2
3
4
5
6
7
8
9
drop to around 225 bar, for an unknown reason, before rebuilding to 280 bar. These
occasional process upsets induced atypical scatter in the data collected. Nonetheless, the
results clearly show a high average liquid hydrocarbon yield and no significant catalyst
deactivation over the 24 hr reaction. The stability of this catalyst was surprising due to the
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
reported instability of γ-Al
2
O
under hydrothermal conditions 8-10. It is worth noting that
3
the two potential transformations of γ-Al
O
2 3
under the reaction conditions used, γ-Al
2
O to
3
9
44, would have occurred to a significant extent over the 24 hr
may also have been
AlOOH or γ-Al
O
2 3
to α-Al
2 3
O
reaction, if either transformation was going to occur. The γ-Al
O
2 3
stabilized by the presence of o-cresol or other alcohol-containing species in the reaction
mixture 45.
Fig. 10. Results from reacting o-cresol over 10%, 0.5% NiCu/Al
O
2 3
(W/F = 3.22 min, Tinlet
65 °C, Toutlet = 390 °C, 280 bar, feed flow rates (mL/min, ambient temperature): H
.270, o-cresol = 0.030, internal standard solution =0.400, H = 0.006 mol/min, H
catalyst reduction).
=
3
0
2
O =
to o-
2
2
cresol molar ratio = 20:1, 550 °C in situ H
2
Fig. 10 also shows that the yields of the three major liquid hydrocarbon products:
cyclohexane, toluene, and methylcyclohexane. These three products accounted for > 85%
Dickinson 27
ACS Paragon Plus Environment

ACS Catalysis
Page 28 of 37
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
of the liquid hydrocarbon yield, but small yields of benzene (≤ 3%) and methylcyclohexene
(≤ 2%) were also observed. This finding indicates that this supported catalyst is much less
active for isomerization than the calcined Raney Ni catalyst.
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 yields of oxygenated and gas products were generally low. The oxygenated
products were phenol, cyclohexanone, cyclohexanol, 2-methylcyclohexanol, and 2-
methylcyclohexanone, and the average yield for each oxygenated molecule was ≤ 2% (table
S5). The gas products were methane and CO
2
with average yields of 5 ± 2% and 3 ± 2%,
respectively.
3
.2.2.3 Calcined 5% Raney NiCu
The calcined Raney Ni catalyst tested in both batch and flow reactors showed very
promising results because of the increased liquid hydrocarbon yield when compared with
2
1
previous work . Examination of the products revealed that this catalyst was still too active
for gasification, because, as shown in fig. 9 and table S4, between 20 and 40% of the carbon
from o-cresol formed methane. Therefore, to suppress this gasification activity, we added
5
% Cu to the calcined Raney Ni catalyst, and reacted o-cresol over this catalyst at a W/F of
9.9 min.
1
Dickinson 28
ACS Paragon Plus Environment

Page 29 of 37
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
Fig. 11. Results from reacting o-cresol over 5% calcined Raney NiCu (W/F = 19.9 min, Tinlet
=
0
365 °C, Toutlet = 390 °C, 280 bar, feed flow rates (mL/min, ambient temperature): H
.270, o-cresol = 0.030, internal standard solution =0.400, H = 0.006 mol/min, H to o-
catalyst reduction).
2
O =
2
2
cresol molar ratio = 20:1, 550 °C in situ H
2
Fig. 11 shows that the o-cresol was completely converted over the calcined 5%
Raney NiCu catalyst throughout the reaction. The liquid hydrocarbon yield was ~40% early
in the reaction and increased linearly to ~60% after 1200 min on stream. As with the
calcined Raney Ni catalyst, the two major liquid hydrocarbon products were
methylcyclohexane and cyclohexane. The methylcyclohexane yield increased linearly with
TOS from ~20 to ~40% while the cyclohexane yield remained steady at 10 ± 2% over the
course of the reaction. The remaining liquid hydrocarbon and oxygenated products (table
S6) were the same as those observed with the calcined Raney Ni catalyst, and each product
had a low yield (≤ 2%).
Fig. 11 also shows the methane yield from the calcined 5% Raney NiCu catalyst and
indicates that at early TOS, the catalyst was active for gasification, but after 800 min on
Dickinson 29
ACS Paragon Plus Environment