Hydrodeoxygenation of Furylmethane Oxygenates to Jet and Diesel Range Fuels: Probing the Reaction Network with Supported Palladium Catalyst and Hafnium Triflate Promoter

Article abstract of DOI:10.1021/acscatal.7b00986

Catalytic hydrodeoxygenation of furylmethane oxygenates to high carbon branched chain jet and diesel fuel range alkanes under mild reaction conditions is a promising strategy for energy-efficient production of fuels with minimal C-C cracking to undesired products. Here, we report that a strong Lewis acidic promoter can overcome the energy barrier for furylmethane hydrodeoxygenation at lower temperature. Furan rings of furylmethanes are first hydrogenated to fully saturated cyclic ethers by a hydrogenation catalyst, which then undergo facile ring opening and deoxygenation by the promoter. A cyclic intermediate between ethereal O and the Lewis acidic metal center, assisted by the triflate ligand of the promoter, is formed in the ring-opening step. Probing the reaction pathway with symmetric single furan ring surrogate molecules suggests that the promoter is necessary for the ring opening. Deoxygenation of ring-opened oxygenates takes place more quickly for single furan ring surrogates than for the multiple furan ring furylmethanes. A maximum 97% jet fuel range alkanes with 93% selectivity in C₁₅H₃₂ and C₁₄H₃₀ is achieved from C₁₅-furylmethane under optimal conditions. The yield and selectivity of alkanes with desired carbon numbers can be tuned using furylmethanes with tailored carbon chains, furan numbers, and a carbon center that minimizes C-C cracking. (Chemical Equation Presented).

Full text of DOI:10.1021/acscatal.7b00986

Subscriber access provided by HKU Libraries
Article
Hydrodeoxygenation of Furylmethane Oxygenates to Jet and
Diesel Ranged Fuels: Probing the Reaction Network with
Supported Palladium Catalyst and Hafnium Triflate Promoter
Saikat Dutta, and Basudeb Saha
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.7b00986 • Publication Date (Web): 10 Jul 2017
Downloaded from http://pubs.acs.org on July 16, 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 23
ACS Catalysis
1
2
3
4
5
6
Hydrodeoxygenation of Furylmethane Oxygenates to Jet and Diesel Ranged Fuels: Probing
the Reaction Network with Supported Palladium Catalyst and Hafnium Triflate Promoter
7
8
Saikat Dutta, and Basudeb Saha*
9
Catalysis Center for Energy Innovation and Department of Chemical and Biomolecular
Engineering, University of Delaware, Newark, DE 19716, USA
Corresponding author: B. Saha (bsaha@udel.edu)
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
Abstract
Catalytic hydrodeoxygenation of furylmethane oxygenates to high carbon branched chain jet and
diesel fuel ranged alkanes at mild reaction conditions is a promising strategy for energy-efficient
production of fuels with minimal C-C cracking to undesired products. Here, we report that a
strong Lewis acidic promoter can overcome the energy barrier for furylmethanes
hydrodeoxygenation at lower temperature. Furan rings of furylmethanes are first hydrogenated
to fully saturated cyclic ethers by a hydrogenation catalyst, which then undergo facile ring-
opening and deoxygenation by the promoter. A cyclic intermediate between etheric ‘O’ and the
Lewis acidic metal center, assisted by the triflate ligand of the promoter, is formed in the ring-
opening step. Probing the reaction pathway with symmetric single furan ring surrogate molecules
suggest that the promoter is necessary for the ring-opening. Deoxygenation of ring-opened
oxygenates takes place faster for single ring furan surrogates than the multi-furan rings
furylmethanes. Maximum 97% jet fuel ranged alkanes with 93% selectivity in C15H32 and C14H30
is achieved from C15-furylmethane at optimal conditions. The yield and selectivity of alkanes
with desired carbon numbers can be tuned by two-stage heating, and using furylmethanes with
tailored carbon-chain, furan numbers and the carbon center that minimizes C-C cracking.
Keywords: hydrodeoxygenation, furylmethane, jet fuel, branched alkane, 2-5-dimethylfuran,
ring-opening.
1
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 23
1
2
3
4
Introduction
5
6
7
8
According to International Energy Agency’s Oil Market Report, global crude oil
consumption in 2015 reached about 1.4 trillion gallons of which the United States alone
consumed about one fifth. Direct carbon emissions from crude oil (~92 g CO2 E/MJ) are high
when compared with renewable fuel alternative, e.g., bioethanol from the 1st generation
biomass.1 The United States Renewable Fuel Standard (RFS) act mandates blends of renewable
biofuels with conventional fuel to minimize carbon emissions, leading to a rapid expansion of
bioethanol and other high octane advanced biofuels production from non-food lignocellulosic
biomass. Among advanced biofuels, production of high carbon branched and long chain alkanes
for jet (C8-15) or diesel (C12-22) ranged fuels has received considerable attention in recent years
because the US Federal Aviation Administration (FAA) has set a goal for the US aviation
industry to consume renewable jet fuel from 2018.2 As a result, extensive research and
commercialization activities have been initiated to meet increased renewable jet fuels demand.3
Currently, renewable jet fuels are produced via alcohol to jet (ATJ), hydroprocessing esters and
fatty acids (HEFA), Fischer Tropsch (FT), hydrotreated depolymerized cellulosic jet (HDCJ) and
catalytic conversion of sugars-to jet (CCSTJ) processes.4 Among these processes, the CCSTJ
pathway is of great interest because of abundant supply of non-food biomass, low temperature
processing compared to other high temperature hydrotreatment and high product selectivity.
A major limitation in the direct conversion of carbohydrates to fuels is the low carbon
sugar units (C5/C6) in cellulose and hemicellulose, which are unsuited for jet or diesel ranged
fuels. As a result, direct catalytic hydroprocessing of lignocellulose or cellulose involving
depolymerization, dehydration and hydrodeoxygenation (HDO) pathways produce low carbon
alkanes (hexane and pentane) as the major products.5 A strategy for building such fuel precursors
is C-C coupling of low carbon biomass intermediates by hydroxyalkylation/alkylation (HAA),6
Garcia-Gonzalez (GG)7 and aldol condensation reactions.8 Subsequent hydrogenation, ring-
opening (RO), and deoxygenation of the resulting molecules over heterogeneous catalysts can
produce alkanes of desired carbon numbers.
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
Catalytic HDO of such fuel precursors over metal catalysts, e.g., Pt/C/TaOPO4,9
Pd/Zeolite-,10 Pt/C-Pt/TiO2,11 Pd/H-ZSM-5, Pd/SiO2-Al2O3, Pd/Al2O3,12 Pt/MCM-41,13 Ni-
14a
Mo2C/SiO2
and Pt/C,6a has been reported. These reactions involve multiple steps
(hydrogenation, RO, deoxygenation) and severe reaction conditions (up to 400 C). For example,
2
ACS Paragon Plus Environment

Page 3 of 23
ACS Catalysis
1
2
3
4
5
6
7
8
9
Pt/C alone could not fully deoxygenate the precursor without TaOPO4 at 300 C.9 While
deoxygenation of the precursor occurred over Pt/C in a flow reactor at higher temperature (350 -
400 C) and H2 pressure of 50 bar,6a, 11 Pt on an acidic supports (Al2O3, TiO2) enhanced C-C
cracking at these severe conditions.6a C-C cracking was also observed with the metal-acid
bifunctional catalysts containing weakly acidic sites, especially at temperatures above 300 C
and H2 pressure of 40-50 bar.6a Gradient heating of the reaction mixture from 120 C to 350 C
in the flow reactor suppressed C-C cracking slightly (5-7%). However, mechanistic
understanding, such as the order in which hydrogenation, RO and deoxygenation steps take
place, remains unclear. Recently, Pd/NbOPO4 has been reported as an efficient catalyst for C15
branched alkanes.14b
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
To overcome the challenges associated with high temperature HDO and to minimize the
energy input of the process, an alternate strategy has been devised to deoxygenate the precursors
at reasonably lower temperature (<225 C) using a physical mixture of a hydrogenation catalyst
(Pd/C) and a promoter, e.g., a metal triflate salt.15 The authors proposed the reaction progressed
via RO of unsaturated furan ring to a ketonic intermediate followed by HDO of C=O bonds, and
that ring hydrogenation of the furan first results in a dead-end intermediate. In contrary, Song et
al.16 observed that ring hydrogenation of furan occurs first and the triflate salt acts as an oxygen
remover in opening the hydrogenated ring.
Recently we developed a novel and efficient catalytic pathway to synthesize oxygenated
furylmethane type fuel precursors (Scheme 1) of carbon numbers 12-21 with 95% yield via HAA
reaction of low carbon biomass furans and carbonyl compounds.17 This reaction, operating at
low temperature (60 C) and neat conditions without involving any complex separation and
purification steps, is scalable to produce fuel precursors. Here, we use a C15 fuel precursor, 5,5-
bis(2-methylfuranyl)furan-2-ylmethane (1) as a model furylmethane to study HDO with the Pd/C
catalyst in the presence of a small amount of Hf-triflate (Hf(OTf)4) salt, to elucidate sequence of
steps (hydrogenation, RO and deoxygenation). Compound 1 is an ideal model substrate because
it produces C15 branched alkane, which is preferred for jet fuel application due to their better
cold flow properties than their straight chain analogues (n-alkanes).
We show that 1 undergoes hydrogenation of furan rings to fully saturated furan derivative
with branched carbon backbone that is transformed to a hydroxylated intermediate via RO and
eventually to C15 and C14 alkanes upon deoxygenation. We demonstrate that RO of furan ring
3
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 23
1
2
3
4
5
6
7
8
9
and cleavage of CO bonds are the key steps in post-hydrogenation of furan rings. Single ring
furan molecules, 2,5-demethylfuran (DMF) and 2,5-dimethyltetrahydrofuran (DMTHF) were
used as probe systems to assess the reaction network. The role of Hf(OTf)4 in RO of furan
moieties is elucidated by comparing its activity with the other Hf-salts, zeolite and blank
(without Hf-salt) systems. The reaction conditions are systematically optimized to achieve
maximum 97% yield of branched alkanes containing over 90% C15 and C14 alkanes as the major
products. Structurally diverse fuel precursors (2-4) are used to understand how the nature of
carbon backbone of furylmethane influences RO, HDO and alkanes selectivity.
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
Scheme 1. Hydrodeoxygenation (HDO) of furylmethanes to branched chain alkanes using Pd/C-
Hf(OTf)4 as catalyst.
Experimental Section
Chemicals. Furylmethanes 1, 2, 3, and 4, were synthesized by catalytic HAA of 2-methylfuran
(2-MF) with furfural, butanal, acetone, and 5-hydroxymethylfurfural (HMF), respectively, using
improve graphene oxide (IGO) as a catalyst under neat condition by following our earlier
publication.17 In brief, 2-MF and a carbonyl compound were reacted in 2:1 molar ratio in the
presence of the IGO catalyst at 60 C for 6 h without any solvent. After reaction, the catalyst was
separated by centrifugation and the condensation products were collected by decantation. Pd10/C
(10 wt% Pd), hafnium triflate (Hf(OTf)4), cyclohexane, hafnium chloride (HfCl4), hafnium oxide
(HfO2), triflic acid (TfOH), 2,5-dimethylfuran (DMF), 2,5-dimethyltetrahydrofruan (DMTHF)
were purchased from Sigma-Aldrich, USA. H-BEA-25 was purchased from Zyolist, USA.
Standard C8 – C19 alkane samples (1000 L/g) and standard oil samples (C18 - C35) were
purchased from Absolute Standards INC, USA. These chemicals were used as received without
further purification.
4
ACS Paragon Plus Environment

Page 5 of 23
ACS Catalysis
1
2
3
4
5
6
7
8
9
Catalyst Characterization. Powder X-ray diffraction (XRD) pattern of Pd-samples were
collected at room temperature using a Bruker D8 Advance X-ray diffractometer equipped with a
monochromatized Cu Kα radiation ( = 1.54055 Å) source operating at 45 V and 40 mA.
Surface measurements were carried out on a Micromeritics ASAP 2020 surface area analyzer
using N2 gas physisorption technique after degassing the samples overnight at 80 C. Pulse CO
chemisorption after reduction in H2 was performed on an Altamira AM-200ip instrument. Before
CO chemisorption, the sample was first heated at 120 C and then cooled down to 30 C.
Subsequently, the sample was reduced with H2 at 300 C by ramping the temperature at a heating
rate of 5 C/min and then holding the temperature for 2 h. Pulses of CO in a He carrier gas were
injected at 30 C. The signal was monitored by a thermal conductivity detector (TCD). TEM
images of the catalyst samples were collected on a JEM-2010F (Keck CAMM) aberration
corrected transmission electron microscope (TEM) operating at an accelerating voltage of 120
kV. Samples for TEM images were prepared by applying one drop of dilute suspension of the
catalyst samples dispersed in acetone onto the carbon coated Cu grid and allowing the solvent to
evaporate at room temperature.
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
Catalysis. All reactions were performed in a 50 mL stainless steel Parr reactor (model
4792HP/HT(FH) equipped with a thermocouple, pressure gauge, rupture disk and gas release
valve. In a typical experiment, 1 (1.23 mmol), Pd10/C (0.106 g, 0.1 mmol Pd), Hf(OTf)4 (0.04 g,
0.05 mmol) and cyclohexane (25 mL) were loaded in the reactor. Subsequently, the reactor head
was sealed, purged with ultra pure H2 for three times and then pressurized to the desired
pressure. The pressurized reactor was placed in a band heater connected to a heating controller
and heated at 15 C/min heating rate and held at desired reaction temperature for the set time.
The solution was stirred at a constant rpm using a magnetic bar. Upon reaction, the reactor was
cooled down in an ice bath and released the pressure. The reaction mass was transferred to a
centrifuge tube to separate the solid catalyst by centrifugation. The supernatant liquid was
decanted and used for GC-MS or GC analysis upon appropriate dilution with cyclohexane.
Similar experimental procedure was followed to test furylmethanes 2-4.
Product analysis. The conversion of precursors and the yields of alkanes and other oxygenated
intermediates were determined on an Agilent 7890A GC connected to a MS Agilent 5977A
MSD. A DB-5 column of dimension 30 m  250 μm  0.25 μm was used for analysis. The
operating conditions of the GC-MS are as follows: injection volume 1.0 L, inlet temperature
5
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 23
1
2
3
4
5
6
7
8
9
250 C, detector temperature 250 C, and a split ratio 1:20. The initial column temperature was
40 C (5 min) which was ramped at a rate of 15 C/min and the final temperature was held at 320
C for 10 min. Alkanes and other oxygenated intermediates were identified by their m/z numbers
obtained from the mass hunter program and by comparing their retention times with standard
samples. Actual concentrations of each identified products were quantified from their pre-
calibrated plots of count numbers versus concentrations developed using standard alkane
samples. HDO products of DMF and DMTHF were analyzed on a Shimadzu-GC-MS (model
QP2010) and a GC (Agilent 7890A) equipped with an FID detector and HP-INNOWAX
capillary column of dimension 0.25 m ID  250 m  30 m. The operating conditions of the
GC are as follows: injection volume 1.0 L, inlet temperature 250 C, detector temperature 250
C, and a split ratio of 1:10. The initial column temperature was 30 C (hold time of 5 min)
which was ramped at a rate of 15 C/min and the final temperature was held at 250 C for 5 min.
Products were identified comparing their retention time with the retention time of the standard
samples and by mass numbers from GC-MS analysis. Unless otherwise mentioned, the yields of
alkanes and identified oxygenated intermediates were calculated on molar basis of total carbons.
Total yield is the sum of all quantified alkanes and oxygenated intermediates.
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

Alkanes with less than 8 carbon numbers and some oxygenated intermediates were not
identified, resulting in a loss of carbon balance.
Results and Discussion
Screening reaction conditions and promoters. Key steps in converting furylmethanes to the
corresponding branched chain alkanes (Scheme 1) include hydrogenation, RO and
deoxygenation of oxygenated RO intermediates. Mechanistic understanding of the sequence of
these steps can make the process energy-efficient and improve the yield of alkanes of desired
carbon numbers. To achieve this, we studied HDO of 1 using a physical mixture of Pd/C and a
Lewis acidic metal triflate salt as a promoter. Metal triflate salts, M(OTf)3, (M = L, Ce, Sm, Yb,
Sc, Fe, Hf, Al) have been used to promote RO and oxygen removal of furans.15-16, 18 However,
the exact role of the promoter remains elusive and contradictory. While one group found the
ring-hydrogenated furan is a dead-end intermediate,15 another group16 reported furan ring
6
ACS Paragon Plus Environment

Page 7 of 23
ACS Catalysis
1
2
3
4
5
6
7
8
hydrogenation occurs first before RO and oxygen removal by the triflate salt. We choose
Hf(OTf)4 as a promoter because of its high effective charge density at the metal center to
facilitate RO by activating C-O linkages of furan rings of 1.
First, we screened various reaction conditions (H2 pressure, reaction time and
temperature) with and without Hf(OTf)4. Pd10/C alone yields only 22% C15H32 and low carbon
alkanes (C14H30, C11H24, C9H20) via C-C cracking. The total yield of all identified alkanes and
fully furan ring saturated oxygenated intermediate (C15H25O3, 6) is 53% (entry 2, Table 1). The
addition of a small amount of Hf(OTf)4 (0.05 mmol) results in significantly higher C15H32 (62%)
and total alkane yields (95%; entry 7 of Table 1), with a very small amount of unconverted
oxygenated intermediates and minimal C-C cracking (producing C14H30 as the only smaller
alkane). A blank reaction without Pd/C and Hf(OTf)4 shows no conversion of 1 (entry 1, Table
1). A reaction using half the amount of Hf(OTf)4 (0.05 mmol) yields less alkanes (entry 8). A
two-stage heating experiment with Pd10/C-Hf(OTf)4 (1st stage at 50 C for 15 h followed by 2nd
stage at 180 C for 20 h) improves C15H32 (47%) and total alkane yields (70%) with fewer C-C
cracking products (entry 4, Table 1) when compared with one-stage experiment (180 C for 24
h). The screening experiments (Table 1) suggest that the reaction at 225 C offers the highest
alkane yield and C15H32 selectivity and minimal unconverted intermediate oxygenates by
overcoming the energy barrier for C-O bond cleavage.19 Our maximum operating temperature is
still lower than those reported at higher temperatures for similar HDO reactions of substrates
with branched carbon backbone.11,13,14 This reduction in temperature, and hence the
minimization in C-C cracking, is possible because of efficient C-O cleavage and oxygen removal
by the triflate salt. Noteworthy, our reaction temperature is comparable to those used for HDO of
aldol condensation derived fuel precursors containing one furan ring in which linear alkanes or
oxygenates were formed as products.15,16 Increasing H2 pressure from 25 to 50 atm accelerated
HDO of intermediate 6, resulting in increased yield of desired C15 and C14 alkanes (entries 5-7,
Table 1).
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
Since the triflate salt can dissociate to triflic acid (HOTf) by water formed during HDO,
we also used HOTf as a promoter. The results in Table 1 (entry 9) reveal that HOTf promotes
C15H32 (35%) and C14H30 (20%) yields when compared with Pd10/C alone because of proton
assisted RO of furan moieties of 1. Another reaction using 4 equivalent of HOTf, maximum
7
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 23
1
2
3
4
Table 1. Screening of reaction conditions and promoters for HDO of 1 with Pd10/C.
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
Entry Promoter
H2 (atm)
Yields (mol %)
T (C)
C9H20 C10H22 C11H24 C14H30 C15H32
6
-
Total yield (mol %)
a
1
2
-
225
225
180
50
50
50
50
25
30
50
50
50
50
50
50
50
-
-
-
-
-
-
-
4.2
4.5
4.5
2.8
4.3
1
4
8.0
5.8
2.0
-
17.7
8.2
18
11
63
67
73
78
91
97
76
88
94
74
50
59
3
Hf(OTf)4
Hf(OTf)4 2-stage heatingb
3.2
5.4
4.7
5.4
2.5
4
22.3 21.5
47.1 3.3
4
10.4
15.0
19.5
28
5
Hf(OTf)4
Hf(OTf)4
Hf(OTf)4
225
225
225
225
225
225
225
225
225
19.1 35.1
32.9 24.8
6
2.7
2
7
62
55
35
44
15
7
2.5
7
c
8
Hf(OTf)4
HOTf
0
1
18
9
2.4
4.0
4.2
1.8
1.8
3.3
5.0
4
4.4
3.0
7.3
8.8
3.5
19.5
22
22
15
32
28
7
10
11
12
13
HOTfd
HfCl4
11.3
9.3
HfO2
4
H-BEA-25
5.5
18.5
21
Pd10/C and promoters were physically mixed. aNo Pd10/C. bTwo staged heating – 1st stage at 50 C for 15 h and 2nd stage at 180 C for
20 h. cHf(OTf)4 = 0.025 mmol. dHOTf = 0.05 mmol. Other reaction conditions: Hf(OTf)4 (0.040 mg; 0.05 mmol), HOTf (7.5 mg, 0.05
mmol), HfCl4 (0.015 g, 0.05 mmol), HfO2 (0.010 g, 0.05 mmol), H-BEA-25 (0.050 g), cyclohexane (25 mL), 24 h.
8
ACS Paragon Plus Environment

Page 9 of 23
ACS Catalysis
1
2
3
4
5
6
7
8
9
amount of HOTf could be formed from complete dissociation of one equivalent of Hf(OTf)4, the
yield of jet fuel ranged alkanes improved slightly (entry 10, Table 1) than the reaction using 1
equivalent HOTf. However, HOTf is not as effective as triflate salt in total alkane yields and C15
alkane selectivity, indicating that the Lewis acidic metal center of Hf(OTf)4 predominantly
catalyzes the C-O cleavage of the furan rings. Heterogeneous acidic zeolite (HBEA-25)
performed worse than the homogeneous acid due to potentially lower accessibility of the acid
sites in its micropores by the bulky molecules 1; importantly, the zeolite promotes significant C-
C cracking. Another Lewis acidic Hf salt, HfCl4, formed much less C15H32 (15%) and C14H30
(11%) products than those obtained with Hf(OTf)4 or acid (HOTf, HBEA-25) promoters, and the
majority of intermediate oxygenate, 6, remained unconverted (32%). This suggests that HfCl4
could not activate C-O bonds of furan rings through its binding with ‘O’, presumably due to its
poor Lewis acidity in cyclohexane or poor dissociation to generate Brønsted acid (HCl) in water
formed during HDO. Similarly, HfO2 also could not promote HDO, leading to a significant
amount of 6 (28%).
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
Noteworthy, Hf(OTf)4 is the most effective among the promoters tested because of its
unique characteristics. It contains four strong electron withdrawing triflate ligands (OTf = -
OSO3CF3) with three fluorine atoms in each ligand that makes the Hf center highly electron
deficient and strongly Lewis acidic, with strong affinity to bind the nucleophilic ‘O’ of –C-O
linkages of furan moieties; it thus activates the etheric ‘O’ towards RO (Scheme 2). Further, ‘O’
atoms of sulfonate group (-SO3) of triflate ligand could facilitate cleavage of -C-H bond
through formation of an intermediate (species B) and accelerate RO of Hf-bonded ether to Hf-
bonded alcohol (species C). Marks et al. have shown a similar pathway for opening of C-O
linkages of 5-and 6-members hydrogenated rings with Hf(OTf)4.18a
Scheme 2. Hf(OTf)4 assisted ring opening (RO) of hydrogenated furan ring oxygenate 6. F and
MF are furan and 2-methylfuran, respectively.
9
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 23
1
2
3
4
5
6
7
8
9
Next we studied the effect of Pd loading by varying the molar ratios of 1 to Pd. These
experiments at 1:Pd molar ratios of 12 - 44 reveals that the maximum yield of C15 alkane is
achieved at the molar ratio of 12 (Figure 1a). A gradual decrease in C15 alkane yield with high
accumulation of unconverted oxygenate 6 is observed with decreasing Pd amount. In addition,
several other oxygenate intermediates remain undetected, contributing to carbon loss at low Pd
amount.
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
Upon optimization of reaction temperature (225 C), H2 pressure and Pd amount (1:Pd
(Pd10/C) = 12), we varied the reaction time from 5 h to 24 h (Figure 1b). Intermediate 6 is
formed as the major product in 5 h with total detected carbon yield of little over 80%. This
intermediate is fully converted to C15 and C14 alkanes after prolonged reaction (24 h).
Interestingly, the total detected carbon yield for 24 h reaction is much higher than that of 5 h
reaction. This suggests that the product solution in 5 h reaction contained some undetected
oxygenates (Figure 2 shows peaks for undetected oxygenates between RT 16-18 min in GC
chromatogram) that are also converted to alkanes at long times.
Figure 1. Hydrodeoxygenation (HDO) results of compound 1 with Pd10/C-Hf(OTf)4 at different
(a) molar ratios of 1 to Pd and (b) reaction time. Reaction conditions: 1 (0.300 g; 1.23 mmol),
Hf(OTf)4 (0.05 mmol), H2 (50 atm), cyclohexane (25 mL), 225 ºC. Pd amount was varied from
0.028 mmol to 0.1 mmol. Exact amount of Pd10/C used for 1:Pd variation experiment are in the
supporting information (Table S1). Table S2 shows data used in Figure b. Reaction time for 1:Pd
experiments is 24 h.
10
ACS Paragon Plus Environment

Page 11 of 23
ACS Catalysis
1
2
3
4
5
6
7
8
9
Time Profile of HDO with and without Hf(OTf)4. To understand what happens in the early
stage of the reaction pathway towards alkane formation, especially ring hydrogenation, RO and
oxygen removal, we investigate the reaction profile at early stages at room temperature (25 C)
and at 225 C. The reactor was quenched at specific times in an ice-bath and the solution was
analyzed by GC-MS. First, 1, catalyst (Pd10/C-Hf(OTf)4) and solvent were added in the reactor,
stirred the mixture for 10 min without pressurizing the reactor with H2. The chromatogram in
Figure 2 shows a single peak for 1, indicating no reaction occurs without H2. Then we purged the
reactor with ultrapure H2 for three times and hold the pressure at 50 atm and 25 C for 10 min
with continuous stirring. Two new peaks appear, corresponding to intermediate 6 and a partially
furan ring hydrogenated oxygenated intermediate 5. This result indicates that hydrogenation of
furan ring is facile at 25 C.
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
RO of 6 to ring-opened oxygenate (7), alkanes and other unidentified products occurred
during heating to 225 C (heating rate 15 C/min). A significant amount of 7 with a small
amount of alkanes formation during heating indicates that RO of 6 to 7 is fast. RO step is
depicted in Scheme 2. At long times, C15 and C14 alkanes form via deoxygenation,
dehydrogenative decarbonylation and hydrogenation, along with small amounts of C9-C11
alkanes via C-C cracking. Intermediates with di or tri hydroxyl groups are also expected to form
from 7 via opening of two other furan rings (Scheme 3). However, the latter two oxygenates
were not identified, possibly due to their faster conversion to alkanes as they are formed from 7.
Notably, the GC chromatogram after 24 h reaction shows fewer, small peaks for unidentified
species, accounting for a small loss in carbon balance (3%), compared to shorter times (e.g., 5 h).
The corresponding reaction profile showing yields of various products as a function of time is
shown in Figure S1.
11
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 23
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 2. GC chromatograms obtained during HDO of 1 with Pd10/C-Hf(OTf)4; major
compounds and retention times on the right. Reaction conditions: 1 (0.300 g; 1.23 mmol),
Hf(OTf)4 (0.05 mmol), H2 (50 atm), cyclohexane (25 mL), temperature and time as indicated in
the graph. 0 h is the time when the temperature of the reaction mixture reached 225 C. The
heating rate is 15 C/min.
In the absence of Hf(OTf)4, ring hydrogenation of 1 to 6 follows a similar trend but RO
of 6 to 7 is not evident at 225 C, even after 1 h of reaction at 225 C (Figure S2). In contrast,
two unidentified peaks are observed at slightly higher retention times of 6. Thus, the RO
pathway without Hf(OTf)4 is significantly different from that observed with Hf(OTf)4. In the
latter case, facile RO of 5 to 7 and subsequent HDO of 7 to alkanes occur at 225 C. This
difference in results with and without Hf(OTf)4 further confirms significant roles of the promoter
in RO. Prolonged reaction (24 h) without the promoter at 225 C shows peaks for alkanes (entry
2, Table 1) and several unidentified peaks for oxygenates, which were not observed in the
reaction with Hf(OTf)4, and account for higher carbon loss (37%).
12
ACS Paragon Plus Environment

Page 13 of 23
ACS Catalysis
1
2
3
4
5
6
7
8
9
Reaction pathway and network. A possible reaction network is proposed in Scheme 3. First,
hydrogenation of furan rings of 1 over Pd catalyst gives the fully ring hydrogenated intermediate
6. Second, RO of 6 to 7 occurs through formation of an adduct between strong Lewis acidic Hf
center of Hf(OTf)4 and the ‘O’ of the furan ring and concurrent elimination of a H at the 
carbon of the ‘O’. Subsequent CO cleavage forms a hydroxyl intermediate with an olefinic
C=C group, consistent with the prior work.18a A controlled experiment using lower H2 pressure
(6 atm) reveals the formation of 7 is independent on the H2 pressure (Figure S3), which
strengthens our proposition of -H elimination during ring-opening. Two other ring saturated
furans of 7 are also expected to be similarly opened to hydroxylated intermediates (8). Third,
cleavage of Hf-bonded hydroxylated intermediate to Hf-H2O and concurrent hydrogenation of
intermediate forms C15 alkane. Fourth, we observed formation of a significant amount of C14
alkane, which can be explained by dehydrogenative decarbonylation of primary alcohol moiety
of 7. Dehydrogenative decarbonylation of primary alcohols such as benzyl alcohol and HMF on
Pd catalysts 20a, 20b and butanol over Pt, Ru catalysts20c have been reported. Proposed pathway for
dehydrogenative decarbonylation of benzyl alcohol involved adsorption of primary alcohol
moiety of benzyl alcohol (CH2OH) onto the Pd(111) surface, followed by dehydrogenation of
the alcohol to an aldehyde and finally decarbonylation of derived aldehyde. Both
dehydrogenative decarbonylation20a and decarbonylation20c can occur on the Pd surface. In
analogy to benzyl alcohol conversion, we propose the primary alcohol group of a portion of
intermediate 7 is dehydrogenated an intermediate 9, which is subsequently converted to
intermediate 10 via RO of the two remaining furan rings, similar to formation of intermediate 8.
Intermediate 10 is then decarbonylated to C14 alkane as shown in Scheme 3. Noteworthy, RO of
two other methyl substituted furan rings will form secondary alcohols, which are unable to
decarbonylate via dehydrogenation like the primary alcohol. We believe CC cracking of furan
rings at tertiary carbon of intermediate 8 also occurs, which forms intermediates (oxygenated or
alkenes (11 and 12)) for C10 and C11 alkanes. Alkenes are hydrogenated over Pd/C to the
corresponding alkanes. C9 and other smaller alkanes could form from C-C cracking of high
carbon alkanes.
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
13
ACS Paragon Plus Environment

ACS Catalysis
Page 14 of 23
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
Scheme 3. Possible reaction network for conversion of 1 to alkanes using Pd/C-Hf(OTf)4
catalyst.
Probing reaction pathway using single furan molecules. To further elucidate the mechanism
over Pd/C-Hf(OTf)4 catalyst, we use simple surrogates of 1 such as DMF and DMTHF. A recent
study describes full ring hydrogenation of DMF over a Pd/C catalyst without RO,21 consistent
with the findings of Sutton et al.15 RO of unsaturated furan over a Pt/C catalyst was observed,
which parallels with computational and experimental work on RO of substituted (-CHO, -
CH2OH, no side group) single furan ring molecules over a Ir(111) single crystal catalyst. In the
latter case, the RO paths depend on the degree of ring saturation and the electonegativity of the
side-chain functional groups.22 DMF conversion over Pt/C and bimetallic PtCo/C (Pt3Co/C and
Pt3Co2/C) catalysts showed23 that DMF hydrogenation to DMTHF and RO to 2-hexanone, 2-
hexanol, 2-propoxyhexane and hexane are facile over Pt/C due to strong chemisorption of furans
on the metal surface. Monolayer oxide (Co3O2) of Pt3Co2/C caused weaker interaction of furans
on the surface, resulting in lower DMF conversion. Pt3Co exhibited intermediate catalytic
activity. These results suggested furan RO occur via ring hydrogenation upon accessing exposed
metal surface. Another study reported RO of DMF over Ru/C occurs via parallel pathways in
which both partial saturated and fully saturated furan rings opened,24 an observation supported
14
ACS Paragon Plus Environment

Page 15 of 23
ACS Catalysis
1
2
3
4
5
6
7
8
by DFT calculations and microkinetic modeling. Catalytic conversions of furfural and HMF to
low carbon alkanes involving hydrogenation of C=O bonds and furan-rings and C-O
hydrogenolysis have also been reported.25
We observe that the DMF conversion follows a similar pathway as 1 with Pd10C-
Hf(OTf)4 leading to hexane as the major product (Figure 3A). First, DMF undergoes rapid
hydrogenation to DMTHF and partially ring hydrogenated DMF-H2 at 25 C and 50 atm H2 with
full conversion of DMF (Scheme 4). A significant fraction of hexane forms during heating the
reaction mixture to 225 C at 50 atm H2. Both fully hydrogenated and partially hydrogenated
furans could undergo RO by Hf(OTf)4 during heating as the GC peak intensity of both species
decreased, although RO of partial ring-saturated furan is believed to be faster as per energy
calculation reported previously.22 The remaining furans (partially and fully saturated)
quantitatively convert to hexane in significantly shorter time (1 h) than that of multi-furans
furylmethane 1 (24 h) and with much cleaner chromatogram. Unlike furylmethane, ring-opened
hydroxylated intermediate was not observed for DMF. This could be attributed to the faster
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
cleavage of the -COH∙∙∙∙Hf linkage of the hydroxylated species (Scheme 2). Comparison of
furylmethane conversion with its surrogate molecule suggests that Hf(OTf)4 is much more
effective and selective for RO of partially or fully hydrogenated furan of a single ring (DMF).
This may be due to steric hindrance for stable adduct formation between Hf4+, furan ‘O’ and –
OSO3CF3 groups by the bulky furylmethane molecule as shown in Scheme 2.
A room temperature reaction of DMF with Pd10/C alone shows formation of only
DMTHF and DMF-H2 in 2 h at 25 C and 50 atm H2. RO and deoxygenation start upon addition
of Hf(OTf)4 and heating the reaction mixture to 225 C (Figure 3B). Similarly, DMTHF is the
main product from DMF at high temperature (225 ºC) without Hf(OTf)4 in 1 h (Figure S4). The
RO step starts only after addition of Hf(OTf)4 into the resulting DMTHF leading to formation of
hexane. This result further illustrates that Hf(OTf)4 is necessary to open fully hydrogenated ring.
15
ACS Paragon Plus Environment

ACS Catalysis
Page 16 of 23
1
2
3
4
5
Hexane
(A)
(B)
6
7
Hexane
cis-DMTHF
cis-DMTHF
O
O
O
trans-
DMTHF
O
trans-
DMTHF
8
9
DMF
O
Hexane
DMF
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
DMF-H2
O
(a)
(a)
(b)
(b)
(c)
(c)
(d)
(d)
(e)
0
1
2
3
4
5
1
2
3
4
5
Retention Time (min)
Retention Time (min)
Figure 3. (A) GC chromatograms obtained during HDO of DMF with Pd10/C-Hf(OTf)4: (a)
without H2, (b) with H2 added at 25 C after 10 min, (c) after the reaction mixture reached to 225
C, and (d) after reaction for 1 h at 225 C. Other conditions: Pd10/C (106 mg; 0.1 mmol),
Hf(OTf)4 (0.040 g; 0.05 mmol), DMF (0.118 g, 1.23 mmol), cyclohexane (10 mL), 50 atm H2.
(B) GC chromatograms obtained during HDO of DMF (0.123 g, 1.23 mmol) with Pd10/C (0.106
g, 0.1 mmol Pd) in the absence of Hf(OTf)4: (a) without H2, (b) with H2 added at 25 C after 40
min, (c) after 1 h, (d) after 2 h and (e) after addition of Hf(OTf)4 (0.040 g, 0.05 mmol) and
heating to 225 C in cyclohexane (10 mL). DMF-H2, trans- and cis-DMTHF and hexane have
the same RT in all chromatograms.
Scheme 4. Dimethylfuran conversion with and without Hf(OTf)4.
Effect of furylmethane structure on alkane selectivity. We found C-C cracking of oxygenated
intermediate of 1 caused some selectivity loss of desired C15H32 product. It is hypothesized that
16
ACS Paragon Plus Environment

Page 17 of 23
ACS Catalysis
1
2
3
4
5
6
7
8
9
C-C cracking can be minimized by tuning molecular structures of furylmethanes with tailored
carbon-chain, furan numbers and the carbon center. Answers to this hypothesis will be critical in
designing and developing appropriate fuel precursors for selective production of high carbon
branched chain alkanes. In addition to 1, we choose three more furylmethanes (2, 3 and 4;
Scheme 1) with different numbers of furan rings and nature of carbon center (tertiary versus
quaternary) for the comparative study. These furylmethanes were subjected to react with Pd10/C-
Hf(OTf)4 at the optimal reaction conditions developed for 1 (entry 7, Table 1). Figure 4
compares the products distribution of alkanes at the full furylmethanes’ conversion. C14
furylmethane (2) with two furan rings and a butane chain yields the corresponding C14 alkane
with the highest selectivity (92% yield). Precursor 4, C21 furylmethane with four furan rings,
yields 60% C20 alkane out of total detected carbon yield of 63%, indicating that the larger
molecule with several furan rings could form several undetected oxygenated intermediates.
Precursor 3 containing two furan rings and a quaternary carbon gives more C-C cracked alkanes
than 4 with alkane yield of maximum carbon number (C13) of 52%. Figure 4 suggests that
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
120
C15H26O3
O
O
O
C18H38
C16H34
C15H32
C14H30
C13H29
C11H24
C10H22
C9H20
C8H18
C7H16
1
100
80
60
40
20
0
O
2
O
O
O
3
O
4 O
Furylmethanes
Figure 4. Yields of alkanes during HDO of different furylmethanes over Pd10/C-Hf(OTf)4.
Reaction conditions: Pd10/C (106 mg; 0.1 mmol), Hf(OTf)4 (0.040 g; 0.05 mmol), furylmethanes
(1.23 mmol), cyclohexane (25 mL), 50 atm H2, 225 C, 24 h. Table S3 tabulates the raw data.
furylmethanes with fewer furan rings and a straight carbon-chain unit are beneficial to produce
alkanes with high selectivity, e.g., 2. In contrast, precursors with multiple furan rings possibly
17
ACS Paragon Plus Environment

ACS Catalysis
Page 18 of 23
1
2
3
4
5
6
form many ring-opened oxygenates, from which multiple C-C cracking reactions could occur
during deoxygenation.
7
8
9
Another important observation is that 1 forms significant amount of alkane with one
carbon less than its parent molecule, which was not observed for other furylmethanes. This can
be explained by the presence of one un-substituted furan in 1, which forms a primary alcohol
upon RO and undergoes dehydrogenative decarbonylation to C14 alkane. Other furylmethanes
containing methyl substituted furan rings in their molecular structure will form secondary
alcohols upon RO and render their dehydrogenative decarbonylation to alkanes with one carbon
less than their parent molecules.
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
Comparative analysis. A comparative analysis of C12-C21 furylmethanes HDO to alkanes with
Pd/C-Hf(OTf)4 and other reported catalysts is presented in Table 2. It shows the yield of alkanes
depends on the reaction conditions and the nature of the catalysts. A Pt catalyst containing mixed
Al2O3 and carbon supports achieved low alkane yield (68%) from C14 furylmethane at high
temperature (350 C; entry 2).6a At the same temperature, C15 furylmethane over the Pt catalysts
on acidic Al2O3 and weakly acidic TiO2 supports, along with Pt/C, yielded slightly higher
alkanes (85%; entries 1 and 3). In contrast, the Pd/C-Hf(OTf)4 catalyst in our study yields much
higher alkanes (94-97%) from both C14 and C15 furylmethanes at lower temperature (entries 8-9).
A Eu(OTf)3-Pd/C catalyst has also achieved high alkane yield from C12 furylmethane, indicating
the triflate salt is beneficial for RO.26 A Pd/NbOPO4 catalyst gave similar C15 alkane yield (84%
C15H32, entry 6) as the Pt catalysts over alumina and titania supports from C15 furylmethane.14b
Notably, the Pd/NbOPO4 catalysis was conducted at considerably mild temperature like us.
NbOx species activated the C-O RO of furans, consisting with a prior report,27 and Brønsted
acidic phosphate sites promoted the dehydration step. Another NbOPO4 catalyst with weaker
hydrogenation metal (Ru) gave much less C15H32 (26%) and total alkanes (67%),14b indicating
hydrogenation activity of the metal is important for ring hydrogenation and post-hydrogenation
of RO intermediates. For the same reason, Ni/NbOPO4 catalyst produced only 16% jet fuel
ranged alkanes (entry 8). In another work, a Ni catalyst over ZSM-5 support achieved higher C15
alkane (50%) owing to higher reaction temperature (260 C), higher H2 pressure (60 atm) and
mesoporous zeolite support (entry 9).28 The authors found that higher Ni dispersion on the ZSM-
18
ACS Paragon Plus Environment

Page 19 of 23
ACS Catalysis
1
2
3
4
Table 2. Hydrodeoxygenation results of furylmethanes with different catalysts.
5
6
7
8
Alkane
yield
(%)
85
T
(C)
C numbers of
furylmethane
PH2
(atm)
Conv.
(%)
Entry
Catalysts
t (h)
Solvents
Reference
a
1
2
3
C15
C14
C15
Pt/C-TiO2
Pt/C-Al2O3
Pt/C-Al2O3
350
350
350
-
-
-
-
50
50
50
>99
>99
>99
11
11
11
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
a
-
94
85
a
-
a
4
C14
Pd/C (3%Pd)
350
-
-
50
>99
20
11
5
6
7
8
9
C12
C15
C15
C15
C15
Pd/C- Eu(OTf)3
Pd/NbOPO4
Ru/NbOPO4
Ni/NbOPO4
Ni/H-ZSM-5
Ni/H-ZSM-5-
200
200
200
200
260
16
12
12
12
EtOAc
cyclohexane
cyclohexane
cyclohexane
-
40
40
40
40
60
>99
100
>99
>99
>50
98
99
67
16
50
26
14b
14b
14b
28
a
-
a
C15
260
-
-
60
>99
70
28
10
EG
11
12
13
14
C15
C14
C13
C21
Pd/C-Hf(OTf)4
225
225
225
225
24
24
24
24
cyclohexane
cyclohexane
cyclohexane
cyclohexane
50
50
50
50
>99
>99
>99
>99
97
94
75
62
This work
This work
This work
This work
Pd/C-Hf(OTf)4
Pd/C-Hf(OTf)4
Pd/C-Hf(OTf)4
areactions in a flow reactor. EG = Ethylene glycol.
19
ACS Paragon Plus Environment

ACS Catalysis
Page 20 of 23
1
2
3
4
5
6
5 support via ethylene glycol (EG) assisted synthesis of the catalyst enhanced C15 alkane yield to
70% (entry 10).28
7
8
9
Conclusions
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
We demonstrated that commercial Pd10/C catalyst with a small amount of Hf(OTf)4 promoter is
effective for conversion of C13-C21 furylmethane oxygenates to the corresponding branched
chain alkanes with high yield. The reaction occurs at considerably lower temperature because of
facile ring-opening of furan units by strong Lewis acidic Hf center of the promoter. Time profile
and controlled experiments depict that the reaction progresses sequentially via furan ring
hydrogenation, ring-opening, deoxygenation and hydrogenation pathways. Hf(OTf)4 facilitates
ring-opening of furan via formation of an adduct between Hf4+ and nucelophilic ‘O’ of the ring
and activating –C-O linkages. ‘O’ atoms of sulfonate group (-SO3) of triflate ligand likely
facilitate cleavage of  C-H bond through formation of a cyclic intermediate and accelerate the
ring-opening of Hf4+-bonded ether to Hf4+-bonded hydroxyl intermediates. Subsequent cleavage
of –C-OH linkages of –C-OH∙∙∙Hf4+ adduct and hydrogenation of intermediates form branched
alkanes. The intermediate with primary alcohol also undergoes dehydrogenative decarbonylation
reaction to alkane. Probe reactions with single furan ring surrogate molecules suggest that the
reaction is faster for single ring molecules than the multi-ring furylmethanes and Hf(OTf)4 is
necessary for ring-opening. An optimal reaction condition achieves 97% yield of jet fuel ranged
alkanes with 93% selectivity in C15H32 and C14H30.
Acknowledgements
This work is conducted with financial support as part of the Catalysis Center for Energy
Innovation, an Energy Frontier Research Center funded by the U. S. Department of Energy,
Office of Science, Office of Basic Energy Sciences under Award Number DE-SC0001004.
Authors acknowledge Professor Dion Vlachos, Catalysis Center for Energy Innovation,
University of Delaware for his valuable time providing helpful technical suggestions and
feedback on the manuscript.
20
ACS Paragon Plus Environment

Page 21 of 23
ACS Catalysis
1
2
3
4
5
6
7
8
Electronic supplementary material (ESI). HDO experimental results at varying Pd10/C:1
molar ratio and reaction time, carbon-chain-length of substrates, reaction profile and GC
chromatograms.
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
References
1. Liska, A. J.; Yang, H. S.; Bremer, V. R.; Klopfenstein, T. J.; Walters, D. T.; Erickson, G. E.;
Cassman, K. G. J. Ind. Ecol. 2009, 13, 58-74.
2. Davidson, C.; Newes, E.; Schwab, A.; Vimmerstedt, L. National Renewable Energy
Laboratory, Retrieved September 2014, 5, 2014.
3. Hari, T. K.; Yaakob, Z.; Binitha, N. N., Aviation biofuel from renewable resources: routes,
opportunities and challenges. Renew. Sustainable Energy Rev. 2015, 42, 1234-1244.
4. Blakey, S.; Rye, L.; Wilson, C. W., Aviation gas turbine alternative fuels: A review.
Proceedings of the combustion institute 2011, 33, 2863-2885.
5. (a) Op de Beeck, B.; Dusselier, M.; Geboers, J.; Holsbeek, J.; Morre, E.; Oswald, S.; Giebeler,
L.; Sels, B. F. Energ. Environ. Sci. 2015, 8, 230-240; (b) Liu, S.; Tamura, M.; Nakagawa, Y.;
Tomishige, K.. ACS Sustainable Chem. Eng. 2014, 2, 1819-1827; (c) Xia, Q.; Chen, Z.; Shao,
Y.; Gong, X.; Wang, H.; Liu, X.; Parker, S. F.; Han, X.; Yang, S.; Wang, Y. Nat Commun 2016,
7, 11162.
6. (a) Corma, A.; de la Torre, O.; Renz, M. Energ. Environ. Sci. 2012, 5, 6328-6344; (b) Li, G.;
Li, N.; Wang, Z.; Li, C.; Wang, A.; Wang, X.; Cong, Y.; Zhang, T. ChemSusChem 2012, 5,
1958-1966.
7. Sutton, A. D.; Kim, J. K.; Wu, R.; Hoyt, C. B.; Kimball, D. B.; Silks, L. A.; Gordon, J. C.
ChemSusChem 2016, 9, 2298-2300.
8. (a) Mascal, M.; Dutta, S.; Gandarias, I. Angew. Chem. Int. Ed. 2014, 53, 1854-1857; (b) Yang,
J.; Li, N.; Li, G.; Wang, W.; Wang, A.; Wang, X.; Cong, Y.; Zhang, T. ChemSusChem 2013, 6,
1149-1152.
9. Liu, D.; Chen, E. Y. X. ChemSusChem 2013, 6, 2236-2239.
10. (a) Bohre, A.; Saha, B.; Abu‐Omar, M. M. ChemSusChem 2015, 8, 4022-4029; (b) Bohre,
A.; Dutta, S.; Saha, B.; Abu-Omar, M. M. ACS Sustainable Chem. Eng. 2015, 3, 1263-1277.
11. Corma, A.; de la Torre, O.; Renz, M.; Villandier, N. Angew. Chem. Int. Ed. 2011, 123, 2423-
2426.
21
ACS Paragon Plus Environment

ACS Catalysis
Page 22 of 23
1
2
3
4
5
6
12. Yang, J.; Li, N.; Li, S.; Wang, W.; Li, L.; Wang, A.; Wang, X.; Cong, Y.; Zhang, T. Green
Chem. 2014, 16, 4879-4884.
7
8
9
13. Wen, C.; Barrow, E.; Hattrick-Simpers, J.; Lauterbach, J. Phy. Chem. Chem. Phy. 2014, 16,
3047-3054.
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
14. (a) Li, S.; Li, N.; Li, G.; Wang, A.; Cong, Y.; Wang, X.; Zhang, T. Catal. Today 2014, 234,
91-99; (b) Xia, Q.; Xia, Y.; Xi, J.; Liu, X.; Zhang, Y.; Guo, Y.; Wang, Y. ChemSusChem 2017,
10, 747-753.
15. Sutton, A. D.; Waldie, F. D.; Wu, R.; Schlaf, M.; Louis III, A.; Gordon, J. C. Nature Chem.
2013, 5, 428-432.
16. Song, H. J.; Deng, J.; Cui, M. S.; Li, X. L.; Liu, X. X.; Zhu, R.; Wu, W. P.; Fu, Y.
ChemSusChem 2015, 8, 4250-4255.
17. Dutta, S.; Bohre, A.; Zheng, W.; Jenness, G. R.; Núñez, M.; Saha, B.; Vlachos, D. G. ACS
Catal. 2017, 7, 3905-3915.
18. (a) Li, Z.; Assary, R. S.; Atesin, A. C.; Curtiss, L. A.; Marks, T. J. J. Am. Chem. Soc. 2013,
136, 104-107; (b) Atesin, A. C.; Ray, N. A.; Stair. P. C.; Marks, T. J. J. Am. Chem. Soc. 2012,
134, 14682-14685. (c) Assary, R. S.; Atesin, A. C.; Li, Z.; Curtiss, L. A.; Marks, T. J. ACS
Catal. 2013, 3, 1908-1914.
19. Wang, Y.; Xia, Q.; Xia, Y.; Xi, J.; Zhang, Y.; Guo, Y.; Liu, X. ChemSusChem 2016, 10, 747-
753.
20. a) Pang, S. H.; Roman, A. M.; Medlin, J. W. J. Phy. Chem. C 2012, 16, 13654-13660, b)
Mitra, J.; Zhou, X.; Rauchfuss, T. Green Chem. 2015, 17, 307-313; c) Gürbüz, E. I.; Hibbitts, D.
D.; Iglesia, E. I. J. Am. Chem. Soc. 2015, 137, 11984-11995.
21. Louie, Y. L.; Tang, J.; Hell, A. M.; Bell, A. T. Appl. Catal. B: Environ. 2017, 202, 557-568.
22. Jenness, G. R.; Wan, W.; Chen, J. G.; Vlachos, D. G. ACS Catal. 2016, 6, 7002-7009.
23. Luo, J.; Yun, H.; Mironenko, A. V.; Goulas, K.; Lee, J. D.; Monai, M.; Wang, C.;
Vorotnikov, V.; Murray, C. B.; Vlachos, D. G.; Fornasiero, P.; Gorte, R. J. ACS Catal. 2016, 6,
4095-4104.
24. Gilkey, M. J.; Mironenko, A. V.; Yang, L.; Vlachos, D. G.; Xu, B. ChemSusChem 2016, 9,
3113-3121.
25. a) Nakagawa, Y.; Tamura, M.; Tomishige, K. ACS Catal. 2013, 3, 2655-2668. b) Nakagawa,
Y.; Liu, S.; Tamura, M.; Tomishige, K. ChemSusChem 2015, 8, 1114-1132.
22
ACS Paragon Plus Environment

Page 23 of 23
ACS Catalysis
1
2
3
4
5
6
7
8
9
26. Keskiväli, J.; Wrigstedt, P.; Lagerblom, K.; Repo, T. Appl. Catal. A: Gen. 2017, 534, 40-45.
27. Xia, Q.-N.; Cuan, Q.; Liu, X.-H.; Gong, X.-Q.; Lu, G.-Z.; Wang, Y.-Q. Angew. Chem.
Angew. Chem. Int. Ed. 2014, 53, 9755-9760.
28. Li, S.; Li, N.; Li, G.; Li, L.; Wang, A.; Cong, Y.; Wang, X.; Zhang, T. Green Chem. 2015,
17, 3644-3652.
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
TOC
23
ACS Paragon Plus Environment