Reactions of 2-Methyltetrahydropyran on Silica-Supported Nickel Phosphide in Comparison with 2-Methyltetrahydrofuran

Article abstract of DOI:10.1021/acscatal.6b01033

The reactions of 2-methyltetrahydropyran (2-MTHP, C₆H₁₂O) on Ni₂P/SiO₂ provide insights on the interactions between a cyclic ether, an abundant component of biomass feedstock, with a transition-metal phosphide, an effective hydrotreating catalyst. At atmospheric pressure and a low contact time, conditions similar to those of a fast pyrolysis process, 70% of products formed from the reaction of 2-MTHP on Ni₂P/SiO₂ were deoxygenated products, 2-hexene and 2-pentenes, indicating a good oxygen removal capacity. Deprotonation, hydrogenolysis, dehydration, and decarbonylation were the main reaction routes. The reaction sequence started with the adsorption of 2-MTHP, followed by ring-opening steps on either the methyl substituted side (Path I) or the unsubstituted side (Path II) to produce adsorbed alkoxide species. In Path I, a primary alkoxide was oxidized at the α-carbon to produce an aldehyde, which subsequently underwent decarbonylation to 2-pentenes. The primary alkoxide could also be protonated to give a primary alcohol which could desorb or form the final product 2-hexene. In Path II, a secondary alkoxide was oxidized to produce a ketone or was protonated to a secondary alcohol that was dehydrated to give 2-hexene. The active sites for the adsorption of 2-MTHP and O-intermediates were likely to be Ni sites.

Full text of DOI:10.1021/acscatal.6b01033

Subscriber access provided by Weizmann Institute of Science
Article
Reactions of 2-Methyltetrahydropyran on Silica Supported
Nickel Phosphide, in Comparison with 2-Methyltetrahydrofuran
Phuong P. Bui, S. Ted Oyama, Atsushi Takagaki, Brad P. Carrow, and Kyoko Nozaki
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.6b01033 • Publication Date (Web): 01 Jun 2016
Downloaded from http://pubs.acs.org on June 7, 2016
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 35
ACS Catalysis
1
2
3
4
5
6
7
8
9
Reactions of 2-Methyltetrahydropyran on Silica Supported Nickel Phosphide, in
Comparison with 2-Methyltetrahydrofuran
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
Phuong P. Bui1,2, S. Ted Oyama1,2*, Atsushi Takagaki1, Brad P. Carrow3,4, Kyoko Nozaki3
1. The University of Tokyo, Department of Chemical Systems Engineering,
7ꢀ3ꢀ1 Hongo, Tokyo, 13ꢀ8656, Japan.
2. Virginia Tech, Department of Chemical Engineering, Blacksburg, VA, USA
3. The University of Tokyo, Department Department of Chemistry and Biotechnology,
7ꢀ3ꢀ1 Hongo, Tokyo, 13ꢀ8656, Japan..
4. Princeton University, Department of Chemistry, Princeton, NJ, USA.
Corresponding author email: oyama@vt.edu
1
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 35
1
2
3
4
Abstract
5
6
7
8
9
The reactions of 2ꢀmethyltetrahydropyran (2ꢀMTHP, C6H12O) on Ni2P/SiO2 provide insights
on the interactions between a cyclic ether, an abundant component of biomass feedstock, with a
transition metal phosphide, an effective hydrotreating catalyst. At atmospheric pressure and a
low contact time, conditions similar to those of a fast pyrolysis process, 70% of products formed
from the reaction of 2ꢀMTHP on Ni2P/SiO2 were deoxygenated products, 2ꢀhexene and 2ꢀ
pentenes, indicating a good oxygen removal capacity. Deprotonation, hydrogenolysis,
dehydration, and decarbonylation were the main reaction routes. The reaction sequence started
with the adsorption of 2ꢀMTHP, followed by ringꢀopening steps on either the methyl substituted
side (Path I) or the unsubstituted side (Path II) to produce adsorbed alkoxide species. In Path I, a
primary alkoxide was oxidized at the αꢀcarbon to produce an aldehyde which subsequently
underwent decarbonylation to 2ꢀpentenes. The primary alkoxide could also be protonated to give
a primary alcohol which could desorb or form the final product 2ꢀhexene. In Path II, a secondary
alkoxide was oxidized to produce a ketone or was protonated to a secondary alcohol that was
dehydrated to give 2ꢀhexene. The active sites for the adsorption of 2ꢀMTHP and Oꢀintermediates
were likely to be Ni sites.
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
Keywords: Hydrodeoxygenation; 2ꢀmethyltetrahydropyran; nickel phosphide; contact time
studies; reaction network.
2
ACS Paragon Plus Environment

Page 3 of 35
ACS Catalysis
1
2
3
4
1. Introduction
5
6
7
8
9
Meeting the everꢀgrowing demand for transportation fuels and chemical feedstocks is of
significant worldwide importance. Among the various renewable energy types, like solar, wind,
or wave, only bioresources satisfy both needs, and continuous efforts have been expended in
developing biomass processing technology [1,2,3]. Data obtained from different sciences such as
analytical chemistry [4,5], fundamental catalysis [6,7,8], and mathematical modelling [9,10,11]
are used to improve the technology for biomass conversion. Our work contributes to
understanding the interactions between a catalyst and a compound, 2ꢀmethyltetrahydropyran
(C6H12O), whose structure is close to those of components in the biomass resources. Specifically,
many compounds derived from cellulose and hemicellulose have sixꢀmembered rings with
internal ether linkages like the pyran ring.
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
A central issue of biomass and bioꢀoil treatments is the removal of large amounts of
multifunctional oxygenate compounds as they decrease heating value and cause chemical and
physical instability in the final fuel products [12]. The complex composition requires thorough
research on how each component reacts on a catalyst, thus mechanistic studies on model
compounds of these constituents provide valuable insights to this end. This study describes the
reactions of 2ꢀmethyltetrahydropyran (2ꢀMTHP) on silicaꢀsupported nickel phosphide. The
reactant has a sixꢀmember ether ring with a methyl group attached to one alphaꢀcarbon.
Analogous compounds with fiveꢀmember ether rings such as furan and furan derivatives were
investigated on various catalysts including conventional hydrotreating catalysts [13ꢀ19], noble
metals [20ꢀ31], metals [32ꢀ34], and metal phosphides [35ꢀ39].
3
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 35
1
2
3
4
5
6
7
8
9
Metal phosphides are a promising class of materials that show excellent reactivity in
hydrodesulfurization, hydrodenitrogenation, and hydrodeoxygenation (HDO) [40ꢀ42]. The
phosphides have physical properties similar to those of both ceramics such as high strength and
hardness, and of metals such as good thermal and electrical conductivity. The catalysts provide
active sites that facilitate hydrogenation, hydrogenolysis as well as nucleophilic substitution.
Moreover, the relatively strong metal to phosphorus interaction enhances the phosphides’
resistance to Sꢀ or Nꢀ compounds. It has been shown that Ru2P/SiO2 is nearly three times more
active than Ru/SiO2 and significantly more active than the traditional hydrotreating catalyst
CoMo/Al2O3 in the HDO of furan [39]. The phosphide catalysts show preference towards
hydrogenolysis products, i.e. C4 hydrocarbons while the metal catalysts favor decarbonylation
products, i.e. C3 hydrocarbons. The HDO of 2ꢀmethyltetrahydrofuran on phosphide catalysts
shows higher reactivity than Ru/C and Pd/Al2O3 [35,36,37,38]. The proposed reaction network
involves the adsorption of 2ꢀMTHF on a metallic site, followed by a rateꢀdetermining opening of
the ether ring to produce the adsorbed intermediates 1ꢀpentoxide or 2ꢀpentoxide, depending on
which side the ring is opened. Similar reactions occur with 2ꢀmethyltetrahydropyran on
Ni2P/SiO2 (vide infra) but with distinctive differences, such as lower reactivity.
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. Experimental
2.1 Synthesis of 2-methyltetrahydropyran
Amounts (~250 mL) of 2ꢀmethyltetrahydropyran (2ꢀMTHP) were synthesized in three
steps. First, hydrogen chloride was added to the starting material, 3,4ꢀdihydroꢀ2Hꢀpyran (DHP)
4
ACS Paragon Plus Environment

Page 5 of 35
ACS Catalysis
1
2
3
4
5
6
7
8
9
to produce 2ꢀchlorotetrahydropyran (ClꢀTHP), second, the Grignard reagent was prepared, and
third, the Grignard reagent was reacted with 2ꢀchlorotetrahydropyran to produce 2ꢀ
methyltetrahydropyran as final product. Details are given in the Supplementary Information.
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 Materials and characterization
The Ni2P catalyst was synthesized via temperature programmed reduction of a phosphate
precursor. Details of the method were reported earlier [35, 43, 44]. Briefly, 2 mL of an aqueous
solution containing 337 mg of nickel (II) nitrate (Ni(NO3)2.6H2O – Alfa Aesar, 99.6%) and 306
mg of ammonium hydrogen phosphate ((NH4)2HPO4 – Aldrich) was impregnated onto 1 g of
silica (Cabꢀosil ® EH5). After drying at 120 oC for 4 h and calcining at 600 oC for 6 h, the
mixture was cooled and then pressed and pelletized to 16/20 mesh size (0.8ꢀ1.2 mm in diameter)
particles. The precursor particles were reduced to nickel phosphide in a H2 stream (1 L minꢀ1 of
H2 per gram of precursor) as the temperature was increased from 25 oC to 615 oC at a rate of 3 oC
minꢀ1. After reduction the sample was cooled to room temperature in a He stream and was
passivated in a 0.5 % O2/He stream for 2 h.
The surface area of the catalyst was estimated by N2 physisorption using the Brunauer
Emmett Teller (BET) method. The active phase of nickel phosphide was determined by Xꢀray
diffraction and the surface sites were quantified by carbon monoxide (CO) chemisorption.
The amount of acid sites on the catalyst was probed by temperature programmed
desorption of ammonia. First a sample of 100 mg of passivated Ni2P/SiO2 was activated in H2
(100 cm3 minꢀ1) at 500 oC (3 oC minꢀ1 heating rate) for 2 h. After the sample was cooled in He (50
cm3 minꢀ1) to 120 oC and held at this temperature for 0.5 h, a 10% NH3/He stream (50 cm3 minꢀ1)
5
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 35
1
2
was introduced for 0.5 h. The saturated sample was flushed with pure He (50 cm3 minꢀ1) for 0.5 h
and then the temperature was increased from 120 oC to 700 oC at a heating rate of 10 oC minꢀ1.
Masses 16 and 17 were monitored during the desorption by a mass spectrometer.
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
2.3 Reactivity studies
For the reactivity tests of 2ꢀmethyltetrahydropyran (2ꢀMTHP), an amount of Ni2P/SiO2
equivalent to 12.6 µmol of active sites was mixed with 1.5 g of quartz chips of the same size and
the mixture was packed in a tubular quartz reactor of 8 mm in diameter. The catalyst was reꢀ
activated in H2 with the temperature raised from 25 oC to 500 oC (5 oC minꢀ1) and held at this
level for 2 h, followed by cooling to 350 oC. A liquid pump was used to inject a solution of 96
mol% 2ꢀMTHP and 4 mol% benzene as internal standard at a rate of 0.02 g minꢀ1 (3.5 ꢁmol sꢀ1
total) into a vaporizer kept at 300 oC where mixing with a hydrogen stream occurred. Benzene
was easily distinguished in the gas chromatograms and it was verified that benzene was not
hydrogenated at the reaction conditions. The feed concentration was held at 2.6 mol% of
reactant in a H2 flow (180 cm3 minꢀ1, 123 µmol sꢀ1). The catalyst was stabilized for 12 h after the
introduction of 2ꢀMTHP then temperature was varied in the following order: 350 oC ꢀ 400 oC
ꢀ 450 oC ꢀ 500 oC ꢀ 475 oC ꢀ 425 oC ꢀ 375 oC ꢀ 350 oC, with the temperature held for 3h
at each temperature. The components in the exit stream were analyzed with a gas chromatograph
equipped with a flame ionization detector and an HPꢀ1 (0.25mm x 100m) column. In the
ꢁ
experiments, the contact time was kept constant at 3.9 s. Contact time is defined as ꢀ =
, in
ꢂ
ꢃꢄ
which τ is the contact time (s) based on active sites, FA0 is the molar flow rate of 2ꢀMTHP (3.2
µmol sꢀ1), L is the number of active sites (12.6 µmol) approximated by the CO uptake (45 µmol
6
ACS Paragon Plus Environment

Page 7 of 35
ACS Catalysis
1
2
3
4
5
ꢅ
gꢀ1 x 0.28 g). Note that the contact time based on catalyst bed volume ꢀꢅ = has a value of 0.5 s,
ꢆ
where V is the bed volume (1.5 cm3) and v is the volumetric flow rate (180 cm3 minꢀ1).
6
7
8
9
The contact time study for 2ꢀMTHP reactions was carried out in the same manner, except
at a fixed temperature of 450 oC and various contact times from 0.6 to 14.0 s. The contact time
was adjusted by changing the flow rate of the feed while keeping the feed concentration constant.
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
For the reactivity tests of 2ꢀmethyltetrahydrofuran (2ꢀMTHF) a catalyst amount
equivalent to 30 µmol of active sites mixed with 1.5 g of quartz chips were packed into a quartz
tubular reactor. The catalyst was pretreated in the same manner as with 2ꢀMTHP. A stream of
3.3 mol% 2ꢀMTHF in H2 (200 cm3 minꢀ1) made by flowing H2 through a bubbler carried 2ꢀ
MTHF to the reactor. The bubbler had 95 volume% of 2ꢀMTHF and 5 volume% of heptane as
internal standard and was kept at 0 oC. The nꢀheptane was easily distinguished in the gas
chromatograms and it was verified that it did not undergo cracking at the reaction conditions.
The reaction was stabilized for 12h then temperature was varied as follows: 350 oC ꢀ 300 oC ꢀ
250 oC ꢀ 275 oC ꢀ 325 oC ꢀ 350 oC ꢀ 375 oC, with each temperature was held for 4 h. The
contact time was kept at 6.8 s. The contact time of the 2ꢀMTHP was lower at 3.9 s. It would
have been better to have used the same contact time, but it’s lower reactivity was unexpected,
and there were supply limitations because of the need for a lengthy synthesis. In fact, we do not
recommend studies of 2ꢀMTHP because of this preparation limitation (it is not available
commercially) and the complexity of the product mixture.
Mathematical modelling was carried out to evaluate the proposed reaction network.
Assuming that the reactions were firstꢀorder, a set of ordinary differential equations (ODEs)
describing the consumption rate of 2ꢀMTHP (A) and the formation rates of products (Pi) was
7
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 35
1
2
3
4
5
established, following a general form of ꢇ[ꢈ] ꢊꢋ ꢇ[ꢌ ] = ꢎ(ꢏ , ꢑ , [ꢒ ]) where ki are the
ꢍ
[ ]
ꢐ
ꢐ
ꢇꢉ
ꢇꢉ
6
7
corresponding rate constants and [A] and [Pi] are the gaseous concentration of the reactant 2ꢀ
8
9
MTHP and the products in the effluent stream. The surface reactions involving adsorbed reactant
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
A* and surface species Pi* were also modelled in a similar manner, ꢇ[ꢈ∗] ꢊꢋ ꢇ[ꢌ ∗]
=
ꢍ
ꢇꢉ
ꢇꢉ
[ ] [ ] [ ] [
]
ꢓ(ꢏꢐ, ꢑ , ꢒ , ∗ , ꢑ ∗ , [ꢒ ∗]) where [A*] and [Pi*] are the surface concentration of the
ꢐ
ꢐ
adsorbates. The concentration of total active sites was set constant and equal to the sum of vacant
sites [*] and occupied sites [A*] and [Pi*]. Details of the modelling equations, initial conditions,
and units are given in the supplementary information. A programming language, Python 2.7, was
used to solve the ODEs simultaneously while adjusting the rate constants to minimize the
difference between the experimental data and the simulated data. The objective function for the
ꢙꢚ.ꢛ
minimization was the sum of timeꢀweighted residuals: ꢔꢔꢉꢕꢖꢗ
=
ꢐ ꢘꢐ
ꢎ − ꢜꢐ ꢝ where ti is the
( )
ꢐ
∑
contact time value, fi is the predicted data, and yi is the experimental data. The function was
defined so as to achieve better fitting at lower contact times. Two minimization methods were
applied: 1. the sequential least square programming method, SLSQP, and 2. the limitedꢀmemory
BroydenꢀFletcherꢀGoldfarbꢀShanno, LꢀBFGSꢀB, method.
The fit between the model and experimental data was reported using the coefficient of
ꢟꢟ
ꢠꢡꢢ
determination defined as ꢞꢝ =
where SSreg is the regression sum of squares and SSres
ꢠꢡꢢ
ꢟꢟ ꢤꢟꢟ
ꢠꢡꢣ
is the residual sum of squares. The regression sum of squares expresses the variance of the
ꢎ − ꢜꢦ ꢝ where fi is the predicted value and
∑ ( )
ꢐ ꢐ
model’s predictions to the observed data: ꢔꢔꢕꢖꢥ
=
ꢜꢦ is the mean of the observed data. The residual sum of squares shows the difference between
the predicted data and the experimental data: ꢔꢔꢕꢖꢗ
=
ꢎ − ꢜꢐ ꢝ in which fi is the predicted
∑ ( )
ꢐ ꢐ
8
ACS Paragon Plus Environment

Page 9 of 35
ACS Catalysis
1
2
value and yi is the experimental data. The R2 value has a range from 0 to 1. A good fit occurs
when the SSres value is minimized and so the R2 value is maximized (~1).
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
3
Results and Discussion
3.1 Synthesis of 2-methyltetrahydropyran
For the synthesis of 2ꢀmethyltetrahydropyran the first step consisted of the addition of
hydrogen chloride (HCl) to the starting material 3,4ꢀdihydroꢀ2Hꢀpyran (DHP) under dry
conditions to avoid water addition to the double bond. A yield of 90% of the 2ꢀClꢀTHP product
was obtained, the 1H NMR spectrum of which is shown in the Supplementary Information. The
2ꢀClꢀTHP product was converted to the 2ꢀMTHP by use of a methyl Grignard reagent.
9
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 35
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 1. 1H NMR spectra of 2-methyltetrahydropyran.
The proton NMR spectrum of the 2ꢀMTHP (Figure 1) shows strong proton peaks of the
methyl group at 1.1 ppm, followed by protons of the methylene groups of aliphatic carbons at
1.8~1.2 ppm, and finally the protons of the alpha carbons at 3.4 and 3.9 ppm. The spectrum
confirms the successful synthesis of 2ꢀMTHP.
.
3.2 Characterization and reactivity results
10
ACS Paragon Plus Environment

Page 11 of 35
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
Fresh
Used
PDF 00-003-0953
20 40
60
80
2
θ
Figure 2. X-ray diffraction pattern of Ni2P/SiO2 before (top) and after (bottom) reactions.
The synthesized Ni2P/SiO2 had a BET surface area of 138 m2 gꢀ1, a CO uptake of 45
µmol gꢀ1 and a NH3 uptake of 202 µmol gꢀ1. The phosphide phase showed signature XRD peaks
at 42, 44, 47, and 55 o, confirming its presence on the catalyst (Figure 2). After reaction, the
crystal structure of Ni2P/SiO2 was retained, indicating stability throughout the reactions (Figure
2). From the Scherrer equation the crystallite size of the reduced, passivated catalyst was 10 nm
and that of the used catalyst was 11 nm. In accordance with this, the surface areadecreased from
138 to 121 m2 gꢀ1 and the CO uptake from 45 to 27 µmol gꢀ1, indicating some sintering and
blockage of the support and some loss of active sites during the reactions. Previous studies on
guaiacol deoxygenation showed the nickel phosphide catalyst was slightly oxidized to nickel
phosphate over the silica support, one of the reasons leading to the loss of active sites [7].
11
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 35
1
2
3
4
5
6
7
8
9
At atmospheric pressure, the conversion of 2ꢀMTHP increased with increasing
temperature as expected (Figure 3). The sixꢀmembered cyclic oxygenate showed a low reactivity
even though the reaction temperatures were in a moderate to high range. For example, the total
conversion was less than 10 % at temperatures up to 400 oC and rose to the highest 41 % at the
maximum tested temperature of 500 oC. Using the same catalyst system, 2ꢀMTHF, a fiveꢀ
membered ring compound homologous with 2ꢀMTHP, showed a much higher reactivity with a
total conversion over 90 % at 325 oC.
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
Turnover frequency indicates the amount of reactant converted per surface metal atom
active site as estimated by CO chemisorption and gives a useful estimate of the intrinsic
ꢂ
ꢃꢄ
reactivity of the catalyst. The turnover frequency is calculated as ꢧꢨꢩ = ꢁ ꢫꢚꢪꢚ, where FA0 is the
molar flow rate of 2ꢀMTHP (3.2 µmol sꢀ1), X is total conversion (%), and L is the number of
active sites (12.6 µmol). The profiles of turnover frequency of 2ꢀMTHP and 2ꢀMTHF are shown
in the bottom graph of Figure 3. The observed trend reflects the conversion profile, which
showed a lower reactivity of 2ꢀMTHP compared to 2ꢀMTHF. It should be noted that the contact
time for 2ꢀMTHP (3.9 s) was shorter than for 2ꢀMTHF (6.8 s), so 2ꢀMTHP molecules had less
time to approach and interact with the active sites. This difference in contact time could give at
most a factor of two greater conversion of 2ꢀMTHF and does not account for the reactivity
difference (vide infra).
The Arrhenius plots are linear at low temperature, and the apparent activation energies
were 133 kJ molꢀ1 for 2ꢀMTHP and 116 kJ molꢀ1 for 2ꢀMTHF. The slightly higher value for 2ꢀ
MTHP can be attributed to the lower ring strain, and thus lower ground state energy, of 2ꢀMTHP
that possesses a sixꢀmembered ring compared to 2ꢀMTHF [45]. This results in a relatively larger
activation energy in the ring opening reaction of 2ꢀMTHP compared to 2ꢀMTHF. At high
12
ACS Paragon Plus Environment

Page 13 of 35
ACS Catalysis
1
2
3
4
5
6
7
8
9
temperature the turnover frequencies of both molecules level off due to mass transfer limitations
(Figure 3). To compare the reactivity at the same temperature, for example at 400 oC, the
turnover frequency of 2ꢀMTHF can be extrapolated and it is found to be about two orders of
magnitude larger than that of 2ꢀMTHP at this temperature. The difference can be attributed to the
ring size and the ringꢀstrain energy of the compounds. For instance, tetrahydropyran with a sixꢀ
membered ring is reported to have a strain energy of 2.1 kJ molꢀ1 while tetrahydrofuran with a
fiveꢀmembered ring has a strain energy of 24.7 kJ molꢀ1 [45]. Thus, for 2ꢀMTHP ring opening is
not as favorable as for 2ꢀMTHF, and higher temperatures are needed. The 2ꢀMTHF is a lessꢀ
hindered Oꢀligand, because the molecule’s planar structure has the neighboring Hꢀatoms pointing
away from the oxygen atom, while the 2ꢀMTHP in the chair conformation has two of the Hꢀ
atoms pointing in the same direction as the oxygen atom. The geometry causes tetrahydrofuran
to have a smaller dipole moment of 1.74 compared to tetrahydropyran which has a dipole
moment of 1.87 [46]. The nonplanarity of the 2ꢀMTHP may lead to weaker binding and easier
desorption from catalytic sites.
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 35
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 3. Reactivity of 2-MTHP and 2-MTHF on Ni2P/SiO2 in H2, 1 atm: a) conversion vs.
temperature and b) turnover frequency vs. inverse temperature.
A WeiszꢀPrater analysis was carried out for all reactivity experiments to determine the
influence of pore diffusion on the reaction rates. The detailed analysis is provided in the
Supplementary Information (SI). In Figure 3a, the lower conversion data satisfied the Weiszꢀ
Prater criterion (CWP), indicating that internal diffusion limitations were negligible, whereas, the
higher conversion data for 2ꢀMTHF showed values of CWP of ~ 1.4, indicating mass transfer
limitations [47,48]. The Arrhenius plots in Figure 3b duly showed a flatter region for the higher
temperature, higher conversion points on the left side of each plot, characteristic of diffusion
limitations. It should be noted that the contact time studies to be presented later were carried out
in the kinetic regime.
14
ACS Paragon Plus Environment

Page 15 of 35
ACS Catalysis
1
2
3
4
5
6
7
8
9
The reactions of 2ꢀMTHP on Ni2P/SiO2 produced mostly unsaturated products (Figure 4).
This result is understandable thermodynamically since dehydrogenation is generally favored at
high temperatures and the reactions of 2ꢀMTHP at 1 atm occurred only at high temperatures
(Figure 3). The products included C1ꢀC4 hydrocarbons (6 %), 2ꢀpentenes (10 %), 2ꢀhexene
(49 %), 2ꢀhexanone (19 %), and dehydrogenated pyran derivatives (11%). Some hexꢀ4ꢀenal and
hexꢀ4ꢀenꢀ1ꢀol (3%) were formed at temperatures lower than 450 oC but were not observed at
higher temperatures. Analogous products were observed from the reactions of 2ꢀMTHF such as
the cracking product, nꢀbutane, the deoxygenation product, nꢀpentane, and the ketone, 2ꢀ
pentanone. All of these products had saturated hydrocarbon chains, and in fact, very few
unsaturated products were produced with the reactant 2ꢀMTHF in contrast to 2ꢀMTHP. This is
partly because reactions of 2ꢀMTHF proceeded at much lower temperatures with conversion
more than 60% at 300 oC, where hydrogenation reactions dominated.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
As temperature was raised, 2ꢀMTHP was converted to increasing amounts of C5 products
like 2ꢀpentenes and decreasing quantities of deoxygenated products like 2ꢀhexenes, suggesting
that decarbonylation was favored with increasing temperatures. At the same time, the alcohol
species, mostly primary, diminished while 2ꢀhexanone remained significant. This indicates that
these two species were formed by independent pathways.
The product distribution for the reaction of 2ꢀMTHP indicates that ring opening can
occur on both sides of the oxygen atom to give rise to primaryꢀsubstituted (e.g. 1ꢀhexenal/ol) and
secondaryꢀsubstituted (e.g. 2ꢀhexanone) products. These different pathways will be denoted as
Path I for the primary products and Path II for the secondary products in subsequent discussion.
Overall at conditions similar to those of fast pyrolysis with low contact time and high
temperatures, the total deoxygenation of 2ꢀMTHP attained moderately high levels of 70%.
15
ACS Paragon Plus Environment

ACS Catalysis
Page 16 of 35
1
2
3
4
5
6
Higher deoxygenation levels require higher pressures, but not too high as 0.5 MPa already gives
>98% deoxygenated products for 2ꢀMTHF at 300 oC [37].
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 4. Product selectivities of reactions of 2-MTHP (top) and 2-MTHF (bottom) on
Ni2P/SiO2 at 0.1 MPa.
3.3 Simple reaction network
The reaction sequence of 2ꢀMTHP on Ni2P/SiO2 can be inferred from the results of
contact time measurements where the product selectivities reveal the sequence of product
formation. Initial products are formed first and thus show nonꢀzero selectivities at short contact
time that decrease with increasing contact time. Intermediate products are produced from the
initial species so they start out at zero then increase to a maximum before finally abating as they
form final products. Final products in the sequence also start at zero but rise steadily with
16
ACS Paragon Plus Environment

Page 17 of 35
ACS Catalysis
1
2
3
4
5
6
7
8
9
contact time. In Figure 5 the selectivities (left) and yields (right) of the products are presented
for the four main groups: a) initial oxygenated products produced from the ringꢀopening on the
methyl substituted side, in Path I: hexꢀ4ꢀenꢀ1ꢀol and hexꢀ4ꢀenal, b) intermediate oxygenated
products from the ringꢀopening on the unsubstituted side, in Path II: 2ꢀhexanone, c) final
deoxygenated products: 2ꢀhexenes and 2ꢀpentenes, and d) dehydrogenated products. The total
conversion (indicated by the symbol X) is also shown in panel d). The curves presented are
results from a simple kinetic analysis to be discussed later.
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
On the left side of Figure 5, the primary alcohol, hexꢀ4ꢀenꢀ1ꢀol, and the aldehyde, hexꢀ4ꢀ
enal, showed decreasing selectivities and yields with increasing contact time, exhibiting apparent
initial product behavior. However, the amounts formed are very small (see scale on left of
figure), and it’s difficult to make an exact determination. In contrast, the ketone, 2ꢀhexanone,
exhibited increasing selectivity with contact time, and with high selectivities and yields. The
deoxygenated products, 2ꢀhexenes and 2ꢀpentenes, also demonstrated the same trends. On the
right side of Figure 5, the yields of 2ꢀhexanone, 2ꢀhexenes, and 2ꢀpentenes rose slowly at first
when contact time was less than 4 s, then increased rapidly. This growth lag at the start suggests
that these products were formed later in the sequence. The secondary alcohol 2ꢀhexanol was not
detected in the gas phase probably because its formation amount was below the detection limit,
as a result of the compound’s affinity for the active sites that inhibited desorption. Meanwhile,
the formation of 2ꢀhexanone was preferred, with the desorption of 2ꢀhexanone being dominant.
The pyran derivatives included mostly dehydrogenated products and a small amount of
isomerized products of 2ꢀMTHP, which probably existed in equilibrium with 2ꢀMTHP.
17
ACS Paragon Plus Environment

ACS Catalysis
Page 18 of 35
1
2
3
4
5
6
7
8
Hex-4-enal
Hex-4-en-1-ol
6
4
2
0.6
0.4
0.2
O
HO
a)
b)
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
0
0.0
15
2
2
2
4
4
4
6
6
6
8
8
8
10
10
10
12
12
12
14
14
14
0
0
0
2
4
6
8
8
8
10
12
12
12
14
14
14
300
2-Hexanone
20
10
00
10
5
O
0
2
4
6
10
30
2-Hexenes
2-Pentenes
40
20
10
c)
d)
20
0
60
0
2
4
6
10
600
Conversion
Pyran Derivatives
40
20
40
20
0
O
0
0
2
4
6
8 10 12 14
0
2
4
6
8 10 12 14
Contact Time / s
Figure 5. Contact time study of 2-MTHP reactions on Ni2P/SiO2 in H2 at 0.1 MPa, 450 oC.
Points are experimental data and lines are calculated using the rate constants in Scheme 1.
a) Product selectivity and yield for hex-4-enal and hex-4-en-1-ol, b) Product selectivity and
yieldfor 2-hexanone, c) Product selectivity and yieldfor 2-hexenes and 2-pentenes, d)
Product selectivity and yield for pyran derivatives and conversion of 2-MTHP
Based on the contact time results, a preliminary reaction sequence is proposed as follows
(Scheme 1). The first step is the adsorption of 2ꢀMTHP on an active site followed by ringꢀ
opening to give either primary or secondary alkoxide species depending on the cleavage position
of the intraꢀring CꢀO bond. The adsorbed primary alkoxide species in Path I, formed from ringꢀ
opening on the methylꢀsubstituted side, undergoes the following reactions: I a) protonation to
18
ACS Paragon Plus Environment

Page 19 of 35
ACS Catalysis
1
2
3
4
5
6
7
8
9
yield a primary alcohol, I b) hydrogen elimination to give an aldehyde which can be
decarbonylated in a later step to 2ꢀpentene, and I c) deoxygenation to 2ꢀhexene. In path II, the
adsorbed secondary alkoxide produced from cleavage on the unhindered side can II a) eliminate
hydrogen to give 2ꢀhexanone or can be II b) dehydrated to 2ꢀhexene.
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
Hex-4-en-1-ol
Hex-4-enal
k1d
2-Pentene
117
HO
k-1a
O
k1b
I a)
4.8
45.3
I b)
4.0
k1a
3.3
k-1b
k1
0.31
Path I
k1c
25
O
2-MTHP
2-Hexene
kA
*
I c)
0.0039
O+
II b)
23
k2b
0.21
k-A
O
Path II
0.35
k2
*
k-PYD
46
kPYD
4.5
O
*
k2a
93
II a)
2-Hexanone
O
C6O - Pyran Derivatives
O
Scheme 1. Proposed simple reaction network of 2-MTHP on Ni2P/SiO2, 1atm, 450 oC. Gas
phase species are shown in color, adsorbed species are in black. The rate constant values
are indicated. Units are s-1 for the surface reaction and desorption rate constants and cm3
µmol-1 s-1 for the adsorption rate constants.
Fitting results for the simple reaction network are shown in Figure 5 and the rate
constants in Table 1. The network takes into consideration the main observed gas phase species
with the most minimal possible consideration of surface species. The R2 value for the fitting of
19
ACS Paragon Plus Environment

ACS Catalysis
Page 20 of 35
1
2
3
4
5
6
7
8
9
the yields was 0.78 and for the selectivities was 0.75. The results are consistent with the
observation that the simulated yields of the main products, 2ꢀhexanone, 2ꢀhexenes and 2ꢀ
pentenes match more closely to the experimental data (Figure 5, two middle plots on the right
side) compared to the other calculated yields and selectivities. It is realized that this simple
scheme is inadequate in describing the reaction of 2ꢀMTHP on Ni2P/SiO2 and that a more
detailed model is required.
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
Table 1: Rate constants for the simple reaction network (Scheme 1)
Reactant
kA kꢀA
Path I
Path II
k2a k2b
k1
k1a kꢀ1a k1b kꢀ1b k1c k1d
k2
kPYD
4.5
kꢀPYD
46
0.0039 0.21 0.31 4.8 4.0 3.3
45
25 117 0.35 93 23
cm3
µmolꢀ1
sꢀ1
sꢀ1
sꢀ1
sꢀ1
sꢀ1
sꢀ1
sꢀ1
sꢀ1 sꢀ1 sꢀ1 sꢀ1 sꢀ1
cm3
sꢀ1
µmolꢀ1
sꢀ1
3.4 Detailed reaction network
A comprehensive sequence is proposed in Scheme 2. The adsorbed alkoxide species
progresses to form other intermediate surface species prior to desorption into the observed
products. In Path I the primary alkoxide species transforms in two ways. It can convert to an
adsorbed aldehyde or to an adsorbed primary alcohol. The aldehyde can desorb or go through
decarbonylation into 2ꢀpentenes and the alcohol may desorb or form 2ꢀhexenes. In path II the
secondary alkoxide species reacts to produce 2ꢀhexanone or undergoes protonation then
dehydration to generate 2ꢀhexenes.
20
ACS Paragon Plus Environment

Page 21 of 35
ACS Catalysis
1
2
3
4
Hex-4-enal
5
6
7
8
9
k-ALD
Hex-4-en-1-ol
O
15
0.2
kALD
kPEN
6
O
H
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
O
*
2-MTHP
2-Pentene
k1b
k-1b
k-OL1
kOL1
22 29
8.6 0.3
k1EN
k1a
O
k1
kA
k-A
22
6.1
2.8
Path I
15 0.012
+
2-Hexene
18
O
H
O
k-1a
*
*
+
k2
O
k2EN
k2b
3.2
Path II
35
*
11
+
kPYD
k-PYD
O H
O
5.4
3.5
0.014
k-2b
*
*
k-ONE
30
O
C6O - Pyran Derivatives
O
2-Hexanone
Scheme 2. Proposed detailed reaction network for 2-MTHP reactions on Ni2P/SiO2 at 1atm,
450 oC. Gas phase species are shown in color, adsorbed species are in black. Units for the
rate constants are s-1 for the surface reaction and desorption rate constants and cm3 µmol-1
s-1 for the adsorption rate constants.
The fitting results for the detail scheme (Scheme 2) are plotted in Figure 6. The R2 value
for the fitting of the yields is 0.96 and for the selectivities is 0.90. The continuous lines showing
the simulation results fit the experimental points well, indicating that the proposed network is
consistent with the data.
21
ACS Paragon Plus Environment

ACS Catalysis
Page 22 of 35
1
2
3
4
5
6
7
8
Hex-4-enal
Hex-4-en-1-ol
6
4
2
0.6
0.4
0.2
a)
O
HO
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
0
0.0
15
2
2
2
4
4
4
6
6
6
8
8
8
10
10
10
12
12
12
14
14
14
0
0
0
2
4
6
8
8
8
10
12
12
12
14
14
14
300
2-Hexanone
20
10
00
10
5
b)
c)
d)
O
0
2
4
6
10
30
2-Hexenes
2-Pentenes
40
20
10
20
0
60
0
600
2
4
6
10
Conversion
Pyran Derivatives
40
20
0
40
20
0
O
0
2
4
6
8 10 12 14
0
2
4
6
8 10 12 14
Contact Time / s
Figure 6. Fitting results for the detailed reaction network of 2-MTHP on Ni2P/SiO2 in H2 at
1 atm, 450 oC. Points are experimental data and lines are simulated data using the
estimated rate constants in Scheme 2. a) Product selectivity and yield for hex-4-enal and
hex-4-en-1-ol, b) Product selectivity and yieldfor 2-hexanone, c) Product selectivity and
yieldfor 2-hexenes and 2-pentenes, d) Product selectivity and yield for pyran derivatives
and conversion of 2-MTHP.
The estimated values for the rate constants in Scheme 2 indicate a slow adsorption of the 2ꢀ
MTHP on a vacant active sites (Table 2). The ringꢀopening steps display the smallest values for
rate constants compared to all subsequent surface reaction steps, suggesting that these are rate
22
ACS Paragon Plus Environment

Page 23 of 35
ACS Catalysis
1
2
3
4
5
6
7
8
9
determining steps. The rate constant for the ringꢀopening step at the less hindered side of 2ꢀ
MTHP in Path II is slightly larger than the rate constant for reaction at the more hindered side of
2ꢀMTHP in Path I. The trends of the values of rate constants observed here with the sixꢀ
memberedꢀring 2ꢀMTHP molecule show similarity to those of the fiveꢀmembered ring of 2ꢀ
MTHF. The studies of both molecules show that the ringꢀopening steps are rate limiting and that
the ring opening is faster on the less hindered side of 2ꢀMTHP in path II than on the hindered
side.
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
Table 2: Estimated rate constants for detailed reaction network (Scheme 2)
Reactant
Surface reactions Path I
Path II
k2b
k1
2.8
sꢀ1
k1a
22
sꢀ1
kꢀ1a
18
k1b
22
sꢀ1
kꢀ1b
29
k2
3.2
sꢀ1
kꢀ2b
5.4
sꢀ1
35
sꢀ1
sꢀ1
sꢀ1
Adsorption / desorption Path I
Path II
kA
kꢀA
kALD
0.2
kꢀALD
15
kOL1
0.3
kꢀOL1 k1EN kPEN
kꢀONE k2EN
kPYD kꢀPYD
0.012 15
8.6
6.1
6.0
30
11
0.014
3.5
cm3
µmolꢀ1
sꢀ1
sꢀ1
cm3
sꢀ1
cm3
sꢀ1
sꢀ1
sꢀ1
sꢀ1
sꢀ1
cm3
sꢀ1
µmolꢀ1
sꢀ1
µmolꢀ1
sꢀ1
µmolꢀ1
sꢀ1
Concentrations of the surface species are plotted in Figure 7. The results show a low
adsorption of 2ꢀMTHP on the catalyst, probably due to high temperature and low contact time.
The adsorbed reactant is the dominant species, followed by the adsorbed species from Path I, and
finally the adsorbed species from Path II. This order reflects the trend observed in Scheme 2
where Path I includes surface reactions and desorption steps with lower rate constants compared
to those of Path II, so the intermediate species from Path I have longer residence on the active
sites.
23
ACS Paragon Plus Environment

ACS Catalysis
Page 24 of 35
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 7. Concentrations of surface species calculated from the estimated rate constants
from Scheme 2.
The nickel phosphide catalyst offers multiple active centers for hydrogen transfer and
deoxygenation steps as suggested by the following discussion of the crystal structure of the
catalyst [41, 49]. The metalꢀrich Ni2P particles supported on silica have a crystal structure
consisting of two types of Ni atoms, referred to as Ni(1) and Ni(2), as well as two types of P
atoms, P(1) and P(2) [50ꢀ52]. The surface Ni atoms are positively charged by 0.07 e, and the P
atoms are negatively charged by ꢀ0.07 e, so the Ni atoms provide binding sites for electronꢀrich
adsorbates such as Oꢀcompounds. The P atoms may be associated with unreduced passivation
oxygen to form acidic PꢀOH centers, although a recent diffuse reflectance infrared spectroscopy
study suggests that at reducing conditions only the phosphate on the silica is oxidized [53]. The
Ni(1) site has a near tetrahedral structure and favors hydrogenolysis reactions while the Ni(2) site
24
ACS Paragon Plus Environment

Page 25 of 35
ACS Catalysis
1
2
3
4
5
6
7
8
9
has a square pyramidal structure and prefers hydrogenation reactions [54]. The bulk structure of
Ni2P has alternating layers of Ni3P2 and Ni3P stacked along the (0001) direction [41,55,56]. The
Ni3P2 layer contains Ni(1) and P(2) atoms and is reported to be more stable than the Ni3P layer,
which has Ni(2) and P(1) atoms. The surface (0001) should be dominated with Ni3P2 structure.
Yet, the threefold hollow sites in a Ni(1) vacancy on a Ni3P2ꢀterminated surface lack a P
exposing dangling bonds that require stabilization, resulting in a surface with about 80% of Pꢀ
covered Ni3P2 structure (or Ni3PꢀP structure) and 20% of uncovered Ni3P2 structure [57,58]. A
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
ꢦ
similar conclusion was reached for the (1010) plane in which all Ni atoms are Pꢀterminated [59].
So, much of the active surface metallic Ni sites are covered with P atoms, leading to a much
lower apparent number of active sites, quantified by CO chemisorption, than the theoretical
value from the total surface area of Ni2P crystallites.
The Ni(1) sites are active for adsorption of Sꢀcompounds and Oꢀcompounds [41,60] and
are likely to be the adsorption sites for 2ꢀMTHP and the intermediate species. Both Ni(1) and
Ni(2) sites could be responsible for hydrogen dissociative adsorption, assisting the ring opening
reactions [40,61]. Thus, both sites are involved in the reaction.
The role of surface P atoms is both beneficial and detrimental. The P atoms protect the
stability of the nickel phosphide phase and break up Ni ensemble that can produce coke; yet, an
excess of P atoms reduces the availability of active Ni sites. Nevertheless, theoretical studies
indicate that the P sites are capable of adsorbing atomic hydrogen [62, 63] which can serve as a
hydrogen source for various reactions. The hydrogen may spill over from nickel centers. Also,
in oxidized form the P atom can serve as hydrogen acceptor. Such oxidized species have been
observed by FTꢀIR spectroscopy [64]. The CO uptake was 45 µmol gꢀ1 and the NH3 uptake was
202 µmol gꢀ1, indicating that there are more acid sites than metal sites. The NH3 adsorption
25
ACS Paragon Plus Environment

ACS Catalysis
Page 26 of 35
1
2
3
4
5
6
7
8
9
method does not distinguish between Brønsted and Lewis sites, so no definite conclusions can be
made on their role. It is speculated that the Brønsted acid sites are likely associated with oxidized
phosphorus both on the phosphide and any excess on the support and are involved with the
various proton transfer processes in the reactions of the adsorbed 2ꢀMTHF and 2ꢀMTHP. It is
also surmised that the Lewis sites are associated with reduced Ni sites which are involved in
interactions with the oxygen atoms in the adsorbates.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The detailed chemistry of the subsequent steps is presented in Scheme 3. For ring
opening on the substituted side (Scheme 3, Path I), the reaction must involve atoms more distant
from the hindered side, which would have only weak interactions with the active sites due to the
steric blockage. So it is likely that a basic site abstracts a proton from a βꢀcarbon in an E2
elimination, creating a hexꢀ4ꢀenꢀ1ꢀalkoxide species. This elimination requires the hydrogen to be
antiperiplanar with the leaving group so the 2ꢀMTHP molecule likely reacts in a chair
conformation. An αꢀhydrogen elimination from the alkoxide produces an aldehyde species. The
adsorbed aldehyde has a moderate interaction with the active site so it can desorb or go through a
decarbonylation step to produce 2ꢀpentene. The primary alkoxide could also be protonated to
form an adsorbed primary alcohol. The surface adsorbed water and/or PꢀOH groups are possible
proton donors. The adsorbed alcohol could simply desorb or undergo hydrogenolysis to 2ꢀ
hexene. The process requires an attack by a hydride, possibly a metal hydride NiꢀHꢀ, on the αꢀ
carbon and a shift of the electron pair to a nearby proton, creating 2ꢀhexene and an adsorbed
hydroxyl group.
The ring opening on the unsubstituted side (Scheme 3, Path II), has less steric hindrance
than Path I. The transformation is likely initiated by a hydride attack on the αꢀcarbon, which
cleaves the CꢀO bond in an SN2 manner, yielding an adsorbed secondary alkoxide. The 2ꢀ
26
ACS Paragon Plus Environment

Page 27 of 35
ACS Catalysis
1
2
3
4
5
6
7
8
9
hexanone is produced by abstraction of a proton by a basic site from the substituted αꢀcarbon.
The secondary alkoxide can also undergo protonation and subsequent dehydration to form 2ꢀ
hexene. It is interesting that no secondary alcohol was detected. This is possibly because the
adsorbed secondary alcohol precursor is very reactive and readily undergoes dehydration to 2ꢀ
hexene.
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
Hex-4-enal
O
2-Pentene
O
*
Hex-4-en-1-ol
H
H
O
2-MTHP
*
O
H
*
H
O
+ H+
H
+
+
H O
*
O
H
O
+
2-Hexene
O
Path I
*
*
*
*
+
H
H
O
+ H+
*
Path II
+
*
+
+
OH
O
O H
-
O
H
*
*
*
*
*
O
C6O - Pyran Derivatives
O
H
O
*
2-Hexanone
*
27
ACS Paragon Plus Environment

ACS Catalysis
Page 28 of 35
1
2
3
4
5
6
7
8
9
Scheme 3. Proposed detailed chemistry of the surface steps in the reaction network. Gas
phase species are shown in color, adsorbed species are in black. Symbol * represents a
surface site. The species in brackets depict suggested possible reaction steps.
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
In summary this second reaction network gives a good fit to the contact time data. It is
acknowledged that the number of fitting constants is larger than the simpler scheme. However,
what this indicates is that this type of analysis requires a realistic description of the elementary
steps involved in a transformation. It should be recognized that confirmation of the model results
is needed by independent measurements, for example, by spectroscopic observation of the
adsorbed intermediates.
4. Conclusions
The sixꢀmembered ring compound 2ꢀmethyltetrahydropyran with less ring strain was
much less reactive on Ni2P/SiO2 than the analogous fiveꢀmembered ring compound 2ꢀ
methyltetrahydrofuran. At conditions similar to those of fast pyrolysis, i.e. low contact time and
high temperature, nickel phosphide showed good deoxygenation capability (~70%) and produced
2ꢀhexenes and 2ꢀpentenes at all conversion levels. The reaction routes for 2ꢀMTHP were similar
to those for 2ꢀMTHF and involved two main pathways, depending on the CꢀO bond that cleaved
in the first step. Cleavage on the methylꢀsubstituted side produced linear aldehyde and primary
alcohol compounds while cleavage on the unsubstituted side formed a ketone product. Both
routes generated 2ꢀhexene as final products while Path I also produced 2ꢀpentene. Metallic Ni
sites are possibly the main active sites for the adsorption of reactants and the dissociative
28
ACS Paragon Plus Environment

Page 29 of 35
ACS Catalysis
1
2
3
4
5
6
7
8
9
adsorption of hydrogen while PꢀOH sites provide protons. The reaction network was supported
by fitting with a kinetic model. The obtained values of the rate constants showed that the slowest
step was ringꢀopening of adsorbed 2ꢀMTHP, consistent with previous results for the homologue
2ꢀMTHF.
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
Acknowledgments
This work was supported by the US Department of Energy under grant DEꢀFG02ꢀ96ER 14669,
and the Japan Society for the Promotion of Science for a GrantꢀinꢀAid for Scientific Research
(B) (no. 26289301).
Supporting Materials (SI)
The SI section describes 1) The detailed preparation of the 2ꢀmethyltetrahydropyran compound,
2) Calculations of the WeiszꢀPrater criterion for lack of diffusion limitations, 3) The equations
used in the modeling of the reaction network. This material is available free of charge via the
Internet at http://pubs.acs.org.
29
ACS Paragon Plus Environment