Production of 1,6-hexanediol from tetrahydropyran-2-methanol by dehydration-hydration and hydrogenation

Article abstract of DOI:10.1039/c6gc03606f

In this work we present an alternate method for the conversion of tetrahydropyran-2-methanol (THP2M), a cellulose-derived renewable building block, to 1,6-hexanediol (1,6-HDO). Our method is composed of three consecutive steps that either use relatively inexpensive catalysts or no catalyst at all. First, THP2M is catalytically dehydrated to 2,3,4,5-tetrahydrooxepine (THO) in up to 40% yield. THO is then hydrated to 2-oxepanol (OXL) and 6-hydroxyhexanal (6HDHX) with a combined yield of 85% in the absence of a catalyst. OXL and 6HDHX are then quantitatively hydrogenated to 1,6-HDO over a commercially available Ni/C or Ru/C catalyst. Various silicoaluminates were screened for the first acid-catalyzed reaction, and it was found that K-BEA shows the highest THO yield (40% over fresh catalyst, 20% after 25 h on stream). An overall 1,6-HDO yield of 34% from THP2M was obtained.

Full text of DOI:10.1039/c6gc03606f

View Article Online
View Journal
Green
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: S. P. Burt, K. J.
Barnett, D. J. McClelland, P. Wolf, J. A. Dumesic, G. W. Huber and I. Hermans, Green Chem., 2017, DOI:
10.1039/C6GC03606F.
This is an Accepted Manuscript, which has been through the
Volume 18 Number
7 7 April 2016 Pages 1821–2242
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Accepted Manuscripts are published online shortly after
Cutting-edge research for a greener sustainable future
www.rsc.org/greenchem
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1463-9262
CRITICAL REVIEW
G. Chatel et al.
Heterogeneous catalytic oxidation for lignin valorization into valuable
chemicals: what results? What limitations? What trends?
rsc.li/green-chem

Please do not adjust margins
Green Chemistry
Page 1 of 10
View Article Online
DOI: 10.1039/C6GC03606F
Green Chemistry
ARTICLE
Production of 1,6 Hexanediol from Tetrahydropyran-2-methanol
by Dehydration-Hydration and Hydrogenation
Samuel P. Burt,a,b Kevin J. Barnett,b Daniel J. McClelland,b Patrick Wolf,c James A. Dumesic,b
George W. Huber*,b and Ive Hermans*,a,b
Received 00th January 20xx,
Accepted 00th January 20xx
DOI: 10.1039/x0xx00000x
In this work we present an alternate method for the conversion of tetrahydropyran-2-methanol (THP2M), a cellulose-
derived renewable building block, to 1,6-hexanediol (1,6-HDO). Our method is composed of three consecutive steps that
either use relatively inexpensive catalysts or no catalyst at all. First, THP2M is catalytically dehydrated to 2,3,4,5-
tetrahydrooxepine (THO) in up to 40% yield. THO is then hydrated to 2-oxepanol (OXL) and 6-hydroxyhexanal (6HDHX)
with a combined yield of 85% in the absence of a catalyst. OXL and 6HDHX are then quantitatively hydrogenated to 1,6-
HDO over a commercially available Ni/C or Ru/C catalyst. Various silicoaluminates were screened for the first acid-
catalyzed reaction, and it was found that K-BEA shows the highest THO yield (40% over fresh catalyst, 20 % after 25h on
stream). An overall 1,6-HDO yield of 34% from THP2M was obtained.
www.rsc.org/
(THP2M), a biomass-derived feedstock,17 can be converted into 1,6-
HDO with high selectivity. However, the high cost of the catalyst, in
Introduction
combination with
a low productivity, limits the industrial
applicability of those systems.18
Terminal diols such as 1,4-butanediol (1,4-BDO) and 1,6-
hexanediol (1,6-HDO) are important building-block chemicals and
find application in the synthesis of specialty chemicals and a variety
of polymers, primarily in polyesters and polyurethanes, and
polyamides such as nylon-6,6.1,2 Their synthesis from petro-
chemical resources involves multiple reaction steps, and produces
significant amounts of by-products. Therefore, sustainable catalytic
systems that could use a renewable feedstock such as biomass are
desired. In particular for short chain diols such as ethylene glycol
and 1,4-BDO,3,4 renewable alternatives have recently been
proposed in the literature. Currently, 1,6-HDO is produced
industrially from cylohexanone/cyclohexanol (KA oil) by oxidation
with nitric acid to form adipic acid, followed by hydrogenation of
the dimethoxy ester to yield 1,6-HDO. The attractiveness of the
overall process is tempered by low conversions (4-8% for the
oxidation of KA oil), difficult separations, the use of non-renewable
fossil feedstocks, and emission of greenhouse gasses, namely N2O.5
Most recently an alternative route to 1,6-HDO via hydroformylation
of 1,3-butadiene was proposed.6 Although showing high yields, the
use of homogeneous Rh complexes as catalyst and the multistep
process hampers its industrial applicability.6 Various attempts to
synthesize 1,6-HDO, as well as other α,ω-diols, from biomass
resources have been reported. Using supported catalysts that
contain precious metals with a reducible metal oxide such as Pt-
ReOx,7 Rh-ReOx,8-14 and Ir-ReOx,15,16 tetrahydropyran-2-methanol
The dehydration of THP2M to 2,3,4,5-tetrahydrooxepine (THO)
over amorphous silicoaluminates was observed in the 1960s.19,20
THO can then be re-hydrated to 2-oxepanol (OXL) and 6-
hydroxyhexanal (6HDHX), both of which can be subsequently
hydrogenated to 1,6-HDO giving an overall yield as high as 29%.20
Slightly higher 1,6-HDO yields were achieved using a copper
chromite catalyst for the initial dehydration, but this introduces a
toxic metal that could make this system environmentally harmful.20-
22
No time-on-stream data were reported to assess the stability of
these catalysts. Moreover, only amorphous catalysts were tested
for the dehydration reaction. This opens the door for a wide range
of zeolites to be tested, which have been found to outperform their
amorphous counterparts in many reactions.23,24 The shape
selectivity of zeolites is also ideal for biomass conversion, because
the choice of framework can be used to encourage production of
oxygenated products, aromatics, or larger polymers.25
In this study, we present a three-step approach for the
synthesis of 1,6-HDO from THP2M via THO (see Scheme 1).
Scheme 1. Three-step route from tetrahydropyran-2-methanol
(THP2M) to 1,6-hexanediol (1,6-HDO).
In a first step, THP2M is dehydrated to THO over BEA zeolites,
followed by hydration to OXL, which is in equilibrium with 6HDHX.
Finally, 6HDHX and OXL are hydrogenated to 1,6-HDO over Ru/C or
Ni/C. In this work, we present the highest reported overall yields to
1,6-HDO from THP2M using a silicoaluminate as a dehydration
catalyst, and without the use of expensive metals such as Pt, Ir, and
Rh. We show effects of the framework topology and exchange
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 2 of 10
View Article Online
DOI: 10.1039/C6GC03606F
ARTICLE
Journal Name
cation on THO yields and present a catalytic system for a renewable The curves in Figure S2 were fitted to identify individual acid sites
route to 1,6-HDO.
responsible for NH3 desorption at specific temperatures, which is
displayed in Table 1. From this point on, all acid sites that desorb
o
NH3 in the range of 150-230 C will be denoted as weak acid sites.
Those that desorb NH3 in the range of 230-350 oC will be denoted as
moderate-strength acid sites, and we will not show any strong acid
sites.26 According to Table 1, Na-BEA, the framework with the
highest yields to THO, has a weak and a moderate acid site. Na-
ZSM-5 and Na-MOR also have a weak and moderate acid site (in
different concentrations). The latter two catalysts produce much
less THO than Na-BEA, so it is not immediately obvious which is the
active site for THO production. Furthermore, Na-Y has no moderate
acid site, and yet shows higher THO yields than Na-ZSM-5 and Na-
MOR at nearly every time-on-stream. This would seem to indicate
that the weak acid site is active for THO production, and the
moderate site is not.
Results
THP2M Dehydration
Initial experiments on the catalytic conversion of THP2M into
THO were performed over Na-BEA (SiO2/Al2O3 = 25) and compared
with an amorphous SiO2-Al2O3 (SiO2:Al2O3 = 13) benchmark catalyst
similar to that reported earlier20 for this reaction (see Figure 1). The
SiO2-Al2O3 catalyst used in
a previous study is no longer
commercially available.20 The SiO2-Al2O3 used in this study
produced low THO yields over a temperature range from 330 to 430
oC (see Figure S1). Amorphous SiO2 did not show any activity
towards THO, and over Al2O3 primarily cyclopentanecarbaldehyde
(CPC) was produced. Na-BEA was initially selected because sodium
was previously reported to enhance the THO yield in the case of
amorphous SiO2-Al2O3.20 Ion-exchange of commercial H-BEA was
conducted with a 1M aqueous NaNO3 solution resulting in a Na/Al
molar ratio of 0.66 (exchange ratios of all zeolites synthesized in
this manner can be found in Table S1). Na-BEA shows higher THO
yields compared to amorphous SiO2-Al2O3, suggesting that the pore
structure and/or Na+ incorporation influence the catalytic
performance.
☐
Figure 1. THP2M dehydration over Na-BEA ( , SiO2:Al2O3 = 25) and
amorphous SiO2-Al2O3 ( , SiO2:Al2O3 = 12). Reaction conditions: 400

oC, 0.628 mL/hr THP2M (liquid flow rate, STP), 30 mL/min (STP) H2,
1 atm, 150 mg catalyst. Error bars shown are the result of four
identical tests with Na-BEA. Error bars are assumed to be the same
for each catalyst in this study.
Figure 2. THO yields (a) and THP2M conversion (b) over Na-
exchanged zeolites. Reaction conditions: 400 oC, 0.628 mL/hr
THP2M (liquid flow rate, STP), 30 mL/min (STP) H2, 1 atm, 150 mg
catalyst.
Na-exchanged zeolites with SiO2:Al2O3 ratios of 20-30 were
screened for THP2M dehydration to investigate the influence of the
zeolite topology on the catalytic activity. Out of the four tested
framework types—MFI, MOR, FAU, BEA—the latter shows the
highest yields towards THO (see Figure 2). As this is an acid-
catalyzed dehydration, Temperature Programmed Desorption of
NH3 (NH3-TPD) was performed (see Table 1, Figure S2) to rationalize
the differences in activity between the various zeolite frameworks.
Temperature programmed oxidation (TPO) experiments in a
ThermoGravimetric Analyzer (TGA) were done to investigate the
o
cause of deactivation. Spent catalysts were heated to 600 C (ramp
2 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 3 of 10
View Article Online
DOI: 10.1039/C6GC03606F
Journal Name
ARTICLE
rate 5 oC/min) under a flow of air. Each spent catalyst shows a mass
BEA samples exchanged with each alkali metal were prepared
loss in the range of 16 to 37% during heating in air, due to coking to determine the effect of the exchange cation on THO yields.
over the course of the reaction (see Table S2, Figure S4). These Moreover, the THO yields of these catalysts were compared with
experiments show that the less active zeolites (Na-ZSM-5 and Na- those of H-BEA, i.e. the starting material in the synthesis of all BEA
MOR) produce roughly half the amount of coke than do more active zeolites in this study. The catalysts display clear differences in THO
zeolites (Na-Y and Na-BEA).
yields (see Figure 3), however, no clear periodic trend among the
catalysts with regards to THO production is immediately observed.
Table 1. NH3 desorption signals fitted for strong, moderate, and K- and H-BEA show the highest THO yields amongst the tested
weak acid Sites on Na-exchanged zeolites. Ramp Rate = 10 oC/min
catalysts with initial yields of up to 40% for K-BEA. Furthermore, as
shown in Figure 3b, Rb- and Cs-BEA show the lowest conversions at
all times-on-stream. This will be addressed further later.
Catalyst
NH3
Assignment
Quantity
Desorbed
(µmol NH3/g
catalyst)
1080
Desorption
Temp. (oC)
We regenerated the spent K- and H-BEA catalysts under flowing
air at 580 oC and then retested their catalytic activity as shown in
Figure 4. K-BEA has the higher initial and final yields compared to H-
BEA, with the highest single-point yield of 40% obtained after 7
hours on stream in the first cycle. It appears that for both catalysts
the initial activity cannot be fully recovered even after
regeneration. However, after the second regeneration the catalyst
does not continue to lose activity.
Na-MOR
186
257
Weak Acid
Moderate Acid
Weak Acid
264
Na-ZSM5 191
290
606
Moderate Acid
Weak Acid
364
Na-Y
177
226
174
238
367
Moderate Acid
Weak Acid
7.19
Na-BEA
549
Moderate Acid
214
Figure 4. Regeneration of K-BEA ( ) and H-BEA ( ). Reaction
☐ ✕
conditions: 400 oC, 0.628 mL/hr THP2M (liquid flow rate, STP), 30
mL/min (STP) H2, 1 atm, 150 mg catalyst.
NH3-TPD was performed on the alkali-exchanged BEAs (see
Table 2, Figure S3). NH3 desorption below 150 oC is likely due to
physisorbed NH3, as seen previously.27,28 It is clear that Rb- and Cs-
BEA are less acidic than the other catalysts. This correlates with low
conversions obtained with these catalysts. Next, Li- and Na-BEA
have similar TPD profiles to each other, with large quantities of a
weak acid site and comparable quantities of a moderate acid site.
Figure 3a shows that these catalysts produce less THO than K- and
H-BEA, but more than Rb- and Cs-BEA. Finally, K-BEA shows more
weak acid sites than Rb- and Cs-BEA, and essentially no moderate
acid sites. K-BEA is the best of all metal-exchanged catalysts in
terms of THO yield, as shown by the data in Figure 3a. These
observations are in line with our working hypothesis that the weak
acid site is responsible for THO production. Furthermore, the parent
catalyst, H-BEA, shows high levels of this weak acid site, along with
a smaller, but still significant amount of moderate acid sites. H-BEA
Figure 3. THO yields (a) and THP2M conversion (b) over alkali-
exchanged BEA zeolites. Reaction conditions: 400 oC, 0.628 mL/hr
THP2M (liquid flow rate, STP), 30 mL/min (STP) H2, 1 atm, 150 mg
catalyst.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 4 of 10
View Article Online
DOI: 10.1039/C6GC03606F
ARTICLE
Journal Name
has a higher ratio of weak sites to moderate sites than any other related to the concentration of moderate-strength acid sites on the
catalyst except K-BEA. This, along with the fact that K- and H-BEA catalyst. Table S2 and Figure S4 show that this trend does not apply
are the two catalysts in this study that have the highest THO yield, for other frameworks. This could be due to several reasons, such as
further supports the weak acid site being active for THO production. the possibility that the frameworks with five-membered rings (MFI
and MOR) are inaccessible by the reactant or certain intermediates,
Table 2. NH3-desorption signals fitted for strong, moderate, and thus decreasing reactivity and coke formation.
weak acid Sites on alkali-exchanged BEAs. Ramp Rate = 10 oC/min
Table 3. Brønsted/Lewis acid concentrations determined by FTIR of
Catalyst NH3
Assignment
Quantity
pyridine remaining on each catalyst at 250 oC after adsorption at 25
oC. 1540 cm-1 band used for Brønsted site, 1450 cm-1 band used for
Lewis site.29 Extinction coefficients shown in Experimental Section.
Desorption
Temp. (oC)
Desorbed
(µmol NH3/g
catalyst)
H-BEA
Li-BEA
144
161
289
176
258
174
238
174
290
142
168
151
244
Physisorbed NH3 150
Catalyst
Brønsted Acid
Concentration
Lewis Acid
Concentration
Weak Acid
615
89.8
669
135
549
214
399
7.02
Moderate Acid
Weak Acid
(µmol/g catalyst)
(µmol/g catalyst)
H-BEA
Li-BEA
Na-BEA
K-BEA
Rb-BEA
Cs-BEA
Na-Y
Na-MOR
Na-ZSM-5
483
322
222
482
679
394
140
73.9
348
1140
892
Moderate Acid
Weak Acid
84.0
11.4
4.63
3.09
25.7
199
Na-BEA
K-BEA
Moderate Acid
Weak Acid
Moderate Acid
Rb-BEA
Cs-BEA
Physisorbed NH3 74.9
78.9
Weak Acid
145
Weak Acid
55.3
21.7
Complete product distributions for all dehydrations can be
found in the supporting information (see Tables S5-S13 and Table
S4 for explanation of abbreviations). Carbon balances in the range
of 75-90% can be attributed to coke formation and a large amount
of insoluble carbon compounds in the reactor effluent, which could
not be accurately quantified. THO selectivity as high as 58% for H-
BEA at 28% conversion was obtained.
Moderate Acid
FTIR of adsorbed pyridine was performed on each catalyst to
determine the Brønsted/Lewis acid nature of the active site for THO
production. Pyridine was adsorbed on dehydrated samples at 25 oC,
o
which were subsequently heated to 250 C under a flow of dry N2.
Table 3 shows the Brønsted and Lewis acid concentrations for each
catalyst. Concentrations were calculated by multiplying the
Brønsted/Lewis acid ratios—determined by FTIR—by the overall
concentration of acid sites—determined by NH3-TPD (physisorbed
NH3 was not included in this calculation). H-BEA shows much more
Brønsted acid sites (an order of magnitude in most cases) than all
other catalysts. This indicates that the aqueous ion-exchange
performed for the synthesis of the metal-exchanged catalysts
effectively exchanges with Brønsted sites.
Figure 5 shows the major products of THP2M dehydration at
four different contact times over K-BEA. Scheme 2 shows a possible
reaction pathway based on Figure 5. Yields to 6,8-
dioxabicyclooctane (DBO)—while relatively low at all contact
times—decrease with increasing contact time, indicating it is
formed early in the reaction network. Yields to tetrahydropyran-2-
carbaldehyde (THP2C) stay relatively constant with contact time.
However, not shown in the figures, THP2M conversion increases
with increasing contact time, meaning the selectivity to THP2C
drops with increasing contact time. This supports the hypothesis
that THP2C is indeed formed early in the reaction sequence, as
shown in Scheme 2. The fact that yields to THO reach a maximum at
an intermediate contact time indicates that it can be degraded into
further reaction products, also shown in Scheme 2. This would
explain why we were unable to obtain higher THO yields with these
catalysts; at low contact times, THO intermediates are formed,
while at higher contact times, THO degradation occurs. Oxepane
Table 3 also shows that the two catalysts that have the highest
THO yield—K-BEA and H-BEA—have Brønsted acid concentrations
that differ by nearly two orders of magnitude, while their Lewis acid
concentrations differ by only a factor of two. This indicates that
Brønsted acids are irrelevant towards THO production, and Lewis
acids are the active sites. This, along with the data from NH3-TPD,
leads us to the conclusion that THO production from THP2M is
catalyzed by weak Lewis acid sites.
TPO was used to characterize the alkali-exchanged catalysts
after reaction (Table S2, Figure S5). As expected, Rb- and Cs-BEA
show the least amount of mass loss due to coking, as these are the
least active catalysts (similar effect seen with Na-MOR and Na-ZSM-
5). Furthermore, Li- and Na-BEA show 1.2-1.9 times the amount of
coke that other BEA catalysts produce. In fact, Table S2 and Figure
S5 show that the amount of coke formed on BEA catalysts is directly
(OXE),
cyclopentanecarbaldehyde
(CPC),
5-hexenal,
dihydromethylpyran (DMP), and 2-methylcyclopentanone (2MCP)
increase in yield with increasing contact time, indicating that they
are among the degradation products formed late in the reaction
network. Due to the high cost and unavailability of reaction
intermediates (including THO) in their pure forms, we could not
4 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 5 of 10
Journal Name
View Article Online
DOI: 10.1039/C6GC03606F
ARTICLE
perform reactions using each intermediate as a reactant. These seven-membered cyclic ether (see Table S3). Furthermore, the C-O-
experiments would provide a more definitive reaction network, and C bond angle of the oxygen in the five-membered ring of DBO was
should be part of the scope of future work. In our current work, compared to that of the same bond in tetrahydrofuran (THF), the
Scheme 2 represents a possible reaction pathway supported by our fully hydrogenated five-membered cyclic ether. Table S3 indicates
data and similar reactions in the literature.30-40
that the C-O-C bond in the five-membered ring of DBO is more
strained than that of the seven-membered ring. The C-O-C bond
angle of the DBO five-membered ring is 8.0o smaller than that of
THF, while the C-O-C bond angle of the DBO seven-membered ring
is only 2.4o smaller than that of OXE. The higher strain on the
oxygen atom in the five-membered ring of DBO (bridging oxygen)
indicates that it is the less stable bond. Moreover, Table S3 shows
that the C-O bonds in the five-membered ring of DBO is 0.013 Å
longer than the C-O bonds in the seven-membered ring of the same
compound, once again indicating that the oxygen in the five-
membered ring is more reactive. If DBO dehydrates at the oxygen in
the five-membered ring, it will likely form THO. Therefore, our
calculations indicate that DBO could be the precursor to THO.
Figure 5. Product distribution with changing catalyst contact time
(a), same data zoomed in on byproducts (b). Low time-on-stream (4
hrs) was used for all measurements to minimize the effects of
coking, although this led to relatively low carbon balances, in the
range of 77-79% for each measurement. K-BEA used for each test.
Scheme 2. Possible reaction network for dehydration of THP2M
(major products shown).
The final desirable step in our network, that is, the dehydration
of DBO to THO, cannot be found in literature, as the mechanism of
THO formation has not been studied. We have already shown that
DBO is likely formed early in the reaction network, and DBO is the
only product formed early in the reaction network that has a seven-
membered ring. It seems likely, therefore, that THO is formed from
the dehydration of DBO. The portrayal of DBO in Scheme 2 shows
that one oxygen atom of DBO is part of a seven-membered ring,
and the other (bridging) oxygen is part of a five-membered ring. To
find out which of these sites is more prone to dehydration to THO,
quantum-chemical calculations were performed. The C-O-C bond
angle of the oxygen in the seven- membered ring in DBO was
compared to that of the same bond in OXE, the fully hydrogenated
THO Hydration
The product from THP2M dehydration over K-BEA was then
subjected to hydration conditions to produce OXL and 6HDHX from
THO (see Scheme 1). The hydration of the THO in the reaction
mixture proceeded homogeneously in water at 70 °C for 5h in an
autoclave. This resulted in the complete conversion of THO and the
formation of one main gas chromatogram (GC) signal. Nuclear
Magnetic Resonance (NMR) spectroscopy was performed on the
product mixture. As shown in the Quantitative 13C NMR spectrum in
Figure 6, there exists two compounds that are only surpassed in
intensity by unconverted THP2M from the dehydration step.41,42
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 6 of 10
View Article Online
DOI: 10.1039/C6GC03606F
ARTICLE
The relative ratio of peak intensities of THP2M to the two
Journal Name
unknown species in 13C NMR matches those of THP2M to the
hydration product in GC analysis, indicating these are indeed the
hydration products and that their GC peaks overlap one another.
The NMR peak at 91.38 ppm is very likely the hemiacetal carbon of
OXL, the direct hydration product of THO. The NMR peak at 209.31
ppm is indicative of an aldehyde, likely 6HDHX, the ring-opened
tautomer of OXL (see Scheme 1). This is further evidenced by Dept.
135 NMR analysis (see Figure S7),41,42 which indicates both peaks
are either primary or tertiary carbons, likely the primary carbons of
OXL and 6HDHX. Heteronuclear single quantum correlation (HSQC)
and heteronuclear multiple bond correlation (HMBC) NMR analyses
were performed and HMBC correlations confirmed the 6 carbons of
each hydration product (see Figs. 6 and S7).41,42
Figure 7. OXL + 6HDHX (□) and 1,6-HDO (●) concentraꢀon at various
time points in a batch reactor (130 mg Ru/C diluted 20x in Silica gel
[6.5 mg Ru/C], T=120 °C, P=6.4 MPa).
As the ring-opening tautomerization of OXL to 6HDHX is
thermodynamically favorable (as shown by gas-phase
thermochemistry calculations), it is likely that 6HDHX is the main
product of the hydration step and was the main 1,6-HDO precursor.
This is further corroborated by the fact that 1,6-HDO can be formed
via simple hydrogenation with
a monometallic hydrogenation
catalyst; ring-opening of OXL would require the use of a bimetallic-
oxophilically promoted catalyst.9 Additionally, the reaction was
determined to be 1st-order in 6HDHX (see Figure 7 insert). The
Figure 6. Quantitative 13C NMR spectra of reaction solution formed reaction rate of 994 μmol gcat-1 min-1 is an order of magnitude
via hydration of THP2M dehydration product.
higher than achieved in typical C-O-C bond hydrogenolysis
reactions,9 and suggests the hydrogenation of an aldehyde to be
rate determining.43
Hydrogenation to 1,6-HDO
The hydration product was then hydrogenated over a pre-
reduced Ru/C catalyst in a batch reactor. Prior to the reaction,
Quantitative 13C NMR was used to quantify the concentration of
OXL and 6HDHX in the feed (28.5 mM) by using the peak ratio of
THP2M to the hydration products.41,42 The hydrogenation reaction
resulted in the complete conversion of the hydration products and
~98% carbon yield to 1,6-HDO after 22 h (see Figure 7). There was
no appreciable side-product formation in the hydrogenation step. A
maximum of 85% 1,6-HDO yield from THO was achieved. Any loss in
1,6-HDO yields from THO was likely the result of oligomer/polymer
formation in the hydration step. In fact, hydrations at higher
temperature (100 ˚C) resulted in the formation of solid residue in
the reactor. Moreover, essentially identical results over this time
scale were achieved using a Ni/C catalyst instead of Ru/C for the
hydrogenation, under the same reaction and reduction conditions.
However, Inductively Coupled Plasma (ICP) analysis of the product
streams of the Ru/C and Ni/C hydrogenations displayed the
instability of Ni/C catalysts, with 31% of the Ni originally present in
Ni/C leached into the product stream. In contrast, less than 1% of
the Ru originally present in Ru/C leached into the product stream
during its reaction.
Conclusions
An alternative to the conventional Pt, Rh, and Ir-based catalysts
for the conversion of tetrahydropyran-2-methanol (THP2M) to 1,6-
hexanediol (1,6-HDO) is presented. While the existence of this
pathway has been known since at least the 1960s, it has been
essentially ignored over the past 40-50 years. In the first of three
steps, THP2M is dehydrated to 2,3,4,5-tetrahydrooxepine (THO) in a
continuous flow, gas-phase reactor over a silicoaluminate catalyst.
While this study was not able to replicate THO yields over
amorphous SiO2-Al2O3 cited in the literature,19,20 data from this
study indicates that zeolite materials produce nearly three times
higher yields of THO than amorphous SiO2-Al2O3. Various zeolite
frameworks, as well as various alkali metals as exchange ions were
tested in this dehydration reaction. Maximum THO yields of 40%
were achieved over K-BEA. THO production over K-BEA benefits
from the catalyst’s large quantity of moderate-strength Lewis acid
sites (399 cm3 NH3 desorbed/g catalyst), and very few, if any, strong
or weak (Brønsted or Lewis) acid sites. Without the use of any
catalyst, THO was then hydrated in 85% yields to 6-hydroxyhexanal
6 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 7 of 10
Journal Name
View Article Online
DOI: 10.1039/C6GC03606F
ARTICLE
(6HDHX) and 2-oxepanol (OXL), as confirmed by NMR. 6HDHX and spectrum was background subtracted and normalized by the
OXL (likely through 6HDHX) were then hydrogenated to 1,6-HDO intensity of silica framework signals in the range of 1800-2100 cm-1.
with Ru/C or Ni/C nearly quantitatively. Therefore, our pathway The time-zero spectrum was subtracted from each of the spectra
o
affords an overall yield of 34% to 1,6-HDO from THP2M, higher than taken at 250 C so that only signals due to interaction with pyridine
previously reported 1,6-HDO yields from THP2M without using toxic remained. Finally, the signals at 1540 and 1450 cm-1 were
metals.20 In future research, optimizing Lewis acid strength and integrated for Brønsted and Lewis acid sites, respectively. Extinction
concentration could potentially lead to higher THO selectivity, and coefficients of 1.3 × 106 cm-1/mol and 1.5 × 106 cm-1/mol were used
therefore improve the overall viability of this route.
to compare quantities of Brønsted and Lewis acid sites,
respectively.44
Temperature-programmed oxidation experiments with spent
catalysts were carried out with a Mettler Toledo TGA-DSC1.
Samples were heated to 600 oC at 5 oC/min under a 20 mL/min flow
of 20%O2-N2 (Airgas). Samples were then held at 600 oC for two
hours under the same gas flow.
Experimental Details
Materials Synthesis
Amorphous SiO2-Al2O3 (Sigma, grade 135) was used for
comparison to zeolites in THP2M dehydration. All zeolite
frameworks were purchased from Zeolyst in their hydrogen form
with SiO2:Al2O3 ratios of 20-30:1. All alkali exchanges were done at
80 oC in 1 M solutions of the corresponding nitrate in water for 8
hours. After exchange, the catalyst was washed with one liter of
distilled water per gram of catalyst and dried at 110 oC overnight.
This ion-exchange was repeated two additional times, using two
liters of distilled water per gram of catalyst for the final washing.
After the final drying step, the catalyst was calcined under a flow of
Reaction Testing
Dehydration of THP2M was performed in a continuous flow
(down-flow) reactor in the gas phase. The reactor was a quartz tube
(52 cm long, 1.1 cm internal diameter) with a quartz frit in the
middle to support the catalyst. The reactor was packed with 150 mg
catalyst between layers of glass wool. Temperature was monitored
and controlled by a thermocouple placed inside the reactor, just
above the catalyst bed. The reactant was mixed with carrier gas
above the furnace (Carbolite, VST12/300), and preheated in a layer
of glass beads within the furnace before reaching the catalyst bed.
Gas flow was controlled by mass flow controllers (Bronkhorst, EL-
FLOW). Reactant flow was controlled by syringe pump (Chemyx,
Fusion 200). Products were collected in a glass collection vessel
inside a dry ice/acetone bath. Samples were taken at 4, 7, 10, 13,
16, and 25 hours on stream. The 4-hour sample normally showed
lower carbon balances, likely due to the need to fully coat the
reactor at the beginning of the reaction. That being the case, the 4-
hour sample is not reported in this study, except when investigating
the reaction network. The 4-hour sample is used for this purpose to
minimize the effects of coking on our analysis. Products were
analyzed by an Agilent gas chromatograph (HP6890) equipped with
an HP-5 column and a flame ionization detector.
o
o
o
air to 580 C at a rate of 3 C/min, holding at 580 C for 6 hours. 5
wt% Ru/C was purchased from Sigma-Aldrich. 5 wt% Ni/C was
prepared by incipient wetness impregnation of nickel nitrate
hexahydrate (Sigma, >98.5% (KT))in water onto Vulcan XC-72 (Cabot
Corp.)
Catalyst Characterization
Exchange rates of zeolites synthesized in this study were
determined by ICP-AES after digesting the solid samples in HF. NH3-
TPD measurements were carried out with
a Micromeritics
Autochem II Chemisorption Analyzer. Prior to NH3-TPD analysis,
samples were dehydrated at 580 oC for three hours (10 oC/min)
under a flow of He. Samples were then cooled to 100 oC for NH3
adsorption. Adsorption was performed for 30 min at 100 oC under a
flow of 40 mL/min 10%NH3-Ar (Airgas). NH3 was desorbed under a
40 mL/min flow of He while heating at 10 oC/min.
The organic product solution formed from THP2M dehydration
over K-BEA was added dropwise to DI water up to 2wt% organic
fraction in a 50mL Hastelloy Parr reactor. This was done slowly
while stirring in order to prevent organic fraction loss to reactor
walls and the stir bar. An insoluble organic layer was formed upon
addition. The reactor was closed and pressurized to 500 psi with
helium after 2 purge cycles. The reactor was heated to reaction
temperature while stirring (500 rpm). An ice bath was used to
quench the reaction after the desired reaction time. A single,
water-soluble phase was formed after hydration. Liquid product
samples were filtered with Restek PES syringe filters prior to GC
analysis (Shimadzu GC2010 equipped with a flame ionization
detector and an RTX-VMS column).
FTIR experiments were performed with a Bruker Vertex 70
Fourier Transform Infrared Spectrometer equipped with an MCT
detector. 32 scans were averaged to give one spectrum with a
resolution of 4.00 cm-1. A flow-through transmission cell was loaded
with approximately 25 mg of sample that was pressed into a self-
supporting wafer with a diameter of 13 mm. Samples were
o
dehydrated at 350 oC (10 C/min, hold for 1 hour) under 10-7 mbar.
Samples were then cooled to 25 oC, and the time-zero spectrum
was taken. Pyridine was then flowed over the catalyst for 10 min by
way of a bubbler with a carrier gas of dry N2 at atmospheric
pressure. The cell was then flushed with dry N2 for 10 min to
remove physisorbed pyridine. Samples were then heated at 10
oC/min to 250 oC, and held at that temperature for 20 min while
o
The catalyst used for hydrogenation was a 5wt% Ru/C catalyst
(Sigma Aldrich #206180) diluted 20x in Davisil silica gel (Sigma
Aldrich #236845). The catalyst was pre-reduced in UHP H2 (Airgas,
100 mL/min) at 300˚C (4 ˚C/min ramp rate, 1 hr hold time) in a glass
taking spectra of the catalyst. Samples were also heated to 150 C
and 350 oC with the same procedure, showing similar results. The
final spectrum at 250 oC was used for contribution to this paper.
The Opus 7.0 software package was used to analyze the data. Each
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 8 of 10
View Article Online
DOI: 10.1039/C6GC03606F
ARTICLE
Journal Name
tube reactor with Ultratorr fittings and ball valves at the inlet and Magnetic Resonance Facility in the Chemistry Department of the
outlet. Upon cooling to room temperature, the H2 flow was University of Wisconsin-Madison for all NMR experiments.
stopped and the ball valves shut. The catalyst was transferred to a Specifically, we acknowledge the gift of Paul J. Bender and NIH
glove box with He atmosphere at positive pressure and stored. Grant# 210OD012245 for NMR spectrometer funding. We also
Prior to the hydrogenation, 130 mg of total catalyst (6.5 mg Ru/C) acknowledge Zachary J. Brentzel, Juan M. Venegas, Philipp Müller,
was added to a 75 mL Hastelloy Parr reactor equipped with a dip Kefeng Huang, and Christos T. Maravelias for helpful discussion
tub for time-on-stream sampling in the glove box. After isolating throughout this project.
the catalyst and removing the reactor from the glove box, 40 mL of
the product from the hydration step was added to the reactor with REFERENCES
an HPLC pump after removing any residual oxygen from the lines.
The reactor was pressurized to 6.4 MPa with H2, heated to 120 ˚C,
and stirred at 750 rpm. Samples were taken at desired reaction
times with a dip tube and the reactor re-pressurized to 6.4 MPa
after each sample. Samples were filtered with Restek PES syringe
filters prior to GC analysis (Shimadzu GC2010 equipped with a flame
ionization detector and an RTX-VMS column).
1.
2.
P., Werle, M. Morawietz, S. Lundmark, K. Sörensen,
E. Karivinen, and J. Lehtonen, 2012, Alcohols,
Polyhydric. In Ullmann’s Encyclopedia of Industrial
Chemistry.
W. Tianfu, M. S. Ide, M. R. Nolan, R. J. Davis, and B.
H. Shanks, Energy and Environmental Focus, 2016, 5,
13–17.
3.
4.
C. Angelici, B. M. Weckhuysen, and P. C. A.
NMR
Bruijnincx, ChemSusChem, 2013, 6, 1595–1614.
A. Wang and T. Zhang, Accounts of Chemical
Research, 2013, 46, 1377–1386.
The hydration product was characterized by NMR with
quantitative 13C, 13C DEPT-135, 2D HSQC, and 2D HMBC
experiments. D2O was added to the sample (1:9 v/v). The 13C NMR
experiments were acquired on a Bruker Biospin (Billerica, MA)
AVANCE III 500 MHz spectrometer fitted with a DCH (13C-optimized)
cryoprobe. Bruker standard pulse sequence ‘zgig30’ was used for
the quantitative 13C experiments with the following parameters: an
inter-scan relaxation delay of 40 s, a sweep width of 240 ppm
centered at 110 ppm, acquiring 59,520 data points with an
5.
6.
S. Van de Vyver and Y. Roman-Leshkov, Catalysis
Science & Technology, 2013, 3, 1465–1479.
J. Mormul, J. Breitenfeld, O. Trapp, R. Paciello, T.
Schaub, and P. Hoffman, ACS Catalysis, 2016 6,
2802–2810.
7.
8.
A. M. Allgeier, W. I. N. De Silva, K. Ekaterini, C. A.
Menning, J. C. Ritter, S. K. Sengupta, and C. S.
Stauffer, 2014, Process For Preparing 1,6-Hexanediol.
M. Chia, Y. J. Pagán-Torres, D. Hibbitts, Q. Tan, H. N.
Pham, A. K. Datye, M. Neurock, R. J. Davis, and J. A.
Dumesic, Journal of the American Chemical Society,
2011, 133, 12675–12689.
acquisition time of
1
s, and 256 scans. The 13C DEPT-135
experiments used the Bruker standard pulse sequence ‘deptsp135’
with the following parameters: an inter-scan relaxation delay of 2 s,
a sweep width of 240 ppm centered at 110 ppm, acquiring 59,520
data points with an acquisition time of 1 s, and 256 scans.
Mestrelab Research’s MestReNova software was used to process
the spectra.
9.
M. Chia, B. J. O’Neill, R. Alamillo, P. J. Dietrich, F. H.
Ribeiro, J. T. Miller, and J. A. Dumesic, Journal of
Catalysis, 2013, 308, 226–236.
The 2D NMR (HSQC and HMBC) experiments were carried out
on a Bruker Biospin (Billerica, MA) AVANCE III HD 600 MHz
spectrometer fitted with a TCI-F cryoprobe. Bruker standard pulse
sequence ‘hsqcedetgpsisp2p3’ was used for the HSQC experiment
with the following parameters: 14 ppm sweep width in F2 (1H),
centered at 4.7 ppm, acquiring 3,366 data points, 240 ppm sweep
width centered at 110 ppm in F1 (13C) acquiring 1,309 increments, 4
scans per increment, and a 2.0 s relaxation delay. Bruker standard
pulse sequence ‘hmbcgplpndprqf’ was used for the HMBC
experiment with the following parameters: 14 ppm sweep width
centered at 4.7 ppm in F2 (1H) acquiring 3366 data points, 240 ppm
sweep width centered at 110 ppm in F1 (13C) acquiring 1,309
increments, 4 scans per increment, and a 2.0 s relaxation delay.
Bruker’s Topspin 3.5 software was used to process spectra.
10.
P. U. Karanjkar, S. P. Burt, X. Chen, K. J. Barnett, M. R.
Ball, M. D. Kumbhalkar, J. B. Miller, I. Hermans, J. A.
Dumesic, and G. W. Huber, Catalysis Science &
Technology, 2016, 6, 7841–7851.
Y. Kojima, S. Kotani, M. Sano, T. Suzuki, and T.
Miyake, Journal of the Japan Petroleum Institute,
2013, 56, 133–141.
11.
12.
K. Chen, S. Koso, T. Kubota, Y. Nakagawa, and K.
Tomishige, ChemCatChem, 2010, 2, 547–555.
13.
14.
Y. Nakagawa, M. Tamura, and K. Tomishige, ACS
Catalysis, 2013, 3, 2655–2668.
S. Koso, I. Furikado, A. Shimao, T. Miyazawa, K.
Kunimori, and K. Tomishige, Chemical
Communications, 2009, 15, 2035–2037.
Y. Nakagawa, Y. Shinmi, S. Koso, and K. Tomishige,
Journal of Catalysis, 2010, 272, 191–194.
B. Xiao, M. Zheng, X. Li, J. Pang, R. Sun, H. Wang, X.
15.
16.
Acknowledgements
Pang, A. Wang, X. Wang, and T. Zhang, Green
Chemistry, 2016, 18, 2175–2184.
This material is based upon work supported by the Department of
Energy, Office of Energy Efficiency and Renewable Energy (EERE),
under Award Number DE-EE0006878. The authors acknowledge the
8 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 9 of 10
View Article Online
DOI: 10.1039/C6GC03606F
Journal Name
ARTICLE
17.
T. Buntara, S. Noel, P.H. Phua, I. Mellán-Cabrera, J. G.
de Vries, and H. J. Heeres, Angewandte Chemie
International Edition, 2011, 50, 7083–7087.
Z. J. Brentzel, K. J. Barnett, K. Huang, C. T. Maravelias,
J. A. Dumesic, and G. W. Huber, Commodity
Chemicals from Biomass: Combining Ring-opening
Tautomerization and Hydrogenation Reactions to
Produce 1,5-Pentanediol from Furfural, Submitted.
32.
33.
34.
H. Hagiwara and H. Uda, Journal of the Chemical
Society, Perkin Transactions 1, 1985, 0, 283–287.
A. P. Ranney and P. J. Ziemann, The Journal of
Physical Chemistry A, 2016, 120, 2561–2568.
V. A. Kuplenieks, D. R. Kreile, V. A. Slavinskaya, and A.
A. Avots, Chemistry of Heterocyclic Compounds,
1982, 18, 351–354.
18.
19.
20.
21.
D. Tysee, 1972, Preparation of
tetrahydrooxepin.
2
3
4
5-
35.
36.
C. L. Wilson, Journal of the American Chemical
Society, 1947, 69, 3002–3004.
A. Misono, T. Osa, and Y. Sanami, Bulletin of the
Chemical Society of Japan, 1968, 41, 2447–2453.
Y. Inoue, N. Matsumoto, T. Hakushi, and R.
Srinivasan, The Journal of Organic Chemistry, 1981,
46, 2267–2272.
B. Dhal, H. Thatoi, N. Das, and B. D. Pandley, Journal
of Chemical Technology & Biotechnology, 2010, 85,
1471–1479.
L. Guo, J. Zhou, J. Mao, X. Guo, and S. Zhang, Applied
Catalysis A: General, 2009, 367, 93–98.
37.
38.
J. Wang, M. Wang, J. Xiang, H. Xi, and A. Wu,
Tetrahedron, 2015, 71, 7687–7694.
22.
23.
K. Sekizawa, in Successful Design of Catalysts: Future
Requirements and Development, ed. T. Inui, Elsivier
Science Publishers, New York, 1989, 44, 203-217.
M. Ziolek, D. Szuba, and R. Leskowski, Studies in
Surface Science and Catalysis, 1988, 37, 427-434.
P. A. Jacobs, M. Tielen, and J. B. Uytterhoeven,
Journal of Catalysis, 1977, 50, 98-108.
A. Corma, A. Martinez, V. Martinezsoria, and J. B.
Monton, Journal of Catalysis, 1995, 153, 25–31.
M. J. Climent, A. Corma, S. Iborra, M. C. Navarro, and
J. Primo, Journal of Catalysis, 1996, 161, 783–789.
J. Jae, G. A. Tompsett, A. J. Foster, K. D. Hammond, S.
M. Auerbach, R. F. Lobo, and G. W. Huber, Journal of
Catalysis, 2011, 279, 257–268.
39.
40.
41.
24.
25.
R. M. Silverstein, F. X. Webster, and D. Kiemle,
Spectrometric Identification of Organic Compounds,
7th Edition; Wiley, 2005.
26.
27.
D. Liu, P. Yual, H. Liu, J. Cai, D. Tan, H. He, J. Zhu, and
T. Chen, Applied Clay Science, 2013, 80-81, 407-412.
V. D. Kumari, B. P. Prasad, and V. V. Rao, Indian
Journal of Chemical Technology, 1995, 3, 95-100.
E. Dima and L. V. C. Rees, Zeolites, 1990, 10, 8-15.
42.
43.
44.
J. Fisher, Modern NMR Techniques for Synthetic
Chemistry; Taylor & Francis, 2014.
J. Hack, R. Falcone, and R. Davis, ChemCatChem,
2016, 8, 1074-1083.
28.
29.
M. Zaki, M. A. Hasan, F. A. Al-Sagheer, and L.
Pasupulety, Colloids and Surfaces A: Physiochemical
and Engineering Aspects, 2001, 190, 261-274.
V. L. Sushkevich, I. I. Ivanova, and E. Taarning,
I. Kiricsi, C. Flego, G. Pazzuconi, W. O. Parker Jr., R.
Millini, C. Perego, and G. Bellussi, The Journal of
Physical Chemistry, 1994, 98, 4627-4634.
30.
31.
ChemCatChem, 2013, 5, 2367–2373.
T. D. J. D’Silva and D. W. Peck, Journal of Organic
Chemistry, 1972, 37, 1828–1829.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins

Green Chemistry
Page 10 of 10
View Article Online
DOI: 10.1039/C6GC03606F
288x140mm (96 x 96 DPI)