Expanding the scope of biomass-derived chemicals through tandem reactions based on oxorhenium-catalyzed deoxydehydration

Article abstract of DOI:10.1002/anie.201307564

New modes of DODH: Oxorhenium compounds act as deoxydehydration(DODH)/acid dual-purpose catalysts to transform biomass-derived diol substrates into a variety of commodity chemical precursors. The power of this approach is highlighted by a tandem [1,3]-OH shift/DODH of 2-ene-1,4-diols and 2,4-diene-1,6-diols, and by a DODH/esterification sequence of sugar acids to unsaturated esters for the production of polymers and plasticizers. Copyright

Full text of DOI:10.1002/anie.201307564

Angewandte
Chemie
DOI: 10.1002/anie.201307564
Biomass Conversion
Expanding the Scope of Biomass-Derived Chemicals through Tandem
Reactions Based on Oxorhenium-Catalyzed Deoxydehydration**
Mika Shiramizu and F. Dean Toste*
With the growing demand for sustainability, cellulosic bio-
mass has attracted much attention as a renewable, carbon-
neutral and inexpensive feedstock for chemicals and fuels.
However, the conversion of biomass faces the fundamental
challenge that saccharides and their polyol derivatives, the
most basic platform chemicals accessible from cellulose, are
too oxygen-rich to be compatible with the current petroleum-
based infrastructure. The search for efficient deoxygenation
methods has resulted in rapidly growing interest in the
catalytic deoxydehydration (DODH) reaction to remove two
adjacent hydroxyl groups from vicinal diols to afford
alkenes.[1] Although preliminary studies on ruthenium,[2]
vanadium,[3] molybdenum,[4] and rhenium carbonyl catalysts
(Re2(CO)10 or BrRe(CO)5)[5] have been reported, most
precedents of DODH employ high-valent oxorhenium cata-
lysts in conjunction with various reductants, such as phos-
phines,[6] H2,[7] and NaSO3.[8] In 2012, we reported the
Scheme 1. The general DODH reaction. Red.=reductant.
methyltrioxorhenium (MTO)-catalyzed DODH reaction
using a sacrificial alcohol (3-pentanol, 3-octanol, 1-butanol)
as reductant/solvent.[9] This system smoothly converted sugar
alcohols and sugars into linear alkene products and aromatics
with remarkable selectivity. Oxorhenium-catalyzed DODH
of diols using glycerol[10] or benzyl alcohol[11] as reductants
were also independently reported by the Abu-Omar and
Nicholas groups.
first examples of formal 1,4-DODH and 1,6-DODH, thus
opening new possibilities for DODH reaction development.
The inspiration of DODH on non-vicinal diols arose from
our efforts to clarify the polyol DODH mechanism. In our
previous study on the MTO-catalyzed DODH of sugar
alcohols,[9] we generally observed only complete DODH
products. For example, the DODH of erythritol (21) gave
butadiene (2) as a major product and the partial DODH
product (Z)-but-2-ene-1,4-diol (1) was not observed. Like-
wise, d-chiro-inositol, muco-inositol, and allo-inositol all gave
high benzene yields (> 45%), without forming cyclohexa-2,5-
diene-1,4-diol and cyclohexa-3,5-diene-1,2-diol. This led us to
speculate that, in addition to cis-vicinal diols, 1,4- and 1,6-
allylic diols might be reactive toward DODH through the
oxorhenium-catalyzed [1,3]-OH shift of an allylic alcohol.[14]
To examine this hypothesis and gain insights into the origin of
the observed exquisite selectivity in polyol DODH reactions,
we tested a series of 2-ene-1,4-diol and 2,4-diene-1,6-diol
substrates and indeed observed the unprecedented 1,4-
DODH and 1,6-DODH reactions (Table 1).[15] Both cis- and
trans- 2-butene-1,4-diol (1 and 4, respectively) were reactive,
excluding the possibility of the direct coordination of 1,4-diols
to Re forming a 7-membered ring Re diolate (entries 1 and 2).
The observation that no 2 was produced from (Z)-4-methoxy
2-buten-1-ol (5) under the same conditions supports the
requirement for two OH groups (entry 3). Trisubstituted
alkene 6 gave more dehydration product[16] than DODH
product, possibly because the electron-donating alkyl group
stabilizes the allylic carbocation intermediate (entry 4). No
significant reactivity difference was observed between cis-
and trans- cyclic diols, which suggests that [1,3]-OH shift is in
Although the significant potential of the DODH reaction
in the context of biomass conversion has been well recog-
nized, the method development is still in the early stages. As
previously noted, one particular feature of the DODH
reaction is the need for a vicinal cis-diol structure (1,2-
DODH): For example, whereas cis-1,2-cyclohexanediol is
reactive, trans-1,2-cyclohexanediol is not. This observation
has been explained by the required formation of a metal cis-
diolate intermediate,[12] which undergoes the equivalent of the
reverse reaction of the OsO4-catalyzed cis-dihydroxylation of
olefins.[13] However, we have found that oxorhenium com-
plexes catalyze not only the DODH reaction, but also the
[1,3]-OH shift of allylic alcohols[14] in a tandem manner, and
we herein report novel modes of DODH (1,4-DODH and 1,6-
DODH, respectively) from 2-ene-1,4-diols and 2,4-diene-1,6-
diols (Scheme 1). To the best of our knowledge, these are the
[*] M. Shiramizu, Prof. F. D. Toste
Department of Chemistry University of California, Berkeley
Berkeley, CA 94720 (USA)
E-mail: fdtoste@berkeley.edu
[**] This work was supported by the Energy Biosciences Institute.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201307564.
Angew. Chem. Int. Ed. 2013, 52, 12905 –12909
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
12905

.
Angewandte
Communications
Table 1: The DODH of 2-ene-1,4-diol and 2,4-diene-1,6-diol moieties.
Entry Diol
1[a]
DODH product Dehydration product
2[a]
3[a]
Figure 1. Proposed catalytic cycle for the oxorhenium-catalyzed DODH
reaction.
4
ally well suited for polyol DODH compared to other
reductant systems not merely because the substrates have
better solubility, but also because the enhanced proton
transfer[12a] maximizes the benefit of auto-selectivity increase.
Ensured of the efficiency of the polyol DODH reaction,
we then turned to applying our system to the production of
commodity chemicals from biomass. In this regard, sugar
acids, particularly C6 aldaric acids, caught our attention
because the expected product muconic acid[18] has wide utility
in the chemical industry as a precursor to adipic acid,[19]
terephthalic acid,[20] and 1,6-hexanediol.[21] Moreover,
although C6 aldaric acids have four internal hydroxyl
groups, our findings in the 1,4-DODH reaction suggested
that, regardless of whether the first DODH takes place at the
a,b-position or on b,g-position, the second DODH would
furnish muconic acid. We therefore tested mucic acid (23), the
oxidized form of galactose, because the product bearing the
stable trans,trans-stereochemistry (24) was expected both
from b,g-/a,d- DODH (sterics) and from a,b-/g,d- DODH
(cis-diol stereospecificity).[22]
In the initial experiment using MTO and 3-pentanol
(reductant/solvent), we obtained 24 in 43% yield,[23] exclu-
sively with trans,trans-stereochemistry, accompanied by die-
ster 25 (14% yield; Scheme 2a). As 25 was easier to
manipulate and purify than 24, we sought to shift the
selectivity towards the ester product. We thus envisioned
replacing MTO with perrhenic acid (HReO4), which earlier
studies identified as a promising DODH catalyst.[9] While the
Brønsted acidity of perrhenic acid had previously caused
decomposition of sugar substrates, we anticipated that
HReO4 may conveniently catalyze both DODH and the
in situ acid-catalyzed esterification reaction of sugar acids in
a tandem manner. Gratifyingly, by using a slightly higher
temperature and the sterically accessible primary alcohol 1-
butanol instead of a secondary alcohol, we obtained trans,-
trans-dibutyl muconate 26 in 62% yield under air and in 71%
yield under an inert atmosphere (Scheme 2b).[24] The majority
of 1-butanal was trapped as acetal 27 with this system. When
mucic acid was pre-esterified in refluxing 1-butanol/HCl,
mucic acid dibutyl ester 28 was converted into 26 in near-
quantitative yield, thus confirming the efficiency of the
5[b]
6[c]
7[d,e]
8[e]
9[e]
[a] 3-octanol was used instead of 3-pentanol. [b] 5 mol% catalyst
loading. [c] It was not determined whether benzene was formed by
dehydration of the substrate or by oxidation of 1,3-cyclohexadiene.[7b]
[d] Reaction temperature was 2008C. [e] Reaction time was 1 h.
a fast equilibrium (entries 5 and 6). Furthermore, the DODH
reaction was also applicable to cis,cis-muconic alcohol 16,
affording (E)-hexatriene 17 (entry 7). Although aromatiza-
tion to naphthols was inevitable, a small amount of naphtha-
lene was also obtained from 18 (entry 8), in marked contrast
to the non-conjugated trans-diol 20 that produced no trace of
cyclohexene (entry 9).
On the basis of these observations, a plausible catalytic
cycle[9,17] is shown in Figure 1, depicting erythritol (21) as
a representative substrate. We propose that the 1,4- and 1,6-
DODH reactions proceed through a five-membered ring
reduced rhenium–diolate intermediate before the olefin
extrusion, which is identical to the last step of 1,2-DODH
from vicinal cis-diols. This hypothesis suggests that the
DODH reaction is particularly useful for the total deoxyge-
nation of polyols by merging different intermediates into one
product and thereby increasing product selectivity. Therefore,
the solvent-alcohol-driven DODH system may be exception-
12906 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 12905 –12909

Angewandte
Chemie
Scheme 4. Conversion of d-erythronolactone and d-(+)-ribono-1,4-lac-
tone.
through g-crotonolactone (35) (Scheme 4a). Upon treatment
with NH3,[26] 36 furnishes 2-pyrrolidone, which can be
converted into nylon 4[27] or used as a precursor to N-
methylpyrrolidone (NMP), polyvinylpyrrolidone (PVP), and
polyvinylpolypyrrolidone (PVPP).[28] Additionally, d-
(+)-ribono-1,4-lactone (37), a C5 aldonic acid lactone deriv-
ative, was smoothly converted into 38 without optimization
(Scheme 4b).
Finally, we sought to further exploit the unique capability
of HReO4 to serve as a DODH/acid dual-catalyst by
combining DODH with other reactions. To this end, we
obtained the 4-cyclohexene-1,2-dicarboxylic acid ester 41,
a plausible plasticizer precursor, from l-(+)-tartaric acid (39)
and erythritol (21) by a sequence of tandem DODH,
esterification, and Diels–Alder reaction in a simple one-pot
two-step procedure (70% yield, Scheme 5).[29] The alkene
Scheme 2. DODH of mucic acid and mucic acid dibutyl ester.
DODH step (Scheme 2c). To emphasize the utility of
muconic acid and its ester (24–26) in commodity chemical
synthesis, we also demonstrated the one-pot two-step con-
version of 23 into dibutyl adipate 29 (Scheme 2d). There was
no interference from the Re catalyst during the Pd/C-
catalyzed hydrogenation step.[25]
Encouraged by the efficient DODH of mucic acid to
muconic acid, a potential feedstock for nylon 6-6, we sought
to obtain e-caprolactone from C6 aldonic acid for the
production of nylon 6. As shown in Scheme 3, when d-
Scheme 5. One-pot conversion of l-(+)-tartaric acid and erythritol into
plasticizer precursor 41.
functional group of 41 would serve as a good manipulation
handle to either oxidize/disproportionate to phthalates,[30] or
to hydrogenate to cyclohexane-1,2-dicarboxylic acid esters,[31]
which are both common plasticizer skeletons. This reaction
combines the total deoxygenation of the sugar alcohol into
unsaturated hydrocarbons and the selective deoxygenation of
the sugar acid while preserving the ester group, and exem-
plifies the capability of the DODH reaction to construct
structures with very specific functionality from biomass-
derived polyols. It also underscores the advantage of the
DODH reaction over conventional biomass deoxygenation
methods such as hydrogenation and hydrogenolysis, in that
the product alkenes are deoxygenated but still synthetically
versatile.
Scheme 3. DODH of gluconic acid.
gluconic acid 30 was examined, the attempt to preserve the
terminal OH group by using the secondary alcohol–CH3ReO3
conditions afforded 31a/31b in 50% total yield, with ethers
32a/32b being the minor products. In contrast, by using the
acidic HReO4 catalyst, the products converged to the
(2E,4E)-ether 33. While our efforts on the downstream
conversion of these unsaturated products to nylon 6 are
underway, we also demonstrated the one-pot two-step con-
version of d-erythronolactone (34) to g-butyrolactone (36)
Angew. Chem. Int. Ed. 2013, 52, 12905 –12909
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12907

.
Angewandte
Communications
[13] T. Strassner in Computational Modeling of Homogeneous
Catalysis, Vol. 25 (Eds.: F. Maseras, A. Lledꢂs), Springer, New
York, 2002, pp. 253 – 268.
[14] A. T. Herrmann, T. Saito, C. E. Stivala, J. Tom, A. Zakarian, J.
Am. Chem. Soc. 2010, 132, 5962 – 5963.
In summary, we have found that the oxorhenium-cata-
lyzed DODH reaction using a sacrificial alcohol as a reduc-
tant/solvent is viable not only for vicinal diols (1,2-DODH)
but also for 2-ene-1,4-diols (1,4-DODH) and 2,4-diene-1,6-
diols (1,6-DODH) through a tandem [1,3]-OH shift-DODH
process. These unprecedented reactivity modes substantially
expand the future directions of DODH reaction development
and shed light on the polyol DODH mechanism. The strategy
of using oxorhenium for tandem DODH/acid-catalyzed
reactions was further applied to the conversion of sugar
acids into unsaturated esters. In this context, HReO4 was
identified as a particularly interesting catalyst because it well
mediates not only DODH and the above-mentioned [1,3]-OH
shift, but also other Brønsted acid catalyzed reactions, such as
esterification. The power of this approach was best exempli-
fied in the one-pot synthesis of plasticizer precursors from
tartaric acid and erythritol. We believe that this work
demonstrates the high potential of DODH as a unique and
selective key transformation in the context of biomass-
derived chemicals synthesis.
[15] Conversion was complete in all cases except entry 9. Yields were
determined by 1H NMR spectroscopy using mesitylene as an
internal standard. The 1H NMR spectra and the GCMS spectra/
retention time matched those of authentic samples. Compounds
7 and 8 were also isolated (see the Supporting Information).
[16] a) T. J. Korstanje, E. F. de Waard, J. T. B. H. Jastrzebski, R. J. M.
Klein Gebbink, ACS Catal. 2012, 2, 2173 – 2181; b) T. J. Kor-
stanje, J. T. B. H. Jastrzebski, R. J. M. Klein Gebbink, ChemSu-
sChem 2010, 3, 695 – 697; c) T. J. Korstanje, J. T. B. H. Jastrzeb-
ski, R. J. M. Klein Gebbink, Chem. Eur. J. 2013, 19, 13224 –
13234.
[17] Alhough the Re diolate may exist in equilibrium with its
dinuclear form, only the monomer is shown for simplicity. The
order of diol isomerization, Re diolate formation, and the
reduction of Re has not been clarified, and the catalytic cycle
might be based on ReIII–ReV instead of ReV–ReVII as suggested
by Abu-Omar et al. (Ref. [12b]).
[18] For biocatalytic routes to cis,cis-muconic acid from glucose, see:
a) W. Niu, K. M. Draths, J. W. Frost, Biotechnol. Prog. 2002, 12,
201 – 211; b) K. M. Draths, J. W. Frost, J. Am. Chem. Soc. 1994,
116, 399 – 400.
[19] The concept of synthesizing adipic acid from glucose through
DODH was touched on by Metzger in his minireview on biomass
deoxygenation (Ref. [1b]) while we were preparing this manu-
script, but no specific pathway for such a reaction has been
reported.
Received: August 28, 2013
Revised: October 17, 2013
Published online: November 12, 2013
Keywords: biomass · deoxydehydration · rhenium · sugar acids ·
.
sustainable chemistry
[20] a) D. Schweitzer (Amyris, Inc.), WO2012/82725A1, 2012;
b) J. W. Frost, A. Miermont, D. Schweitzer, V. Bui (Amyris,
Inc.), US 20100314243A1, 2010.
[1] a) S. Dutta, ChemSusChem 2012, 5, 2125 – 2127; b) J. O.
Metzger, ChemCatChem 2013, 5, 680 – 682.
[21] V. Bui, J. W. Frost (Amyris, Inc.), WO2012/141993A1, 2012.
[22] Glucaric acid (aldaric acid derived from glucose) is commer-
cially available in salt form only (Na, Ca, K). These salts were
themselves unreactive toward DODH, possibly because of the
basicity. Glucosaminic acid and N,N’-dibenzyl-d-glucaramide
also did not give any appreciable products. When monopotas-
sium glucarate was preacidifed with cation exchange resin and
subjected to DODH conditions in 1-butanoldibutyl muconate
was obtained in 25% yield (trans,trans/cis,trans = 7:18). See the
Supporting Information for details.
[23] In Schemes 2–5, isolated yields are reported. Although con-
version could not be rigorously determined because of the low
solubility of polyol substrates in organic solvents, the disappear-
ance of the solid starting materials suggested complete con-
version. Product 24 was also insoluble in organic solvent, but
1H NMR analysis of the islated compound in [D6]DMSO
confirmed the complete consumption of 23. See the Supporting
Information.
[24] 23 was not produced in an isolable amount. The monoester was
not observable by TLC.
[25] We were not able to convert 28 into 29 or 26 using H2 as the sole
reductant in the absence of alcohol; see Ref. [7a].
[26] a) Dairen Chemical Corporation; S.-C. Chen, J.-M. Lin, C.-L.
Tsai, F.-S. Lin, L.-A. Hsu, US6350883 B1, 2002; b) F. B.
Minnock, P. D. Taylor (ISP Investments Inc.), US5393888,
1995; c) M. Rudloff, P. Stops, E. Henkes, H. Schmidtke, R.-H.
Fischer, M. Julius, R. Lebkuecher, K.-H. Ross (BASF Aktien-
gesellschaft), WO 2003022811, 2003.
[2] a) S. Stanowski, K. M. Nicholas, R. S. Srivastava, Organometal-
lics 2012, 31, 515 – 518; b) S. Murru, K. M. Nicholas, R. S.
Srivastava, J. Mol. Catal. A 2012, 363 – 364, 460 – 464.
[3] K. M. Nicholas, G. Chapman, Jr., Chem. Commun. 2013, 49,
8199 – 8201.
[4] L. Hills, R. Moyano, F. Montilla, A. Pastor, A. Galindo, E.
ꢀlvarez, F. Marchetti, C. Pettinari, Eur. J. Inorg. Chem. 2013,
3352 – 3361.
[5] E. Arceo, J. A. Ellman, R. G. Bergman, J. Am. Chem. Soc. 2010,
132, 11408 – 11409.
[6] a) G. K. Cook, M. A. Andrews, J. Am. Chem. Soc. 1996, 118,
9448 – 9449; b) S. Raju, J. T. B. H. Jastrzebski, M. Lutz, R. J. M.
Klein Gebbink, ChemSusChem 2013, 6, 1673 – 1680.
[7] a) J. E. Ziegler, M. J. Zdilla, A. J. Evans, M. M. Abu-Omar,
Inorg. Chem. 2009, 48, 9998 – 10000; b) A. L. Denning, H. Dang,
Z. Liu, K. M. Nicholas, F. C. Jentoft, ChemCatChem 2013, DOI:
10.1002/cctc.201300545.
[8] a) P. Liu, K. M. Nicholas, Organometallics 2013, 32, 1821 – 1831;
b) I. Ahmad, G. Chapman, K. M. Nicholas, Organometallics
2011, 30, 2810 – 2818; c) S. Vkuturi, G. Chapman, I. Ahmad,
K. M. Nicholas, Inorg. Chem. 2010, 49, 4744 – 4746.
[9] M. Shiramizu, F. D. Toste, Angew. Chem. 2012, 124, 8206 – 8210;
Angew. Chem. Int. Ed. 2012, 51, 8082 – 8086.
[10] J. Yi, S. Liu, M. M. Abu-Omar, ChemSusChem 2012, 5, 1401 –
1404.
[11] C. Boucher-Jacobs, K. M. Nicholas, ChemSusChem 2013, 6, 597 –
599.
[27] C. E. Barnes, Lenzinger Ber. 1987, 62, 62 – 66.
[12] a) S. Qu, Y. Dang, M. Wen, Z.-X. Wang, Chem. Eur. J. 2013, 19,
3827 – 3832; b) S. Liu, A. Senocak, J. L. Smeltz, L. Yang, B.
Wegenhart, J. Yi, H. I. Kenttꢁmaa, E. A. Ison, M. M. Abu-Omar,
Organometallics 2013, 32, 3210 – 3219.
[28] A. L. Harreus, R. Backes, J. O. Eichler, R. Feuerhake, C. Jꢁkel,
U. Mahn, R. Pinkos, R. Vogelsang, Ullmannꢀs Encyclopedia of
Industrial Chemistry, Wiley-VCH, Weinheim, 2000.
12908 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 12905 –12909

Angewandte
Chemie
[29] Higher yields were obtained using either butadiene instead of
erythritol, or a two-pot two-step procedure to separately convert
39 into 40 and 21 into 2 before combining the two streams for the
Diels–Alder reaction. See the Supporting Information.
[30] a) C. M. Buchan, J. I. G. Cadogan, I. Gosney, B. J. Hamill, S. F.
Newlands, D. A. Whan, J. Chem. Soc. Chem. Commun. 1983,
725 – 726; b) X. Zhang, L. Xu, X. Wang, N. Ma, F. Sun, Chin. J.
Chem. 2012, 30, 1525 – 1530.
[31] M. Brunner, A. Bottcher, B. Breitscheidel, K. Halbritter, J.
Henkelmann, L. Thil, R. Pinkos (BASF Aktiengesellschaft),
US6284917 B1, 2001.
Angew. Chem. Int. Ed. 2013, 52, 12905 –12909
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12909