transformation of biomass carbohydrates by transition metal catalysts
TRANSCRIPT
Purdue UniversityPurdue e-Pubs
Open Access Dissertations Theses and Dissertations
Fall 2014
Transformation of biomass carbohydrates bytransition metal catalystsChristine M BohnPurdue University
Follow this and additional works at: https://docs.lib.purdue.edu/open_access_dissertations
Part of the Biochemistry Commons, and the Inorganic Chemistry Commons
This document has been made available through Purdue e-Pubs, a service of the Purdue University Libraries. Please contact [email protected] foradditional information.
Recommended CitationBohn, Christine M, "Transformation of biomass carbohydrates by transition metal catalysts" (2014). Open Access Dissertations. 233.https://docs.lib.purdue.edu/open_access_dissertations/233
Graduate School ETD Form 9 (Revised 12/07)
PURDUE UNIVERSITY GRADUATE SCHOOL
Thesis/Dissertation Acceptance
This is to certify that the thesis/dissertation prepared
By
Entitled
For the degree of
Is approved by the final examining committee:
Chair
To the best of my knowledge and as understood by the student in the Research Integrity and Copyright Disclaimer (Graduate School Form 20), this thesis/dissertation adheres to the provisions of Purdue University’s “Policy on Integrity in Research” and the use of copyrighted material.
Approved by Major Professor(s): ____________________________________
____________________________________
Approved by: Head of the Graduate Program Date
Christine M. Bohn
TRANSFORMATION OF BIOMASS CARBOHYDRATES BYTRANSITION METAL CATALYSTS
Doctor of Philosophy
Mahdi Abu-Omar
Garth J. Simpson
Nathan Mosier
Tong Ren
Mahdi Abu-Omar
R. E. Wild 09/05/2014
TRANSFORMATION OF BIOMASS CARBOHYDRATES BY
TRANSITION METAL CATALYSTS
A Dissertation
Submitted to the Faculty
of
Purdue University
by
Christine M. Bohn
In Partial Fulfillment of the
Requirements for the Degree
of
Doctor of Philosophy
December 2014
Purdue University
West Lafayette, Indiana
ii
I dedicate this dissertation in loving memory of Mark Bohn, Louis
Gruttadauria, and Sophie Ciuni.
iii
ACKNOWLEDGMENTS
There are many people I need to thank because without them, this
dissertation and my journey to become a scientist would not be possible.
First, thank you to my research advisor, Mahdi, thank you for your
guidance and support, and for teaching me to become an independent
researcher. I look forward to seeing where this PhD in Chemistry takes me. The
Abu-Omar group: thank you to Jeanette, Nick, Mike, Erin, Eurick, Mary, Judy,
and Jennifer who all welcomed me into the lab all those years ago and taught me
the ways of inorganic chemistry. Thank you to the current and recently graduated
group members for your support and friendship: Basu, Trenton, Ram, Bobby,
Curt, Isaac, TG, Scott, Yuan, Ian, Ben, Shuo, Mike, Paul, Andrew, Manasa, Keith,
Jing, Hao, Shou, and Kelsey.
My committee, Dr. Ren, Dr. Simpson, and Dr. Mosier, thank you for your
support and guidance.
I also need to give a big thanks to my family. Without their love and
guidance I would not be the woman I am today. Thanks to my Nana, Aunt
Chrissy, Uncle Paul and Aunt Chris, and my brother Tommy; your support and
love have meant the world to me. An extra special thank you needs to go to my
mom, Louise, who gave me the tools I needed to go into the world and find
iv
myself and taught me to not give up when I got lost. I want to thank you for
everything that you provided and sacrificed for me.
I want to thank my STERIS co-workers, Iain, Eric, and Rick, thank you for
being my fan club and for making my first job one of the best experiences. Thank
you for encouraging me to pursue my PhD.
To my friends, my other family, I don’t think I would have made it all the
way to the end without you support. Thank you for listening and not getting angry
when I didn’t call enough. My JCU crew: Kim, Lisa, Dana, Karen, Christine,
Roberta, Nicole, and Megan. The amazing new friends I made here in Lafayette,
your day to day support has been invaluable. Gabby, Tina and Rich, you have
become some of our closest friends and your support and generosity, especially
during this last year, is something I will always be grateful for. To those who
traveled the PhD road with me, Amanda and Joel, Nathan and Autumn, Frank,
and Sr. Mary Simon, I am truly blessed to have your friendship and to have
shared this trying and crazy part of our lives together.
Finally I need to thank my husband, Christopher Dettmar. I don’t know
what I would have done these last few years without your love and support. You
have stood by me in my weakest times and helped me find the strength to
persevere. I cannot wait to go out and tackle the world together as Dr. and Dr.
L
L
L
A
C
C C
IST OF TA
IST OF FIG
IST OF AB
ABSTRACT
CHAPTER 1
1.11.2
1.311
1.41.5
1.6
CHAPTER 2CATALYST
2.12.2
22
2.32.4
222
2.5
ABLES .......
GURES .....
BBREVIATIO
T ................
1.INTRODU
Renewa1 Plant Su2
Hemic Dehydra3
1.3.1 The R1.3.2 Dehyd Highly V4 C3Bio: T5
Bioma Disserta6
2.A REVIEWTS UPGRAD
Lewis a1 Reactio2
2.2.1 Isome2.2.2 Humin Solvent3 Glucose4
2.4.1 Chrom2.4.2 Alumin2.4.3 Iron a Propose5
TABL
.................
.................
ONS .........
.................
UCTION ....
able Carbonupport Strucellulose ...ation Chem
Role of Metadration of HValuable OrThe Centerass to Biofuation Overv
W OF HOMDING HEXO
and Brønsteon Pathwayserization ....ns and Poly Systems ..e Transformmium Catalynum Catalynd Other Ced Reaction
LE OF CON
.................
.................
.................
.................
.................
n Sources: uctures: Lign.................
mistry .........al Salts in D
Hexoses andrganics fromr for the Diruels ...........view ...........
MOGENEOUOSES AND
ed Acid Dehs ................................ymerization.................
mations .....ysis ...........ysis ...........
Catalysts ....n Mechanis
NTENTS
.................
.................
.................
.................
.................
Plants vs. nin, Cellulo..................................
Dehydrationd Pentosesm Biomassrect Catalyt..................................
US TRANSD PENTOS
hydration C..................................
n .....................................................................................................sms ...........
.................
.................
.................
.................
.................
Hydrocarboose, and ..................................n Reactionss ................s ................tic Convers..................................
SITION METES ............
Catalysis .............................................................................................................................................................
P
.................
.................
.................
.................
.................
ons ...........
.................
.................s ..................................................
sion of ..................................
TAL .................
.................
.................
.................
.................
.................
.................
.................
.................
.................
.................
v
Page
... viii
..... ix
..... xi
.... xii
..... 1
..... 1
..... 2
..... 5
..... 5
..... 6
..... 7
..... 9
..... 9
... 11
... 11
... 12
... 12
... 13
... 13
... 15
... 15
... 16
... 17
... 19
CDC
CW
CHAPTER 3DEHYDRATCARBOHYD
3.13
3.23333
3.333
333
3.43.5
CHAPTER 4WALL WITH
4.14444
4.244444
4.3444
4.44
4.5
3.IRON (III)TION CATADRATES ...
Introduc13.1.1 Microw Method2
3.2.1 Mater3.2.2 Dehyd3.2.3 Analys3.2.4 Calcu Results3
3.3.1 Surve3.3.2 Maxim
Irradia3.3.3 Using 3.3.4 Comp3.3.5 Dehyd Discuss4 Conclus5
4.PROBINGH FLUORES
Introduc14.1.1 The P4.1.2 Utilizin4.1.3 Fluore4.1.4 Click C Method2
4.2.1 Mater4.2.2 Click R4.2.3 Probe4.2.4 Modifi4.2.5 Feedin Results3
4.3.1 Fluore4.3.2 Fluore4.3.3 Fluore Discuss4
4.4.1 Impro Conclus5
) CHLORIDALYST FOR.................
ction ..........wave Irradis ...............ials ...........dration Reasis ............lations ...... and Discusy of Transit
mizing the Dation ..........Brønsted A
paring Iron tdration of Osion ...........sion ...........
G BIOMASSSCENT TA
ction ..........Plant Cell Wng Xyloglucescence ....Chemistry .s ...............ials ...........Reactions a
e Synthesis ied Sugar Sng Plant Pr ................
escence of escence of escent Signsion ...........vements ansions .........
DE HEXAHYR THE UPG.................
.................ation ...........................................
actions ........................................ssion ........tion Metal S
Dehydration.................Acids as a Cto Aluminum
Other Subst..................................
S: FOLLOWGGING US
.................Wall ............can ................................................................................and fluores.................
Synthesis ..rotoplast Ce.................Alkyne ProAzido Suga
nal from Co.................nd Changes.................
YDRATE AGRADE OF .................
.................
.................
.................
.................
.................
.................
.................
.................Salt Activitien Reaction w.................Co-Catalysm and Chrotrates ..........................................
WING FUCOSING CLICK
.................
.................
.................
.................
.................
.................
.................cent detect..................................ells .............................
obe ............ar Fed Planntrols .........................s ................................
AS A SELECBIOMASS.................
.................
.................
.................
.................
.................
.................
.................
.................es .............with Microw.................
st for Iron ...omium Salt ...................................................
OSE INTO K CHEMIST
.................
.................
.................
.................
.................
.................
.................tion ................................................................................................nt Cells ..........................................................................
P
CTIVE .................
.................
.................
.................
.................
.................
.................
.................
.................
.................wave ..................................Catalysts ....................................................
THE CELLTRY ..........
.................
.................
.................
.................
.................
.................
.................
.................
.................
.................
.................
.................
.................
.................
.................
.................
.................
.................
vi
Page
... 22
... 22
... 23
... 26
... 26
... 27
... 28
... 29
... 29
... 29
... 32
... 35
... 37
... 38
... 42
... 47
L ... 48
... 48
... 48
... 48
... 49
... 50
... 52
... 52
... 52
... 53
... 53
... 54
... 55
... 56
... 60
... 64
... 66
... 67
... 68
vii
Page
LIST OF REFERENCES .................................................................................... 69
VITA ................................................................................................................... 75
PUBLICATION .................................................................................................... 76
viii
LIST OF TABLES
Table ............................................................................................................. Page
Table 2-1: Survey of Reactions in the Literature ................................................. 18
Table 3-1: Different Transition Metal Salt Catalysts for Glucose Dehydration .... 30
Table 3-2: Changing the Concentration of Molybdenum Catalyzed Dehydration Reactions ................................................................................................. 31
Table 3-3: Changing Iron Concentrations for Dehydration Reactions ................. 31
Table 3-4: The Best Production of HMF and Levulinic Acid from Glucose ......... 34
Table 3-5: Using Brønsted Acid as a Co-catalyst with Iron ................................. 35
Table 3-6: Different Type of Lewis-Brønsted Acid Systems ............................... 36
Table 3-7: Comparing Single Brønsted Acid Dehydration and Combined for Fe (III) and HBr Acid Systems ....................................................................... 37
Table 3-8: Dehydration of Different Model Compounds ...................................... 39
Table 3-9: Physical Properties for All of the Catalyst Ions Used in this Study .... 45
Table 4-1: Setup of Feeding Reactions .............................................................. 54
Table 4-2: Summary of Maximum Intensities for Probe and Click Reactions ..... 56
Table 4-3: Maximum Intensities for Azido Tagged and Untagged Sugars at Different Concentrations and in Different Solvents .................................. 60
Table 4-4: Summary of Maximum Intensities, Wavelengths, and Tagged/ Untagged Ratios for Click Reactions in Plant Fractions ........................... 63
Table 4-5: Summary of Maximum Intensities for Click Reactions with Control, Tosyl Azide .............................................................................................. 65
ix
LIST OF FIGURES
Figure ............................................................................................................. Page
Figure 1-1: The Repeating Monomers for Cellulose and Hemicellulose Xylan Chains ............................................................................................ 3
Figure 1-2: The Linear and Cyclical Structures for the Three Main Sugars Derived from Biomass .............................................................................. 4
Figure 1-3: The Aldose-Ketose Isomerization Mechanism ................................... 6
Figure 1-4: The Three Main Products Derived from Dehydration Reactions with Glucose ............................................................................ 7
Figure 3-1: Time Profile for 0.1M FeCl3 System Heated to 160°C (100 W) ....... 32
Figure 3-2: Time Profile for 0.1M FeCl3 System Heated to 160°C (200 W) ........ 33
Figure 3-3: Time Profile for 0.1M Fecl3 System Heated to 200°C (200 W) ........ 34
Figure 3-4: Comparing Glucose Transformation for Three Transition Metal Salts .............................................................................................. 38
Figure 3-5: Dehydration of Biomass Glucans using Iron .................................... 40
Figure 3-6: Dehydration of Biomass Glucans using Aluminum ........................... 41
Figure 3-7: Dehydration of Biomass Glucans using Chromium .......................... 42
Figure 4-1: Mechanism for Huisgen 1,3-Dipolar Cycloaddition ........................... 51
Figure 4-2: 600nm OD of Feeding-2 Cultured Arabidopsis Cells ........................ 55
Figure 4-3: Emission Spectrum for 200M Naphthalimide Probe in 56%DMSO Excited at 357nm .................................................................................... 57
Figure 4-4: Emission Spectrum for 100M Naphthalimide Probe in 56%DMSO Excited at 357nm .................................................................................... 57
x
Figure Page
Figure 4-5: Emission Spectrum for 200M of Azide Tagged and Untagged Sugar Clicked in 56%DMSO and Excited at 357nm ............................... 58
Figure 4-6: Emission Spectrum for 100M of Azide Tagged and Untagged Sugar Clicked in 56%DMSO and Excited at 357nm ............................... 59
Figure 4-7: Emission Spectrum for Click Reaction in Media (100M) Fractions Excited at 357nm .................................................................................... 61
Figure 4-8: Emission Spectrum for Protoplast Fractions (200M) Excited at 357nm .................................................................................... 62
Figure 4-9: Emission Spectrum for Wall Fractions (100M) Excited at 357nm .................................................................................... 63
Figure 4-10: Emission Spectrum for Reactions in Plant Feeding Media and Reactions with Feeding Experiment Blank Samples Excited at 357nm .................................................................................... 64
Figure 4-11: Emission Spectrum for 200M Tosyl Azide Clicked in 56%DMSO Excited at 357nm ................................................................. 65
Figure 4-12: The Difference in Fluorescence Intensity Achieved by Sawa et al. ........................................................................................ 66
xi
LIST OF ABBREVIATIONS
HMF 5-(Hydroxymethyl)-2-furfuraldehyde
LA Levulinic Acid
THF Tetrahydrofuran
MeTHF 2-methyltetrohydrofuran
C3Bio Center for the Direct Catalytic Conversion of Biomass to Biofuels
[EMIM]Cl 1-ethyl-3-methylimidazolium chloride
eV electron volt. 1eV = 1.6×10−19 joule
IR infrared
UV ultraviolet
WT wild type
4ENP 4-ethynyl-N-ethyl-1,8-naphthalimide aAF 6-azido-1,2,3,4-tetra-O-acetyl-6-deoxy-a,b-L-
galactopyranose
xii
ABSTRACT
Bohn, Christine M. Ph.D., Purdue University, December 2014. Transformation of Biomass Carbohydrates by Transition Metal Catalysts. Major Professor: Mahdi Abu-Omar.
By selectively removing functional groups from biomass derived
carbohydrates, valuable platform chemicals can be generated from renewable
sources. Through dehydration chemistry glucose can be upgraded into 5-
(Hydroxymethyl)-2-furfuraldehyde (HMF) and levulinic acid. Iron (III) chloride
hexahydrate has shown moderate activity to transform glucose into HMF and has
also shown high yields and selectivity for the production of levulinic acid.
Typically synthesized from acidic solutions made with mineral acids, levulinic
acid has now been produced in high yields with a metal salt. The difference
between maximizing production for HMF or levulinic acid from the same catalyst
relies on the control of the reaction conditions. By using microwave irradiation,
improved collisions and stabilized transition states allow for selective production
of desired products while eliminating undesired reaction pathways.
Understanding the improvement of biomass carbohydrates also requires
understanding how they are incorporated into the plant structures. By utilizing
fluorescent tagging strategies a two part marking system was used to follow the
incorporation of fucose into the plant cell wall. Fucose is limited in its use by
xiii
plants only becoming incorporated into branched xylan chains that help to link
cellulose and hemicellulose together. The synthesis of xylan in the plant Golgi
also utilizes extracellular sugar. By feeding plant cells a specially designed azido
tagged sugar it should become incorporated into the cell wall. On its own, the
azido tagged sugar has very little fluorescent properties; however the attached
azide group becomes important for click reactions. Click reactions involve two
small functional groups that selectively combine. In this case the azide is clicked
to an alkyne group to form a triazole. The alkyne group is part of a naphthalimide
compound that has strong fluorescent properties due to its aromatic structure. By
adding a triazole to the fluorescent naphthalimide complex, a much stronger
fluorescent signal can be generated. By reproducing this strategy with plant cells
it was hoped that the strong fluorescent signals could be imaged after their
incorporation into plants. However issues with reproducibility of fluorescent
emission spectrum and only two-fold intensity differences rather than the desired
order of magnitude difference proved to be experimentally challenging.
s
o
3
p
c
E
re
n
in
a
y
sy
c
b
g
c
1.
Due t
earching fo
f the contin
0 years a s
art of indus
arbon cycle
Energy, less
enewable s
uclear sour
ncluded pet
bout altern
ears, fossil
ynthesis of
Crude
oal are use
y a refinery
as oil, and
rude oil com
Renew1
to our energ
or a sustain
nual decrea
significant a
stry.1 By us
e can be clo
s than 10%
sources like
rces. Fossil
troleum, na
ative energ
fuels are s
f polymers a
e oil gets us
ed mostly fo
y it is separ
residues. T
mpositions.
CHAPTE
wable Carb
gy and con
able and ca
se in crude
amount of b
ing plant m
osed.2 In a
of our prim
e biomass o
l fuels mad
atural gas, a
gy sources a
still our prim
and plastics
sed for ene
or heat and
ated into 5
The amount
. The napht
ER 1. INTRO
bon Sources
sumption n
arbon dioxid
e oil, van Ha
bio-based c
material as a
2009 repor
mary energy
or solar sou
e up 83% o
and coal so
and reducin
mary source
s.1
ergy and ch
electricity n
major fract
ts of these
tha fraction
ODUCTION
s: Plants vs
needs, rese
de-neutral
averen et a
hemicals w
a replaceme
rt from the
y consumpt
urces; and o
or our prima
urces.3 Des
ng consum
es for not on
hemical nee
needs. Whe
tions, naph
fractions w
is used as
N
s. Hydrocar
earchers ha
energy sou
al. predicted
will become
ent for crud
U.S. Depar
tion was pro
only 9% cam
ary energy
spite a rise
ption over t
nly energy
eds while na
en crude oi
tha, gasolin
will vary from
s a feedstoc
rbons
ve been
urce. Becau
d that in 20
a significan
de oil, the
rtment of
ovided from
me from
sources; th
in concern
the last five
but industri
atural gas a
il is process
ne, kerosen
m different
ck for the
1
use
to
nt
m
his
n
e
ial
and
sed
ne,
p
T
to
fe
c
m
s
m
h
p
n
2
b
p
p
c
m
B
roduction o
The six platf
oluene, and
eedstock ch
arbon and
material for
electively fu
Plant
most of their
emicellulos
er year and
on-food ne
0% is oils a
ut some N
ure carbon
resent. By
hemicals th
1.2
Plant
most abunda
Biomass is t
of platform c
form chemi
d xylene. An
hemicals al
hydrogen. T
polymers li
unctionalize
material or
r carbon str
se, and lign
d only 3% o
eds. 70-75
and fatty ac
and S. A d
and hydro
using react
he petroche
Plant Sup
material is
ant and the
the only ren
chemicals w
cals include
nd unlike bi
l have a ve
These six c
ke ethylene
ed with oxy
r biomass,
ructures in
in.2 The tot
of that is be
% of bioma
cids, and 5%
rop-in strat
gen source
tions to sele
emical indus
port Structu
an exclusi
e must unde
newable ca
which go on
e ethylene,
iomass the
ery low oxyg
chemicals c
e, propylene
ygen, nitrog
such as tre
the structur
tal biomass
eing used fo
ass is lignoc
% are prote
egy can be
e, the desira
ectively rem
stry already
ures: Lignin
ve carbon s
erutilized re
rbon resou
n to make a
propylene
naphtha a
gen content
can then for
e and butad
en, or chlo
ees, grasse
ral polymer
s in the wor
or human us
cellulose an
eins and co
e used even
able chemic
move undes
y know how
n, Cellulose
source from
esource for
rce. Plants
all of our bu
, C4-olefins
nd the subs
t and are p
rm solvents
diene; or be
ride.1
s, and crop
rs, cellulose
rld is 17,000
ses like foo
nd structura
ntain not on
n though bio
cal structur
sired functio
w to use can
e, and Hem
m nature.4
our carbon
use a syst
ulk chemica
s, benzene,
sequent
redominant
s; starting
ecome
p leftovers s
e,
0 million ton
od and othe
al material,
nly C, H, an
omass is no
res are alre
onalities, b
n be made.
icellulose
Plants are
n needs.5
tem of polym
2
als.
,
tly
store
ns
er
15-
nd O
ot a
ady
ulk
.1
the
mer
c
a
m
g
a
FC
ti
S
g
s
m
e
in
c
hains made
re lignin, ce
Lignin
made up of
lucose, xylo
nd 40-50%
Figure 1-1: TChains
Prima
mberlands,
Secondary r
eneration b
ugars that c
materials ca
xisting eng
nfrastructur
ommerciall
e up of carb
ellulose, an
n is a poly-p
repeating c
ose and ara
% of dry weig
The Repea
ary biomass
, as well as
resources in
biofuels inc
can be turn
an be blend
ines, or can
e. In 2010,
y produced
bohydrates
nd hemicellu
phenolic ma
carbohydrat
abinose. He
ght biomas
ting Monom
s resources
s agriculture
nclude resid
lude fuels f
ned into bio
ed with pet
n be distrib
50 billion li
d.5 Second
to give the
ulose.
aterial and
tes. The mo
emicellulos
s is cellulos
mers for Ce
s include fo
e resources
dues and w
from materi
diesel or bi
troleum-bas
uted throug
iters of first
generation
em structure
cellulose a
ost abunda
se is 23-35%
se.6,7,8
ellulose and
orest resour
s like grain c
waste from
als generat
io-ethanol.
sed fuels, c
ghout the c
t generation
biofuels ar
e. These na
and hemice
ant sugars i
% of dry we
d Hemicellu
rces, like fu
crops or gr
crops or an
ted from ea
These plan
combust in
urrent indu
n biofuels w
re generate
atural polym
llulose are
n biomass
eight bioma
lose Xylan
uel wood or
asses.
nimals.3 Fir
asily extract
nt based
currently
strial
were being
ed from
3
mers
are
ass
rst
table
lig
fo
is
FD
in
in
a
fu
re
tr
is
b
c
a
a
b
gnocellulos
ood plant m
s not cost e
Figure 1-2: TDerived from
The p
ndustrially t
ndustry gen
s dispersin
uel.10 If prop
efinery coul
ransformati
s to use the
een focusin
arbon, they
romatic pro
nd the rele
iomass ferm
sic materials
material. How
ffective.5 R
The Linear m Biomass
processing
here is little
nerated ove
g or binding
perly plann
ld burn the
on of cellul
e valuable p
ng on this p
y have succ
oducts.11 Ce
ased carbo
mentation p
s, the majo
wever, righ
Recovering 8
and Cyclica
of cellulose
e value to th
er 50 million
g agents, b
ed, isolated
lignin prod
ose and he
polyphenolic
problem, us
cessfully up
ellulose and
ohydrates c
process.12
ority being o
ht now, prod
80% of pro
al Structure
e into paper
his materia
n tons of lig
but due to it
d lignin gen
uced and p
emicellulose
c structures
sing a mixed
pgraded the
d Hemicellu
an easily b
of inexpens
duction of s
ducts can g
es for the T
r or pulp lea
l. In 2004 th
nin. Some
ts low value
nerated from
provide an e
e sugars int
s. The Abu-
d catalyst s
e lignin stru
ulos are mo
be used to m
ive and abu
second gen
give a low u
Three Main
aves behin
he pulp and
of that mat
e it is prima
m an integra
excess of e
to ethanol.2
-Omar rese
system of Z
ctures to va
ore easily d
make ethan
undant non
eration biof
unit cost.9
Sugars
d lignin;
d paper
erial was u
arily burned
ated bio-
energy for th
2 An alterna
earch group
Zn and Pd o
aluable
depolymeriz
nol through
4
n-
fuels
sed
as
he
ative
p has
on
zed
a
a
m
a
p
d
s
fr
in
c
c
lin
b
d
it
th
th
Typic
cid. Until re
medium but
luminum, ti
icture when
ehydration
Direc
accharifica
rom aldoses
nto HMF.14
omposed o
ellulose, is
nked xylose
reak apart
ehydrate th
self must b
hat have sh
hree are av
1.3.1 The
cal dehydra
ecently (200
Fringuelli a
itanium and
n selectivel
of biomass
ct transform
tion or hydr
s into ketos
The plant c
of non-reduc
made up o
e.15 A succe
biomass po
he alcohol f
e effective
hown some
vailable as w
D1.3
e Role of M
tion catalys
01) Lewis a
and cowork
d tin chlorid
y improving
s must go b
ation of bio
rolysis of ce
ses, and fin
cell wall is m
cing glucos
of chains of
essful carb
olymers by
functional g
but still ine
promise in
water solub
Dehydration
Metal Salts i
sts are mine
acids were b
kers catalyz
es.13 It is im
g the value
beyond the
omass into
ellulose into
ally dehydr
made up of
se. The hem
glucosyl an
ohydrate d
hydrolyzing
groups of in
expensive to
the literatu
ble salts and
n Chemistry
in Dehydrat
eral acids li
believed to
zed organic
mportant to
of plant ca
dehydratio
HMF requir
o monosacc
ration of ke
f several typ
micellulose,
nd xylosyl.
ehydration
g the glyco
dividual ca
o use. Ther
ure: Alumin
d cost less
y
tion Reactio
ike sulfuric
be unusab
c reactions i
o consider th
arbohydrate
n of fructos
res 3 steps
charides, is
tose or fura
pes of suga
, crosslinkin
Xylan is co
catalyst m
sidic linkag
rbohydrate
re are seve
um, Iron an
than a doll
ons
or phospho
ble in aqueo
in water us
he whole
es. The
se.
: first a
somerizatio
anose suga
ars. Cellulos
ng the
omposed of
ust be able
ges and
es. The cata
eral metal sa
nd Zinc. All
ar per gram
5
oric
ous
ing
on
ars
se is
f
e to
alyst
alts
m.
a
p
im
b
H
h
d
H
F
F
is
th
e
The m
n undesirab
oor leaving
mprovemen
een shown
HMF was fir
eterogeneo
From
oes not eas
HMF from g
Figure 1-3: T
Figure 1-3 il
somerizatio
In the
he 5 carbon
asily dehyd
1.3.2
many hydro
ble platform
g ability. Pro
nt of hydrox
n to facilitate
rst reported
ous acid ca
fructose, c
sily isomeri
lucose hav
The Aldose
lustrates th
n. More me
e presence
n levulinic a
drate to ma
2 Dehydra
oxyl groups
m chemical
otonation (-
xides as a le
e isomeriza
from react
talysts in a
carbohydrat
ize into fruc
e been pro
e-Ketose Iso
his aldose-k
echanistic d
of water, H
acid and the
ke furfural.
ation of Hex
that are pa
and remov
-OH2+) by B
eaving grou
ation of gluc
tions with o
queous me
tes can eas
ctose witho
posed but a
omerization
ketose isom
details will b
HMF can un
e one carbo
xoses and P
art of glucos
ving them ca
Brønsted ac
up.16 The us
cose into fru
xalic acid a
edia.12
sily dehydra
ut a catalys
are not faci
n Mechanis
merization th
be covered
ndergo a re
on formic ac
Pentoses
se and xylo
an be diffic
cid results in
se of Lewis
uctose. In t
as well as h
ate to form
st. Direct pa
ile reaction
sm
hrough an e
in chapter
hydration re
cid. Pentos
ose make th
cult due to t
n the
s acids have
the 19th cen
homo and
HMF. Gluc
athways to
s.
enolate
two.
eaction to f
ses, like xylo
6
hem
heir
e
ntury
cose
form
ose,
th
th
b
1
fe
FG
s
fo
h
H
re
h
re
O
Fruct
he price of a
he Pacific N
uilding bloc
2 building b
ermentation
Figure 1-4: TGlucose
Furfu
olvent synt
or generatin
eating fuels
HMF can un
esults in 2,5
ydroxyl and
enewable s
Oxidation of
1.4
ose easily d
a fructose o
Northwest N
cks that will
block comp
n and dehyd
The Three
ral, which i
hesis like th
ng valuable
s,20 and has
ndergo oxid
5-diformylfu
d formyl gro
starting mon
f the furan r
Highly Va
dehydrates
one.17 The
National Lab
l easily und
pounds, 8 o
dration reac
Main Produ
s made fro
hat of THF
e organic co
s uses in b
ation react
uran, an im
oup gives 2
nomer for te
ring and or
aluable Org
s into HMF,
National R
boratory ha
dergo transf
f them can
ctions.18
ucts Derived
m xylose, c
and MeTH
ompounds.1
iofuel, bioc
ions where
portant indu
2,5-furandic
erephthalic
the formal
anics from
however a
Renewable
ave reported
formation in
be derived
d from Deh
can be used
F. HMF is a
19 It can be
hemical, an
the oxidati
ustry mono
carboxylic a
acid in pla
group can
Biomass
a glucose fe
Energy Lab
d on 12 sug
nto biochem
d from gluco
hydration R
d as a prec
a good plat
e a good pr
nd biopolym
ion of the h
omer, oxida
acid which c
stics and p
yield
eedstock is
boratory an
gars derive
micals. Of th
ose7 throug
eactions w
cursor for
tform chem
recursor to
mer industri
hydroxyl gro
tion of both
could be a b
olyester.17,
7
half
nd
ed
hese
gh
ith
ical
high
ies.
oup
h the
bio-
12
8
2,5-bis(hydroxymethyl)furan which is a chemical building block for
polymers and polyurethane foams. Reduction of the formyl and hydroxyl groups
gives 2,5-dimethylfuran which has biofuel potential. HMF can also undergo
halogen substitutions giving 5-halomethylfurfurals which make good synthesis
intermediates due to their high reactivity. Finally hydrolysis of the HMF ring in
acidic conditions forms levulinic and formic acid and can follow a polymer path
through a 2,3 water addition or a 4,5 addition of water forming
2,5-dioxo-3-hexenal which then fragments into levulinic and formic acid.12
Levulinic acid is an important platform molecule due to its reactivity.
Levulinic acid is often produced from an acid catalysis which will also produce
large amounts of humins or humic acids, black insoluble materials which are
made from unwanted polymerization reactions. Direct esterification of levulinic
acids with alcohols can make ester compounds that will improve the flow of
biodiesel at low temperatures or as oxygen additives for fuels. Reduction of
levulinic acid results in the formation of gamma-valerolactone, which can be used
for perfumes, food additives, solvent precursors, as well as a fuel additive similar
to ethanol. Reduction can also result in the production of methyltetrahydrofuran.
MeTHF can be blended with fuel up to 60% (v/v) and can be used in current
combustion engines. It can also be used as a green solvent.21 Levulinic acid can
be used to make acrylate polymers as well as being used as a fuel additive.17
When discussing valuable organics from biomass sources it is important
to keep in mind the three paths chemical synthesis can take. The first strategic
route is the Drop-In where bio-derived molecules replace an intermediate
c
s
c
g
in
R
b
a
fo
b
c
c
p
u
m
in
h
urrently obt
econd is Em
onnected to
enerate tar
C31.5
This w
nterdisciplin
Research C
iomass into
pproaches
ocuses on t
iological an
atalysts, an
hapter thre
art of the s
The n
seful carbo
more and m
nsight into t
exoses and
tained from
merging wh
o petrochem
rget chemic
3Bio: The C
work is part
nary group
enter. The
o hydrocarb
. The first in
the assemb
nd genetic e
nd the fourt
e falls unde
econd appr
need to inex
on and hydr
ore of our f
the possibil
d pentoses
m crude oil a
here a bran
micals is us
cals that alr
Center for th
t of the Pur
is a U.S. De
research fo
bon-rich bio
nvolves cat
bly and dec
engineering
h studies fa
er the first a
roach.
D1.6
xpensively
rogen sourc
finite fossil
ities of hom
. Chapter tw
and proceed
d new route
sed. Finally
ready exist.
he Direct CBiofuels
rdue resear
epartment o
ocuses on t
ofuels by us
talytic conve
onstruction
g of biomas
ast-hydropy
approach a
Dissertation
and easily
ces become
fuel resourc
mogeneous
wo will revie
d with exist
e with a pro
y Substitutin
22
Catalytic Cos
rch center,
of Energy f
the convers
sing four dis
ersion of bi
n of cellulos
ss to alter a
yrolisis of b
nd the rese
n Overview
transform b
es more im
ces. This d
catalysis fo
ew the liter
ting techno
oduct that i
ng uses a n
onversion of
C3Bio. This
funded Ene
sion of plan
stinct resea
iomass, the
se, the third
assembly or
iomass. Th
earch in cha
w
biomass ma
mportant as
issertation
or the trans
rature and p
ology.The
s not
new pathwa
f Biomass t
s
ergy Frontie
nt lignocellu
arch
e second
d uses
r to incorpo
he research
apter four is
aterials into
we use up
provides
sformation o
provide an
9
ay to
to
er
losic
orate
h of
s
o
of
10
outline of the previous work with homogeneous transition metal catalysis and the
proposed mechanisms for Brønsted and Lewis acid catalysis. It will highlight
some of the breakthroughs in this area as well as some of the limitations of the
reaction systems. Chapter three will report on the success of using Fe(III)
Chloride as a homogeneous dehydration catalyst for hexose and pentose
transformation into HMF, levulinic acid and Furfural. Finally chapter four will take
a different direction in understanding the transformation of biomass.
Investigations in fluorescently tagged sugars that are followed into the cell wall
will be detailed. The challenges in using of small amounts of fluorescent tags and
probes for detection will also be explained.
e
B
T
b
st
a
c
m
M
c
a
S
a
d
CHAPT
Lewis
lectron pair
Brønsted ba
The dehydra
Brøns
etter leavin
tep, followe
cids have b
oncentratio
mineral acid
Maleic acid
an dehydra
cids indust
Special equi
The a
ctive metal
ehydration
TER 2. A RCATALY
Le2.1
s acids acce
r. Brønsted
ases gain or
ation reactio
sted acids p
ng group. T
ed by more
been used c
ons use dilu
ds. But not j
can easily d
ate into HM
rial concern
ipment can
acidic pH of
that can fa
. In the abs
REVIEW OFYSTS UPGR
wis and Brø
ept lone ele
acids lose
r accept the
on requires
protonate th
he first wat
favored re
commercia
ute acid of 1
ust strong a
dehydrate x
F.24 It also
ns arise abo
be used fo
f Lewis acid
acilitate the
sence of Le
F HOMOGERADING HE
ønsted Acid
ectron pairs
or donate
e H+. Water
s the use of
he hydroxy
er removed
moval for e
ally to synth
1-10% whic
acids like H
xylose and
o must be co
out storage
or acid cata
ds is respon
ring openin
wis acids, n
ENEOUS TEXOSES A
d Dehydrat
s and Lewis
a hydrogen
r is both a B
f Brønsted a
yl groups to
d by dehydr
each subseq
esize HMF
ch equates
HCl and H2S
with a co-c
onsidered t
e and corros
alyst but tha
nsible for fo
ng of gluco
no fructose
TRANSITIOAND PENTO
tion Catalys
s bases don
n cation or p
Brønsted a
and Lewis A
OH2+ ma
ration is a u
quent wate
F. Typical in
to 0.1 to 1M
SO4 dehydr
catalyst like
that when u
sion of equ
at can be ex
orming a ca
se and the
e is observe
ON METALOSES
sis
nate the
proton and
cid and bas
Acids.
king them a
unfavorable
er.23 Brønste
ndustrial
M concentra
rate glucos
e AlCl3, gluc
using miner
uipment.
xpensive.7
atalytically
following
ed from gluc
11
se16
a
e
ed
ated
e.
cose
ral
cose
in
g
ir
b
th
tr
k
a
C
th
p
re
re
c
p
o
n solution. H
roups of glu
ron (III) are
orderline h
Gluco
he metal of
ransition sta
etose struc
nd hemicel
Care must b
heir hydroxi
recipitation
eaction is in
Use o
egeneration
atalyst can
roducts like
The is
ne.27 Fructo
Hard Lewis
ucose whic
known to b
ard/soft Lew
opyranose
the metal s
ate of the ri
ctures. Lew
llulose.26 T
be taken tha
ide species
n. Formation
n an acidic
of homogen
n. For the g
be particul
e insoluble
somerizatio
ose is muc
acids can
ch are them
be hard Lew
wis acids.
must ring o
salts will ac
ng opened
is acids can
They can als
at metal com
s which can
n of hydrox
environme
neous meta
glucose deh
larly suscep
polymers a
2.2
2.2
on process
h more acti
interact stro
mselves hard
wis acids w
open to isom
ct as a Lewi
sugar as it
n also cleav
so coordina
mplexes do
n result in lo
ides can be
nt.
al solutions
hydration re
ptible to the
and humins
Reaction P
2.1 Isomer
between fr
ive for dehy
ongly with o
d Lewis bas
hile iron (II)
merize to fr
is acid meta
t isomerizes
ve glycosid
ate H2O and
o not begin
oss of cataly
e reduced b
can elimina
eaction syst
e potential f
clogging h
Pathways
rization
ructose and
ydration tha
oxygen ato
ses.25 Alum
) and zinc (
uctose, it is
al center an
s between
ic linkages
d act as a n
to hydrolyz
yst activity
by making s
ate the nee
tem, hetero
for unwante
heterogeneo
d glucose is
an glucose
ms of hydro
minum (III) a
(II) are
s believed t
nd stabilize
aldose and
in cellulose
nucleophile
ze and mak
and
sure the
ed for cataly
ogeneous
ed side
ous pores.1
s a reversib
is. This is
12
oxyl
and
hat
e the
d
e
e.
ke
yst
17
ble
b
s
st
c
=
o
c
in
re
L
liq
T
a
th
[E
th
1
ecause glu
olution is lo
tructures w
an increase
3 kJ mol-1
The f
ligosaccha
ross polym
nclude dehy
eactions.28
There
ewis acid c
quids.29 Ion
They are oft
bility to diss
hemselves.
EMIM]Cl an
his catalytic
0% HMF w
ucose forms
ower than th
which accou
e glucose-fr
) isomeriza
formation of
rides which
erize with i
ydration for
e are three
catalysts, a
nic liquids a
ten used fo
solve cellul
Fructose m
nd heated fo
c activity wa
was detecte
s a stable ri
he rate for f
unt for the fa
ructose iso
ation of gluc
2.2.2 Hu
f humins is
h still have r
ntermediate
rming non-f
2.3
types of so
single solv
are a mixtur
r biomass d
ose14 as w
mixed with
or 3 hours a
as not the s
d.30
ng structur
fructose. Fr
acile HMF d
merization
cose.27
mins and P
contributed
reducing gr
es and HM
furan cyclic
Solvent S
olvent syste
vent or mon
re of ions th
dehydration
ell as demo
in 1-ethyl-3
at 120°C ga
same for glu
re therefore
ructose form
dehydration
due to the
Polymerizat
d to the abi
roups. Thes
F.12 Some
ethers and
Systems
ems typicall
nophasic, a
hat are liqui
n reactions
onstrate cat
3-methylimid
ave a yield
ucose, at 18
e the enoliza
ms less sta
n.12 High te
slightly end
tion
ility of gluco
se reducing
competing
d condensa
y used with
biphasic sy
id at room t
because th
talytic activ
dazolium c
of 70% HM
80°C a yiel
ation rate in
able ring
emperatures
dothermic (
ose to form
g groups ca
pathways
ation
h Brønsted
ystem, or io
temperature
hey have th
vity by
hloride or
MF. Howeve
d of less th
13
n
s
(H
an
and
onic
e.
he
er
han
14
Despite these advantages, ionic liquids also have several disadvantages.
Ionic liquids are expensive, they often require many steps and a lot of energy to
synthesize, and they can be easily deactivated by small amounts of water which
will be generated from carbohydrate dehydration. They are also difficult to
separate from products.25,31
Water is usually the ideal solvent especially for cost and environmental
concerns. But using water can be a poor choice because more water will give
lower HMF yields.27 In addition use of organic solvents need to be able to be
easily removed from products. The solvent used by Pagan-Torres, et al. has a
boiling point similar to HMF and requires additional steps beyond distillation and
results in a loss of product when isolating HMF.25
The use of a biphasic system attempts to make the best of solvent
strengths and weaknesses. By using an organic solvent with water, the products
can be extracted away from excess water and the catalyst thus preventing any
unwanted side reactions. To maximize the benefit of the biphasic system with
water, the organic solvent should be immiscible with water and be volatile
enough to easily remove through reduced pressure without the use of heat.
Earlier research groups used THF as their organic solvent. The benefits of THF
included the stability of the furan ring as well as the potential for green synthesis.
The downside of using THF is its complete miscibility with water. However using
a significant amount of salt in the water resulted in adequate separation of the
two solvents due to the increase in their partition coefficient.20
a
u
2
g
h
p
a
io
a
tr
th
w
y
4
th
h
The u
queous and
ps due to t
-methyltetr
reater imm
ave low bo
oint of HMF
bility to be
Most
onic liquid o
nd Cr(II) ha
ransforming
he ionic liqu
workers use
ield of HMF
:1 in the [E
When
he following
ydroxide io
use of high
d organic s
he corrosiv
ohydrofura
iscibility of
iling points
F. THF and
synthesize
of the litera
or non-aque
ave very sim
g glucose w
uid. HMF yi
ed a bi-cata
F from micr
MIM]Cl wa
n a chromiu
g ligand com
on.17 Jia and
concentrat
solvents will
veness of hi
n is a prom
MeTHF wit
at 65°C an
d MeTHF ar
d from carb
Gl2.4
2.4.1
ature has s
eous solven
milar activit
with Cr was
elds of 70%
lyst system
ocrystalline
s used.14
um salt is d
mbinations,
d co-worke
ions of NaC
l become a
igh salt con
mising altern
th water (14
nd 80°C, bo
re both con
bohydrate b
lucose Tran
Chromium
hown Cr is
nt systems.
ies.23 One
from Zhao
% from gluc
m in ionic liq
e cellulose.
issolved in
, 6 aquas, 5
rs analyzed
Cl to improv
an impedanc
ntent in wat
native to TH
4g/100g).32
oth of which
sidered gre
biomass.
nsformation
m Catalysis
most effec
27 It has als
of the earlie
, et al. and
cose were r
quids and w
A combine
water 99%
5 aquas and
d Cr in wate
ve separati
ce as any r
ter. MeTHF
HF due to th
2 THF and M
h are still be
een solvent
ns
ctive for HM
so been sho
er reports o
their work
reported.30
were able to
ed catalyst o
% of Cr is co
d a Cl-, or 5
er, and look
on between
reaction sca
F or
he much
MeTHF bot
elow the bo
ts due to th
MF productio
own that C
of successf
with CrCl2
Kim and co
o generate 6
of CrCl2: Ru
oordinated w
5 aquas and
ked into
15
n
ale-
h
oiling
eir
on in
r(III)
fully
and
o-
60%
uCl3,
with
d a
16
Chromium’s ability to isomerize glucose into fructose. Over the course of an hour,
they analyzed the conversion of glucose and the fructose being made by the
reaction system. The isomerization proceeds in a linear fashion until around 30%
conversion and then HMF formation accelerates after this point. Cellobiose was
detected at the end of the reaction indicating the reactions of undesired inter- or
intra-molecular dehydration of glucose. Comparable pH solutions using HCl were
compared to CrCl3. The HCl catalyst converted very little glucose and supported
that Cr was an active species for isomerization of glucose.27
Finally a mixed Cr and Brønsted acid system was able to produce
significant yields of HMF. Aqueous chromium (III) chloride converted 100% of
the glucose by itself and with a Brønsted acid co-catalyst. When mixed equally,
each 0.02M, KH2PO4 and CrCl3 were able to produce a yield of 50.3% levulinic
acid with no HMF. However there was a significant amount of glucose that
formed undesired products.4
2.4.2 Aluminum Catalysis
The Abu-Omar lab has already done a lot with Al as a dehydration catalyst.
A water ethanol mix was used with glucose and Al to make furan complexes. The
more water that was present in the two solvent system, the more levulinic acid
produced. Using an all ethanol solvent yielded 44% furans while an addition of 10
wt% water gave a 57% yield of furans. When an acid catalyst was used in place
of Cr, isomerization activity was lost and low yields were generated from glucose.
This reaction system also had difficulty with a cellulose substrate due to 160°C
17
not being a high enough temperature to depolymerize or decrystallize the
polymer.33
Next a water and THF solvent system with 0.25M sugar and 0.1M AlCl3
was compared to a reaction with an In chloride salt catalyst. Al gave 48% HMF
while In only gave a 7% yield despite being more acidic (pka=4 vs. pka=5). The
Lewis acid cation was crucial to the dehydration mechanism and did not depend
on the Brønsted nature of the catalyst. 34 Finally the biphasic system with AlCl3
and glucose was revisited again but the tetrahydrofuran-water system used NaCl
to improve the partition coefficient. This system afforded a 61% yield of 5-
hydroxymethylfufural with very little byproducts (1% levulinic acid). 20
2.4.3 Iron and Other Catalysts
Reading through the literature, there are not a lot of groups reporting high
yields of HMF or levulinic acid from Fe catalysts. vom Stein and co-workers
reported a successful dehydration of xylose into furfural in a biphasic reaction
system using FeCl3 as the catalyst. A 64% furfural yield was achieved in the THF
and water with NaCl solvent system.
Meanwhile investigations into Fe and ionic liquids were mixed. Using low
temperatures, 80°C, FeCl3 and glucose in [EMIM]Cl produced no reaction.23
Using fructose with FeCl3 and Et4NBr, reactions were able to generate an HMF
yield of 86% for reactions at 90°C. This reaction was also completed in a single
organic solvent, N-methylpyrrolidone.35
18
Finally it should be mentioned that there is an industrial process, the
Biofine Process, that uses 1.5-3wt% sulfuric acid to catalyze cellulosic material
into HMF and levulinic acid. The process utilizes a two reactor system. The first
reactor runs at 215°C, 25 bar, for 12 seconds, and HMF is continuously removed
into a second reactor. In the second reactor HMF is reacted at 190°C and 14 bar
for 20 min. The levulinic acid produced gives a 50% yield based on starting
hexose content of mixed cellulosic material. Furfural and formic acid are also
collected. The Biofine Process allows for a competitive cost production of
levulinic acid.21
Table 2-1: Survey of Reactions in the Literature 17, 36, 37, 25, 20
Source Substrate, M Catalyst, M Solvents, ratio Heating Method,
Temp., Time Product Yields
vom Stein37 0.4M Xylose 0.16M FeCl3 Water (NaCl),
MeTHF 1:1
Oil Bath 140°C, 4hr
71% Furfural
Zhao30 0.14M
Glucose 0.008M CrCl2 [EMIM]Cl, 0.5g
Batch Reactor 100°C, 3 hr
68-70% HMF
Pagan-Torres25
0.4163 mmoles Glucose
0.0075 mmoles AlCl3 pH=2.5 HCl
Water (NaCl), 2-Sec-butylphenol
1:2
Oil Bath 170°C, 40min
62% HMF
Yang, Yu20 0.25M
Glucose 0.1M
AlCl3*6H2O
Water (NaCl), THF 1:3
Microwave Irradiation
160°C, 10 min 61% HMF
Choudhary17 0.56M
Glucose 0.019M CrCl3
0.1M HCl
Water (NaCl), THF 1:2
Oil bath 140°C, 360 min,
180 min
59% HMF (biphasic) 46% LA
(monophasic)
Yang, Fan4 0.0556M Glucose
0.02M CrCl3 0.02M
KH2PO4 Water
Batch Reactor 170°C, 4.5 hr
50.3%LA 0% HMF
Szabolcs38 0.28M
Glucose 2M HCl Water
Microwave 170°C, 30 min
48.6% LA
Rasrendra39 0.1M Glucose 0.005M AlCl3 Water Oven
140°C, 6 hr
25% HMF
p
u
cy
H
p
w
re
(h
J
g
ri
o
in
o
a
T
fo
d
is
Litera
athway, an
ndergoes a
yclize, unde
HMF and the
athway inv
works sugge
eacts with t
hydroxymet
adhav and
roup of C-3
ng closes a
f glucose to
ntermediate
Qian
pen-chain
C2-O5 bon
Torres expla
ormation of
ehydration
Ståhl
somerizatio
2.
ature propo
nd most pop
a keto-aldos
ergo dehyd
en finally de
olves direc
est glucose
the hydroxy
thyl)-2-fura
co-workers
3 eliminates
and dehydr
o HMF proc
e, three time
propose a
pathway fo
nd. 40 This
ains. Fructo
f the 5 mem
are due to
berg and co
n mechanis
Propo5
ses two pa
pular sugge
se isomeriz
dration at C
ehydrate tw
t conversio
dehydrate
yl group on
ldehyde wh
s propose a
s and then
ates to form
ceeds more
es faster th
mechanism
r H+/H3O+ w
is similar to
ose can be
mbered ring
solvent co
o-workers u
sm via NMR
osed Reacti
thways for
estion, is the
zation to for
-1, isomeriz
wo more tim
on of glucos
s at the C-2
C-5 to cycl
hich then de
a 3-deoxy-2
isomerizes
m HMF. Jad
e readily thr
an fructose
m through a
where proto
o the direct
formed thro
intermedia
mpetition fo
used a deu
R. The C2 p
ion Mechan
glucose to
e fructose p
rm fructose
ze again to
mes to form
se to HMF.
2 position to
lize to tetra
ehydrates f
2-keto pathw
to 3-deoxy
dhav conclu
rough the 3
e.19,25
a direct cycl
onation occ
glucose to
ough a 1,2
ate. Barriers
or protons.4
terated glu
position of
nisms
form HMF.
pathway. H
e. Fructose
o form the a
HMF.25,19 T
Pagan-Tor
o form a ca
hydro-3,4-d
further to fo
way where
yglucosone
udes that th
3-deoxygluc
lic mechani
curs on the
HMF path
hydride shi
s for acid ca
40
cose to stu
the glucose
. The first
Here glucose
will then
aldehyde of
The second
rres and co-
arbocation t
dihydroxy-5
orm HMF.
the hydrox
which then
he dehydra
cosone
ism, not an
C2-OH to f
the Pagan-
ift after the
atalyzed
udy the
e was
19
e
d
-
that
5-
xyl
n
tion
form
-
20
deuterated. If the reaction followed an ene-diol mechanism, most of the
deuterium would be in the solvent. This would support formation of a ketone on
the C2 position of fructose. The second option was a 1,2-hydride shift and the
fructose species would incorporate 100% of the deuterium. Ståhlberg found the
NMR showed less than 5% of deuterium incorporated into HMF, supporting the
first mechanism. However, enzyme isomerization of glucose occurs via the
second mechanism.41
Guan and coworkers propose HMF is formed by an open chain 1,2-
enedion mechanism or a fructofuranosyl cationic intermediate. In an ionic liquid,
[EMIM]Cl, and Cr(II) system the CrCl3- anion is the active site facilitating a
mutarotation in the isomerization of glucopyranose to fructopyranose. In ionic
liquid Cr, the ionic liquid cation, and chloride ion form a cluster where Cr(III) is
stable in a 5-membered ring structure. Reactivities of glucopyranose
isomerization into fructofuranose decrease in order for the following metal
catalysts WCl3, MoCl3, CrCl3, FeCl3.23
Cr is believed to stabilize ring opening of glucose through the Lewis acid
metal center during isomerization.25 Choudhary proposes an active M(H2O)5OH2+
species where M = Cr(III). This species stabilizes the aldo-keto isomerization
through an enolate.17 Pidko and coworkers propose that the active Cr(II) species
is CrCl42- and a binuclear Cr complex forms to stabilize activated anionic
sugars.29
Finally a combined Brønsted and Lewis acid system, or synergistic
catalyst, proposes a mechanism where the Lewis acid metal center and the
21
Brønsted acid participate in isomerization and rehydration of HMF together. The
Cr3+ center draws electrons from the oxygen of the glucose’s carbonyl which
promotes the alpha H to dissociate and leads to the subsequent isomerization
into the enol intermediate. Protons attack the carbon double bond in the enol.
The use of Brønsted acid allows for an increase in H+ concentration to facilitate
isomerization as well as improve selectivity. For HMF it is hydrated and
dehydrated into the 2,5-dioxohex-3-enal intermediate where the Cr3+ interacts
with the oxygens from two carbonyl groups. Then the oxygens of H2PO4- can
attack the carbons of the carbonyls interacting with Cr3+ and facilitate the carbon
cleavage releasing levulinic and formic acid. H2SO4 and HCl are said to not make
good synergistic catalysts due to the lack complex formation with the
2,5-dioxohex-3-enal intermediate.4
g
to
a
c
a
u
d
re
o
c
th
s
a
CHAPTDE
The m
lucose and
o carbon. In
lternative fu
hapter two,
nd xylose i
sed as a de
ue to its co
esulting in t
ffer an alte
haracteristi
In this
he transform
hown to fur
robust cata
TER 3. IRONEHYDRATI
most abund
d xylose. Th
n order to s
uels or fuel
, transition
n aqueous
ehydration
orrosiveness
the formatio
rnative to m
ics in aqueo
s chapter w
mation of gl
rther elucid
alyst that is
N (III) CHLOON CATAL
C
3.1
dant sugars
hese carboh
uccessfully
additives t
metals hav
and organi
catalyst. Ho
s. It is also
on of humin
mineral acid
ous solution
we investiga
lucose into
ate the rea
s capable o
ORIDE HEXLYST FOR CARBOHYD
Introdu1
s from cellul
hydrates ha
y use glucos
hey must b
ve been sho
ic environm
owever sulf
difficult to c
ns alongside
ds due to Le
ns.
ate the use
HMF and l
ction pathw
f dehydratin
XAHYDRATTHE UPGR
DRATES
uction
lose and he
ave a high a
se and xylo
be dehydrat
own to dehy
ments. Indus
furic acid is
control the
e products.
ewis and B
of iron as a
levulinic ac
way for Fe.
ng sugars f
TE AS A SRADE OF B
emicellulos
amount of o
ose as platf
ted. As revi
ydrate fruct
strially, sulf
s difficult to
reaction pa
. Transition
rønsted ac
a dehydrati
cid. Evidenc
It will be sh
from bioma
ELECTIVEBIOMASS
e include
oxygen rela
form chemic
iewed in
tose, glucos
furic acid is
store and
athways
n metal salts
id
on catalyst
ce will be
hown that F
ass materia
22
E
ative
cals,
se,
s
use
s
for
Fe is
l as
23
well as pure glucose. Iron and aluminum offer more cost effective, earth
abundant metal alternatives to the more toxic chromium salts.
3.1.1 Microwave Irradiation
A microwave reactor will be used to quickly and efficiently heat
dehydration reactions. Microwave heating allows for a rapidly heated and cooled
reaction system. Microwaves quickly generate a desired internal temperature
without having to heat through bulk material. The speed of microwave heating
allows for higher yields, cleaner reactions, and lower reaction times.42
The majority of microwave research began during World War II when
researchers discovered high frequency radar using microwaves. Domestic
microwave ovens would not become widespread until the 1970s and 80s.
Current research with microwaves include chemical and biological applications
including catalysis, and synthesis from ligands to nanoparticles.43, 44
Microwave, or dielectric heating, is an alternative to classic thermal
heating techniques like conductive, convective, or radiative heating. The ability of
microwaves to heat materials relies in its ability to transform electromagnetic
energy into heat. It is important to understand that the energy of microwaves is
only enough for creating heat but not for exciting electrons. Wavelengths are in
between infrared and radio frequencies, with wavelengths of 1cm to 1m or
300GHz-300MHz frequencies. High energy microwaves have a high frequency
and small wavelengths. Some photon energies like those of gamma- or x-rays
have enough energy to excite inner core electrons with eV to the sixth or fifth
24
power, ultraviolet and visible light can initiate chemical reactions, and infrared
radiation can excite bond vibrations. Microwaves have an energy of around
0.0016 eV; which is less than that of Brownian motion and only excites molecular
rotations. The energy required to dissociate a hydrogen atom from a hydroxide
group or a carbon requires 5.2 or 4.5 eV of energy. Therefore microwaves cannot
and do not induce chemical changes on their own.45,43
Microwaves behave similar to light waves and will change direction when
traveling from one dielectric material to another. Microwaves are reflected by
metallic objects, absorbed by some materials, and not absorbed by other
materials. Ceramics and thermoplastics only slightly absorb microwaves. Water
and carbon are very good at absorbing microwaves which have resulted in their
use with food. Domestic ovens and laboratory systems need to use microwave
frequencies that will not interfere with telecommunications and cellular phone
frequencies; the ideal range is 2.45 GHZ or 12.2cm.45
Interaction between electromagnetic waves and mater is quantified by two
types of complex physical qualities, magnetic susceptibility, and dielectric
permeability. The magnetic component of the wave structures magnetic
moments.45 Electromagnetic waves can reorganize dipole moments and induce
electric conduction. Absorption of microwaves by a material depends on the
electromagnetic interactions of the specific polar molecules or dipoles of that
material. There are also thermal effects of microwaves that arise from the dipole
inversion. The alternating electric field results in molecular friction and energy
dissipates in the form of heat. Because this energy dissipation occurs at the
25
molecular level, there is a regular and higher temperature distribution than
compared to conventional heating methods. For liquids, only polar molecules
selectively absorb microwaves.46
The effects microwaves have on reactions can be attributed to thermal
and non-thermal effects. Thermal effects include overheating, hot spots creation
and selective heating. Non-thermal effects include highly polarized fields. The
increased mobility and diffusion may create an increase in probability of effective
reaction collisions. Looking at the Arrhenius equation, the reaction rate can be
understood as follows:
(3. 1)
The rate constant, k, is defined by A, the pre exponential factor, that also
includes the probability of molecular impacts or molecular alignment, Ea the
activation energy, T the temperature of the reaction, and R the gas constant. By
influencing the orientation of polar molecules you would influence the collision
efficiency and therefore increase the pre-exponential factor, A, without changing
the activation energy, Ea. Another way the reaction is influenced is by the outright
decreasing of the activation energy, EaThis can be explained when thinking in
terms of enthalpy and entropy.
∆ ∆ ∆ (3. 2)
According to Perreux, the magnitude of – TS would increase for a reaction
heated by microwaves. Microwave heating results in more organization than a
reaction heated conventionally due to the difference in dipolar polarization. In
a
st
b
st
st
(I
(S
(I
9
s
a
le
c
w
s
h
Q
ddition the
tate stabilit
ecomes en
tate, the us
tate in turn
All ch
II) chloride
Sigma Aldri
I) acetate (
9.9%), zinc
ulfate hepta
nhydrous (
evulinic acid
ellobiose (S
with 150-400
ulfuric acid
ydrobromic
Q Advantag
activation e
y. If the tran
nhanced as
se of microw
decreasing
hemicals we
hexahydra
ich, 96%), a
(Aldrich, 99
c bromide (S
ahydrate (A
Mallinckrod
d (Sigma A
Sigma, 98%
0 ppm BHT
(Macron, 9
c acid (Sigm
e A10 Ultra
energy cou
nsition state
the reacta
waves will h
g activation
3
3
ere used as
ate (Alfa Ae
aluminum c
%), molybd
Sigma Aldr
Alfa Aesar,
dt), 5-(hydro
ldrich, 98%
%). The orga
T stabilizer (
98%), hydro
ma Aldrich,
apure Wate
ld also be r
e has a gre
nt move fro
help stabiliz
energy of
Meth3.2
3.2.1 Mate
s received.
esar, 97%),
chloride hex
denum (V) c
rich, 98%),
98%). Subs
oxymethyl)f
%), D-(-)-fruc
anic solven
(Sigma Ald
ochloric aci
48%). All w
er Purificatio
reduced du
eater polarit
om ground s
ze the form
the reactio
hods
erials
Metal cata
chromium
xahydrate (
chloride, an
zinc chlorid
strates incl
furfural (Sig
ctose (Sigm
nt used was
rich, 99.5%
d (Macron,
water used
on System.
ue to increa
ty or the di
state into th
ation of the
n system.42
lysts used i
(III) chlorid
(Fluka, 99%
nhydrous (S
de (Aldrich,
uded D-glu
gma Aldrich
ma Aldrich,
s 2-methylte
%). Acids us
37%), form
was purifie
ased transiti
ipole mome
heir transitio
e transition
2,43,45–47
included iro
e hexahydr
%), mangan
Sigma Aldri
99.9% ), z
ucose,
h, 99%),
99%), D-(+
etrahydrofu
sed include
mic acid (88
ed using a M
26
ion
ent
on
on
rate
nese
ich,
zinc
+)-
uran
d
8%),
Milli-
27
Biomass included poplar and switchgrass samples. Poplar was obtained
from Purdue University’s Department of Forestry and Natural Resources. The
poplar sample was a hybrid aspen INRA 717-1B4 (P. tremula x P. alba, WT-717).
Switchgrass was obtained from Dr. Keith Johnson and was grown at
Throckmorton-Purdue Agricultural Center near West Lafayette, Indiana. Both
biomass samples were dried to about 5% moisture and milled to fine particles
using a No. 20 mesh; 0.841mm openings. Carbohydrate composition was
determined by the Mosier research group using the LAP002 sulfuric acid
digestion method.48 Poplar was composed of 23% xylans and 42% glucans.
Switchgrass was composed of 18% xylan and 45% glucans.
3.2.2 Dehydration Reactions
Dehydration reactions were performed with a CEM Discover SP/S-class
microwave reactor. A one milliliter aqueous solution (made with Millipore filtered
water) of 0.1M FeCl3 6H2O and 0.25M glucose was added to a 10 mL glass
microwave reaction vessel. Three milliliters of 2-methyltetrahydrofuran (MeTHF)
was then added to the vessel along with a small Teflon coated stir bar. Using the
fixed power control method, reactions were heated to 140 - 200°C at 100 or 200
watts of microwave power for up to 60 minutes. Reaction times began once the
set temperature was achieved as registered by the instrument’s IR temperature
sensor. Reactions were quenched by a nitrogen gas flow which cooled the
system to 60°C.
28
3.2.3 Analysis
After the reaction completed the organic layer was separated from the
aqueous layer. The aqueous layer was washed two times with small aliquots of
MeTHF. The wash aliquots were combined with the original organic layer and
filtered through a glass wool plug. The organic layer was concentrated to a few
hundred microliters using a Buchii Rotavapor. The concentrated organic layer
was analyzed by proton NMR (Bruker ARX400, qnp probe, 32 scans) using
deuterated acetone (Cambridge Isotope Laboratories, 99.9%) and an internal
standard of N,N-dimethylformamide (Macron, 99.8%). The NMR data for HMF,
furfural, and levulinic acid is as follows. 5-(hydroxymethyl)-2-furfuraldehyde
(C6H6O3) 1H NMR (400 MHz, acetone-D6): 9.59 (s, 1H), 7.38 (d, J = 3.4 Hz,
1H), 6.58 (d, J = 3.4 Hz, 1H), 4.72 (dd, J = 5.6, 6.6 Hz, 1H), 4.63 (d, J = 6.0 Hz,
2H). Furan-2-carbaldehyde (C5H4O2) 1H NMR (400 MHz, acetone-D6): 9.67 (s,
1H), 7.94 (m, 1H), 7.44 (d J = 3.6 Hz, 1H), 6.74 (dd, J = 3.6, 1.7 Hz, 1H). 4-
oxopentanoic acid (C5H8O3) 1H NMR (400 MHz, acetone-D6): 2.72 (t, J = 6.5
Hz, 2H), 2.50 (t, J = 6.5 Hz, 2H), 2.12 (s, 3H). Integration and analysis of NMR
spectra was performed with Mest Re Nova version 8.1
The aqueous layer was collected, diluted 20 times, and filtered through a
polypropylene syringe filter with 0.2 micrometer pores prior to analysis by HPLC.
A Waters 2695 Separations Module with a Bio-Rad Aminex HPX-87H column
and Waters 2414 Refractive Index Detector was used. Sample analysis used a
0.005 M sulfuric acid or a 5 %( w/w) acetonitrile in 0.005M sulfuric acid mobile
phase at a flow rate of 0.6 mL min-1. Column temperature was 65°C. Quantitative
a
b
p
p
%
%
T
%
S
%
%
c
p
a
nalysis of g
ased on ex
H measure
H units).
The a
%YieldofHM
%YieldofFur
The convers
%Conversion
Selectivity to
%Selectivity
%Selectivity
Looki
atalysts, re
revious Al a
nd sulfate s
glucose, fru
xternal stan
ements wer
amount of p
MForLA
rfural
sion of reac
nofreactant
owards spe
ofHMForLA
ofFurfural
3.3.1
ing to expa
actions we
and Cr yield
salts were a
uctose, HMF
dard curve
e taken usi
3.2
product prod
ctants (i.e. g
ecific produc
A
R3.3
Survey of
nd the cata
re performe
ds found in
also examin
F, and furfu
s made fro
ng an Accu
2.4 Calcul
duced, or p
1
glucose, xy
ct formation
Results and
f Transition
alog of valua
ed with Fe,
the literatu
ned. Microw
ural concen
m known p
ument AB15
lations
percent yiel
100
100
lose) was c
n was calcu
Discussion
n Metal Salt
able transit
Zn, Mn, M
ure. Differen
wave heat s
ntrations we
pure compo
5/15+ pH m
d, was calc
calculated b
100
ulated by:
100
100
n
t Activities
tion metal d
o and comp
nces betwe
settings use
ere calculate
ounds. Aque
meter (+/- 0
culated by:
(
(
by:
(
(
(
dehydration
pared to
een chloride
ed low pow
29
ed
eous
.01
(3. 3)
(3. 4
(3. 5)
(3. 6)
(3. 7)
n
e
wer,
30
100 W, and a temperature set point of 160°C held for 20 minutes. The biphasic
solvent system was used in order to maintain maximum product formation.
Manganese and both zinc halide salts were not explored further due to
their low HMF yields. Fe (III) chloride produced the most HMF out of all of the
metal salt surveyed. Molybdenum (Entry 7 in Table 3-1) showed poor HMF
formation despite transforming almost all of the glucose. However it produced
45.6% levulinic acid while all other entries produced very little to no rehydration
products. Mo was investigated further using lower catalyst concentrations.
Table 3-1: Different Transition Metal Salt Catalysts for Glucose Dehydrationa
Entry Metal Salt Catalyst, 0.1M % Yield of HMF 1 Iron (III) Chloride hexahydrate 14 2 Zinc Sulfate heptahydrate 9 3 Iron (III) Sulfate hydrate 6 4 Iron (II) Chloride, anhydrous 5 5 Zinc Chloride 4 6 Zinc Bromide 4 7 Molybdenum (V) Chloride, anhydrous 3 8 Manganese(II) acetate tetrahydrate 0
aDehydration reactions used 0.25M glucose with the respective catalyst in 1 mL H2O and 3 mL MeTHF.
Reactions were heated to 160°C (100 W) for 20 minutes.
Initial catalyst concentration began at 40 mol% relative to the substrate.
Due to the large amount of rehydration product formed, the amount of Mo was
reduced to 10 and 4%. Table 3-2 shows the product yields for these
concentrations. At 4 mole %, Mo still transforms almost all of the glucose but the
improvement in yields were not as high or selective as to be a competitive with
31
previously reported dehydration catalysts. In addition, all Mo reactions formed
insoluble black precipitate indicative of humin and insoluble polymer formation.
Table 3-2: Changing the Concentration of Molybdenum Catalyzed Dehydration Reactionsa
Entry Conc. of MoCl5, Mb % Yield of HMF % Yield of Levulinic Acid
1 0.100 3 46
2 0.025 18 20
3 0.010 15 15 a
Dehydration reactions used 0.25M glucose with the respective catalyst in 1 mL H2O and 3 mL MeTHF. Reactions were heated to 160°C (100 W) for 20 minutes. b
For all three concentrations conversion of glucose was 90-100%.
The catalyst amount of Fe (III) chloride was also varied. Table 3-3 shows
how the conversion and selectivity change with the amount of catalyst. At
concentrations of 30 mol% the HMF selectivity improved to 80%, where using
40% catalyst reduced HMF selectivity. However at the higher catalyst
concentration we see an almost double increase in the amount of glucose
transformed.
Table 3-3: Changing Iron Concentrations for Dehydration Reactionsa
Entry FeCl3 (mol %) % Yield of HMF
% Yield of Levulinic Acid
% Glucose Converted
% HMF Selectivity
1 40 14 6 49 28
2 30 23 7 28 81 a
Dehydration reactions used 0.25M glucose with the respective catalyst in 1 mL H2O and 3 mL MeTHF. Reactions were heated to 160°C (100 W) for 20 minutes.
32
3.3.2 Maximizing the Dehydration Reaction with Microwave Irradiation
Iron catalyst dehydration reactions used fixed power microwave irradiation.
The CEM reactor program controlled the temperature set point by rapidly cycling
the irradiation on and off. The effects of different temperatures and microwave
powers will be looked at as the dehydration reaction progress over time.
Figure 3-1 shows how the dehydration and rehydration reactions
proceeded over time when using low power, 100W microwaves, to heat an FeCl3
and glucose reaction to 160°C . Maximum glucose conversion is 72% after one
hour. This reaction system was selective towards HMF over LA. The maximum
HMF produced in an hour was around a 30% yield.
Figure 3-1: Time Profile for 0.1M FeCl3 System Heated to 160°C (100 W) Glucose (0.25M) was reacted with 0.1M FeCl3 in 1 mL H2O and 3 mL MeTHF.
When a higher microwave power, 200W, is applied to the FeCl3-glucose
system to achieve 160°C, we begin to see a change in the amount of glucose
0
20
40
60
80
100
5 10 30 60
%
Time, minutes
% Glucose Converted
% Yield of HMF
% Yield of Levulinic Acid
% Yield of Formic Acid
% HMF Selectivity
% LA Selectivity
33
being converted as well as a change in product selectivity (Figure 3-2). The
amount of glucose converted is now around 80% after one hour. The production
of levulinic acid steadily increases over the course of the reaction. And unlike the
previous reaction system, the production of HMF begins to decrease after 30
minutes.
Figure 3-2: Time Profile for 0.1M FeCl3 System Heated to 160°C (200 W) Glucose (0.25M) was reacted with 0.1M FeCl3 in 1 mL H2O and 3 mL MeTHF.
By maintaining high microwave power but increasing the temperature to
200°C a change in conversion and product selectivity was observed
Figure 3-3). Complete glucose conversion occurs at 30 minutes. Product
selectivity began favoring HMF and then switched to very high levulinic acid
selectivity after 5 minutes of reaction time. At 60 minutes more than an 80% yield
of levulinic acid
0
20
40
60
80
100
5 10 30 60
% Glucose Converted
% Yield of HMF
% Yield of Levulinic Acid
% Yield of Formic Acid
% HMF Selectivity
% LA Selectivity
34
Figure 3-3: Time Profile for 0.1M FeCl3 System Heated to 200°C (200 W) Glucose (0.25M) was reacted with 0.1M FeCl3 in 1 mL H2O and 3 mL MeTHF. (Very similar results can also be seen when heated to 180°C (200 W).
has been produced and very little HMF is detected. At this temperature and high
microwave power, the best production of HMF and levulinic acid gives a 39 and
88% yield respectively (Table 3-4).
Table 3-4: The Best Production of HMF and Levulinic Acid from Glucose
Entry % Glucose Converted
% Yield of HMF
% Yield of Levulinic Acid
% Yield of Formic Acid
% HMF Selectivity
% LA Selectivity
1a 70 39 22 23 56 32
2b 100 2 88 68 2 88
aDehydration reactions used 0.25M glucose with 0.1M Fe (III) in 1 mL H2O and 3 mL MeTHF. Reactions were heated
to 200°C (200 W) for 5 minutes. b
Dehydration reactions used 0.25M glucose with 0.1M Fe (III) in 1 mL H2O and 3 mL MeTHF. Reactions were heated to 200°C (200 W) for 60 minutes.
0
20
40
60
80
100
5 10 30 60
%
Time, minutes
% Glucose Converted
% Yield of HMF
% Yield of Levulinic Acid
% Yield of Formic Acid
% HMF Selectivity
% LA Selectivity
35
3.3.3 Using Brønsted Acids as a Co-Catalyst for Iron
Experiments were performed to look at small additions of a Brønsted acid
to the Fe (III) Lewis acid catalyst and its effectiveness. Table 3-5 illustrates how
glucose is transformed when using only a Lewis acid or Brønsted acid catalyst
(entry 1 and 3, respectively) or when both types of catalyst are used together.
The biggest changes between the different types of catalyst systems were
evident in the conversion and selectivity. At high temperatures for a short amount
of time using the Brønsted and Lewis acid together converted the most glucose.
However the best HMF selectivity was only evident when a single catalyst type
was used.
Table 3-5: Using Brønsted Acid as a Co-catalyst with Irona
Entry Catalyst % Yield of HMF
% Yield of Levulinic Acid
% Glucose Converted
% HMF Selectivity
% LA Selectivity
1 FeCl3 • 6H2O 39 22 70 56 32
2 FeCl3 • 6H2O + 0.05M HCl 29 24 83 35 29
3 HCl, pH=1 (0.1M) 27 11 40 66 27
aDehydration reactions used 0.25M glucose with 0.1M Fe (III) in 1 mL H2O and 3 mL MeTHF. Reactions were heated to 200°C (200 W) for 5 minutes.
Different types of Brønsted acids were tested for their co-catalyst activity
(Table 3-6). Again, the different types of acids did not dramatically change
product yields, but changes to glucose conversion can be observed. When we
look closer at hydrobromic acid as a co-catalyst to Fe (III) we see selectivity
differences between the combined system and an individual Brønsted acid
36
catalyst. In entry 3 and 9 of Table 3-7 it can be observed that by using HBr with
iron the production of HMF is suppressed. Again, increased conversion of
glucose is also observed for this combined acid system.
Table 3-6: Different Type of Lewis-Brønsted Acid Systemsa
Entry Co-catalyst % Yield of HMF
% Yield of Levulinic Acid
% Glucose Converted
% HMF Selectivity
% LA Selectivity
1 Sulfuric Acid 23 21 82 29 26
2 Hydrobromic Acid 28 21 80 35 27
3 Hydrochloric Acid 29 24 83 35 29
4 Formic Acid 26 14 65 40 21
5 Maleic Acid 38 22
aDehydration reactions used 0.25M glucose with 0.1M Fe (III) and 0.05M (pH = 1.3) of above respective co-catalyst
in 1 mL H2O and 3 mL MeTHF. Reactions were heated to 200°C (200 W) for 5 minutes.
When Fe(III) and HBr are used together in large concentrations, 0.1M and 0.5M,
glucose conversion increases and HMF yield decreases (Table 3-4). Significant
yields of levulinic acid do not become evident until a concentration of 0.5M HBr is
used for individual or combined systems. Entry 5 shows that for a very low
concentration of HBr, 0.0025M combined with Fe, no changes in selectivities
were observed but there was an increase in glucose conversion.
37
Table 3-7: Comparing Single Brønsted Acid Dehydration and Combined for Fe (III) and HBr Acid Systemsa
No. Concentration of Fe(III), M
Concentration of HBr, M
% Yield of HMF
% Yield of Levulinic Acid
% Glucose Converted
% HMF Selectivity
% LA Selectivity
1 0 0.005 Trace Trace 0 N/A N/A
2 0 0.050 8 5 16 49 29
3 0 0.500 18 68 88 21 78
4 0.1 0 14 6 49 28 12
5 0.1 0.0025 21 10 90 23 12
6 0.1 0.0050 24 21 61 50 21
7 0.1 0.0250 26 18 90 29 20
8 0.1 0.0500 23 24 75 32 31
9 0.1 0.5000 4 69 97 4 71 a
Dehydration reactions used 0.25M glucose with 0.1M respective catalysts in 1 mL H2O and 3 mL MeTHF. Reactions were heated to 160°C (100 W) for 20 minutes.
3.3.4 Comparing Iron to Aluminum and Chromium Salt Catalysts
AlCl3 and CrCl3 were also used as dehydration catalysts in the biphasic
solvent system. Dehydration reactions for Al and Cr were also compared to the
performance of Fe at short times when using 200°C (200 W). Both Cr and Al
were able to transform all of the glucose; however the selectivity towards HMF
and levulinic acid was very low when compared to the selectivity for products
when using Fe. This reactivity difference will be important when we look at
transition metal salts abilities to use biomass as a substrate, which will be
discussed in section 3.4.
38
Figure 3-4: Comparing Glucose Transformation for Three Transition Metal Salts Glucose (0.25M) was reacted with 0.1M of the respective catalyst in 1 mL H2O and 3 mL MeTHF and microwave irradiated for 5 minutes at 200°C (200 W).
When comparing Al and Cr to the Fe reactions it was evident that Al and
Cr were able to easily transform all of the glucose. However when comparisons
between product yields and selectivities are made it is evident that for high power
and high temperature reactions Fe does a much better job than Al and Cr
dehydrating glucose into desirable products.
3.3.5 Dehydration of Other Substrates
Carbohydrate substrates beyond glucose and fructose were studied.
Cellobiose was easily transformed by Fe (III) converting almost all of the
substrate and improving the HMF selectivity for the reaction. Levulinic acid and
HMF are both susceptible to decomposition into other products. Xylose only
0
20
40
60
80
100
FeCl3 • 6H2O AlCl3 • 6H2O CrCl3 • 6H2O
%
Catlayst
% Glucose Converted
% HMF Selectivity
% LA Selectivity
39
forms Furfural, at a 41.8% yield, and no other desired products, entry 4 in
Table 3-8.
Table 3-8: Dehydration of Different Model Compoundsa
Entry Substrate % Monomer remaining in
aqueous layer % Yield of HMF
% Yield of Levulinic Acid
1 Cellobiose 0.2 34 21
2 HMF 3.0 34b 44
3 Levulinic Acid 5.4 0 75c
4 Xylose 2.1 0 0
aDehydration reactions used 0.25M monomer with 0.1M Fe (III) in 1 mL H2O and 3 mL MeTHF. Reactions
were heated to 200°C (200 W) for 5 minutes. bPercentage of starting HMF remaining after reaction
cPercentage of starting LA reaming after reaction
As a catalyst Fe is not only able to dehydrate glucose but it also is able to
hydrolyze the glycosidic bonds that link the glucose polymer chains. Dehydration
reactions using a milled unprocessed biomass have shown significant yields of
desirable products. Figure 3-6 shows the product yields for different reaction
conditions for transformations of a poplar substrate. For short reaction times, a
significant amount of furfural was made from xylans relseased from biomass.
Furfural yields of 60% were around the same value even when reaction times
were extended to 30 minutes. Reactions for 60 minutes proved to be too long,
reducing valuble product yields as well as increasing the amount of solids
present in the reaction system. Solids included undissolved biomass as well as
insoluble humions and polymers. There was also little difference in washed and
unwashed biomass.
40
Figure 3-5: Dehydration of Biomass Glucans using Iron An amount of biomass yielding 0.25M glucans was reacted with 0.1M FeCl3
catalyst in 1 mL H2O and 3 mL MeTHF and microwave irradiated for various times at 200°C (200 W).
When using Al as a catalyst for biomass dehydration there was a large
difference in the amounts of products made. Al was not only able to generate
almost 60% furfural at short times, 5min, but the HMF yield was almost twice of
that made from Fe in the same amount of time (Figure 3-6). However at short
reaction times there was a significant amount of solids remaining in solution. For
30min reaction times the amount of solids decreased but the amounts of HMF
and furfural decreased. At 60 minutes the amount of solids began to increase
again and the only significant product amount was 40% of levulinic acid, 40%
yield.
0
20
40
60
80
100
5 min 30 min 60 min 30 min 40 min 50 min 40 min
washedpoplar
washedpoplar
washedpoplar
unwashedpoplar
unwashedpoplar
unwashedpoplar
unwashedswitch grass
% Yield of HMF % Yield of Levulinic Acid
% Yield of Furfural % Solids Collected
41
Figure 3-6: Dehydration of Biomass Glucans using Aluminum An amount of biomass yielding 0.25M glucans was reacted with 0.1M AlCl3
catalyst in 1 mL H2O and 3 mL MeTHF and microwave irradiated for various times at 200°C (200 W)
At high power and high temperature conditions Cr produced the least
amount of desired products (Figure 3-7) at short and long reaction times. Five
minute and 60 minute reactions all had a large amount of solids remaining in
solution. Short reaction times left behind a lot of un-hydrolyzed material but
longer reaction times formed insoluble humins and polymers in addition to
undissolved biomass.
0
20
40
60
80
100
5 min 30 min 60 min
washed poplar washed poplar washed poplar
% Yield of HMF % Yield of Levulinic Acid
% Yield of Furfural % Solids Collected
FActi
s
m
B
a
m
sy
L
e
Figure 3-7: DAn amount oatalyst in 1 mes at 200
Iron (
uch as HM
mechanistic
Brønsted ac
nd assume
metals. Beca
ystem, diffe
ewis acid.
In Fig
xcept for th
0
20
40
60
80
100
Dehydrationof biomass mL H2O an
0°C (200 W
III) chloride
F and levul
pathway fo
cid catalyst.
es mechanis
ause of larg
erent interm
gure 3-1 an
he microwa
5 min
washed popl
% Y
% Y
n of Biomasyielding 0.2
nd 3 mL Me)
3.
e has the ab
linic acid. T
or Fe that d
Current lite
stic insight
ge differenc
mediates an
d in Figure
ve power a
ar
Yield of HMF
Yield of Furfu
ss Glucans25M glucaneTHF and m
Discus4
bility to con
The above e
differs from
erature lum
for Cr can
ce between
nd different
3-2 the rea
applied to th
30 min
washed po
ral
s using Chrons was reacmicrowave
ssion
nvert glucos
experiments
previously
mps all Lewi
be translate
n Cr and Fe
pathways m
actions con
he system.
n
oplar
% Yield
% Solid
omium cted with 0.irradiated f
se into valu
s highlight
studied Lew
is Acids into
ed to other
e data in the
must be for
nditions are
In Figure 3
60 m
washed
of Levulinic A
s Collected
.1M CrCl3
for various
able produc
a unique
wis and
o one categ
r transition
e same rea
rming for ea
nearly iden
3-1 the
min
poplar
Acid
42
cts
gory,
ction
ach
ntical
43
conversion of glucose as well as the yield of desired products follows a linear
increase over time. However, for Figure 3-2 the production of desired products
and selectivities have a noticeable decrease, almost peaking around 30 minutes,
while the conversion of glucose continues to follow a linear increase over time.
These two results can be explained by the different applications of microwave
power
As outlined in section 3.2, the reaction system used a fixed power
program. The microwaves were cycled on and off in order to maintain
temperature. The reactions in Figure 3-1 were set up to maintain 160°C using
100W of power while the reactions in Figure 3-2 used 200W of power to control
the temperature. In order to maintain 160°C, the instrument cycled 200W of
power less frequently. In other words there was more time when the microwaves
were off for the reactions in Figure 3-2 then they were for reactions in Figure 3-1.
Even though the temperatures were the same, the products reacted differently to
the amount to electromagnetic waves being applied to the reaction system.
The reaction system in Figure 3-3 utilized a maximum temperature as well
as maximum microwave power. 200W was the maximum recommended
microwave power to be used for reactions in water and 200°C was the maximum
value the reaction system could generate. The instrument frequently cycled the
power on and off to control temperature; the cycles for a 200W 200°C reaction
system were two times that of a 200W 160°C reaction system.
The applications of microwaves are effecting the collisions as well as the
polar molecules to result in enhanced reaction pathways for the production of
44
HMF. However if the applications of microwaves was only improving the effective
collisions between catalyst and substrate, we should see similar activity when we
look at other catalysts using high temperature and power reactions. Looking at
Figure 3-4 we can see that the three transition metals, Fe, Al, and Cr have
different responses to the increased microwave power. Al and Cr were able to
transform all of the glucose, however only a small amount of this transformation
resulted in the production of HMF or levulinic acid. Here the effect of increased
collisions did not result in the improvement of desired product pathways.
Microwave irradiation must be stabilizing a polar intermediate that is only
generated by an iron catalyst and is not present in Cr or Al catalysts reactions.
Differences in catalyst interactions with microwaves may be influenced by
the differences in the ion’s physical properties. Table 3-9 lists the physical
properties for the ions that are involved in the transformation of glucose. Entries
1-6 have been shown to transform significant amounts of glucose into valuable
products while the remaining entries have shown less then promising product
yields under mild reaction conditions. The clearest trend that differentiates entries
1-6 from the rest of the catalyst ions are the low pKa values, resulting in the
formation of acidic solutions. Looking closer at Fe(III), Cr(III) and Al (III), Fe has
the higher pKa, electronegativity, as well as polarizability. These three properties
may be responsible for the ability of microwave irradiation to improve iron’s
reaction mechanisms. Glucose to HMF formation increases with decreasing ionic
radii, this allows the catalyst to have a stronger electrostatic interaction between
glucose and the smaller cation.25 This trend seems to follow, especially for the
45
biomass samples, Fe did the best followed by Al and then Cr, which has the
largest radii of the three metals. However other sources indicates
electronegativity is more important for metal ligand interactions than radii.49
Table 3-9: Physical Properties for All of the Catalyst Ions Used in this Study
Entry Catalyst Ion pKaa Radiusb, pm Electronegativityb
Electric Polarizabilityc, Å3
1 SO42- -2.0, 1.9 (S) 184 (S) 2.58 ~5
2 Cl- -7 181 3.16 4.65
3 Br- -9 196 2.96 N/A
4 Fe3+ 2.2 55 1.83 2.14
5 Cr3+ 4 62 1.66 1.565
6 Al3+ 4.85 53 1.61 0.417
7 Fe2+ 9.49 61 1.83 1.339
8 Zn2+ 9.0 74 1.65 1.297
9 Mo5+ N/A 61 2.16 N/A
10 Mn2+ 10.6 67 1.55 1.39 a
Used source 16 for entries 1-3; source 50 for entry 4; source 51 for entries 4, 8, 10; and source 26 for entries 6-7 b
Used source 16 for all entries cUsed source 52 for entry 1 and source 53 for entries 2-10
The pathways to desired and undesired products were influenced by the
amount of catalyst present as well as the amount of glucose transformed. When
comparing entry 1 in Table 3-3 and entry 1 in Table 3-4, the selectivity for HMF
almost doubles; 27.9 to 56% selectivity. And by running the high temperature and
wattage reaction for longer times, the conversion of glucose into HMF follows into
the rehydration path to form levulinic acid. From a sulfuric acid solution a 62%
46
yield of levulinic acid can be achieved from glucose at 140°C.39 In this chapter an
FeCl3 solution was shown to yield 88% levulinic acid at 200°C
The reactions with Fe and the Brønsted acid showed little results for
improved selectivity and desired product yields. The biggest change resulted in
an increase in the amount of glucose transformed. The presence of additional or
excess protons from a Brønsted acid did not result in improved selectivity.
Other substrates were used for dehydration reactions. When using HMF
as a substrate it results in the formation of levulinic acid. And when levulinic acid
is used as a substrate it is relatively stable with 25% transforming into other
products. The addition of formic acid does not result in increased selectivity or
yields, and shows no change in levulinic yields this other literature shows that
Formic acid is produced before levulinic acid and does not promote any Le
Chatelier shift towards levulinic acid production.
When using biomass substrates, Cr continues to struggle to transform
glucose or glucans into valuable products giving low yields. Al, however is able to
produce significant amounts of HMF, and furfural from xylans, in a short amount
of time. However increased reaction times do not improve results. Catalysts
differences are made more complex due to the biomass substrates which bring
other carbohydrates, proteins, lipids and other materials into the reaction system
that can result in helping or hindering the reaction system.
Kamireddy and co-workers used FeCl3 as a pretreatment of biomass. Fe
increased hemicellulose hydrolysis in aqueous solution and increased
hemicellulose removal when compared to a hot water treatment. Pretreatment
w
c
w
w
T
re
te
U
a
le
was 10 min
ellulose. Th
water molec
water as a n
Iron (
Through the
eactivity of
emperature
Unlike other
cid was ab
evulinic acid
at 180°C w
he iron met
cules. Lewis
nucleophile.
III) chloride
e use of a b
Fe was opt
e reactions a
r transition m
le to be ach
d from a no
with 0.1M Fe
al center po
s acids clea
.26
3.
e proved to
iphasic rea
timized to p
also proved
meatal cata
hieved. Res
on-Brønsted
eCl3. Fe tar
olarized an
ave glycosid
Concl5
be a robus
action syste
produce sig
d useful for
alyst a high
sulting in on
d acid catal
rgeted C-O
d withdrew
dic linkages
usion
st and active
m and micr
gnificant am
the dehydr
ly selective
ne of the hi
yst.
-C and C-H
w electron d
s by using c
e dehydrati
rowave irra
mounts of LA
ration of bio
e production
ighest repo
H bonds in
ensity from
coordinated
ion catalyst
adiation the
A. These hi
omass suga
n of levulini
orted yields
47
m
d
t.
igh
ars.
c
of
d
c
st
b
c
c
m
th
m
a
st
CHAPTERWAL
Over
ry weight c
ellulose, he
tructures.15
iomass ma
ompounds
arbohydrat
majority of th
hree major
microfibrils a
Xylog
nd it will inc
tructurally u
R 4. PROBILL WITH FL
90% of bio
can be attrib
emicellulose
5 As discus
aterial can g
derived fro
es can be m
he cell wall
structural p
and connec
glucan is a
corporate fu
unique suga
NG BIOMALUORESCE
4.1
4.1.1
omass sourc
buted to the
e and lignin
ssed in cha
generate ch
om fossil fue
made can h
is compos
polymers is
cts them to
4.1.2
branched s
ucose onto
ar, which ca
ASS: FOLLOENT TAGG
Introdu1
The Plant
ces have a
e cell wall15
n, is home t
pters one a
hemical com
els. In addit
help in the a
ed of gluco
xyloglucan
the cross-li
Utilizing X
structure ma
the ends o
an be introd
OWING FUGING USING
uction
t Cell Wall
cell wall3 a
. The cell w
to many ca
and two, ha
mpounds an
tion, unders
analysis of
ose and xylo
n which hyd
inked hemi
Xyloglucan
ade primari
of the branc
duced to th
UCOSE INTG CLICK C
and up to 3
wall, compo
rbohydrate
arnessing C
nd fuel stoc
standing ho
their decon
ose. In add
drogen bond
cellulose.54
ily of glucos
ches. Fucos
he cell extra
TO THE CECHEMISTRY
0% of a pla
osed of
and aroma
C and H in
cks to repla
ow
nstruction. T
dition to thes
ds to cellulo
4
se and xyla
se is a
acellularly
48
ELL Y
ant’s
atic
ace
The
se
ose
an,
49
where it will become incorporated into the cell’s system through the salvage
pathway. Xyloglucan is made from recycled monosaccharides and extracellular
fucose in the Golgi. Xyloglucan is also made through a de Novo pathway where
fucose is only made from mannose.55 This allows for a modified fucose to be
introduced into a plant cell and allowed to become incorporated in to the
synthesis of the wall.
By making a minor modification to the C-5 position of a sugar and
acetylating its -OH groups to facilitate membrane diffusion, Sawa and co-workers
demonstrated the incorporation of azido fucose into mammalian cells. Sawa used
a small, mostly chemically inert azide group to tag the fucose.55
4.1.3 Fluorescence
Fluorescent techniques offer high specificity while usually being non-
destructive to the sample. Fluorescent active compounds will absorb UV or
visible light which will cause electrons to become excited to a higher energy
orbital. The absorbance of energy promotes a ground state electron into a
vibrational state and results in excited electronic and vibrational states.
Absorption also occurs in a very short amount of time, around 10-15 seconds.
After absorption of light the vibrational modes will relax very quickly and the
compound will undergo internal conversion as a radiationless transfer occurs.
Non-fluorescent molecules will return to ground state efficiently be transferring
heat to the surrounding solvent. Fluorescent molecules will give off their energy
as light when they return to ground state. The lifetime of this excited state is
50
dependent on many factors including the molecules’ environment which includes
solvent and nearby molecules that can quench the fluorescence. Emitted
fluorescent light will be of lower energy and longer wavelength than the light that
was absorbed. Biologically proteins like tryptophan and cofactors like flavins,
pyridine nucleotides and pyridoxal are naturally fluorescent. Fluorescent dyes
can be used to mark proteins or membranes. Studying a biological system with
fluorescent compounds may involve looking for the formation of or loss of
fluorescent compounds. Fluorogenic reaction can also be used in a similar way
to generating colored compounds for spectrophotometry.56
4.1.4 Click Chemistry
Rostovtsev and co-workers demonstrated the benefit of using an azide
group to undergo a Huisgen 1,3-dipolar cycloaddition with an alkyne to create a
fluorescent probe.57 Using the techniques proposed by Sawa and co-workers, 4-
ethynyl-N-ethyl-1,8-naphthalimide (4ENP) and peracetylated azido fucose (aAF)
can be clicked together to generate a fluorescent species with a signal stronger
then the individual reactants.
F
p
c
a
c
re
lig
th
C
Figure 4-1: M
The m
resence of
Figur
opper (I). In
scorbate. A
opper (I) sp
eaction rate
gand speci
hen coordin
Cu-substitut
Mechanism
mechanism
a copper (
e 4-1) copp
n solution th
A ligand, lik
pecies. Acc
e between t
es will form
nate the azi
ted 1, 2, 3-t
m for Huisge
to cyclize t
I) catalyst (
per (II) sulfa
he copper (
e histidine,
cording to S
the azido fu
m a Cu (I) ac
de group. T
triazole is p
en 1,3-Dipo
the azide a
(see
ate pentahy
(II) species
is often us
Sawa, histid
ucose and n
cetylide wit
To release
protonated.5
olar Cycload
and alkyne i
ydrate is us
is reduced
sed to stabi
dine was us
naphthalim
th the naph
the produc
58
ddition
into a triazo
sed as a pre
d to copper
lize the solu
sed to incre
ide probe. T
thalimide p
ct, the Cu-C
ole requires
ecursor to
(I) by sodiu
ution gener
ase the
The copper
probe which
C bond of th
51
s the
um
rated
r-
h can
he 5-
u
A
A
B
a
e
h
th
w
fl
c
d
fl
c
H
Subst
sed include
Aesar, 98%)
Aldrich, 99%
Biotech), as
The c
nd synthes
thynyl grou
istidine for
he amount
were in a so
uorimeter,
ollected. In
emonstratin
uorimeter w
omplication
Horiba Fluor
trate synthe
ed 6-deoxy-
). The Huis
%), copper(I
corbic acid
4.2.2
click reactio
sized napht
up, two equ
every copp
of copper. T
olution of 56
the reaction
itial reactio
ng a signal
was switche
ns and fluor
rolog-3 Fluo
4
4
esis began
-L-galactos
gen 1,3-dip
II) sulfate p
(Sigma Ald
Click Reac
ons were pe
halimide pr
ivalents of
per, and sod
The reactio
6% (v/v) DM
ns were exc
ons showed
4 times tha
ed to a well
rescence ex
orimeter wi
Meth4.2
4.2.1 Mate
with L-gala
se (MP Biom
polar cycloa
entahydrate
drich), and
ctions and
erformed be
robe. Typica
copper for
dium ascor
ons were lef
MSO and w
cited at 357
promise w
at of the pro
plate read
xperiments
th a 450W
hods
erials
actose (Car
medicals), L
addition req
e (Mallinck
dimethyl su
fluorescent
etween a sy
al reactions
every alkyn
rbate used
ft at room t
water. Using
7nm and th
with the click
obe by itse
er there we
s were cont
Xe arc lam
rbosynth), o
L-(-)-galact
quires L-his
krodt), HEP
ulfoxide (M
t detection
ynthesized
s used one
ne, 2 equiva
in an exces
temperature
g a Cary Ec
he emission
ked materia
lf. However
ere alignme
inued with
mp.
other sugar
ose (Alfa
stidine (Sigm
ES (Fisher
acron, 99.9
azido fuco
azide for e
alents of
ss of five tim
e and usua
clipse
n spectra we
al
r, when the
ent
a Jobin-Yvo
52
rs
ma
9%).
se
every
mes
lly
ere
e
on
53
4.2.3 Probe Synthesis
Synthesis of the alkyne probe structure began with the commercially
available 4-Bromo-N-ethyl-1,8-napthalimide compound and followed the
procedure outlined in the supporting information from Sawa and co-workers.55
The structure was confirmed through NMR analysis using a Bruker ARX400 with
a qnp probe. 4-ethynyl-N-ethyl-1,8-naphthalimide (C16H13N1O2): 1H NMR (400
MHz, CD2Cl2): 8.68 (dd, J = 1.11, 8.40 Hz, 1H), 8.61 (dd, J = 1.13, 7.29 Hz, 1H),
8.51 (d, J = 7.57, 1H), 7.95 (d, J = 7.56 Hz, 1H), 7.85 (dd, J=7.31, 8.34 Hz, 1H),
4.20 (q, J = =7.11, 7.11, 7.10 Hz, 2H), 3.81 (s, 1H), 1.29 (t, J = 7.08, 7.08Hz, 3H)
4.2.4 Modified Sugar Synthesis
Starting from the commercially available L-galactose, and azido tagged
fucose was synthesized by following the previous published methods.59 The
structure was confirmed through NMR analysis using a Bruker ARX400 with a
qnp probe. 6-azido-1,2,3,4-tetra-O-acetyl-6-deoxy-a,b-L-galactopyranose
(C14H19N3O9): 1H NMR (400 MHz, CD2Cl2) 5.50 (d, J = 5.0 Hz, 1H), 4.61 (dd, J
= 7.9, 2.5 Hz, 1H), 4.32 (dd, J = 5.0, 2.5 Hz, 1H), 3.93 – 3.87 (m, 1H), 3.46 (dd, J
= 12.8, 8.2 Hz, 1H), 3.30 (dd, J = 12.8, 4.8 Hz, 1H), 1.31 (s, 3H), 1.42 (s, 3H),
1.51 (s, 3H), 1.55 (s, 3H). 13C NMR (400 MHz, CDCl3) 170.23-169.06, 92.26,
73.32, 70.96, 67.81, 67.64, 50.19, 21.01-20.65. 1,2,3,4-tetra-O-acetyl-6-deoxy-
a,b-L-galactopyranose (C15H22O9): 13C NMR (400 MHz, CDCl3) 170.68-169.29,
90.12, 70.73, 67.97, 67.43, 66.62, 21.06-16.07.
54
4.2.5 Feeding Plant Protoplast Cells
Many of the click reaction experiments performed in the literature use
millimolar concentrations of the material. Due to the expensive starting material,
and expected fluorescent intensity, micromolar concentrations of reactants were
used for these experiments. The first feeding experiment used 25M azido
fucose added to the feeding media. But due to poor click reactions, the feeding
experiment was repeated with a much larger amount of azido fucose, 200M
(See Table 4-1) to increase activity.
Table 4-1: Setup of Feeding Reactions
(+) aAF (+) Fucose (-) Fucose
Feeding 1
Culture Media Cultured Cell
Suspension 25uM Peracetylated
– Azido Fucose
Culture Media Cultured Cell
Suspension 25uM Peracetylated
-Fucose
Culture Media
Cultured Cell Suspension
Feeding 2
Culture Media 3mL of washed
Cultured Cell Suspension
200uM Peracetylated –Azido Fucose
Culture Media 3mL of washed
Cultured Cell Suspension
200uM Peracetylated -Fucose
Not used
The growth of protoplasts was monitored by observing the amount of light
that would pass through a suspension of protoplasts. As the cells continued to
grow, the amount in solution would increase and less light would pass through
the solution. This offered a quick and easy way to monitor cell growth over time.
F
T
te
A
s
w
o
c
w
fl
s
a
Figure 4-2: 6
The Arabido
echniques,
After five da
amples of m
The a
were each s
f the individ
ounts per s
without suga
uorescent s
ugar a sign
nd compar
0.0
0.2
0.4
0.6
0.8
1.0
Ab
sorb
ance
600nm OD
opsis cells w
the cells w
ys of growt
media, prot
alkyne prob
suspended
dual compo
second. The
ars. Howev
signal incre
nal of 18E+0
ing the high
00
20
40
60
80
00
0 1
of Feeding
were culture
ere allowed
th (see Figu
toplast and
be and azide
in a 56% (v
onents of th
e 4ENP pro
er in the pr
eased to 12
06 counts p
hest and low
2 3
Days
g-2 Cultured
ed by the S
d to incorpo
ure 4-2), ce
wall.
Res4.3
e tagged su
v/v) DMSO
e Click rea
obe gave al
resence of a
E+06. And
per second
west achiev
3 4
d Arabidops
Szymanski l
orate the az
ells were ha
ults
ugar were s
and analyz
action gave
lmost 9E+0
a non-tagge
in the pres
was achiev
ved maxim
5
F 2fuc
AFfuc
F 1fucafte
AFfucafte
sis Cells
lab. Using s
zido fucose
arvested an
synthesized
zed with a f
less than 1
06 counts w
ed sugar, fu
sence of the
ved. Lookin
um intensit
2, peracetylatcose, sampled
F 2, peracetylacose, sampled
1, peracetylatcose, sampleder 5days
F 1, peracetylacose, sampleder 5 days
sterile
over five d
nd divided in
d, and they
fluorimeter.
1.00E+06
when in solu
ucose, the
e azide tag
ng at Table
ty, the
edd daily
ated azidod daily
edd once
ated azidod once
55
days.
nto
All
ution
ged
4-2
56
fluorescent probe, an aAF clicked reaction and a controlled fucose clicked
reaction, the range of intensities has a very broad range for the azide-alkyne click
reaction. This inconsistency between fluorescent readings of samples is evident
throughout all experiments.
Table 4-2: Summary of Maximum Intensities for Probe and Click Reactions
Compound Highest Maximum
Intensity Lowest Maximum
Intensity Difference
Naphthalimide Probe
13.7E+06 8.2E+06 05.5E+06
Reacted Azido Fucose
20.5E+06 6.3E+06 14.2E+06
Reacted Fucose 14.5E+06 8.5E+06 6.0E+06
4.3.1 Fluorescence of Alkyne Probe
The individual alkyne naphthalimide probe will fluoresce on its own. When
the azide ring is formed the fluorescence of the aromatic structure and the azide
ring should be much greater than that of the structure without the azide ring.
4ENP was dissolved in a DMSO solution and fluoresced. Looking at six different
trials of fluoresced 4ENP solutions in Figure 4-3 you can see most trials fall
around a maximum emission of 8E+06 counts per second (c.p.s) with one trial
emitting a maximum intensity of 13E+06.
57
Figure 4-3: Emission Spectrum for 200M Naphthalimide Probe in 56%DMSO Excited at 357nm
When the concentration of 4ENP was cut in half the variability in emission
remained with maximum intensities of 8, 9 at 10E+06 c.p.s. (Figure 4-4).
Figure 4-4: Emission Spectrum for 100M Naphthalimide Probe in 56%DMSO Excited at 357nm
00.0E+00
04.0E+06
08.0E+06
12.0E+06
16.0E+06
20.0E+06
372
390
408
426
444
462
480
498
516
534
552
570
588
606
624
642
660
678
696
c.p.s.
Wavelength, nm
Trial 1
Trial 2
Trial 3
Trial 4
Trial 5
Trial 6
00.0E+00
04.0E+06
08.0E+06
12.0E+06
16.0E+06
20.0E+06
372
388
404
420
436
452
468
484
500
516
532
548
564
580
596
612
628
644
660
676
692
c.p.s.
Wavelength, nm
Trial 1
Trial 2
Trial 3
58
The forms of the peaks also changed with concentration, from 200M to 100M,
going from one or two features to three peaks.
When the sugars were added to the 4ENP solution the shape of the peak
changes but the variability in maximum emissions remains the same. For the
higher concentration of azido sugar and probe, 200M, the highest fluorescent
emission was 20E+06 c.p.s (Figure 4-5).
Figure 4-5: Emission Spectrum for 200M of Azide Tagged and Untagged Sugar Clicked in 56%DMSO and Excited at 357nm
The lowest emission for a clicked trial was 12E+06, which is very close in
value to the maximum for 200M 4ENP by itself (13.7E+06 c.p.s.). For the lower
concentration of azido sugar and probe, the maximum fluorescent emission is
00.0E+00
04.0E+06
08.0E+06
12.0E+06
16.0E+06
20.0E+06
372
388
404
420
436
452
468
484
500
516
532
548
564
580
596
612
628
644
660
676
692
c.p
.s.
Wavelength, nm
aAF Click Trial 1
aAF Click Trial 2
aAF Click Trial 3
Fucose Click Trial 1
Fucose Click Trial 2
59
19.5E+06 c.p.s. Repeating this trial again resulted in a maximum emission
of 16.5E+06. For both concentrations, adding an azido sugar to the 4ENP
solution produced a large signal than when a fucose sugar, without an azido
group, was added to the reaction.
Figure 4-6: Emission Spectrum for 100M of Azide Tagged and Untagged Sugar Clicked in 56%DMSO and Excited at 357nm
00.0E+00
04.0E+06
08.0E+06
12.0E+06
16.0E+06
20.0E+06
372
388
404
420
436
452
468
484
500
516
532
548
564
580
596
612
628
644
660
676
692
c.p
.s.
Wavelength, nm
aAF Click Trial 1
aAF Click Trial 2
Fucose Click Trial 1
60
Table 4-3: Maximum Intensities for Azido Tagged and Untagged Sugars at Different Concentrations and in Different Solvents
Entry Clicked sugar type
Concentration, M
Max Intensity, c.p.s. Solvent
1 aAF 200 20.4E+06 DMSO 2 aAF 200 12.6E+06 DMSO 3 aAF 200 20.4E+06 DMSO 4 aAF 100 19.5E+06 DMSO 5 aAF 100 16.5E+06 DMSO 6 aAF 200 16.1E+06 HEPES 7 aAF 200 20.5E+06 HEPES 8 aAF 100 6.30E+06 HEPES 9 aAF 100 18.9E+06 HEPES 10 Fucose 200 8.78E+06 DMSO 11 Fucose 200 11.2E+06 DMSO 12 Fucose 100 8.50E+06 DMSO 13 Fucose 200 14.0E+06 HEPES 14 Fucose 200 14.5E+06 HEPES 15 Fucose 200 11.1E+06 HEPES
These experiments were also repeated in a buffered solution of HEPES instead
of a DMSO solution. High and low maximum signals (20E+06 and 16E+06 for
200M reactants) were also achieved in the different solvent. Table 4-2
summarizes the maximum intensities for azido fucose and plain fucose reactions
with 4ENP in DMSO and HEPES.
4.3.2 Fluorescence of Azido Sugar Fed Plant Cells
As described in the methods section, protoplast plant cells were cultured
in a media infused with the azido tagged fucose or untagged fucose. Once the
cells multiplied in the sugar the mixtures were separated. The media was
decanted from the pelleted cells. The cells were then crushed to release their
61
insides and the internal protoplasts were separated from their walls. Each of
these three fractions was used in a click reaction to detect for the presence of
azido fucose.
The media fraction was analyzed in a dilution of DMSO and with HEPES
buffer (Figure 4-7). The different solvents caused a slight shift in the maximum
intensity wavelength emission; otherwise only a small change was observed
between the azido clicked and controlled reaction (17E+06 vs. 15E+06 c.p.s.)
Figure 4-7: Emission Spectrum for Click Reaction in Media (100M) Fractions Excited at 357nm
When the azido fucose protoplast fraction and control fraction were
subject to click reaction conditions, both fluorescent trials produced a maximum
emission of around 19E+06 c.p.s.
00.0E+00
04.0E+06
08.0E+06
12.0E+06
16.0E+06
20.0E+06
372
388
404
420
436
452
468
484
500
516
532
548
564
580
596
612
628
644
660
676
692
c.p
.s.
Wavelength, nm
aAF Media Fraction Clicked in HEPES
Fucose Media Fraction Clicked in HEPES
aAF Media Fraction Clicked in DMSO
Fucose Media Fraction Clicked in DMSO
62
Figure 4-8: Emission Spectrum for Protoplast Fractions (200M) Excited at 357nm
As shown in Figure 4-8, both peaks had the same basic shape. In addition the
maximum emissions for the protoplast fractions were well above the intensity
achieved for the probe by itself in a DMSO solution (average of 9.6E+06 c.p.s.).
Finally, the fractions with the wall fragments were analyzed for click
fluorescence. Just like the peaks in Figure 4-8 the azido reaction and control
reaction gave very similar peaks at relatively high intensities. In Figure 4-9 the
two trials have a maximum intensity of 14E+06 c.p.s. which is lower than the
intensity from the protoplast fractions. The protoplast and wall fractions also differ
in the shapes of their peaks.
00.0E+00
04.0E+06
08.0E+06
12.0E+06
16.0E+06
20.0E+06
372
390
408
426
444
462
480
498
516
534
552
570
588
606
624
642
660
678
696
c.p
.s.
Wavelength, nm
aAF Protoplast Fraction Clicked in HEPES
Fucose Protoplast Fraction Clicked in HEPES
63
Figure 4-9: Emission Spectrum for Wall Fractions (100M) Excited at 357nm
The maximum intensities and the wavelengths emitted are shown in
Table 4-4. The ratios of azido fucose to control fucose intensities were also
calculated.
Table 4-4: Summary of Maximum Intensities, Wavelengths, and Tagged/ Untagged Ratios for Click Reactions in Plant Fractions
Azido Fucose Fucose
Fraction Max Wavelength
Max Intensity
Max Wavelength
Max Intensity
Ratio of Intensities
Media (in HEPES) 452 17.4E+06 450 16.2E+06 1.08
Media (in DMSO) 454 17.0E+06 454 14.8E+06 1.15
Protoplast 418 18.7E+06 414 19.3E+06 0.97
Wall 430 13.7E+06 430 14.4E+06 0.96
00.0E+00
04.0E+06
08.0E+06
12.0E+06
16.0E+06
20.0E+06
372
386
400
414
428
442
456
470
484
498
512
526
540
554
568
582
596
610
624
638
652
666
680
694
c.p
.s.
Wavelength, nm
aAF Wall Fraction Clicked in DMSO
Fucose Wall Fraction Clicked in DMSO
64
And all of the reactions give an almost 1:1 ratio indicating there was no large
change in the amount of fluorescence given from an azide-alkyne clicked
reaction.
Figure 4-10: Emission Spectrum for Reactions in Plant Feeding Media and Reactions with Feeding Experiment Blank Samples Excited at 357nm
Lastly when the emission spectrums for the media samples and model
click reaction material without plant material are compared we can see the
similarities of all their emission spectrum. There is no clear delineation between
the trials that have an azide and trials that do not have one.
4.3.3 Fluorescent Signal from Controls
In an effort to investigate further the limits of the click reaction a different
azide complex was used to analyze the reactivity of the probe. These trials would
offer some insight to the possibility of an inactive azido tagged sugar. Figure 4-11
shows the emission for a tosyl azide click reaction with the 4ENP probe. When
00.0E+00
04.0E+06
08.0E+06
12.0E+06
16.0E+06
20.0E+0637
238
840
442
043
645
246
848
450
051
653
254
856
458
059
661
262
864
466
067
669
2
c.p
.s.
Wavelength, nm
200uM Azido Fucose in Media
100uM Azido Fucose in Media
Azido Fucose Blank
200uM Fucose in Media
Fucose Blank
65
that emission is compared to different emissions of the probe by itself, it
becomes difficult to determine if the reaction worked or not.
Figure 4-11: Emission Spectrum for 200M Tosyl Azide Clicked in 56%DMSO Excited at 357nm
However compared to previous click reactions with just azido sugar
(Figure 4-3) all of the maximum intensities are very low (8E+06 vs. 20E+06
c.p.s.). Table 4-5 compares four trials of tosyl azide reacted with the 4ENP probe
and it appears the reaction works one in four attempts.
Table 4-5: Summary of Maximum Intensities for Click Reactions with Control, Tosyl Azide
Entry Tosyl Azide Click Reaction Intensity, c.p.s.
4ENP in DMSO Intensity, c.p.s. Ratio
1 8.68E+06 8.73E+05 9.94
2 1.01E+07 9.96E+06 1.01
3 1.00E+07 1.01E+07 0.99
4 9.15E+06 5.92E+06 1.55
00.0E+00
04.0E+06
08.0E+06
12.0E+06
16.0E+06
20.0E+06
372
390
408
426
444
462
480
498
516
534
552
570
588
606
624
642
660
678
696
c.p.s.
wavelength, nm
TS‐N3 Clicked in DMSO
4ENP Trial 1
4ENP Trial 2
D
c
o
m
s
b
c
F(1tr
th
c
The b
DMSO and w
licked prob
rder of mag
many structu
ignal is not
ackground
licked and
Figure 4-12:1a) is the Nriazole55
When
he results d
ompared a
best results
water. Ther
e, however
gnitude (1E
ures within
large enou
signal. Fig
unclicked m
: The DifferNaphthalimid
n trying to r
did not show
click react
4.
s were gene
re was a dif
r that differe
E+07 to 1E+
the comple
ugh any sig
ure 4-12 sh
material from
rence in Flude probe a
react the az
w any cons
ion to a bla
Discus4
erated from
fference in
ence was s
+08) the pla
ex plant cel
nals from th
hows the di
m Sawa an
uorescence nd (3a) is th
zide and alk
istent react
ank reaction
ssion
the model
signal betw
small (9E+0
ant system
l that will flu
he click rea
fference in
nd co-worke
Intensity Ahe reacted
kyne togeth
tion results.
n, or a react
compound
ween the pr
06 to 18E+0
would need
uoresce an
action will b
intensities
ers.
Achieved by
material fo
her in a stan
. Furthermo
tion without
ds mixed in
robe and th
06) and not
d. There ar
nd if the pro
be lost in the
between
y Sawa et aorming the
ndard react
ore when yo
t the azide,
66
he
the
re
obe
e
al.
tion,
ou
, the
67
fluorescent intensity normally only doubled. As of now, no increase of ten times
achieved by Sawa has been able to be reproduced.
All of the individual components, when excited at 357nm, are well below
the intensity of the unclicked naphthalimide probe. A previously synthesized tosyl
azide compound was used to test the fidelity of the naphthalimide probe (see
Figure 4-11). However results from these experiments showed inconsistencies
as well.
4.4.1 Improvements and Changes
One of the major impedances to these experiments was the amounts of
material used. Due to the cost of synthesizing the azide tagged sugar, small
amounts of sugar were used with the cells. Further collaborations and study of
the amount of sugar becoming incorporated into the cells may have indicated
that very small amounts of azide were incorporating into the cells. The use of
very small concentrations of the control compounds indicated unreliable results
and weak signals. Also the low intensities from the azide tagged sugar and the
control tosyl azide, could have indicated inadequacies with the synthesized probe.
Further purification may have been necessary as impurities may have been
interfering with the click reactions.
Finally, commerically available probes can be purchased to be used with
the Sawa reaction conditions. If intensities of 10 fold or greater cannot be
achieved, commercial azides as well as alkynes may need to be used to further
this work.
c
in
s
to
re
Inves
hallenging.
n emission
ugar was in
ool and rep
esponses fo
stigations in
The Cu ca
intensities f
ntroduced t
eated react
or clicked a
4.5
nto a fluores
atalyzed clic
for during id
o plant mat
tions were
and unclicke
Conclu5
scent taggin
ck reactions
dealized re
terial the cl
unable to p
ed material
usions
ng system f
s showed s
action cond
ick reaction
produce con
.
for plant ce
some mode
ditions. How
ns became
nsistent fluo
ell walls pro
erate chang
wever when
an unreliab
orescence
68
oved
es
n the
ble
1
LIST OF REFERENCES
69
LIST OF REFERENCES
(1) Van Haveren, J.; Scott, E. L.; Sanders, J. Bulk Chemicals from Biomass. Biofuels, Bioprod. Biorefining 2008, 2, 41–57.
(2) Alonso, D. M.; Bond, J. Q.; Dumesic, J. A. Catalytic Conversion of Biomass to Biofuels. Green Chem. 2010, 12, 1493.
(3) OAK RIDGE NATIONAL LABORATORY. Biomass Program: About the Program. U.S. Department of Energy: Energy Efficiency & Renewable Energy http://www1.eere.energy.gov/biomass/about.html (accessed Oct 28, 2010).
(4) Yang, F.; Fu, J.; Mo, J.; Lu, X. Synergy of Lewis and Brønsted Acids on Catalytic Hydrothermal Decomposition of Hexose to Levulinic Acid. Energy & Fuels 2013, 27, 6973–6978.
(5) Naik, S. N.; Goud, V. V.; Rout, P. K.; Dalai, A. K. Production of First and Second Generation Biofuels: A Comprehensive Review. Renew. Sustain. Energy Rev. 2010, 14, 578–597.
(6) Smith, C. J. Chapter 4, Carbohydrate Biochemistry, The Cell Wall. In Plant Biochemistry and Molecular Biology; Lea, P. J.; Leegood, R. C., Eds.; John Wiley & Sons Ltd.: New York, NY, 1999; pp. 108–118.
(7) Kupiainen, L.; Ahola, J.; Tanskanen, J. Kinetics of Glucose Decomposition in Formic Acid. Chem. Eng. Res. Des. 2011, 89, 2706–2713.
(8) Mosier, N.; Wyman, C.; Dale, B.; Elander, R.; Lee, Y. Y.; Holtzapple, M.; Ladisch, M. Features of Promising Technologies for Pretreatment of Lignocellulosic Biomass. Bioresour. Technol. 2005, 96, 673–686.
(9) Cai, C. M.; Nagane, N.; Kumar, R.; Wyman, C. E. Coupling Metal Halides with a Co-Solvent to Produce Furfural and 5-HMF at High Yields Directly from Lignocellulosic Biomass as an Integrated Biofuels Strategy. Green Chem. 2014, 16, 3819–3829.
70
(10) Zakzeski, J.; Bruijnincx, P. C. A.; Jongerius, A. L.; Weckhuysen, B. M. The Catalytic Valorization of Lignin for the Production of Renewable Chemicals. Chem. Rev. 2010, 110, 3552–3599.
(11) Parsell, T. H.; Owen, B. C.; Klein, I.; Jarrell, T. M.; Marcum, C. L.; Haupert, L. J.; Amundson, L. M.; Kenttämaa, H. I.; Ribeiro, F.; Miller, J. T.; et al. Cleavage and Hydrodeoxygenation (HDO) of C–O Bonds Relevant to Lignin Conversion Using Pd/Zn Synergistic Catalysis. Chem. Sci. 2013, 4, 806.
(12) Rosatella, A. A.; Simeonov, S. P.; Frade, R. F. M.; Afonso, C. A. M. 5-Hydroxymethylfurfural (HMF) as a Building Block Platform: Biological Properties, Synthesis and Synthetic Applications. Green Chem. 2011, 13, 754.
(13) Fringuelli, F.; Pizzo, F.; Vaccaro, L. Lewis-Acid Catalyzed Organic Reactions in Water. The Case of AlCl 3 , TiCl 4 , and SnCl 4 Believed To Be Unusable in Aqueous Medium †. J. Org. Chem. 2001, 66, 4719–4722.
(14) Kim, B.; Jeong, J.; Lee, D.; Kim, S.; Yoon, H.-J.; Lee, Y.-S.; Cho, J. K. Direct Transformation of Cellulose into 5-Hydroxymethyl-2-Furfural Using a Combination of Metal Chlorides in Imidazolium Ionic Liquid. Green Chem. 2011, 13, 1503–1506.
(15) P. Lea; R.C. Leegood. Plant Biochemistry and Molecular Biology; 2nd Ed.; John Wiley and Sons Ltd.: New York, NY, 1999; pp. 108–117.
(16) Atkins, P.; Overton, T.; Rourke, J.; Weller, M.; Armstrong, F. Shriver & Atkins Inorganic Chemistry; 4th ed.; Oxford University Press: New York, NY, 2006; p. 822.
(17) Choudhary, V.; Mushrif, S. H.; Ho, C.; Anderko, A.; Nikolakis, V.; Marinković, N. S.; Frenkel, A. I.; Sandler, S. I.; Vlachos, D. G. Insights into the Interplay of Lewis and Brønsted Acid Catalysts in Glucose and Fructose Conversion to 5-(Hydroxymethyl)furfural and Levulinic Acid in Aqueous Media. J. Am. Chem. Soc. 2013, 135, 3997–4006.
(18) Corma, A.; Iborra, S.; Velty, A. Chemical Routes for the Transformation of Biomass into Chemicals. Chem. Rev. 2007, 107, 2411–2502.
(19) Jadhav, H.; Pedersen, C. M.; Sølling, T.; Bols, M. 3-Deoxy-Glucosone Is an Intermediate in the Formation of Furfurals from D-Glucose. ChemSusChem 2011, 4, 1049–1051.
71
(20) Yang, Y.; Hu, C.; Abu-Omar, M. M. Conversion of Carbohydrates and Lignocellulosic Biomass into 5-Hydroxymethylfurfural Using AlCl3·6H2O Catalyst in a Biphasic Solvent System. Green Chem. 2012, 14, 509.
(21) Climent, M. J.; Corma, A.; Iborra, S. Conversion of Biomass Platform Molecules into Fuel Additives and Liquid Hydrocarbon Fuels. Green Chem. 2014, 16, 516.
(22) Dapsens, P. Y.; Mondelli, C.; Pe, J. Biobased Chemicals from Conception toward Industrial Reality : Lessons Learned and To Be Learned. ACS Catal. 2012, 1487–1499.
(23) Guan, J.; Cao, Q.; Guo, X.; Mu, X. The Mechanism of Glucose Conversion to 5-Hydroxymethylfurfural Catalyzed by Metal Chlorides in Ionic Liquid: A Theoretical Study. Comput. Theor. Chem. 2011, 963, 453–462.
(24) Kim, E. S.; Liu, S.; Abu-Omar, M. M.; Mosier, N. S. Selective Conversion of Biomass Hemicellulose to Furfural Using Maleic Acid with Microwave Heating. Supporting Info. Energy & Fuels 2012, 26, 1298–1304.
(25) Pagan-Torres, Y. J.; Wang, T.; Gallo, J. M.; Shanks, B. H.; Dumesic, J. A. Production of 5-Hydroxymethylfurfural from Glucose Using a Combination of Lewis and Brønsted Acid Catalysts in Water in a Biphasic Reactor with an Alkylphenol Solvent. ACS Catal. 2012, 930–934.
(26) Kamireddy, S. R.; Li, J.; Tucker, M.; Degenstein, J.; Ji, Y. Effects and Mechanism of Metal Chloride Salts on Pretreatment and Enzymatic Digestibility of Corn Stover. Ind. Eng. Chem. Res. 2013, 52, 1775–1782.
(27) Jia, S.; Liu, K.; Xu, Z.; Yan, P.; Xu, W.; Liu, X.; Zhang, Z. C. Reaction Media Dominated Product Selectivity in the Isomerization of Glucose by Chromium Trichloride: From Aqueous to Non-Aqueous Systems. Catal. Today 2014, 234, 83–90.
(28) Zheng, B.; Fang, Z.; Cheng, J.; Jiang, Y. Microwave-Assisted Conversion of Carbohydrates into 5-Hydroxymethylfurfural Catalyzed by ZnCl2. Z. Naturforsch 2010, 65b, 168–172.
(29) Pidko, E. a.; Degirmenci, V.; Hensen, E. J. M. On the Mechanism of Lewis Acid Catalyzed Glucose Transformations in Ionic Liquids. ChemCatChem 2012, n/a–n/a.
(30) Zhao, H.; Holladay, J. E.; Brown, H.; Zhang, Z. C. Metal Chlorides in Ionic Liquid Solvents Convert Sugars to 5-Hydroxymethylfurfural. Science 2007, 316, 1597–1600.
72
(31) Jessop, P. G. Searching for Green Solvents. Green Chem. 2011, 13, 1391.
(32) Watanabe, K.; Yamagiwa, N.; Torisawa, Y. Cyclopentyl Methyl Ether as a New and Alternative Process Solvent. Org. Process Res. Dev. 2007, 11, 251–258.
(33) Yang, Y.; Hu, C.; Abu-Omar, M. M. Conversion of Glucose into Furans in the Presence of AlCl3 in an Ethanol-Water Solvent System. Bioresour. Technol. 2012, 116, 190–194.
(34) Yang, Y.; Hu, C.; Abu-Omar, M. M. The Effect of Hydrochloric Acid on the Conversion of Glucose to 5-Hydroxymethylfurfural in AlCl3–H2O/THF Biphasic Medium. J. Mol. Catal. A Chem. 2013, 376, 98–102.
(35) Tong, X.; Li, M.; Yan, N.; Ma, Y.; Dyson, P. J.; Li, Y. Defunctionalization of Fructose and Sucrose: Iron-Catalyzed Production of 5-Hydroxymethylfurfural from Fructose and Sucrose. Catal. Today 2011, 175, 524–527.
(36) Zhao, H.; Holladay, J. E.; Brown, H.; Zhang, Z. C. Metal Chlorides in Ionic Liquid Solvents Convert Sugars to 5-Hydroxymethylfurfural. Science 2007, 316, 1597–1600.
(37) Vom Stein, T.; Grande, P. M.; Leitner, W.; de María, P. D. Iron-Catalyzed Furfural Production in Biobased Biphasic Systems: From Pure Sugars to Direct Use of Crude Xylose Effluents as Feedstock. ChemSusChem 2011, 4, 1592–1594.
(38) Szabolcs, Á.; Molnár, M.; Dibó, G.; Mika, L. T. Microwave-Assisted Conversion of Carbohydrates to Levulinic Acid: An Essential Step in Biomass Conversion. Green Chem. 2013, 15, 439.
(39) Rasrendra, C. B.; Makertihartha, I. G. B. N.; Adisasmito, S.; Heeres, H. J. Green Chemicals from D-Glucose: Systematic Studies on Catalytic Effects of Inorganic Salts on the Chemo-Selectivity and Yield in Aqueous Solutions. Top. Catal. 2010, 53, 1241–1247.
(40) Qian, X. Mechanisms and Energetics for Brønsted Acid-Catalyzed Glucose Condensation, Dehydration and Isomerization Reactions. Top. Catal. 2012, 55, 218–226.
(41) Ståhlberg, T.; Rodriguez-Rodriguez, S.; Fristrup, P.; Riisager, A. Metal-Free Dehydration of Glucose to 5-(hydroxymethyl)furfural in Ionic Liquids with Boric Acid as a Promoter. Chemistry 2011, 17, 1456–1464.
73
(42) Strauss, C. R.; Rooney, D. W. Accounting for Clean, Fast and High Yielding Reactions under Microwave Conditions. Green Chem. 2010, 12, 1340.
(43) Zhao, J.; Yan, W. Microwave-Assisted Inorganic Syntheses. In Modern Inorganic Synthetic Chemistry; XU, R.; PANG, W.; HUO, Q., Eds.; Elsevier: Amsterdam, 2011; pp. 173–195.
(44) Gerbec, J. A.; Magana, D.; Washington, A.; Strouse, G. F. Microwave-Enhanced Reaction Rates for Nanoparticle Synthesis. J. Am. Chem. Soc. 2005, 127, 15791–15800.
(45) Stuerga, D. Microwave – Materials Interactions and Dielectric Properties : From Molecules and Macromolecules to Solids and Colloidal Suspensions. In Microwaves in Organic Synthesis; Hoz, A. de la; Loupy, A., Eds.; Wiley-VCH: Weinheim, Germany, 2012; pp. 1–56.
(46) Perreux, L.; Loupy, A.; Petit, A. Nonthermal Effects of Microwaves in Organic Synthesis. In Microwaves in Organic Synthesis; Hoz, A. de la; Loupy, A., Eds.; Wiley-VCH: Weinheim, Germany, 2012; pp. 127–207.
(47) Ritter, S. K. Microwave Chemistry Remains Hot, Fast, And A Tad Mystical | January 27, 2014 Issue - Vol. 92 Issue 4 | Chemical & Engineering News http://cen.acs.org/articles/92/i4/Microwave-Chemistry-Remains-Hot-Fast.html (accessed Mar 31, 2014).
(48) Kootstra, A. M. J.; Mosier, N. S.; Scott, E. L.; Beeftink, H. H.; Sanders, J. P. M. Differential Effects of Mineral and Organic Acids on the Kinetics of Arabinose Degradation under Lignocellulose Pretreatment Conditions. Biochem. Eng. J. 2009, 43, 92–97.
(49) Crumbliss, A. L.; Garrison, J. M. A Comparison of Some Aspects of the Aqueous Coordination Chemistry of Aluminum(III) and Iron(III). Comments Inorg. Chem. A J. Crit. Discuss. Curr. Lit. 1988, 8:1-2, 1–26.
(50) De Abreu, H. A.; Guimarães, L.; Duarte, H. A. Density-Functional Theory Study of Iron(III) Hydrolysis in Aqueous Solution. J. Phys. Chem. A 2006, 110, 7713–7718.
(51) Wulfsberg, G. Inorganic Chemistry; Sausalito CA, 2000; p. 978.
(52) Wu, G. J.; Frech, R. The Optical and Spectroscopic Properties of the Sulfate Ion in Various Crystalline Environments. J. Chem. Phys. 1977, 66, 1352.
74
(53) Shannon, R.; Fischer, R. Empirical Electronic Polarizabilities in Oxides, Hydroxides, Oxyfluorides, and Oxychlorides. Phys. Rev. B 2006, 73, 235111.
(54) Albersheim, P.; Darvill, A.; Roberts, K.; Serderoff, R.; Staehelin, A. Plant Cell Walls: From Chemistry to Biology; First.; Garland Science: New York, NY, 2011; p. 430.
(55) Sawa, M.; Hsu, T.-L.; Itoh, T.; Sugiyama, M.; Hanson, S. R.; Vogt, P. K.; Wong, C.-H. Glycoproteomic Probes for Fluorescent Imaging of Fucosylated Glycans in Vivo. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 12371–12376.
(56) Ninfa, A. J.; Ballou, D. P. Fundamental Laboratory Approaches for Biochemistry and Biotechnology; Fitzgerald Science Press. Inc.: Bethesda, MD, 1998; p. 372.
(57) Rostovtsev, V. V; Green, L. G.; Fokin, V. V; Sharpless, K. B. A Stepwise Huisgen Cycloaddition Process: copper(I)-Catalyzed Regioselective “Ligation” of Azides and Terminal Alkynes. Angew. Chem. Int. Ed. Engl. 2002, 41, 2596–2599.
(58) Shao, C.; Wang, X.; Xu, J.; Zhao, J.; Zhang, Q.; Hu, Y. Carboxylic Acid-Promoted copper(I)-Catalyzed Azide-Alkyne Cycloaddition. J. Org. Chem. 2010, 75, 7002–7005.
(59) Laughlin, S. T.; Bertozzi, C. R. Metabolic Labeling of Glycans with Azido Sugars and Subsequent Glycan-Profiling and Visualization via Staudinger Ligation. Nat. Protoc. 2007, 2, 2930–2944.
75
VITA
75
VITA
Christine Marie Bohn is from Solon, a suburb of Cleveland (Cleveland
rocks!) in Ohio where she grew up with her mom, Louise, and brother, Tom. She
studied Chemistry at John Carroll University and received her Bachelors of
Science in Chemistry (and minor in English) in 2006. While at JCU she worked
with some copper ligands under the advisement of Dr. Catherine Miller. After
graduation she worked for STERIS Corporation for two years as an associate
scientist creating a portable electrochemical system to generate hydrogen
peroxide for the military. With encouragement from her supervisors she left to
continue her education at Purdue University working on catalytic dehydration of
sugars under Professor Mahdi Abu-Omar. In addition to Chemistry, she enjoys
running, knitting, and theater. While at Purdue, she also met and married her
loving husband, Chris.
PUBLICATION
DOI: 10.1002/cssc.201402530
Zinc-Assisted Hydrodeoxygenation of Biomass-Derived 5-Hydroxymethylfurfural to 2,5-DimethylfuranBasudeb Saha,*[a] Christine M. Bohn,[a] and Mahdi M. Abu-Omar*[a, b]
Introduction
Rapid depletion of nonrenewable petroleum reserves, the highvolatility of the crude oil price, and global warming due to in-creased carbon dioxide emissions have directed current re-search efforts towards the development of renewable alterna-tives to meet the growing energy demand for the future gener-ation.[1] Therefore, innovation of new synthetic routes and relat-ed technologies for generating fuels from biorenewable resour-ces is a cutting-edge research area. Currently, ethanol is theonly renewable liquid fuel produced on a commercial scale, pri-marily from food crops such as grains, sugar beet, and oilseeds. Due to competition with other land requirements and itsadverse effect on food production, it is realized that 1st-genera-tion bioethanol will not achieve targets for oil-product substitu-tion, climate change mitigation, and economic growth.[2] There-fore, the production of 2nd-generation biofuels from lignocellu-lose, such as low-cost crop and forest residues, wood chips,and municipal waste, has been targeted in recent years tomeet a significant portion of the 36 billion gallons (1 gallon=3.785 L) of cellulosic biofuels target to be produced by 2022.
Although ethanol production from 2nd-generation lignocellu-lose is gaining momentum to meet the production target, itshigh oxygen content (O/C=0.5) and low energy density(23.4 MJL�1 vs. 31 MJL�1 for gasoline) are seen as disadvantag-es. While the lower energy density of ethanol is largely offsetby its higher octane number (RON=110) as compared to gaso-line (RON=87–93), the average fuel economy of E15 fuel isstill about 5% lower than regular gasoline. Therefore, research-ers in both industry and academia are developing technologiesfor the next-generation advanced liquid fuels based on biore-newable platform chemicals, 5 hydroxymethylfurfural (HMF)and furfural (Ff).[3–4] In this context, the development of eco-nomically and environmentally viable routes for producing 2,5-dimethylfuran (DMF), 5-ethoxymethylfurtutal (EMF), ethyl levu-linate (EL), g-valerolactone (VL), and long-chain hydrocarbonsof diesel fractions has received significant attention.[5–9] It hasbeen reported that DMF is a superior liquid fuel compared toethanol because of its higher energy density (30 MJL�1),higher octane number (RON=119), and lower oxygen content(O/C=0.17).[10,11] Additionally, DMF is immiscible with waterand is easier to blend with gasoline than ethanol. Recently,DMF has been tested as a biofuel on a single-cylinder gasolinedirect-injection (GDI) research engine.[11] Test results of DMFwere satisfactory against gasoline in terms of combustion, igni-tion, and emission characteristics.
The chemical transformation of lignocellulose into DMF isa multistep process, involving (i) pretreatment of lignocelluloseinto glucose, (ii) acid-catalyzed dehydration of fructose to HMF,and (iii) catalytic hydrodeoxygenation (HDO) of HMF to DMF.[12]
Dumesic et al. reported a two-step process for the conversionof fructose to DMF (71% yield) involving Cu–Ru/C-catalyzed
2,5-Dimethylfuran (DMF), a promising cellulosic biofuel candi-date from biomass derived intermediates, has received signifi-cant attention because of its low oxygen content, high energydensity, and high octane value. A bimetallic catalyst combina-tion containing a Lewis-acidic ZnII and Pd/C components is ef-fective for 5-hydroxymethylfurfural (HMF) hydrodeoxygenation(HDO) to DMF with high conversion (99%) and selectivity(85% DMF). Control experiments for evaluating the roles ofzinc and palladium revealed that ZnCl2 alone did not catalyzethe reaction, whereas Pd/C produced 60% less DMF than thecombination of both metals. The presence of Lewis acidic com-ponent (Zn) was also found to be beneficial for HMF HDO with
Ru/C catalyst, but the synergistic effect between the two metalcomponents is more pronounced for the Pd/Zn system thanthe Ru/Zn. A comparative analysis of the Pd/Zn/C catalyst topreviously reported catalytic systems show that the Pd/Znsystem containing at least four times less precious metal thanthe reported catalysts gives comparable or better DMF yields.The catalyst shows excellent recyclability up to 4 cycles, fol-lowed by a deactivation, which could be due to coke forma-tion on the catalyst surface. The effectiveness of this combinedbimetallic catalyst has also been tested for one-pot conversionof fructose to DMF.
[a] Dr. B. Saha, C. M. Bohn, Prof. Dr. M. M. Abu-OmarDepartment of Chemistry and theCenter for Catalytic Conversion of Biomass to Biofuels (C3Bio)Purdue University560 Oval Drive, West Lafayette, IN 47907 (USA)E-mail : [email protected]
[b] Prof. Dr. M. M. Abu-OmarSchool of Chemical EngineeringPurdue University560 Oval Drive, West Lafayette, IN 47907 (USA)
Supporting Information for this article is available on the WWW underhttp://dx.doi.org/10.1002/cssc.201402530.
� 2014 Wiley-VCH Verlag GmbH&Co. KGaA, Weinheim ChemSusChem
0000,
00, 1 – 8
&1
CHEMSUSCHEMFULL PAPERS
76
HDO of HMF under 6.8 bar H2.[5] Raines et al. reported the pro-
duction of DMF from untreated corn stover, giving 9% DMFyield based on the cellulose content of the corn stover.[6]
Raines’s two-step process of DMF preparation involved theCrCl3–HCl catalyzed transformation of corn stover into HMF,followed by HDO of HMF to DMF using Cu–Ru/C catalyst inthe presence of H2. In this process, toxic chromium salt alongwith mineral acid was used as a catalyst for the degradation ofcorn stover into HMF. Sen et al. reported the conversion of car-bohydrates to 2,5-dimethyltetrahydrofuran (DMTHF)[13] in 81%yield using homogeneous RhCl3/HI catalyst. In the same year,Bell et al. attempted HMF HDO with activated carbon (AC) sup-ported palladium, ruthenium, and rhodium catalysts in 1-ethyl-3-methylimidazolium chloride ([EMIM]Cl) solvent under 62 barH2 pressure.[14] Among several reactions, the Pd/C-catalyzedHDO reported maximum 19% HMF conversion with 13% DMFselectivity. Although both HMF conversion and DMF selectivityimproved to 47% and 32%, respectively, when acetonitrile wasmixed with [EMIM]Cl solvent, total DMF yield (15%) was nothigh enough. An additional drawback of this method was thationic liquid decreased the solubility of H2. Hence, high pres-sures of H2 (62 bar) were required, making the process energy-intensive. Under similar reaction conditions, the authorsshowed that Ru/C catalyst failed to produce DMF from HMF.Most recently, Ru/Co3O4 has been reported as an effective cata-lyst for HMF HDO in THF.[15] The only disadvantage of this pro-cess is that the reaction requires high catalyst loading (40 wt%based on substrate loading).
Recently, we have demonstrated that Pd/C is very effectivecatalyst in the presence of a small amount of Lewis-acidic ZnCl2for HDO of monomeric lignin surrogate molecules.[16] Herein,we demonstrate the benefit of a Pd/Zn/C catalytic system forthe conversion of HMF to DMF under mild reaction conditionsand low H2 pressures. From the two solvent systems studied,tetrahydrofuran (THF) is a more effective solvent than metha-nol. The synergistic effect of ZnCl2 with Ru/C and Ni/C catalystshas also been investigated for HMF HDO under comparable re-action conditions. The results show that the Pd/Zn/C catalyst ismost effective for quantitative conversion of HMF to DMF withvery high selectivity. The effectiveness of the later catalyst isalso examined for one-pot conversion of fructose to DMF.
Results and Discussion
AC-supported palladium, rhodium, and ruthenium catalysts areeffective for hydrogenation of organic substrates at low tem-perature and pressure. However, these catalysts have provedineffective for hydrogenation of HMF even at high pressures ofH2 (62 bar).[14] A recent study has demonstrated a synergisticeffect between zinc and palladium, and reported a significantimprovement in yields and selectivity of the corresponding hy-drodeoxygenation products from their respective phenolic al-cohol and aldehyde precursors.[16] Mechanistic studies havesuggested Zn2+ adsorption onto AC, and the resulting bimetal-lic catalyst activated phenolic substrates via binding to �OHgroups and inducing reactivity with Pd�H sites on the surfacevia hydrogen spillover. To examine the catalytic effectiveness
of the combined ZnCl2–Pd/C catalyst for a simple reactionsystem, we investigated HMF HDO under mild reaction condi-tions.
Catalyst screening
A preliminary reaction for HMF HDO with combined Pd/C andZnCl2 catalytic species was carried out in MeOH solvent at150 8C and 20 bar H2 pressure by using 0.04 g Pd/C, 0.04 gZnCl2, and 0.2 g HMF (1.58 mmol). The yield of DMF was 39%after 2 h with about 75% conversion of HMF. The yield of DMFremained unchanged when the same reaction was repeatedfor 4 and 8 h. Based on previous reports,[18] it is hypothesizedthat etherification of HMF with MeOH in the presence of theLewis acid ZnCl2 results in the formation of 5-methoxymethyl-furfural (MMF), and hence blocks the hydroxymethyl group ofHMF from the desired HDO reaction. Therefore, we used THFas a solvent for HMF conversion by adopting the following ex-perimental conditions: HMF=3.96 mmol (0.50 g), Pd/C=0.05 g, ZnCl2=0.05 g, and THF=15 mL at 150 8C and 22 bar ofH2. The later reaction achieved 85% DMF yield with completeHMF conversion in 8 h (Supporting Information, Figure S3).This DMF yield is about 82% higher than that observed by Belland co-workers[14] using Pd/C catalyst alone at 62 bar H2 pres-sure. The GC–MS analysis of the product solution showed thepeaks for MTHFA and HD, accounting for their respectiveyields of 2.6% and 1.6%. Other products (OP) accounting for9–10% of total carbon mass balance corresponds to the un-identified peaks between retention times 9.6 to 12 min asshown in the GC chromatogram (Figure S3). Another reactionat lower H2 pressure (8 bar) gave similar DMF yield (84%). Acontrol experiment without ZnCl2 produced only 27% DMFunder comparable reaction conditions between HMF and Pd/Ccatalyst (Supporting Information, Figure S4). As evident by thecomparison in Figure 1, a small amount of ZnCl2 improvedDMF yield as much as 60%. ZnCl2 alone did not catalyze thereaction as HMF remained unconverted without Pd/C.
Assuming that the acidity of ZnCl2 could be the reason forthe enhanced activity of the Pd/Zn system, we tested the activ-ity of the Pd/C catalyst in the presence of a small amount ofAmberlyst-15 as co-catalyst. Under comparable conditions, a re-action using 0.025 g Amberlyst-15 in place of ZnCl2 resulted inthe formation of the furan ring-hydrogenation product of DMF,2,5-dimethyl tertrahydrofuran (DMTHF), as a major product(13%), along with small amount of DMF (6%) and other ring-hydrogenation intermediate products such as 2-hexanone(1.6%), tetrahydrofurfural alcohol (THFA) (3%), MTHFA (6%), 2-methyl-tetrahydrofuran-5-aldehyde (4.5%), 2,5-bis(dihydroxy)tetrahydromethyfuran (13%), and some unidentified peaks(Supporting Information, Figure S5). The formation of multiplering-hydrogenation products has also been observed for HMFHDO using HI/RhI3
[19] and ruthenium–porous metal oxide (Ru–PMO)[20] catalysts. This result demonstrates that the acidity ofZnCl2 is not the only reason for the enhanced activity of thecatalyst, resulting in high DMF yield, as the reaction in thepresence of Amberlyst-15 favored furan ring hydrogenationwith the formation of multiple products.
� 2014 Wiley-VCH Verlag GmbH&Co. KGaA, Weinheim ChemSusChem 0000,
00, 1 – 8
CHEMSUSCHEMFULL PAPERS www.chemsuschem.org
77
The beneficial effect of ZnCl2 has been further extended toRu/C- and Ni/C-catalyzed HDO of HMF. Bell et al. have shownthat Ru/C catalyst was ineffective for HMF HDO at high pres-sures of H2.
[14] Although Ru/C facilitates DMF production usingformic acid as a source of hydrogen in THF,[8] the present workshows that a combined catalytic system comprising ZnCl2 andRu/C is more effective than the earlier report. Under compara-ble reaction conditions, the combined ZnCl2-Ru/C catalyst ena-bles 41% DMF yield along with 52% yield of BHMF intermedi-ate product and quantitative conversion of HMF. In the ab-sence of ZnCl2, HMF HDO with Ru/C catalyst alone producesonly 3% DMF, which is significantly lower than the yield ob-tained with Pd/C catalyst (27%). While the percentage of HMFconversion with the Ru/C catalyst is moderately higher thanthe Pd/C, the former catalyst predominantly facilitates hydro-genation of HMF, resulting in the formation of a significantamount of BHMF (45%) intermediate and poor DMF selectivity.
When Lewis-acidic ZnCl2 was added along with Ru/C, theyield of DMF distinctly improved (41%), which is presumablydue to enhanced deoxygenation of BHMF in the presence ofZn2+ . Noteworthy, the conversion of HMF has also increased inthe presence of Zn2+ , suggesting that synergism of Zn2+ withruthenium plays a role in enhancing hydrogenation of HMF aswell. The latter reaction with the ZnCl2–Ru/C catalyst alsoshowed a significant amount of unconverted BHMF intermedi-ate (52%) for 8 h reaction. Assuming that continuing this reac-tion for a longer time could deoxygenate unconverted BHMFto DMF, we repeated this reaction for 20 h, which revealeda further conversion of BHMF and hence lowered the concen-tration of BHMF intermediate from 52% to 31%, but the over-all DMF yield (45%) did not improve to the same extent dueto the formation of other products (11% MTHFA and 7% 2,5-bis(dihydroxy) tetrahydromethyfuran). This result suggests thatBHMF deoxygenation with the Ru/Zn/C catalyst is slower thanthe Pd/Zn/C catalyst. The latter catalyst resulted in completeconversion of BHMF in 8 h with the formation of 88% DMF.
When comparing the activity of the Ru/C system with that ofPd/C, the latter catalyst provided a higher DMF yield than theformer (Figure 1). Therefore, it can be inferred that the palladiumcatalyst alone can catalyze both hydrogenation of HMF to BHMF
and deoxygenation of BHMF toDMF. In the presence of ZnCl2,the activity of the palladium cata-lyst for both hydrogenation ofHMF and deoxygenation of BHMFimproved to achieve significantlyhigher DMF yield (85%). Althoughtotal HMF conversions for bothZnCl2-Ru/C and ZnCl2-Pd/C cata-lyzed reactions were similar, thehigher DMF yield in the reactioncatalyzed by the latter is attribut-ed to a more pronounced syner-gistic effect for the Pd–Zn systemthan the Ru–Zn one.
To further understand the roleof the hydrogenation metal com-
ponent of the bifunctional catalyst in determining the overallHMF HDO, we studied HMF conversion with Ni/C catalyst in theabsence and presence of ZnCl2. Under comparable conditions,of 8 bar H2 and 1508C, a reaction between 3.96 mmol HMF and0.05 g Ni/C showed only 10% HMF conversion with the forma-tion of a trace amount of DMF. In the presence of ZnCl2(0.05 g), HMF conversion increased to 26%, although the yieldof DMF remained poor (7%). This comparison indicates that thebifunctional catalyst containing a poor hydrogenation metalcomponent, which is the case for Ni/C, shows poor synergismbetween the hydrogenation and Lewis acidic metal sites of thebifunctional catalyst. Based on a previous report showing thesynergy between tungsten oxide (WO3) and hydrogenationmetals,[21] we tested the activity of Pd/C catalyst in the presenceof a small amount of WO3 for HDO of HMF in THF. A reactionof 3.96 mmol HMF with a mixture of 0.1 g WO3 and 0.05 g Pd/Cat 1508C and 19 bar H2 for 8 h achieved 66 mol% DMF, 2% HD,14% BHMF and 1.4% 2-MF. Notably, the WO3 componentmostly remained undissolved in the solution, as was confirmedby performing a parallel experiment without using Pd/C. Thelatter reaction confirmed that WO3 did not catalyze the reactionby itself, but partially dissolved WVI species enhanced the activi-ty of the Pd/C component. A comparison of HMF conversionsand product distributions using different catalytic systems issummarized in Figure 1.
The effect of reaction temperature
The effectiveness of the ZnCl2-Pd/C catalyst was evaluated byperforming HMF HDO at varying reaction parameters such asH2 pressure and reaction temperature. The reaction tempera-ture was varied in the range of 120 to 2008C. As shown inFigure 2, the DMF yield increased from 69 to 84% upon increas-ing the reaction temperature from 120 to 1508C, but the yielddecreased to 58% when the temperature was further increasedto 200 8C. Overhydrogenation of the aromatic furan ring of DMFand other intermediate furanic species was the reason for thelower DMF yield at higher temperature. This interpretation wassupported by the fact that a GC chromatogram of the productsolution showed a new peak with a significant peak area close
Figure 1. The results of HMF hydrodeoxygenation using different catalysts. DMF=2,5-dimethylfuran; BHMF=2,5-bis(hydroxymethyl)furan; MTHFA=5-methyltetrahydrofurfural alcohol; HD=2,5-hexane dione; OP=otherproducts.
� 2014 Wiley-VCH Verlag GmbH&Co. KGaA, Weinheim ChemSusChem
0000,
00, 1 – 8 &3&
CHEMSUSCHEMFULL PAPERS www.chemsuschem.org
78
to the peak for DMF, along with other small peaks in the reten-tion time region 9–12 min. GC–MS revealed the mass numberof the new peak as 100, which could be either due to 2-methyltetrahydro-5-furanone or dihydrofurfural alcohol. The formationof similar dihydro species has been noted previously in the liter-ature.[13,14] However, further analysis of the reaction product by1H NMR spectroscopy did not provide sufficient informationabout the identity of the new species because of overcrowdedproton signals from multiple products.
The effect of H2 pressure
More experiments were designed for optimizing the reactionconditions, by varying the H2 pressure in the range of 2–40 barat a fixed reaction temperature (150 8C). The percentage ofHMF conversion and the corresponding DMF yields are shownin Figure 3. The yield of DMF increased from 29 to 84% uponincreasing H2 pressure from 2 to 8 bar, and remained un-changed up to 20 bar. A further increase of H2 pressure to40 bar did not significantly influence the overall DMF yield.
The effect of reaction time
To elucidate the reaction sequence, we conducted a reactionbetween 3.96 mmol HMF and ZnCl2 (0.05 g)–Pd/C (0.05 g) cata-lyst at 150 8C for multiple times by varying the reaction timefrom 1 h to 20 h. As shown in Figure 4, HMF HDO for 1 h pro-
duced 35% BHMF as intermediate along with 55% DMF. Theyield of BHMF intermediate decreased from 35% to 12% uponincreasing the reaction time from 1 h to 2.5 h, during whichthe yield of DMF also increased from 55% to 71%. A further in-crease in reaction time from 2.5 h to 8 h resulted in completeconversion of BHMF to DMF. The yield of DMF and the concen-tration of MTHFA and HD by-products remained unchanged
up to 20 h. To validate the formation of BHMF as intermediatein HMF HDO pathway, we performed a separate experimentusing BHMF as a starting substrate. Under comparable reactionconditions, BHMF deoxygenation with the ZnCl2-Pd//C catalystachieved 88% conversion in 8 h and produced 82% DMF (Sup-porting Information, Figure S6). This observation of HMF HDOsequence partly agrees with previous findings by Bell et al. ,[14]
in which the �CHO group of HMF is hydrogenated first toform BHMF as an intermediate (Scheme 1). Subsequent deoxy-genation of hydroxymethyl groups of BHMF results in the for-mation of DMF via 5-methylfurfural alcohol (MFA) as intermedi-ate. As MFA was not detected in the GC–MS spectra of productsolutions, it is believed that MFA was rapidly deoxygenated toDMF under the reaction conditions, as evidenced in earlier re-ports.[21b] The formation of small amounts of MTHFA and HDattributes to the fact that hydrogenation of furan ring of MFAand hydration of DMF took place during the reaction.[22–24] Theyields of MTHFA and HD were, however, below 3% even after20 h, which suggests that these side reactions occur at muchslower rates. Mechanistically, we postulate that the zinc com-ponent facilitates hydrogenolysis of BHMF to DMF by cleavageof C�O bonds, as discussed in our earlier report involving HDOof phenolic alcohols and aldehydes.[16] A similar observationwas made in the HDO of HMF with Ru/Co3O4 as catalyst, inwhich the ruthenium component hydrogenated HMF to BHMFand CoOx facilitated hydrogenolysis of BHMF to DMF.[15]
Comparison of catalyst activity
To explore the commercial potential of this process for DMFproduction, we prepared the Pd/Zn/C catalyst by impregnatingZn2+ on commercially purchased Pd/C by the incipient wet-ness method, as discussed in the Experimental section. Insteadof adding the Pd/C and ZnCl2 components separately, we ex-amined the effectiveness of as-synthesized Pd/Zn/C catalyst by
Figure 2. The effect of reaction temperature on HMF hydrodeoxygenation toDMF.
Figure 3. HMF hydrodeoxygenation as a function of H2 pressure.
Figure 4. Reaction profile of HMF hydrodeoxygenation with ZnCl2–Pd/C ascatalyst.
� 2014 Wiley-VCH Verlag GmbH&Co. KGaA, Weinheim ChemSusChem
0000,
00, 1 – 8 &4&
CHEMSUSCHEMFULL PAPERS www.chemsuschem.org
79
performing an experiment in which 3.96 mmol HMF was react-ed with 0.05 g catalyst at 150 8C and 8 bar H2 for 8 h. The activ-ity of the Pd/Zn/C catalyst was comparable to the ZnCl2-Pd/Ccatalytic system, giving 85% DMF. Additionally, the inherentfuel properties of small amounts of overhydrogenation prod-ucts, accounting for the carbon mass balance, eliminate thenecessity of an expensive separation step of DMF from theproduct stream.
The effectiveness of the Pd/Zn/C catalyst was also comparedto previously reported catalytic systems (Table 1). Among othereffective catalytic systems reported in the literature, Pd/C andRu/C catalysts (entries 7 and 8) using alternate hydrogen sour-ces, generated in situ from isopropanol and formic acid, gave70 and 81% DMF, respectively. However, in both cases the con-centrations of active metals in the catalysts were at least4 times higher than the present catalyst. Similarly, the dosageof both ruthenium and copper metals in the Cu-Ru/C catalystwas at least 50 times higher than the metal loadings in the Pd/Zn/C catalyst. This comparison suggests that our Pd/Zn/C cata-lyst, containing the least amount of palladium (0.013 mmol Pd)but giving comparable or higher DMF yields, is certainly supe-rior to the other reported systems.
Recyclability studies
As catalytic effectiveness of Zn2+-preloaded Pd/Zn/C material issimilar to that of combined ZnCl2-Pd/C catalyst, we used as-syn-thesized Pd/Zn/C catalyst for recyclability studies. In this experi-ment, 3.96 mmol (0.5 g) of HMF was reacted with 0.05 g Pd/Zn/C catalyst in 15 mL THF at 1508C and 8 bar H2 for 8 h. After the1st reaction cycle, the solid catalyst was recovered by simple fil-tration, washed with about 10–15 mL THF, dried in air, andreused for the next cycle. The next cycle was started by adding3.96 mmol of HMF and 15 mL THF. Some amount of fresh cata-lyst (10–15 wt% of total catalyst) was also added to replenishthe loss of mass in recycled catalyst during recovery. The aliquotfrom each cycle was analyzed by GC to quantify DMF yields.The results, as shown in Figure 5, reveal a deactivation of thecatalyst, in terms of DMF yield, after the 4th cycle.
To examine the catalyst deactivation, we analyzed the freshPd/Zn/C catalyst and the recovered catalyst after the 3rd and5th cycles by ICP–AES to measure palladium and zinc contents.There was a loss of zinc metal in the recovered catalyst ob-tained after the 3rd cycle. While this loss of zinc did not signifi-cantly influence activity in the 4th cycle, we found a noticeabledeactivation in catalyst activity in the 5th cycle. ICP–AES analy-sis of the recovered catalyst after the 5th cycle showed that thezinc content in the catalyst was almost the same as after the3rd cycle, but the amount of palladium was decreased byabout 0.54%.
While catalyst deactivation in the 5th cycle can be interpretedeither by (i) loss of palladium metal in the solution, or by (ii) di-lution of palladium concentration in the catalyst due to cokeformation on the catalyst surface, we performed a separate ex-periment to examine the possibility of palladium loss and in-volvement of homogeneous palladium metal in HMF HDO. Inthis experiment, 0.05 g Pd/Zn/C catalyst was heated in 15 mLTHF at 150 8C and 8 bar H2 for 22 h, which is almost the timeneeded for recycling the catalyst for 4 cycles. Upon cooling thereactor, the catalyst was separated by filtration and 3.96 mmolHMF was added into the filtrate, and the reaction was contin-ued at 150 8C and 8 bar H2 without any solid catalyst. A GCchromatogram of the aliquot showed no peaks for DMF orother hydrogenated intermediates. The peak area of HMF ac-counted for 95% of unconverted HMF. This result precludes
the hypothesis of palladiummetal leaching from the catalystduring 22 h heating with the sol-vent. Additionally, ICP analysis ofthe filtrate after five cyclesshowed that palladium loss inthe filtrate was very negligible(0.0065 mg or 0.34% based on1.9 mg palladium metal presentin the synthesized Pd/Zn/C cata-lyst). These results indirectly sup-port the hypothesis of coke for-mation on the catalyst surface ascause of deactivation and dilu-tion of the palladium concentra-
Table 1. Results of HMF HDO with different catalysts as a function of temperature and H2 pressure.
Entry Catalyst T[8C]
t [h] Solvent PH2
[bar]Conv.[%]
DMF yield[%]
Ref.
1 Pd/C/Zn 150 8 THF 8 >99 85 this work[a]
2 Rh/C 120 1 [EMIM]Cl 62 16 1 [14]3 Pd/C 120 1 [EMIM]Cl 62 19 2 [14]4 Pt/C 120 1 [EMIM]Cl 62 11 <1 [14]5 Ru/C 120 1 [EMIM]Cl 62 23 0 [14]6 Cu-Ru/C 220 10 2-butanol 6.8 – 71 [5][b]
7 Pd/C 150 15 THF generated in situ – 70 [9][c]
8 Ru/C 190 6 isopropanol generated in situ >99 81 [25][d]
[a] HMF=0.5 g, Pd/Zn/C=0.05 g (0.018 mmol Pd). [b] 30 wt% catalyst loading based on HMF (Cu=1.78 mmol;Ru=0.74 mmol). [c] 160 wt% catalyst loading based on HMF (Pd=0.188 mmol). [d] 35 wt% catalyst loadingbased on HMF (Ru=0.049 mmol).
Scheme 1. Proposed reaction pathway for HMF hydrodeoxygenation withthe ZnCl2-Pd/C catalyst.
� 2014 Wiley-VCH Verlag GmbH&Co. KGaA, Weinheim ChemSusChem
0000,
00, 1 – 8 &5&
CHEMSUSCHEMFULL PAPERS www.chemsuschem.org
80
tion in the recycled catalyst. Transmission electron microscopy(TEM) images of the as-synthesized and the recovered cata-lysts, showing their nanostructure and lattice fringes, areshown in Figure S7 (Supporting Information).
Substrate scope
The scope of the present investigation was further extendedto the direct conversion of fructose to DMF in one pot. Thetwo-step conversion of fructose to DMF involves dehydrationof fructose to HMF in the first step, followed by HDO of HMFto DMF in the 2nd step. As the present catalytic system containsboth a Lewis-acidic (ZnCl2) and hydrogenation component (Pd/C), we investigated the effectiveness of the ZnCl2-Pd/C catalystfor the conversion of fructose in THF at 150 8C under 8 bar H2.A reaction between 1.8 mmol (0.324 g) of fructose and ZnCl2(0.2 g)-Pd/C (0.05 g) for 8 h produced 22% DMF, 2% HMF, and3% 5-methylfurfural. Some unidentified GC peaks, accountingin total for half the peak area relative to that of DMF, were alsoobserved. Under comparable reaction conditions, a control ex-periment using only ZnCl2 as a dehydration catalyst producedonly 13 mol% HMF. This result suggests that fructose conver-sion to HMF and its subsequent HDO to DMF increased in thepresence of Pd/C. Assuming high pressure of H2 is a limitationfor HMF production in the beginning of experiment, we de-signed an experiment in which 1.8 mmol fructose was reactedwith ZnCl2 (0.2 g)-Pd/C (0.05 g) in 15 mL THF at 150 8C for 1.5 hand then the reactor was pressurized at 8 bar H2 to continuethe reaction for another 8 h. The latter reaction showed no im-provement in DMF yield. Considering the solvent to be a barrierfor effective dehydration of fructose to HMF, we conducteda similar experiment in a water/THF mixture (1 mL water+14 mL THF). However, total yields of DMF (19%) and HMF(6.5%) in the later experiment were similar.
Conclusions
2,5-Dimethylfuran (DMF) is a promising biofuel candidate, withfeatures that are more desirable when compared to ethanol. Al-though several research articles demonstrating different catalyt-ic technologies for DMF production have been published inrecent years, selective production of DMF in high yields remains
a challenge. We demonstrate a bimetallic palladium and zinccatalyst that achieves a DMF yield as high as 85% under mildreaction conditions (150 8C and 8 bar H2). Under comparable re-action conditions, Pd/C alone produces only 27% DMF. ZnCl2alone does not catalyze deoxygenation or hydrogenation of 5-hydroxymethylfurfural (HMF), suggesting a strong synergisticeffect between the two metals. Separate experiments using as-synthesized Pd/Zn/C catalyst give similar yields of DMF as theZnCl2-Pd/C combination. A reaction catalyzed by Ru/C produces2,5-bis(hydroxymethyl)furan (BHMF) as its major product. Theeffectiveness of the Ru/C catalyst improves in the presence ofZnCl2, but the overall DMF yield is much lower (41%) thanwhen using the Pd/Zn/C system. The less-effective hydrogena-tion catalyst Ni/C shows little improvement with ZnCl2, givingonly 7% DMF. The Pd/C-catalyzed reaction in the presence ofBrønsted-acidic Amberlyst 15 produces multiple products con-taining overhydrogenated tetrahydrofuran rings. The as-synthe-sized Pd/Zn/C catalyst shows excellent recyclability up to 4cycles, followed by some deactivation. Control experiments sug-gest that coke deposition on the catalyst surface is the reasonfor catalyst deactivation. The effectiveness of the Pd/Zn/C cata-lyst for the one-pot conversion of fructose to DMF is also inves-tigated. The zinc component facilitates hydrogenolysis of BHMFto DMF. Further experiments to understand the exact role ofthe zinc component in the bifunctional catalysts are underway.
Experimental Section
Materials
HMF, DMF, methanol, and fructose were purchased from Sigma–Al-drich and used as-received. Pd/C and Ru/C catalysts containing5 wt% of the respective metal loading, supported on activatedcarbon (AC) with a Brunauer–Emmett–Teller (BET) surface area of1200 m2g�1, were purchased from Strem Chemicals. However, in-ductively coupled plasma–atomic emission spectroscopic (ICP-AES)analysis by Galbraith Laboratory showed the palladium loading inthe purchased catalyst was 3.9%. THF containing butylated hydrox-yl toluene (BHT) as stabilizer was purchased from Fisher Chemicals.ZnCl2 was purchased from Acros Organics and used without furtherpurification. Ni/C catalyst was prepared by following a literaturemethod.[17] According to this method, nickel(II) nitrate hexahydrate(Ni(NO3)2·6H2O, 1.3 g) was dissolved in 6 mL water in a beaker.About 2 g AC support of 100 mesh particle size was added into theNiII solution and the mixture was stirred for 24 h. The sample wasdried overnight at 1208C. The dried sample was then reduced ina U-shaped quartz dried tube (Schwartz).[17] The resultant powderwas washed with water to remove any free nickel, filtered, anddried overnight at 1208C. 2,5-Bis(hydroxymethyl)furan (BHMF) wasprepared by NaBH4 reduction of HMF in ethanol. In this method,0.17 g NaBH4 was dissolved in 10 mL ethanol and the solution wascooled in an ice bath for 15 min. To this solution, 1 g HMF wasslowly added while the solution was continuously stirred. Upon stir-ring at 08C for 1 h, the solution was warmed to room temperatureand stirred overnight. The pH of the solution was adjusted to ap-proximately 7 by dropwise addition of dilute HCl and the solutionwas decanted to another round-bottom flask. BHMF was isolated(90% yield) after rotary evaporation of solvent and its purity wasdetermined by 1H NMR spectroscopy (Supporting Information, Fig-ure S1). As-synthesized Pd/Zn/C catalyst was prepared by incipient
Figure 5. Recyclability of the Pd/Zn/C catalyst for HMF hydrodeoxygenationto DMF.
� 2014 Wiley-VCH Verlag GmbH&Co. KGaA, Weinheim ChemSusChem
0000,
00, 1 – 8 &6&
CHEMSUSCHEMFULL PAPERS www.chemsuschem.org
81
wetness method. In this method, Pd/C and ZnCl2 in 1:3 ratio (w/w)were mixed in water and the mixture was stirred overnight at roomtemperature. The resultant solid was filtered and washed severaltimes with water to remove any free zinc salt and dried undervacuum. ICP–AES analysis of as-synthesized Pd/Zn/C catalystshowed the presence of palladium and zinc metals. Further analysisof this catalyst using SEM elemental mapping (NOVA nanoSEM FEI)confirmed that both metals are uniformly distributed throughoutthe sample (Supporting Information, Figure S2).
Catalysis
Catalytic HDO of HMF was carried out in a stainless steel Parr reac-tor equipped with a stirring impeller, gas line, and programmablecontrol device for setting reaction time and temperature. In a typi-cal experiment, the reactor was loaded with 0.5 g HMF(3.96 mmol), 0.05 g Pd/C, 0.05 g ZnCl2, and 15 mL solvent, andthen properly sealed with the reactor head. After purging the mix-ture with UHP-grade H2 for a couple of minutes with continuousstirring, the reactor was pressurized with H2 in the pressure rangeof 2–40 bar and heated at 150 8C for the desired time (1–20 h).Upon completion of reaction for the set time, the reactor wascooled to room temperature and the pressure released. The reac-tor vessel was removed and the solution was filtered. The collectedfiltrate was analyzed by GC and GC–MS.
Recyclability study
Catalyst reusability studies were performed by using as-synthesizedPd/Zn/C material by following the aforementioned procedure. Afterthe 1st cycle of reaction between 3.96 mmol of HMF and 0.05 g Pd/Zn/C catalyst at 1508C and 8 bar H2 for 8 h, the aliquot was filtered.The recovered catalyst was washed several times with THF, dried inair, and loaded in the reactor for the next cycle. The aliquot of eachcycle was analyzed by GC for quantification of products.
Determination of DMF yield
The yield of DMF and other products were measured by analyzingthe product solutions on a GC instrument (Agilent 6890N)equipped with a FID detector and DB-5 capillary column of dimen-sion 0.25 mm ID�0.25 mm�30 m. Essential parameters of GC anal-ysis are as follows: injection volume 1.0 mL, inlet temperature250 8C, detector temperature 250 8C, and a split ratio 1:5. Initialcolumn temperature was 50 8C (2 min) with a temperature rise of10 8Cmin�1 and final temperature was 300 8C. DMF was identifiedby its retention time in comparison with an authentic sample andby GC–MS analysis. 5-methyltetrahydrofurfural alcohol (MTHFA)and 2,5-hexane dione (HD) peaks were characterized by GC–MSanalysis. Each peak of the GC chromatogram was properly integrat-ed and the actual concentration of DMF was obtained from a pre-calibrated plot of peak area against concentrations. Unless other-wise mentioned, all yields are reported in mol%. GC–MS spectrom-etry analyses were carried out using an Agilent 5975C (AgilentLabs, Santa Clara, CA) mass spectrometer system. Typical electronenergy was 70 eV with the ion source temperature maintained at250 8C. The individual components were separated using a 30meter DB-5 capillary column (250 mm i.d. � 0.25 mm film thick-ness).The initial column temperature was set at 35 8C (for 3 min)and programmed to 280 8C at 10.0 8Cmin�1. The flow rate was typi-cally set at 1 mLmin�1. The injector temperature was set at 250 8C.
Acknowledgements
The authors acknowledge financial support from the Center fordirect Catalytic Conversion of Biomass to Biofuels (C3Bio), anEnergy Frontier Research Center funded by the US Department ofEnergy, Office of Science, and Office of Basic Energy Sciencesunder Award Number DE-SC0000997. We thank Dr. Trenton Par-sell (Purdue University) for helpful discussions and Ian Klein(Purdue University) for assisting ICP analysis.
Keywords: biomass conversion · heterogeneous catalysis ·hydrodeoxygenation · Lewis acids · ruthenium
[1] J. Goldemberg, Science 2007, 315, 808–810.[2] R. Sims, M. Taylor, J. Saddler, W. Mabee, From 1st to 2nd Generation Biofuel
Technologies, in An Overview of Current Industry and R&D Activities, Inter-national Energy Agency, November 2008. See http://www.iea.org/publi-cations/freepublications/publication/2nd_Biofuel_Gen.pdf (accessedAugust 2014).
[3] G. W. Huber, S. Iborra, A. Corma, Chem. Rev. 2006, 106, 4044–4098.[4] J. O. Metzger, Angew. Chem. Int. Ed. 2006, 45, 696–698; Angew. Chem.
2006, 118, 710–713.[5] Y. Rom�n-Leshkov, C. J. Barrett, Z. Y. Liu, J. A. Dumesic, Nature 2007, 447,
982–985.[6] J. B. Binder, R. T. Raines, J. Am. Chem. Soc. 2009, 131, 1979–1985.[7] S. Dutta, S. De, I. Alam, B. Saha, M. M. Abu-Omar, J. Catal. 2012, 288, 8–
15.[8] S. De, S. Dutta, B. Saha, ChemSusChem 2012, 5, 1826–1833.[9] T. Thananatthanachon, T. B. Rauchfuss, Angew. Chem. Int. Ed. 2010, 49,
6616–6618; Angew. Chem. 2010, 122, 6766–6768.[10] M. T. Barlow, D. J. Smith, D. G. Steward, Eur. Pat. EP0082689, 1983.[11] S. Song, R. Daniel, H. Xu, J. Zhang, D. Turner, M. L. Wyszynski, P. Ri-
chards, Energy Fuels 2010, 24, 2891–2899.[12] R. Rinaldi, F. Sch�th, ChemSusChem 2009, 2, 1096–1107.[13] W. Yang, A. Sen, ChemSusChem 2010, 3, 597–603.[14] M. Chidambaram, A. T. Bell, Green Chem. 2010, 12, 1253–1262.[15] Y. Zu, P. Yang, J. Wang, X. Liu, J. Ren, G. Lu, Y. Wang, Appl. Catal. B 2014,
146, 244–248.[16] T. H. Parsell, B. C. Owen, I. Klein, T. Jerrell, M. Marcum, L. J. Haupert, L. M.
Amundson, H. I. Kenttamaa, F. Ribeiro, J. T. Miller, M. M. Abu-Omar,Chem. Sci. 2013, 4, 806–813.
[17] Q. Song, F. Wang, J. Cai, Y. Wang, J. Zhang, W. Yua, J. Xu, Energy Environ.Sci. 2013, 6, 994–1007.
[18] Z. Hong, C. Quan, L. Chunhu, M. Xindong, Carbohydr. Res. 2011, 346,2016–2018.
[19] M. G. Grochowski, W. Yang, A. Sen, Chem. Eur. J. 2012, 18, 12363–12371.
[20] H. S. Thomas, B. Katalin, P. T. Anastas, P. C. Ford, A. Riisager, Green Chem.2012, 14, 2457–2461.
[21] a) A. Wang, R. Zhang, Acc. Chem. Res. 2013, 46, 1377–1386; b) Y. K.Kwon, E. D. Jong, S. Raoufmoghaddam, M. T. M. Koper, ChemSusChem2013, 6, 1659–1667.
[22] Y. Zhang, S. Wei, Chinese Patent, CN101423467, 2009.[23] T. Noguchi, K. Takayama, M. Nakano, Biochem. Biophys. Res. Commun.
1977, 78, 418–423.[24] G. Wang, Z. Guan, R. Tang, Y. He, Synth. Commun. 2010, 40, 370–377.[25] J. Jae, W. Zheng, R. F. D. G. Vlachos, ChemSusChem 2013, 6, 1158–1161.
Received: June 10, 2014Revised: July 24, 2014Published online on && &&, 0000
� 2014 Wiley-VCH Verlag GmbH&Co. KGaA, Weinheim ChemSusChem
0000,
00, 1 – 8 &7&
CHEMSUSCHEMFULL PAPERS www.chemsuschem.org
82
FULL PAPERS
B. Saha,* C. M. Bohn, M. M. Abu-Omar*
&& –&&
Zinc-Assisted Hydrodeoxygenation ofBiomass-Derived 5-Hydroxymethylfurfural to 2,5-Dimethylfuran
Are you thinking what I’m zincing?The addition of zinc salt enhances thecatalytic effectiveness of a palladium–carbon (Pd/C) catalyst in the hydrodeox-ygenation (HDO) of 5-hydroxymethylfur-fural (HMF) into 2,5-dimethylfuran(DMF) by a factor of more than 3. Thefinding allows to use catalysts with lesspalladium. The synergistic effect of thezinc salt with different hydrodeoxygena-tion catalysts is compared to elucidatethe role of the zinc component.
� 2014 Wiley-VCH Verlag GmbH&Co. KGaA, Weinheim ChemSusChem 0000, 00, 1 – 8 &8&
These are not the final page numbers! ��
83
Supporting Information� Copyright Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, 2014
Zinc-Assisted Hydrodeoxygenation of Biomass-Derived 5-Hydroxymethylfurfural to 2,5-DimethylfuranBasudeb Saha,*[a] Christine M. Bohn,[a] and Mahdi M. Abu-Omar*[a, b]
cssc_201402530_sm_miscellaneous_information.pdf
84
1
Table S1. Results of HMF HDO with different catalysts. a
Entry Catalyst T (ºC) H2 (bar)
Conv. (%) Products (%)
O
DMF
OOHHO
BHMF
OOH
MTHFA
O
OHD
O CHO
5-MF
1 Pd/C + ZnCl2 150 22 >99 85 - 2.6 1.6 - 2 Pd/C 150 22 40 26.5 - 4 1.5 -
Ru/C 150 8 60 3 45 2 - - Ni/C 150 8 10 2 2 - - -
3 ZnCl2 150 10 ~ 5 - 3 - - - 4 Ru/C + ZnCl2 150 15 99 41 52 1.5 - - 5 Ni/C + ZnCl2 150 8 26 7 4 - - 6.7% 6 Pd/C + ZnCl2 150 8 >99 84 - 3 2 - 7 Pd/C + WO3 150 19 95 66 14 - 2 <1 8b Pd/C + ZnCl2 150 10 88 82 - - - -
aHMF = 0.5 g, Pd/C = 0.05 g, ZnCl2 = 0.05 g, 15 mL THF, 8 hr; b BHMF = 0.2 g;
Table S2 ICP-AES analysis data of the as-synthesized and the recovered catalysts after the 3rd and 5th catalytic cycles.
Materials ICP-AES analysis
Zn (%) Pd (%)
As synthesized Pd/Zn/C 9.67 3.90
Recovered catalyst after 3rd cycle 0.40 3.84
Recovered catalyst after 5th cycle 0.47 3.30
85
2
Figure S1. 1H NMR spectrum of isolated BHMF product obtained from HMF reduction.
86
3
Figure S2. (a) SEM image (b) Zn mapping (C) Pd mapping and (d) elemental distribution of as-synthesized Pd/Zn/C
catalyst.
(b)
(c) (d)
(a)
100 nm
87
4
Figure S3. GC chromatogram of HMF HDO product.
Figure S4. GC chromatogram of reaction product obtained from HMF HDO with Pd/C catalyst.
O
OOH
O
OOH
O
OOHCOH
88
5
Figure S5. GC chromatogram of reaction product obtained from HMF HDO with Pd/C/ Amberlyst-15 catalyst.
Figure S6. GC chromatogram of reaction product obtained from BHMF HDO with the ZnCl2- Pd/C catalyst.
O
O
OO
OH
OOH
O CHOO
OHHO
89
6
Figure S7. TEM images of the as-synthesized Pd/Zn/C (a) and the recovered catalyst after the 3rd cycle (b) showing their nanostructure. Images (c) and (d) are the corresponding magnified view of this nanostructure showing the lattice fringes.
Cycle 2 Cycle 3
a b
c d10 nm 10 nm
1 nm 1 nm
90