Next-Generation Water-Soluble Homogeneous Catalysts for Conversion of Glycerol to Lactic Acid

Article abstract of DOI:10.1021/acs.organomet.8b00081

An attractive method for valorization of glycerol is the catalytic transformation to lactic acid. By overcoming the solubility challenge associated with known homogeneous catalysts for this reaction, we show that thermally robust Ir(I), Ir(III), and Ru(II

Full text of DOI:10.1021/acs.organomet.8b00081

This is an open access article published under an ACS AuthorChoice License, which permits
copying and redistribution of the article or any adaptations for non-commercial purposes.
Article
Cite This: Organometallics XXXX, XXX, XXX−XXX
Next-Generation Water-Soluble Homogeneous Catalysts for
Conversion of Glycerol to Lactic Acid
Matthew Finn, J. August Ridenour, Jacob Heltzel, Christopher Cahill,
and Adelina Voutchkova-Kostal*
Department of Chemistry, The George Washington University, 800 22nd Street NW, Washington, DC 20052, United States
S Supporting Information
*
ABSTRACT: An attractive method for valorization of
glycerol is the catalytic transformation to lactic acid. By
overcoming the solubility challenge associated with known
homogeneous catalysts for this reaction, we show that
thermally robust Ir(I), Ir(III), and Ru(II) N-heterocyclic
carbene (NHC) complexes with sulfonate-functionalized
wingtips are highly prolific for this process, requiring no
cosolvents other than aqueous base. The activity of the
catalysts is compared under both conventional heating and
microwave conditions. The most active catalyst reaches a TOF
of 45 592 h− (microwave) and 3477 h (conventional) with
equiv of KOH, and proceeds at a constant rate for at least 8
h. Although higher activity is observed with KOH, the catalysts
1
−1
1
−1
are also highly active with the weaker base, K CO (13 000 h and concurrent formation of formate). The protocol can be
2
3
modified to achieve quantitative conversion of glycerol in only 3 h. The high activity of these catalysts compared to
nonsulfonated analogs is attributed to the stabilization the lactate product in aqueous media. The most active catalyst retains
equal activity for crude glycerol. A mechanism is proposed for the most active catalyst precursor involving O−H oxidative
addition of glycerol.
INTRODUCTION
which then undergo dehydration and intramolecular Cannizaro
reaction to afford lactic acid. The most active catalyst identified
was an Ir(I) N-heterocyclic carbene (NHC) complex, which
■
The last few decades have seen a rapid rise in the production of
biodiesel from crops: In 2011, worldwide production exceeded
.5 million tons, and continues to grow. Glycerol, the major
13
1
afforded 337 turnovers per hour at 115 °C (see Scheme 1).
1
1
,2
Subsequently, Williams et al. reported that a highly robust
NHC-pyridine Ir(I) complex reaches 4.5 million turnovers in
byproduct, amounts for ∼10 wt % of biodiesel produced. The
abundance of glycerol produced and the lack of viable
valorization technologies has caused its market price to
−
1
3
2 days, with an average TOF of 6000 h at 145 °C under neat
14
3
conditions and base. A polymeric iridium catalyst reported by
Tu et al. allowed recyclability, with an average TOF of 3444
turnovers/h. In addition to electron-donating NHC ligands,
approach zero. This has fueled research into new valorization
4
5
processes, such as conversion to propane diols, glyceric acid,
15
16
6
7
cyclic acetals, and acroelin. One particularly attractive value-
added product of glycerol is lactic acid: a versatile platform
chemical with a rich portfolio of applications, including food,
phosphine ligands have also been found useful in designing
catalysts for this process: for example, Beller reported that a Ru
PNP pincer compound reaches 10 318 turnovers per hour with
67% selectivity for lactic acid at 140 °C in the presence of
8
cosmetics, pharmaceuticals, fine chemicals, and the synthesis
of polylactic acid (PLA). Use of PLA is growing rapidly due to
the shift toward biodegradable and renewable alternatives to
petrochemical-derived plastics. Furthermore, the traditional
9
17
diglyme as cosolvent. A similar PNP ligand was utilized by
Hazari et al. to develop an iron catalyst that averages 293
10
18
2
b,10,11
turnovers per hour using NMP as cosolvent.
fermentation process for producing lactic acid
is difficult
Although some of these catalysts are highly active, to develop
a homogeneous catalyst that is potentially viable for large-scale
conversion of a low-value chemical such as glycerol, exceptional
activity must be achieved. Improvements of catalytic activity
can be made by addressing the following: (i) the low solubility
of these homogeneous catalysts in polar protic media; (ii) poor
to scale, requires complicated purification and workup, and has
12
limited productivity. The production of lactic acid from
glycerol could offer scalable, atom-economical processes that
utilize a low-value feedstock.
A summary of the reported homogeneous catalysts for this
transformation is presented in Scheme 1. The seminal report of
the catalytic conversion of glycerol to lactic acid by Crabtree et
al. (Scheme 1) proposed that acceptorless dehydrogenation
of glycerol forms dihydroxyactone (DHA) or glyceraldehyde,
1
8
Received: February 8, 2018
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.organomet.8b00081
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 1. Catalysts Reported for the Synthesis of Lactic
Acid from Glycerol via Acceptorless Dehydrogenation
complexes have high aqueous solubility and are highly charged,
enabling them to take advantage of microwave heating effects.
All complexes share a motif of at least one NHC ligand
functionalized with sulfonates on the N-wingtip. We propose
that sulfonate groups are particularly well-suited to enhance
solubility in polar protic solvents, while having minimal
influence on the electron-donating ability of the NHC ligand
26
and metal coordination. This report builds on initial findings
by Peris et al. and by our group, which demonstrate the activity
of sulfonated NHC complexes for transfer hydrogenation of
22,27
22,27
27
ketones,
aldehydes,
and imines from glycerol, where
we observed formation of lactic acid.
RESULTS AND DISCUSSION
■
Catalyst Synthesis. Eleven iridium and ruthenium
complexes bearing alkyl- and arylsulfonate-functionalized
NHC ligands (Charts 1 and 2) were synthesized and
Chart 1. Catalysts for Acceptorless Dehydrogenation of
a
Glycerol
stabilization of lactate product in aprotic solvents; (iii) the
formation of deactivated polyhydride clusters, which have been
19
identified for Ir-NHC catalysts; and (iv) mass transfer
limitations due to high viscosity of medium. To address these
issues, we set out to design catalysts that are (i) soluble in polar
media, thus allowing reactions to be carried out under aqueous
conditions without added organic solvents, and (ii) active at
very low catalyst loadings (high dilution decreases concen-
tration-dependent cluster formation).
Microwave heating allows rapid, accurate, and uniform
20
heating to a set temperature, which can improve dispersion
21
of the base in aqueous glycerol. These advantages stem from
the fact that microwave heating uses an alternating electro-
magnetic field (AEMF) that interacts with polar molecules and
ions to cause oscillation and rotation, thus converting the
22
electric energy to kinetic energy, and subsequently to heat.
Each reaction component has unique dielectric properties (ε′
dielectric constant and ε″ dielectric loss), which determine its
ability to capture energy from the AEMF (ε′) and to convert
the energy into heat (ε″). The ratio of these two is an energy
dissipation factor (or loss tangent, tan δ), which can be used
compare the efficiency of conversion of microwave energy to
heat by polar molecules (although other factors, such as
relaxation times, should also be considered). Generally, the
reaction component with the most favorable dielectric
a
New compounds designated with asterisk.
characterized with the goal of assessing their activity for the
conversion of glycerol to lactic acid. The syntheses of
compounds 1 and 3−5 have been previously described,
while compounds 2 and 6−9 are first reported here.
28
23
properties can undergo selective heating. Microwave heating
is also particularly well-suited to reactions with polyols, such as
glycerol, which have high loss tangents (for glycerol the loss
24
Chart 2. Non-Sulfonated Ir-NHC Catalysts Utilized for
Acceptorless Dehydrogenation of Glycerol
tangent at 2.45 GHz is 0.651, compared to 0.123 for water)
a
even though the 2.45 GHz frequency of commercial micro-
waves is optimized based on the relaxation time of water.
Heating efficiency in reactions in glycerol can thus be further
enhanced by addition of water and ionic components, such as
base and catalyst. As highlighted by Kappe in his seminal study
on microwave effects, all of these factors contribute to bulk
thermal effects, rather than any “non-thermal microwave
2
5
effects”.
Herein we report the design, synthesis, characterization, and
application to the conversion of glycerol to lactic acid of a series
of iridium and ruthenium homogeneous catalysts. The
a
New compound designated with asterisk.
B
DOI: 10.1021/acs.organomet.8b00081
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 2. Synthesis of Compound 8 via Alkoxide Dimer route
Compounds 1−3 and 5 were prepared by silver transmetalation
from the corresponding imidazolium salts, while 4 was
synthesized by direct metalation using [Ir(cod)OMe]2.
2.048(5) Å, respectively, comparable to those reported for
31
analogous Ir-NHC compounds; these angles also compare
well to those in ([Ir(cod)(EMIm) ]NTf (2.067 and 2.059
2
7
2
2
30
Compounds 6−9 were prepared from the [Ir(cod)OEt]2
Å). The phenyl wingtips stagger in opposite directions with
torsion angles measuring 60.40 and 45.02°. The differences in
torsion angle between the two ligands can be attributed to the
coordinating sodium ion and closely associated water molecules
that act as a bridge between two molecular units (see Figures
S2 and S3 for more information).
dimer using a protocol analogous to that reported by Hermann
29
and Jimenez (Scheme 2). Addition of the imidazolium salt to
the dimer at room temperature resulted in clean formation of
the bis-NHC Ir(I) complex, whereas Ag transmetalation with
[
Ir(cod)Cl] afforded poor yields.
Compounds 1, 3, and 4 were included in light of their
2
To assess the effect of the sulfonate groups on catalytic
activity we also included two non-sulfonated analogs (Chart 2,
previously reported activity for transfer hydrogenation of
22,27
aldehydes, imines, and ketones from glycerol,
whereas 5
1
0 and 11). Compound 10, a nonsulfonated analogue of 5, was
was included in order to evaluate the effect of the
pentamethylcyclopentadinyl (Cp*) ligand relative to that of
the acetate and iodide ligands in 4. Iridium(I) compounds 6−9
are included in light of previously reported activity of Ir(I)
32
prepared using a procedure adapted from Heinkey et al.,
while 11 was prepared via transmetalation from the respective
13,33
Ag-NHC precursors.
Compound 11 was screened by
13
1
4,19
Crabtree et al. in the original report of this reaction and is a
close non-sulfonated analog of 6. Single crystal X-ray diffraction
of 10 shows Ir−C bond lengths of 2.039(5) and 2.026(5) Å for
the NHC ligands, Ir−Cl bond length of 2.4157(12) Å, and an
average Ir−C bond lengths for Cp* ligand of 2.220 Å (Figure
complexes.
Finally, compounds 7 and 9, which feature aryl-
sulfonate wingtips, allow us to compare the effect of the N-
sulfonate spacer on catalytic activity and stability, as well as the
effect of stronger π-acceptor CO ligands in 9 versus the weaker
and more labile cyclooctadiene (COD) ligand in 7.
The molecular structure of 7 was determined by single
crystal X-ray diffraction (Figure 1). The crystal structure of 7
2
). The chelating bond angle of 84.53(19)° between the
32
iridium and two NHC ligands fall into the expected ranges.
The carbonyl stretching frequencies (νCO) of the iridium
carbonyl compounds 8, 9, and 12 (where 12 is the carbonyl
analog of 11) were compared in order to assess the potential
Figure 1. ORTEP diagram of 7 with ellipsoids shown at 50%
probability level (noncoordinating cations, disordered lattice waters,
and hydrogen atoms omitted for clarity). Select bond lengths (Å) and
angles (deg): C(10)−Ir(1) 2.044(5), C(20)−Ir(1) 2.048(5), and
C(10)−Ir(1)−C(20) 87.33(18).
shows a disordered square planar coordination with two phenyl
wingtips in a staggered orientation with a closely coordinating
sodium ion and several cocrystallized water molecules. The
distortion angles between the COD-Ir plane and the carbene
carbons (C(20) and C(10) respectively) are 10.00(15) and
Figure 2. ORTEP diagram of 10 with ellipsoids shown at 50%
probability level (noncoordinating cations and hydrogen atoms
omitted for clarity). Select bond lengths (Å) and angles (deg):
Ir(1)-cent 1.8471(4), Ir(1)−C(11) 2.026(5), Ir(1)−C(7) 2.039(5),
Ir(1)−Cl(1) 2.4157(12), C(11)−Ir(1)−Cl(1) 91.19(13), C(7)−
Ir(1)−Cl(1) 88.62(13), C(7)−Ir(1)−C(11) 84.53(19).
1
2.78(15)°, respectively (Figure S1). These angles suggest
greater disorder than that observed in an analogous N-methyl,
N′-ethyl analog ([Ir(cod)(EMIm) ]NTf (where EMIm = 1-
2
2
ethyl-3-methylimidazole-2-ylidene) (distortion angles of 0.66°
3
0
and 0.53°). The Ir−C bond distances measure 2.044(5) and
C
DOI: 10.1021/acs.organomet.8b00081
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
a
Chart 3. FTIR νCO Stretches of Compounds 8, 9, and 12
a
18
Asterisks denote data for compound 12 reported by Crabtree et al.
effect of N-sulfonate spacers on the electron density of iridium.
The symmetric νCO of 8, 9, and 12 differ by 5 cm− in total
of the reaction vessels at such temperatures. The reaction
showed relatively little sensitivity to the ratio of glycerol to
water in the range of 2:1 to 1:2 glycerol/water (v/v) (entries 4,
6, and 7), where the differences observed are likely due to both
the changes in glycerol concentration and base equivalents (as a
result of keeping the total volume constant). Complete
elimination of water by using neat glycerol, however, decreased
yields, likely because water reduces the viscosity of the reaction
mixture and facilitates better mass transfer of base. A 1:1 ratio
of glycerol/water proved optimal, providing a balance between
glycerol concentration and viscosity of the reaction mixture.
The reaction conditions optimized for catalyst 1 were
subsequently extended to catalysts 2−11 under both conven-
tional and microwave heating conditions (Figure 3 and Tables
S4 and S5). The TOFs observed under microwave conditions
were significantly higher than those obtained by conventional
heating due to more efficient heating and improved mass
transfer, but the relative trends in activity among the catalysts
were almost identical. The three sulfonated ruthenium catalysts
1
(Chart 3), suggesting electronic differences are negligible.
Glycerol Dehydrogenation and Conversion to Lactic
Acid. To identify optimal reaction conditions, we examined the
activity of catalyst 1 using different temperatures, base
concentrations and solvent ratios (Table 1). For ease of
optimization these reactions were carried out using microwave
heating and fixed to a reaction time of 3 hours.
Table 1. Optimization of Reaction Conditions for
Conversion of Glycerol to Lactic Acid Using Catalyst 1
temp
(°C)
KOH
(mmol)
KOH
LA
a
entry
(equiv) Gly/H O (v/v) (mmol) TON
2
1
2
3
4
5
6
7
150
150
150
150
115
150
150
0
1.5
6
0
1:1
1:1
1:1
1:1
1:1
2:1
1:2
0
0
442
(
9
1−3) were less active than the sulfonated iridium catalysts (4−
0.04
0.14
1.02
1.02
0.77
1.53
1.5
2.6
3.0
0.17
2.1
2.4
). The Ru(bis-NHC) (1) and Ru(pyr-NHC) (2) complexes
1643
1983
342
afforded comparable TOFs, suggesting that the difference in
electron-donating ability of the pyridine versus NHC ligand, as
well as the hemilabile nature of the NHC-pyridine ligand in 2,
do not result in a significant effect on activity. Ru compound 3,
which only differs from 2 by a propylsulfonate imidazolium
moiety substituted on the pyridine, afforded 2−3 times higher
TOF compared to 1 and 2. The higher activity of 3 could be
the result of the in situ formation of a more catalytically active
42
42
42
42
1501
1708
a
Conditions: Catalyst 1 (12 ppm relative to glycerol), glycerol, and
water (6 mL total in ratio indicated), KOH (quantity indicated),
microwave heating in sealed 10 mL vial for 3 h at the indicated
temperature; yields determined by NMR using NaOAc as an internal
standard.
34
CNC Ru complex upon displacement of the p-cymene ligand.
The higher activity likely for a CNC Ru complex is supported
by the report by Beller et al. of an active Ru PNP pincer catalyst
The quantity of base (KOH) required is stoichiometric
relative to glycerol (i.e., 42 mmol), as lower quantities of base
and weaker bases (such as carbonates), result in lower yields of
lactic acid (LA) with this catalyst, (entries 1−4). As further
evidence, in entry 2 we observe the yield of lactic acid (LA) is
limited by the molar quantity of base. All reactions with
stoichiometric KOH afforded solely lactic acid, with no side
products of glycerol oxidation (Figure S4 shows NMR of crude
reaction). However, in reactions with substoichiometric KOH
traces of 1,2-propane diol (PDO) were observed. Control
reactions without base did not produce any lactic acid, and no
conversion of glycerol was observed (entry 1).
17
for this process. Although further optimization of the ligand
architecture around the ruthenium species is likely to result in
improvements of catalytic activity, in light of the significantly
higher activity observed with iridium catalysts, we did not
pursue further optimization of ruthenium complexes.
We next examine the activity of iridium(III) sulfonated
catalysts 4 and 5. The initial screen indicated that 4 is more
active than 5, with the difference being more drastic under
conventional relative to microwave conditions. To test whether
initial substitution of the chloride ligand is responsible for the
inferior activity of 5, we compared the activity with and without
an excess of NaOAc (100 equiv to catalyst), which is expected
to accelerate chloride substitution. The reaction with excess
NaOAc doubled the activity under microwave conditions (5743
vs 2850 turnovers/h). As a control, a reaction was also
performed with catalyst 4 and excess NaOAc, which resulted in
The activity observed with catalyst 1 was sensitive to reaction
temperature: decreasing temperature from 150 to 115 °C
resulted in a significant decrease in TON (1983 to 342, entries
4
and 5). Temperatures higher than 150 °C were not explored
because the high base concentration results in severe corrosion
D
DOI: 10.1021/acs.organomet.8b00081
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
This process would be favored under the highly basic and high-
temperature conditions employed and would make 6 less
35
thermally robust. We expect that more thermally robust
catalysts would have a significant advantage in catalyst lifetime
under microwave conditions, where superheating can cause
hotspots.
A prior report by Crabtree et al. indicated that substituting
the cod ligand of nonsulfonated Ir(I) complexes with carbonyls
13
affords an increase in glycerol dehydrogenation activity. We
thus prepared corresponding dicarbonyl complexes 8 and 9 by
passing CO gas through solutions of compounds 6 and 7.
While carbonyl derivatives 8 and 9 are significantly more active
than the cod precursors (6 and 7) (Figure 4), the relative
Figure 4. Bar graph comparing average turnover frequencies (TOF,
turnovers/h) obtained with catalysts 6−9 under microwave con-
ditions.
improvements were not equal. For the N-propylsulfonate
analogs (6 and 8) the improvement observed is significantly
smaller than that for the N-phenylsulfonate analogs (7 and 9);
furthermore, the activity of 9 relative to 7 increases ∼3-fold
(Figure 4). We attribute this to the more thermally robust
nature of the phenylsulfonate analogs.
Figure 3. Turnover numbers (labeled, upper horizontal axis) and
frequencies (lower horizontal axis) afforded by catalysts 1−11 for
conversion of glycerol to lactic acid (based on lactic acid yield) under
−1
The TOF afforded by 9 (42 592 h ) is approximately 4-fold
(
a) microwave conditions at 150 °C for 3 h and (b) conventional
higher than the best reported to-date for pure glycerol (11 055
heating to 150 °C for 24 h (see Tables S4 and S5 for details and
yields). Legend: green, nonsulfonated Ir catalysts; blue, sulfonated Ir
catalysts; red, sulfonated Ru catalysts.
−1
h
observed for glycerol/NMP by Beller et al. at 140 °C in the
36
first hour of reaction). To probe the effect of temperature on
reaction rate, we performed reactions at 115 °C using
microwave and conventional heating. Under these conditions
1
−1
a significant decrease in activity (553 vs 3122 turnovers/h), as
in this case free acetate slows down the Ir-OAc dissociation,
which is needed to generate the active catalyst.
catalyst 9 afforded 2241 h− and 416 h using microwave and
conventional heating, respectively.
To demonstrate the longevity of catalyst 9, the reaction was
extended to 8 h (Figure 5). Over that time, the reaction rate
remains constant, as evidenced by the linear increase in
concentration of product, reaching 293 249 turnovers. The
constant reaction rate is likely due to pseudoconstant
concentration of substrate and negligible catalyst deactivation
under these conditions. Given that very low catalyst loading is
employed, after 8 h the total conversion of glycerol was 35%.
In order to demonstrate that this catalyst can achieve full
conversion of glycerol in 1:1 glycerol/water we tested the
activity of 9 at an increased catalyst loading (18 ppm Ir,
0.00215 mol %). To avoid the pressure limitations observed in
the initial tests with high catalyst loading, the total volume of
The Ir(I) catalysts (6−9) were significantly more active than
the Ir(III) catalysts (4, 5). In order to avoid exceeding the
maximum pressure limit (250 psi) of the microwave vials, the
catalyst amount was reduced 10-fold (from 18 to 1.8 ppm Ir).
This decreased catalyst loading was sufficient to prevent the
pressure from reaching the maximum limit of the reactor. The
two catalysts with cyclooctadiene (cod) ligands (6 and 7)
achieve 7383 and 13 586 turnovers/h under microwave
conditions and 840 and 1635 turnovers/h respectively with
conventional heating. The significantly higher TOF and total
TONs of 7 relative to 6 is likely due to resistance of the N-
phenylsulfonate in 7 to degradation via Hoffman elimination.
E
DOI: 10.1021/acs.organomet.8b00081
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
analogs; the improvement is even more pronounced when
ligand degradation can be minimized.
Microwave versus Conventional Heating. In all cases
reported herein, the microwave protocol afforded significant
increase in TOF relative to conventional heating at the same set
temperature without sacrificing catalyst lifetime (reactions
performed for up to 8 h with catalyst 9 showed no decrease
in rate). The average magnitude of this increase was estimated
from a plot of average TOFs observed for conventional versus
microwave reactions at ∼11-fold. To demonstrate that this
increase is merely the result of thermal effects (improved
heating efficiency), we carried out the reaction at 150 °C in a
3
7
silicon carbide vial (designed by Kappe). These vials
efficiently convert microwave energy to heat but do not allow
microwaves to penetrate the vial. The turnover numbers
achieved under these reaction conditions were within
experimental error of those with the glass vial, confirming the
origin of the rate acceleration as thermal efficiency.
Figure 5. Extended reaction for the dehydrogenation of glycerol to
lactic acid using catalyst 9. Reaction conditions: 2 mL of glycerol (27.4
mmol), 2 mL of aqueous KOH (14 M,), TSP (0.11 mmol, internal
std)catalyst 9 (29.5 nmol), microwave heating to 150 °C in a sealed
microwave reaction tube. TONs determined by NMR using TSP as an
internal standard. TSP: 3-(trimethylsilyl)propionic-2,2,3,3-d4 acid
sodium salt.
Toward Milder Base Conditions for Lactic Acid
Formation. In an effort to explore the activity of the catalyst
under milder base conditions, the reaction with catalyst (9) was
attempted with lower base concentration and weaker bases.
First, the activity of 9 was tested with potassium carbonate
instead of KOH. Due to the lower solubility of K CO in 1:1
reaction was reduced, allowing for increased headspace.
Catalyst 9 afforded near-quantitative conversion of glycerol
and 91% yield of LA in 3 h (46 400 turnovers). The remaining
2
3
glycerol/water, we were only able to achieve a maximum
concentration of 2.7 M (0.4 equiv). Interestingly, we observed
the production of not only lactic acid, but also 1,2-propanediol,
as well as formic acid (Figure S5). Although the turnover
numbers were lower than for the reaction with 1 equiv of KOH
9
% of glycerol was converted to 1,2-propane diol (PDO),
which is formed by hydrogenation of the pyruvaldehyde. The
pyruvaldehyde in turn is formed via dehydration of
glyceraldehyde (Scheme 3). These results demonstrate the
exceptional activity and selectivity for the reduction of glycerol
to lactic acid in a short reaction times.
(
TONs: 39 000 LA, 8 300 PDO, 630 FA), this activity is highly
promising and offers possibility of generating other products
from glycerol. We also explored the effect of lower KOH
concentration: Upon lowering the KOH loading to 0.22 equiv
Scheme 3. Quantitative Conversion of Glycerol Using 18
ppm (2.15 × 10− mol %) Catalyst 9
3
(
1.7 M), the turnovers and glycerol conversion decreased
significantly (27 801 vs 127 777 turnovers; 3 % vs 13% glycerol
conversion). This is consistent with prior reports that show the
rate dependence on concentration of strong base.
Crude Glycerol as Substrate. Given that purification of
crude glycerol from biodiesel synthesis is the ultimate
application of the catalysts in question, we tested the activity
of catalyst 9 for dehydrogenation of crude glycerol (80−85%
glycerin purchased from Inkemia). Catalyst 9 affords almost
identical activity and selectivity for lactic acid with crude and
pure glycerol, suggesting that the impurities do not deactivate
the catalyst (Table 2). The slight decrease in TON can be
attributed to slightly higher dilution in crude glycerol, as one of
the major impurities is water.
Sulfonate Effect. In order to probe the effect of the
sulfonate groups on catalytic activity we tested two non-
sulfonated analogs in Chart 2 under identical conditions.
Compound 10, analog of 5, is approximately 50% less active
than 5 (Figure 4 and Table S4 and S5), likely due to its lower
solubility. Furthermore, unlike the acceleration observed for 5
upon addition of NaOAc, in the case of 10, acetate addition had
no favorable effect (Table S4 entries 13−14), likely due to its
the low solubility in the aqueous reaction medium. Similarly,
compound 11, a nonsulfonated (methyl) analog of 6 and 7, was
significantly less active than 6 and 7 under microwave
conditions (6554, 7383, and 13 586 turnovers/h for 11, 6,
and 7 respectively), and a similar trend was observed using
conventional heating (390, 840, and 1635 turnovers/h for 11,
Effect of Stir Rate for Microwave Reactions. In
optimizing the microwave protocol, we found that the catalytic
Table 2. Comparison of Catalyst 9 with Pure and Crude
a
Glycerol
entry
reaction
TON
avg. TOF
1
2
pure glycerol (99.5%)
127777
125679
42592
41893
crude glycerol (Inkemia, 80%)
a
6
, and 7 respectively). Thus, we observe multiple examples
where sulfonate functionalization of the NHC wingtips
Conditions: 3 mL of glycerol, 3 mL of H O, 1 equiv of KOH, 18 ppm
2
catalyst 9, microwave heating in closed 10 mL vial to 150 °C for 3 h;
significantly increases catalytic activity relative to nonsulfonated
yields determined by NMR using NaOAc as an internal standard.
F
DOI: 10.1021/acs.organomet.8b00081
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
activity is affected by stir rate, which is often omitted from
reported microwave protocols. We hypothesized that changing
the stir rate resulted in fluctuations of the internal reaction
temperature, which were not captured by the external IR
thermometer in the microwave cavity. Thus, we monitored the
internal reaction temperature at different stir rates using an in
situ fiber optic thermometer (Figure 6).
alkoxide, β-hydride elimination, and finally hydride abstraction
+
38
using H to release H . Given that CO substitution in 9 is not
2
39
facile, we propose the catalytic cycle in Figure 6. First, the 16-
electron Ir(I) species undergoes O−H oxidative addition with
glycerol to form an alkoxy hydride complex; this step has
40
precedents for Ir(I) complexes. The resulting Ir(III) alkoxy
hydride complex could then eliminate H via σ-bond metathesis
2
with an adjacent hydroxyl of glycerol, a reaction with theoretical
41
precedent for Pt-group metals. The chelating bis-alkoxy
complex can then undergo β-hydride elimination at the primary
or secondary position by labializing one of the alkoxy groups,
and release the product (glyceraldehyde or dihydroxyacetone)
via O−H reductive elimination. Neither of these intermediates
has been observed under basic conditions, as both are expected
to undergo dehydration and intramolecular Cannizzaro
reaction.
−
The difference in activity observed for [(NHC-ph-SO ) Ir-
3
2
−
−
−
(
cod)] (7) versus [(NHC-ph-SO ) Ir(CO) ] (9) could be
3 2 2
due to (i) more favorable H elimination or O−H reductive
2
elimination from Ir center in 9 with stronger π-acceptor (CO)
ligands or (ii) the fact that 7 undergoes an entirely different
mechanism than 9 (cycooctadiene ligand in 7 could easily be
displaced, opening coordination sites, which is not the case for
Figure 6. Effect of stir rate on internal temperature of microwave
reactions for glycerol conversion to lactic acid (internal temperature
measured using fiber optic thermometer).
39
9
). Although the presence of carbonyl ligands also opens the
35a,39
possibility for cluster formation,
in the case of 9 cluster
formation is not likely at the low catalyst loadings used,
compared to previous reports.
Figure 6 shows the relationship between stir rate, TON, and
internal reaction temperature. When the stir rate is set to the
maximum (1000 rpm), significant splashing can be seen with
the internal camera, potentially resulting in inefficient heating
and lower TONs. At 1000 rpm stir rate the internal probe
measured 143 °C, compared to 150 °C on the wall of the vial.
As stated previously, we compared the catalytic activity at 150
13,19
Although a detailed mechanistic investigation will be the
subject of a future report, we performed an NMR experiment to
identify the Ir−hydride species formed. Upon addition of
glycerol and KOH(aq) to 9 at room temperature no change is
observed, but heating to 50 °C for 30 min results in a color
change from yellow to green, and appearance of two hydride
peaks at −10.8 and −11.1 ppm respectively (relative
integrations 1:1.3, Figure S6). For comparison, we have
previously reported that catalyst 4 (Ir(III) complex) under
the same conditions affords a single hydride signal at −12.8
°
C and 1000 rpm stir rate in a silicon carbide vial. These vials
efficiently convert microwave energy to heat, but do not allow
microwaves to penetrate the vial. The turnover numbers
achieved under these reaction conditions were 49 645, within
experimental error of those with the glass vial.
27
ppm at room temperature. The observation of two hydride
resonances for the reaction with 9 is consistent with the
catalytic cycle proposed (Figure 7).
When the stir rate is decreased to 300−700 rpm, heating
becomes more uniform, and the internal temperature was
steady at 147 °C. As expected, no difference in TON was
achieved at these two stir rates (700 and 300 rpm). However,
when stirring is decreased to 100 rpm the internal temperatures
reaches 180 °C, which results in a vast increase in catalytic
activity and rapid hydrogen generation. The maximum pressure
CONCLUSIONS
■
Here we report the design, synthesis, and application of highly
active catalysts for the conversion of glycerol to lactic acid via
acceptorless dehydrogenation of glycerol in aqueous base under
microwave and conventional heating conditions. The catalysts
consist of Ir(I), Ir(III) and Ru(II) complexes bearing sulfonate-
functionalized N-heterocyclic carbene ligands (NHCs). The
sulfonate functionalization renders the catalyst very soluble in
the aqueous reaction medium, which results in significant
improvement in activity compared to nonsulfonated analogs.
Further improvements result from the use of N-substituents on
the NHC that disfavor Hoffman elimination, thus enhancing
thermal stability. The significant improvement in apparent
activity of the catalysts under microwave conditions (11-fold on
average) are likely due to more efficient heating and improved
dispersion of base in viscous glycerol. The most active catalyst
(30 bar) of the reactor is quickly reached under these
conditions, and significant etching of the glass vial is observed.
While the use of this microwave vial does not allow us to carry
out these reactions at temperatures above 150 °C (due to high
base concentration required), this result signals that catalytic
activity can further be increased by using a reactor suitable for
high temperatures, pressures, and pH. For the existing
apparatus, we identified an optimal stir rate of ∼600 rpm.
The reaction was repeated in a 30 mL vial; under these
conditions, catalyst 9 achieved 124 438 turnovers (vs 127 777
in the 10 mL vial), suggesting that pressure build-up in the 10
mL vials does not affect the productivity of the catalysts under
these conditions.
Mechanism. The identification of Ir(I) compound 9 as the
most active catalyst poses the question of whether its
mechanism for acceptorless dehydrogenation significantly
differs from that of more extensively studied Ir(III) catalysts.
As proposed by Xiao et al, Ir(III) compounds likely undergo
ligand substitution by an alcohol, deprotonation to form Ir-
−
−
is [(NHC-ph-SO ) Ir(CO) ] (9), which affords 45 592
3 2 2
turnovers/h under microwave conditions and was observed to
afford up to 293 249 turnovers in 8 hours. Under these
conditions, 9 achieves average turnover frequencies approx-
imately 4-fold higher than the best reported catalyst to-date for
pure glycerol. Although higher activity is observed with strong
G
DOI: 10.1021/acs.organomet.8b00081
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
modeled. In complex 7, PART, SIMU, and RIGU commands were
used to model two-part disorder in both of the sulfonate groups and
two of the five water molecules (OW1A/OW1B, OW3A/OW3B), and
PART and ISOR commands were used to model disorder in a third
water molecule (OW5A/OW5B). Additionally, the OMIT 1 0 0
command was used in complex 10 to remove a reflection affected by
the beamstop. Structures were checked for additional symmetry using
4
9
PLATON. Thermal ellipsoid figures were prepared with ORTEP
5
0
III. Data collection and refinement details for complexes 7 and 10
are included in Table S1, and select bond lengths can be found in
tables S2 and S3.
Pyr-NHC Ligand (2a). 2-(Imidazol-1-yl)pyridine (0.4734 g, 3.26
mmol) and 1,2-propane sultone (0.5924 g, 0.485 mmol) were
measured into a pyrex pressure tube with MeCN (7 mL). The tube
was sealed and heated at 100 °C for 16 h at which point a white
precipitate is observed. After cooling to room temperature, the solid
was collected by filtration and washed with 3 × 10 mL of CH Cl and
2
2
dried in vacuo at room temperature. Ligand 2a was recovered as a
1
white solid (828 mg, 3.1 mmol, 94% yield). H NMR (400 MHz,
D O) δ 9.65 (t, J = 1.7 Hz, 1H), 8.61 (ddd, J = 4.9, 1.8, 0.8 Hz, 1H),
2
8
7
.20 (dd, J = 2.2, 1.7 Hz, 1H), 8.17 (ddd, J = 8.2, 7.6, 1.8 Hz, 1H),
.85−7.79 (m, 2H), 7.65 (ddd, J = 7.6, 5.0, 0.9 Hz, 1H), 4.55 (t, J =
13
Figure 7. Proposed catalytic cycle for glycerol dehydrogenation by
catalyst 9.
7.2 Hz, 2H), 3.07−2.99 (m, 2H), 2.50−2.39 (m, 2H). C NMR (101
MHz, D O) δ 194.92, 149.18, 146.20, 140.89, 134.48, 125.48, 123.23,
2
1
20.00, 114.94, 48.47, 47.15, 24.92. HRMS(ESI/Q-TOF) m/z: [M +
+
H] Calcd for C H N O S 268.0756. Found 268.0759.
11
14
3
3
bases, 9 is also highly active with the weaker base, K CO
2
3
Ru(Pyr-NHC-iPrSO )(p-cymene)Cl (2). Ligand 2a (75 mg, 0.28
13 000 h−1 and concurrent formation of formate). Activity is
3
(
mmol) was dissolved in a degassed solution of 16 mL of MeOH and 3
also retained for crude glycerol. The low catalyst loadings used
disfavor previously reported catalyst deactivation via cluster
formation. We propose a mechanism for 9 that involves O−H
oxidative addition of glycerol, σ-bond metathesis to eliminate
hydrogen, β-hydride elimination and O−H reductive elimi-
nation to expel the dehydrogenation products. The results
presented here highlight the importance of enhanced catalyst
solubility for reactions in highly polar media, and the substantial
enhancement that can be afforded for by microwave heating for
viscous and highly base-concentrated reactions.
mL of H O. The flask was wrapped in foil, and Ag O (65 mg, 0.28
2
2
mmol) was added in darkness and the solution stirred at 50 °C for 90
min. A solution of NaCl (35 mg, 0.60 mmol) in 1 mL of H O was
2
added and the solution stirred for another 15 min. The silver solution
was then filtered and transferred to a solution of [Ru(p-cymene)Cl2]
2
(
100 mg, 0.16 mmol) in 15 mL of H O and stirred at room
2
temperature overnight. The solution was filtered and the filtrate dried
in vacuo at 50 °C. Then, 50 mL of MeCN was added to the crude
product, stirred for 1h at room temperature and filtered over Celite.
The clear yellow filtrate was dried, dissolved in a minimum amount of
MeOH, and transferred to a silica column. Elution with 800 mL of
acetone/MeOH (80/20) and 400 mL of acetone/MeOH (50/50)
afforded a bright orange band. Removal of the solvent in vacuo
afforded 2 as a highly hygroscopic orange solid (45 mg, 0.0837 mmol,
30% yield).
EXPERIMENTAL SECTION
■
General Considerations. The syntheses of the catalysts were
carried out under nitrogen using standard Schlenk technique, unless
otherwise stated. Commercial chemicals were used without further
1H NMR (400 MHz, CD
OD) δ 9.32 (ddd, J = 5.8, 1.5, 0.7 Hz,
3
purification. [RuCl (p-cymene)]2 and IrCl3 were purchased from
1H), 8.24−8.17 (m, 2H), 8.02 (ddd, J = 8.3, 1.2, 0.7 Hz, 1H), 7.78 (d,
J = 2.3 Hz, 1H), 7.52 (ddd, J = 7.5, 5.8, 1.3 Hz, 1H), 6.47 (dd, J = 6.1,
1.3 Hz, 1H), 6.31 (dd, J = 6.3, 1.3 Hz, 1H), 6.13 (dd, J = 6.4, 1.3 Hz,
1H), 5.88 (dd, J = 6.1, 1.4 Hz, 1H), 4.85−4.78 (m, 1H), 4.72 (ddd, J =
13.8, 10.6, 5.8 Hz, 1H), 3.06 (dt, J = 7.5, 5.1 Hz, 2H), 2.65−2.54 (m,
1H), 2.43 (p, J = 7.0 Hz, 2H), 2.27 (s, 3H), 0.98 (dd, J = 6.9, 3.8 Hz,
2
Acros-Organics. Solvents were dried using a solvent purification
system (SPS MBraun) or 4 Å molecular sieves. Glycerol (>99%, Alfa
Aesar) was dried over activated 4 Å molecular sieves. Noncommercial
reagents: ligands, catalyst precursors including, [Ir(cod)Cl]2 and
[
IrCp*Cl ] , and catalysts 1, 3−5, and 11 were synthesized via
2
2
6H). 13C NMR (101 MHz, CD
OD) δ 185.30, 156.75, 153.38, 142.66,
previously reported procedures; for synthetic details, please refer to
the Supporting Information. NMR spectra were recorded on an
Agilent NMR spectrometer operating at 400 MHz.
3
126.19, 124.25, 118.06, 113.51, 92.80, 92.40, 88.33, 82.60, 51.35,
49.29, 32.48, 27.77, 22.90, 22.52, 19.32. HRMS(ESI/Q-TOF) m/z:
+
Single Crystal X-ray Diffraction. Single crystals from the bulk
samples of compounds 7 and 10 were isolated and mounted on
MiTeGen micromounts. Reflection data were collected at 100(2)K
with 0.5° ω scans on a Bruker SMART diffractometer equipped with
an APEX II CCD detector using Mo Kα (λ = 0.71073 Å) radiation.
[M + H] Calcd for C21
H
27
N
3
O
3
SRuCl 538.0505. Found 538.0504
General Procedure for Synthesis of Ir(I) Catalysts. Under a
nitrogen atmosphere, [Ir(cod)Cl] (66 mg, 0.098 mmol) was dissolved
2
in 8 mL of degassed EtOH (200 proof). A solution of NaH (25 mg, 60
wt % in mineral oil) in 2 mL of EtOH was added dropwise; the
solution quickly from orange to yellow and was stirred for 30 min at
25 °C. A suspension of the imidazolium ligand (0.37 mmol) in 3 mL
42
The data were integrated using the SAINT program within the
43
APEX II software suite, and absorption corrections were applied
44
using SADABS. Complex 7 was solved via direct methods using
of EtOH and 0.5 mL of H O was slowly added to the iridium solution
2
45
Superflip, and complex 10 was solved via direct methods using SIR
via syringe and stirred at 25 °C for 72 h. The solvent was then
removed under reduced pressure and the crude product redissolved in
4
6
47
9
2. Both complexes 7 and 10 were refined using SHELXL-2014 in
48
the WinGX software suite. In each structure, all non-hydrogen atoms
were located in difference Fourier maps and were refined anisotropi-
cally. Aromatic hydrogen atoms were placed in idealized positions by
utilizing the HFIX43 command and allowed to ride on the coordinates
a minimum amount of MeOH (3 mL). Addition of Et O afforded a
2
precipitate, which was collected, washed 3 times more with Et O and
2
dried under vacuum.
Na[Ir(NHC-nPrSO ) (cod)] (6). Following the general procedure for
3
2
of the parent atom with isotropic thermal parameters (U ) fixed at
Ir(I) catalysts, complex 6 was recovered as a hygroscopic red solid (78
iso
1
1
.2U . The hydrogen atoms on all the water molecules in complex 7
mg, 0.106 mmol, 57% yield). H NMR (400 MHz, CD OD) δ 7.26
eq
3
could not be located in the difference Fourier map and were thus not
(dd, J = 14.0, 2.0 Hz, 2H), 7.17 (dd, J = 10.3, 2.0 Hz, 2H), 4.57−4.45
H
DOI: 10.1021/acs.organomet.8b00081
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(
m, 2H), 4.24−4.12 (m, 2H), 4.07 (m, 1H), 4.05 (s, 6H), 3.89 (m,
33.25, 30.93, 20.22, 13.80, 9.40. HRMS(ESI/Q-TOF) m/z: [M −
+
1
2
2
1
2
H), 3.73−3.65 (m, 2H), 2.99−2.89 (m, 4H), 2.50−2.41 (m, 2H),
PF ] Calcd for C H ClN Ir 623.2492. Found 623.2483.
6
25 39
4
.35−2.24 (m, 2H), 2.24−2.14 (m, 4H), 2.05−1.95 (p, J = 7.5 Hz,
General Procedure for Glycerol Dehydrogenation. A 10 mL
microwave tube was charged with glycerol (3 mL, 41 mmol), aqueous
KOH (3 mL, 14 M, 42 mmol), catalyst (0.5−0.7 μmol), and a
magnetic stir bar. The reaction mixture was thoroughly stirred to
ensure adequate mixing before being placed in the cavity of a CEM
Discover or Anton Parr Monowave microwave reactor. Heating with
150 W power and a maximum pressure of 250 PSI allowed the desired
temperature (150 °C) to be reached in 1 min on average. After cooling
to room temperature, the reaction was opened, and aqueous solution
1
3
H), 1.91−1.71 (m, 2H). C NMR (101 MHz, CD OD) δ 176.60,
3
61.59, 123.12, 120.66, 109.99, 77.36, 74.54, 49.10, 37.40, 32.01,
9.46, 26.15. HRMS(ESI/Q-TOF) m/z: [M + H]+ Calcd for
C H N S O IrNa 731.1530. Found 731.1535
22
35
4
2
6
Na[Ir(NHC-PhSO ) (cod)] (7). Following the general procedure for
3
2
Ir(I) catalysts, complex 7 was recovered as a bright red solid (69 mg,
0
diffusion of acetone into a concentrated solution of 7 in water. H
NMR (400 MHz, CD OD) δ 8.02−7.97 (m, 4H), 7.37−7.32 (m, 4H),
7
3
=
2
1
Q-TOF) m/z: [M − Na] Calcd for C H N S O Ir 775.1236.
Found 775.1227.
Ir(NHC-nPrSO )(CO) (8). Complex 6 (54 mg, 0.073 mmol) was
dissolved in 10 mL of degassed MeOH, and the system was flushed
with N for 10 min. CO(g) was bubbled through the solution at room
.086 mmol, 53% yield). X-ray quality crystals were grown via vapor
1
of standard was added (NaOAc or TSP, approximately 0.1 M solution
3
1
in D O). Lactic acid formation was determined via H NMR
.24 (d, J = 2.0 Hz, 2H), 7.14 (d, J = 2.0 Hz, 2H), 4.78−4.70 (m, 2H),
.61 (q, J = 7.7 Hz, 2H), 3.22 (s, 6H), 2.51−2.38 (m, 2H), 2.28 (dd, J
2
spectroscopy using D O as solvent and NaOAc or TSP as an internal
2
standard. For reactions performed using conventional heating, the
same procedure was followed, but using 20 mL reaction tubes that
were placed in a Heidolph 12 Carousel and heated to 150 °C for 24 h.
15.2, 7.8 Hz, 2H), 2.17−2.05 (m, 2H), 1.69 (td, J = 14.4, 7.3 Hz,
H). 13C NMR (101 MHz, CD OD) δ 177.58, 147.47, 142.61, 128.28,
3
27.14, 124.30, 124.25, 80.73, 73.91, 38.00, 35.94, 28.26. HRMS(ESI/
−
28
30
4
2
6
ASSOCIATED CONTENT
■
3
2
*
S
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.organo-
met.8b00081.
2
temperature for 90 min, and the solution changed from orange to
yellow in color. The solvent was reduced to about 2 mL in vacuo at
room temperature, at which point addition of Et O afforded the
formation of a yellow precipitate. The solid was washed 3 × 5 mL
2
Synthetic details for previously reported ligands and
catalysts, NMR mechanistic studies, crystallographic
parameters, NMR spectra for new catalysts (PDF)
further with Et O and dried in vacuo at room temperature to yield
2
1
complex 8 as a yellow solid (35 mg, 0.051 mmol, 69% yield). H NMR
(
400 MHz, CD OD) δ 7.45 (d, J = 2.0 Hz, 2H), 7.41 (d, J = 2.0 Hz,
3
Accession Codes
2
2
1
H), 4.06 (t, J = 7.6 Hz, 4H), 3.98 (s, 6H), 2.78 (t, J = 7.2 Hz, 4H),
.13 (m, 4H). 13C NMR (101 MHz, CD OD) δ 179.88, 166.83,
CCDC 1586716−1586717 contain the supplementary crystal-
lographic data for this paper. These data can be obtained free of
charge via www.ccdc.cam.ac.uk/data_request/cif, or by email-
ing data_request@ccdc.cam.ac.uk, or by contacting The
Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
3
24.26, 122.46, 49.33, 47.71, 37.74, 26.28. HRMS(ESI/Q-TOF) m/z:
+
[
M − Na] Calcd for C H N S O Ir 655.0508. Found 655.0505
16
22
4
2
8
Ir(NHC-PhSO ) (CO) (9). Complex 7 (69 mg, 0.086 mmol) was
3
2
2
dissolved in 10 mL of degassed MeOH, and the system was flushed
with N for 10 min. CO(g) was bubbled through the solution at room
temperature for 90 min, and a solution changed from red/orange to an
2
orange/yellow color. The solvent was reduced to about 2 mL in vacuo,
AUTHOR INFORMATION
■
after which Et O was added until a yellow precipitate was formed. The
2
Corresponding Author
*E-mail: avoutchkova@gwu.edu.
ORCID
Matthew Finn: 0000-0001-9474-7022
Christopher Cahill: 0000-0002-2015-3595
Adelina Voutchkova-Kostal: 0000-0002-7016-5244
solid was washed 3 times with Et O and dried in vacuo to yield
2
1
complex 9 as a yellow solid (36 mg, 0.047 mmol, 54% yield). H NMR
(
=
400 MHz, D O) δ 7.90−7.85 (m, 4H), 7.50−7.46 (m, 4H), 7.24 (d, J
2
13
2.0 Hz, 2H), 7.00 (d, J = 2.0 Hz, 2H), 3.44 (s, 6H). C NMR (101
MHz, D O) δ 179.49, 168.73, 163.16, 142.84, 140.92, 126.64, 125.27,
2
+
1
23.95, 122.39, 38.31. HRMS(ESI/Q-TOF) m/z: [M + H] Calcd for
C H N S O IrNa 747.0177. Found 747.0155.
22
19
4
2
8
[
Ir(bis-NHC)Cp*Cl]PF (10). [IrCp*Cl ] (30.6 mg, 0.038 mmol)
6 2 2
Notes
and 1,1′-methylenebis(3-butyl-1H-imidazol-3-ium) dichloride (31.8
mg, 0.081 mmol) were added to a Schlenk flask under a nitrogen
The authors declare no competing financial interest.
atmosphere. Et N (0.1 mL, 0.717 mmol) was added via syringe,
3
followed by 30 mL of dry and degassed MeCN. The mixture was
refluxed under a nitrogen atmosphere overnight. After cooling to room
temperature, the solvent was removed under vacuum. The yellow
residue was washed with 3 × 5 mL of acetone and filtered over Celite.
A small amount of silica was added to the filtrate, and solvent was
removed under vacuum. The yellow silica cake was loaded onto a silica
column conditioned with acetone, elution of a yellow band was
achieved with acetone/KPF6. The yellow band was collected and
solvent was removed under vacuum, leaving a yellow residue with
excess KPF . Washing with CH Cl and filtration over Celite leaves 10
ACKNOWLEDGMENTS
■
We thank the NSF for funding through NSF CAREER award
no. 1554963).
(
REFERENCES
■
(1) Quispe, C. A. G.; Coronado, C. J. R.; Carvalho, J. A., Jr. Renewable
Sustainable Energy Rev. 2013, 27, 475−493.
(2) (a) Tan, H. W.; Abdul Aziz, A. R.; Aroua, M. K. Renewable
Sustainable Energy Rev. 2013, 27, 118−127. (b) Corma, A.; Iborra, S.;
Velty, A. Chem. Rev. 2007, 107 (6), 2411−2502.
(3) Gholami, Z.; Abdullah, A. Z.; Lee, K.-T. Renewable Sustainable
Energy Rev. 2014, 39, 327−341.
6
2
2
in solution, which can be dried under vacuum to a yellow powder (54
mg, 0.07 mmol, 92%). Single crystals suitable for X-ray analysis were
grown via vapor diffusion of Et O into a concentrated solution of 10 in
2
1
CH Cl at room temperature. H NMR (400 MHz, CDCl ) δ 7.45
(4) ten Dam, J.; Hanefeld, U. ChemSusChem 2011, 4 (8), 1017−
1034.
2
2
3
(
1
dd, J = 2.1, 0.8 Hz, 2H), 7.09−7.07 (m, 2H), 6.24 (d, J = 13.6 Hz,
H), 5.58 (d, J = 13.6 Hz, 1H), 4.26 (td, J = 11.9, 5.6 Hz, 2H), 3.83
ddd, J = 12.5, 10.9, 5.7 Hz, 2H), 1.84 (dd, J = 17.1, 7.1 Hz, 4H), 1.79
(5) Villa, A.; Veith, G. M.; Prati, L. Angew. Chem., Int. Ed. 2010, 49
(26), 4499−4502.
(
(
13
d, J = 0.7 Hz, 15H), 1.44 (h, J = 7.5 Hz, 4H), 1.02−0.93 (m, 6H).
C
́
(6) Ruiz, V. R.; Velty, A.; Santos, L. L.; Leyva-Perez, A.; Sabater, M.
NMR (101 MHz, CDCl ) δ 165.62, 122.19, 120.79, 92.91, 49.94,
J.; Iborra, S.; Corma, A. J. Catal. 2010, 271 (2), 351−357.
3
I
DOI: 10.1021/acs.organomet.8b00081
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(
7) Haider, M. H.; Dummer, N. F.; Zhang, D.; Miedziak, P.; Davies,
(34) Li, X. W.; Wang, X. B.; Xu, H. J.; Yu, X. J.; Zhuo, S. P. Chin. J.
Inorg. Chem. 2014, 30 (6), 1381−1387.
T. E.; Taylor, S. H.; Willock, D. J.; Knight, D. W.; Chadwick, D.;
Hutchings, G. J. J. Catal. 2012, 286, 206−213.
(
(35) (a) Manas, M. G.; Campos, J.; Sharninghausen, L. S.; Lin, E.;
Crabtree, R. H. Green Chem. 2015, 17 (1), 594−600. (b) Salman, A.
W.; Haque, R. A.; Budagumpi, S.; Zetty Zulikha, H. Polyhedron 2013,
49 (1), 200−206. (c) Budagumpi, S.; Haque, R. A.; Salman, A. W.;
Ghdhayeb, M. Z. Inorg. Chim. Acta 2012, 392, 61−72.
8) Wee, Y.-J.; Kim, J.-N.; Ryu, H.-W. Food Technol. Biotechnol. 2006,
4
(
4 (2), 163−172.
9) (a) Dusselier, M.; Van Wouwe, P.; Dewaele, A.; Makshina, E.;
Sels, B. F. Energy Environ. Sci. 2013, 6 (5), 1415−1442. (b) Amass, W.;
Amass, A.; Tighe, B. Polym. Int. 1998, 47 (2), 89−144.
(36) Li, Y.; Nielsen, M.; Li, B.; Dixneuf, P. H.; Junge, H.; Beller, M.
Green Chem. 2015, 17 (1), 193−198.
(
10) Jamshidian, M.; Tehrany, E. A.; Imran, M.; Jacquot, M.;
Desobry, S. Compr. Rev. Food Sci. Food Saf. 2010, 9 (5), 552−571.
11) Castillo Martinez, F. A.; Balciunas, E. M.; Salgado, J. M.;
Domínguez Gonzalez, J. M.; Converti, A.; Oliveira, R. P. d. S. Trends
Food Sci. Technol. 2013, 30 (1), 70−83.
12) Abdel-Rahman, M. A.; Sonomoto, K. J. Biotechnol. 2016, 236,
76−192.
13) Sharninghausen, L. S.; Campos, J.; Manas, M. G.; Crabtree, R.
H. Nat. Commun. 2014, 5, 5084.
14) Lu, Z.; Demianets, I.; Hamze, R.; Terrile, N. J.; Williams, T. J.
ACS Catal. 2016, 6 (3), 2014−2017.
15) Sun, Z. M.; Liu, Y. Q.; Chen, J. B.; Huang, C. Y.; Tu, T. ACS
Catal. 2015, 5, 6573.
16) (a) Hahn, F. E.; Jahnke, M. C. Angew. Chem., Int. Ed. 2008, 47
17), 3122−3172. (b) Nelson, D. J.; Nolan, S. P. Chem. Soc. Rev. 2013,
2 (16), 6723−6753. (c) Peris, E. Chem. Rev. 2017, DOI: 10.1021/
acs.chemrev.6b00695.
17) Li, Y.; Nielsen, M.; Li, B.; Dixneuf, P. H.; Junge, H.; Beller, M.
Green Chem. 2015, 17 (1), 193−198.
18) Sharninghausen, L. S.; Mercado, B. Q.; Crabtree, R. H.; Hazari,
N. Chem. Commun. 2015, 51 (90), 16201−16204.
19) Sharninghausen, L. S.; Crabtree, R. H. Isr. J. Chem. 2017, 57,
37.
20) (a) Larhed, M.; Moberg, C.; Hallberg, A. Acc. Chem. Res. 2002,
5 (9), 717−727. (b) Gronnow, M. J.; White, R. J.; Clark, J. H.;
Macquarrie, D. J. Org. Process Res. Dev. 2005, 9 (4), 516−518.
21) Hayes, B. L. Microwave Synthesis: Chemistry at the Speed of Light;
CEM Publising: Matthews, NC, 2002.
22) Azua, A.; Mata, J. A.; Peris, E.; Lamaty, F.; Martinez, J.;
Colacino, E. Organometallics 2012, 31 (10), 3911−3919.
23) Wang, K.; Dimitrakis, G.; Irvine, D. J. Chem. Eng. Process. 2017,
22, 389−396.
24) Gabriel, C.; Gabriel, S.; Grant, E. H.; Grant, E. H.; Halstead, B.
S. J.; Mingos, D. M. P. Chem. Soc. Rev. 1998, 27 (3), 213−223.
25) Kappe, C. O.; Pieber, B.; Dallinger, D. Angew. Chem., Int. Ed.
013, 52 (4), 1088−1094.
26) Schaper, L.-A.; Hock, S. J.; Herrmann, W. A.; Ku
Chem., Int. Ed. 2013, 52 (1), 270−289.
27) Azua, A.; Finn, M.; Yi, H.; Beatriz Dantas, A.; Voutchkova-
Kostal, A. ACS Sustainable Chem. Eng. 2017, 5 (5), 3963−3972.
28) (a) Jantke, D.; Cokoja, M.; Pothig, A.; Herrmann, W. A.; Ku
(37) Kappe, C. O. Acc. Chem. Res. 2013, 46 (7), 1579−1587.
(
̈ ̈
(38) Gulcemal, S.; Gulcemal, D.; Whitehead, G. F. S.; Xiao, J. Chem. -
́
Eur. J. 2016, 22 (30), 10513−10522.
(39) Zhang, S. J.; Foyle, S. D.; Okrut, A.; Solovyov, A.; Katz, A.;
Gates, B. C.; Dixon, D. A. J. Phys. Chem. A 2017, 121 (26), 5029−
5044.
(
1
(
(40) (a) Ladipo, F. T.; Kooti, M.; Merola, J. S. Inorg. Chem. 1993, 32
(9), 1681−1688. (b) Blum, O.; Milstein, D. J. Am. Chem. Soc. 2002,
(
́
124 (38), 11456−11467. (c) Morales-Morales, D.; Redon, R.; Wang,
Z.; Lee, D. W.; Yung, C.; Magnuson, K.; Jensen, C. M. Can. J. Chem.
001, 79 (5−6), 823−829.
41) Milet, A.; Dedieu, A.; Kapteijn, G.; vanKoten, G. Inorg. Chem.
2
(
(
1
(
(
(
997, 36 (15), 3223−3231.
(
42) SAINT Software; Bruker AXS Inc.: Madison, WI, 2007.
43) APEX II Software; Bruker AXS Inc.: Madison, WI, 2008.
44) Krause, L.; Herbst-Irmer, R.; Sheldrick, G. M.; Stalke, D. J. Appl.
(
4
Crystallogr. 2015, 48 (48), 3−10.
45) Palatinus, L.; Chapuis, G. J. Appl. Crystallogr. 2007, 40 (4), 786−
90.
46) Altomare, A.; Cascarano, G.; Giacovazzo, C.; Guagliardi, A.;
Burla, M. C.; Polidori, G.; Camalli, M. J. Appl. Crystallogr. 1994, 27,
35.
47) Sheldrick, G. M. Acta Crystallogr., Sect. C: Struct. Chem. 2015, 71,
−8.
(
(
7
(
(
(
9
(
3
4
(
3
(
(
(
48) Farrugia, L. J. J. Appl. Crystallogr. 2012, 45, 849−854.
49) Spek, A. L. J. Appl. Crystallogr. 2003, 36 (1), 7−13.
(
50) Crystal Maker; Crystal Maker Software Limited: Bicester, U.K.,
2
009.
(
(
1
(
(
2
(
̈
hn, F. E. Angew.
(
(
̈
̈
hn,
F. E. Organometallics 2013, 32 (3), 741−744. (b) Azua, A.; Sanz, S.;
Peris, E. Chem. - Eur. J. 2011, 17 (14), 3963−3967. (c) Jantke, D.;
̈
Pardatscher, L.; Drees, M.; Cokoja, M.; Herrmann, W. A.; Kuhn, F. E.
ChemSusChem 2016, 9 (19), 2849−2854.
29) (a) Rentzsch, C. F.; Tosh, E.; Herrmann, W. A.; Kuhn, F. E.
Green Chem. 2009, 11 (10), 1610−1617. (b) Jimenez, M. V.;
Fernandez-Tornos, J.; Perez-Torrente, J. J.; Modrego, F. J.; García-
Orduna, P.; Oro, L. A. Organometallics 2015, 34 (5), 926−940.
30) Hintermair, U.; Englert, U.; Leitner, W. Organometallics 2011,
0 (14), 3726−3731.
31) (a) Chianese, A. R.; Li, X.; Janzen, M. C.; Faller, J. W.; Crabtree,
(
́
́
́
̃
(
3
(
R. H. Organometallics 2003, 22 (8), 1663−1667. (b) Frey, G. D.;
Rentzsch, C. F.; von Preysing, D.; Scherg, T.; Muhlhofer, M.;
Herdtweck, E.; Herrmann, W. A. J. Organomet. Chem. 2006, 691 (26),
5
(
(
725−5738.
32) Vogt, M.; Pons, V.; Heinekey, D. M. Organometallics 2005, 24
8), 1832−1836.
33) Gnanamgari, D.; Sauer, E. L. O.; Schley, N. D.; Butler, C.;
Incarvito, C. D.; Crabtree, R. H. Organometallics 2009, 28 (1), 321−
25.
(
3
J
DOI: 10.1021/acs.organomet.8b00081
Organometallics XXXX, XXX, XXX−XXX