Partial deoxygenation of glycerol catalyzed by iridium pincer complexes

Article abstract of DOI:10.1021/cs400551g

Iridium pincer complexes (POCOP)Ir(CO) (POCOP = κ³-C ₆H₃-1,3-[OP(^tBu)₂]₂) and substituted POCOP derivatives catalyze deoxygenation of glycerol to n-propanol and 1,3-propanediol in good yield under moderate conditions (acidic aqueous dioxane, 200 C, 80 bar H₂). Catalyst solubility in the polar reaction mixture is improved by incorporation of a polar moiety in the para position of the POCOP phenyl ring, with the best results obtained with a dimethylamino substituent.

Full text of DOI:10.1021/cs400551g

Research Article
pubs.acs.org/acscatalysis
Partial Deoxygenation of Glycerol Catalyzed by Iridium Pincer
Complexes
David B. Lao, Alisa C. E. Owens, D. Michael Heinekey,* and Karen I. Goldberg*
Department of Chemistry, University of Washington, Box 351700, Seattle, Washington 98195-1700, United States
S
* Supporting Information
ABSTRACT: Iridium pincer complexes (POCOP)Ir(CO)
(POCOP = κ3-C6H3-1,3-[OP(tBu)2]2) and substituted
POCOP derivatives catalyze deoxygenation of glycerol to n-
propanol and 1,3-propanediol in good yield under moderate
conditions (acidic aqueous dioxane, 200 °C, 80 bar H2).
Catalyst solubility in the polar reaction mixture is improved by incorporation of a polar moiety in the para position of the
POCOP phenyl ring, with the best results obtained with a dimethylamino substituent.
KEYWORDS: glycerol, catalysis, deoxygenation
examples of heterogeneous catalysts, there are relatively few
examples of homogeneous catalysts. Braca has reported a
ruthenium iodocarbonyl catalyst for the deoxygenation of
glycerol to 1-PO with up to 90% selectivity.13 Schlaf has
recently reported ruthenium aqua complexes with cocatalyst
triflic acid for the deoxygenation of glycerol to 1-PO with up to
18% yield.14,15
We recently reported the deoxygenation of the glycerol
model substrate 1,2-propanediol (1,2-PD) to 1-PO using a
bis(phosphinite) (POCOP) iridium pincer catalyst (1).16
Speciation studies showed that under the reaction conditions,
the iridium carbonyl complex (2) was formed by alcohol
decarbonylation. Complex 2 was then shown to be an effective
catalyst for the deoxygenation of 1,2-PD.
INTRODUCTION
■
Dwindling petroleum reserves have led to significant focus on
the development of renewable sources for fuels and commodity
chemicals. Because of its infrastructure compatibility with
current transportation fuels, biodiesel has drawn considerable
interest, and production has grown significantly in recent years.
The rapid growth in biodiesel production has resulted in the
generation of large quantities of byproduct glycerol. While
glycerol has many uses as a commodity chemical, the
production scale of transportation fuels far exceeds that of
commodity chemicals. As a result of high production quantities
that exceed market demand and high purification costs, waste
glycerol from biodiesel production is often burned onsite as fuel
for the production process. The transformation of waste stream
glycerol into a C-3 platform chemical would improve upon the
atom efficiency and the economics of the biodiesel industry.1−5
Potential C-3 target chemicals from glycerol are 1,3-
propanediol (1,3-PD) and 1-propanol (1-PO). 1,3-PD is
utilized in the production of polyesters, polyurethanes, and
polyethers. The conversion of glycerol to 1,3-PD can be
envisioned as a net deoxygenation via a tandem catalytic
sequence with an initial acid catalyzed dehydration followed by
metal catalyzed hydrogenation (eq 1).
Herein, we report an extension of this work for catalytic
deoxygenation of glycerol to 1-PO and 1,3-PD using POCOP
iridium carbonyl complex 2 and derivatives.
RESULTS AND DISCUSSION
■
Deoxygenation of glycerol in the C2 and C3 positions would
afford 1-PO, a useful solvent. Although 1-PO has lower value
than 1,3-PD, it is still a value-added chemical made from
glycerol, particularly so if generated from crude biodiesel
byproduct glycerol.
Catalysts capable of selectively deoxygenating glycerol must
be stable in aqueous acidic solutions at high temperatures.
There have been many recent heterogeneous catalysts reported
for glycerol deoxygenation with various transition metals
including Pt,6−9 Rh,10−12 Ru,4,5 and Cu.4,5 Despite all the
Increasing Solubility in Polar Reaction Mixtures.
Consistent with our previous studies of deoxygenation of 1,2-
PD, it was found that dioxane is a suitable solvent for this
chemistry. However, in glycerol/water/dioxane mixtures,
complex 2 has limited solubility at room temperature and at
Received: July 12, 2013
Revised: August 30, 2013
© XXXX American Chemical Society
2391
dx.doi.org/10.1021/cs400551g | ACS Catal. 2013, 3, 2391−2396

ACS Catalysis
Research Article
elevated temperatures. A series of pincer ligands were prepared
to increase catalyst solubility in polar reaction mixtures
containing glycerol and water. Novel ligands containing
diethylene glycol monomethyl ether (3a) and dimethyl amine
(3b) were prepared by synthesizing the corresponding 5-
substituted resorcinol17,18 followed by diphosphorylation.19
We propose that 1,2-PD was not observed because of the
rapid deoxygenation of 1,2-PD to 1-PO as determined in our
previous studies16 and in control reactions. Control reactions
have also shown that under our reaction conditions, 1,3-PD was
not deoxygenated to 1-PO. It is reasonable to then suggest that
1-PO is formed via deoxygenation at the glycerol C1 position
followed by a subsequent deoxygenation at the C2 position of
1,2-PD. Our previous studies have shown that deoxygenation of
1,2-PD to 1-PO occurs at much lower temperatures (125 °C)
than the reaction conditions used for the deoxygenation of
glycerol. Deoxygenation at the glycerol C2 position leads to
formation of 1,3-PD.
The selectivity of product formation appears to change over
time. In the first 6 h, 1,3-PD and 1-PO are formed in an
approximate 1:2 ratio, respectively. Over the course of 48 h, the
1,3-PD: 1-PO selectivity diminishes to 1:4. As noted previously,
this decrease in 1,3-PD selectivity is not because of conversion
of 1,3-PD to 1-PO or decomposition of products as shown by
overall mass balance. One possible explanation for the change
in selectivity may be a change in the catalyst over the course of
the reaction. A change in the reaction medium over the course
of the reaction may also result in the different selectivity with
time.
Iridium carbonyl catalysts prepared from these ligands
showed improved solubility over 2 in reaction mixtures
containing glycerol in aqueous dioxane. Complete catalyst
solubility in reaction mixtures at room temperature enabled
reliable reaction optimization experiments to be performed.
Because of ease of ligand synthesis and improved solubility in
acidic polar solutions, the dimethyl amine complex 4 formed
from ligand 3b was used in all catalytic reactions described here.
The effect of reaction temperature on conversion and
selectivity was investigated (Figure 2). Lowering the reaction
Deoxygenation of Glycerol. In a typical experiment,
glycerol was converted to 1-PO and 1,3-PD in aqueous dioxane
solutions with 0.125 mol % 4 relative to glycerol and 1 mol %
sulfuric acid relative to glycerol at 200 °C under 80 bar H2.
Reaction progress was monitored by quantitative 13C{1H}
NMR spectroscopy of neat reaction solutions (see Exper-
imental Section for details). Approximately 45% conversion of
glycerol to 1-PO and 1,3-PD was observed after 48 h (Figure
1). Bubbling the reaction headspace gas through acetone-d6 and
1
characterizing by H NMR spectroscopy showed formation of
small amounts of propane and propylene. Weighing reaction
vessels before and after heating and venting showed good mass
balance suggesting that propane and propylene were formed in
trace amounts. No appreciable amounts of 1,2-PD were
observed in these reactions.
Figure 2. Conversion of glycerol to 1-propanol and 1,3-propanediol at
different temperatures. Reactions with 0.0065 mmol of 4, 5.2 mmol of
glycerol, and 0.052 mmol of H2SO4 in deionized water (0.197 g) and
dioxane (1.22 g) under 80 bar H2. Reactions heated for 24 h.
Concentrations measured by qualitative 13C NMR spectroscopy (see
Experimental Section for details).
temperature from 200 to 180 °C for 24 h resulted in the
reduction of glycerol conversion from about 35% to 9%.
Further lowering the temperature to 170 °C resulted in
reduction of glycerol conversion to 4%.
The reaction selectivity is also a function of temperature.
Reactions carried out at 180 °C show a 1:1 ratio of 1,3-PD to 1-
PO. The selectivity increased to a 2:1 ratio of 1,3-PD to 1-PO
when the reaction was carried out at 170 °C.
We have previously shown that deoxygenation of 1,2-
propanediol at 185 °C requires H2 pressures above 40 bar to
achieve high selectivities (90%) to 1-PO.16 Comparable
selectivities and conversions for glycerol deoxygenation at
200 °C for 20 h were observed when H2 pressures between 40
and 80 bar were used (Figure 3). Notably, the use of a mixture
of 40 bar CO and 40 bar H2 in our reactions resulted in a
decrease of conversion from 24% to 3%, but maintained a 1,3-
Figure 1. Time course of glycerol conversion to 1-propanol and 1,3-
propanediol. Reactions with 0.0065 mmol of 4, 5.2 mmol of glycerol,
and 0.052 mmol of H2SO4 in deionized water (0.197 g) and dioxane
(1.22 g) under 80 bar H2. Reactions heated to 200 °C. Concentrations
measured by quantitative 13C NMR spectroscopy (see Experimental
Section for details).
2392
dx.doi.org/10.1021/cs400551g | ACS Catal. 2013, 3, 2391−2396

ACS Catalysis
Research Article
To test the applicability of our reaction conditions to the
development of a practical process, the deoxygenation of waste
glycerol from biodiesel production was also studied. Waste
stream glycerol contains impurities such as sodium hydroxide
and methanol necessary for the transesterification process as
well as carboxylates. While distillation is the most common
method for purification on an industrial scale, it is an energy
intensive process. In lieu of distillation, we have found that
waste glycerol can be purified by acidifying with 1 M sulfuric
acid, allowing the biphasic mixture to settle, and decanting the
impurity layer containing various free acids. The resulting
solution of water, methanol, and glycerol can be diluted in
dioxane and used directly in deoxygenation reactions. In our
initial test reactions, recycled waste glycerol was deoxygenated
in 20% yield over 24 h, with a 1,3-PD to 1-PO selectivity of 1:9.
The high selectivity for 1-PO in these reactions is due to the
high acid concentration employed in the purification of crude
glycerol.
Mechanistic Studies. In our previous work, we reported
that the air and water stable iridium carbonyl precatalyst,
(POCOP)Ir(CO), could be used in place of (POCOP)Ir(H)2
for the deoxygenation of 1,2-propanediol. We also discovered
that upon the completion of these reactions, both (POCOP)-
Ir(CO) and trans-(POCOP)Ir(CO)(H)2 were present in the
reaction mixture. The presence of these two species was
confirmed by NMR (1H and 31P{1H}) and IR spectroscopy.
The proposed mechanism for formation of trans-(POCOP)Ir-
(CO)(H)2 from (POCOP)Ir(CO) is shown in Scheme 1.
Figure 3. Conversion of glycerol to 1-propanol and 1,3-propanediol at
different headspace pressures. Reactions with 0.0065 mmol of 4, 5.2
mmol of glycerol, and 0.052 mmol of H2SO4 in deionized water (0.197
g) and dioxane (1.22 g). Reactions heated at 200 °C for 20 h.
Concentrations measured by qualitative 13C NMR spectroscopy (see
Experimental Section for details).
PD: 1-PO selectivity of about 1:4. The decrease in conversion
in the presence of CO suggests that catalysis is hindered by CO
coordination to the catalyst. The proposed reaction mechanism
is discussed below.
The effect of acid concentration was also studied (Figure 4).
Acid concentration was varied from 0.25% to 4.0% versus
Scheme 1. Proposed Mechanism for Formation of trans-
(POCOP)Ir(CO)(H)2 from (POCOP)Ir(CO)
Initial protonation of (POCOP)Ir(CO) forms a cationic
monohydride species. Dihydrogen association and deprotona-
tion affords trans-(POCOP)Ir(CO)(H)2. It is hypothesized
that the iridium dihydride and/or dihydrogen-hydride species
are the active catalysts for the deoxygenation of glycerol.
The formation of trans-(POCOP)Ir(CO)(H)2 from
(POCOP)Ir(CO) has been verified by pressurizing a
tetrahydrofuran (THF) solution of (POCOP)Ir(CO) under
H2 for 24 h in the presence of 0.5 equiv of anilinium
tetrafluoroborate. In the absence of catalytic acid, the formation
of trans-(POCOP)Ir(CO)(H)2 is not observed. A linear
relationship was observed for H2 pressure and conversion to
trans-(POCOP)Ir(CO)(H)2. Under 110 bar of H2, up to 80%
conversion to trans-(POCOP)Ir(CO)(H)2 could be achieved,
but complete conversion was never observed. These observa-
tions are consistent with an equilibrium between trans-
(POCOP)Ir(CO)(H)2 and (POCOP)Ir(CO), with equilibra-
tion facilitated by acid. Upon neutralization of the acid, the
resulting mixture is stable for weeks at ambient temperature
without any change in composition.
Figure 4. Conversion of glycerol to 1-propanol, 1,2-propanediol, and
1,3-propanediol with varying acid equivalents. Reactions with 0.0065
mmol of 4 and 5.2 mmol of glycerol in deionized water (0.197 g) and
dioxane (1.22 g) under 80 bar H2. Reactions heated to 200 °C for 24
h. Concentrations measured by quantitative 13C NMR spectroscopy
(see Experimental Section for details).
glycerol. Acid concentrations below 4.0% were used to limit the
formation of acid catalyzed side products. An increase in acid
concentration results in an increase in glycerol conversion from
8 to 38% over 24 h at 200 °C. The increase in conversion with
acid concentration is consistent with an initial acid catalyzed
dehydration step. Selectivity for 1,3-PD decreased with
increasing acid concentration. Increasing acid concentration
from 0.25% to 4.0% led to a decrease in 1,3-PD:1-PO selectivity
from 2:1 to 1:4. Notably, with 0.25% acid, formation of 1,2-PD
was also observed as a product. 1,2-PD was not observed with
higher acid concentrations.
The dependence of selectivity on acid concentration is not
fully understood at this time. A change in the protonation level
of the catalyst may account for the increased selectivity for 1-
PO at higher acid concentration. Studies are currently
underway to determine the role of the catalyst on product
selectivity.
Reaction of (POCOP)Ir(CO) with the bis-etherate of
tetrakis[3,5-bis(trifluoromethyl)phenyl]boric acid ([H-
(OEt2)2]B(C6H3(CF3)2)4) affords the cationic Ir(III) complex
[(POCOP)Ir(CO)(H)]+, identified by the observation of a
1
triplet resonance in the H NMR spectrum at −36 ppm (J =
10.4 Hz). The structure of this cationic complex was
determined by X-ray diffraction (Figure 5).
2393
dx.doi.org/10.1021/cs400551g | ACS Catal. 2013, 3, 2391−2396

ACS Catalysis
Research Article
Scheme 2. Proposed Mechanism for Formation of 1,3-
Propanediol from Glycerol with trans-
[(POCOP)Ir(CO)(H2)(H)]+
Figure 5. ORTEP of [(POCOP)Ir(CO)(H)]+.(50% probability
ellipsoids) Counter-anion and non hydride H atoms omitted for
clarity. Significant bond distances: Ir−P = 2.3209(14) Å, 2.3196(12)
Å; Ir−CPh = 2.033(5) Å; Ir−CCO = 1.929(6) Å; Ir−H = 1.58(7) Å.
The cationic complex [(POCOP)Ir(CO)(H)]+ is a rare
example of a structurally characterized five coordinate Ir(III)
complex.20,21 For comparison, the structure of the neutral Ir(I)
complex 2 was also determined. The structure of 2 is very
reaction temperature, and the reaction time. Whether this is
due to changes in the reaction medium as the reaction
progresses or decomposition of the catalyst is not yet clear and
is under further investigation.
similar, with Ir−P = 2.2785(3) Å, 2.2800(3) Å; Ir−CPh
=
2.0488(12) Å; Ir−CCO = 1.8658(14) Å. Details of the X-ray
data collection are found in the Supporting Information.
The possible intermediacy of cationic trans-[(POCOP)Ir-
(CO)(H2)(H)]+ was investigated by low temperature 1H NMR
spectroscopy. At 180 K, a solution containing trans-(POCOP)-
Ir(CO)(H)2 in CD2Cl2 was treated with 1 equiv of
[H(OEt2)2]B(C6H3(CF3)2)4. New broad resonances at −18.7,
−23.0, and −23.3 ppm are observed upon mixing. Upon
warming the solution to 223 K, hydrogen was evolved, and the
resonances at −23.0 and −23.3 ppm disappear. The resonance
at −18.7 ppm disappears upon warming the sample to 273 K.
Complete formation of [(POCOP)Ir(CO)(H)]+ was indicated
CONCLUSION
■
The catalyzed deoxygenation of glycerol to 1,3-propanediol and
1-propanol has been achieved with sulfuric acid and an iridium
pincer complex. Conversions of up to 45% glycerol and
selectivities of 1:4 1,3-propanol to 1-propanol have been
achieved. The catalytic reactions can also be performed using
acidified waste glycerol from biodiesel production.
The transient trans-[(POCOP)Ir(CO)(H2)(H)]+ is pro-
posed to be the active catalyst for these reactions.
1
by the appearance of a resonance at −36 ppm in the H NMR
spectrum. The highly transient nature of these intermediates
hampers accurate determination of the relaxation times (T1) of
the hydride resonances, but an approximate value of about 300
ms for all three hydride signals was measured at 180 K. This
value is not consistent with the presence of a dihydrogen ligand.
Further studies are currently underway to confirm the
intermediacy of the proposed dihydrogen complex.
Proposed Mechanism for Deoxygenation. trans-
[(POCOP)Ir(CO)(H2)(H)]+ is hypothesized to be a possible
proton and hydride source for the deoxygenation of glycerol. A
proposed catalytic cycle for glycerol deoxygenation to 1,3-
propanediol via trans-[(POCOP)Ir(CO)(H2)(H)]+ is shown in
the top portion of Scheme 2. Initial protonation of (POCOP)-
Ir(CO) to [(POCOP)Ir(CO)(H)]+ followed by reaction with
hydrogen generates trans-[(POCOP)Ir(CO)(H2)(H)]+. trans-
[(POCOP)Ir(CO)(H2)(H)]+ can act as a proton and hydride
source for the deoxygenation of glycerol. An alternative
pathway is shown in the bottom of Scheme 2. Deprotonation
of trans-[(POCOP)Ir(CO)(H2)(H)]+ affords trans-(POCOP)-
Ir(CO)(H)2. An external acid can serve as a proton source, and
trans-(POCOP)Ir(CO)(H)2 can serve as a hydride source for
the deoxygenation of glycerol.
EXPERIMENTAL SECTION
■
General Procedures. All experiments were performed
under an inert atmosphere using standard glovebox or Schlenk
techniques. All NMR solvents were dried using appropriate
drying agents. All reaction solvents and reagents were used as
received.
1H, 31P, and 13C NMR spectra were obtained using a Bruker
500 MHz spectrometer. IR spectra were obtained using a
Bruker Vector 33 FT-IR spectrometer. X-ray crystal structures
were collected at −173 °C on a Bruker APEX II single crystal
X-ray diffractometer, Mo-radiation.
Reaction Quantification by 13C NMR Spectroscopy.
Qualitative studies were performed by collecting 13C NMR
spectra of neat reaction mixtures. Intensity enhancement due to
the nuclear Overhauser effect (NOE) was nullified by
integration of carbons with an equal number of protons.
Reaction mixtures were quantified by collecting 13C NMR
spectra of neat reaction mixtures. A similar procedure for
quantitative analysis of reaction mixtures was recently reported
by Gagne et al.22 Quantitative spectra were obtained by
collecting inverse-gated decoupled NMR spectra to avoid NOE
enhancement. To minimize integration problems associated
with long relaxation times, delay times of 3 s were employed,
and the paramagnetic additive Cr(acac)3 was used as a
relaxation agent.23 Sodium p-toluenesulfonate was used as an
internal standard.
Formation of 1-propanol is hypothesized to proceed through
an analogous mechanism for deoxygenation at the C1 position,
leading to 1,2-PD, which is then rapidly deoxygenated to 1-
propanol. The selectivity of the reaction (1,3-PD vs 1-PO
product) was found to depend on the acid concentration, the
2394
dx.doi.org/10.1021/cs400551g | ACS Catal. 2013, 3, 2391−2396

ACS Catalysis
Research Article
After the reaction is complete, approximately 300−400 mg of
material comes to rest in the reactor head assembly. Rinsing the
reactor head with D2O and examining the resulting solution by
1H NMR spectroscopy showed 1-propanol and solvent dioxane.
Since the exact amount of lost material was not quantified,
reported conversion percentages only account for the material
in the reactor vessels.
the volatiles were removed under vacuum, and the solid was
washed with small portions of cold pentane and methanol.
(5-Diethylene Glycol Monomethyl Ether POCOP)Ir(CO).
Because of product solubility in pentane, the condensed residue
was purified by neutral alumina column chromatography, eluted
with 10:90 Et2O:hexanes. From 0.192 g (0.37 mmol) of 5-
diethylene glycol monomethyl ether POCOP and 0.132 g of
Ir(CO)2(Cl)(p-toluidine) in 8 mL of toluene was obtained
0.081 g (30% yield) of (5-diethylene glycol monomethyl ether
POCOP)Ir(CO) as a yellow solid. 1H NMR (500 MHz, 25 °C,
CD2Cl2): δ 6.20 (s, 2H, 3- and 5-H); glycol: δ 4.05 (t, 2 H), δ
3.76 (t, 2 H), δ 3.65 (t, 2 H), δ 3.52 (t, 2 H), δ 3.34 (s, 3 H); δ
1.35 (v t, J = 7.3 Hz, 36 H, PCC-H). 31P{1H} NMR (202 MHz,
25 °C, CD2Cl2): δ 200. IR (pentane, cm−1): 1946 (νCO).
(5-Dimethylamine POCOP)Ir(CO) (4). From 0.300 g (0.68
mmol) of 5-dimethylamine POCOP and 0.266 g (0.68 mmol)
of Ir(CO)2(Cl)(p-toluidine) in 10 mL of toluene was obtained
Synthesis. (POCOP)Ir(H)(Cl),19 (POCOP)Ir(CO),24 Ir-
26
(CO)2(Cl)(p-toluidine),25 and [H(OEt2)2]B(C6H3(CF3)2)4
were prepared as previously described in the literature.
5-Diethylene Glycol Monomethyl Ether Resorcinol. Follow-
ing a similar procedure as Marzo et al.,17 to a stirring solution of
phloroglucinol (0.63 g, 5.0 mmol), potassium fluoride (0.29 g,
5.0 mmol), and potassium carbonate (0.35 g, 2.5 mmol) in 4
mL of dimethylsulfoxide (DMSO) under argon was added 2-
bromoethyl 2-methoxyethyl ether (0.92 g, 5.0 mmol) in 2 mL
of DMSO. The slurry was stirred under argon. After 48 h, 10
mL of deionized water was added to the mixture to afford a
clear orange solution. The solution was neutralized to pH 7
with 10% HCl, and the organic material was extracted with
chloroform. The organic extractions were combined, dried over
MgSO4, and condensed under vacuum to afford an orange
liquid. The crude product was purified by column chromatog-
raphy on silica, eluted with 60:40 ethyl acetate:hexane to collect
1
0.160 g (35% yield) of 4 as a bright yellow powder. H NMR
(500 MHz, 25 °C, CD2Cl2): δ 6.03 (s, 2H, 3- and 5-H); δ 2.90
(s, 6 H, NC-H); δ 1.36 (v t, J = 7.0 Hz, 36 H, PCC-H).
31P{1H} NMR (202 MHz, 25 °C, CD2Cl2): δ 200. IR (pentane,
cm-1): 1941 (νCO).
trans-(POCOP)Ir(CO)(H)2. To a stirring solution of 0.31 g
(0.50 mmol) of (POCOP)Ir(H)(Cl) and 0.056 g (0.50 mmol)
of potassium tert-butoxide in 6 mL of benzene under argon was
added 0.015 g (0.50 mmol) of paraformaldehyde. The dark red
solution was heated under static argon at 80 °C for 3 h. After
cooling, the solution was filtered through Celite and condensed
to afford a brown-yellow solid that was used without further
purification. The isolated product consisted of an about 1.6:1
mixture of (POCOP)Ir(CO) to trans-(POCOP)Ir(CO)(H)2
1
0.090 g (8% yield) of an off-white solid. H NMR (500 MHz,
25 °C, DMSO-d6): δ 9.17 (s, 2H, 2 × OH); δ 5.82 (m, 1 H, 2-
H); δ 5.77 (m, 2 H, 4- and 6-H); glycol: δ 3.92 (t, 2 H), δ 3.67
(t, 2 H), δ 3.56 (t, 2 H), δ 3.45 (t, 2 H), δ 3.25 (s, 3 H).
5-Dimethylamine Resorcinol. Following a similar procedure
as Petrzilka and Lusuardi,18 to a stirring solution of
phloroglucinol (1.0 g, 7.9 mmol) in 9 mL of H2O and 14
mL of DMF was added an aqueous solution of dimethylamine
(1.3 mL 40% w:w solution, 10 mmol). The solution was stirred
under argon for 48 h and then concentrated under vacuum to
afford a dark red oil. A pink precipitate formed upon addition of
about 30 mL of chloroform to the dark red oil. The mixture was
filtered and washed with small portions of cold chloroform to
1
1
by H- and 31P NMR integration. H NMR (500 MHz, 25 °C,
CD2Cl2): δ 6.68 (t, J = 8.2 Hz, 1H, 4-H); δ 6.37 (d, J = 7.3 Hz,
2H, 3- and 5-H); δ 1.29 (v t, J = 7.1 Hz, 36 H, PCC-H); δ
−10.08 (t, J = 14.8 Hz, 2 H, Ir−H). 31P{1H} NMR (202 MHz,
25 °C, CD2Cl2): δ 183.
[(POCOP)Ir(CO)(H)]+. Reaction of 1 equiv of trans-
(POCOP)Ir(CO)(H)2 with 1 equiv of [H(OEt2)2]B-
(C6H3(CF3)2)4 in dichloromethane affords [(POCOP)Ir(CO)-
(H)]+. X-ray quality crystals were obtained upon standing for
24 h.
1
afford 0.59 g (49% yield) of a pink solid. H NMR (500 MHz,
25 °C, DMSO-d6): δ 8.81 (s, 2 H, 2 × OH); δ 5.59 (s, 3 H, 2-,
4- and 6-H); δ 2.77 (s, 6 H, NC-H).
General Procedure for POCOP Ligand Synthesis by
Brookhart.24 5-Diethylene glycol monomethyl ether POCOP
(3a). From 0.090 g (0.39 mmol) of 5-diethylene glycol
monomethyl ether resorcinol, 0.020 g (0.83 mmol) of NaH,
and 0.150 g (0.83 mmol) of di-tert-butylchlorophosphine was
obtained 0.20 g (99% yield) of 3c as a light yellow oil. 1H NMR
(500 MHz, 25 °C, C6D6): δ 7.22 (m, 1 H, 2-H); δ 6.79 (m, 2
H, 4- and 6-H); glycol: δ 3.82 (t, 2 H), δ 3.50 (t, 2 H), δ 3.44
(t, 2 H), δ 3.32 (t, 2 H), δ 1.13 (s, 3 H); δ 3.82 (d, J = 11.7 Hz,
36 H, PCC-H). 31P{1H} NMR (202 MHz, 25 °C, C6D6): δ
151.
Glycerol Deoxygenation. Deoxygenation reactions were
carried out in Parr reactors outfitted with gas-inlet valves and
pressure sensors. In a typical reaction, 45 mL reactors equipped
with Teflon stirbars were charged with (N(CH3)2POCOP)-
IrCO (4.3 mg, 0.0065 mmol) and a solution containing glycerol
(0.479 g, 5.2 mmol), deionized water (0.197 g), and dioxane
(1.22 g). A sulfuric acid solution (1.0 M) was added with a
micropipet (52 μL, 0.052 mmol). Reactors were stirred at 400
rpm, purged with H2 for 4 min, pressurized to 80 bar, and
heated to 200 °C over 20 min (pressure increases to 110 bar).
After 24 h, the reactor was allowed to cool to room temperature
and then cooled in an ice water bath for 20 min before venting
slowly. To analyze gaseous products, the vented gas was
bubbled slowly through acetone-d6 in a J. Young NMR tube
5-Dimethylamine POCOP (3b). From 0.153 g (1.0 mmol)
of 5-dimethylamine resorcinol, 0.048 g (2.0 mmol) of NaH,
and 0.36 g (2.0 mmol) of di-tert-butylchlorophosphine was
1
1
obtained 0.44 g (99% yield) of 3b as a light orange oil. H
and analyzed by H NMR spectroscopy.
NMR (500 MHz, 25 °C, C6D6): δ 7.10 (m, 1 H, 2-H); δ 6.52
(m, 2 H, 4- and 6-H); δ 2.55 (s, 6 H, NC-H); δ 1.18 (d, J = 12
Hz, 36 H, PCC-H). 31P{1H} NMR (202 MHz, 25 °C, C6D6): δ
149.
General Procedure for (5-X-POCOP)Ir(CO). A solution of 1
equiv of 5-X-POCOP and 1 equiv of Ir(CO)2(Cl)(p-toluidine)
in toluene was heated to reflux under argon for 16 h. After 16 h,
To ensure gaseous products were not produced in
appreciable amounts, mass changes were detected using a
solution balance (August Sauter KG, Ebingen, Germany). Prior
to reaction, the reactor was charged with glycerol, solvent, and
catalyst and balanced against an equivalent mass of water. After
reaction, the reactor was cooled and vented, and this procedure
was repeated. No change in mass was observed. Independent
2395
dx.doi.org/10.1021/cs400551g | ACS Catal. 2013, 3, 2391−2396

ACS Catalysis
Research Article
(6) Daniel, O. M.; DeLaRiva, A.; Kunkes, E. L.; Datye, A. K.;
Dumesic, J. A.; Davis, R. J. ChemCatChem 2010, 2, 1107−1114.
(7) Gong, L.; Lu, Y.; Ding, Y.; Lin, R.; Li, J.; Dong, W.; Wang, T.;
Chen, W. Appl. Catal., A 2010, 390, 119−126.
experiments established that changes in mass of 5 mg could
be detected.
Reaction conditions of heating time, gas pressure, acid
concentration, and reaction temperature were varied. Time
course studies were performed by setting up a series of identical
reactions and stopping reactions at different times. For
reactions with CO and H2 mixtures, the reactors were purged
with CO for 4 min, pressurized to 40 bar of CO followed by 40
bar of H2 to increase the total pressure to 80 bar. For studies
involving varying acid concentration, the mass of deionized
water was adjusted to maintain the total water concentration.
Crude Glycerol Purification and Deoxygenation. To a
stirring solution of crude glycerol from biodiesel production
was added concentrated sulfuric acid until the solution became
cloudy (ca. pH = 0). The cloudy solution was left overnight to
allow formation of two distinct layers. The clear, colorless
bottom layer was collected and filtered to remove any
particulates.
The solution of purified glycerol (1.0 g) was diluted with
dioxane (5.1 g). Standard deoxygenation procedures were
carried out with 2.0 g of this solution and (N(CH3)2POCOP)-
Ir(CO) (6.8 mg, 0.010 mmol).
Low Temperature Protonation of trans-(POCOP)Ir-
(CO)(H)2. A J. Young NMR tube was charged with solid
[H(OEt2)2]B(C6H3(CF3)2)4 (9.8 mg, 0.0097 mmol) and trans-
(POCOP)Ir(CO)(H)2 (6.0 mg, 0.0097 mmol). The J. Young
NMR tube was evacuated, and CD2Cl2 was vacuum distilled
into the tube. The solution was kept frozen until immediately
before NMR studies at which time, tubes were partially thawed
quickly in a dry ice/isopropanol bath, inverted, and
immediately injected into an NMR probe cooled to 180 K.
(8) Oh, J.; Dash, S.; Lee, H. Green Chem. 2011, 13, 2004−2007.
(9) Qin, L.; Song, M.; Chen, C. Green Chem. 2010, 12, 1466−1472.
(10) Shimao, A.; Koso, S.; Ueda, N.; Shinmi, Y.; Furikado, I.;
Tomishige, K. Chem. Lett. 2009, 38, 540−541.
(11) Chaminand, J.; Djakovitch, L.; Gallezot, P.; Marion, P.; Pinel,
C.; Rosier, C. Green Chem. 2004, 6, 359−361.
(12) Furikado, I.; Miyazawa, T.; Koso, S.; Shimao, A.; Kunimori, K.;
Tomishige, K. Green Chem. 2007, 9, 582−588.
(13) Braca, G.; Galletti, A. M. R.; Sbrana, G. J. Organomet. Chem.
1991, 417, 41−49.
(14) Thibault, M. E.; DiMondo, D. V.; Jennings, M.; Abdelnur, P. V.;
Eberlin, M. N.; Schlaf, M. Green Chem. 2011, 13, 357−366.
(15) Taher, D.; Thibault, M. E.; DiMondo, D.; Jennings, M.; Schlaf,
M. Chem.Eur. J. 2009, 15, 10132−10143.
(16) Ahmed-Foskey, T. J.; Heinekey, D. M.; Goldberg, K. I. ACS
Catal. 2012, 2, 1285−1289.
(17) Brizzi, A.; Brizzi, V.; Cascio, M. G.; Corelli, F.; Guida, F.;
Ligresti, A.; Maione, S.; Martinelli, A.; Pasquini, S.; Tuccinardi, T.;
Marzo, V. D. J. Med. Chem. 2009, 52, 2506−2514.
(18) Petrzilka, T.; Lusuardi, W. G. Helv. Chim. Acta 1973, 56, 510−
518.
(19) Gottker-Schnetmann, I.; White, P.; Brookhart, M. J. Am. Chem.
̈
Soc. 2004, 126, 1804−1811.
(20) Ghosh, R.; Emge, T. J.; Krogh-Jespersen, K.; Goldman, A. S. J.
Am. Chem. Soc. 2008, 130, 11317−11327.
(21) Bernskoetter, W. H.; Hanson, S. K.; Buzak, S. K.; Davis, Z.;
White, P. S.; Swartz, R.; Goldberg, K. I.; Brookhart, M. J. Am. Chem.
Soc. 2009, 131, 8603−8613.
(22) McLaughlin, M. P.; Adduci, L. L.; Becker, J. J.; Gagne, M. R. J.
Am. Chem. Soc. 2013, 135, 1225−1227.
(23) Lamar, G. N. Chem. Phys. Lett. 1971, 10, 230−232.
(24) Goettker-Schenetmann, I.; White, P.; Brookhart, M. Organo-
metallics 2004, 23, 1766−1776.
ASSOCIATED CONTENT
■
S
* Supporting Information
(25) Roberto, D.; Cariati, E.; Psaro, R.; Ugo, R. Organometallics 1994,
13, 4227−4231.
(26) Brookhart, M.; Grant, B.; Volpe, A. F., Jr. Organometallics 1992,
11, 3920−3922.
Text, tables, and CIF files giving X-ray crystallographic data for
(POCOP)Ir(CO) and [(POCOP)Ir(CO)(H)]+ and details of
crystallographic data collection. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Authors
*E-mail: heinekey@chem.washington.edu (M.H.).
*E-mail: goldberg@chem.washington.edu (K.I.G).
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the Center for Enabling New
Technologies through Catalysis (CENTC, CHE-1205189). We
thank Dr. Gene Wong, Mr. Alex Wallner, Dr. N. Malathi
Weliange, and Mr. Eric Camp, for technical assistance and
useful discussions. We thank Mr. William Levin for the waste
glycerol from biodiesel production.
REFERENCES
■
(1) Schlaf, M. Dalton Trans. 2006, 39, 4645−4653.
(2) Johnson, T. D.; Taconi, K. A. Environ. Prog. 2007, 26, 338−348.
(3) Cheda, J. N.; Huber, G. W.; Dumesic, J. A. Angew. Chem., Int. Ed.
2007, 46, 7164−7183.
(4) Dam, J. T.; Hanefeld, U. ChemSusChem 2011, 4, 1017−1034.
(5) Ruppert, A, M.; Weinberg, K.; Palkovits, R. Angew. Chem., Int. Ed.
2012, 51, 2564−2601.
2396
dx.doi.org/10.1021/cs400551g | ACS Catal. 2013, 3, 2391−2396