Alcohol-assisted base-free hydrogenation of acetophenone catalyzed by OsH(NHCMe2CMe2NH2)(PPh3)2

Article abstract of DOI:10.1139/cjc-2014-0060

The hydrido-amido complex OsH(NHCMe₂CMe₂NH ₂)(PPh₃)₂ (1) catalyzes the base-free hydrogenation of ketones in benzene. Kinetic studies using acetophenone revealed that the system has an induction period, after which the rate of the reaction increases. A constant rate was observed when a critical amount of the product alcohol was added, indicating that the reaction is autocatalytic in 1-phenylethanol. Varying the initial conditions showed that the reaction rate is dependent on hydrogen and catalyst concentration and independent of ketone concentration. Above the critical concentrations of 1-phenylethanol, the reaction rate is independent of alcohol concentration. The rate law for pressures up to 5 atm was found to be rate = d[alcohol]/dt = -d[ketone]/dt = k[Os][H₂], with k = 30 mol L^-1s^-1 at 293 K and the temperature dependence provided energy of activation parameters. Therefore, the heterolytic splitting of dihydrogen is rate determining under these conditions. Only small kinetic isotope effects were measured in contrast with the analogous ruthenium system. Complex 1 reacts with the product alcohol 1-phenylethanol and is partially converted into the dihydride complex trans-OsH₂(NH₂CMe₂CMe₂NH ₂)(PPh₃)₂ and acetophenone; with excess alcohol, an osmium alkoxide is observed at low temperature. As expected from these results, 1 is a ketone transfer hydrogenation catalyst in isopropanol.

Full text of DOI:10.1139/cjc-2014-0060

731
ARTICLE
Alcohol-assisted base-free hydrogenation of acetophenone
catalyzed by OsH(NHCMe2CMe2NH2)(PPh3)2
Sean E. Clapham, Marco Zimmer-De Iuliis, Katharina Mack, Demyan E. Prokopchuk,
and Robert H. Morris
Abstract: The hydrido–amido complex OsH(NHCMe2CMe2NH2)(PPh3)2 (1) catalyzes the base-free hydrogenation of ketones in
benzene. Kinetic studies using acetophenone revealed that the system has an induction period, after which the rate of the reaction
increases. A constant rate was observed when a critical amount of the product alcohol was added, indicating that the reaction is
autocatalytic in 1-phenylethanol. Varying the initial conditions showed that the reaction rate is dependent on hydrogen and catalyst
concentration and independent of ketone concentration. Above the critical concentrations of 1-phenylethanol, the reaction rate is
independent of alcohol concentration. The rate law for pressures up to 5 atm was found to be rate = d[alcohol]/dt = −d[ketone]/dt =
k[Os][H2], with k = 30 mol L−1 s−1 at 293 K and the temperature dependence provided energy of activation parameters. Therefore, the
heterolytic splitting of dihydrogen is rate determining under these conditions. Only small kinetic isotope effects were measured
in contrast with the analogous ruthenium system. Complex 1 reacts with the product alcohol 1-phenylethanol and is partially
converted into the dihydride complex trans-OsH2(NH2CMe2CMe2NH2)(PPh3)2 and acetophenone; with excess alcohol, an osmium
alkoxide is observed at low temperature. As expected from these results, 1 is a ketone transfer hydrogenation catalyst in
isopropanol.
Key words: catalytic hydrogenation, hydride, ketone, autocatalysis, isotope effect, osmium complex.
Résumé : Le complexe d’hydrure et d’amidure OsH(NHCMe2CMe2NH2)(PPh3)2 (1) catalyse l’hydrogénation non basique des
cétones dans le benzène. Des études cinétiques réalisées en présence d’acétophénone ont révélé que le système réactionnel
présentait une période d’induction après laquelle le taux de la réaction augmente. Un taux constant a été observé lorsque qu’une
quantité importante d’alcool est ajoutée, ce qui montre que la réaction est autocatalytique en présence de 1-phényléthanol. En
faisant varier les conditions initiales, on montre que le taux de la réaction est indépendant de la concentration en cétone et qu’il
dépend des concentrations en hydrogène et en catalyseur. Lorsque l’on dépasse des valeurs de concentration élevées en
1-phényléthanol, le taux de la réaction devient indépendant de la concentration en alcool. Pour des pressions allant jusqu’a`
5 atm, on a constaté que la loi de vitesse de la réaction est la suivante : vitesse = d[alcool]/dt = −d[cétone]/dt = k[Os][H2], où k =
30 mol L−1 s−1, a` 293 K. La relation de dépendance entre k et la température a permis d’obtenir les paramètres liés a` l’énergie
d’activation de la réaction. Ainsi, la rupture hétérolytique influe sur la vitesse de réaction dans ces conditions. Seuls de faibles
effets isotopiques cinétiques ont été mesurés, en comparaison a` ceux du système analogue comportant du ruthénium. Le complexe 1
réagit avec l’alcool 1-phényléthanol et se transforme partiellement en le complexe de dihydrure trans-OsH2(NH2CMe2CMe2NH2)(PPh3)2
et en acétophénone. Lorsque l’alcool est en excès, on constate a` basse température la présence d’un alcoolate d’osmium. Comme on
peut s’y attendre au vu de ces résultats, 1 est un catalyseur de l’hydrogénation de cétone par transfert dans l’isopropanol. [Traduit par
la Rédaction]
Mots-clés : hydrogénation catalytique, hydrure, cétone, autocatalyse, effet isotopique, complexe d’osmium.
the ligand from ␤-hydride elimination and thus preventing the
catalyst from converting to a dihydride with an imine form of the
Introduction
The mechanisms of the homogenous H2 hydrogenation under
ligand. The concept of metal−ligand bifunctional catalysis, first
base-free conditions of ketones catalyzed by well-defined ruthe-
proposed by Noyori, applies well to these catalytic systems.5,6 That
nium complexes has been studied previously by our research
group. These complexes include trans-RuH2(R-binap)(tmen), where
is, both the metal and the ligand, in this case an amido group,
participate in the heterolytic splitting of H2, the rate-determining
step of the reaction. Noyori’s group studied the mechanism of
hydrogenation of the benchmark substrate acetophenone in iso-
propanol using a well-defined precursor RuH(BH4)(S-tolbinap)(S,S-
dpen) and found that under basic conditions, it also proceeded via
rate-determining splitting of dihydrogen across the ruthenium−
amido bond with a kinetic isotope effect (KIE) kH2/kD2 of 2.0.7 We
tmen is (NH2CMe2CMe2NH2),1,2 RuH(NHCMe2CMe2NH2)(PPh3)2,1
(S,S)-RuHCl(PPh2C6H4CH2NHC6H10NHCH2C6H4PPh2)/KOiPr,3 and
RuH(app)(S-binap), where app is the amide version of the ligand
2-amino-2-(2-pyridyl)-propane.4 An essential catalytic structure
contains a ruthenium−amido bond that causes the heterolytic
splitting of H2 producing a dihydride and an amine group on
ruthenium. The diamine ligands tmen and app used in these stud-
ies do not possess ␤-hydrogen atoms to protect the amido form of
reported a kH2/kD2 of 2.1
0.1 for the hydrogenation of ace-
Received 3 February 2014. Accepted 11 March 2014.
S.E. Clapham, M.Z.-D. Iuliis, K. Mack, D.E. Prokopchuk, and R.H. Morris. Department of Chemistry, University of Toronto, 80 St. George Street,
Toronto, ON M5S 3H6, Canada.
Corresponding author: Robert Morris (e-mail: rmorris@chem.utoronto.ca).
This article is part of a Special Issue dedicated to Professor Barry Lever in recognition of his contributions to inorganic chemistry across Canada and beyond.
Can. J. Chem. 92: 731–738 (2014) dx.doi.org/10.1139/cjc-2014-0060
Published at www.nrcresearchpress.com/cjc on 17 March 2014.

732
Can. J. Chem. Vol. 92, 2014
Fig. 1. Alcohol-assisted transition states proposed by Noyori and
co-workers (I), Ikariya and co-workers (II), Casey and co-workers (III),
and Morris and co-workers (IV).
tophenone in benzene catalyzed by our trans-RuH2(tmen)(R-binap)
catalyst.2 The use of KIE in the study of organometallic reaction
mechanisms has recently been reviewed.8
To further explore the reactivity of the group 8 metal triad, the
osmium(II) hydrido−amido complex OsH(HNCMe2CMe2NH2)(PPh3)2
(1) and the trans-dihydride complex OsH2(H2NCMe2CMe2NH2)(PPh3)2
(2) were examined as catalysts in the hydrogenation of ace-
tophenone to 1-phenylethanol in benzene. Initial experiments
using 1 as a catalyst revealed an increase in the rate of reaction as
the reaction proceeded.9 Such behaviour could indicate the oper-
ation of autocatalysis as first proposed for the system RuH(app)
(S-binap),4 that is, the product alcohol speeds up the reaction by
lowering the activation barrier for the rate-determining step.
The autocatalytic behaviour of our RuH(app)(S-binap) system
was attributed to the phenylethanol product assisting in the split-
ting of the H−H bond by inserting its −OH moiety into the four-
membered ring of the Ru(H2)(amido) transition state structure to
produce a lower energy six-membered transition state, as shown in
Fig. 1, IV.4 Noyori and co-workers found that isopropanol was the best
solvent for the H2 hydrogenation of ketones. They had previously
proposed a similar transition state structure.7 Alcohol-assisted
transition states for other ruthenium-based bifunctional H2
hydrogenation catalysts were also suggested by Ikariya10 and
Casey.11
We describe here our investigation of the kinetics of catalysis by
the OsH(NHCMe2CMe2NH2)(PPh3)2 (1) complex.
J. Young NMR tube and dissolved in C6D6 to give an orange solu-
tion. 1H NMR (600 MHz) and 31P{1H} NMR spectra were measured
at 10 min, 7 h, and 24 h at 25 °C and were found to be identical
apart from the appearance of some minor peaks in the last spectrum.
Experimental details
General
Unless otherwise stated, all manipulations were carried out
under argon using standard Schlenk and glovebox techniques.
OsHCl(tmen)(PPh3)2 and OsH(NHCMe2CMe2NH2)(PPh3)2 were
prepared as described.9 All solvents were distilled under argon
using sodium/benzophenone ketyl. Acetophenone was purchased
from Sigma Aldrich, dried over P2O5, and distilled under argon.
Acetophenone-CD3 (99% CD3) was purchased from Aldrich and
was freeze−thaw degassed and dried over 4 Å molecular sieves.
Toluene-d8 was purchased in ampoules from Aldrich, dried with
activated 3 Å molecular sieves, and stored under argon. NMR spec-
tra were recorded on Agilent 600 and Varian 400 and 300 MHz
spectrometers. The 1H NMR spectra were recorded at 600, 400 or
300 MHz, the 13C NMR spectra at 100 or 75 MHz and the 31P NMR
spectra at 242, 161 or 121 MHz. The 1H NMR and 13C NMR spectrum
peaks were referenced relative to partially deuterated solvent
peaks (mostly benzene-d6), and the 31P NMR chemical shifts were
measured relative to the external reference 85% H3PO4.
In each case, resonances assigned to 1 2
(0.005 mol L−1),
(0.005 mol L−1), acetophenone (0.005 mol L−1), and 1-phenylethanol
(0.2 mol L−1). The reaction was repeated with wet 1-phenylethanol.
In this case, broad hydride resonances were also present at −20
and −21 ppm and these are thought to be associated with
OsH(OH)(tmen)(PPh3)2 solvated with alcohol. After 6 h, the J. Young
1
NMR tube was filled with hydrogen gas (ϳ1 atm) and H NMR
spectra were obtained again immediately and after several hours.
A significant decrease in the signals of 1 was observed, leaving the
dihydride 2 as the major species. The signals due to the hydroxide
species disappeared more slowly. The acetophenone resonance
disappeared leaving only 2 and 1-phenylethanol.
(c) Observation of OsH(OCHPhMe)(tmen)(PPh3)2 (3). The toluene-d8
solution of b was cooled to −80 °C and spectra were collected. The
concentrations of species observed are reported in the calculation in
entry 3 of Table 1 (see below). 3: 1H NMR: −22.5 br (OsH), 0.51, 0.60,
0.87, 0.88 (br, diastereotopic methyl of tmen due to chiral alkox-
ide), 1.42, 2.30, 4.27, 5.30 (br, four inequivalent NH), 5.18 (br,
OCHMePh), 6.8–8.1 ppm (Ph). 31P{1H} NMR: 25.0 (br), 13.3 ppm (br).
Attempted reaction of OsH(NHCMe2CMe2NH2)(PPh3)2 (1)
with acetophenone
OsH(NHCMe2CMe2NH2)(PPh3)2 (20 mg, 0.0230 mmol) and ace-
tophenone (11 mg, 0.0916 mmol) were added to an NMR tube and
Hydrogenation of acetophenone using OsHCl(tmen)(PPh3)2
and base
1
dissolved in C6D6. An orange-red solution was observed. The H
Under nitrogen, OsHCl(tmen)(PPh3)2 (5 mg, 0.00494 mmol), po-
tassium tert-butoxide (10 mg, 0.0891 mg), and acetophenone
(1.000 g, 8.323 mmol) were added to a Schlenk flask resulting in an
orange solution. The solution was purged with hydrogen and
stirred under 1 atm hydrogen gas at room temperature for 3 h.
After this time, an aliquot was dissolved in CDCl3 and the 1H NMR
spectrum was taken. Integration of the methyl peaks of ace-
tophenone and 1-phenylethanol was used to determine a conver-
sion of 16%.
and 31P{1H} NMR spectra were obtained and only the presence of
the starting materials was detected.
Reaction of OsH(NHCMe2CMe2NH2)(PPh3)2 with 1-phenylethanol
(a) Under argon, OsH(NHCMe2CMe2NH2)(PPh3)2 (20 mg, 0.0230 mmol)
and 1-phenylethanol (10 mg, 0.0819 mmol, ϳ3.6 equiv.) were
added to an NMR tube and dissolved in C6D6. An orange-red solu-
tion was observed. The 1H and 31P{1H} NMR spectra were obtained.
The presence of 1 (1H NMR (OsH) −24 ppm, br; 31P{1H} NMR
29 ppm, br) and 1-phenylethanol were detected as well as the
presence of trans-OsH2(tmen)(PPh3)2 (2) (1H NMR (OsH) −7.1 ppm;
31P{1H} NMR 37 ppm) and acetophenone. Integration of the hy-
Transfer hydrogenation
Under argon, OsH(NHCMe2CMe2NH2)(PPh3)2 (13 mg, 0.01565 mmol)
and acetophenone (104 mg, 0.8656 mmol) were dissolved in
2-propanol (1.005 g, 1.280 mL) and stirred in a round-bottom flask.
The catalyst and substrate concentrations were 0.01223 and
0.676 mol L−1, respectively, giving a substrate to catalyst ratio of
55:1. An orange suspension formed due to the insolubility of the
1
dride peaks in the H NMR spectrum of 1 and 2 provided a 1:1
hydride ratio, indicating a 2:1 ratio of 1 to 2.
(b) Under argon, OsH(NHCMe2CMe2NH2)(PPh3)2 (9 mg, 0.0108 mmol)
and 1-phenylethanol (26 mg, 1.21 mmol, 20 equiv.) were added to a
Published by NRC Research Press

Clapham et al.
733
Table 1. Equilibria by adding dry 1-phenylethanol to complex 1 in benzene-d6 or toluene-d8 under argon in a sealed
NMR tube as observed by 1H and 31P NMR.
Entry
Equilibrium
Equationa
Temperature (°C)
K (mol L−1
)
1
2
3
1 + ROH = 2 + AC
1 + ROH = 3
1 + ROH = 3
K4 = ([2][AC])/([1][ROH]) = ((0.005)(0.005))/((0.005)(0.2))
K3 = [3]/([1][ROH]) = (0.005)/((0.0005)(0.2))
K3 = [3]/([1][ROH]) = (0.0004)/((0.005)(0.2))
25
−80
25
0.03
50
0.4
aAC, acetophenone; ROH, 1-phenylethanol.
osmium complex in isopropanol. The orange suspension gradu-
ally became a yellow solution with a small amount of suspended
yellow solid. After 3.5 h, the reaction mixture was filtered through
celite in air and diluted to twice the volume with acetone. The
sample was analyzed by GC to determine an 80% conversion of
acetophenone to 1-phenylethanol.
As well, the new sampling method allowed for double the number
of data points to be obtained, leading to more accurate results.
This setup minimized the loss of precious deuterium gas. Stan-
dard solutions of catalyst, acetophenone, and 1-phenylethanol in
benzene were prepared by dissolving the desired amount in 10 or
5 mL volumetric flasks in an argon-filled glove box. Reaction mix-
tures were prepared by pipetting the appropriate amount of sub-
strate and alcohol solutions into a 3 mL volumetric flask and catalyst
solution into a 2 mL volumetric flask and diluting with benzene to
give a total volume of 5 mL upon mixing. The substrate−alcohol
solution and catalyst solution were taken up in a single 10 mL syringe
and then injected into the reactor against a flow of gas. Samples were
taken every 2 min. Concentrations of 1-phenylethanol were deter-
mined via GC. To ensure proper stirring, a specialized neodymium
stirring bar was utilized and the stirring was checked before each
reaction by placing the bottom of the reactor over the stir plate to
ensure proper alignment of the stir plate and stir bar.
Preparation of racemic 1-phenylethanol-d5
1-Phenylethanol-d5 was prepared by deuterating 5 mL of neat
acetophenone-CD3 catalyzed by50 mg of RuH(NHCMe2CMe2NH2)
(PPh3)2 at 15 atm of D2 at 25 °C for 24 h. The resulting mixture was
then distilled at reduced pressure under argon and stored over
molecular sieves. Attempts to prepare CH3CD(OD)Ph from protio
acetophenone using the same conditions resulted in deuterium
scrambling into the methyl group of the product alcohol. The
1H NMR spectrum using C6D6 and 2H NMR spectrum using C6H6 of
nondeuterated 1-phenylethanol was compared to the synthesized
1-phenylethanol-d5 to confirm that the correct protons had been
labelled.
Without adequate stirring, diffusion of H2 or D2 into the solu-
tion does not occur as efficiently and the rate of the reaction
proceeds extremely slowly.
Kinetics: concentration dependence studies
The hydrogenation reactions were carried out in a 50 mL Parr
reactor fitted with a glass sleeve and magnetic stirrer. Standard solu-
tions of 1, acetophenone, and 1-phenylethanol in benzene were
made to concentrations of 2.407 × 10−3, 0.8323, and 0.8210 mol L−1,
respectively. Stock solutions were diluted by the required amount
of benzene to achieve the desired concentrations. Catalyst solu-
tions were made up to 2 mL separately from the ketone−alcohol
solutions that were made up to 3 mL and drawn up into separate
syringes. The reactor was degassed and pressurized to the desired
pressure with hydrogen gas. The reactor was kept at a desired
constant temperature with a Fischer Scientific Isotemp constant
temperature bath. The ketone−alcohol solution was injected into
the reactor under a flow of hydrogen followed by the catalyst
solution, totaling 5 mL. Samples were removed at intervals by
temporarily venting the reactor and drawing up a small aliquot
by syringe and exposing it to air. The ratio of 1-phenylethanol to
acetophenone in each aliquot was determined by GC.
Results
Like the ruthenium analogue RuH(NHCMe2CMe2NH2)(PPh3)2,
the new osmium–amido complex 1 is an active ketone hydrogenation
catalyst. With an acetophenone concentration of 0.17 mol L−1 in
benzene and a ratio of acetophenone to 1 of 347:1 and the mild
conditions of 5 atm of hydrogen and 20 °C, the turnover frequency
(TOF) continuously increases, reaching a maximum of 0.4 s−1
with 98% conversion to 1-phenylethanol in 20 min. Under the
same conditions using RuH(NHCMe2CMe2NH2)(PPh3)2, the TOF
continuously decreases from an initial maximum TOF of 1 s−1,
achieving 99% conversion in 20 min (Fig. 2). The hydrido–chloro
complex OsHCl(tmen)(PPh3)2 is a precatalyst that is activated by
reaction with KOtBu. This combination can be used to hydrogenate
neat acetophenone to 1-phenylethanol under the mild conditions of
1 atm of dihydrogen and room temperature. Reported osmium cata-
lysts for the hydrogenation of ketones require higher temperatures
(>80 °C).12–18 An initial comparison of the kinetic behaviour of the
osmium– and ruthenium–amido catalysts showed a distinct differ-
ence in the shape of the kinetic curves (Fig. 2). The major qualitative
difference observed between the two different catalysts is that while
the RuH(NHCMe2CMe2NH2)(PPh3)2 starts with a higher rate and
slows down as the reaction progresses, the osmium catalyst 1 dis-
plays an induction period with a relatively low rate, which then
increases as the reaction progresses. The decrease of the rate ob-
served in reactions using the ruthenium catalyst is due to the in-
crease in alcohol concentration, which reacts with the catalyst to
form an alkoxide complex. The decrease in available catalyst is re-
sponsible for the decrease in reaction rate. To test whether alcohol
concentration was responsible for the increasing rate observed when
using the osmium catalyst 1, the reaction was repeated with an ini-
tial concentration of 1-phenylethanol added to the reaction mixture.
The addition of 1-phenylethanol to the reaction mixture resulted
in a marked increase in reaction rate compared to the initial reaction
with no alcohol added. In Fig. 3, the plots of 1-phenylethanol concen-
tration versus time are given for the reactions with and without
alcohol. The observed induction time for the reaction without an
Isotope effect studies
All of the hydrogenations were carried out at constant pres-
sures of H2 or D2 using a stainless steel 50 mL Parr hydrogenation
reactor fitted with a glass sleeve and a magnetic stirrer. The reac-
tor was modified such that in one of the ports, a stainless steel
sampling tube was added. The tubing was 30 cm in length with an
inner diameter of 0.01 in. The tube was inserted to the bottom of
the reaction vessel and then secured in place using a gas-tight
compression fitting. A swing valve was then attached to the other
end of the sampling loop. Samples were taken during the reaction
by opening the swing valve, allowing an aliquot of the reaction
mixture to be sampled.2 For each data point, two aliquots of equal
volume were removed from the reactor, the first of which was
discarded. The hydrogenation kinetic study described above uti-
lized a different sampling method that required the pressure in
the reactor to be dropped to atmospheric pressure so that an
aliquot of the reaction mixture could be extracted via syringe.
Utilizing this newly developed sampling tube method, the data
points could be obtained at smaller time intervals and also closer
to the start of the reaction with minimal changes to the pressure.
Published by NRC Research Press

734
Can. J. Chem. Vol. 92, 2014
Free hydrogen gas could not be detected in the 1H NMR spectrum
due to overlap with the methine hydrogen of 1-phenylethanol. No
hydrogen is expected because the calculated concentration of ace-
tophenone equals that of the concentration of the dihydride 2
within experimental error:
Fig. 2. A comparison of the course of the hydrogenation of ace-
tophenone using the catalysts RuH(NHCMe2CMe2NH2)(PPh3)2 and
OsH(NHCMe2CMe2NH2)(PPh3)2. [M] = 4.8 × 10−4 mol L−1
,
[acetophenone]i = 0.1665 mol L−1, 5 atm H2, 5 mL of C6H6.
(2)
When 20 equiv. of 1-phenylethanol was added to 1 in C6D6 or
toluene-d8, the equilibrium shifted toward the dihydride 2, now
in a ratio of 1:1 with 1. A broad resonance of the amido hydride is
shifted downfield to −23.75 ppm from −23.85 in the absence of
alcohol. This is actually the weighted average of the hydride
chemical shifts of the amido complex 1 at −23.85 and a small
amount of the alkoxide complex OsH(OCHMePh)(tmen)(PPh3)2 3
at −22.5 ppm, which are in fast exchange. The last resonance can
be observed when the solution is cooled to −80 °C. In fact, the equi-
librium shifts dramatically toward the alkoxide complex at low tem-
perature, as indicated by the equilibrium constants reported in
Table 1. These were determined from the integration of proton
resonances and relating these to the starting concentrations of 1
and alcohol. The alkoxide complex 3 has spectra very similar to
those of RuH(OR)(tmen)(PPh3)2,1 RuH(OCHMePh)(R-binap)(tmen),1
and RuH(OCHMePh)(R-binap)((R,R)-NH2CHPhCHPhNH2)19 reported
previously. The reaction of the osmium–amido complex with wa-
ter to produce a hydride hydroxide complex OsH(OH)(tmen)(PPh3)2
has also been reported but this is only observed when the added
alcohol contains water.20
Fig. 3. The effect of 1-phenylethanol concentration on the rate of
hydrogenation using complex 1 and acetophenone in benzene.
[Os] = 3.19 × 10−4, [acetophenone] = 0.1634 mol L−1, 5 atm H2, 20 °C,
5 mL of C6H6. [1-Phenylethanol]i = 0.1604 mol L−1 (squares), no initial
1-phenylethanol added (triangles).
The ability of 1 to dehydrogenate 1-phenylethanol implied its
activity as a transfer hydrogenation catalyst. To test its effective-
ness as a transfer hydrogenation catalyst, 1 and acetophenone were
stirred in 2-propanol under argon in a ratio of 55:1. After a period
of 3.5 h, 80% conversion of acetophenone to 1-phenylethanol was
achieved with a TOF of 13 h−1. It is of note that this is much slower
than the TOF of 0.4 s−1 achieved for the preliminary H2 hy-
drogenation described above (Figs. 2 and 3). Turnover frequencies
of 120 and 60 h−1 were achieved for the transfer hydrogenation of
acetophenone using OsH2Cl2(PiPr3)2 and OsH2Cl2(PMetBu2)2, respec-
tively, as catalysts in Werner’s group.17 Although both reported TOF
are higher, it should be noted that they performed their reactions at
83 °C, while we performed ours at room temperature.
initial concentration of alcohol has been eliminated in the reaction
with alcohol added.
To examine the effects of the substrate and product on the
catalyst, stoichiometric reactions of the amido–hydrido complex
1, and product 1-phenylethanol and the substrate acetophenone
were conducted. The results were unlike those of the analogous
ruthenium complex RuH(NHCMe2CMe2NH2)(PPh3)2, which re-
sulted in alkoxide and enolate complexes.1 Instead, no reaction
was observed between acetophenone and 1:
Kinetics
Kinetic experiments were performed to determine the effect of
the catalyst, substrate, and hydrogen concentrations on the rate
of hydrogenation of acetophenone to 1-phenylethanol using 1 as a
catalyst. Previous results from preliminary hydrogenation exper-
iments using 1 indicated that the rate of hydrogenation was in-
creased by the presence of the product alcohol (Fig. 3). Therefore,
kinetic experiments were needed with an alcohol present at the
start. The fact that 1 is also a transfer hydrogenation catalyst
precluded the use of 2-propanol as a solvent due to competing
transfer and H2 hydrogenation processes. Instead, the kinetic ex-
periments were performed in benzene as the solvent, and nearly
equal concentrations of the substrate acetophenone and the
product 1-phenylethanol were added before beginning the hy-
drogenation. The faster alcohol-assisted rate (Table 2) was thereby
achieved throughout the hydrogenation and any transfer hy-
drogenation from 1-phenylethanol to acetophenone had no net
effect on the concentration of the substrate or product.
(1)
When 1 was reacted with 4 equiv. of 1-phenylethanol in C6D6, as in
eq. 2, an equilibrium was established where some 1-phenylethanol
was dehydrogenated to produce the dihydride 2 and acetophenone.
Integration of the hydride peaks of 1 and 2 in the 1H NMR spectrum
indicate a 2:1 ratio of 1 to 2. While the 31P{1H} signal at 37.1 ppm for
1 was sharp, the amido peak at 29.3 ppm was broadened. The
experiment was performed in a closed NMR tube, which pre-
vented any loss of hydrogen gas from 2 from leaving the system.
The effect of the catalyst concentration on the rate of the reac-
tion can be seen in Fig. 4. As the concentration of 1 is increased,
keeping the hydrogen pressure and initial substrate and alcohol
Published by NRC Research Press

Clapham et al.
735
Table 2. Dependence of the rate of production of alcohol catalyzed by 1 on the concentrations of 1, H2, acetophenone,
and temperature.
Pressure H2 (atm);
[H2] (mol L−1
[Ketone]
Rate r
(mol L−1 s−1
k = r/([Os][H2])
(mol L−1 s−1
)
Run
[1] (mol L−1
)
)
(mol L−1
)
Temperature (K)
)
1
2
3
4
5
6
7
8
9.6×10−5
4.8×10−5
1.4×10−4
9.6×10−5
9.6×10−5
1.4×10−4
1.4×10−4
1.4×10−4
1.4×10−4
2.7×10−4
2.7×10−4
2.7×10−4
2.7×10−4
2.7×10−4
5.0; 1.32×10−2
5.0; 1.32×10−2
5.0; 1.32×10−2
5.0; 1.32×10−2
5.0; 1.32×10−2
2.0; 5.28×10−3
5.0; 1.32×10−2
9.0; 2.38×10−2
14.0; 3.8×10−2
5.0; 1.32×10−2
5.0; 1.36×10−2
5.0; 1.39×10−2
5.0; 1.24×10−2
5.0; 1.19×10−2
0.1665
0.1665
0.1665
0.1332
0.1998
0.1665
0.1665
0.1665
0.1665
0.1665
0.1665
0.1665
0.1665
0.1665
293
293
293
293
293
293
293
293
293
293
298
303
288
283
4.6×10−5
2.1×10−5
8.3×10−5
4.5×10−5
5.8×10−5
4×10−5
36
33
45
36
46
26
32
26
26
26
28
31
6.0×10−5
9×10−5
9
10×10−5
9.3×10−5
1.03×10−4
1.16×10−4
7.4×10−5
6.1×10−5
10a
11a
12a
13a
14a
22
19
Note: Initial 1-phenylethanol concentration = 0.164 mol L−1
aTemperature varied; different setup and catalyst batch.
.
Fig. 4. The effect of osmium concentration on the initial rate of ace-
tophenone hydrogenation (runs 1−3 in Table 2). [1-Phenylethanol]i =
0.1642 mol L−1, 5 atm H2, [acetophenone]i = 0.1665 mol L−1, 5 mL of
C6H6. Molar osmium concentrations are given in the legend.
Fig. 5. The effect of hydrogen concentration on the initial rate of
acetophenone hydrogenation (runs 6−8 in Table 2). [Os] =
1.44 × 10−4 mol L−1, [acetophenone]i = 0.1665 mol L−1, [1-phenyletha-
nol]i = 0.1642 mol L−1, 5 mL of C6H6. Hydrogen pressures 2, 5, 9, and
14 atm, respectively.
concentrations constant, the rate of hydrogenation increases pro-
portionally.
When the concentration of 1 and the initial concentration of
product, along with the pressure of hydrogen, were kept constant
and the substrate concentration was varied, no change in the rate
of hydrogenation was observed. The straight lines observed in all
of the kinetic experiments reinforces the fact that there is no
dependence on the substrate concentration. The linearity also
means that, although the presence of alcohol increases the rate of
reaction compared to the rate of reaction in the absence of alco-
hol, the reaction is not directly dependent on the concentration of
the alcohol. This is consistent with the observation of constant
rate to complete conversion. Varying the hydrogen pressure (and
thereby the hydrogen concentration) and keeping the catalyst
concentration and the initial substrate and alcohol concentra-
tions constant caused an increased rate of hydrogenation with
higher hydrogen pressures (Fig. 5). The hydrogen concentration
dependence is more complicated than that of previously reported
ruthenium systems. At pressures lower than 5 atm, we assume
that the rate is first order in hydrogen concentration but that at
higher pressures, another step becomes rate determining and the
hydrogen pressure dependence levels off.
Fig. 6. An Eyring plot for temperatures ranging from 283 to 303 K
(runs 10–14 in Table 1).
Thus, the rate law for the alcohol-assisted hydrogenation of
acetophenone using 1 as a catalyst at up to 5 atm H2 is given by
At a temperature of 20 °C, a rate constant k of 34 mol L−1 s−1 was
obtained. By contrast, the analogous ruthenium complex has the
same rate law with k of 1.1 × 102 mol L−1 s−1.1 Thus, the ruthenium
complex is three times more active under comparable conditions.
The temperature dependence (and isotope effect studies) were
(3)
Rate ϭ d[1 Ϫ phenylethanol]/dt
ϭ Ϫd[acetophenone]/dt ϭ k[1][H2]
Published by NRC Research Press

736
Can. J. Chem. Vol. 92, 2014
Table 3. Average initial reaction rates, rate constants k, and the isotope effects using catalyst 1.
Kinetic run
H*2/PhCOCH*3/ROH*
Rate of reaction,
␯0 (10−5 mol L−1 s−1
Rate constant k =
␯0/([Os][H2 ]) (mol L−1 s−1
Isotope effect
(kH/kD)
)
)
HHH
HHD
HDD
DDD
11 1
11 1
11 1
10 1
26
26
26
24
1.0 0.1
1.0 0.1
1.1 0.1
Note: Runs were carried out using a single syringe in which the catalyst, substrate, and alcohol were mixed prior
to injection into the reactor. [Os] = 3.18 × 10−4 mol L−1, [acetophenone] or [acetophenone-d3] = 0.163 mol L−1
[1-phenylethanol] or [1-phenylethanol-d5] = 0.160 mol L−1, 5 atm of H2 or D2 = 1.32 × 10−2 mol L−1, 20 °C, 5 mL of C6H6.
,
Fig. 7. Sample plot of the HHH (squares) and DDD (circles) runs. The
error bars represent one standard deviation.
determined using a different reactor setup and batch of catalyst.
The rate constant in this case was 26 mol L−1 s−1 at 293 K. The
Eyring plot (Fig. 6) used five rates from temperatures spaced 5°
apart from 10 to 30 °C. The thermodynamic values for activation
enthalpy and entropy are ⌬H‡ = 3.7 kcal mol−1 and ⌬S‡
=
−39 cal mol−1 with a Gibbs free energy of activation of ⌬G‡ (293 K) =
15 kcal mol−1. The values correspond to a transition state that is
more ordered than the reactants due to the loss of entropy of the
gaseous hydrogen.
Isotope effect measurements
The rate data to probe isotope effects were collected in at least
triplicate to improve the accuracy of the data (Table 3). The error
in the rates is 10%. The experiments requiring deuterium gas were
measured at 5 atm D2 pressure. A few runs that gave a value that
deviated by more than 10%, probably because of leak or impurity,
were rejected. The purity of the catalyst was verified weekly by 1H
and 31P NMR and by the standard rate generated by the all-protio
system.
To differentiate between the protio and deutero gas, ace-
tophenone, and 1-phenylethanol for each run, a three-letter des-
ignation code was used such that the first letter of the code
represents the gas, the second letter the ketone, and the third
letter the alcohol that was used in the experiment. H is used for
protio species, while D is used for deutero species. For example, a
set of runs that used H2 gas, protio acetophenone, and protio
1-phenylethanol would be designated HHH.
Scheme 1. A possible mechanism for the hydrogenation of ace-
tophenone to 1-phenylethanol using OsH(NHCMe2CMe2NH2)(PPh3)2
as a catalyst: the nonassisted splitting of dihydrogen (I) and the
alcohol-assisted splitting of dihydrogen (II).
A typical comparison of plots for the HHH and DDD runs is
shown in Fig. 7. The slight delay in the first 60 s is likely to be
associated with temperature changes when the solution is added
to the reactor and dissolution of gas into the solution, slower for
deuterium. The lower initial pressure of gas allows the reverse of
reaction 4, the dehydrogenation/dedeuteration of the added alco-
hol, to occur, resulting in an initial drop in alcohol concentration.
Discussion
The rate law for the hydrogenation of acetophenone catalyzed
by 1 in the presence of excess product alcohol with pressures less
than 5 atm displays a dependence on the concentration of catalyst
and the concentration of hydrogen. This indicates that the rate-
limiting step involves the addition of hydrogen to the catalyst, a
feature similar to the system using the ruthenium hydrido–amido
complex RuH(NHCMe2CMe2NH2)(PPh3)2 as a catalyst.1 What is not
taken into account is the effect of alcohol on the rate. At relatively
low concentrations of alcohol (ϳ <0.02 mol L−1), the rate of hy-
drogenation is slower than that when higher concentrations of
alcohol are present (ϳ >0.02 mol L−1) (Fig. 3). Yet at higher alcohol
concentrations, the rate of hydrogenation is independent of alco-
hol concentration. The alcohol is changing the nature of the rate-
limiting step in such a way as to lower the activation energy of
that transition state. At low concentrations of alcohol, the hy-
drogenation reaction operates by a slow mechanism, and at
higher concentrations of alcohol, it operates by a faster “alcohol-
assisted” mechanism. The non-alcohol-assisted mechanism (I in
Scheme 1) is the same as that proposed for the analogous ruthe-
nium system,1 where hydrogen is heterolytically split across a
metal–amido bond generating a metal hydride and amine. The
alcohol-assisted mechanism (II in Scheme 1) involves a six-
membered ring consisting of the metal center, a molecule of hy-
drogen, the OH of a molecule of alcohol, and the amido nitrogen
(Fig. 1). Rather than splitting hydrogen directly across the osmium–
amido bond, the hydrogen is split between the metal and alcohol
oxygen, while the amido nitrogen is protonated by the alcohol pro-
Published by NRC Research Press

Clapham et al.
737
Scheme 2. Routes to explain the formation of the alkoxide complex 3.
the current work is that the alkoxide forms quickly from the
amido complex and alcohol and is more prevalent at low temper-
atures. In the low dielectric constant solvent benzene, hydride
transfer from 2 to acetophenone to give a transient free alkoxide
seems unfavourable, making a concerted H+/H− transfer route
more likely, as proposed previously for catalysts operating in
benzene−alcohol mixtures.
Conclusions
The synthesis of the amino–amido complex of osmium
OsH(NHCMe2CMe2NH2)(PPh3)2 has allowed a detailed study of the
(base-free) hydrogenation of acetophenone in benzene catalyzed
by this well-defined complex. This osmium complex can split dihy-
drogen across its metal−amido bond to produce a trans-dihydride
2 and this behaviour is somewhat analogous to the ruthenium
complex RuH(NHCMe2CMe2NH2)(PPh3)2. Where the osmium com-
plex differs from the ruthenium complex is in the reactivity of the
amido nitrogen. Stronger bonding between osmium and the
amido nitrogen makes it less reactive than the ruthenium
counterpart.9 The osmium–amido complex can act as a transfer
hydrogenation catalyst as well as a H2 hydrogenation catalyst. It
acts most efficiently as a direct hydrogenation catalyst for ace-
tophenone when alcohol is present.
The kinetic isotope effect has been determined experimen-
tally for the homogeneous hydrogenation of ketones using
OsH(NHCMe2CMe2NH2)(PPh3)2 (1) and found to be about 1.1 0.1.
The next step will be to use density functional theory calculations
to assist in finding a mechanism that explains the lower KIE for
osmium than the value of 2.1 found for the related ruthenium
system RuH2(NH2CMe2CMe2NH2)(R-binap).
ton. This proton shuttle mechanism is proposed to facilitate the
addition of hydrogen to the catalyst, thereby raising the rate of
hydrogenation. This alcohol effect was not observed for the re-
lated ruthenium system,1 and this likely reflects the higher reac-
tivity of the Ru−N(amido) bond compared to the Os−N bond
toward dihydrogen and other small molecules.9,21
The observed KIEs are negligible or small, with a value of 1.1
0.1 for the DDD system, comparing the all-protio system with
the system using deuterium gas, deuterated acetophenone (at
the methyl), and deuterated alcohol (except for the phenyl group)
(Table 3). Deuteration of the gas and substrate does not seem to
significantly affect the rates (the DDH and DHH systems have a
KIE of 1.0 0.1). The value of 1.1 contrasts with the ruthenium
system trans-RuH2(tmen)(binap), which has a kH/kD of 2.1 0.1
for the direct heterolytic splitting of dihydrogen/dideuterium
across the Ru−N(amido) bond in benzene.2 The proton shuttle
mechanism like that shown in Scheme 1 appears to lack a KIE.
Density functional theory is currently being applied to further
interpret this result and understand this mechanism more
deeply.
Acknowledgements
Funding was provided by a Discovery Grant to R.H.M. from
NSERC and Deutscher Akademischer Austauschdienst (DAAD) to
K.M.
References
(1) Abdur-Rashid, K.; Clapham, S. E.; Hadzovic, A.; Harvey, J. N.; Lough, A. J.;
Morris, R. H. J. Am. Chem. Soc. 2002, 124, 15104. doi:10.1021/ja016817p.
(2) Zimmer-De Iuliis, M.; Morris, R. H. J. Am. Chem. Soc. 2009, 131, 11263. doi:10.
1021/ja9043104.
(3) Rautenstrauch, V.; Hoang-Cong, X.; Churlaud, R.; Abdur-Rashid, K.;
Morris, R. H. Chem. Eur. J. 2003, 9, 4954. doi:10.1002/chem.200304884.
(4) Hadzovic, A.; Song, D.; MacLaughlin, C. M.; Morris, R. H. Organometallics
2007, 26, 5987. doi:10.1021/om700849w.
(5) Noyori, R.; Ohkuma, T. Angew. Chemie Int. Ed. Engl. 2001, 40, 40. doi:10.1002/
1521-3773(20010105)40:1<40::AID-ANIE40>3.0.CO;2-5.
(6) Noyori, R.; Yamakawa, M.; Hashiguchi, S. J. Org. Chem. 2001, 66, 7931. doi:
10.1021/jo010721w.
(7) Sandoval, C. A.; Ohkuma, T.; Muniz, K.; Noyori, R. J. Am. Chem. Soc. 2003, 125,
13490. doi:10.1021/ja030272c.
NMR experiments to examine the reaction of alcohol with com-
plex 1 have revealed several details. The triplet hydride resonance
of the dihydride 2 in the presence of alcohol is broadened to a
peak width of 18 Hz at 600 MHz at 25 °C. This is attributed to the
formation of an OsH···HOR dihydrogen bond (hydridic protonic
interaction) as shown in 2.ROH in Scheme 2; this causes a short-
(8) Gomez-Gallego, M.; Sierra, M. A. Chem. Rev. 2011, 111, 4857. doi:10.1021/
cr100436k.
(9) Clapham, S. E.; Morris, R. H. Organometallics 2005, 24, 479. doi:10.1021/
om048998k.
(10) Ito, M.; Hirakawa, M.; Murata, K.; Ikariya, T. Organometallics 2001, 20, 379.
doi:10.1021/om000912+.
(11) Casey, C. P.; Johnson, J. B.; Singer, S. W.; Cui, Q. J. Am. Chem. Soc. 2005, 127,
3100. doi:10.1021/ja043460r.
(12) Sánchez-Delgado, R. A.; Rosales, M.; Esteruelas, M. A.; Oro, L. A. J. Mol. Cat. A
1995, 96, 231. doi:10.1016/1381-1169(94)00039-5.
(13) Schluenken, C.; Esteruelas, M. A.; Lahoz, F. J.; Oro, L. A.; Werner, H. Eur. J.
Inorg. Chem. 2004, 2477. doi:10.1002/ejic.200300898.
(14) Rosales, M.; Gonzalez, A.; Navarro, J.; Soscun, H.; Zarraga, J. Inorg. Chim. Acta
1997, 257, 131. doi:10.1016/S0020-1693(96)05434-5.
(15) Sánchez-Delgado, R. A.; Medina, M.; Lopez-Linares, F.; Fuentes, A. J. Mol. Cat.
A 1997, 116, 167. doi:10.1016/S1381-1169(96)00191-4.
(16) Cheong, M.-K.; Jun, M.-J. Bull. Korean Chem. Soc. 1996, 17, 569.
(17) Aracama, M.; Esteruelas, M. A.; Lahoz, F. J.; Lopez, J. A.; Meyer, U.; Oro, L. A.;
Werner, H. Inorg. Chem. 1991, 30, 288. doi:10.1021/ic00002a025.
(18) Sánchez-Delgado, R. A.; Lee, W.; Choi, S. R.; Cho, Y.; Jun, M. J. Transition Met.
Chem. 1991, 16, 241. doi:10.1007/BF01032844.
(19) Hamilton, R. J.; Bergens, S. H. J. Am. Chem. Soc. 2008, 130, 11979. doi:10.1021/
ja8034812.
ening of the T1 and broadening of the hydride resonance22–29
A
detail still to be determined is the route followed when 1 reacts
with alcohol to give the dihydride 2 and the alkoxide 3 (Scheme 2).
This is the equivalent question, discussed in the literature, as to
whether such dihydride complexes react with acetophenone to
give directly the alcohol and the amido complex 1 via H+/H− trans-
fer to the ketone in the transition state (path 2 to a polarized TS,
then path 3 to 1, and then path 4 to the alkoxide complex 3
of Scheme 2)1,2,4,5,7,30 or to give an alkoxide complex 3 via an
initial hydride transfer (path 2 to an ion pair TS that includes a
hydrogen-bonded alkoxide and then path 5 to complex 3 in
Scheme 2).19,31–36 Arguments have been made for each pathway
depending on the solvent and other conditions. An observation in
Published by NRC Research Press

738
Can. J. Chem. Vol. 92, 2014
(20) Prokopchuk, D. E.; Collado, A.; Lough, A. J.; Morris, R. H. Dalton Trans. 2013,
42, 10214. doi:10.1039/c3dt50452b.
(21) Clapham, S. E.; Guo, R.; Zimmer-De Iuliis, M.; Rasool, N.; Lough, A.;
Morris, R. H. Organometallics 2006, 25, 5477. doi:10.1021/om060661q.
(22) Lough, A. J.; Park, S.; Ramachandran, R.; Morris, R. H. J. Am. Chem. Soc. 1994,
116, 8356. doi:10.1021/ja00097a049.
(29) Custelcean, R.; Jackson, J. E. Chem. Rev. 2001, 101, 1963. doi:10.1021/
cr000021b.
(30) Haack, K. J.; Hashiguchi, S.; Fujii, A.; Ikariya, T.; Noyori, R. Angew. Chem. Int.
Ed. Engl. 1997, 36, 285. doi:10.1002/anie.199702851.
(31) Hamilton, R. J.; Bergens, S. H. J. Am. Chem. Soc. 2006, 128, 13700. doi:10.1021/
ja065460s.
(23) Lee, J. C., Jr.; Rheingold, A. L.; Muller, B.; Pregosin, P. S.; Crabtree, R. H.
J. Chem. Soc. Chem. Comm. 1994, 1021. doi:10.1039/C39940001021.
(24) Shubina, E. S.; Belkova, N. V.; Epstein, L. M. J. Organomet. Chem. 1997,
536–537, 17. doi:10.1016/S0022-328X(96)06791-5.
(25) Crabtree, R. H.; Eisenstein, O.; Sini, G.; Peris, E. J. Organomet. Chem. 1998, 567,
7. doi:10.1016/S0022-328X(98)00661-5.
(32) Takebayashi, S.; Dabral, N.; Miskolzie, M.; Bergens, S. H. J. Am. Chem. Soc.
2011, 133, 9666. doi:10.1021/ja202732q.
(33) O, W. W. N.; Lough, A. J.; Morris, R. H. Organometallics 2012, 31, 2137. doi:10.
1021/om300108p.
(34) Prokopchuk, D. E.; Morris, R. H. Organometallics 2012, 31, 7375. doi:10.1021/
om300572v.
(26) Klooster, W. T.; Koetzle, T. F.; Siegbahn, P. E. M.; Richardson, T. B.;
Crabtree, R. H. J. Am. Chem. Soc. 1999, 121, 6337. doi:10.1021/ja9825332.
(27) Silantyev, G. A.; Filippov, O. A.; Tolstoy, P. M.; Belkova, N. V.; Epstein, L. M.;
Weisz, K.; Shubina, E. S. Inorg. Chem. 2013, 52, 1787. doi:10.1021/ic301585k.
(28) McGrady, G. S.; Guilera, G. Chem. Soc. Rev. 2003, 32, 383. doi:10.1039/
b207999m.
(35) Hasanayn, F.; Morris, R. H. Inorg. Chem. 2012, 51, 10808. doi:10.1021/ic301233j.
(36) Bertoli, M.; Choualeb, A.; Lough, A. J.; Moore, B.; Spasyuk, D.;
Gusev, D. G. Organometallics 2011, 30, 3479. doi:10.1021/om200437n.
Published by NRC Research Press