Effect of the structure of the diamine backbone of P-N-N-P ligands in iron(II) complexes on catalytic activity in the transfer hydrogenation of acetophenone

Article abstract of DOI:10.1021/ic101548j

The asymmetric transfer hydrogenation of aromatic ketones can be efficiently accomplished using catalysts that are based on platinum group metals which are more toxic and less abundant than iron. For that reason the discovery of iron based catalysts for t

Full text of DOI:10.1021/ic101548j

Inorg. Chem. 2010, 49, 11039–11044 11039
DOI: 10.1021/ic101548j
Effect of the Structure of the Diamine Backbone of P-N-N-P ligands in Iron(II)
Complexes on Catalytic Activity in the Transfer Hydrogenation of Acetophenone
Alexandre A. Mikhailine and Robert H. Morris*
Davenport Laboratory, Department of Chemistry, University of Toronto, 80 St. George Street, Toronto,
Ontario M5S 3H6, Canada
Received August 1, 2010
The asymmetric transfer hydrogenation of aromatic ketones can be efficiently accomplished using catalysts that are
based on platinum group metals which are more toxic and less abundant than iron. For that reason the discovery of iron
based catalysts for the use in this transformation is important. To address this issue, we synthesized a new series of
iron(II)-based precatalysts trans-[Fe(Br)(CO)(PPh2CH2CHdNCHRCHRNdCHCH2PPh2)]BPh4 (5a-5d) contain-
ing P-N-N-P ligands with the diamines (R,R)-1,2-diaminocyclohexane (a), (R,R)-1,2-diphenyl-1,2-diaminoethane
(b), (R,R)-1,2-di(4-methoxyphenyl)-1,2-diaminoethane (c), and ethylenediamine (d) incorporated in the backbone
using a convenient one-pot synthesis using readily available starting materials. All of the complexes, when activated
with a base, show a very high activity in the transfer hydrogenation catalysis of acetophenone, using 2-propanol as a
reducing agent under mild conditions. A comparison of the TOF of complexes 5a-5d show that the catalytic activity of
complexes increase as the size of the substituents in the backbone of ligands increases (d < a < b = c).
Introduction
group metals (PGM) in combination with different reducing
agents.3-33 On the other hand, the use of less toxic, more
abundant and inexpensive metals would be highly beneficial.
Recently, several groups have replaced PGM-containing
catalysts with those based on first row transition metals.
Iron, in particular, is an interesting alternative because of its
chemical properties such as easily varied oxidation states,
The reduction of pro-chiral ketones to valuable chiral
alcohols is an important process for both industrial and
academic purposes.1-3 This transformation can be efficiently
accomplished using catalysts that are based on platinum
*To whom correspondence should be addressed. E-mail: rmorris@chem.
utoronto.ca.
(17) Ohkuma, T.; Utsumi, N.; Tsutsumi, K.; Murata, K.; Sandoval, C.;
Noyori, R. J. Am. Chem. Soc. 2006, 128, 8724–8725.
(18) Ooka, H.; Arai, N.; Azuma, K.; Kurono, N.; Ohkuma, T. J. Org.
Chem. 2008, 73, 9084–9093.
(19) Palmer, A. M.; Zanotti-Gerosa, A.; Nedden, H. Tetrahedron: Asym-
metry 2008, 19, 1310–1327.
(20) Praetorius, J. M.; Wang, R.; Crudden, C. M. Organometallics 2010,
29, 554–561.
(21) Qiu, L. Q.; Kwong, F. Y.; Wu, J.; Lam, W. H.; Chan, S.; Yu, W. Y.;
Li, Y. M.; Guo, R. W.; Zhou, Z.; Chan, A. S. C. J. Am. Chem. Soc. 2006, 128,
5955–5965.
(22) Reetz, M. T.; Li, X. G. J. Am. Chem. Soc. 2006, 128, 1044–1045.
(23) Sandoval, C. A.; Li, Y. H.; Ding, K. L.; Noyori, R. Chem. Asian J.
2008, 3, 1801–1810.
(24) Sandoval, C. A.; Bie, F. S.; Matsuoka, A.; Yamaguchi, Y.; Naka, H.;
Li, Y. H.; Kato, K.; Utsumi, N.; Tsutsumi, K.; Ohkuma, T.; Murata, K.;
Noyori, R. Chem. Asian J. 2010, 5, 806–816.
(25) Schlatter, A.; Woggon, W. D. Adv. Synth. Catal. 2008, 350, 995–
1000.
(26) Shen, W. Y.; Li, Y. Y.; Dong, Z. R.; Gao, J. X. Synthesis, Stuttgart
2009, 2413–2417.
(1) The Handbook of Homogeneous Hydrogenation; de Vries, J. G., Elsevier,
C. J., Eds; Wiley-VCH: Weinheim, 2007; Vols. 1-3.
(2) Blaser, H.-U.; Studer, M. Acc. Chem. Res. 2007, 40, 1348–1356.
(3) Klingler, F. D. Acc. Chem. Res. 2007, 40, 1367–1376.
(4) Arai, N.; Azuma, K.; Nii, N.; Ohkuma, T. Angew. Chem., Int. Ed.
2008, 47, 7457–7460.
(5) Baratta, W.; Chelucci, G.; Magnolia, S.; Siega, K.; Rigo, P. Chem.—
Eur. J. 2009, 15, 726–732.
(6) Baratta, W.; Barbato, C.; Magnolia, S.; Siega, K.; Rigo, P. Chem.—
Eur. J. 2010, 16, 3201–3206.
(7) Cho, B. T. Chem. Soc. Rev. 2009, 38, 443–452.
(8) Enthaler, S.; Hagemann, B.; Bhor, S.; Anilkumar, G.; Tse, M. K.;
Bitterlich, B.; Junge, K.; Erre, G.; Beller, M. Adv. Synth. Catal. 2007, 349,
853–860.
(9) Hamilton, R. J.; Bergens, S. H. J. Am. Chem. Soc. 2008, 130, 11979–
11987.
(10) Ikariya, T.; Blacker, A. J. Acc. Chem. Res. 2007, 40, 1300–1308.
(11) Ito, M.; Endo, Y.; Ikariya, T. Organometallics 2008, 27, 6053–6055.
(12) Liu, D. L.; Xie, F.; Zhao, X. H.; Zhang, W. B. Tetrahedron 2008, 64,
3561–3566.
(13) Liu, J. T.; Zhou, D.; Jia, X.; Huang, L.; Li, X. S.; Chan, A. S. C.
Tetrahedron: Asymmetry 2008, 19, 1824–1828.
(27) Wang, C.; Wu, X. F.; Xiao, J. L. Chem. Asian J. 2008, 3, 1750–1770.
(28) Wettergren, J.; Buitrago, E.; Ryberg, P.; Adolfsson, H. Chem.—Eur.
J. 2009, 15, 5709–5718.
(14) Martins, J. E. D.; Clarkson, G. J.; Wills, M. Org. Lett. 2009, 11, 847–
850.
(15) Naud, F.; Malan, C.; Spindler, F.; Ruggeberg, C.; Schmidt, A. T.;
Blaser, H. U. Adv. Synth. Catal. 2006, 348, 47–50.
(16) Ohkuma, T.; Tsutsumi, K.; Utsumi, N.; Arai, N.; Noyori, R.;
Murata, K. Org. Lett. 2007, 9, 255–257.
(29) Wu, X. F.; Li, X. H.; Zanotti-Gerosa, A.; Pettman, A.; Liu, J. K.;
Mills, A. J.; Xiao, J. L. Chem.—Eur. J. 2008, 14, 2209–2222.
(30) Guo, R.; Elpelt, C.; Chen, X.; Song, D.; Morris, R. H. Org. Lett.
2005, 7, 1757–1759.
r
2010 American Chemical Society
Published on Web 11/10/2010
pubs.acs.org/IC

11040 Inorganic Chemistry, Vol. 49, No. 23, 2010
Mikhailine and Morris
Scheme 1. Examples of Well-Defined Iron-Based Asymmetric Trans-
fer Hydrogenation Precatalysts
Scheme 2. Reported Synthesis of the Precatalyst 2
P-N-N-P ligand were mixed in situ to give a catalyst
of lower activity.34 The P-N-N-P ligands in these systems
were prepared from a commercially available phosphine
aldehyde and enantiopure diamines. In the complexes of
our initial study, the tetradentate ligands form 6-, 5-, and
6- membered rings with the iron. Although the catalyst
activity was promising, the enantiomeric excess (ee) of the
product alcohols were low, and the starting organopho-
sphorus compound was limited in availability. For this
reason, we prepared by means of a template synthesis method
a second generation of iron precatalyst 2 with a P-N-N-P
ligand that has a smaller numbers of atoms in the ligand-
metal rings (5, 5, 5).36 The diamine used in the template
synthesis, 1,2-diphenyl-1,2-diaminoethane (dpen), provided
the chirality for the catalyst. This turned out to be more active
and enantioselective than the first generation catalyst 1b with
(6, 5, 6) rings. In the current report we investigate the effect of
incorporating different diamines into the backbone of the
P-N-N-P ligand on the activity of the resulting catalysts in
the transfer hydrogenation of the aromatic ketones.
well investigated coordination properties and Lewis acidity
as well as low toxicity and high abundance.34-43 Some other
reactions such as cross-coupling, cycloadditions, and polym-
erization that are important for industrial and synthetic
applications can now be accomplished using iron-based
catalytic systems,44 although harsh conditions such as high
temperature, high catalyst loading, and additives are often
required for a successful reaction. Hand-in-hand with the
development of new catalytic processes must come studies
designed to probe the mechanism of action of these new iron
catalysts. Our approach is the preparation and utilization of
well-defined homogeneous catalysts to gain an understand-
ing of their mechanism.
Our group reported the first highly active, well-defined
iron-based asymmetric precatalysts for the reduction of
pro-chiral aromatic ketones to chiral alcohols (Scheme 1).35
Gao and co-workers had previously reported the use of a
less well-defined system where an iron carbonyl cluster and a
Results and Discussion
(31) Guo, R.; Elpelt, C.; Chen, X.; Song, D.; Morris, R. H. Chem.
Commun. 2005, 3050–3052.
(32) Guo, R.; Morris, R. H.; Song, D. J. Am. Chem. Soc. 2005, 127, 516–
517.
(33) Jerphagnon, T.; Haak, R.; Berthiol, F.; Gayet, A. J. A.; Ritleng, V.;
Holuigue, A.; Pannetier, N.; Pfeffer, M.; Voelklin, A.; Lefort, L.; Verzijl, G.;
Tarabiono, C.; Janssen, D. B.; Minnaard, A. J.; Feringa, B.; de Vries, J. G.
Top. Catal. 2010, 53, 1002–1008.
(34) Chen, J.-S.; Chen, L.-L.; Xing, Y.; Chen, G.; Shen, W.-Y.; Dong,
Z.-R.; Li, Y.-Y.; Gao, J.-X. Huaxue Xuebao 2004, 62, 1745–1750.
(35) Sui-Seng, C.; Freutel, F.; Lough, A. J.; Morris, R. H. Angew. Chem.,
Int. Ed. 2008, 47, 940–943.
(36) Mikhailine, A. A.; Lough, A. J.; Morris, R. H. J. Am. Chem. Soc.
2009, 131, 1394–1395.
(37) Meyer, N.; Lough, A. J.; Morris, R. H. Chem.—Eur. J. 2009, 15,
5605–5610.
(38) Morris, R. H. Chem. Soc. Rev. 2009, 38, 2282–2291.
(39) Gaillard, S.; Renaud, J.-L. ChemSusChem 2008, 1, 505–509.
(40) Enthaler, S.; Junge, K.; Beller, M. Angew. Chem., Int. Ed. 2008, 47,
3317–3321.
(41) Shaikh, N. S.; Enthaler, S.; Junge, K.; Beller, M. Angew. Chem., Int.
Ed. 2008, 47, 2497–2501.
Synthesis of the Complexes. Imine formation is usually
accomplished in organic synthesis via the condensation of
aldehydes with amines. Stabilizing groups such as electron-
withdrawing substituents on nitrogen, or unsaturated
groups in conjugation with the imine group that is formed,
favor the reaction.45 These factors make the synthesis of
the P-N-N-P ligand of complex 1 easier than that of 2.
On the other hand, we have reported that iron(II) serves
as a template in the condensation of diamines with the
unstable phosphine aldehyde PPh2CH2CHO released by
the action of the strong base from the phosphonium salt 3
(Scheme 2).36,46 This leads to the efficient formation of
the desired complex containing a P-N-N-P ligand and
two trans acetonitrile ligands. One problem that detracts
from the convenience of the method is the separation of the
complex trans-[Fe(NCMe)2(PNNP)]2þ from stoichiometric
(42) Naik, A.; Maji, T.; Reiser, O. Chem. Commun. 2010, 46, 4475–4477.
(43) Kandepi, V.; Cardoso, J. M. S.; Peris, E.; Royo, B. Organometallics
2010, 29, 2777–2782.
(44) Bolm, C.; Legros, J.; Le Paih, J.; Zani, L. Chem. Rev. 2004, 104, 6217–
6254.
(45) Patai, S. The Chemistry of the Carbon-Nitrogen Double Bond; Wiley:
London, 1970.
(46) Mikhailine, A. A.; Kim, E.; Dingels, C.; Lough, A. J.; Morris, R. H.
Inorg. Chem. 2008, 47, 6587–6589.

Article
Inorganic Chemistry, Vol. 49, No. 23, 2010 11041
Scheme 3. One Pot Synthesis of Iron(II) Precatalysts 5a-5d Starting
from Dimer 3
Table 1. Yields of Complexes 5a-5d Using the Method Described in Scheme 3
and Selected IR and NMR Data
complex
yield (%)
48
IR CO stretch (cm-1
1974
)
31P NMR ppm
5a
65.2(d), 66.2(d)
JPP = 38.2 Hz
64.2(d), 65.6(d)
JPP = 39.5 Hz
64.6(d), 66.0(d)
JPP = 39.1 Hz
67.6
5b
5c
53
55
82
1975
1974
5d
2a
1981
2001
69.3(d), 65.7(d)
JPP = 30 Hz
a Reference 36.
amounts of salts that are formed in the reaction. We
reported that for the dpen-derived complex this can be
overcome by the selective precipitation of the bis(aceto-
nitrile) complex as the tetraphenylborate salt.36 Then the
carbonyl-containing precatalyst was simply prepared by
stirring a solution of the bis(acetonitrile) complex 4 in
acetone under an atmosphere of carbon monoxide.
An attempt to synthesize similar iron(II) complexes
with other diamines using the template procedure led to
the formation of a byproduct in the last step of the
synthesis as was determined by 31P NMR. The electro-
spray ionization mass spectrometry (ESI-MS) spectrum
of the mixture had a significant peak at 697.0828 m/z,
showing that one of the products contained bromide
coordinated to the iron along with carbonyl and P-N-
N-P ligands. Our group has recently reported the synthesis
and characterization by single crystal X-ray diffraction of
similar iron(II) complexes with bromide trans to carbonyl.47
This observation directed us to a new optimized one pot
synthesis of active iron(II) complexes 5 (Scheme 3).
The template synthesis starting from the dimer 3, di-
amine, and iron(II) precursor was performed in an identical
fashion to the previously reported template procedure,
but the isolation of the bis(acetonitrile) complex 4 was
omitted. The solvent mixture of acetonitrile and methanol
was evaporated and replaced with acetone, and the resulting
solution was subjected to a CO atmosphere. The Br-/CO
complexes 5 are more thermodynamically stable compared
with the CH3CN/CO complexes 2, since the formation of 5
was observed when Br- was present in small quantity in the
reaction mixture containing a large excess of acetonitrile.
The greater stability of complex 5 compared with that of
complex 2 can be explained by the strong, ionic Fe2þ-Br-
bond relative to the weaker ion-dipole Fe2þ-Nδ- bond of
an acetonitrile ligand. The selective precipitation of the
resulting complex with NaBPh4 from methanol yielded
the desired compounds 5 contaminated with salts. A con-
venient purification method is the filtration of a dichlor-
omethane solution of the crude product through Celite. This
general procedure was used to successfully synthesize and
purify complexes 5a-5d in moderate to good yields using a
range of diamines as shown in Scheme 3.
other common solvents. The complexes were character-
ized by 1H, 31P, 13C NMR, HRMS, and FTIR. Selected
properties are listed in Table 1. Complexes 5a-5c that
contain a chiral diamine incorporated into the back-
bone have inequivalent phosphorus atoms because of
the lack of symmetry. This was signaled in the 31P{1H}
NMR spectra by the appearance of two doublets as
opposed to the singlet observed for the C2- symmetrical
bis(acetonitrile) complex 4. The carbonyl 13C NMR reso-
nance was only detected for complexes 5b (at 214.9 ppm)
and 5c (215.1 ppm). Complex 5c gave the expected two
doublets due to coupling to two inequivalent phosphorus
atoms. The long relaxation time of the carbonyl-carbon
of 5a and 5d presumably prevented the detection of their
carbonyl resonances. On the other hand, the presence of
this functionality was supported by the strong peak of the
carbonyl stretch in the IR spectrum and the detection of
the parent ion peak by mass spectroscopy. The carbonyl
stretching wavenumber of the compounds was observed
in a range from 1974 to 1981 cm-1. The monocationic
bromide-carbonyl complex (5b) has a lower carbonyl
stretching frequency compared to the dicationic acetonitrile-
carbonyl complex 2 at 2001 cm-1, consistent with an
increase in iron-carbon backdonation on going from the
dication to the monocation. Attempts were unsuccessful
in obtaining crystals of complexes suitable for X-ray
diffraction analysis.
Catalytic Activity. The catalytic hydrogenation of aceto-
phenone to 1-phenethanol was conducted in an argon
glovebox using precatalysts 5a-5d. The reaction was
performed at 28-30 °C using acetophenone, catalyst,
and base (KOtBu) in a ratio 6000/1/8, respectively. After
precatalysts 5a-5d are activated by the base they become
extremely air sensitive and must be handled under N2 or
Ar. Samples were taken by injection of small portions of
the reaction mixture into a sealed vial containing air and
aerated 2-propanol to ensure successful quenching of the
oxygen-sensitive catalyst system. They were then ana-
lyzed using gas chromatography. The reaction using each
precatalyst was performed three times. The rate of the
reaction was obtained by plotting the concentration of
1-phenethanol versus time and determining the slope of
the linear part of the graph. The turnover frequency
(TOF) values were calculated using least-squared fit
equations at 15-50% conversion and are summarized
in Table 2.
Properties of the Complexes. Complexes 5a-5d were
isolated as yellow, slightly air-sensitive solids. They dem-
onstrate similar solubility properties: soluble in dichloro-
methane, acetone, and acetophenone but insoluble in
The catalytic activity of the precatalyst 2 was redeter-
mined under the conditions used for 5b to identify the
effect of the ligand in the trans position relative to the
(47) Lagaditis, P. O.; Lough, A. J.; Morris, R. H. Inorg. Chem. 2010,
49, 10057-10066.

11042 Inorganic Chemistry, Vol. 49, No. 23, 2010
Mikhailine and Morris
Table 2. Transfer Hydrogenation of Acetophenone Catalyzed by Complexes
5a-5d in Basic 2-Propanol
complexes
TOF (h-1 a
)
ee (% S)
5a
5b
5c
5db
2
4.9 ꢀ 103
2.0 ꢀ 104
2.0 ꢀ 104
2.1 ꢀ 103
2.1 ꢀ 104
60
81
82
82
a TOF at 15-50% conversion. b 15% conversion.
carbonyl (Table 2). The activation and the maximum
constant rate of the transfer hydrogenation using 5b
and 2 were found to be the same within experimental
error. A lower TOF of 3.6 ꢀ 103 h-1 was reported earlier
for 2 (Scheme 1), but this refers to a lower temperature
and a lower substrate loading than for the present condi-
tions. This observation also shows that the dissociation of
the ligand trans to carbonyl most likely is taking place
during the activation of the precatalyst and that bromide
and acetonitrile do not participate in the process of the
reduction of acetophenone to 1-phenethanol.
The incorporation of different diamines into the back-
bone of the ligand of complexes 5a-5e influences both the
electronic and the steric properties of the complexes. The
basicity of the nitrogen donors decrease in order dach =
MeO-dpen = dpen > en as interpreted from the wave-
numbers of the carbonyl absorptions listed in Table 1—
the more basic the donor, the lower the CO stretching
wavenumber. The TOF values of the corresponding
complexes do not correlate well with the basicity of the
donors. The sterically bulky diamines might be expected
to increase the enantioselectivity but decrease the activity
of ruthenium complexes. This was the observation when
the sterically more crowded precatalyst RuCl2(daipen)-
(xylBinap) was used in place of RuCl2(dpen)(tolBinap) in
the catalytic hydrogenation of acetophenone.48 However
there is also a report where the substitution of the diamine
backbone with phenyl groups in the 1- and 2-positions
results in an increase in ee and an increase in rate
compared to diamines with only monosubstitution at
the 1- or 2-positions.49 No explanation was provided for
the rate enhancement. But the activity trend within this
group of iron catalysts is different with the more bulky
phenyl and substituted phenyl ligands of 5b and 5c
producing more active precatalysts than the less bulky
diaminocyclohexane- and ethylenediamine-derived ligands
of 5aand 5d. This showsthatthe steric effect of the diamine
has a larger influence on the activity of the catalyst than
electronic factors.
Figure 1. Reaction profile of the catalytic reduction of acetophenone
to 1-phenethanol catalyzed by the precatalyst 5b under Ar at 28-30 °C.
[(a) cat/sub/base =1/6000/8, (b) 4800 equiv of acetophenone was added
to the reaction mixture after 23 min of reaction, (c) reaction conducted
with cat/sub/acetone/base = 1/6000/6000/8.]
the profile for complex 5a is depicted in Figure 1a. The
reaction starts with the activation of the precatalyst to the
so far unknown active catalyst, a process with continu-
ously increasing rate. The activation period varies for
different catalysts, but in general takes place during the
first 5 min of the reaction. The following turnover step
produces 1-phenethanol at a constant rate before the
reaction reaches 60% conversion in case of complexes
5a-5c and 15% conversion in case of the complex 5d. A
reduction of the rate is observed later in the reaction to
give, at the end, more than 80% of the product in case of
the complexes 5a-5c and 40% in case of the complex 5d.
We wondered whether the observed rate reduction is
associated with the decomposition of the catalyst after a
certain number of turnovers in the catalytic cycle. To
check that assumption, extra acetophenone was added
after 23 min of the reaction (Figure 1b). Since the rate
increases to the value comparable to that observed during
the turnover step, we conclude that the catalyst decom-
position cannot be the reason for the rate reduction.
As the reaction progresses toward equilibrium, the
concentration of the acetophenone decreases as the con-
centration of the acetone and 1-phenethanol increases.
Therefore, the rate of the 1-phenethanol formation can
decrease because of its dehydrogenation back to the
acetophenone. Another explanation for the rate decrease
is that the reduction of the acetone back to the 2-propanol
is taking place in a non-productive cycle which competes
with the hydrogenation of the acetophenone (Scheme 4).
This process is reasonable, especially if the acetone con-
centration becomes higher than the concentration of the
To gain a deeper understanding of this process, the
conversion of the substrate to the product was monitored
by taking samples of the reaction mixture at short time
intervals. Catalysts 5a-5d showed similar reaction pro-
files in the reduction of acetophenone to 1-phenethanol;
(48) Noyori, R.; Ohkuma, T. Angew. Chem., Int. Ed. 2001, 40, 40–73.
(49) Hayes, A.; Clarkson, G.; Wills, M. Tetrahedron: Asymmetry 2004,
15, 2079–2084.

Article
Inorganic Chemistry, Vol. 49, No. 23, 2010 11043
Scheme 4. Catalytic Cycle of the Acetophenone Reduction Where
for activity. Further studies of the mechanism of these
interesting catalytic systems are underway.
[Fe] and [Fe]HþH- Are Unknown Active Iron Complexes
Experimental Section
General Comments. All manipulations that involved air- or
moisture-sensitive materials were performed using Schlenk
techniques or a glovebox under an argon or nitrogen atmo-
sphere. Solvents were degassed and dried using standard pro-
cedures prior to all manipulations and reactions. Deuterated
solvents were distilled and dried over activated molecular sieves.
The phosphonium dimer 3 was synthesized according to the
procedure previously reported by our group.46 Other reagents
used were purchased from commercial sources and utilized
without further purifications. NMR spectra of the samples that
were prepared under argon in degassed solvents were recorded
at ambient temperature and pressure using a 400 MHz Varian
Gemini [1H (400 MHz), 13C{1H} (100 MHz), and 31P{1H} (161
MHz)]. The infrared spectra of the KBr pellets containing
precatalysts 5a-5d was measured on Paragon 500 (Spectral
Range 4600 cm-1 to 400 cm-1) using Perkin-Elmer’s SPECTRUM
for Windows system for data collection and processing at 25 °C.
The ESI-MS data on samples in methanol/water were done on
an AB/Sciex QStar mass spectrometer with an ESI source.
The elemental analyses were performed at the Department of
Chemistry, University of Toronto, on a Perkin-Elmer 2400
CHN elemental analyzer. Some complexes gave a low carbon
analysis but acceptable hydrogen and nitrogen contents because
of a combustion problem that is common for the compounds
containing boron atoms.50
acetophenone. [Fe]HþH- in Scheme 4 represents a so-far
unknown intermediate in catalytic cycle that transfers an
H2 equivalent from 2-propanol to the acetophenone.
A mixture (50-50% by mol) of acetophenone and
acetone was subjected to the catalytic conditions to see how
acetone influences the rate of the reaction (Figure 1c).
There was almost a 3-fold decrease in rate for the reaction
with extra acetone compared to the rate of the initial
reaction. This unambiguously shows that acetone is
competing with acetophenone for the active site on the
catalyst.
Conclusions
In conclusion, we were able to synthesize and characterize
iron(II) complexes (5a-5d) containing P-N-N-P ligands
with different diamines incorporated into the backbone to
compare their activity and enantioselectivity in asymmetric
catalytic transfer hydrogenation of acetophenone to 1-
phenethanol using 2-propanol as a reducing agent. A highly
efficient one pot procedure allowed us to prepare those
compounds in high purity and in moderate to good yields
starting from readily available starting materials. Similar
activities, previously reported by our group, of precatalyst
2 and 5b indicate that ligand trans to CO has no effect on the
reaction rate. All of the complexes show very high TOF (up to
2.0 ꢀ 104 h-1) and enantioselectivity (up to 82% ee), using
very low catalyst loadings (0.016% relative to the substrate).
Monitoring the formation of product in a course of the
reaction illustrated that the catalytic process consists of
three stages: activation of the precatalyst to an active
catalyst (increasing rate of product formation), turnover step
(formation of the product at a constant rate), and termina-
tion step (decreasing rate of product formation). A series of
experiments has proven that the decrease in the rate when
approaching the end of the reaction may be associated with
the competing reduction by isopropanol of acetone versus
acetophenone. The comparison of the TOF of complexes
5a-5d showed that the catalytic activity of the complexes
increases as the steric influence of the groups in the backbone
of ligand increases.
(R,R)-[Fe(Ph2PCH2CHN(C6H10)NCHCH2PPh2)(Br)(CO)]-
[BPh4] (5a). In an Ar glovebox dimer 3 (5.41 g, 8.75 mmol),
[Fe(H2O)6][BF4]2 (4.43 g, 13.1 mmol) and NaOMe (0.946 g, 17.5
mmol) were dissolved in 20 mL of MeOH to give a slightly
yellow transparent solution after 15 min of stirring. A solution
of (1R,2R)-diaminocyclohexane (1.00 g, 8.75 mmol) in aceto-
nitrile (5 mL) was added to the reaction mixture to give a bright
purple solution instantaneously. After 48 h the solution became
orange in color with an orange-pink precipitate. The solvent was
removed under vacuum, and the resulting orange-pink solid was
washed with diethyl ether (10 mL) and redissolved in 15 mL of
acetone. The resulting pink-orange solution with a white precipi-
tate was stirred under 1 atm of carbon monoxide at 30 °C for 48 h
to give a bright yellow solution with a white precipitate that was
filtered. The solvent of the eluate was removed under vacuum, and
the resulting solid was redissolved in 40 mL of MeOH. The solution
of NaBPh4 (2.99 g, 8.75 mmol) in 5 mL of MeOH was slowly added
to a yellow solution to give yellow precipitate that was filtered and
washed with diethyl ether (10 mL) three times. The resulting solid
was dissolved in 5 mL of dichloromethane to give a brown-yellow
solution and white precipitate that was filtered through the glass
frit and then through a bed of Celite. The solvent of eluate was
removed under vacuum to give a bright-yellow precipitate that was
mixed with diethyl ether, and the mixture stirred for 20 h. The
precipitate was filtered and washed with 10 mL of diethyl ether
three times to give microcrystals containing diethylether (NMR
evidence). Yield: 48% (4.26 mg); 1H NMR (400 MHz, CD2Cl2) δ:
1.2-1.4 (m, 1H, NCHCH2; 2H, NCHCH2CH2), 1.4-1.6 (m, 1H,
NCHCH2), 1.7-2.8 (m, 1H, NCHCH2; 2H, NCHCH2CH2),
2.2-2.5 (m, 1H, NCHCH2), 3.3-3.5 (m, 1H, NCH; 1H, PCH2),
3.6-3.8 (m, 1H, PCH2), 3.8-4.0 (m, 1H, NCH; 1H, PCH2),
4.1-4.2 (m, 1H, PCH2), 6.7-7.8 (m, 40H, ArH), 7.9-7.7 (m,
1H, NdCH), 7.9-8.1 (m, 1H, NdCH); 13C{H} NMR (100 MHz;
CD2Cl2) δ: 23.7 (s, CH2CH2), 23.7 (s, CH2CH2), 29.4 (s, CHCH2),
30.0 (s, CHCH2), 47.2-47.9 (m, PCH2), 70.1 (s, NCH), 74.0 (s,
NCH), 122 (s, BPhCH), 126 (m, BPhCH), 127-134 (m, ArCH),
In an accompanying paper we showed the effect on the
activity of such iron catalysts for transfer hydrogenation
when the substituents on phosphorus were varied from Ph to
i
Et to Pr to Cy.47 The last two changes produced inactive
complexes and therefore indicated that there cannot be too
much steric bulk on the phosphorus. The activity of the
complexes and the enantioselectivity dropped on going from
Ph to Et groups on phosphorus. Thus less donating groups
that are not too bulky are required on the phosphorus atoms
(50) Marco, A.; Compano, R.; Rubio, R.; Casals, I. Microchim. Acta
2003, 142, 13–19.

11044 Inorganic Chemistry, Vol. 49, No. 23, 2010
Mikhailine and Morris
131 (s, BPhCH), 136 (s, BPhCH), 164 (m, JCB = 49.3 Hz,
BPhCH), 169-170 (m, N=CH); 31P{H} NMR (161 MHz;
CD2Cl2): 65.2 (d, JPP = 38.2 Hz), 66.2 (d, JPP = 38.2 Hz); HRMS
(ESI-TOF) m/z calculated for [C35H36N2P2FeOBr]þ: 697.0830,
found: 697.0828 m/z. Anal. Calcd for C59H56N2P2FeBrOB
C4H10O: C, 69.31; H, 6.09; N, 2.56. Found: C, 70.79; H, 6.06;
N, 2.93.
[Fe(Ph2PCH2CHNCH2CH2NCHCH2PPh2)(Br)(CO)][BPh4]
(5d). Complex 5d was synthesized similarly to the complex 5a
using the following amounts of starting materials: dimer 3 (1.20 g,
1.94 mmol), [Fe(H2O)6][BF4]2 (0.98 g, 2.91 mmol), NaOMe (0.21
g, 3.88 mmol), ethylenediamine (0.12 g, 1.94 mmol) and NaBPh4
1
(0.66 g, 1.94 mmol). Yield: 82% (1.53 g); H NMR (400 MHz,
CD2Cl2) δ: 3.47-3.59 (m, 2H, NCH2), 3.58-3.69(m, 2H, PCH2),
3.75-3.86 (m, 2H, PCH2), 3.85-3.97 (m, 2H, NCH2), 6.75-7.68
(m, 40H, ArH; 2H, NCH); 13C{H} NMR (100 MHz; CD2Cl2) δ:
48.3-48.8 (m, PCH2), 61.1 (s, NCH2), 121.6 (s, BPhCH), 125.4 (m,
BPhCH), 127.5 (PPhCH), 128.3 (PPhCH), 130.7 (PPhCH), 131.0
(PPhCH), 131.4 (PPhCH), 133.3 (PPhCH), 135.3 (s, BPhCH),
162.9 (m, JCB = 48.5 Hz, BPhCH), 172.5 (s, NCH); 31P {H} NMR
(161 MHz; CD2Cl2): 67.58 ppm (s); HRMS (ESI-TOF) m/z
calculated for [C31H30N2P2FeOBr]þ: 643.0360 m/z. Found:
643.0356 m/z. Anal. Calcd for C55H50N2P2FeBrOB: C, 68.56; H,
5.23; N, 2.91. Found: C, 66.70; H, 5.55; N, 3.27.
(S,S)-[Fe(Ph2PCH2CHNC(Ph)HC(Ph)HNCHCH2PPh2)(Br)-
(CO)][BPh4] (5b). Complex 5b was synthesized similarly to the
complex 5a using the following amounts of starting materials:
dimer 3 (5.24 g, 8.47 mmol), [Fe(H2O)6][BF4]2 (4.29 g, 12.7 mmol),
NaOMe (0.916 g, 16.9 mmol), (1S,2S)-1,2-diphenyldiaminoethane
(1.80 g, 8.47 mmol), and NaBPh4 (2.90 g, 8.47 mmol). Yield: 53%
1
(5.01 g); H NMR (400 MHz, CD2Cl2) δ: 3.41-3.55 (m, 1H,
PCH2), 3.59-3.74 (m, 1H, PCH2), 3.76-3.91 (m, 1H, PCH2),
3.99-4.13 (m, 1H, PCH2), 5.15-5.25 (m, 1H, NC(Ph)H),
5.62-5.72 (m, 1H, NC(Ph)H), 6.70-7.57 (m, 50H, ArH),
7.52-7.66 (m, 1H, NCH), 7.73-7.87 (m, 1H, NCH); 13C{H}
NMR (100 MHz; CD2Cl2) δ: 47.05-47.71 (m, PCH2), 78.07 (s,
NC(Ph)H), 81.42 (s, NC(Ph)H), 122.03 (s, BPh), 125.96 (m, BPh),
127.9-129.7 (m, ArCH), 130.17 (s, BPh), 130.3-136.52 (m, ArCH),
136.2 (s, BPh), 164.62 (m, JCB = 49.3 Hz, BPh), 174.2-174.5 (m,
NCH), 214.89 (s, weak, CO); 31P{H} NMR (161 MHz; CD2Cl2):
64.24 (d, JPP = 39.5 Hz), 65.62 (d, JPP = 39.5 Hz); HRMS (ESI-
TOF) m/z calculated for [C43H38N2P2FeOBr]þ: 795.0986, found:
795.0983. Anal. Calcd for C67H58N2P2FeBrOB: C, 72.13; H, 5.24;
N, 2.51. Found: C, 69.21; H, 5.30; N, 2.75.
General Procedure for the Catalytic Reduction of the Aceto-
phenone Using Catalysts 5a-5d and 2-Propanol As a Reducing
Agent. In an Ar glovebox, two stock solutions were prepared.
The stock solution 1 (1SS) contains 4.48 ꢀ 10-6 moles of the
catalyst (5a-5d) dissolved in 4.64 ꢀ 10-3 mole of the acetophe-
none. The stock solution 2 (2SS) contains 8.91 ꢀ 10-5 mole of
the KOtBu dissolved in 1.70 ꢀ 10-2 moles of 2-propanol. For
every catalytic run, 0.110 g of 1SS was diluted with acetophe-
none (0.527 g) and 2-propanol (8.0 g). The 0.081 g 2SS was
diluted with 2-propanol (0.50 g). Resulting solutions were mixed
to initiate the reaction. Final concentrations of reagents in
the reaction mixture were as follows: [cat] = 7.7 (mol/L),
[substrate]=0.64 (mol/L). Samples were taken by injection of
small portions of the reaction mixture into a sealed vial contain-
ing air and aerated 2-propanol to ensure successful quenching
of the oxygen-sensitive catalyst system. They were then ana-
lyzed using gas chromatography using a Perkin-Elmer Auto-
system XL chromatograph with a chiral column (CP chirasil-
Dex CB 25 m ꢀ 2.5 mm). Hydrogen was used as a mobile phase
at a column pressure of 5 psi. The injector temperature was
250 °C, a FID temperature was 275 °C, and an oven tempera-
ture was 130 °C. GC retention times were 5.02 min, 8.73, and
9.42 for ketone, alcohol R-isomer, and alcohol S-isomer,
respectively.
(R,R)-[Fe(Ph2PCH2CHNC(p-MeO-Ph)HC(p-MeO-Ph)HN-
CHCH2PPh2)(Br)(CO)][BPh4] (5c). Complex 5c was synthesized
similarly to the complex 5a using the following amounts of starting
materials: dimer 3 (2.25 g, 3.64 mmol), [Fe(H2O)6][BF4]2 (1.83 g,
5.45 mmol), NaOMe (0.393 g, 7.27 mmol) (1R,2R)-1,2-di(40-
methoxyphenyl)-1,2-diaminoethane (0.990 g, 3.64 mmol) and
NaBPh4 (1.24 g, 3.64 mmol). Yield: 55% (2.35 g); 1H NMR (400
MHz, CD2Cl2) δ: 3.40-3.57 (m, 1H, PCH2), 3.60-3.90 (m, 2H,
PCH2), 3.74 (s, 3H, OCH3), 3.77 (s, 3H, OCH3), 3.94-4.11 (m, 1H,
PCH2), 5.00-5.23 (m, 1H, NC(Ph)H), 5.50-5.73 (m, 1H, NC-
(Ph)H), 6.50-7.70 (m, 48H, ArH), 7.73-7.97 (m, 1H, NCH),
7.48-7.72 (m, 1H, NCH); 13C{H} NMR (100 MHz; CD2Cl2) δ:
47.14-47.84 (m, PCH2), 55.8 (s, OCH3), 55.9 (s, OCH3), 77.36 (s,
NC(Ph)H), 81.00 (s, NC(Ph)H), 122.1 (s, BPh), 124.9-136.2 (m,
ArCH), 125.4 (m, BPh), 136.16(s, BPh), 164.62 (m, JCB = 49.3 Hz,
BPh), 173.7-174.5 (m, NCH), 215.1 (dd, JCP = 30.2 Hz, CO);
31P{H} NMR (161 MHz; CD2Cl2): 64.6 (d, JPP = 39.1 Hz),
66.0(d, JPP = 39.1 Hz); HRMS (ESI-TOF) m/z calculated for
[C45H42N2P2FeO3Br]þ: 855.1197, found: 855.1196 m/z. Anal.
Calcd for C69H62N2P2FeBrO3B: C, 70.49; H, 5.32; N, 2.38. Found:
C, 68.49; H, 5.48; N, 2.34.
Acknowledgment. R.H.M. thanks NSERC for a Discovery
grant.
Supporting Information Available: Further details are pro-
vided in Tables 1S-4S and Figures 1S-3S. This material is
available free of charge via the Internet at http://pubs.acs.org.