Low-valent ene-amido iron complexes for the asymmetric transfer hydrogenation of acetophenone without base

Article abstract of DOI:10.1021/ja202375y

Examination of the role of base in the activation of our previously reported iron(II) complexes having the general formula [Fe(CO)(Br)(PNNP)] [BPh₄] revealed a five-coordinate iron(II) complex in which the tetradentate PNNP ligand had been doubly deprotonated. The new iron(II) complexes were used in the transfer hydrogenation of acetophenone in isopropanol in the absence of added base, and certain analogues showed catalytic activity.

Full text of DOI:10.1021/ja202375y

COMMUNICATION
pubs.acs.org/JACS
Low-Valent EneꢀAmido Iron Complexes for the Asymmetric Transfer
Hydrogenation of Acetophenone without Base
Paraskevi O. Lagaditis, Alan J. Lough, and Robert H. Morris*
Department of Chemistry, University of Toronto, Toronto, Ontario, Canada M5S 3H6
S Supporting Information
b
Scheme 1. Synthesis of Doubly Deprotonated Iron(II)
Complexes 2
ABSTRACT: Examination of the role of base in the
activation of our previously reported iron(II) complexes
having the general formula [Fe(CO)(Br)(PNNP)][BPh4]
revealed a five-coordinate iron(II) complex in which the
tetradentate PNNP ligand had been doubly deprotonated.
The new iron(II) complexes were used in the transfer
hydrogenation of acetophenone in isopropanol in the
absence of added base, and certain analogues showed
catalytic activity.
latter could be conveniently filtered to remove any residual
excess base. The green compounds 2 were obtained as sticky
residues despite extensive drying. They were air-sensitive and
turned brown when exposed to air. Benzene solutions of
compounds 2aꢀd were stable for months in a glovebox under
an inert atmosphere, while 2e and (S,S)-2f were less stable, with a
noticeable loss of color from green to brown after a couple
of weeks.
he replacement of precious platinum metal-based complexes
T
in catalysis with first-row transition metals, especially iron, is
a growing phenomenon.1 The impetus for this lies in the fact
that these metals are more abundant and benign, thereby
creating cost-effective and “green” catalysts. Our recent research
has been focused on the development of iron(II) catalysts,
and we have demonstrated complexes of the type trans-[Fe-
(CO)(Br)(PNNP)]þ (PNNP = PR2CH2CHdNCH(R)ꢀCH-
(R)NdCHCH2PR2) form active and selective catalysts for the
asymmetric transfer hydrogenation (ATH) of ketones in basic
isopropanol.2 However, the mechanism of this transformation is
still elusive. As a starting point, we examined the role of base in
the transfer hydrogenation reaction because these iron(II) com-
plexes become active only upon the addition of a strong base (the
use of 8 equiv of KOtBu was found to be optimal). An obser-
vation indicating that base plays an important role in the
activation of the catalyst was the immediate color change of
the reaction mixture from yellow to green when base was added.
Here we report that deprotonation of the tetradentate ligand
backbone, specifically at the carbon R to the phosphorus, is an
important part of the catalyst activation process. This allows the
first iron-based catalytic ATH of acetophenone without the need
for addition of base.3
Compounds 2aꢀf were analyzed by 1H,31P{1H}, and 11B NMR
spectroscopy in C6D6. The 31P NMR spectra in all cases revealed
a downfield shift of up to 10 ppm for the phosphorus signal
relative to that for the starting iron(II) complex.2b The absences
of BPh4ꢀ aromatic protons and boron resonances in the 1H and
11B NMR spectra, respectively, are consistent with 2aꢀf being
neutral compounds. The 1H NMR data (integrations, chemical
1
shifts, and 1Hꢀ H COSY results) revealed that while the carbons
adjacent (i.e., R) to the phosphorus of the PNNP ligand back-
bone in 1aꢀf had two inequivalent hydrogens, those in com-
pounds 2aꢀf had only one hydrogen (Figure 1). Hence, the base
deprotonated the ligand at both R carbons. The diastereotopic
nature of the ꢀCH2ꢀ groups in the ethylene backbones of 2a,
2c, and 2e was retained, with peaks appearing in a 1:1 ratio
(Figure 1 bottom). In addition, the ꢀCH(Ph) hydrogens were
still present in the chiral compounds (S,S)-2b, (S,S)-2d, and
(S,S)-2f.
We observed that the green color occurred even without the
presence of the ketone substrate; thus, the iron(II) precatalysts,
1a-f, (10 mg) were mixed with 8 equiv of KOtBu to mimic the
conditions for catalysis (Scheme 1). However, when this was
done in isopropanol, the loss of the green color was evident
within minutes or overnight, which implied that the green species
was reactive toward that solvent. Hence, benzene was chosen
instead. Both the base and the iron(II) precatalysts had limited
solubility in benzene but reacted fully after 30 min to give a dark-
green solution with disappearance of the yellow starting material.
All of the solutions were filtered to remove an off-white pre-
cipitate, and evaporation in vacuo gave a green residue that was
soluble in numerous solvents, such as alkanes. Solutions of the
Regardless whether 2 or 8 equiv of base was used in the
synthesis of 2aꢀe, the same ligand double deprotonation pattern
was observed in the 1H NMR spectra. However, the use of excess
base in the synthesis of (S,S)-2f produced some side products
according to the NMR spectra. The cleanest spectra obtained
were from samples in which 2 equiv of base were used. Here a
second, unidentified product was present in a ratio of 1:3 with
(S,S)-2f. The 31P{1H} NMR spectra also showed the presence of
two compounds by means of a doublet of doublets at 77.1 and
Received: March 16, 2011
Published: May 31, 2011
r
2011 American Chemical Society
9662
dx.doi.org/10.1021/ja202375y J. Am. Chem. Soc. 2011, 133, 9662–9665
|

Journal of the American Chemical Society
COMMUNICATION
Scheme 3. Synthesis of Doubly Deprotonated Iron(II)
Complexes 5
Figure 1. 1H NMR spectra (400 MHz) of hydrogens of interest in the
iron(II) complexes 1a in CD2Cl2 (top) and 2a in C6D6 (bottom).
Scheme 2. Synthesis of Deuterated Iron(II) Complex 3
1
68.3 ppm and a broad signal at 67.7 ppm. The 31Pꢀ H HMBC
NMR spectrum showed that the doublet of doublets belonged to
the minor product while the broad signal belonged to (S,S)-2f. A
31P variable-temperature NMR (VT-NMR) experiment to as low
as ꢀ80 °C resulted in the disappearance of the broad peak at
67.7 ppm to the baseline without the appearance of a new set of
signals. Although the 31P NMR spectrum of 2e displayed a
singlet, the resonance was rather broad in comparison with the
sharp singlet resonances of 2a and 2c. Hence, we attribute the
broad resonances in the 31P NMR spectra of 2e and (S,S)-2f to
fluxional behavior of the Et groups on phosphorus.
To further support deprotonation at each carbon R to the
phosphorus on the PNNP ligand backbone, DCl (1 M in Et2O)
was added to a filtered green solution of 2e (Scheme 2). This
gave an immediate change in the color of the reaction solution
from green to yellow. The 1H NMR spectrum showed that the
carbon R to the phosphorus on the PNNP ligand was deuterated;
a crystal of 3 suitable for X-ray diffraction was obtained (see the
Supporting Information). This experiment implies a five-coordi-
nate iron species upon deprotonation and also demonstrates the
ease of deprotonation and protonation of the PNNP ligand.
To further support the inference that the carbon R to the
phosphorus on the ligand backbone was deprotonated, iron(II)
complexes 4aꢀc with o-phenylene groups as NN linkers of the
PNNP ligand (Scheme 3) were synthesized. The crystal structure
of 4b is shown in Figure 2. No protons exist on the NN linker
backbone of the ligand, ruling out any possibility of deprotona-
tion of that part of the ligand. The reaction of 4aꢀc with KOtBu
followed by removal of KBr and KBPh4 as well as excess base
(Scheme 3) and subsequent NMR characterization of the isolated
product again showed evidence of the double deprotonation at the
same R carbon on the PNNP ligand. The doubly deprotonated
compounds 5aꢀc in these cases were not green but rather black
unless diluted significantly to give purple (5a, 5b) or blue (5c)
solutions. The color is due to multiple absorbances at wavelengths
Figure 2. ORTEP plot (thermal ellipsoids at 50% probability) of the
X-ray crystal structure of 4b. The BPh4 anion and most of the H atoms
have been removed for clarity.
across the visible spectrum, as determined from UVꢀvis spectro-
scopic measurements (see the Supporting Information).
The proposed structure of the neutral, five-coordinate iron
complexes was confirmed by a single-crystal X-ray diffraction
study of 5b (Figure 3). The structure has a pseudo-square-
pyramidal geometry. There is no anion present. The CO ligand is
apical; there is no spectroscopic or crystallographic evidence of a
hydride trans to CO. On going from 4b to 5b, the PCHxꢀCHN
distance contracts from 1.481(4) to 1.342(4) Å. All of the other
distances in the five-membered FeꢀPꢀCꢀCꢀN rings also
contract upon deprotonation except for the distances corre-
sponding to N1ꢀC2, which lengthen from 1.277(5) to 1.385(3) Å.
All of these changes suggest that there is extensive delocaliza-
tion of bonding in 5b. Square-pyramidal, five-coordinate iron
compounds are not prevalent.4 Of all of the examples, two were
found that contain a tetradentate ligand and an apical CO ligand
analogous to the compounds presented in this report. The νCtO
of the reported compounds [Fe(CO)(C21H22N4)]5a and
[Fe(CO)(C36H44N4)]5b containing macrocyclic ligands were
found at 1921 or 1948 cmꢀ1, respectively. These are higher than
those for compounds 2aꢀf and 5aꢀc, which were found to be in
the 1880ꢀ1898 cmꢀ1 range. The νCtO of the parent com-
pounds 1aꢀf and 4a-c range between 1945 and 1958 cmꢀ1 2b
.
Hence, the lower wavenumbers of 2aꢀf and 5aꢀc are believed to
be due to not only extensive delocalization of the negative charge
across the ligand but delocalization into the metal as well.
9663
dx.doi.org/10.1021/ja202375y |J. Am. Chem. Soc. 2011, 133, 9662–9665

Journal of the American Chemical Society
COMMUNICATION
inactive while complexes 2e and (S,S)-2f were active for transfer
hydrogenation. The fact that 2aꢀd were not active was not
surprising because we previously ascribed the inactivity of their
parent compounds under transfer hydrogenation to the steric
crowding of the Cy or iPr substituents on phosphorus.2b The
catalytic activities of 2e and (S,S)-2f were similar to those of their
parent compounds 1e and (S,S)-1f, respectively, upon activation
with base (Figure 4). However, in the case of (S,S)-2f, the order
of addition of substrate caused a significant change in the activity.
When isopropanol was added first and then the acetophenone
substrate, the activity profile was identical to that for (S,S)-1f and
base (Figure 4). However, when acetophenone was added first
and then isopropanol, the catalysis was slower (see the Support-
ing Information). The observed difference signifies that for
productive catalyst activation, interaction with isopropanol must
occur first, before interaction with the substrate.
A similar enantiomeric excess of (R)-1-phenylethanol was
obtained after transfer hydrogenation using (S,S)-2f (52%) in
comparison to its parent compound (S,S)-1f (52%) at 50% con-
version.10 This observation also implies that, in situ, the activated
precatalyst and compounds 2e and 2f both lead to similar active
species and that base does not play any further role in the
catalysis.
Figure 3. ORTEP plot (thermal ellipsoids at 50% probability) of the
X-ray crystal structure of 5b. Most of the H atoms have been removed for
clarity.
The base-sensitive substrate 4-acetylbenzoate ethyl ester3a was
tested for transfer hydrogenation using (S,S)-2f. The compound
was reduced, but both the ethyl ester and the isopropanol ester of
the alcohol were produced. Therefore, base must be generated
when (S,S)-2f reacts with isopropanol, and this catalyzes the
transesterification reaction.
In conclusion, we have isolated a series of five-coordinate
iron(II) complexes formed from the deprotonation of the ligand
by base, specifically at the carbon R to the phosphorus on the
PNNP ligand backbone. The complexes that were active for the
transfer hydrogenation of acetophenone in isopropanol had
catalytic behavior similar to that for catalysts generated from
precursor complexes by reaction with excess base. Hence, we
have identified intermediates that are close to, or within, the
catalytic cycle. These compounds may be of use in the activation
of other small molecules and are currently under investigation in
our group.
Figure 4. Catalytic transfer hydrogenation of acetophenone to 1-phe-
nylethanol in isopropanol at 50 °C in the presence of (9) (S,S)-1f,
KOtBu (catalyst/base/substrate = 1/8/500) (see ref 9); (4) (S,S)-2f
(catalyst/substrate = 1/500); (b) 1e, KOtBu (catalyst/base/substrate =
1/8/200) (see ref 9); (O) 2e (catalyst/substrate = 1/200).
’ ASSOCIATED CONTENT
S
Supporting Information. Experimental procedures,
b
spectroscopic data for new compounds, and complete crystal-
lographic data (CIF) for complexes 3, 4a, 4b, and 5b. This
material is available free of charge via the Internet at http://
pubs.acs.org.
Ligands chelating via phosphine and eneꢀamido are rare.
Complexes with bidentate R2PꢀCRdCHꢀNAr ligands are of
interest for their possible use in the Shell higher olefin process.6
Platinum group metal complexes with tridentate pincer ligands7
8
such as iPr2PꢀCHdCHꢀNꢀCH2CH2ꢀPiPr2 and iPr2Pꢀ
’ AUTHOR INFORMATION
9
C6H3MeꢀNꢀC6H3MeꢀPiPr2 are active toward small-mole-
Corresponding Author
rmorris@chem.utoronto.ca
cule activation. Milstein and co-workers have shown that
PNP- and PNN-type ligands [where PNN is 2-(di-tert-butyl-
phosphinomethyl)-6-(diethylaminomethyl)pyridine] are rever-
sibly converted into eneꢀamido forms during catalysis in the
ruthenium-catalyzed direct synthesis of amides from alcohols
and amines and other dehydrogenative processes.10 Compounds
2aꢀf and 5aꢀc in this report are the first examples of tetra-
dentate eneꢀamido phosphine ligands.
’ ACKNOWLEDGMENT
R.H.M. thanks NSERC for a Discovery Grant. P.O.L. thanks
Digital Specialty Chemicals for an OGSST Grant.
’ REFERENCES
(1) (a) Correa, A.; Mancheno, O. G.; Bolm, C. Chem. Soc. Rev. 2008,
37, 1108–1117. (b) Chakraborty, S.; Guan, H. Dalton Trans. 2010,
Complexes 2aꢀf were tested as catalysts for the transfer
hydrogenation of acetophenone to 1-phenylethanol in isopropa-
nol at 50 °C without the addition of base. Complexes 2aꢀd were
9664
dx.doi.org/10.1021/ja202375y |J. Am. Chem. Soc. 2011, 133, 9662–9665

Journal of the American Chemical Society
COMMUNICATION
39, 7427–7436. (c) Ellis, W. C.; Tran, C. T.; Roy, R.; Rusten, M.;
Fischer, A.; Ryabov, A. D.; Blumberg, B.; Collins, T. J. J. Am. Chem. Soc.
2010, 132, 9774–9781. (d) Enthaler, S.; Junge, K.; Beller, M. Angew.
Chem., Int. Ed. 2008, 47, 3317–3321. (e) Casey, C. P.; Guan, H. J. Am.
Chem. Soc. 2007, 129, 5816–5817.(f) Chirik, P. J. In Catalysis without
Precious Metals; Bullock, R. M., Ed.; Wiley-VCH: Weinheim, Germany,
2010; pp 83ꢀ110. (g) Wang, C.; Wu, X. F.; Xiao, J. L. Chem.—Asian
J. 2008, 3, 1750–1770. (h) Langer, R.; Leitus, G.; Ben-David, Y.;
Milstein, D. Angew. Chem., Int. Ed. 2011, 50, 2120–2124.
(2) (a) Morris, R. H. Chem. Soc. Rev. 2009, 38, 2282–2291. (b)
Lagaditis, P. O.; Lough, A. J.; Morris, R. H. Inorg. Chem. 2010,
49, 10057–10066. (c) Mikhailine, A. A.; Morris, R. H. Inorg. Chem.
2010, 49, 11039–11044.
(3) (a) Guo, R.; Chen, X.; Elpelt, C.; Song, D.; Morris, R. H. Org.
Lett. 2005, 7, 1757–1759. (b) Ohkuma, T.; Koizumi, M.; Mu~niz, K.; Hilt,
G.; Kabuto, C.; Noyori, R. J. Am. Chem. Soc. 2002, 124, 6508–6509. (c)
Zweifel, T.; Naubron, J.-V.; B€uttner, T.; Ott, T.; Gr€utzmacher, H. Angew.
Chem., Int. Ed. 2008, 47, 3245–3249. (d) Dong, Z.-R.; Li, Y.-Y.; Chen,
J.-S.; Li, B.-Z.; Xing, Y.; Gao, J.-X. Org. Lett. 2005, 7, 1043–1045. (e)
Abdur-Rashid, K.; Clapham, S. E.; Hadzovic, A.; Harvey, J. N.; Lough,
A. J.; Morris, R. H. J. Am. Chem. Soc. 2002, 124, 15104–15118. (f) Clarke,
Z. E.; Maragh, P. T.; Dasgupta, T. P.; Gusev, D. G.; Lough, A. J.;
Abdur-Rashid, K. Organometallics 2006, 25, 4113–4117. (g) Corberꢀan,
R.; Peris, E. Organometallics 2008, 27, 1954–1958.
(4) (a) Scheidt, W. R.; Frisse, M. E. J. Am. Chem. Soc. 1975,
97, 17–21. (b) Pelczar, E. M.; Emge, T. J.; Krogh-Jespersen, K.;
Goldman, A. S. Organometallics 2008, 27, 5759–5767. (c) Taktak, S.;
Ye, W.; Herrera, A. M.; Rybak-Akimova, E. V. Inorg. Chem. 2007, 46
2929–2942. (d) Kohl, S. W.; Heinemann, F. W.; Hummert, M.; Bauer,
W.; Grohmann, A. Chem.—Eur. J. 2006, 12, 4313–4320. (e) Bouwkamp,
M. W.; Lobkovsky, E.; Chirik, P. J. J. Am. Chem. Soc. 2005,
127, 9660–9661. (f) Bart, S. C.; Chzopek, K.; Bill, E.; Bouwkamp,
M. W.; Lobkovsky, E.; Neese, F.; Wieghardt, K.; Chirik, P. J. J. Am. Chem.
Soc. 2006, 128, 13901–13912. (g) Bart, S. C.; Lobkovsky, E.; Chirik, P. J.
J. Am. Chem. Soc. 2004, 126, 13794–13807. (h) Wu, J. Y.; Stanzl, B. N.;
Ritter, T. J. Am. Chem. Soc. 2010, 132, 13214–13216.
(5) (a) Goedken, V. L.; Peng, S.-M.; Molin-Norris, J. A.; Park, Y.-A.
J. Am. Chem. Soc. 1976, 98, 8391–8400. (b) Silvernail, N. J.; Noll, B. C.;
Schulz, C. E.; Scheidt, W. R. Inorg. Chem. 2006, 45, 7050–7052.
(6) Braunstein, P. Chem. Rev. 2006, 106, 134–159.
(7) van der Vlugt, J. I.; Reek, J. N. H. Angew. Chem., Int. Ed. 2009,
48, 8832–8846.
(8) Friedrich, A.; Drees, M.; K€ass, M.; Herdtweck, E.; Schneider, S.
Inorg. Chem. 2010, 49, 5482–5494.
(9) (a) Ozerov, O. V.; Guo, C.; Papkov, V. A.; Foxman, B. M. J. Am.
Chem. Soc. 2004, 126, 4792–4793. (b) Gregor, L. C.; Chen, C.-H.;
Fafard, C. M.; Fan, L.; Guo, C.; Foxman, B. M.; Gusev, D. G.; Ozerov,
O. V. Dalton Trans. 2010, 39, 3195–3202. (c) C-elenligil-C-etin, R.;
Watson, L. A.; Guo, C.; Foxman, B. M.; Ozerov, O. V. Organometallics
2005, 24, 186–189.
(10) (a) Gunanathan, C.; Ben-David, Y.; Milstein, D. Science 2007,
317, 790–792. (b) Milstein, D. Top. Catal. 2010, 53, 915–923.
9665
dx.doi.org/10.1021/ja202375y |J. Am. Chem. Soc. 2011, 133, 9662–9665