Iron(II) complexes containing unsymmetrical P-N-P′ pincer ligands for the catalytic asymmetric hydrogenation of ketones and imines

Article

pubs.acs.org/JACS

Iron(II) Complexes Containing Unsymmetrical P−N−P′ Pincer

Ligands for the Catalytic Asymmetric Hydrogenation of Ketones and

Imines

Paraskevi O. Lagaditis, Peter E. Sues, Jessica F. Sonnenberg, Kai Yang Wan, Alan J. Lough,

and Robert H. Morris*

Department of Chemistry, University of Toronto, 80 St. George Street, Toronto, Ontario M5S 3H6 Canada

S

* Supporting Information

ABSTRACT: After their treatment with LiAlH₄and then

alcohol, new iron dicarbonyl complexes mer-trans-[Fe(Br)-

(CO)₂(P−CHN−P′)][BF₄] (where P−CHN−P′ =

R₂PCH₂CHNCH₂CH₂PPh₂and R = Cy or iPr or P−

CHN−P′ = (S,S)- Cy₂PCH₂CHNCH(Me)CH(Ph)-

PPh₂) are catalysts for the hydrogenation of ketones in THF

solvent with added KOtBu at 50 °C and 5 atm H₂. Complexes

with R = Ph are not active. With the enantiopure complex,

alcohols are produced with an enantiomeric excess of up to

85% (S) at TOF up to 2000 h⁻¹, TON of up to 5000, for a

range of ketones. An activated imine is hydrogenated to the amine in 90% ee at a TOF 20 h⁻¹and TON 99. This is a signiﬁcant

advance in asymmetric pressure hydrogenation using iron. The complexes are prepared in two steps: (1) a one-pot reaction of

phosphonium dimers ([cyclo-(PR₂CH₂CH(OH)⁻)₂][Br]₂), KOtBu, FeBr₂, and Ph₂PCH₂CH₂NH₂(or (S,S)-Ph₂PCH(Ph)CH-

(Me)NH₂for the enantiopure complex) in THF under a CO atmosphere to produce the complexes cis- and trans-

[Fe(Br)₂(CO)(P−CHN−P′)]; (2) the reaction of these with AgBF₄under CO(g) to aﬀord the dicarbonyl complexes in high

yield (50−90%). NMR and DFT studies of the process of precatalyst activation show that the dicarbonyl complexes are

converted ﬁrst to hydride−aluminum hydride complexes where the imine of the P−CHN−P′ ligand is reduced to an amide

[P−CH₂N−P′]⁻with aluminum hydrides still bound to the nitrogen. These hydride species react with alcohol to give

monohydride amine iron compounds FeH(OR′)(CO)(P−CH₂NH−P′), R′ = Me, CMe₂Et as well as the iron(0) complex

Fe(CO)₂(P−CH₂NH−P′) under certain conditions.

and later Beller and co-workers showed that it also worked for

activated imines.²⁹Since then we have focused on optimizing

the transfer hydrogenation catalysts and have developed highly

active and enantioselective iron(II) catalysts, [Fe(P−N−N−

P)(CO)(Br)][BPh₄], where P−N−N−P is a unique tetraden-

tate ligand, (S,S)-Ph₂PCH₂CHNC(H)PhC(H)-

PhNCHCH₂PPh₂formed by the condensation of an (S,S)-

diamine with phosphine aldehydes templated by iron(II).^30,31

We have also modiﬁed the ligand by substituting the phenyl

substituents on phosphorus with alkyl or substituted phenyl

substituents to examine their eﬀects on the catalytic

behavior.^32,33The precatalysts are activated in situ by reaction

with base (KOtBu) to form bis-eneamido iron(II) complexes

which are then subsequently half reduced by isopropanol.^34,35

The catalytic reduction of ketones involves a bifunctional

mechanism where the ligand is directly involved in the

catalysis.^35,36Our catalysts are exceedingly eﬃcient for the

reduction of prochiral ketones by transfer of hydrogen from

isopropanol achieving TOFs of up to 200 s⁻¹, conversions of

INTRODUCTION

■

Current research in the ﬁeld of catalytic hydrogenation has

moved toward developing complexes that do not employ

platinum group metals such as Ir, Rh, or Ru.¹These platinum

group metals are expensive because of their low abundance and

are toxic; hence, they are undesirable for some applications.

Several examples of catalysts based on 3d metals have appeared

in the literature in recent years that have been shown to be

competitive with these precious metal-based catalysts for the

asymmetric catalytic reduction of unsaturated bonds.²⁻¹¹Some

systems based on enzymes,¹²⁻¹⁵and metal-free compounds are

also promising.¹⁶⁻²²

In recent years, a variety of iron-based hydrogenation

catalysts have been developed because iron is abundant,

cheap, and nontoxic.²³⁻²⁷Our group reported that an iron

compound with a tetradentate ligand was moderately active for

the asymmetric hydrogenation of acetophenone under basic

conditions with a turnover frequency (TOF) of 5 h⁻¹resulting

in an enantiomeric excess (ee) of 27% (S) toward 1-

phenylethanol (M1 of Figure 1). We found that the

monocarbonyl derivative of M1 was a good asymmetric transfer

hydrogenation precatalyst for ketones at room temperature,²⁸

Received: August 13, 2013

A

dx.doi.org/10.1021/ja4082233 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Figure 1. Active iron catalysts for ketone (M1−3 and B1) and imine (B2) hydrogenation.

99%, with enantioselectivity (ee) toward one enantiomer of the

alcohol as high as 98% at room temperature.^35,37They are also

active and exceptionally enantioselective for the transfer

hydrogenation of certain activated imines.^37,38Transfer hydro-

genation of ketones in this way is an equilibrium process, and

sometimes racemization of the product alcohol is observed.

Therefore, we sought an active asymmetric catalyst that utilizes

H₂gas, thereby enabling the irreversible hydrogenation of

substrates in order to conveniently achieve complete

conversion without racemization of the product alcohol.

Recent work by Milstein and co-workers demonstrated that

tridentate P−N−P pincer ligands can be eﬀective. Their

iron(II) complexes, Fe{2,6-(PiPr₂CH₂)₂C₅H₃N}(H)(CO)(Br)

(M2)³⁹and Fe{2,6-PiPr₂CH₂)₂C₅H₃N}(H)(CO)(HBH₃)

(M3)⁴⁰shown in Figure 1, were found to be active for the

catalytic hydrogenation of ketones with 4.1 atm H₂, achieving

turnover numbers (TON) for acetophenone hydrogenation of

up to 1720 with TOF of approximately 430 h⁻¹at 40 °C for

M2 and up to 1980 with TOF 300 h⁻¹at 40 °C for M3. The

latter complex did not require activation by the addition of

base. A mechanistic investigation using experimental and DFT

methods led to the conclusion that the catalysts operate by

Milstein’s well-established aromatization−dearomatization of

the P−N−P ligand that has been previously demonstrated to

occur with Ru analogues.⁴¹⁻⁴³More recently, Kirchner and co-

workers have presented an analogue to M1, Fe{2,6-

(PiPr₂CH₂)₂C₅H₃N}(H)(CO)(Cl) with −NH in place of the

methylene linkers between the pyridine and phosphorus groups

on the P−N−P ligand. The have also observed dearomatization

and aromatization under hydrogen; however, no report of

catalytic activity in hydrogenation of ketones was reported.⁴⁴In

the current work we generate asymmetric catalysts related to

these using our templated ligand synthesis methodology.

tions.⁴⁵⁻⁴⁸More recently Berkessel et al. synthesized a chiral

analogue of Knolker’s complex by the replacement of a CO

̈

ligand with chiral phosphoramidite ligands (B1, Figure 1),⁴⁹

while Beller and co-workers added a chiral phosphoric acid as a

cocatalyst (B2, Figure 1).^50,51The Berkessel system hydro-

genated acetophenone under UV irradiation with 10 atm H₂at

25 °C with nine turnovers in 24 h to give 1-phenylethanol in

30% ee, while the Beller system produced chiral amines in up to

96% ee using 50 atm H₂and 65 °C with 16 turnovers based on

iron in 24 h. Herein we report our progress toward discovering

more active iron(II) catalysts for the asymmetric H₂hydro-

genation of polar bonds.

RESULTS AND DISCUSSION

■

Template Synthesis of Iron(II) Complexes. We

previously reported the synthesis of the iron(II)−P−CH

N−P′ complexes, [Fe(Ph₂PCH₂CH₂NCHCH₂PR₂)-

(NCCH₃)₃]²⁺(where R = Ph or Cy), by means of a

multicomponent template synthesis using a cyclic phospho-

nium salt as a source of phosphine-aldehyde, [Fe(H₂O)₆]-

[BF₄]₂, KOtBu, and 2-(diphenylphosphino)ethylamine in

acetonitrile.⁵²These P−CHN−P′-pincer ligands have

inequivalent phosphorus donors, a feature which is not easy

to achieve by the conventional means of synthesis of P−N−P-

pincer ligands.⁵³⁻⁵⁵The ³¹P{¹H} NMR spectra of both

complexes displayed two AB doublets with a large ²J_PPcoupling

of 160 Hz for R = Ph and 148 Hz for R = Cy, indicating that

the P are trans; this indicates a mer-arrangement of the P−

CHN−P′ ligand about iron.

With this strategy in mind, we modiﬁed our template

synthesis procedure and used FeBr₂as our iron source. Initially

we conducted the reaction shown in Scheme 1 in THF under a

N₂atm in hopes of isolating an Fe(P−CHN−P′)(Br)₂

complex; however, no reaction occurred. Upon exposure of

the reaction mixture to carbon monoxide, the pale-yellow slurry

Knolker’s Fe(II) complex was found by Casey and co-

̈

workers to catalyze the hydrogenation of ketones via a

bifunctional mechanism using H₂gas under mild condi-

B

dx.doi.org/10.1021/ja4082233 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

selectively form one isomer. When Fe(CO)₄(Br)₂was used in

place of FeBr₂, there was an immediate release of gas as well as

a color change to orange. A ³¹P{¹H} NMR spectrum of this

mixture revealed an intractable mixture. This was then exposed

to UV light. After at least 5 h, the solution turned dark purple

and a ³¹P{¹H} NMR spectrum of the solid isolated upon

Scheme 1. Template Synthesis of Iron Complexes, trans-

[Fe(P−CHN−P′)(CO)₂(Br)][BF₄]

workup showed that the isomer at 67.8 and 71.9 ppm (²J_PP

=

175 Hz) (R = Cy) was the major species with, on average, less

than 10% of other species. Crystals suitable for X-ray

crystallography were isolated and conﬁrmed selective formation

of the trans-Fe(Cy₂PCH₂CHNCH₂CH₂PPh₂)(CO)(Br)₂

complex, 2a (Figure 2). Complex trans-Fe(iPr₂PCH₂CH

Figure 2. ORTEP plot (thermal ellipsoids at 30% probability) of the

X-ray crystal structure of trans-Fe(Cy₂PCH₂CHNCH₂CH₂PPh₂)-

(CO)(Br)₂, 2a. Hydrogen atoms of Ph and Cy substituents removed

for clarity. Selected bond lengths (Å) and angles (deg): Fe(1)−P(1):

2.2680(9); Fe(1)−P(2): 2.2613(9); Fe(1)−N(1): 2.011(2); Fe(1)−

Br(1): 2.4545(5); N(1)−C(2): 1.269(4); N(1)−C(3): 1.479(4):

O(1)−C(5): 1.138(4); C(5)−Fe(1)−N(1): 177.6(1); P(2)−Fe(1)−

P(1): 167.75(3); Br(1)−Fe(1)−Br(2): 175.12(2).

immediately turned red-purple. After stirring for suﬃcient time

(5 h) and removal of salts (KBr and excess FeBr₂) a red-purple

solid was isolated. Starting with the phosphonium dimer 1a

with R = Cy, iron compounds were produced in approximately

a 1:1 ratio based on the ³¹P{¹H} NMR spectra. One set of

doublets was observed at 67.8 and 71.9 ppm (²J_PP= 172 Hz)

and a second set at 39.6 and 60.5 ppm (²J_PP= 201 Hz). The

large coupling constants are indicative that the P atoms of the

P−CHN−P′ ligand are trans about the iron metal center.

Similar results were obtained from 1b with R = iPr and 1c with

R = Ph (Table S1 in Supporting Material [SI]). Hence, two

isomers of iron complexes formed: trans-Br (2a-c) and cis-Br

(3a-c) (Scheme 1). The formation according to Scheme 1 of

two mer-Fe(P−CHN−P′)(CO)(Br)₂isomers contrasts with

Milstein’s system where only the trans Br isomer was isolated.³⁹

Our results closely resemble those of Kirchner and co-

workers who have developed the synthesis of Fe(P−N−

P)(CO)(X)₂complexes, where X is Cl or Br, based on the

pincer ligand, 2,6-(PiPr₂NH)₂C₅H₃N.^56,57They have found

that when X is Cl the cis isomer formed under solvent-free

conditions while the trans isomer formed in solution. However,

in the case where X is Br, they did not observe the same

selectivity and always obtained a mixture of cis and trans

isomers. We attempted to make the Cl analogue as well using

the Cl salt of 1b and FeCl₂in the template synthesis but both

cis and trans isomers formed as well based on the ³¹P{¹H}

NMR spectrum of the reaction mixture (see the SI, Table S1).

We postulate that in our case there is no control of cis- and

trans-Br or -Cl isomers since we are employing a multi-

component reaction where the ligand is made in situ.

NCH₂CH₂PPh₂)(CO)(Br)₂, 2b, using phosphonium dimer 1b

was also made in this manner. However, in the case of

phosphonium dimer 1c, a mixture of cis- and trans-isomers, 2c

and 3c, formed in every attempt. These photochemical

syntheses have several limitations such as low yields and the

formation of cis-[Fe(P−CHN−P′)(CO)₂(Br)]⁺as described

in the SI and thus were not pursued further.

Kirchner and co-workers have reported the formation of

trans-[Fe(P−N−P)(CO)₂(Br)][BF₄] compounds selectively

from the reaction of a halide abstractor (such as AgBF₄) with

a mixture of cis- and trans-Fe(P−N−P)(CO)(Br)₂isomers.⁵⁹

Hence, upon inclusion of AgBF₄in our procedure there was an

immediate color change from red-purple to bright purple, the

color of complexes 4a−c (Scheme 1). The ³¹P{¹H} NMR

spectra of isolated compounds showed the formation of one

new iron complex with AB doublets with smaller ²J_PPcouplings

of 82 Hz (R = Cy), 85 Hz (R = iPr) and 94.5 Hz (R = Ph),

signiﬁcantly smaller than that of cis-[Fe(P−CHN−P′)-

2

(CO)₂(Br)]⁺with J_PP= 145 Hz (see the SI). Crystals of 4a

and 4b suitable for X-ray crystallography were obtained (Figure

3) and conﬁrmed that the trans-CO geometry about iron also

occurred with our compounds. We believe the mechanism for

this selective trans conﬁguration follows the one proposed by

Kirchner and co-workers because we also found that this

reaction can occur in the absence of carbon monoxide;⁵⁹

however, to achieve high yields (85−90%) we conducted the

halide abstraction step under one atmosphere of CO. The

58

We tested the compound Fe(CO)₄(Br)₂as a starting iron

source, reasoning that, if the halides are already coordinated to

iron and the CO ligands are removed under UV light, we could

C

dx.doi.org/10.1021/ja4082233 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

214.7 ppm. The IR spectra of complexes 4a−d displayed only

one ν_COabsorption in the range from 2000 to 2011 cm⁻¹,

similar to those of the trans-CO iron complexes reported by

Kirchner and coworkers.

Catalytic Hydrogenation. The discovery of the activation

of the precatalysts 4a−d for hydrogenation was made while

treating these well-deﬁned iron dicarbonyl complexes with

various hydride reagents. The objective was to produce iron

hydride amine complexes with the bifunctional HN-FeH group

known to eﬃciently reduce CO and CN bonds.^23,35,37

Reactions with NaHBEt₃or NaBH₄resulted in the formation of

intractable mixtures. The use of LiAlH₄was more promising.

To generate iron hydrides, often a precursor is reacted with a

slight excess of LiAlH₄in THF,⁶⁰⁻⁶⁴sometimes with the

addition of a protic solvent.⁶⁵⁻⁶⁹In the present case, we found

that the addition of at least 6 equiv of LiAlH₄was necessary to

produce a solution of iron hydride complexes reproducibly (see

below). The mixture was then treated with alcohol until gas

evolution had ceased. Methanol, ethanol, and tert-amyl alcohol

(2-methyl-2-butanol, tAmylOH) were found to give active

catalyst preparations. Then the ketone or imine substrate and

additional THF were added, and the entire solution was

injected into a prepared pressure reactor. This procedure then

allowed facile and eﬃcient screening of optimal conditions for

catalysis (Table 1).

It was discovered that, while 25 atm H₂was an eﬀective

pressure for catalysis, the use of 5 atm still resulted in full

conversion of acetophenone to 1-phenylethanol within 15 min.

Unsurprisingly, faster conversions were observed at 50 °C than

at 25 °C (Table 1, entry 2 vs 3). Substitution of LiAlH₄with

NaAlH₄did not aﬀect the rate of reaction (entry 4). No

catalysis was observed in the absence of base (entry 7) which

prevents the testing of base-sensitive substrates. The use of a

tertiary alcohol, tAmylOH, in the catalyst activation process was

observed to create a more active system than the primary

alcohols, MeOH or ethanol (entry 2 vs 5 vs 6). There are at

least two possible explanations for this alcohol eﬀect; either the

catalyst activation period could be slower with methoxide or

ethoxide as a ligand, or there are other deactivation processes

that are more pronounced with a primary alkoxide.⁷⁰The use of

a tertiary alcohol also has the advantage of preventing a

nonselective transfer hydrogenation mediated by Al(OR)₃. It

appears to be imperative for the P−N−P′ ligand to contain at

least one large alkyl-substituted phosphorus atom to create an

active catalyst (Table 1, entries 1 and 2 vs 9). We do not yet

know whether this is a steric or electronic eﬀect and at which

stage in catalysis it applies. The TOF value for the

hydrogenation of acetophenone using the 4a precatalyst

(entry 8, TON 2000, TOF 1980 h⁻¹at 50 °C and 5 atm H₂)

is comparable to that reported for Milstein’s complexes M2,

430 h⁻¹at 40 °C, 4 atm H₂. A diﬀerence of the present system

is that THF can be used in place of the alcohol solvent needed

for catalysis with M2 or M3 of Figure 1.

Figure 3. ORTEP plot (thermal ellipsoids at 30% probability) of the

−

X-ray crystal structures of 4a (left) and 4b (right). The BPh₄anions

−

of 4a, the BF₄anions of 4b and hydrogen atoms of Ph, Cy and iPr

substituents removed for clarity. Selected bond lengths (Å) and angles

(deg) of 4a: Fe(1)−C(6): 1.829(4); Fe(1)−N(1): 1.980(3); Fe(1)−

P(2): 2.271(1); Fe(1)−Br(1): 2.4416(6); O(1)−C(5): 1.104(4);

O(2)−C(6): 1.132(4); N(1)−C(3): 1.275(5); N(1)−C(2): 1.482(4);

C(6)−Fe(1)−C(5): 172.2(2); P(2)−Fe(1)−P(1): 168.38(4); N(1)−

Fe(1)−Br(1): 175.35(9); and of 4b: Fe(1)−P(1): 2.265(2); Fe(1)−

N(1): 1.980(4); Fe(1)−C(12): 1.803(6); Br(1)−Fe(1): 2.4530(8);

O(2)−C(12): 1.144(6); N(1)−C(2): 1.270(6); N(1)−C(3):

1.477(7); C(12)−Fe(1)−C(11): 170.4(2); P(1)−Fe(1)−P(2):

167.40(5); N(1)−Fe(1)−Br(1): 174.9(1).

¹³C{¹H} NMR spectra of complexes 4a−c displayed only one

broad triplet or a doublet of doublets (dd) for the CO ligands

at approximately 210 ppm which indicates that the complexes

have the expected C_ssymmetry. The ¹³CO analogue of 4b

(henceforth referred to as 4b-¹³CO) was synthesized to further

conﬁrm via NMR spectroscopy that two CO ligands are

coordinated to iron. The ³¹P{¹H} NMR (CD₂Cl₂) spectrum of

4b-¹³CO displayed the same two phosphorus resonances as 4b

2

at 45.5 and 78.3 ppm (²J_PP= 86 Hz, J_PC= 22 Hz) but with a

triplet of doublet multiplicity instead of a doublet for each

phosphorus atom. The ¹³C{¹H} NMR spectrum of 4b-¹³CO

displayed one intense CO triplet resonance at 211.7 ppm (²J_CP

= 24 Hz).

The chiral aminophosphine, (S,S)-2-amino-1-phenylpropyl-

diphenylphosphine was successfully used in the template

synthesis method to generate the chiral complex, (S,S)-4d

(Scheme 2). (S,S)-4d was designed with Cy substituents on the

ligand in anticipation that the bulky groups would enhance

enantiomeric interactions with the substrate during catalysis.

The ¹³C{¹H} NMR spectrum of (S,S)-4d showed two

inequivalent CO resonances with a dd pattern at 210.5 and

Scheme 2. Synthesis of the chiral iron complex, (S,S)-

[Fe(P−CHN−P′)(CO)₂(Br)][BF₄], (S,S)-4d

In the case of the hydrogenation of benzaldehyde, it was

found that the higher pressure of 10 atm of H₂gas was needed

to achieve 90% conversion in 2.5 h vs 10 h with 5 atm H₂gas

(Scheme 3).

Finally, the substrate scope of the asymmetric hydrogenation

reaction of ketones was investigated with the conditions of 50

°C and pressure of 5 atm H₂using 0.1 mol % of the activated

mixture prepared from (S,S)-4d (Table 2). Many of the aryl

ketones were converted to the alcohols with good enantiose-

lectivity, typically about 80% (S). This system represents a

D

dx.doi.org/10.1021/ja4082233 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Table 1. Hydrogenation of Acetophenone Catalyzed by Achiral Complexes 4a,b Once Activated by Reaction with LiAlH₄and

a

Then Alcohol

a

b

a

c

entry

precatalyst

base

temp. (°C)

alcohol

C/S ratio

time (min, 99% conv)

TOF (h⁻¹

)

1

2

3

4

5

6

7

8

9

4a

4b

KOtBu

None

50

25

50

tAmylOH

MeOH

1/500

1/2000

1/500

15

1980

990

550

990

−

15

30

d

4b

30

4b

4a

4c

55

30

EtOH

tAmylOH

no conversion

60

KOtBu

1980

no conversion

−

a

b

Precatalyst activated in situ: 5 mg (4a, b, or c), 4 equiv LiAlH₄(0.05 mL 1 M in THF), followed by 0.5 mL alcohol; 6 mL THF. Base (C/B = 1/

10) dissolved in 4 mL THF and added into the reactor preloaded with THF solution with precatalyst and substrate to commence catalysis.

Calculated at 99% conversion. NaAlH₄used instead.

c

d

formation due to the presence of base. Fortunately, the

inhibitory eﬀect of enolate, which is likely to be a good ligand

for iron, does not poison the catalyst completely. Precatalyst

(S,S)-4d was found to give the moderately eﬃcient hydro-

genation of nonaromatic ketones (entry 17) but with inferior

enantioselectivity (ee =46%). The system was also found to

catalyze the reduction of 2-acetylthiophene (entry 18) and 2-

acetylfuran (entry 19) to near completion. However, the

catalysis was slower (TOF = 240 h⁻¹for 2-acetylthiophene and

220 h⁻¹for 2-acetylfuran) when compared to acetophenone

(TOF = 990 h⁻¹). The diﬀerence is believed to be due to the

stronger binding to the iron of the chelating substrates via the

O or S of the heterocycle. However, since catalysis can go to

completion, this chelation eﬀect must be reversible. The

chelation eﬀect is more pronounced in the case of the substrate

2-acetylpyridine (entry 20) such that catalysis ceased at 20%

conversion (or 200 TON) after one hour. It is presumed that

the substrate, the alcohol product, or both completely

deactivate the catalyst. An increase of the H₂pressure to 10

atm enabled the hydrogenation of 2-acetylpyridine to 60%

conversion (or 600 TON) in two hours and maintained a 74%

ee of the hydrogenated product.

Unlike Milstein’s iron complexes M2 or M3, the system

could not hydrogenate the unsaturated ketone, trans-4-phenyl-

3-buten-2-one, at 5 or 10 atm H₂pressure (Table 2, entry 21).

The presence of an oleﬁn does not poison the catalyst because

the (S,S)-4d system was able to fully hydrogenate 5-hexen-2-

one (entry 22) to 5-hexen-2-ol without aﬀecting the oleﬁn as

determined by NMR spectroscopy. However, in comparison

with benzylacetone (entry 15) where catalytic hydrogenation

was complete in one hour, the hydrogenation of 5-hexen-2-one

was complete in four hours. Furthermore, upon addition of 1-

hexene (500 equiv) to a hydrogenation reaction of

acetophenone (1000 equiv), catalysis was complete in 90 min

as opposed to 30 min. These results imply there is some degree

of reversible oleﬁn coordination analogous to the situation with

the heterocyclic substrates (entries 18−20). Enone and diene

compounds of iron are known and have been well studied.⁷⁴⁻⁷⁸

It is postulated that there is some degree of reversible

coordination with enones and dienes as well with our iron

catalysts because upon addition of trans-4-phenyl-3-buten-2-

one (500 equiv) to a hydrogenation reaction of acetophenone

Scheme 3. H₂hydrogenation of benzaldehyde with an in situ-

generated catalyst derived from 4b

signiﬁcant improvement in activity over the other iron-based

catalysts shown in Figure 1 for the enantioselective hydro-

genation of ketones that employ H₂gas. It is unusual that the

(S,S) chiral catalyst system produces the (S) alcohol

enantiomer. Typically the (R) enantiomer of the alcohol is

produced using ligands derived from the (S,S)-diamine, at least

for our iron-based asymmetric transfer hydrogenation catalysts

and for most ruthenium diamine-based hydrogenation

catalysts.⁷¹⁻⁷³

The catalyst mixture starting with (S,S)-4d was tested for

acetophenone hydrogenation at both 50 and 25 °C. The

enantioselectivity of the reaction increased from 80% to 89%

(S), while the time for complete conversion lengthened from

30 to 90 min, respectively. The catalytic mixture at 50 °C can

accommodate additional loadings of substrate. Four additional

batches of acetophenone (1000 equiv) were added in 30-min

intervals without slowing down the catalyst, and the ee of the

product (S)-1-phenylethanol was unaﬀected. A ﬁfth batch

slowed the catalytic reaction, but it still went to completion

resulting in a net TON of 5000 (see the Supporting

Information, Figure S37). The enantioselectivity of the (S,S)-

4d system decreases along with the activity as steric hindrance

due to bulky substituents on the ketone substrate increases

(Table 2, entry 7 vs 8 vs 9). The low enantioselectivity is not

due to racemization, as the reactions were monitored

periodically and the ee remained constant until completion.

In the case of the less bulky substrate benzylacetone (entry 15),

the catalytic reduction by the (S,S)-4d system was rapid, but

enantioselectivity was lost (ee = 5%). In the case of the

substrate 1-phenyl-2-butanone (entry 16) that is more sterically

hindered than benzylacetone, the ee increased to 30%, while

there was a substantial decrease in catalytic activity (TOF = 90

h⁻¹). This lower catalytic activity is the result of enolate

E

dx.doi.org/10.1021/ja4082233 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Table 2. Reactivity of various ketones in the asymmetric hydrogenation reaction using an in situ-generated catalyst derived from

a

(S,S)-4d.

a

Catalyst prepared in situ: (S,S)-4d (5 mg, 0.007 mmol) mixed with 6 equiv LiAlH₄(0.05 mL of 1 M LiAlH₄in THF) followed by 0.5 mL tAmylOH

b

c

d

1

in 6 mL THF. Determined by GC. Determined by H NMR spectroscopy. Determined by HPLC.

(1000 equiv), catalysis was complete in 90 min with zero

conversion of the enone additive.

Scheme 4. H₂Hydrogenation of N-

(Diphenylphosphinoyl)propiophenoneimine Using an in

Situ-Generated Catalyst Derived from (S,S)-4d

The hydrogenation of imines was also investigated using

activated (S,S)-4d under the harsher conditions of 20 atm H₂

gas and at 50 °C at substrate loadings of 1 mol % (based on

(S,S)-4d) and 10 mol % base. The system was inactive for the

hydrogenation of the imine substrates, N-(1-phenylethylidene)-

aniline or phenyl-N-(1-phenylethylidene)methanamine, or the

nitrile substrate, benzylnitrile. However, it did catalyze the

complete hydrogenation of the activated imine, N-

(diphenylphosphonyl)propiophenoneimine in 22 h with a

TOF of 5 h⁻¹(Scheme 4). The activity of this system is

higher than that of Beller and co-worker’s cooperative system

B2 which had a TOF of about 1 h⁻¹under harsher conditions.⁵⁰

The enantioselectivity of activated (S,S)-4d was high (ee = 90%

F

dx.doi.org/10.1021/ja4082233 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Scheme 5. Synthesis and Proposed Structures of Hydride−Aluminum Hydride Iron Complexes 5a−d, 6a−d and the

Monohydride Complexes 7b (R‴ Me) and 8b and (S,S)-8d (R‴ tAmyl)

(S)), although not as high as the 99% (R) obtained for our

active transfer hydrogenation iron precatalyst, (S,S)-[Fe(CO)-

(Br)(Ph₂PCH₂CHNC(H)PhCH(Ph)NCHCH₂PPh₂)][BPh₄]

with TOF = 200 s⁻¹.³⁴

Since the catalytic hydrogenation activity depended on the

nature of the alcohol used in the catalyst activation, the

treatment of the iron hydride−aluminum hydride complexes

with MeOH and tAmylOH were separately examined by NMR.

Upon addition of an excess of alcohol to a C₆H₆or THF

solution of complexes 5 and 6, iron precatalysts mer-

Fe(H)(OR‴)(CO)(P−CH₂NH−P′) (7 with R‴ = Me; 8

with tAmyl) were formed (Scheme 5). The color of the

solution changed from yellow to orange along with gas

evolution.

Study of the catalyst activation steps. Purple THF

solutions of 4a-d turned immediately dark brown and evolved

gas upon addition of the LiAlH₄activator (Scheme 5). Removal

of the THF solvent and subsequent removal of a black

precipitate with Et₂O aﬀorded a dark brown-yellow solution.

The isolated residue was examined by NMR spectroscopy and

found to contain two iron hydride-aluminum hydride species

with either trans hydrogens with a mer−P−N−P′ ligand on iron

(5a-d) or cis hydrogens with a fac−P−N−P′ ligand (6a-d) on

iron as shown in Scheme 5. For example 5b is thought to have a

mer conﬁguration because it has a large J_PPof 105 Hz while 6b

is fac with a small J_PPof 20 Hz. Since no proton source was

added to the reaction, we postulate that Al is still bound to the

amide as AlH₃. The ²⁷Al NMR spectrum supports this postulate

since it displays a large broad signal at 100 ppm. So also does

the IR spectrum which has ν_Al−Habsorptions at 1783 and 1648

cm⁻¹.⁷⁹The compound H₂Al[N(CH₂CH₂NMe₂)₂]AlH₃pro-

vides a precedent for such an alane-amide adduct.⁸⁰

Unfortunately, many crystallization attempts failed to yield

crystals. We rely on extensive NMR spectroscopic and density

functional theory analysis to deduce possible structures of the

complexes.

There are potentially up to two conﬁgurations of structures

5a-d (H trans to H, and H trans to N). However only the ﬁrst is

the major one as drawn. For structures 6a-d, there is only one

conﬁguration and its enantiomer for 6a and 6b and only one

for 6c (no enantiomer) but there are two diastereomers for

(S,S)-6d with the nitrogen in the R or S conﬁguration. The

chemical shifts, coupling constants and analysis including DFT

calculations that support the tentative structures of the iron

hydride species 5 and 6 shown in Scheme 5 are provided in the

SI.

The NMR spectra of the isolated residues showed multiple

iron hydride complexes with similar patterns; however, they

were iron monohydride complexes. The ¹H, ³¹P, and ¹³C NMR

spectra of complexes 7b, 7b-¹³CO, 8b and (S,S)-8d were

examined in detail to distinguish diﬀerences between the

reaction with MeOH as compared with that with tAmylOH.

Properties of the major isomers are listed in Table 3. The

³¹P{¹H} NMR spectrum demonstrated that there is no

dissociation of the P−N−P′ ligands and that there is a

downﬁeld shift of the phosphorus resonances for solutions of 8

vs solutions of 7, thus providing evidence for an alkoxide ligand.

The major iron monohydride compounds are postulated to

have a mer-structure with trans-phosphorus groups of the P−

2

N−P′ ligand since the J_PPare in the range of 120−150 Hz.

2

1

The J_HPvalues of the hydride resonance in the H NMR

spectrum are greater than 50 Hz, also consistent with structures

1

with hydride cis to phosphorus (see the SI). The H−³¹P

HMBC NMR spectrum enabled the correlation of the hydride

resonances to the appropriate phosphorus resonances for 7b,

1

7b-¹³CO, and (S,S)-8d (discussed later). The H−¹³C HMBC

NMR spectrum of 7b-¹³CO showed unique correlations for

each hydride to one CO ligand (NMR spectra are provided in

1

SI). Due to the complexity of the 1−4 ppm region of the H

NMR spectrum with strong solvent and crown ether

1

resonances, 2D experiments to locate the H NMR resonance

of the hydrogen on nitrogen, for example using ¹⁵N−¹H

1

HMQC or H−¹H NOESY, failed.

An important observation is that the complexes 5c and 6c

that form starting with complex 4c contain symmetrical

PPh₂CH₂CH₂NCH₂CH₂PPh₂ligands. This is apparent from

the ³¹P{¹H} NMR spectrum where 5c and 6c give singlets due

to the equivalent phosphorus nuclei (see the SI, Figure S14).

This is evidence that the imine group is reduced in the starting

complexes 4a-d during catalyst activation.

Therefore, for each isomer there is a mer-conﬁguration of the

complex with one P−NH−P′, hydride, alkoxide and carbonyl

ligand. The trans positions for the monodentate ligands are

inequivalent because of the position of the NH group. This

makes the total number of possible methoxide isomers: six for

7a and 7b (and six enantiomers), three for 7c and six for (S,S)-

7d. The tAmyl derivatives 8 have the same numbers of isomers.

G

dx.doi.org/10.1021/ja4082233 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

³¹P doublet of doublets at 94.8 and 75.3 ppm, with J_PH= 30 and

J_PP= 137 Hz. This hydride was also correlated to ¹³C{¹H}

NMR signals at 222.8 ppm for CO, 139.7 ppm for P−C(Ph)

and 24.5 ppm for P−C(iPr). Similarly, the −22.7 ppm hydride

is a doublet of doublets with J_HP= 52 and 56 Hz, and correlates

to ³¹P doublet of doublets at 95.7 and 75.0 ppm, with J_PH= 21

and J_PP= 136 Hz, and ¹³C{¹H} NMR signals at 222.4 ppm for

CO, 139.5 ppm for P−C(Ph) and 24.6 ppm for P−C(iPr).

These species are quite similar and are known to be

monohydride, monocarbonyl, iron P−N−P species and are

therefore proposed to be mer-Fe(H)(OMe)(CO)(P−

CH₂NH−P′), with the N−H up and down, relative to the

hydride, for each isomer, respectively. The −21.6 ppm hydride

is therefore a kinetic product and the −22.7 ppm hydride, a

thermodynamic product.

Table 3. NMR Chemical Shifts (THF-d₈) of Iron Precatalysts

of Interest

The labeled alcohol, ¹³CH₃OH, was utilized in the

preparation of 7b in order to attempt to verify the presence

of the methoxide ligand. However, only the resonance for free

¹³CH₃OH was detected. This is likely to be due to a dynamic

averaging of the methoxide and methanol resonances since

alkoxide ligands are usually hydrogen bonded to an alcohol as

hydrogen-bonded anions [ROH···OR]⁻.^81,82No enhancement

of the Fe−¹³CO signals were detected, and this shows that the

isomers of 7b and any 9 present do not contain carbonyl

ligands derived from the alcohol.^83,84

Preformed solutions of 7b containing only the two hydride

isomers at −21.6 and −22.7 ppm were tested for H₂

hydrogenation of acetophenone to 1-phenylethanol under 5

atm H₂pressure and 50 °C, as optimized. These solutions were

highly active, converting 1000 equiv of substrate in less than 15

min, but only in the presence of base. The role of the base is

under active investigation.

a

A mixture of hydride complexes 8b without contamination

by complex 9 was also tested for the hydrogenation of

acetophenone at 50 °C and H₂pressure of 25 atm in THF in

the presence of base (KOtBu). Full conversion of acetophe-

none to 1-phenylethanol occurred in 10 min with 0.2 mol %

catalyst loading based on the initial amount of 4b used to

synthesize 8b. The synthesis and isolation of the mixture of

hydrides 8b is tedious, and the resulting oily residue is diﬃcult

to handle. It is more practical to activate the precatalyst 4b

without the isolation of the hydride intermediates as described

above.

Residual coupling due to incomplete decoupling of the high-ﬁeld

hydride resonance.

The two isomers with hydride trans to carbonyl are known to

be too high in energy (see the DFT results below), whereas the

one with hydride trans to alkoxide with the alkoxide next to the

NH is lowest and is likely to be one of the isomers listed in

Table 3. We have not been able to distinguish between the

other isomers but see at least three of them.

The relative ratios of the various hydride species 7b appear to

vary slightly, depending on exact reaction conditions; the use of

slightly less alcohol favors the formation of the hydride species

with ¹H NMR signals at −18.6 ppm (Table 3), whereas the use

of excess alcohol yields much less of the −18.6 ppm hydride

1

The H NMR hydride resonances of the alkoxide species 7

and 8 (Table 3) are similar in chemical shift and coupling

constant to those of the hydride alkoxide complex Fe{2,6-

(PiPr₂CH₂)₂C₅H₃N}(H)(CO)(OCHPh ₂) reported by Mil-

stein and co-workers.³⁹The latter displays a triplet hydride

resonance at −19.9 ppm with J_HP= 53 and is thought to be a

catalytically active form of precatalyst M2 of Figure 1.

1

species and a new species with H NMR hydride signals at

−22.7 ppm. The presence of a separate species which does not

correlate to any hydrides is also often detected. This has been

assigned as a zero-valent Fe(CO)₂(P−CH₂NH−P′) complex 9

with ³¹P{¹H} doublets at 101.0 and 79.7 ppm with J_PPof 73 Hz.

The formation of this complex is associated with the presence

of base (see below for its independent synthesis), in this case

probably LiOH produced by reaction of the hydrides with

traces of water in the solvents. The presence of this complex

shows that ligand redistribution reactions are possible. Over a

period of 24−48 h the hydride signals at −18.6 ppm and all

minor hydride species disappear leaving a spectrum with only

The chiral precatalyst (S,S)-8d also exists as a mixture of iron

monohydride compounds with two major isomers (Table 3).

2

Again the large J_PPindicates trans phosphine coordination of

2

the P−CH₂N−P′ ligand while the large J_HPindicates cis

hydride and phosphine ligands. The chemical shifts of the

minor iron monohydride compounds are provided in the SI.

However, the phosphorus chemicals shifts of these minor

compounds were only identiﬁed via the ¹H−³¹P HMBC

spectrum and this implies they are in low concentration relative

to the two major isomers of (S,S)-8d. Despite the fact that

these complexes exist as a mixture of isomers, they are

surprisingly active for the hydrogenation of ketones and, in the

case of (S,S)-8d, they have surprisingly good enantioselectivity.

two hydride species at −21.6 and −22.7 in a 1:1 ratio. H−³¹P

1

HMBC and ¹H−¹³C HMBC enabled correlation of the hydride

to other nuclei for each isomer. The −21.6 ppm hydride is a

doublet of doublets with J_HP= 52 and 57 Hz, and correlates to

H

dx.doi.org/10.1021/ja4082233 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

For example the mixture of hydrides catalyzes the hydro-

genation of acetophenone to 1-phenylethanol with same ee and

activity as the precatalyst (S,S)-4d activated in situ (Scheme 6).

Scheme 6. Asymmetric Hydrogenation of Acetophenone

Catalyzed by the Mixture of Isomers of Alkoxide Hydrides

(S,S)-8d

The Properties of Isomeric Alkoxide Hydrides (S,S)-7d

As Determined by DFT Calculations. In order to support

our structural assignments for these alkoxides we investigated

the relative stabilities of the possible isomers by DFT studies.

The ﬁrst study utilized a large basis set to treat simpliﬁed

structures of the methoxide isomers where the phenyls on the

phosphorus were replaced with hydrogen, and the cyclohexyl

groups and backbone phenyl, with methyl groups. This study

was designed to reveal the electronically preferred isomers with

the mer-P−NH−P′ ligand stereochemistry that has already

been established by our NMR studies. Of the six possible

diastereomers, the two isomers with hydride trans to carbonyl

were found to be high in energy and thus too unstable to form

(see the SI, Figure S26). The isomer with the hydride trans to

methoxide and with the methoxide next to the NH group is the

most stable. This is the stereochemistry drawn for (S,S)-8d in

Scheme 6.

Figure 4. The relative energies of ﬁve possible diastereomers of (S,S)-

7d. A sixth high-energy diastereomer, D, with hydride trans to CO is

not shown. Bond lengths for structure F: Fe−H = 1.55, Fe−C = 1.69,

Fe−O = 1.99, Fe-N = 2.07, Fe−P (PPh₂) = 2.23, Fe−P (PCy₂) = 2.24

Å, O···HN 1.75 Å. Bond Angles: H−Fe−C = 100.9°, H−Fe−P(PCy₂)

= 83.7°, H−Fe−P(PPh₂) = 77.8°, H−Fe−N = 89.6°, O−Fe−C =

95.7°, O−Fe−P(PCy₂) = 92.4°, O−Fe−P(PPh₂) = 102.3°, O−Fe−N

= 73.7°, Fe−N−H = 88.5°, Fe−O−H (N−H) = 75.2°, N−H−O =

119.7°, Fe−O−C = 126.5°, Fe−C−O = 173.7°, C−Fe-N = 169.4°,

P(PPh₂)−Fe−P(PCy₂) = 158.9°, H−Fe−O = 163.2°.

Scheme 7. Synthesis of 9

The eﬀect of sterics on the relative stabilities of the isomers

was explored using a gas phase calculation of the full structures

treated with a smaller basis set. A summary of the results is

shown in Figure 4. There are four diastereomers that are low in

energy with the isomer F being the most stable. The methoxide

oxygen is within hydrogen-bonding distance of the N−H group

(O···H 1.75 Å).

Reaction of the Methoxide Hydrides 7b with Base.

Base is required for catalysis, but the methoxide hydride

complexes 7b are unstable in the presence of base when the

substrate and hydrogen are absent. When KOtBu was added to

a C₆H₆solution of prepared precatalyst 7b the orange solution

turned red after stirring for 10 min. After removal of a white

precipitate, the ³¹P{¹H} NMR (C₆D₆) spectrum of the isolated

red residue showed one major compound as two doublets at

79.7 and 101.0 ppm (J_PP= 73 Hz) along with minor

compounds and free P−CH₂NH−P′ ligand. Once crystals

suitable for X-ray diﬀraction were isolated by slow diﬀusion of

pentane into the C₆D₆solution, the major compound was

identiﬁed to be the neutral Fe(0) complex Fe-

(Ph₂PCH₂CH₂NHCH₂CH₂PiPr₂)(CO)₂, 9 (Scheme 7, Figure

5).

Complex 9 has a trigonal bipyramidal geometry. The

phosphorus atoms of the P−CH₂NH−P′ ligand and a CO

ligand are in the equatorial plane, while the second CO ligand

and N atom of the P−CH₂NH−P′ ligand occupy the axial

positions. The P−Fe−P bond angle is 119.55(2)° and is larger

than that observed for the [Fe(CO)(Br)(P−N−N−P)]⁺or

[Fe(NCCH₃)₂(P−N−N−P)]²⁺complexes whose P−Fe−P

bond angles ranged between 108 and 112°. The existence of

the amine in complex 9 provides additional evidence for the

reduction of the P−CHNH−P′ ligand by LiAlH₄. The IR

(KBr) spectrum of the isolated residue of 9 showed two intense

carbonyl absorptions at 1884 and 1791 cm⁻¹. A similar

complex, Fe{2,6-(Ph₂PCH₂)₂(C₅H₃N)}(CO)₂, was synthesized

and characterized by Chirik and co-workers and found to also

have a trigonal bipyramidal geometry about the Fe(0) and have

a similar IR spectrum.⁸⁵However, the P−N−P ligand of their

complex is rigid such that phosphorus atoms are found at the

axial positions about the iron center and create a large P−Fe−P

bond angle of 165.99(1)°. We were initially surprised that 9

was an Fe(0) compound that had acquired an additional CO

ligand, but we found that Milstein and co-workers also

observed the formation of the Chirik Fe(0) complex along

with free P−N−P ligand and black precipitate when they

performed a similar reaction with their well-deﬁned iron

hydride precatalyst with base.³⁹It is not clear how the reaction

to form 9 proceeds. The reaction was conducted under an

atmosphere of argon which rules out the possibility of

dinitrogen coordination. It is presumed that the second

carbonyl ligand is obtained via an intermolecular ligand

redistribution reaction still to be characterized. Complex 9

was tested and found to be inactive as a hydrogenation catalyst;

hence, we presume this complex forms when H₂or the

substrate is absent.

I

dx.doi.org/10.1021/ja4082233 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

or Sigma-Aldrich, degassed, and dried over activated molecular

sieves prior to use. All liquid ketone substrates were vacuum

distilled, degassed, and stored over activated molecular sieves.

87

Phosphonium dimers, 1a−c,^52,86and Fe(CO)₄(Br)₂were

synthesized according to literature procedures. 2-

(Diphenylphosphino)ethylamine was donated by Digital

Specialty Chemicals. All other reagents were purchased from

Sigma-Aldrich or Strem Chemicals and utilized without further

puriﬁcations. NMR spectra were recorded at ambient temper-

ature and pressure using Varian VnmrS-400 and Agilent DD2-

600 [¹H (400, 600 MHz), ¹³C{¹H} (100, 150 MHz), ³¹P{¹H}

(161, 242 MHz), ¹⁹F{¹H} (356 MHz)]. The ³¹P chemical shifts

were measured relative to 85% H₃PO₄as an external reference.

The ¹⁹F chemical shifts were referenced relative to CFCl₃. In

the synthesis of 2a and 2b, photolysis was performed using a

450 W mercury vapor lamp (model: Hanovia UV medium

pressure 450 W immersion lamp). The elemental analyses were

performed on a Perkin-Elmer 2400 CHN elemental analyzer.

Some complexes gave unsatisfactory carbon analyses but

acceptable hydrogen and nitrogen content because of a

combustion problem due to the tetraﬂuoroborate anion.⁸⁸

trans-Fe(Ph₂PCH₂CH₂NCHCH₂PCy₂)(CO)(Br)₂, 2a. A vial

was charged with 1a (80 mg, 0.125 mmol) and KOtBu (28 mg,

0.249 mmol) and 25 mL benzene. The slurry was allowed to

stir for 10 min by which time the mixture turned cloudy and

then was ﬁltered into a Schlenk ﬂask. To this solution was

added 2-(diphenylphosphino)ethylamine (57 mg, 0.249 mmol)

followed by Fe(CO)₄(Br)₂(81 mg, 0.249 mmol). The mixture

immediately evolved gas in addition to turning orange in color.

The ﬂask was immediately exposed to UV light and was

allowed to stir for 6 h or until the solution turned red-purple.

The ﬂask was removed from the UV source and ﬁltered

through a pad Celite to remove all precipitates. The solvent was

concentrated, and pentane (10 mL) was added to cause

precipitation of a pale-pink solid. The solid was washed with

pentane (5 mL) and dried under vacuum. Yield: 35% (60 mg).

¹H NMR (400 MHz, C₆D₆) δ: 1.18−1.95 (m, HCy), 2.47 (m,

2H, CH₂PCy₂), 2.66 (m, 2H, CHCy), 2.79 (m, 2H, CH₂PPh₂),

4.11 (m, 2H, CH₂N), 6.94 (m, 2H, HPh), 7.05 (m, 4H, HPh),

7.91 (m, 5H, HPh, CHN). ¹³C {¹H} NMR (100 MHz, C₆D₆)

δ: 26.5 (CCy), 27.9 (CCy), 28.7 (CH₂PPh₂), 29.9 (CCy), 30.6

(CH₂PCy₂), 36.4 (CCy), 62.6 (CH₂N), 129.5 (CPh), 133.1

(CPh), 135.7 (CPh), 146.6 (CPh), 173.3 (CHN), 227.8 (br t,

Figure 5. ORTEP plot (thermal ellipsoids at 30% probability) of the

X-ray crystal structure of 9. Hydrogen atoms of phenyl and isopropyl

substituents removed for clarity. Selected bond lengths (Å) and angles

(deg): Fe(1)−C(6): 1.723(2); Fe(1)−C(5): 1.797(2); Fe(1)−N(1):

2.088(2); Fe(1)−P(2): 2.1735(6); Fe(1)−P(1): 2.2038(6); C(6)−

Fe(1)−C(5): 93.94(9); C(6)−Fe(1)−N(1): 172.41(9); C(5)−

Fe(1)−N(1): 93.07(8); C(6)−Fe(1)−P(2): 89.62(7); C(5)−

Fe(1)−P(2): 120.34(6); C(5)−Fe(1)−P(1): 119.37(6); P(2)−

Fe(1)−P(1): 119.55(2); N(1)−Fe(1)−P(2): 84.34(5); C(6)−

Fe(1)−P(1): 95.00(7).

CONCLUSIONS

■

We have developed an eﬀective synthesis templated by iron(II)

to make unsymmetrical P−N−P′ ligands by the condensation

of phosphine-amines with phosphine-aldehydes, generated

from phosphonium dimers. A mixture of trans- and cis-iron

complexes, Fe(CO)(Br)₂(P−CHN−P′), 2 and 3, were

initially synthesized from the one-pot reaction with phospho-

nium dimers (1a−c), KOtBu, FeBr₂and PhP₂C₂H₄NH₂in

THF under a CO(g) atmosphere. Upon addition of AgBF₄to

these complexes under a CO atmosphere, the new complexes

trans-[Fe(CO)₂(Br)(P−CHN−P′)][BF₄] (4a−d) were synthe-

sized in high yield. Complexes 4a−d were reacted with LiAlH₄,

followed by alcohol, to generate a mixture of iron hydride

complexes with the proposed structures mer-Fe(H)(CO)(OR)-

(P−CH₂NH−P′) (where R = Me or tAmyl). The mixture of

monohydride iron precatalysts 7a, 7b, or (S,S)-8d that have

alkyl substituents on one of the phosphorus atoms of the P−

CH₂NH−P′ ligand are active for the H₂hydrogenation of

ketones and aldehydes under mild basic conditions (T = 50 °C,

p(H₂) = 5 atm). Catalytic performance reached TOF up to

2000 h⁻¹, TON up to 5000, and enantioselectivities up to 85%

(S). It has not been established yet how many of the

diastereomeric hydrides lead to active catalysts. Nevertheless,

we have shown that a variety of such iron hydride compounds

can be prepared in a few steps from aldehyde and amine

building blocks on Fe(II). A more well-deﬁned hydride system

will be needed to study the mechanism of catalysis in more

detail and further improve the catalyst performance.

J

_PC= 24.0 Hz, CO). ³¹P {¹H} NMR (161 MHz, C₆D₆) δ: 68.2

(d, J_PP= 174 Hz), 71.2 (d, J_PP= 174 Hz) ppm. IR (KBr) 1945

cm⁻¹(ν_CO). Anal. Calcd for C₂₉H₃₉NOP₂FeBr₂: C, 50.10; H,

5.65; N, 2.01; Found: C, 49.52; H, 5.88; N, 1.75. MS (ESI,

methanol/water; m/z⁺): 586.1 [C₂₈H₃₉NP₂FeBr]⁺.

trans-Fe(Ph₂PCH₂CH₂NCHCH₂PiPr₂)(CO)(Br)₂, 2b. The

pale-pink solid product was synthesized and isolated using the

procedure outlined for 2a: 1b (80 mg, 0.166 mmol); KOtBu

(38 mg, 0.331 mmol); 2-(diphenylphosphino)ethylamine (76

mg, 0.331 mmol); Fe(CO)₄(Br)₂(108 mg, 0.331 mmol). Yield:

25% (50 mg). ¹H NMR (400 MHz, C₆D₆) δ: 0.90 (dd, J_HP= 7,

14 Hz, 1H, CH(CH₃)₂), 1.21 (dd, J_HP= 7, 13 Hz, 6H, CH₃),

1.44 (dd, J_HP= 7, 13 Hz, 6H, CH₃), 1.70 (dd, J_HP= 7, 16 Hz,

1H, CH(CH₃)₂), 2.74 (m, 2H, CH₂PiPr₂) 2.77 (m, 2H,

CH₂PPh₂), 4.13 (m, J_HP= 20.6 Hz, 2H, CH₂N), 6.93 (m, 4H,

HPh), 7.04 (m, 2H, HPh), 7.22 (indirectly determined from

¹H−¹³C HSQC, CHN), 7.89 (m, 2H, HPh), 7.98 (m, 2H,

HPh), 8.45 (m, 1H, HPh). ¹³C {¹H} NMR (100 MHz, C₆D₆)

δ: 19.35 (CH₃), 19.5 (CH(CH₃)₂), 20.5 (CH₃), 25.8

EXPERIMENTAL SECTION

■

General Comments. All procedures and manipulations

involving air-sensitive materials were performed under an argon

or nitrogen atmosphere using Schlenk techniques or a glovebox

with N₂(g). Solvents were degassed and dried using standard

procedures prior to all manipulations and reactions. Deuterated

solvents were purchased from Cambridge Isotope Laboratories

(CH(CH₃)₂), 28.2 (d, J_CP= 20 Hz, CH₂PPh₂), 38.0 (d, J_CP

=

J

dx.doi.org/10.1021/ja4082233 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

17 Hz, CH₂PiPr₂), 61.3 (CH₂N), 129.4 (CPh), 131.5 (CPh),

132.8 (CPh), 135.3 (CPh), 171.8 (CHN), 227.9 (t, J_CP= 23.8

(73 mg, 0.648 mmol); 2-(diphenylphosphino)ethylamine (149

mg, 0.648 mmol); FeBr₂(208 mg, 0.971 mmol); AgBF₄(157

mg, 0.809 mmol). Yield: 50% (230 mg). H NMR (400 MHz,

Hz, CO). ³¹P {¹H} NMR (161 MHz, C₆D₆) δ: 67.4 (d, J_PP

=

1

176 Hz), 79.7 (d, J_PP= 176 Hz) ppm. IR (KBr) 1945 cm⁻¹

(ν_CO). Anal. Calcd for C₂₃H₃₁NOP₂FeBr₂: C, 44.91; H, 5.08;

N, 2.28; Found: C, 43.99; H, 5.12; N, 2.30. MS (ESI,

methanol/water; m/z⁺): 507.2 [C₂₂H₃₁NP₂FeBr]⁺.

CD₂Cl₂) δ: 2.16 (m, 2H, CH₂PPh₂), 2.99 (m, 2H, CH₂N), 3.77

(m, 2H, CH₂PPh₂), 7.54−7.94 (m, 21H, HPh, CHN). ¹³C

{¹H} NMR (100 MHz, CD₂Cl₂) δ: 29.2 (CH₂PPh₂), 31.7

(CH₂Ph₂), 62.5 (CH₂N), 129.4 (CPh), 131-.6−132.2 (CPh),

182.4 (CHN), 208.8 (br t, J_CP= 24 Hz, CO). ³¹P {¹H} NMR

(161 MHz, CD₂Cl₂) δ: 47.4 (d, J_PP= 95 Hz, PPh₂), 51.9 (d, J_PP

= 95 Hz, PCy₂). ¹⁹F{¹H} NMR (356 MHz, CD₂Cl₂) δ: −155.5

(s, BF₄) ppm. IR (KBr) 2016 cm⁻¹(ν_CO). Anal. Calcd for

C₃₀H₂₇NO₂P₂FeBrBF₄: C 50.18; H, 3.79; N, 1.92; Found: C,

44.38; H, 4.21; N, 1.58. MS (ESI, methanol/water; m/z⁺):

632.0 [C₃₀H₂₇NO₂P₂FeBr]⁺.

trans-[Fe(Ph₂PCH₂CH₂NCHCH₂PCy₂)(CO)₂(Br)][BF₄], 4a.

A Schlenk ﬂask was charged with 1a (200 mg, 0.311 mmol) and

KOtBu (70 mg, 0.623 mmol) and 25 mL THF. The slurry was

allowed to stir for 10 min by which time the mixture turned

cloudy. To this solution 2-(diphenylphosphino)ethylamine

(143 mg, 0.623 mmol) followed by FeBr₂(204 mg, 0.934

mmol). The Schlenk ﬂask was then exposed to an atmosphere

of CO (∼2 atm); upon exposure to CO, the pale yellow slurry

immediately turned dark purple. The reaction was allowed to

stir for 5 h by which time the reaction mixture was red-purple

in color. The solvent was removed and the residue was taken

up in 25 mL dichloromethane. The solution was ﬁltered

through a pad of Celite into a new Schlenk ﬂask and exposed

again to a CO atm. AgBF₄(130 mg, 0.668 mmol) in 5 mL THF

was injected into the reaction mixture. The solution

immediately changed to a bright purple color. After stirring

for 30 min, the solvent was removed, taken up in dichloro-

methane and ﬁltered through a pad of Celite to remove a gray

precipitate. The solvent was concentrated and pentane (10 mL)

was added to cause precipitation of a purple solid. The solid

was washed with pentane (5 mL) and dried under vacuum.

trans-(S,S)-[Fe(Ph₂PCH(Ph)CH(Me)NCHCH₂PCy₂)-

(CO)₂(Br)][BF₄], (S,S)-4d. The red-purple solid product was

synthesized and isolated using the procedure outlined for 4a:

1a (100 mg, 0.156 mmol); KOtBu (35 mg, 0.312 mmol); (S,S)-

Ph₂PCH(Ph)CH(Me)NH₂(100 mg, 0.312 mmol); FeBr₂(100

mg, 0.312 mmol); AgBF₄(70 mg, 0.359 mmol). Yield: 82%

1

(200 mg). H NMR (400 MHz, THF-d₈) δ: 0.72−2.05 (m,

1

HCy), 1.11 (CH₃, indirectly determined via H−¹H COSY),

2.21 (m, 1H, HCy), 2.61 (m, 1H, HCy), 3.43 (CH₂PCy₂,

indirectly determined via ¹H−¹H COSY), 3.67 (CH₂PCy₂,

indirectly determined via ¹H−¹H COSY), 3.85 (m, 1H,

CH(Me)), 4.14 (m, 1H, CH(Ph)), 6.80−7.97 (m, 16H,

1

HPh), 7.96 (d, indirectly determined via H−¹H COSY, J_HP

= 20 Hz, CHN). ¹³C {¹H} NMR (100 MHz, THF-d₈) δ: 12.5

(CCy), 17.2 (CCy), 21.3 (CCy), 27.1 (CCy), 28.4 (CCy), 35.4

(CH₂PCy₂), 37.3 (CCy), 38.2 (CCy), 51.8 (CH(Ph)), 70.4

(CH(Me)), 126.0−135.3 (CPh), 179.3 (CHN), 210.5 (br t, J_CP

= 23.5 Hz, CO), 214.7 (br t, J_CP= 21.3 Hz, CO). ³¹P {¹H}

NMR (161 MHz, THF-d₈) δ: 69.2 (d, J_PP= 81 Hz), 67.8 (d, J_PP

= 81 Hz). ¹⁹F{¹H} NMR (356 MHz, THF-d₈) δ: −155.9 (s,

BF₄) ppm. IR (KBr) 2000.0 cm⁻¹(ν_CO). Anal. Calcd for

C₃₇H₄₅NO₂P₂FeBrBF₄: C 54.17; H, 5.53; N, 1.71; Found: C,

46.31; H, 6.08; N, 1.05. MS (ESI, methanol/water; m/z⁺):

734.1 [C₃₇H₄₅NO₂P₂FeBr]⁺.

1

Yield: 89% (400 mg). H NMR (400 MHz, CD₂Cl₂) δ: 1.16−

2.50 (m, 22H, HCy), 2.50 (m, 2H, CH₂PCy₂), 2.92 (m, 2H,

CH₂PPh₂), 3.63 (m, 2H, CH₂N), 7.56−7.95 (m, 11H, HPh,

HCN). ¹³C {¹H} NMR (100 MHz, CD₂Cl₂) δ: 25.7 (CCy),

27.2 (CCy), 28.5 (CCy), 28.8 (CCy), 37.5 (CH₂PPh₂), 37.7

(CH₂PCy₂), 63.6 (CH₂N), 129.4 (CPh), 131.0 (CPh), 131.8

(CPh), 182.0 (HCN), 211.5 (dd, J_CP= 22, 25 Hz, CO). ³¹P

{¹H} NMR (161 MHz, CD₂Cl₂) δ: 45.7 (d, J_PP= 85 Hz, PPh₂),

70.8 (d, J_PP= 85 Hz, PCy₂). ¹⁹F{¹H} NMR (356 MHz,

CD₂Cl₂) δ: −155.5 (s, BF₄) ppm. IR (KBr) 2005 cm⁻¹(ν_CO).

Anal. Calcd for C₃₀H₃₉NO₂P₂FeBrBF₄: C, 49.35; H, 5.38; N,

1.92; Found: C, 45.94; H, 5.47; N, 1.52. MS (ESI, methanol/

water; m/z⁺): 644.1 [C₃₀H₃₉NO₂P₂FeBr]⁺.

General Procedure for the Synthesis of Complexes

5a−d and 6a−d. A vial was charged with [Fe(CO)₂(Br)(P−

CHN−P′)][BF₄] (∼20 mg) in THF (5 mL) to yield a bright

purple solution to which LiAlH₄was added until the solution

turned dark yellow-brown (∼20 mg). After stirring for 10 min,

the solvent was removed and the residue was taken up with

ether (5−10 mL) to remove the gray-black precipitate. 12-

crown-4 (∼4−5 drops) was added to cause the precipitation of

an oﬀ-white solid ([Li(12-crown-4)]BH₄]). The resulting

solution was ﬁltered and dried in vacuo to yield a yellow residue.

General Procedure for the Synthesis of Complexes 7

and 8. Following the same procedure as outlined for the

synthesis of 5a−d and 6a−d, an excess of alcohol (MeOH or

tAmylOH) was added dropwise to the ﬁnal Et₂O solution until

gas evolution ceased (∼10 drops). The yellow solution turned

orange. The solvent was removed, and the residue was taken up

with pentane and ﬁltered. The solution was dried in vacuo to

aﬀord an orange residue.

trans-[Fe(Ph₂PCH₂CH₂NCHCH₂PiPr₂)(CO)₂(Br)][BF₄], 4b.

The purple solid product was synthesized and isolated using the

procedure outlined for 4a: 1b (200 mg, 0.415 mmol); KOtBu

(93 mg, 0.830 mmol); 2-(diphenylphosphino)ethylamine (190

mg, 0.830 mmol); FeBr₂(271 mg, 1.245 mmol); AgBF₄(201

1

mg, 1.035 mmol). Yield: 83% (450 mg). H NMR (400 MHz,

CD₂Cl₂) δ: 1.13 (m, 2H, CH(CH₃)₂), 1.45 (m, 12H, CH₃),

2.78 (m, 2H, CH₂PiPr₂), 2.90 (m, 2H, CH₂PPh₂), 3.62 (m, 2H,

CH₂N), 7.53 (m, 4H, HPh), 7.91 (m, 7H, HPh, CHN). ¹³C

{¹H} NMR (100 MHz, CD₂Cl₂) δ: 15.5 (CH(CH₃)₂), 19.1

(CH₃), 28.2 (CH₂PiPr₂), 28.3 (CH₂PPh₂), 63.0 (CH₂N), 129.8

(CPh), 131.8 (CPh), 132.1 (CPh), 181.7 (HCN), 211.5 (dd,

J_CP= 22, 25 Hz, CO). ³¹P {¹H} NMR (161 MHz, CD₂Cl₂) δ:

45.5 (d, J_PP= 86 Hz, PPh₂), 78.2 (d, J_PP= 86 Hz, PiPr₂).

¹⁹F{¹H} NMR (356 MHz, CD₂Cl₂) δ: −155.5 (s, BF₄) ppm. IR

(KBr) 2011 cm ^{− 1}(ν_{C  O}). Anal. Calcd for

C₂₄H₃₁NO₂P₂FeBr₂BF₄: C, 44.35; H, 4.81; N, 2.15; Found:

C, 43.21; H, 4.88; N, 2.09. MS (ESI, methanol/water; m/z⁺):

563.2 [C₂₄H₃₁NP₂O₂FeBr]⁺.

General Procedure for Hydrogenation Studies. All of

the hydrogenation reactions were performed at constant

pressures using a stainless steel 50 mL Parr hydrogenation

reactor. The temperature was maintained at 50 °C using a

constant temperature water bath. The reactor was ﬂushed

several times with hydrogen gas at 5 atm prior to the addition

of catalyst and substrate, and base solutions. For standard

trans-[Fe(Ph₂PCH₂CH₂NCHCH₂PPh₂)(CO)₂(Br)][BF₄], 4c.

The purple solid product was synthesized and isolated using the

procedure outlined for 4a: 1c (200 mg, 0.324 mmol); KOtBu

K

dx.doi.org/10.1021/ja4082233 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

catalysis with in situ prepared catalysts, a vial was charged with

[Fe(CO)₂(Br)(P−CHN−P′)]BF₄(5 mg, 0.006 mmol) and

3 mL THF. To this solution, 0.05 mL of LiAlH₄(1 M in THF)

was added, and the color of the solution immediately changed

from purple to a golden brown. After stirring for 5 min, 2-

methyl-2-butanol (0.5 mL) was added; the solution was

allowed to stir for 10 min or until gas evolution ceased. The

solution was transferred to a syringe equipped with a 12 in.

needle. The same vial was then charged with substrate (6.095

mmol) and 3 mL THF. The solution was taken up into the

same syringe that already contained the precatalyst solution;

the needle was then stoppered. For catalysis with preformed

species 7b and (S,S)-8d, NMR solutions (THF-d₈) were

transferred to a vial in a nitrogen-ﬁlled glovebox, and THF (6

mL) and substrate (6.095 mmol) were added. The solution was

transferred to a syringe equipped with a 12 in. needle and

stoppered. A second vial was charged with KOtBu (10 mg,

0.089 mmol) and 3 mL THF. This solution was transferred to a

second syringe equipped with a 12 in. needle and stoppered as

well. Both syringes were taken out of the glovebox and injected

to the prepared Parr reactor against a ﬂow of hydrogen gas.

Small aliquots of the reaction mixture were quickly withdrawn

with a syringe and needle under a ﬂow of hydrogen at timed

intervals. Alternatively, small aliquots of the reaction mixture

were sampled from a stainless-steel sampling dip tube attached

to a modiﬁed Parr reactor. The dip tube was 30 cm in length

with an inner diameter of 0.01 in., and a swing valve was

attached to the end of the sampling tube. All samples for gas

chromatography (GC) analyses were diluted to a total volume

of approximately 1 mL using oxygenated ethanol. All

conversions were reported as an average of two runs. The

reported conversions were reproducible. The conversion and

enantiomeric excess of hydrogenated ketones were analyzed by

a Perkin-Elmer Clarus 400 chromatograph equipped with a

chiral column (CP chirasil-Dex CB 25 m × 2.5 mm) with an

autosampling capability used for GC analyses. Hydrogen was

used as a mobile phase at a column pressure of 5 psi with a split

ﬂow rate of 50 mL/min. The injector temperature was 250 °C,

and the FID temperature was 275 °C. The oven temperatures

and the retention times (t_SM, t_R, t_S/min) for substrates are

provided in the SI.

Computation. Density functional theory calculations on

the models of 5 and 6, and simpliﬁed models of (S,S)-7d were

performed using Gaussian 09.⁸⁹The M06 hybrid functional was

used for all calculations.^90,91All atoms were treated with the 6-

31++G(d,p) basis set. A pruned (99,590) integration grid was

used throughout (Grid = UltraFine). Optimizations were

performed in tetrahydrofuran using the integral equation

formalism polarizable continuum model (IEF-PCM) with

radii and non-electrostatic terms from the SMD solvation

model.⁹²Full vibrational and thermochemical analyses (1 atm,

298 K) were performed on optimized structures to obtain

solvent-corrected free energies (G°) and enthalpies (H°).

Optimized ground states were found to have zero imaginary

frequencies. DFT gas phase calculations on the full structures of

the isomers of (S,S)-7d were performed using GAMESS.^93,94

The M06 functional was used. Iron was treated with the

LANL2DZ basis set with an eﬀective core potential.⁹⁵Atoms C,

H, N, O, and P were treated with the 6-31G basis set.

ASSOCIATED CONTENT

■

S

* Supporting Information

Detailed information regarding limitations of the UV template

reaction, analysis of the NMR spectra of the iron hydride-

aluminum hydride compounds 5a−c and 6a−c, and alkoxide

complexes 7b, 7b−¹³CO, (S,S)-8d, DFT calculation details, the

complete reference 89, energies, structures and coordinates and

CIF ﬁle of all crystallographic data. This material is available

free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION

■

Corresponding Author

rmorris@chem.utoronto.ca

Notes

The authors declare no competing ﬁnancial interest.

ACKNOWLEDGMENTS

■

NSERC is thanked for a Discovery Grant to R.H.M., and OGS

and OGSST are thanked for scholarships to P.O.L. Digital

Specialty Chemicals donated the 2-(diphenylphosphino)ethyl-

amine. Demyan E. Prokopchuk is thanked for preliminary DFT

calculations of iron dihydride compounds. Dmitry Pichugin is

thanked for aiding with NMR spectroscopy. Computations

were performed on the General Purpose Cluster super-

computer at the SciNet HPC Consortium. SciNet is funded

by the Canada Foundation for Innovation under the auspices of

Compute Canada, the Government of Ontario, Ontario

Research Fund - Research Excellence, and the University of

Toronto.

REFERENCES

■

(1) Catalysis without Precious Metals; Bullock, R. M., Ed.; Wiley-VCH:

Hoboken, NJ, 2010.

(2) Tondreau, A. M.; Darmon, J. M.; Wile, B. M.; Floyd, S. K.;

Lobkovsky, E.; Chirik, P. J. Organometallics 2009, 28, 3928−3940.

(3) Monfette, S.; Turner, Z. R.; Semproni, S. P.; Chirik, P. J. J. Am.

Chem. Soc. 2012, 134, 4561−4564.

(4) Sylvester, K. T.; Chirik, P. J. J. Am. Chem. Soc. 2009, 131, 8772−

8774.

(5) Wu, J. Y.; Stanzl, B. N.; Ritter, T. J. Am. Chem. Soc. 2010, 132,

13214−13216.

(6) Dong, Z. R.; Li, Y. Y.; Yu, S. L.; Sun, G. S.; Gao, J. X. Chin. Chem.

Lett. 2012, 23, 533−536.

(7) Vasudevan, K. V.; Scott, B. L.; Hanson, S. K. Eur. J. Inorg. Chem.

2012, 4898−4906.

(8) Zhang, G.; Scott, B. L.; Hanson, S. K. Angew. Chem., Int. Ed.

2012, 51, 12102−12106.

(9) Harman, W. H.; Peters, J. C. J. Am. Chem. Soc. 2012, 134, 5080−

5082.

(10) Federsel, C.; Ziebart, C.; Jackstell, R.; Baumann, W.; Beller, M.

Chem.Eur. J. 2012, 18, 72−75.

(11) Werkmeister, S.; Fleischer, S.; Junge, K.; Beller, M. Chem.

Asian J. 2012, 7, 2562−2568.

(12) Huisman, G. W.; Liang, J.; Krebber, A. Curr. Opin. Chem. Biol.

2010, 14, 122−129.

(13) Savile, C. K.; Janey, J. M.; Mundorff, E. C.; Moore, J. C.; Tam,

S.; Jarvis, W. R.; Colbeck, J. C.; Krebber, A.; Fleitz, F. J.; Brands, J.;

Devine, P. N.; Huisman, G. W.; Hughes, G. J. Science 2010, 329, 305−

309.

(14) Ringenberg, M. R.; Ward, T. R. Chem. Commun. 2011, 47,

8470−8476.

(15) Matsuda, T.; Yamanaka, R.; Nakamura, K. Tetrahedron:

Asymmetry 2009, 20, 513−557.

L

dx.doi.org/10.1021/ja4082233 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

(16) Stephan, D. W.; Greenberg, S.; Graham, T. W.; Chase, P.;

Hastie, J. J.; Geier, S. J.; Farrell, J. M.; Brown, C. C.; Heiden, Z. M.;

Welch, G. C.; Ullrich, M. Inorg. Chem. 2012, 50, 12338−12348.

(17) Rueping, M.; Dufour, J.; Schoepke, F. R. Green Chem. 2011, 13,

1084−1105.

(53) Liang, L.-C.; Li, C.-W.; Lee, P.-Y.; Chang, C.-H.; Man Lee, H.

Dalton Trans. 2011, 40, 9004−9011.

(54) Liang, L.-C.; Chien, P.-S.; Lee, P.-Y. Organometallics 2008, 27,

3082−3093.

(55) Lansing, R. B., Jr.; Goldberg, K. I.; Kemp, R. A. Dalton Trans.

2011, 40, 8950−8958.

(56) Benito-Garagorri, D.; Alves, L. G.; Puchberger, M.; Mereiter, K.;

Veiros, L. F.; Calhorda, M. J.; Carvalho, M. D.; Ferreira, L. P.;

Godinho, M.; Kirchner, K. Organometallics 2009, 28, 6902−6914.

(57) Benito-Garagorri, D.; Puchberger, M.; Mereiter, K.; Kirchner, K.

Angew. Chem., Int. Ed. 2008, 47, 9142−9145.

(18) Sumerin, V.; Chernichenko, K.; Nieger, M.; Leskela, M.; Rieger,

̈

B.; Repo, T. Adv. Synth. Catal. 2011, 353, 2093−2110.

(19) Farrell, J. M.; Hatnean, J. A.; Stephan, D. W. J. Am. Chem. Soc.

2012, 134, 15728−15731.

(20) Farrell, J. M.; Heiden, Z. M.; Stephan, D. W. Organometallics

2011, 30, 4497−4500.

(58) Benito-Garagorri, D.; Wiedermann, J.; Pollak, M.; Mereiter, K.;

Kirchner, K. Organometallics 2007, 26, 217−222.

(59) Benito-Garagorri, D.; Alves, L. G.; Veiros, L. F.; Standfest-

Hauser, C. M.; Tanaka, S.; Mereiter, K.; Kirchner, K. Organometallics

2010, 29, 4932−4942.

(21) Mahdi, T.; Heiden, Z. M.; Grimme, S.; Stephan, D. W. J. Am.

Chem. Soc. 2012, 134, 4088−4091.

(22) Reddy, J. S.; Xu, B.-H.; Mahdi, T.; Frohlich, R.; Kehr, G.;

̈

Stephan, D. W.; Erker, G. Organometallics 2012, 31, 5638−5649.

(23) Morris, R. H. Chem. Soc. Rev. 2009, 38, 2282−2291.

(24) Junge, K.; Schroder, K.; Beller, M. Chem. Commun. 2011, 47,

4849−4859.

(60) Antberg, M.; Frosin, K. M.; Dahlenburg, L. J. Organomet. Chem.

1988, 338, 319−327.

(61) Antberg, M.; Dahlenburg, L. Z. Naturforsch., B: Chem. Sci. 1985,

40, 1485−1489.

(25) Darwish, M.; Wills, M. Catal. Sci. Technol. 2012, 2, 243−255.

(26) Gopalaiah, K. Chem. Rev. 2013, 113, 3248−3296.

(27) Le Bailly, B. A. F.; Thomas, S. P. RSC Adv. 2012, 1, 1435−1445.

(28) Sui-Seng, C.; Freutel, F.; Lough, A. J.; Morris, R. H. Angew.

Chem., Int. Ed. 2008, 47, 940−943.

(29) Zhou, S.; Fleischer, S.; Junge, K.; Das, S.; Addis, D.; Beller, M.

Angew. Chem., Int. Ed. 2010, 49, 8121−8125.

(30) Mikhailine, A.; Lough, A. J.; Morris, R. H. J. Am. Chem. Soc.

2009, 131, 1394−1395.

(31) Mikhailine, A. A.; Morris, R. H. Inorg. Chem. 2010, 49, 11039−

11044.

(62) Roger, C.; Marseille, P.; Salus, C.; Hamon, J.-R.; Lapinte, C. J.

Organomet. Chem. 1987, 336, C13−C16.

(63) Field, L. D.; Messerle, B. A.; Smernik, R. J.; Hambley, T. W.;

Turner, P. Inorg. Chem. 1997, 36, 2884−2892.

(64) Liu, T.; Chen, S.; O’Hagan, M. J.; Rakowski DuBois, M.;

Bullock, R. M.; DuBois, D. L. J. Am. Chem. Soc. 2012, 134, 6257−6272.

(65) Gao, Y.; Holah, D. G.; Hughes, A. N.; Spivak, G. J.; Havighurst,

M. D.; Magnuson, V. R.; Polyakov, V. Polyhedron 1997, 16, 2797−

2807.

(66) Ohki, Y.; Suzuki, H. Angew. Chem., Int. Ed. 2000, 39, 3120−

3122.

(67) Argouarch, G.; Hamon, P.; Toupet, L.; Hamon, J.-R.; Lapinte,

C. Organometallics 2002, 21, 1341−1348.

(68) Sellmann, D.; Weber, W. J. Organomet. Chem. 1986, 304, 195−

201.

(32) Lagaditis, P. O.; Lough, A. J.; Morris, R. H. Inorg. Chem. 2010,

49, 10057−10066.

(33) Sues, P. E.; Lough, A. J.; Morris, R. H. Organometallics 2011, 30,

4418−4431.

(34) Lagaditis, P. O.; Lough, A. J.; Morris, R. H. J. Am. Chem. Soc.

2011, 133, 9662−9665.

(69) Jia, G.; Drouin, S. D.; Jessop, P. G.; Lough, A. J.; Morris, R. H.

(35) Mikhailine, A. A.; Maishan, M. I.; Lough, A. J.; Morris, R. H. J.

Am. Chem. Soc. 2012, 134, 12266−12280.

Organometallics 1993, 12, 906−916.

(70) Fiedler, A.; Schroder, D.; Schwarz, H.; Tjelta, B. L.; Armentrout,

̈

(36) Prokopchuk, D. E.; Morris, R. H. Organometallics 2012, 31,

7375−7385.

P. B. J. Am. Chem. Soc. 1996, 118, 5047−5055.

(71) Noyori, R.; Ohkuma, T. Angew. Chem., Int. Ed. 2001, 40, 40−73.

(72) Abdur-Rashid, K.; Faatz, M.; Lough, A. J.; Morris, R. H. J. Am.

Chem. Soc. 2001, 123, 7473−7474.

(37) Zuo, W.; Li, Y.; Lough, A. J.; Morris, R. H. Science 2013, 342,

1080−1083.

(38) Mikhailine, A. A.; Maishan, M. I.; Morris, R. H. Org. Lett. 2012,

14, 4638−4641.

(73) Clapham, S. E.; Hadzovic, A.; Morris, R. H. Coord. Chem. Rev.

2004, 248, 2201−2237.

(39) Langer, R.; Leitus, G.; Ben-David, Y.; Milstein, D. Angew. Chem.,

Int. Ed. 2011, 50, 2120−2124.

(74) Brookhart, M.; Nelson, G. O. J. Organomet. Chem. 1979, 164,

193−202.

(40) Langer, R.; Iron, M. A.; Konstantinovski, L.; Diskin-Posner, Y.;

Leitus, G.; Ben-David, Y.; Milstein, D. Chem.Eur. J. 2012, 18, 7196−

7209.

(75) Moulton, B. E.; Duhme-Klair, A. K.; Fairlamb, I. J. S.; Lynam, J.

M.; Whitwood, A. C. Organometallics 2007, 26, 6354−6365.

(76) Knolker, H.-J. Chem. Rev. 2000, 100, 2941−2962.

̈

(41) Yang, X. Inorg. Chem. 2011, 50, 12836−12843.

(42) Gunanathan, C.; Ben-David, Y.; Milstein, D. Science 2007, 317,

790−792.

(43) Gunanathan, C.; Milstein, D. Acc. Chem. Res. 2011, 44, 588−

602.

(77) Russell, S. K.; Milsmann, C.; Lobkovsky, E.; Weyhermuller, T.;

Chirik, P. J. Inorg. Chem. 2011, 50, 3159−3169.

(78) Luo, L.; Nolan, S. P. Organometallics 1992, 11, 3483−3486.

(79) Ares, J. R.; Aguey-Zinsou, K. F.; Porcu, M.; Sykes, J. M.;

Dornheim, M.; Klassen, T.; Bormann, R. Mater. Res. Bull. 2008, 43,

1263−1275.

(44) Bichler, B.; Holzhacker, C.; Stoger, B.; Puchberger, M.; Veiros,

̈

L. F.; Kirchner, K. Organometallics 2013, 32, 4114−4121.

(80) Luo, B.; Kucera, B. E.; Gladfelter, W. L. Dalton Trans. 2006, 0,

4491−4498.

(81) Clapham, S. E.; Guo, R.; Iuliis, M. Z.-D.; Rasool, N.; Lough, A.;

Morris, R. H. Organometallics 2006, 25, 5477−5486.

(82) Baratta, W.; Ballico, M.; Esposito, G.; Rigo, P. Chem.Eur. J.

2008, 14, 5588−5595.

(83) Zhang, J.; Gandelman, M.; Shimon, L. J. W.; Rozenberg, H.;

Milstein, D. Organometallics 2004, 23, 4026−4033.

(84) Kloek, S. M.; Heinekey, D. M.; Goldberg, K. I. Organometallics

2006, 25, 3007−3011.

(85) Trovitch, R. J.; Lobkovsky, E.; Chirik, P. J. Inorg. Chem. 2006,

45, 7252−7260.

(86) Mikhailine, A. A.; Lagaditis, P. O.; Sues, P. E.; Lough, A. J.;

Morris, R. H. J. Organomet. Chem. 2010, 695, 1824−1830.

(45) Knolker, H.-J.; Heber, J. Synlett 1993, 1993, 924−926.

̈

(46) Knolker, H.-J.; Heber, J.; Mahler, C. H. Synlett 1992, 1992,

̈

1002−1004.

(47) Casey, C. P.; Guan, H. J. Am. Chem. Soc. 2007, 129, 5816−5817.

(48) Casey, C. P.; Guan, H. J. Am. Chem. Soc. 2009, 131, 2499−2507.

(49) Berkessel, A.; Reichau, S.; von der Hoh, A.; Leconte, N.;

̈

Neudorfl, J.-M. Organometallics 2011, 30, 3880−3887.

̈

(50) Zhou, S.; Fleischer, S.; Junge, K.; Beller, M. Angew. Chem., Int.

Ed. 2011, 50, 5120−5124.

(51) Fleischer, S.; Werkmeister, S.; Zhou, S.; Junge, K.; Beller, M.

Chem.Eur. J. 2012, 18, 9005−9010.

(52) Lagaditis, P. O.; Mikhailine, A. A.; Lough, A. J.; Morris, R. H.

Inorg. Chem. 2010, 49, 1094−1102.

M

dx.doi.org/10.1021/ja4082233 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

(87) Turrell, P. J.; Wright, J. A.; Peck, J. N. T.; Oganesyan, V. S.;

Pickett, C. J. Angew. Chem., Int. Ed. 2010, 49, 7508−7511.

́

(88) Marco, A.; Compano, R.; Rubio, R.; Casals, I. Microchim. Acta

2003, 142, 13−19.

(89) Frisch, M. J. et al. Gaussian 2009, Revision B.2001; Gaussian

Inc.: Wallingford CT, 2009.

(90) Zhao, Y.; Truhlar, D. Theor. Chem. Acc. 2008, 120, 215−241.

(91) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157−167.

(92) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B

2009, 113, 6378−6396.

(93) GAMESS, The General Atomic and Molecular Electronic

Structure System; http://www.msg.chem.iastate.edu/GAMESS/

GAMESS.html.

(94) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.;

Gordon, M. S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.;

Su, S. J.; Windus, T. L.; Dupuis, M.; Montgomery, J. A. J. Comput.

Chem. 1993, 14, 1347.

(95) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270. 284. 299.

N

dx.doi.org/10.1021/ja4082233 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article Doi

Iron(II) complexes containing unsymmetrical P-N-P′ pincer ligands for the catalytic asymmetric hydrogenation of ketones and imines

DOI: 10.1021/ja4082233

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ja4082233

Full text of DOI:10.1021/ja4082233

Products guided by the article

R&D Labs maybe for 1542826-37-8

Relevant to this article

Hot Product