A Fortuitous, Mild Catalytic Carbon-Carbon Bond Hydrogenolysis by a Phosphine-Free Catalyst

Article abstract of DOI:10.1071/CH15792

The putative catalyst trans-[Ru((S,S)-skewphos)(H)₂((R,R)-dpen)] (skewphos=2,4-bis(diphenylphosphino)pentane; dpen=1,2-diphenylethylenediamine) transforms the trifluoroacetyl amide 2,2,2-trifluoro-1-(piperidin-1-yl)ethanone under mild conditions (4 atm H₂, room temperature, 4-24h, 1 mol-% Ru, 15 mol-% KO^tBu in tetrahydrofuran) to generate the formylated amine 1-formylpiperidine and fluoroform via C-C bond hydrogenolysis. Catalysts are also prepared by reacting cis-[Ru(η³-C₃H₅)(MeCN)₂(COD)]BF₄ (COD=1,5-cyclooctadiene) with diamine ligands in situ. Lowerature NMR studies provided insight into this reaction.

Full text of DOI:10.1071/CH15792

RESEARCH FRONT
CSIRO PUBLISHING
Aust. J. Chem.
Full Paper
http://dx.doi.org/10.1071/CH15792
A Fortuitous, Mild Catalytic Carbon–Carbon Bond
Hydrogenolysis by a Phosphine-Free Catalyst
A
A
A
Loorthuraja Rasu, Ben Rennie, Mark Miskolzie,
A B
,
and Steven H. Bergens
A
Department of Chemistry, University of Alberta, 11227 Saskatchewan Drive,
Edmonton, Alberta T6G 2G2, Canada.
B
Corresponding author. Email: sbergens@ualberta.ca
The putative catalyst trans-[Ru((S,S)-skewphos)(H)2((R,R)-dpen)] (skewphos ¼ 2,4-bis(diphenylphosphino)pentane;
dpen ¼ 1,2-diphenylethylenediamine) transforms the trifluoroacetyl amide 2,2,2-trifluoro-1-(piperidin-1-yl)ethanone
under mild conditions (4 atm H2, room temperature, 4–24 h, 1 mol-% Ru, 15 mol-% KOtBu in tetrahydrofuran) to
generate the formylated amine 1-formylpiperidine and fluoroform via C–C bond hydrogenolysis. Catalysts are also
prepared by reacting cis-[Ru(Z3-C3H5)(MeCN)2(COD)]BF4 (COD ¼ 1,5-cyclooctadiene) with diamine ligands in situ.
Low-temperature NMR studies provided insight into this reaction.
Manuscript received: 17 December 2015.
Manuscript accepted: 3 March 2016.
Published online: 6 April 2016.
by reaction of cis-[Ru(Z3-C3H5)(MeCN)2(COD)]BF4 (1; COD
is 1,5-cyclooctadiene) with P–P and N–N ligands.[2,9] Typically,
1 and 1 equivalent of the P–P and N–N ligands are allowed to
react for 30 min at 608C to displace MeCN and COD. The
resulting allylic-Ru complexes are then mixed with KOtBu and
the amide and placed under H2. The putative Ru-dihydride
catalyst is formed in situ for hydrogenation to proceed. For this
study, we chose Bosnich’s ligand (S,S)-2,4-bis(diphenylpho-
sphino)pentane ((S,S)-skewphos)[10] as the diphosphine and
(R,R)-1,2-diphenylethylenediamine ((R,R)-dpen) as the diamine.
We found that the catalytic reaction between the trifluoroacetyl
amide 2,2,2-trifluoro-1-(piperidin-1-yl)ethanone (2) and 1 mol-%
of trans-Ru((S,S)-skewphos)(H)2((R,R)-dpen) (3) proceeded in
100 % yield after 4 h under mild conditions (15 mol-% KOtBu,
4 atm H2, THF, room temperature (rt)). To our surprise, rather than
the expected amide hydrogenation products, 1-formylpiperidine
was the sole detectable (NMR) product of the reaction (Scheme 1).
The related acyclic amide 2,2,2-trifluoro-1-(diethylaminyl)
ethanone formed diethylformamide as sole detectable product in
68.7 % conversion.
Introduction
Wereporttheunexpectedhydrogenolysisof ansp3–sp2 C–Cbond
with homogeneous catalysis under mild conditions. Homoge-
neouscatalyticC–Cbond functionalizationisanefficientstrategy
to modify the skeletal structures of organic molecules. Tangible
progress has been reported on such catalytic systems,[1] but
further developments are required to broaden the net applicability
of this methodology. The intrinsic difficulty is to devise a catalyst
that overcomes the inherent kinetic inertness and thermodynamic
stability of a specific C–C bond, while accommodating the
presence of spectator functional groups. We recently reported
the catalytic hydrogenation of amides with the bifunctional
ꢀ
catalyst precursors [Ru(Ph2PCH2CH2NH2)2(Z3-C3H5)]þ(BF4
)
(for hydrogenations under basic conditions)[2] and trans-Ru
(Ph2PCH2CH2NH2)2(H)(BH4) (for hydrogenations under neutral
conditions).[3] These hydrogenations occur with cleavage of
the amide C–N bond to generate the corresponding amine and
primary alcohol products. The reactions proceed with high turn-
over numbers (up to 6700) under moderate conditions (1008C, 50
atm reaction pressure), and they tolerate a propitious variety
of functional groups. We now report the fortuitous discovery
of a catalytic C–C hydrogenolysis that selectively converts a
trifluoroacetyl amide into the corresponding formylated amine.
Formylated amines are used as solvents and they are inter-
mediates in syntheses of pharmaceuticals.[4,5] They are typically
prepared from the parent amine using formic acid or its deriva-
tives as the formylating reagent.[5] There are several reports of
Ru complexes that catalyze the formylation of amines using
CO,[6] methanol,[7] and quite recently, CO2/H2[8] as the source
of the formyl group.
Assuming the putative catalyst in this reaction is trans-Ru
((S,S)-skewphos)(H)2((R,R)-dpen) (3), we prepared the stable
dichloride trans-Ru((S,S)-skewphos)(Cl)2((R,R)-dpen) (4) as a
structural model for the catalyst. Fig. 1 shows the solid-state
structure of 4 as determined by X-ray diffraction. The backbone
of (S,S)-skewphos is in the chair conformation with one methyl
group in the axial orientation and the other group in the
O
H2 (4 atm)
O
KOtBu (15 %)
N
N
CF3
H
3 (1 %)
Results and Discussion
THF (rt)
2
We showed previously that Ru-dihydride catalysts of the type
trans-Ru(P–P)(H)2(N–N) and related complexes are prepared
Scheme 1.
Journal compilation Ó CSIRO 2016
www.publish.csiro.au/journals/ajc

B
L. Rasu et al.
C34
C33
C35
C23
C24
C32
C36
Cl2
C31
C22
C46
C45
C44
C56
C55
C65
C54
C25
C66
C61
C21
C41
P2
N1
C26
C6
C51
C52
P1
RU
N2
C43
C53
C3
C42
Cl1
C7
C12
C2
C64
C5
C4
C1
C62
C11
C63
C13
C16
C14
C15
Fig. 1. Solid-state structure of trans-Ru((S,S)-skewphos)(Cl)2((R,R)-dpen) (4) as determined by X-ray diffraction.
The positions of the hydrogen atoms are based upon idealized geometries.
equatorial orientation. This conformation holds the phenyl rings
on phosphorus in a pseudo achiral spatial array. Perhaps the
exigencies of crystal packing prevented the ligand from adopt-
ing the more stable d-skew conformation which would place
both methyl groups in the equatorial position and the phenyl
rings in a chiral disposition.[10] As expected, the backbone of (R,
R)-dpen is puckered in the l-conformation with both phenyl
rings equatorially disposed.
Me
C
N
BF4
Ph
Me
H2 (~1 atm)
Ph
C
NH2
NH2
Ph
Ph
NH
KOtBu (8 equiv.)
N
2
H2N
Ru
ϩ
Ru
THF, Ϫ80 to Ϫ40ЊC
1
6
Scheme 2.
Remarkably, (S,S)-skewphos could be omitted, and another
equivalent of (R,R)-dpen was used during the reaction with 1 to
generate the phosphine-free catalyst 5; the latter also catalyzed
the conversion of 2 into the formyl amine in 100 % yield after
23 h under the same mild conditions. The corresponding ethy-
lenediamine catalyst, made by reacting 1 with 2 equivalents of
ethylenediamine, was also active, however, with only 11 %
conversion after 15 h.
Ph
H2 (~1 atm)
Ph
H2N
H2
Ph
NH
Ru
H
KOtBu (8 equiv.)
N
Ru
THF, Ϫ40 to Ϫ0ЊC
N
Ph
H2
Ϫ
6
7 (tenative)
Scheme 3.
The reaction between the bis(acetonitrile) precursor 1 with 2
equivalents of (R,R)-dpen and 8 equivalents of KOtBu under ,1
atm H2 in [D8]THF at ꢀ788C was monitored by NMR spectros-
copy to investigate the reaction. At ꢀ788C, the MeCN ligands
were displaced by 1 equivalent of (R,R)-dpen to form the Z3-
C3H5-amido complex 6, with one NH2 group in (R,R)-dpen
deprotonated by the excess KOtBu (Scheme 2). The compound
was fully characterized in solution by 1H NMR, 1H–1H gCOSY,
resulted in the hydrogenation of propylene to propane, the
hydrogenation of the acetonitrile to ethylamine, and the forma-
tion of unidentified ruthenium-containing products. All
attempts to identify the ruthenium-containing products formed
by warming the mixture of 7 and (R,R)-dpen failed. Addition of
the trifluoroacetyl amide 2 to this mixture, however, resulted in
the formation of fluoroform, CHF3, as the major fluoride-
containing product (identified by 19F NMR and by a headspace
analysis by gas chromatography mass spectrometry). Thus, the
net catalytic reaction between 2,2,2-trifluoro-1-(piperidin-1-yl)
ethanone (2) and the catalysts 3 and 5 is the hydrogenolysis of
the C–C bond to form 1-formylpiperidine and CHF3 (Scheme 4).
While investigating the nature of 5, we found that elemental
mercury (200 equiv.) did not inhibit the hydrogenolysis cata-
lyzed by 5, and the activity of reduced Ru black was minimal
(,1 % conversion) under these conditions. Taken together,
these results suggest that 5 is a homogeneous Ru-H-dpen
compound.
1
1H–13C gHSQC, H–15N gHSQC and TOSCY, and ROESY
NMR experiments.
1
Notably, the H NMR chemical shift for the amido NꢀH
group was ꢀ1.37 ppm. The shifts for the NH2 protons were 4.87
and 5.13 ppm. Compound 6 loses propylene upon slow warming
to approximately ꢀ208C. We tentatively propose that the transi-
tion metal product is the hydride 7, with the C8 ring bonded to Ru
through an olefin and a Z3-allyl group (Scheme 3).[9]
Overlap of peaks prevented conclusive identification of
compound 7. The proposed structure would result from allylic
activation of the 1,5-cyclooctadiene ring in 6, followed by
elimination of propylene. Further warming to room temperature

A Fortuitous, Catalytic Carbon–Carbon Hydrogenolysis
C
H2 (4 atm)
O
O
The percentage conversions were determined by 1H NMR
spectroscopy. The catalyst was removed by passing the solution
through a Florisil plug using CH2Cl2 as the rinsing solvent. The
solvent was then removed under reduced pressure using a rotary
evaporator, and the NMR spectra were recorded using CDCl3.
KOtBu (15 %)
CHF3
ϩ
CF3
N
H
N
3 or 5 (1 %)
THF (rt)
2
Scheme 4.
NMR Study of the Reaction between 1, ((R,R)-dpen),
KOtBu, H2, and 2
Conclusion
Cis-[Ru(Z3-C3H5)(COD)(MeCN)2]BF4 (1) (0.03 mmol, 12.6 mg),
(R,R)-dpen (0.06 mmol, 12.7 mg), and KOtBu (0.24 mmol,
26.9 mg) were weighed into three separate NMR tubes inside the
glove box. Distilled [D8]THF (0.5 mL) was added to (R,R)-dpen
by cannula under argon, and the 1H NMR spectrum was
recorded. Then, the solution was transferred to the NMR tube
containing 1 under argon. The resulting solution was then heated
at 608C for 30 min with occasional shaking. After 30 min, the 1H
and 19F NMR spectra were recorded. Then, the solution
temperature was decreased to ꢀ808C by immersing the NMR
tube in dry ice/acetone mixture. The base in [D8]THF (0.2 mL)
was added to Ru-dpen mixture under hydrogen at ꢀ808C. The
product was characterized using 1H, 19F, 1H–15N HSQC,
gTOCSY, gCOSY, and gROESY NMR experiments in [D8]
THF at variable temperatures (ꢀ808C to room temperature).
The cis-[Ru(Z3-C3H5)(COD)(MeCN)2]BF4/(R,R)-dpen/ KOtBu
precatalyst was prepared again at ꢀ808C as described above and
10 equiv. 2,2,2-trifluoro-1-(piperidin-1-yl)ethanone (2) were
added at ꢀ608C. The reaction was monitored from ꢀ608C to
room temperature.
To our knowledge, the catalytic transformation of 2 into 1-for-
mylpiperidine and fluoroform is the first catalytic hydro-
genolysis of the sp3–sp2 C–C bond in a trifluoro amide. The mild
conditions, along with the absence of phosphine ligands
are attractive features of this novel transformation. Efforts are
underway in our laboratories to investigate the scope, the
identity of the catalysts, and the mechanism of this reaction.
Experimental
General Information
Deuterated solvents were obtained from Aldrich and Cambridge
Isotope Laboratories. Both [D8]THF and THF were dried using
Na/benzophenone just before each experiment. All pressurized
reactionswerecarried outina steel pressure reactor equipped with
a magnetic stir bar. Potassium tert-butoxide was sublimed before
use.
Piperidine, potassium tert-butoxide, and trifluoroacetic
anhydride were obtained from Aldrich and (1R, 2R)-1,2-diphe-
nylethylenediamine was obtained from Alfa Aesar. 1H, 13C, and
19F NMR spectra were recorded using 400 and 600 MHz Varian
Inova, and 500 and 700 MHz Varian DirectDrive spectrometers.
1H and 13C NMR chemical shifts (d) are reported in parts per
million (ppm) relative to TMS with the deuterated solvent as the
internal reference. 19F chemical shifts are reported in parts per
million relative to CFCl3 as the external reference. NMR peak
assignments were made using 1H NMR, 1H–1H gCOSY, 1H–13C
gHSQC, 1H–15N gHSQC and TOSCY, and ROESY NMR
experiments. Gas chromatography mass spectrometry analysis
was performed by using a Hewlett Packard 5890 chromatograph
equipped with a 5970B mass-selective detector and Supelco
Beta DEX 225 capillary column (30 m ꢁ 0.25 mm ꢁ 0.25 mm
film thickness). Elemental analysis data were obtained using a
Carlo Erba CHNS-O EA1108 elemental analyzer.
Supplementary Material
Additional experimental and spectroscopic details, as well as
structural data and parameters from the X-ray crystallography
study are available on the Journal’s website.
Acknowledgements
This work was supported by the Natural Sciences and Engineering Research
Council of Canada (NSERC), GreenCentre Canada, and the University of
Alberta. We sincerely appreciate the assistance of the University of Alberta
High Field NMR Laboratory.
References
Hydrogenation Procedure
[1] (a) Topics in Current Chemistry 346, C–C Bond Activation (Ed.
G. Dong) 2014 (Springer: Heidelberg).
Cis-[Ru(Z3-C3H5)(COD)(MeCN)2]BF4 (1; 0.025 mmol, 10.5 mg),
2 equiv. (R,R)-dpen (0.05 mmol, 11.0 mg) (or 1 equiv. (R,R)-
dpen and 1 equiv. of (S,S)-skewphos), and KOtBu (0.375 mmol,
42.1 mg) were weighed out into three NMR tubes in a glove box.
Freshly distilled THF (0.8 mL) was then added by cannula under
argon pressure into the tube containing (R,R)-dpen. The resul-
tant solution was then cannulated into the NMR tube containing
1 under argon. It was then heated at 608C for 30 min with
occasional shaking (pale brown, clear liquid). Meanwhile, 100
equiv. substrate (2.5 mmol, 453 mg) was dissolved in THF
(1.0 mL) in a NMR tube, and the resultant solution was then
cannulated into a stainless steel autoclave under H2 pressure and
purged with hydrogen for 20 min. After 30 min, KOtBu in THF
(0.7 mL) was added to the pre-heated NMR tube. The resultant
orange-red solution was transferred to the autoclave under
hydrogen pressure followed by washing with THF (2.5 mL). The
autoclave was then pressurized to 4 atm H2 and stirred at room
temperature for 4–24 h. The reaction was stopped by slowly
de-pressurizing the autoclave and opening it to the atmosphere.
(b) K. Ruhland, Eur. J. Org. Chem. 2012, 2683. doi:10.1002/EJOC.
201101616
(c) B. L. Edelbach, D. A. Vicic, R. J. Lachicotte, W. D. Jones,
Organometallics 1998, 17, 4784. doi:10.1021/OM9805281
[2] J. M. John, S. H. Bergens, Angew. Chem. 2011, 50, 10377.
doi:10.1002/ANIE.201103137
[3] J. M. John, R. Loorthuraja, E. Antoniuk, S. H. Bergens, Catal. Sci.
Technol. 2015, 5, 1181. doi:10.1039/C4CY01227E
[4] A. Jackson, O. Meth-Cohn, J. Chem. Soc., Chem. Commun. 1995,
1319. doi:10.1039/C39950001319
[5] C. J. Gerack, L. McElwee-White, Molecules 2014, 19, 7689.
doi:10.3390/MOLECULES19067689
[6] (a) G. Jenner, G. Bitsi, Appl. Catal. 1987, 32, 293. doi:10.1016/S0166-
9834(00)80632-6
(b) G. Bitsi, G. Jenner, J. Organomet. Chem. 1987, 330, 429.
doi:10.1016/S0022-328X(00)99055-7
(c) J. J. Byerley, G. L. Rempel, N. Takebe, B. R. James, J. Chem. Soc.
D 1971, 1482. doi:10.1039/C29710001482
(d) Y. Tsuji, T. Ohsumi, T. Kondo, Y. Watanbe, J. Organomet. Chem.
1986, 309, 333. doi:10.1016/S0022-328X(00)99635-9

D
L. Rasu et al.
[7] N. Ortega, C. Richter, F. Glorius, Org. Lett. 2013, 15, 1776.
bonded to Ru as an Z3-allyl. See: J. A. Wiles, C. J. A. Daley,
R. J. Hamilton, C. G. Leong, S. H. Bergens, Organometallics 2004, 23,
4564. doi:10.1021/JA8034812
doi:10.1021/OL400639M
[8] L. Zhang, Z. Han, X. Zhao, Z. Wang, K. Ding, Angew. Chem., Int. Ed.
2015, 54, 6186. doi:10.1002/ANIE.201500939
[10] P. A. MacNeil, N. K. Roberts, B. Bosnich, J. Am. Chem. Soc. 1981,
[9] A related pathway is allylic activation of the COD ring followed by the
elimination of propylene. The resulting C8H11 fragment is in part
103, 2273. doi:10.1021/JA00399A023