Dehydrogenation of n-alkanes catalyzed by iridium 'pincer' complexes: Regioselective formation of α-olefins

Full text of DOI:10.1021/ja983460p

4086
J. Am. Chem. Soc. 1999, 121, 4086-4087
based complexes were discovered that could catalyze the transfer
of hydrogen from alkanes to sacrificial olefinic hydrogen-
acceptors (eq 1).8,9 Unfortunately, yields were severely limited
Dehydrogenation of n-Alkanes Catalyzed by Iridium
“Pincer” Complexes: Regioselective Formation of
r-Olefins
Fuchen Liu,† Esther B. Pak,‡ Bharat Singh,†
Craig M. Jensen,*,‡ and Alan S. Goldman*,†
(1)
Department of Chemistry, Rutgers
The State UniVersity of New Jersey
Piscataway, New Jersey 08854
by ligand degradation. Nonetheless, unusual and potentially very
valuable kinetic regioselectivity was observed including, in some
cases, dehydrogenation with preference for the less substituted
sites of the alkane, consistent with the selectivity of the C-H
oxidative additions noted above.8,9 In one case dehydrogenation
of an n-alkane (n-hexane) was reported to give selectivity for the
corresponding 1-alkene; however, after reaching only a minute
level (78% of 0.48 mM total olefin), the concentration of 1-hexene
decreased due to isomerization.8
Department of Chemistry, UniVersity of Hawaii
Honolulu, Hawaii 96822
ReceiVed September 29, 1998
The development of methods for the functionalization of
alkanes is of cardinal importance in catalytic chemistry. A specific
functionalization of particularly great potential value is the
conversion of n-alkanes to the corresponding 1-alkenes (R-olefins)
since these serve as precursors for a wide range of commodity-
scale chemicals (>2 × 109 kg/yr).1,2 Such a conversion is also
an intriguing challenge as viewed from a fundamental perspective.
n-Alkanes are the simplest organic molecules with the potential
to undergo regioselective transformations; R-olefins are the
thermodynamically least stable of the corresponding double-bond
isomers and any mechanism for their formation must presumably
involve activation of the strongest bond (primary C-H) in the
molecule. Herein we report the first system to efficiently catalyze
the dehydrogenation of n-alkanes to give R-olefins. Indeed, to
our knowledge this is the first system reported to thermochemi-
cally catalyze any functionalization of the terminal position of
n-alkanes with high efficiency and regioselectivity.3,4
Recently, an efficient catalyst for cycloalkane transfer-dehy-
drogenation was reported: the “pincer” complex (t-BuPCP)IrH2
(1) (t-BuPCP ) 2,6-bis[di(t-butyl)phosphinomethyl]phenyl).10
Initial results with n-alkanes suggested that transfer-dehydroge-
nation catalyzed by 1 did not give significant yields of the
corresponding R-olefins.11 However, we report herein that when
the analogous (i-PrPCP)IrH2 (2) (i-PrPCP ) 2,6-bis[di(i-propyl)-
phosphinomethyl]phenyl) complex was used,12,13 it was apparent
that the major kinetic product is the R-olefin. Yields of R-olefin
much greater than those from any previously reported system can
be obtained, although subsequent isomerization leads ultimately
to the formation of internal olefins. Reexamination of catalysis
using 1 reveals qualitatively similar results although R-olefin
yields are more severely limited by isomerization under typical
conditions.
In the early 1980s examples were discovered of oxidative
addition of C-H bonds to late-metal systems.5 Perhaps the most
remarkable and potentially valuable aspect of this chemistry was
the regioselectivity, which favored reaction of the stronger C-H
bonds (possibly due largely to steric effects): CH4 > 1 > 2 .
3°.6,7 Within the same time frame, soluble late-transition-metal
† Rutgers University.
‡ University of Hawaii.
(1) Lappin, G. R.; Nemec, L. H.; Sauer, J. D.; Wagner, J. D. In Kirk-
Othmer Encyclopedia of Chemical Technology, 4th ed.; Kroschwitz, J. I.,
Howe-Grant, M., Eds.; Wiley-Interscience: New York, 1996; Vol. 17; pp
839-858.
Table 1 shows representative results of the transfer-dehydro-
genation of n-octane catalyzed by 1 and 2 (150 ° C; 1.0 mM
catalyst in n-octane solution in all cases) using various sacrificial
acceptors.14 It can be seen that in several cases 1-octene initially
constitutes g90% of the octene product though the fractions
decrease with time due to olefin isomerization. The combined
fractions of 1- and 2-octenes remains at >95% of total long after
1-octene is no longer the major product; apparently isomerization
to 2-octenes is much more rapid than subsequent isomerization
to 3- or 4-octenes.
(2) Behr, A. In Ullmann’s Encyclopedia of Industrial Chemistry, 5th ed.;
Elvers, B., Hawkins, S., Russey, W., Eds.; VCH Verlagsgesellschaft:
Weinheim, 1989; Vol. A13; pp 240-251.
(3) Despite the much greater bond strengths of primary vs secondary C-H
bonds of n-alkanes, catalysts have been reported capable of oxidation of the
former. However, to our knowledge, selectivity for primary oxidation is at
best only approximately 1:1 on a per bond basis, in contrast with that of the
present work (> 20:1). For some lead references see: (a) Sen, A. Acc. Chem.
Res. 1998, 31, 550-557. (b) Shilov, A. E.; Shul’pin, G. B. Chem. ReV. 1997,
97, 2879-2932. (c) Metalloporphyrins in Catalytic Oxidations; Sheldon, R.
A., Ed.; Marcel Dekker: New York, 1994.
(4) A photochemical system first reported by Tanaka has been modified
to give efficient and regioselective carbonylation of the terminal position of
n-alkanes: (a) Sakakura, T.; Sodeyama, T.; Sasaki, K.; Wada, K.; Tanaka,
M. J. Am. Chem. Soc. 1990, 112, 7221-7229. (b) Rosini, G. P.; Zhu, K.;
Goldman, A. S. J. Organomet. Chem. 1995, 504, 115-121.
(5) (a) Janowicz, A. H.; Bergman, R. G. J. Am. Chem. Soc. 1982, 104,
352-354. (b) Hoyano, J. K.; Graham, W. A. G. J. Am. Chem. Soc. 1982,
104, 3723.
In comparing results obtained using catalyst 2 under varying
conditions it appears that the overall reaction rate is fairly
insensitive to the nature of the olefinic hydrogen-acceptor.
(9) (a) Burk, M. J.; Crabtree, R. H.; Parnell, C. P.; Uriarte, R. J.
Organometallics 1984, 3, 816-817. (b) Burk, M. J.; Crabtree, R. H. J. Am.
Chem. Soc. 1987, 109, 8025-8032.
(10) (a) Gupta, M.; Hagen, C.; Flesher, R. J.; Kaska, W. C.; Jensen, C. M.
Chem. Commun. 1996, 2083-2084. (b) Gupta, M.; Hagen, C.; Kaska, W. C.;
Cramer, R. E.; Jensen, C. M. J. Am. Chem. Soc. 1997, 119, 840-841.
(11) Hagen, C.; Gupta, M.; Kaska, W. C.; Jensen, C. M. Paper INOR 221,
presented at the 213th American Chemical Society National Meeting, San
Francisco, CA, April 13, 1997.
(6) (a) Arndtsen, B. A.; Bergman, R. G.; Mobley, T. A.; Peterson, T. H.
Acc. Chem. Res. 1995, 28, 154-162. (b) Jones, W. D.; Feher, F. J. Acc. Chem.
Res. 1989, 22, 91-100. (c) Harper, T. G. P.; Desrosiers, P. J.; Flood, T. C.
Organometallics 1990, 9, 2523-2528.
(7) For excellent discussions of the kinetic and thermodynamic selectivity
exhibited by transition metal complexes toward C-H bonds, particularly in
terms of factors other than steric, see the following and references therein:
(a) Bennett, J. L.; Vaid, T. P.; Wolczanski, P. T. Inorg. Chim. Acta 1998,
270(1-2), 414-423 (b) Wick, D. D.; Jones, W. D. Organometallics 1999,
18, 495-505.
(12) The protonated i-PrPCP ligand was synthesized according to Milstein13
and reacted with [Ir(cyclooctene)2Cl] in refluxing toluene for 3 days. The
resulting (i-PrPCP)IrHCl was isolated and converted to (i-PrPCP)IrH2, using
the procedure previously reported for (t-BuPCP)IrH2.10
(13) Rybtchinski, B.; Vigalok, A.; Bendavid, Y.; Milstein, D. Organome-
tallics 1997, 16, 3786-3793.
(14) Concentrations were determined by GC (see ref 10). In all cases the
measured concentration of hydrogenated sacrificial acceptor was within 20%
of ([octenes] + 2 × [dienes]).
(8) (a) Baudry, D.; Ephritikine, M.; Felkin, H.; Holmes-Smith, R. J. Chem.
Soc., Chem. Commun. 1983, 788-789. (b) Felkin, H.; Fillebeen-Khan, T.;
Holmes-Smith, R.; Lin, Y. Tetrahedron Lett. 1985, 26, 1999-2000.
10.1021/ja983460p CCC: $18.00 © 1999 American Chemical Society
Published on Web 04/28/1999

Communications to the Editor
J. Am. Chem. Soc., Vol. 121, No. 16, 1999 4087
Table 1. Octene Distributions (Concentrations in mM) from
Transfer-Dehydrogenation of n-Octane Catalyzed by 1 or 2 Using
Various Hydrogen Acceptors. All Runs Conducted at 150 °C with
1.0 mM Catalyst in n-Octane Solution
cata
accep
min 1-oct trans-2 cis-2 other total
%
2
nbeb
5
10
30
60
5
10
30
60
5
10
30
60
10
30
60
6
15
30
60
15
30
60
15
30
60
90
120
11
23
40
6
0.5
4
0.6
3
0
0
12
30
91
0.2 M
76
30
45
43
3
132
208
8
82
0
2
40
0
1
63
0
3
2
2
nbe
0.5 M
8
>90
19
59
59
21
27
44
41
10
43
10
10
18
20
18
23
27
30
13
34
74
94
97
0
22
87
56
105
3
40
71
3
0
154
238
27
38
3
25
tbec
0.5 M
0
78
6
6
0
40
68
65
103
0
45
78
0
1
155
250
10
28
Figure 1. Proposed mechanism of both transfer-dehydrogenation of
n-octane and octene-isomerization catalyzed by the (PCP)Ir catalysts.
19
0
16
2
1
1-dec
0.5 M
>90
31
64
0
21
40
0
0
95
45
obtained with nbe versus tbe may be due to a greater ability of
nbe to “trap” (PCP)IrH2. This might seem to be in contradiction
with the observation that the reaction with tbe (0.5 M) is no slower
than the reaction with nbe (0.5 M, Table 1). However, the rate-
determining step of the catalysis is presumably not the reaction
of olefin with dihydride (cf. the approximately zero-order kinetics
in [nbe]); therefore, no such contradiction exists. Accordingly,
when a 1:1 mixture of nbe:tbe is used as acceptor, the rate of
nbe hydrogenation is approximately double that of tbe hydrogena-
tion; i.e., nbe is apparently the more effective trap for (PCP)-
IrH2.
13
0
134
10
8
tbe
0.2 M
>90
19
41
47
4
8
0
45
40
20
26
2
0
81
25
0
91
20
1
1
nbe
0.5 M
0
29
79
7
3
0
37
73
15
0
5
0
50
60
1-decd
0.5 M
0
0
13
>95
1
0.6
0
36
95
7
4
9
14
0
86
87
9
0
111
143
84
32
0
68
The t-BuPCP complex 1 gives lower reaction rates than the
i-PrPCP analogue 2 and lower fractions of 1-octene when nbe or
tbe is used as the acceptor. The poorer apparent regioselectivity
is surprising as the bulkier t-butyl groups would be expected to
enhance regioselectivity for the terminal position. However, the
results are consistent with Figure 1 and the proposal that the
1-octene percentage reflects a competition between transfer-
hydrogenation and isomerization, i.e., a competition between
acceptor and 1-octene for reaction with dihydride. The bulkier
t-BuPCP complex should be more selective for 1-octene versus
either tbe or nbe, ultimately resulting in lower fractions of 1-octene
even if the kinetic regioselectivity of dehydrogenation is equal
to or greater than that of the i-PrPCP complex. In accord with
this reasoning, when the acceptor used is 1-decene (0.5 M) (which
would react with (PCP)IrH2 as efficiently as 1-octene), we find
that the t-BuPCP complex gives greater selectivity than the i-PrPCP
complex (Table 1); indeed it gives the best regioselectivity and
1-octene yield (97 turnovers) of any catalyst/acceptor system we
have thus far investigated.15
In summary, we report the first catalytic system for the efficient
and selective dehydrogenation of linear alkanes to give R-olefins.
Secondary isomerization limits the yield of R-olefin obtained
under our reaction conditions; however, our results suggest that
modification of the catalyst/acceptor combination or efficient
separation of R-olefin to prevent isomerization may permit high
yields. The oxidative addition of carbon-hydrogen bonds was
first reported 16 years ago and has since held the attention of
organometallic and organic chemists. This system may represent
the closest example to date of a practical catalysis which exploits
that intriguing mode of activation for the functionalization of
alkanes.
a cat ) catalyst; accep ) acceptor; 1-oct ) 1-octene; other ) (other
octenes + 2 × [dienes]); % ) 100 × [1-octene]/total. b Norbornene.
c t-Butylethene. d 1-Decene.
Furthermore, runs conducted with 0.5 and 0.2 M norbornene (nbe),
respectively, gave virtually identical rates of transfer-dehydro-
genation. Thus, neither the nature nor the concentration of the
hydrogen-acceptor greatly affects the reaction rate. Surprisingly,
however, both of these factors clearly and reproducibly affect
the obserVed distribution of double-bond isomers produced (i.e.,
the observed regioselectivity) and in particular the percentage of
R-olefin product. For example, when 0.5 M nbe is the acceptor,
after 30 min thermolysis, the concentration of 1-octene is 59 mM
which constitutes 38% of the total octene. When 0.2 M nbe is
used, under identical conditions and after the same time the
concentration of 1-octene is 40 mM, representing 30% of the total
octenes. After 60 min the difference is even more pronounced:
1-octene is present as 25 and 3% of the product in the reactions
using 0.5 and 0.2 M nbe, respectively. When t-butylethene (tbe,
0.5 M) is used as the acceptor, the observed selectivity for
1-octene is significantly lower than that found with an equal
concentration of nbe.
It would appear that in all cases the kinetic regioselectivity
for R-olefin formation is high and the observed isomer distribution
is determined by the relative rates of isomerization and dehydro-
genation. Figure 1 shows a possible mechanism (in accord with
previously proposed mechanisms9) for the catalysis of both
transfer-dehydrogenation and isomerization. The mechanism
involves essentially only two steps and their microscopic re-
verse: (i) oxidative addition of an alkane C-H bond (with high
selectivity for the terminal H), and (ii) â-hydrogen elimination.
Despite its simplicity, the mechanism may be used to account
for a number of observations, including the dependence of isomer
distribution on nature and concentration of acceptor. We propose
that the observed isomer distribution is largely determined by the
competition between acceptor and 1-octene for insertion into the
Ir-H bond of (PCP)IrH2. Thus, the fraction of 1-octene product
will correlate with both the concentration and the intrinsic
reactivity of the acceptor. For example, the better regioselectivity
Acknowledgment. Support for this research by the Division of
Chemical Sciences, BES, OER, U.S. Department of Energy and the U.S.
Department of Energy Hydrogen Program is gratefully acknowledged.
We thank Johnson-Matthey for a generous loan of iridium. Mira
Kanzelberger is thanked for generous experimental assistance.
Supporting Information Available: Detailed table of octene distribu-
tions for experiments outlined in Table 1, including additional times and
concentrations of dienes and trans-3-octene (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
(15) The attempted use of ethene as acceptor resulted in complete inhibition
of catalytic activity by either 1 or 2, apparently due to the formation of a
stable complex (RPCP)Ir(C2H4).
JA983460P