A Tertiary Phosphonium Salt as a Promoter for the Hydrogenation of CO

Full text of DOI:10.1002/cctc.201200176

DOI: 10.1002/cctc.201200176
A Tertiary Phosphonium Salt as a Promoter for the Hydrogenation of CO
[
a]
[b]
[b]
[a]
Jan H. Blank, Robert Hembre, James Ponasik, and David J. Cole-Hamilton*
Small amounts of [HPBu ]Br that are either present as an im-
[Ru (CO) ] as a catalyst precursor at 2008C, but under milder
3 12
3
purity in commercial [Bu P]Br or are added to it promote the
pressures (CO/H , 1:1, 250 bar), MeOH and EtOH are observed
4
2
hydrogenation of CO catalysed by [Ru (CO) ]. [HPBu ]Br may
as the main products, together with smaller amounts of propa-
nol and 1,2-ethanediol. Qualitative analysis of the gas phase
before or after condensing any condensable compounds
shows significant amounts of dimethyl-, methylethyl-, and di-
ethyethers, which we have shown elsewhere are formed from
the acid-catalysed dehydration of the alcohol products. How-
ever, the reproducibility of this system was poor. Marked
changes in activity were observed whenever different batches
(lot number) of tetrabutylphosphonium bromide were em-
ployed. Some batches showed good activity whereas others
gave much-poorer activity. Scrutiny of concentrated samples of
3
12
3
be responsible for the irreproducibility that is sometimes ob-
served in similar CO-hydrogenation reactions.
The homogeneous conversion of synthesis gas into function-
al chemicals was first reported by Gresham in the 1950s. In
these reactions, metal sources were subjected to high temper-
atures and very high pressures of syngas to form alcohols and
[
1]
polyols. One goal in the research of syngas has been to form
CÀC bonds from individual molecules of CO whilst retaining
some or all of the oxygen functionality. The heterogeneous hy-
drogenation of CO tends to promote cleavage of the CO bond
with the formation of alkanes and alkenes (Fischer–Tropsch
31
these different batches of [PBu ]Br by P NMR spectroscopy re-
4
[
2–5]
chemistry)
but the early examples from Gresham showed
vealed three peaks (Figure 1). The strongest signal (d=
that homogeneous ruthenium catalysts are able to provide
oxygenates. Subsequent research by Dombek (Union Carbide)
and Bradley (Exxon) found that the activity could be greatly
enhanced by the addition of halide promoters, preferably
[
6–12]
iodide.
Knifton and co-workers independently found good
yields when using molten tetraalkylphosphonium halide salts
[
13–15]
as solvents instead of the usual organic media.
Whereas
Dombek’s system with N-methylpyrrolidone and iodide salts
was particularly useful for spectroscopic- and mechanistic anal-
[
10,16]
ysis,
Knifton focused on tuning the selectivity towards
[
15,17]
[18]
a wide scope of products, such as MeOH,
EtOH,
1,2-
[23]
[
19–21]
[13,22]
ethanediol,
and acetic acid
and their derivatives.
Following Dombek’s work, Ono et al. found that a remarkable
increase in selectivity towards EtOH could be achieved by
using phosphoric acid or trimethylphosphate as a promot-
[
24,25]
31
1
er.
Although these systems work well, the problem re-
4]
Figure 1. P{ H} NMR spectrum of an active batch of [PBu Br].
Inset: H-coupled signal for the resonance at d=11.55 ppm.
mains that very high pressures and temperatures are required
to achieve reasonable conversions and, therefore, considerable
emphasis must continue to be directed towards elucidating
the mechanisms of all of these processes and towards finding
factors that increase the reaction rates. Herein, we report an in-
teresting promoter that increases the rate of MeOH produc-
tion.
[
26]
33.56 ppm) was from tetrabutylphosphonium bromide.
In
most batches, another peak was present at d=37.49 ppm,
owing to tri-n-butyl(sec-butyl)phosphonium bromide, which
was formed by Markovnikoff addition of the PÀH bond across
1-butene during the synthesis of PBu from PH and 1-butene.
[15]
By using the melt chemistry reported by Knifton et al.
with tetrabutylphosphonium bromide as the solvent and
3
3
This peak was more intense in the less-active batch of [PBu ]Br.
4
31
1
The P{ H} NMR spectrum of the active batch revealed an
[
a] Dr. J. H. Blank, Prof. D. J. Cole-Hamilton
EaStCHEM, School of Chemistry
University of St. Andrews
St. Andrews, Fife, KY16 9ST, Scotland (UK)
Fax: (+44)1334-463808)
additional peak at d=11.55 ppm, which split into a doublet
31
(
(
J(H,P)=487 Hz) in the proton-coupled P NMR spectrum
1
Figure 1). Likewise, careful scrutiny of the H NMR spectrum
(see the Supporting Information, Figure S2) of the same
E-mail: djc@st-and.ac.uk
sample revealed a low-intensity doublet (d=6.83 ppm, J(H,P)=
[
b] Dr. R. Hembre, Dr. J. Ponasik
Eastman Chemical Company
Kingsport, TN 3766 (USA)
4
87 Hz) in addition to the signals from the butyl groups. These
signals were assigned to tributylphosphonium bromide,
HPBu ]Br, which is a protonated form of tributyl phosphine.
[
3
Supporting information for this article, including experimental details
and additional spectra, is available on the WWW under http://dx.doi.org/
This impurity is presumably formed during the synthesis of tet-
rabutylphosphonium bromide when tributylphoshine reacts
10.1002/cctc.201200176.
ChemCatChem 0000, 00, 1 – 4
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

At the end of the reactions that are depicted in Figure 2a,
the colour varied markedly from dark red (no [HPBu ]Br)
3
through orange to yellow ([HPBu ]Br/Ru, 1:1). These final solu-
3
tions were studied by IR spectroscopy. When no [HPBu ]Br was
3
Scheme 1. Process for the production of [Bu
lead to the formation of small amounts of [HPBu
4
P]Br and how this process can
]Br as an impurity.
added (dark-red colour after the reaction, Figure 3a), the IR
3
with free HBr or when n-butylbromide reacts with traces of
di(n-butyl)phosphine (Scheme 1); it is not present in the batch-
es of [PBu ]Br that provide lower hydrogenation activity. The
4
sec-butyl isomer of [Bu P]Br should have little effect on the cat-
4
alysis in the system, but [HPBu ]Br was not so innocent during
3
the catalysis.
Because the batch of [Bu P]Br that contained [HPBu ]Br was
4
3
more active, we purified a sample of the active batch by re-
precipitating it from acetone with Et O. This procedure effec-
2
tively removed almost all of the [HPBu ]Br. We also synthesised
3
pure [HPBu ]Br from tri(n-butyl)phosphine and HBr and used it
3
in a series of reactions in which we increased the amount of
[
HPBu ]Br that was present in [PBu ]Br. The results of this series
3
4
of experiments (Figure 2, empty symbols) clearly demonstrate
Figure 3. IR spectra that were obtained after the hydrogenation of CO cata-
that [HPBu ]Br, in a limited- and low concentration range, acts
lysed by [Ru
3
(CO)12] (0.25 g, 0.39 mmol) in [PBu
4
]Br (15 g) with a) no added
]Br (1.25 mol(mol
3
À1
[
HPBu
À1
3
]Br; b) [HPBu
3
]Br (0.75 mol(mol Ru) ); or c) [HPBu
3
as a promoter for the CO-hydrogenation reaction in this
Ru) ). The complete spectra are shown in the Supporting Information,
Figures S7–S9.
system. The activity towards MeOH increases as the [HPBu ]Br/
3
Ru ratio is increased to 0.5 and then falls at higher ratios.
spectrum showed peaks at 2112, 2073, 2015, 1988, and
À1
1
952 cm and there was a peak at d=À12.67 ppm in the hy-
1
dride region of the H NMR spectrum. These spectroscopic fea-
À [27]
11
tures can be assigned to [HRu (CO) ] .
However, yellow
3
post-reaction solutions that were obtained from reactions with
[
HPBu ]Br/Ruꢀ1 (Figure 3c) had IR absorptions at 2111, 2048,
3
À1
2
036, 1970, and 1940 cm , a less-intense hydride resonance at
d=À12.67 ppm, and a new triplet hydride resonance at d=
The IR absorptions from
[
37]
À6.30 ppm ([RuHBr(CO) (PBu ) ].
2
3 2
these solutions with [HPBu ]Br/Ruꢀ1 mainly correspond to
3
À [27–29]
3
[
Ru(CO) Br ] .
The spectra (Figure 3b) from successful re-
3
actions that contain added [HPBu ]Br can be assigned as aris-
3
À
ing from mixtures of [HRu (CO) ] , [RuHBr(CO) (PBu ) ], and
3
11
2
3 2
À
[
RuBr (CO) ] ; the relative intensities of the peaks that belong
3 3
to each compound are dependent on the amount of [HPBu ]Br
3
that is added. This result suggests that, for higher reactivity,
À
a catalyst composition that contains both [HRu (CO) ] and
3
11
À
[
[
Ru(CO) Br ] is needed and that [HPBu ]Br reacts with
3 3 3
À
À
HRu (CO) ] to give [Ru(CO) Br ] .
3
11
3
3
In a related system with [Ru (CO) ] and KI in N-methylpyrro-
3
12
lidinone (NMP) as a solvent, Dombek has also shown that both
À
À
[
HRu (CO) ] and [RuI (CO) ] must be present in the solution
3 11 3 3
[10,11]
for it to be active towards CO-hydrogenation.
He argued
À
À
that [HRu (CO) ] donates a hydride moiety to [RuI (CO) ] (or
3
11
3
3
a complex that is derived from it) to make the key formyl inter-
mediate, which is not readily made by the direct intramolecu-
lar migration of a hydride group onto CO. In support of this ar-
gument, Dombek and Harrison showed that a rhenium–formyl
complex could be formed by intermolecular hydride-transfer
Figure 2. a) Yields of MeOH (^, ^) and EtOH (
&
,
&
) from the hydrogenation
]Br (15 g) with
of CO catalysed by [Ru (CO)12] (0.25 g, 0.39 mmol) in [PBu
]Br (empty shapes) or HBr (filled shapes); CO/H (1:2, 250 bar),
2
008C. b) Yield of MeOH+EtOH from the reactions shown in Figure 2a with
3
4
added [HPBu
3
À
[30]
f
r
o
m
[
H
R
u
(
C
O
)
]
t
o
[
C
p
R
e
(
N
O
)
(
C
O
)
]
.
Some of us also pro-
2
4
2
added [HPBu
3
]Br (*) or HBr (*).
vided support for this intermolecular hydride-transfer mecha-
&2
&
www.chemcatchem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 4
ÝÝ
These are not the final page numbers!

nism by synthesising [Ru(CHO)(CO)(dppe) ][SbF ] (dppe=1,2-
2
6
bis(diphenylphosphino)ethane) through hydride-transfer from
À
[31,32]
[
HRu(CO) (dppe) ] to [Ru(CO) (dppe) ][SbF ]
9 2 2 2 6 2.
Interestingly, Ono et al. have shown that [Ru (CO) ] reacts
3
12
with bis(triphenylphosphine)iminium chloride ([PPN]Cl) under
À
CO/H2 pressure to give [HRu CO ] and that, when HCl is
3
11
À
added, [Ru(CO) Cl ] forms. This latter complex is inactive to-
3
3
wards CO-hydrogenation, but the active species in that system
À [24]
is believed to be [RuH(CO) ] .
4
One possible mechanism for the activation of the ruthenium
catalysts by [HPBu ]Br would be for the tertiary phosphonium
3
salt to act as a source of HBr, which might then react with
À
À
[
HRu (CO) ] to give [RuBr (CO) ] . To test this possibility, we
3 11 3 3
performed a series of reactions with equivalent amounts of
HBr (Figure 2a, filled symbols) to those of [HPBu ]Br. The shape
3
3
1
1
Figure 4. P{ H} NMR spectrum of the solution that was recovered from the
hydrogenation of CO in the presence of [Ru (CO)12] ([HPBu ]Br/Ru. 1.0).
of the graph of MeOH-production is similar to that when using
3
3
[
HPBu ]Br, except that the yield of MeOH decreases more
3
quickly at higher [HBr].
13
[33]
By using C-labelling studies, we have shown that EtOH is
formed through MeOH as an intermediate; thus, the overall
rate of MeOH-production is represented by the rate of forma-
tion of MeOH+EtOH. These data are presented in Figure 2b
for the systems that are promoted by [HPBu ]Br (open sym-
3
bols) and by HBr (filled symbols). The very close correspond-
ence of these graphs very strongly suggests that the role of
Scheme 2. Proposed outline mechanism for the formation of MeOH and
EtOH from CO/H , which shows why HBr and [HPBu ]Br act in a similar
2
3
[
HPBu ]Br is to act as a source of HBr. The trend in the yield of
3
manner as promoters in the formation MeOH, but that the formation of
EtOH is inhibited when using [HPBu ]Br; PBu , which is formed alongside
MeOH+EtOH, together with IR studies, which show that
3
3
À
À
HBr, scavenges MeBr, which is an intermediate in the formation of EtOH.
[
HRu (CO) ] is smoothly converted into [Ru (CO) ] as HBr is
3
1
1
3
3
À
11
added, also reinforce the view that both [HRu (CO) ] and
3
À
[
RuBr (CO) ] must be present in the solution to afford good
3 3
activity in MeOH-production.
HBr or [HPBu ]Br added), the yield of EtOH is lower by
3
The situation for EtOH is different when using HBr compared
0.01 mol, which is more than 10 times the concentration of the
additive. Another contributor to the loss in activity towards
EtOH may be that the formed [RuHBr(CO) (PBu ) ] is inactive
with that when using [HPBu ]Br. In the presence of HBr, the
3
yield of EtOH increases as [HBr] is increased and only falls
when the yield of MeOH becomes low, although the ratio of
EtOH/MeOH continues to increase. These observations suggest
2
3 2
À
and removes [RuBr (CO) ] from the system. (We thank a -
3
3
referee for suggesting this alternative).
À
that [RuBr (CO) ] , which also increases at the expense of
3
3
À
[
HRu (CO) ] as HBr is added, is the major species that is re-
3 11
Conclusions
sponsible for the conversion of MeOH into EtOH. In contrast,
when [HPBu ]Br is used, the yield of EtOH remains fairly con-
We conclude that the irreproducibility that is often observed
when studying CO-hydrogenation reactions, especially in
molten phosphonium halides, may arise because of minor im-
purities that are present in the salt. We have discovered that
3
stant as increasing amounts of [HPBu ]Br are added, before fall-
3
ing at higher concentrations of [HPBu ]Br. [MePBu ]Br is ob-
3
3
served by NMR spectroscopy (Figure 4) at the end of the reac-
tion when [HPBu ]Br is used as the promoter, but not when
one such impurity, [HPBu ]Br, can act as a promoter of the re-
3
3
HBr is added. This difference suggests that the free PBu that is
action when added in small amounts (sub-stoichiometric with
respect to ruthenium). To the best of our knowledge, the use
of such compounds as promoters for catalytic reactions has
not been reported before, although PÀH phosphonium salts
have been used as air-stable alternatives to highly basic phos-
3
liberated when HBr is formed from [HPBu ]Br acts to scavenge
3
MeBr, which is an intermediate in the formation of EtOH, thus
lowering the rate of EtOH-formation as more [HPBu ]Br is
3
added. In a separate experiments, it has been shown that nei-
ther PBu nor [MePBu ]Br acts as a promoter, these reactions
[34]
phines, especially when both phosphines and acids are re-
3
3
[
34–36]
are outlined in Scheme 2.
quired in the system.
In the CO-hydrogenation reactions,
À
À
Quantitatively, if the formation of [MePBu ]Br were solely re-
[HPBu ]Br acts to convert [Ru (CO) ] into [Ru(CO) Br ] and
3
3
3
11
3
3
sponsible for the drop in yield of EtOH when using [HPBu ]Br
both of these ruthenium complexes are required for active cat-
3
instead of HBr, this drop in yield should be equal to the con-
alysis to occur. [HPBu ]Br appears to act as a source of HBr,
3
[24,25]
centration of PBu that is produced or to the concentration of
which others have shown (in other solvents)
has similar
3
[
HPBu ]Br that is added. However, this result is not the case.
promoting effects on the production of MeOH. We have also
3
For example, with an additive/Ru ratio of 0.75 (0.0009 mol of
confirmed that this effect is the case in molten [PBu ]Br. The
4
ChemCatChem 0000, 00, 1 – 4
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemcatchem.org
&
3
&
ÞÞ
These are not the final page numbers!

effect of [PBu ]Br on the production of EtOH is different to that
[17] J. F. Knifton, US, 4,332,914, 1982.
3
[
[
[
18] J. F. Knifton, US, 4,605,677, 1986.
19] J. F. Knifton, J. Chem. Soc. Chem. Commun. 1983, 729–730.
20] J. F. Knifton, US 4,315,993, 1982.
of HBr, partly because PBu that is formed by the loss of HBr
3
sequesters MeBr, the key intermediate in the production of
EtOH, and partly because inactive Ru/PBu3 complexes are
formed.
[21] J. F. Knifton, J. Am. Chem. Soc. 1981, 103, 3959–3961.
[22] J. F. Knifton, Hydrocarbon Process. 1981, 60, 113–117.
[
[
23] J. F. Knifton, Platinum Met. Rev. 1985, 29, 63–72.
24] H. Ono, K. Fujiwara, M. Hashimoto, H. Watanabe, K. Yoshida, J. Mol.
Catal. 1990, 58, 289–297.
Acknowledgements
[
25] H. Ono, M. Hashimoto, K. Fujiwara, E. Sugiyama, K. Yoshida, J. Organo-
met. Chem. 1987, 331, 387–395.
We thank Eastman Chemical Company for a studentship (J.H.B.)
and other support.
[
26] A. Robertson, Cytec, Private Communication.
[27] E. A. Seddon, K. R. Seddon, The Chemistry of Ruthenium, Elsevier, Amster-
dam, 1984.
[
28] D. F. Gill, B. E. Mann, B. L. Shaw, J. Chem. Soc. Dalton Trans. 1973, 311–
317.
Keywords: carbon monoxide · hydrogenation · molten salts ·
ruthenium · salt effect
[
[
[
29] M. J. Cleare, W. P. Griffith, J. Chem. Soc. A 1969, 372–380.
30] B. D. Dombek, A. M. Harrison, J. Am. Chem. Soc. 1983, 105, 2485–2486.
31] D. S. Barratt, D. J. Cole-Hamilton, J. Chem. Soc. Chem. Commun. 1985,
[1] W. F. Gresham, C. E. Schweizer, US Patent 2,534,018, 1950.
[2] I. Wender, Fuel Process. Technol. 1996, 48, 189–297.
[3] R. C. Brady, R. Pettit, J. Am. Chem. Soc. 1981, 103, 1287–1289.
[4] L. M. Tau, H. A. Dabbagh, B. H. Davis, Energy Fuels 1991, 5, 174–179.
[5] G. Henrici-Olivꢁ, S. Olive, Angew. Chem. 1976, 88, 144–150; Angew.
Chem. Int. Ed. Engl. 1976, 15, 136–141.
1559–1560.
[
[
32] D. S. Barratt, D. J. Cole-Hamilton, J. Organomet. Chem. 1986, 306, C41–
C44.
33] J. H. Blank, R. Hembre, J. Ponasik, D. J. Cole-Hamilton, unpublished ob-
servations.
[
[
34] M. R. Netherton, G. C. Fu, Org. Lett. 2001, 3, 4295–4298.
35] A. A. Nunez-Magro, L. M. Robb, P. J. Pogorzelec, A. M. Z. Slawin, G. R.
Eastham, D. J. Cole-Hamilton, Chem. Sci. 2010, 1, 723–730.
36] D. J. Cole-Hamilton, A. J. Rucklidge, G. E. Morris, A. M. Z. Slawin, Helv.
Chim. Acta 2006, 89, 1783–1800.
[
[
[
[
6] B. D. Dombek, Adv. Catal. 1983, 32, 325–416.
7] J. S. Bradley, US 4,421,862, 1983.
8] J. S. Bradley, J. Am. Chem. Soc. 1979, 101, 7419–7421.
9] B. D. Dombek, J. Am. Chem. Soc. 1981, 103, 6508–6510.
[
[
[10] B. D. Dombek, J. Organomet. Chem. 1983, 250, 467–483.
[11] B. D. Dombek, Organometallics 1985, 4, 1707–1712.
[12] B. D. Dombek, J. Organomet. Chem. 1989, 372, 151, and references
therein.
37] Although this compound has not been reported previously, the related
[
RuHBr(CO)
Moers, R. W. M. Tan Hoedt, J. P. Langhort, J. Inorg. Nucl. Chem. 1974, 36,
279.
2
(PCy
3
)
2
] has a hydride resonance at d=À5.9 ppm; see: F. G.
2
[
[
[
[
13] J. F. Knifton, J. Catal. 1985, 96, 439–453.
14] J. F. Knifton, ACS Symp. Ser. 1987, 328, 98–107.
15] J. F. Knifton, R. A. Grigsby, J. J. Lin, Organometallics 1984, 3, 62–69.
16] S. H. Han, G. L. Geoffroy, B. D. Dombek, A. L. Rheingold, Inorg. Chem.
Received: March 23, 2012
Revised: June 20, 2012
1988, 27, 4355–4361.
Published online on && &&, 0000
&4
&
www.chemcatchem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 4
ÝÝ
These are not the final page numbers!

COMMUNICATIONS
A big promotion: One reason for the ir-
reproducibility that is sometimes found
in hydrogenation reactions of CO cata-
lysed by [Ru (CO) ] in molten [PBu ]Br
J. H. Blank, R. Hembre, J. Ponasik,
D. J. Cole-Hamilton*
&& – &&
3
12
4
is the possible presence of [HPBu ]Br,
A Tertiary Phosphonium Salt as
a Promoter for the Hydrogenation of
CO
3
which acts as a promoter by providing
a source of HBr.
ChemCatChem 0000, 00, 1 – 4
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemcatchem.org
&
5
&
ÞÞ
These are not the final page numbers!