Liquid-biphase hydrogenolysis of benzo[b]thiophene by rhodium catalysis

Article abstract of DOI:10.1021/ja9641517

The catalytic activity of the zwitterionic complex [(sulphos)Rh(cod)] for the hydrogenation and hydrogenolysis reactions of benzo[b]thiophene (BT) has been studied in either methanol or liquid-biphase systems comprising MeOH or MeOH-H₂O as the polar phase and n-heptane as the organic phase [sulphos = - O₃S(C₆H₄)CH₂C(CH₂PPh₂)₃]. The catalyst activity is independent of the phase variation. Under neutral conditions, the slow but selective hydrogenation of BT to 2,3- dihydrobenzo[b]thiophene is observed. Conversely, in the presence of NaOH or other strong bases, the fast and selective hydrogenolysis of BT to 2-ethylthiophenol sodium salt occurs. In a typical liquid-biphase hydrogenolysis reaction [35 mg (0.035 mmol) of catalyst, 470 mg (3.5 mmol) of BT, 180 mg (4.5 mmol) of NaOH, 5 mL of MeOH, 5 mL of H₂O, 10 mL of n-heptane. 30 bar of H₂, 160 °C], all the substrate is practically consumed in ca. 5 h to give the 2-ethylthiophenolate product. The strong base plays a dual role in the hydrogenolysis reaction: it promotes the formation of the Rh-H species (which is necessary for the C-S insertion step) by heterolytic splitting of H₂ and accelerates the conversion of BT by aiding the inductive elimination of the hydrogenolysis product. The effect of the H₂ pressure and of the substrate, catalyst, and base concentrations on the conversion rate of BT has been studied under liquid-hiphase conditions. High- pressure NMR experiments in sapphire tubes have provided mechanistic information on the catalysis cycle for the hydrogenolysis of BT in MeOH. Under catalytic conditions, the phosphorus-containing rhodium compounds (visible on the NMR time scale) are the trihydride [(sulphos)RhH₃]^- and the dihydride 2- ethylthiophenolate complex [(sulphos)Rh(H)₂(o-S(C₆H₄)C₂H₅)]^- . Consistent with previous studies, the inductive elimination of the thiol from the metal center is suggested to be the rate- determining step of the hydrogenolysis reaction of BT catalyzed by the 16e^- fragment [(sulphos)RhH]^-.

Full text of DOI:10.1021/ja9641517

J. Am. Chem. Soc. 1997, 119, 4945-4954
4945
Liquid-Biphase Hydrogenolysis of Benzo[b]thiophene by
Rhodium Catalysis
Claudio Bianchini,* Andrea Meli, Veronique Patinec, Volker Sernau, and
Francesco Vizza
Contribution from the Istituto per lo Studio della Stereochimica ed Energetica dei Composti di
Coordinazione, ISSECC-CNR, 50132 Firenze, Italy
ReceiVed December 2, 1996X
Abstract: The catalytic activity of the zwitterionic complex [(sulphos)Rh(cod)] for the hydrogenation and
hydrogenolysis reactions of benzo[b]thiophene (BT) has been studied in either methanol or liquid-biphase systems
comprising MeOH or MeOH-H2O as the polar phase and n-heptane as the organic phase [sulphos )
-O3S(C6H4)CH2C(CH2PPh2)3]. The catalyst activity is independent of the phase variation. Under neutral conditions,
the slow but selective hydrogenation of BT to 2,3-dihydrobenzo[b]thiophene is observed. Conversely, in the presence
of NaOH or other strong bases, the fast and selective hydrogenolysis of BT to 2-ethylthiophenol sodium salt occurs.
In a typical liquid-biphase hydrogenolysis reaction [35 mg (0.035 mmol) of catalyst, 470 mg (3.5 mmol) of BT, 180
mg (4.5 mmol) of NaOH, 5 mL of MeOH, 5 mL of H2O, 10 mL of n-heptane, 30 bar of H2, 160 °C], all the substrate
is practically consumed in ca. 5 h to give the 2-ethylthiophenolate product. The strong base plays a dual role in the
hydrogenolysis reaction: it promotes the formation of the Rh-H species (which is necessary for the C-S insertion
step) by heterolytic splitting of H2 and accelerates the conversion of BT by aiding the reductive elimination of the
hydrogenolysis product. The effect of the H2 pressure and of the substrate, catalyst, and base concentrations on the
conversion rate of BT has been studied under liquid-biphase conditions. High-pressure NMR experiments in sapphire
tubes have provided mechanistic information on the catalysis cycle for the hydrogenolysis of BT in MeOH. Under
catalytic conditions, the phosphorus-containing rhodium compounds (visible on the NMR time scale) are the trihydride
[(sulphos)RhH3]- and the dihydride 2-ethylthiophenolate complex [(sulphos)Rh(H)2(o-S(C6H4)C2H5)]-. Consistent
with previous studies, the reductive elimination of the thiol from the metal center is suggested to be the rate-determining
step of the hydrogenolysis reaction of BT catalyzed by the 16e- fragment [(sulphos)RhH]-.
Introduction
Although HDS is routinely carried out in oil refineries
worlwide, a clear need exists for more active catalysts capable
of removing a greater proportion of sulfur. Also, different
refined fuels contain different types of organosulfur compounds,
and thus more specific catalysts for each type of feedstocks need
to be developed.
Considering the current world production of petroleum, its
hydroprocessing is probably the largest volume industrial
application of transition-metal catalysis. In this procedure, fossil
fuel feedstocks are treated with a high pressure of hydrogen
(35-170 bar) over hot heterogeneous catalysts (300-425 °C)
to remove sulfur, nitrogen, and residual metals prior to further
processing.1
Sulfur in fossil materials is contained in various organic
compounds, which include thiols, sulfides, disulfides, and the
more refractory thiophenes, benzothiophenes, and diben-
zothiophenes. The removal of sulfur from fossil materials is
commonly referred to as hydrodesulfurization (HDS). There
are two major reasons for reducing sulfur levels in petroleum
feedstocks: (i) to minimize the amounts of sulfur introduced
into the atmosphere by combustion of petroleum-based fuels
(contribution to acid rain) and (ii) to prevent the poisoning of
reforming and cracking catalysts that are used to upgrade
feedstocks to product fuels.2
The commercial HDS catalysts are generally prepared by
coimpregnation of Mo or W salts on a γ-alumina support,
followed by sulfidation with H2S/H2. Increased efficiency is
achieved by introduction of late transition metals like Co, Ni,
Ru, Rh, and Ir as promoters.3 According to a recent mechanistic
interpretation,4 the active centers for the activation of the
thiophenic molecules are just the promoter atoms located at the
edge plane of a MoS2 (or WS2) single slab, whereas H2
activation occurs on MoS2 (or WS2). The key role of the metal
promoter in the activation of the thiophenes has also been
supported by several homogeneous modeling studies which
clearly show that C-S bond scission is almost exclusively
brought about by promoter metal complexes.5 For the follow-
ing, crucial desulfurization step, however, soluble metal com-
plexes are generally inefficient, unless either W (or Mo) or
hydride ligands bound to a second metallic center are comprised
in the complex framework.6-8 In the absence of these two
conditions, the ultimate activation of thiophenes by soluble metal
complexes seems to be the conversion to thiolate ligands, which
may eventually be eliminated as thiols upon treatment with
X Abstract published in AdVance ACS Abstracts, May 1, 1997.
(1) (a) Leffler, W. L. Petroleum Refining; PennWell Publishing Co.:
Tulsa, OK, 1985. (b) Scherzer, J.; Gruia, A. J. Hydrocracking Science and
Technology; Marcel Dekker: New York, 1996.
(2) (a) Gates, B. C.; Katzer, J. R.; Schuit, G. C. A. Chemistry of Catalytic
Processes; McGraw-Hill: New York, 1979. (b) Satterfield, C. N. Hetero-
geneous Catalysis in Industrial Practice; McGraw-Hill: New York, 1980.
(c) Topsøe, H.; Clausen, B. S. Catal. ReV.-Sci. Eng. 1984, 26, 395. (d)
Chianelli, R. R. Catal. ReV.-Sci. Eng. 1984, 26, 361. (e) Friend, C. M.;
Roberts, J. T. Acc. Chem. Res. 1988, 21, 394. (f) Prins, R.; deBeer, V. H.
J.; Somorjai, G. A. Catal. ReV.-Sci. Eng. 1989, 31, 1. (g) Gates B. C.
Catalytic Chemistry; John Wiley: New York, 1992; Chapter 5, p 390. (h)
Hydrotreating Technology for Pollution Control; Occelli, M. L., Chianelli,
R., Eds.; Marcel Dekker: New York, 1996.
(3) (a) Pecoraro, T. A.; Chianelli, R. R. J. Catal. 1981, 67, 430. (b) Harris,
S.; Chianelli, R. R. J. Catal. 1984, 86, 400. (c) Girgis, M. J.; Gates, B. C.
Ind. Eng. Chem. Res. 1991, 30, 2021. (d) Wiegand, B. C.; Friend, C. M.
Chem. ReV. 1992, 92, 491. (e) Benson, J. W.; Schrader, G. L.; Angelici, R.
J. J. Mol. Catal., A 1995, 96, 283.
(4) Startsev, A. N. Catal. ReV.-Sci. Eng. 1995, 37, 353.
S0002-7863(96)04151-0 CCC: $14.00 © 1997 American Chemical Society

4946 J. Am. Chem. Soc., Vol. 119, No. 21, 1997
Bianchini et al.
H2.9-12 Homogeneous catalytic reactions for the hydrogenolysis
of thiophene, benzo[b]thiophene, dibenzo[b,d]thiophene, and
dinaphtho[2,1-b:1′,2′-d]thiophene have already been reported.9c,d,13
The proven capability of promoter metal complexes of
catalyzing the hydrogenolysis of thiophenes to thiols under mild
reaction conditions suggests an alternative two-step procedure
to HDS in which petroleum is firstly treated with H2 in the
presence of a late-transition-metal catalyst and then is hydro-
genated over conventional heterogeneous catalysts. These
catalysts, in fact, are capable of catalyzing the desulfurization
of the thiols under much milder conditions than those currently
employed to accomplish the overall HDS of the thiophenic
precursors.2,3
As reported above, the hydrogenolysis of thiophenes to thiols
may not necessarily imply a heterogeneous process, but certainly
cannot be performed in the homogeneous phase at the industrial
level. The impressive progress recently achieved in the field
of water-soluble catalysts, however, paves the way to the
application of aqueous biphase catalysis to large volume
reactions such as the hydrotreating of distillates.14 Indeed, a
two-step (aqueous biphase/heterogeneous) approach to HDS of
thiophenes would have some advantages over conventional
heterogeneous processes: (i) Molecular metal catalysts are more
amenable to mechanistic studies applying spectroscopy than
supported catalysts. Hence, a fine tuning of the liquid-biphase
catalyst and of its activity may be anticipated, particularly as
regards the design of catalysts specifically tailored for different
types of thiophenes. (ii) The application of heterogeneous
catalysis to the desulfurization of the thiols allows the use of
mild reaction conditions under which the benzene rings of the
benzothiophenes and dibenzothiophenes are not affected, and
hence a higher octane rating may be obtained. (iii) The direct
extraction of the thiol products into the water phase, resulting
in the net desulfurization of the fuel, may be a feasible process
by addition of bases. On the other hand, disadvantages of the
two-step process proposed here may be envisaged in the
increased cost as well as its application to very large volume
reactions. The economical question may become of secondary
importance, particularly for those countries which are rich in
very heavy crudes, as one considers that international regulations
will soon require reducing the sulfur contents in fuels to less
than 100 ppm (60 ppm in some countries).15 Indeed, this level
is not easily attained with the conventional catalysts, even by
repeated treatments. The two-step process might thus be
convenient for the purification of distillates from residual sulfur
up to the limit of commercial fuels.
We have recently reported the synthesis of a zwitterionic
rhodium(I) complex, [(sulphos)Rh(cod)] (1), that exhibits inher-
ent surface-active attributes [sulphos ) -O3S(C6H4)CH2C-
(CH2PPh2)3, cod ) cycloocta-1,5-diene].16 This complex is
soluble in polar solvents (MeOH or 1:1 (v/v) MeOH-H2O
mixtures) but not in hydrocarbons. However, a 1:1 mixture of
MeOH and n-heptane containing 1 gives a unique phase already
at 60 °C, but at room temperature complete phase separation
occurs. In the MeOH-H2O/n-heptane system, 1 behaves as a
catalyst precursor for the hydrogenation and hydroformylation
of olefins, at the end of which all the rhodium is recovered in
the polar phase and the products are found in the hydrocarbon
phase.16
These properties of 1 prompted us to study its application to
the liquid-biphase hydrogenolysis of benzo[b]thiophene, which
is one of the most difficult thiophenic substrates to degrade.
The results of our study are described here.
(5) (a) Angelici, R. J. Acc. Chem. Res. 1988, 21, 387. (b) Angelici, R.
J. Coord. Chem. ReV. 1990, 105, 61. (c) Rauchfuss, T. B. Prog. Inorg.
Chem. 1991, 39, 259. (d) Reynolds, J. G.Chem. Ind. (London) 1991, 570.
(e) Sa´nchez-Delgado, R. A. J. Mol. Catal. 1994, 86, 287. (f) Angelici, R.
J. In Encyclopedia of Inorganic Chemistry; King, R. B., Ed.; John Wiley:
New York, 1994; Vol. 3, p 1433. (g) Angelici, R. J. Bull. Soc. Chim. Belg.
1995, 104, 265. (h) Bianchini, C.; Meli, A. J. Chem. Soc., Dalton Trans.
1996, 801. (i) Bianchini, C.; Meli, A. In Applied Homogeneous Catalysis
with Organometallic Compounds; Cornils, B., Herrmann, W. A., Eds.;
VCH: Weinheim, 1996; Vol. 2, p 969.
(6) (a) Riaz, U.; Curnow O. J.; Curtis, M. D. J. Am. Chem. Soc. 1995,
117, 6366. (b) Riaz, U.; Curnow O. J.; Curtis, M. D. J. Am. Chem. Soc.
1995, 116, 4357. (c) Riaz, U.; Curnow O. J.; Curtis, M. D. J. Am. Chem.
Soc. 1995, 113, 1416.
(7) Jones, W. D.; Chin, R. M. J. Am. Chem. Soc. 1994, 116, 198.
(8) Bianchini, C.; Jime´nez, M. V.; Mealli, C.; Meli, A.; Moneti, S.;
Patinec, V.; Vizza, F. Angew. Chem., Int. Ed. Engl. 1996, 35, 1706.
(9) (a) Bianchini, C.; Meli, A.; Peruzzini, M.; Vizza, F.; Moneti, S.;
Herrera, V; Sa´nchez-Delgado, R. A. J. Am. Chem. Soc. 1994, 116, 4370.
(b) Bianchini, C.; Jime´nez, M. V.; Meli, A.; Moneti, S.; Vizza, F.; Herrera,
V; Sa´nchez-Delgado, R. A. Organometallics 1995, 14, 2342. (c) Bianchini,
C.; Herrera, V; Jime´nez, M. V.; Meli, A.; Sa´nchez-Delgado, R. A.; Vizza,
F. J. Am. Chem. Soc. 1995, 117, 8567. (d) Bianchini, C.; Fabbri, D.; Gladiali,
S.; Meli, A.; Pohl, W.; Vizza, F. Organometallics 1996, 15, 4604.
(10) Rosini, G. P.; Jones, W. D. J. Am. Chem. Soc. 1992, 114, 10767.
(11) Garcia, J. J.; Mann, B. E.; Adams, H.; Bailey, N. A.; Maitlis, P. M.
J. Am. Chem. Soc. 1995, 117, 2179.
(12) Ogilvy, A. E.; Draganjac, M.; Rauchfuss, T. B.; Wilson, S. R.
Organometallics 1988, 7, 1171.
(13) Bianchini, C.; Casares, J. A.; Meli, A.; Vizza, F.; Sa´nchez-Delgado,
R. A. Polyhedron, in press.
(14) (a) Barton, M.; Atwood, J. D. J. Coord. Chem. 1991, 24, 43. (b)
Kalck, P.; Monteil, F. AdV. Organomet. Chem. 1992, 34, 219. (c) Herrmann,
W. A.; Kohlpaintner, C. W. Angew. Chem., Int. Ed. Engl. 1993, 32, 1524.
(d) Cornils, B.; Wiebus, E. CHEMTECH 1995, 25, 33. (e) Buhling, A.;
Elgersma, J. W.; Nkrumah, S.; Kamer, P. C. J.; van Leeuwen, P. W. N. M.
J. Chem. Soc., Dalton Trans. 1996, 2143. (f) Cornils, B.; Herrmann, W. A.
In Applied Homogeneous Catalysis with Organometallic Compounds;
Cornils, B., Herrmann, W. A., Eds.; VCH: Weinheim, 1996; Vol. 2, p
575.
Experimental Section
General Information. All reactions and manipulations were
routinely performed, unless otherwise stated, under a nitrogen atmo-
sphere by using standard Schlenk techniques. High-pressure, high-
temperature reactions under a controlled pressure of hydrogen were
performed with a stainless steel Parr 4565 reactor equipped with a Parr
4842 temperature and pressure controller. Tetrahydrofuran (THF) was
purified by distillation under nitrogen from LiAlH4. Benzo[b]thiophene
(99%, Aldrich) was sublimed prior to use. Potassium tert-butoxide
(KOBut, 95%), methanol, and n-heptane were purchased from Aldrich
and used without further purification. All the other reagents and
chemicals were reagent grade and were used as received by commercial
suppliers. The rhodium complex [(sulphos)Rh(cod)] (1) was prepared
as previously described.16 Deuterated solvents for NMR measurements
(Merck) were dried over molecular sieves. 1H , 13C{1H}, and 31P{1H}
NMR spectra were obtained on either a Bruker ACP 200 (200.13, 50.32,
and 81.01 MHz, respectively) or a Varian VXR 300 (299.94, 75.43,
and 121.42 MHz, respectively) spectrometer. All chemical shifts are
reported in parts per million (δ) relative to tetramethylsilane, referenced
to the chemical shifts of residual solvent resonances (1H, 13C) or 85%
H3PO4 (31P). The 10 mm sapphire NMR tube was purchased from
Saphikon (Milford, NH), while the titanium high-pressure charging head
was constructed at ISSECC-CNR (Firenze, Italy). GC analyses were
performed on a Shimadzu GC-14 A gas chromatograph equipped with
a flame ionization detector and a 30 m (0.25 mm i.d., 0.25 µm film
thickness) SPB-1 Supelco fused silica capillary column. GC/MS
analyses were performed on a Shimadzu QP 5000 apparatus equipped
(15) (a) Takatsuka, T.; Wada, Y.; Suzuki, H.; Komatsu, S.; Morimura,
Y. J. Jpn. Pet. Inst. 1992, 35, 179. (b) Amorelli, A.; Amos, Y. D.; Haisig,
C. P.; Koaman, I. J.; Jonker, R. S.; de Wind, M.; Vrieling, J. Hydrocarbon
Proc. 1992, June, 93.
(16) Bianchini, C.; Frediani, P.; Sernau, V. Organometallics 1995, 14,
5458.

Liquid-Biphase Hydrogenolysis of Benzo[b]thiophene
J. Am. Chem. Soc., Vol. 119, No. 21, 1997 4947
with a column identical to that used for GC analyses. Elemental
analyses (C, H, N) were performed using a Carlo Erba Model 1106
elemental analyzer. Atomic absorption analyses were performed with
a Perkin-Elmer 5000 instrument.
1.19 (t, 3H, CH2CH3); 13C{1H} NMR (CDCl3, 20 °C) δ 142.5 (C),
137.8 (C), 128.8 (CH), 127.2 (CH), 126.0 (CH), 125.7 (CH), 27.4 (CH2-
CH3), 16.3 (SCH3), 15.0 (CH2CH3); GC/MS, EIMS, 70 eV, m/e (%)
+
152 (100) M+, 137 (95) M - Me+, 105 (24) M - SMe+, 91 (22) C7H7
,
+
77 (30) C6H5
.
Hydrogenation of Benzo[b]thiophene to 2,3-Dihydrobenzo[b]-
thiophene. In a typical experiment, the catalyst 1 and benzo[b]-
thiophene (BT) were dissolved in the appropriate mixture of deareated
organic solvents in a Schlenk flask. The mixture was then transferred
under nitrogen into a Parr reactor. In the reactions with water, this
was introduced into the reactor in a second step due to the poor
solubility of the reaction components in either MeOH-H2O or MeOH-
H2O/n-heptane at ambient temperature. A total amount of 20 mL of
solvents was generally used. After the reactor was pressurized to the
desired H2 pressure at room temperature, the temperature was increased
with stirring (650-1800 rpm). After the chosen time, the reactor was
gently cooled to room temperature over ca. 1 h (initially by exposure
to air, and then, when the internal temperature was lower than 100 °C,
by means of an ice-water bath). At this point, the reactor was
depressurized (in some experiments the gaseous phase was bubbled
into an aqueous solution of Pb(OAc)2 at pH = 4.5 for eventual trapping
of H2S). The reactor was then opened in the air, and its contents were
transferred Via syringe into a flask and analyzed by GC and GC/MS.
In the liquid-biphase reactions, each phase was analyzed by GC.
Afterward, THF (ca. 30 mL) was added until a unique phase was
observed, which was analyzed to obtain the total distribution of
products. Selected catalytic reactions were also carried out in the
presence of excess elemental Hg (2000:1, 1800 rpm) to test the
homogeneous character of the reactions.17 Upon hydrogenation, the
cod ligand in the catalyst precursor was liberated in solution as a mixture
of cyclooctene and cyclooctane.
Hydrogenolysis of BT to 2-Ethylthiophenolate. A. Parr Reactor
Experiments. The procedure described above for the hydrogenation
of BT to 2,3-dihydrobenzo[b]thiophene (DHBT) was followed also for
the hydrogenolysis reactions, the only difference being that a basic
coreagent (generally NaOH) was used. In a typical experiment, the
desired amount of base was introduced into the reactor as a solid prior
to the addition of the solvents and reagents. After each catalytic run,
the reactor was cooled to room temperature and its contents were
collected in a flask containing 1-2 mL of HCl(aq). This acidification
step converts the sodium salt of 2-ethylthiophenolate to 2-ethylthio-
phenol (ETP). This procedure is necessary to have reliable GC analyses
as well as to prevent the base-promoted oxidation of the thiophenolate
product to bis(2-ethylphenyl) sulfide. In the case of liquid-biphase
reactions, ca. 50 mL of THF was added to obtain a unique phase for
GC and GC/MS analyses. Experimental conditions for a typical
reaction in the liquid-biphase system MeOH-H2O/n-heptane are as
follows: 35 mg (0.035 mmol) of catalyst, 470 mg (3.5 mmol) of BT,
180 mg (4.5 mmol) of NaOH, 5 mL of MeOH, 5 mL of H2O, 10 mL
of n-heptane, 30 bar of H2, 160 °C.
B. Sapphire Tube HPNMR Experiments. A 10 mm sapphire
HPNMR (HP ) high pressure) tube was charged with a MeOH-d4 (2
mL) solution of 1 (60 mg, 0.06 mmol) and a 10-fold excess of both
BT (80 mg, 0.6 mmol) and KOBut (67 mg, 0.6 mmol) under nitrogen,
pressurized with hydrogen to 30 bar at room temperature, and then
placed into a NMR probe at room temperature. The reaction was
followed by variable-temperature 31P{1H} NMR spectroscopy.
A
sequence of selected 31P{1H} NMR spectra is reported in Figure 3.
The transformation of 1 (trace a, at 20 °C) already occurred at 60 °C,
yielding four new sets of resonances (trace b, after 1.5 h), three of
which were assigned to [(sulphos)Rh(η3-S(C6H4)CHdCH2)]- (2),
[(sulphos)Rh(H)2(o-S(C6H4)C2H5)]- (3), and [(sulphos)Rh(H)3]- (4).
A fourth signal (doublet at 15.0 ppm) is attributed to a rhodium complex
of still undefined structure that does not participate in any transformation
of BT. Indeed, this species also forms by hydrogenation of 1 in MeOH-
d4 in the presence of KOBut (see below) and does not react with either
BT or ETP. Compounds 2-4 were identified by comparison of their
31P{1H} NMR spectra to those of the analogous triphos complexes
[(triphos)Rh(η3-S(C6H4)CHdCH2)] (2*),18 [(triphos)Rh(H)2(o-S(C6H4)-
C2H5)] (3*),9c and [(triphos)Rh(H)3] (4*),19 respectively. Increasing
the temperature to 120 °C (a temperature to which the catalysis occurs,
see entry 5 in Table 2) led to a considerable increase of the concentration
of the thiolate dihydride 3 at the expense of both 2 and 4. The trihydride
did not disappear even for prolonged heating (2 h), which indicated
the attainment of a thermostationary state (trace c). After the catalytic
reaction was quenched by cooling to room temperature, compounds
2-4 were detected in ratios of 33:48:19 (trace d). The tube was then
removed from the spectrometer and depressurized to ambient pressure.
The contents of the tube were then transferred into a Schlenk-type flask
and acidified with HCl(aq). A sample of the solution, withdrawn and
analyzed by GC and GC/MS, was found to contain ETP (35%), BT
(53%), and DHBT (12%) as mixtures of H/D isotopomers. The
following spectroscopic properties are illustrative of compounds 2-4.
Data for 2: 31P{1H} NMR AMQX pattern, δ 31.3 (PA, J(PAPM) ) 34.0
Hz, J(PAPQ) ) 32.5 Hz, J(PARh) ) 108.0 Hz), -0.8 (PM, J(PMPQ) )
48.1 Hz, J(PMRh) ) 116.2 Hz), -9.8 (PQ, J(PQRh) ) 105.9 Hz). Data
for 3: 31P{1H} NMR AM2X pattern, δ 37.2 (PA, J(PAPM) ) 25.2 Hz,
J(PARh) ) 114.6 Hz), 1.8 (PM, J(PMRh) ) 80.3 Hz); 1H NMR δ -7.85
(dm, J(HPtrans) ) 179 Hz, RhH). Data for 4: 31P{1H} NMR A3X
pattern, δ 22.8 (PA, J(PARh) ) 86.1 Hz); 1H NMR δ -8.68 (dm,
J(HPtrans) ) 146 Hz, RhH). Due to H/D exchange caused by the
MeOH-d4 solvent, the hydride resonances in the 1H NMR spectra of 3
and 4 appear as low-intensity, broad signals. The coupling to deuterium
similarly affects the 31P{1H} NMR resonances of the phosphorus nuclei
trans to the hydride ligands.
In selected experiments carried out in MeOH-H2O/n-heptane
mixtures, the phases were collected without HCl addition. Both phases
were analyzed by GC which showed that all the produced thiophenolate
accumulates in the polar phase, while the unreacted substrate, eventual
byproducts [DHBT, ethylbenzene (EB)], cyclooctane, and cyclooctene
remain in the n-heptane phase. The total product distribution was
confirmed by analysis of the unique phase obtained by adding HCl
and THF. In several biphasic experiments, after the hydrocarbon layer
was separated, the solvent was removed in vacuo and the residue was
analyzed by 31P NMR spectroscopy and atomic absorption spectrometry.
No appreciable amount of either phosphorus compounds or rhodium
was detected. Selected catalytic reactions were also carried out in the
presence of either an excess of elemental Hg (2000:1, 1800 rpm) or a
commercial surfactant of wide use in biphase reactions (Pluronic L101).
S-Methyl-2-ethylthiophenol (METP) is formed as a byproduct already
at 160 °C by reaction of sodium ethylthiophenolate with MeOH. The
concentration of METP increases with time and temperature. The
formation of METP was confirmed by the independent reaction of
sodium ethylthiophenolate with MeOH at 200 °C in the absence of Rh
catalyst. Data for S-Methyl-2-ethylthiophenol: 1H NMR (CDCl3, 20
°C) δ 7.2-7.0 (m, 4H, Ph), 2.68 (q, 2H, CH2CH3), 2.32 (s, 3H, SCH3),
An analogous variable-temperature HPNMR experiment carried out
at higher pressures of H2 (50-70 bar) showed an identical sequence
of products. It was observed, however, that the concentration of the
trihydride 4 under catalytic conditions (120 °C) sligthly increases with
pressure.
HPNMR Reaction of 1 with H2, Followed by Sequential Treat-
ment with BT and H2. A 10 mm sapphire HPNMR tube was charged
with a MeOH-d4 (2 mL) solution of 1 (60 mg, 0.06 mmol) and a 10-
fold excess of KOBut (67 mg, 0.6 mmol) under nitrogen, pressurized
with hydrogen to 30 bar at room temperature, and then placed into a
NMR probe at room temperature. The reaction was followed by
variable-temperature 31P{1H} NMR spectroscopy. A sequence of
selected 31P{1H} NMR spectra is reported in Figure 4. The reaction
of 1 (trace a, at 20 °C) with hydrogen already occurred at 60 °C,
yielding almost quantitatively the trihydride 4 (trace b, at 20 °C after
3 h at 60 °C). The tube was depressurized to ambient pressure, and
then all the hydrogen was replaced with nitrogen by three freeze/pump/
(18) (a) Bianchini, C.; Frediani, P.; Herrera, V; Jime´nez, M. V.; Meli,
A.; Rinco´n, L.; Sa´nchez-Delgado, R. A.; Vizza, F. J. Am. Chem. Soc. 1995,
117, 4333.
(19) Ott, J.; Venanzi, L. M.; Ghilardi, C. A.; Midollini, S.; Orlandini,
A. J. Organomet. Chem. 1985, 291, 89.
(17) Lin, Y.; Finke, R. G. Inorg. Chem. 1994, 33, 4891.

4948 J. Am. Chem. Soc., Vol. 119, No. 21, 1997
Bianchini et al.
Table 1. Selective Hydrogenation of Benzo[b]thiophene Catalyzed by [(sulphos)Rh(cod)]a
reaction mixture composition (%)c
entry
solvent
T (°C)b
t (h)
BT
DHBT
METP
ETP
EB
1
2
3
4
MeOH
MeOH
200
200
200
175
17
70
17
70
50
60
75
30
43
38
24
<1
7
1
1
MeOH-H2O (1:1, v/v)
MeOH-H2O/n-heptane (1:1:2, v/v/v)
1
110
a Reaction conditions: catalyst, 35 mg, 0.035 mmol; BT, 470 mg, 3.5 mmol; solvent, 20 mL; H2 pressure, 30 bar. b At 200 °C partial destruction
of the catalyst was observed. c Key: benzo[b]thiophene (BT), 2,3-dihydrobenzo[b]thiophene (DHBT), S-methyl-2-ethylthiophenol (METP),
2-ethylthiophenol (ETP), ethylbenzene (EB).
thaw cycles. After 10 equiv of solid BT (80 mg, 0.6 mmol) was
Scheme 1
introduced into the tube, this was placed into the probe and heated to
100 °C for 1 h. After the probe was cooled to room temperature, the
quantitative conversion of 4 to 2 was observed (trace c). Finally, the
tube was pressurized with hydrogen to 30 bar and then heated to 120
°C. The spectrum, acquired at this temperature after 40 min, showed
the conversion of 2 to 3 and 4 (trace d). Trace e shows the spectrum
after the probe was slowly cooled to room temperature.
Scheme 2
Synthesis of [(sulphos)Rh(η3-S(C6H4)CHdCH2)]- as the Potas-
sium Salt (K2). This synthesis was performed by using a 10 mm
sapphire HPNMR tube as a small autoclave. The tube was charged
with a MeOH (2 mL) solution of 1 (60 mg, 0.06 mmol) and a 2-fold
excess of KOBut (13 mg, 0.12 mmol) under nitrogen, pressurized with
hydrogen to 30 bar at room temperature, and then placed into a water
bath at 60 °C. After 3 h, the tube was cooled to room temperature and
depressurized to ambient pressure, and then all the hydrogen was
replaced with nitrogen by three freeze/pump/thaw cycles. The tube
was then opened under nitrogen, charged with a solution of BT (80
mg, 0.6 mmol) in MeOH, and heated to 100 °C (ethylene glycol/water
bath). After 1 h, the tube was allowed to reach room temperature and
the contents were transferred Via syringe into a Schlenk-type flask.
After the volatiles were removed under vacuum, the residue was washed
with diethyl ether to eliminate the excess BT. A sample of the residue,
dissolved in MeOH-d4 and characterized by 1H and 31P{1H} NMR
spectroscopy, was found to contain K2 in ca. 90% yield (31P{1H} NMR
integration): 31P{1H} NMR (MeOH-d4, 20 °C), AMQX pattern, δ 31.3
(PA, J(PAPM) ) 34.0 Hz, J(PAPQ) ) 32.5 Hz, J(PARh) ) 108.0 Hz),
-0.8 (PM, J(PMPQ) ) 48.1 Hz, J(PMRh) ) 116.2 Hz), -9.8 (PQ, J(PQ-
(BT) and H2 were carried out in high-pressure reactors in the
presence of catalytic amounts of the zwitterionic Rh(I) complex
[(sulphos)Rh(cod)] (1) (Scheme 1).16 In a typical homogeneous
reaction, the catalyst precursor and the substrate in a 1:100 ratio
are dissolved in methanol and the resulting solution is introduced
into the reactor under a stream of nitrogen. The reactor is
pressurized with hydrogen to 30 bar and then is heated to the
desired temperature with stirring.
The results obtained are reported in Table 1. At 200 °C for
17 h, no C-S bond scission of BT occurs, the only reaction
product being 2,3-dihydrobenzo[b]thiophene (DHBT) in 30%
yield (entry 1). Under these conditions, the slow decomposition
of the catalyst occurs, however, as shown by the formation of
a dark, insoluble material in a reaction lasting 70 h, which
converted only 50% of the substrate and also produced an
appreciable amount of ethylbenzene (EB) (7%) together with
traces of the hydrogenolysis product S-methyl-2-ethylthiophe-
nolate (METP) (Vide infra) (entry 2). The formation of EB is
totally suppressed when the reactions are carried out in the
presence of a large excess of elemental mercury, consistent with
a heterogeneous desulfurization reaction, most likely catalyzed
by rhodium particles.17
The use of a 1:1 mixture of methanol and water as solvent
(entry 3) slightly increases the formation of DHBT (38%) and
produces trace amounts of 2-ethylthiophenol (ETP) and EB. The
selective formation of DHBT also occurs when the reaction is
carried out at 175 °C under biphasic conditions (MeOH-H2O/
n-heptane) (entry 4). At this temperature, no appreciable
decomposition of the catalyst is observed, but the rate decreases
drastically (25% conversion in 110 h). The DHBT product and
the unreacted BT are exclusively found in the n-heptane phase,
while all the rhodium remains in the polar phase. Irrespective
of the phase variation, neither the rate of BT conversion nor
the selectivity changes when the stirring rate is increased from
650 to 1800 rpm.
1
Rh) ) 105.9 Hz); H NMR (MeOH-d4, 20 °C) δ 3.1 (m, vinyl CH-
CHH), 1.61 (m, vinyl CHH); 1H{31P} NMR (MeOH-d4, 20 °C) δ 3.16
(d, J(H3H2′) ) 8.6 Hz, J(H3H2) ) 7.4 Hz, vinyl CH), 3.08 (d, vinyl
CHH), 1.61 (d, vinyl CHH).
Catalyst Stability Tests. After a typical hydrogenolysis reaction
performed in MeOH-H2O/n-heptane for 5 h under typical conditions
(35 mg catalyst, 470 mg of BT, 180 mg of NaOH, 30 bar of H2, 160
°C), the reactor was depressurized and cooled to room temperature.
The reactor was disassembled under a positive pressure of hydrogen
for sampling and recharging (NaOH, 180 mg; BT, 470 mg), and then
it was repressurized with H2 to 30 bar and heated to 160 °C. After 5
h, the polar phase was separated and, after acidification, washed twice
with 10 mL of n-heptane to remove all ETP. The organic phases were
collected together and analyzed by GC and GC/MS, showing that 85%
of the overall amount of BT (940 mg) had been consumed. The acidic
MeOH-H2O phase, exclusively containing soluble rhodium species
and NaCl, was neutralized with NaOH and used for another catalytic
run under standard conditions. The analysis of the reaction products
showed a decrease of activity of only 30% notwithstanding the treatment
of the catalytic phase with excess HCl and its workup in air.
Attempted Hydrogenolysis of DHBT. A solution of 200 mg (1.5
mmol) of DHBT in 10 mL of n-heptane was added to a solution of 35
mg (0.035 mmol) of catalyst in 10 mL of MeOH. The mixture was
degassed with N2 and introduced into a degassed Parr reactor containing
60 mg (1.5 mmol) of NaOH. The reactor was pressurized with 30 bar
of H2 and then heated to 200 °C for 19 h. Analysis of the reaction
products by GC and GC/MS, after the usual workup, showed no
consumption of DHBT.
Base-Assisted Hydrogenolysis of Benzo[b]thiophene to
2-Ethylthiophenol. The addition of NaOH to the catalytic
system dramatically changes the selectivity of the reaction
between BT and H2 and also remarkably accelerates the
degradation of the substrate (Scheme 2).
Results
Hydrogenation of Benzo[b]thiophene to 2,3-Dihydrobenzo-
[b]thiophene. All the reactions between benzo[b]thiophene
In a first group of experiments, methanol was the solvent of

Liquid-Biphase Hydrogenolysis of Benzo[b]thiophene
J. Am. Chem. Soc., Vol. 119, No. 21, 1997 4949
Table 2. Base-Assisted Hydrogenolysis of Benzo[b]thiophene Catalyzed by [(sulphos)Rh(cod)] at e160 °Ca
reaction mixture composition (%)b
entry
solvent
base, mmol T (°C) t (h) BT DHBT METP ETP EB cat:base:BT rate ×10c
1
MeOHe
MeOH
MeOH
MeOH
MeOH
NaOH, 4.5
KOBut, 3.6
NaOH, 4.5
NaOH, 9.0
NaOH, 4.5
NaOH, 4.5
NaOH, 4.5
160
160
160
160
110
160
160
160
160
5
5
6
8
1
1
<1
<1
1
93
91
53
66
63
84
95
89
46
<1 1:130:100
<1 1:100:100
6.5
6.4
7.4
9.2
2
3d
4d
5
16
16
288
5
5
5
45
31
34
11
4
<1
1
1
1:540:840
1:1100:840
1:130:100
1
1
3
5
1
2
6
7
8
9
MeOH-H2O (1:1, v/v)
<1
<1
<1
<1
<1 1:130:100
5.9
6.6
6.2
3.2
MeOH/n-heptane (1:1, v/v)e
<1 1:130:100
MeOH-H2O/n-heptane (1:1:2, v/v/v)e NaOH, 4.5
MeOH-H2O/n-heptane (1:1:2, v/v/v)f NaOH, 4.5
8
2
1
1:130:100
1:130:100
5
51
2
a Reaction conditions: catalyst, 35 mg, 0.035 mmol; BT, 470 mg, 3.5 mmol; solvent, 20 mL; H2 pressure, 30 bar. b Key: benzo[b]thiophene
(BT), 2,3-dihydrobenzo[b]thiophene (DHBT), S-methyl-2-ethylthiophenol (METP), 2-ethylthiophenol (ETP), ethylbenzene (EB). c Average rate
expressed as (mmol of ETP) h-1 d Reaction conditions: catalyst, 8.2 mg, 0.008 mmol; BT, 940 mg, 7 mmol; solvent, 20 mL; H2 pressure, 30 bar.
.
e This reaction, repeated in the presence of a large excess of elemental mercury (Hg/Rh ) 2000), gave identical conversion of BT to ETP and
DHBT. f Reaction carried out in the presence of the surfactant Pluronic L101.
Figure 1. (a) Rate data for the hydrogenolysis of BT catalyzed by 1 in the presence of NaOH at 110 °C (MeOH, 30 bar of H2, NaOH:BT:1 )
130:100:1). (b) ln [BT] vs time.
choice (Table 2). Under the conditions employed in the
hydrogenation reactions but in the presence of a concentration
of NaOH equal to or slightly greater than that of the substrate,
almost all BT (94%) is already consumed after 5 h at 160 °C
(entry 1). The reaction gives the 2-ethylthiophenolate anion
and traces (e1%) of both DHBT and METP. Almost negligible
decomposition of the catalyst occurs as shown by the formation
of only trace amounts of EB (Vide infra). The hydrogenolysis
product may be recovered as the sodium salt Na[o-S(C6H4)-
C2H5]; for product identification and quantification, however,
it is convenient to transform the thiolate salt into ETP by
treatment with aqueous HCl. Accordingly, from now on, the
hydrogenolysis product is referred to as ETP. If the final
reaction mixture is not acidified, spontaneous oxidation of the
thiophenolate product to bis(2-ethylphenyl) sulfide takes place
upon exposure to air (Scheme 2).20 The use of other strong
bases such as KOBut (entry 2) or KOH affects neither the
selectivity nor the activity. Increasing the BT:1 ratio to 840
does not change the selectivity, but the reaction, although much
faster, does not go to completion within 16 h (entry 3) even
when a large excess of base is employed (entry 4).
Figure 1, the catalytic activity is maintained over 12 days and
a linear plot is obtained by plotting the ln [BT] vs time.
Neither the rate nor the selectivity of the reactions is affected
by the use of different solvent mixtures such as MeOH-H2O
(entry 6), MeOH/n-heptane (entry 7), or MeOH-H2O/n-heptane
(entry 8) in the range of stirring rates from 650 to 1800 rpm.
Similarly, there is no appreciable change of catalytic activity
for reactions carried out in the presence of an excess of elemental
mercury, while the addition of an external surfactant decreases
the hydrogenolysis rate (entry 9). Reactions in H2O/n-heptane
biphasic systems were unsuccessful, most likely due to the very
poor solubility of 1 and BT in water.
Interestingly, in the MeOH/n-heptane or MeOH-H2O/n-
heptane biphase reactions, the hydrogenolysis product is com-
pletely found in the polar phase as sodium 2-ethylthiophenolate
so as to leave the hydrocarbon phase almost completely
desulfurized. No trace of rhodium was found in the hydrocarbon
phase by atomic absorption spectrometry.
Increasing the reaction temperature to 200 °C remarkably
accelerates the conversion of BT to ETP irrespective of the
solvent or mixture of solvents employed (Table 3). Under
liquid-biphase conditions (entry 1), all BT is practically
consumed in 30 min with formation of ETP and of an
appreciable amount of EB (4%). The formation of the latter
product is ascribed to increased decomposition of the catalyst
at 200 °C. Indeed, when the reaction of entry 1 was carried
out in the presence of a large excess of elemental mercury, the
production of EB was almost totally suppressed. In the absence
of Hg, the concentration of EB increases with time (entry 2),
A reaction in methanol was carried out at significantly lower
temperature (110 °C, entry 5) at which quite long reaction times
are needed to achieve a high conversion (after 288 h, ETP 75%,
DHBT 3%). During this experiment, the reaction was periodi-
cally sampled (with negligible loss of internal pressure) and
the product formation was followed by GC. As is shown in
(20) March, J. AdVanced Organic Chemistry, 4th ed.; Wiley-Inter-
science: New York, 1992; pp 1199-1205.

4950 J. Am. Chem. Soc., Vol. 119, No. 21, 1997
Bianchini et al.
Table 3. Base-Assisted Hydrogenolysis of Benzo[b]thiophene Catalyzed by [(sulphos)Rh(cod)] at 200 °Ca
reaction mixture composition (%)b
entry
solvent
base, mmol
t (h)
BT
DHBT
MEPT
ETP
EB
1
2
3
4
MeOH-H2O/n-heptane (1:1:2, v/v/v)
MeOH-H2O/n-heptane (1:1:2, v/v/v)
MeOH
NaOH, 4.5
NaOH, 3.75
NaOH, 3.75
NaOH, 3.75
0.5
17
17
5
1
3
3
91
81
48
78
4
7
6
6
11
43
13
<1
<1
MeOH-H2O (1:1, v/v)
17
a Reaction conditions: catalyst, 35 mg, 0.035 mmol; BT, 470 mg, 3.5 mmol; solvent, 20 mL; H2 pressure, 30 bar. Partial destruction of the
catalyst was observed. b Key: benzo[b]thiophene (BT), 2,3-dihydrobenzo[b]thiophene (DHBT), S-methyl-2-ethylthiophenol (METP), 2-ethylthiophenol
(ETP), ethylbenzene (EB).
Table 4. Base-Assisted Liquid-Biphase Hydrogenolysis of Benzo[b]thiophene Catalyzed by [(sulphos)Rh(cod)]: Dependence on Catalyst and
Base Concentrationa
reaction mixture composition (%)b
[cat] (M × 103)
[base] (M × 10)
entry
[mol (L of aq phase)-1
]
[mol (L of aq phase)-1
]
BT
DHBT
2
ETP
EB
cat:base:BT
rate × 10c
1
2
3
4
5
6
7
8
9
3.52
2.76
2.02
1.00
3.52
3.52
3.52
3.52
3.52
4.50
4.50
4.50
4.50
3.52
2.24
1.42
0.72
0.10
8
29
43
72
18
24
50
65
84
89
70
56
27
81
74
47
33
16
1
1
1
1
1
2
2
1
<1
1:130:100
0.8:130:100
0.6:130:100
0.3:130:100
1:100:100
1:60:100
1:40:100
1:20:100
1:3:100
6.2
4.9
3.9
1.9
5.7
5.2
3.3
2.3
1.1
1
1
a Reaction conditions: BT, 470 mg, 3.5 mmol, 3.5 × 10-1 mol (L of org phase)-1; except entry 9 (KOBut) NaOH was employed as base;
solvent, 20 mL, MeOH-H2O/n-heptane (1:1:2, v/v/v); H2 pressure, 30 bar; temperature, 160 °C; time, 5 h. b Key: benzo[b]thiophene (BT), 2,3-
dihydro[b]thiophene (DHBT), 2-ethylthiophenol (ETP), ethylbenzene (EB). c Average rate expressed as (mmol of ETP) h-1
.
Table 5. Base-Assisted Hydrogenolysis of Benzo[b]thiophene Catalyzed by [(sulphos)Rh(cod)]: Dependence on Hydrogen Pressure and
Substrate Concentrationa
reaction mixture composition (%)b
[BT] (M × 10)
PH2 pressure
(bar)
entry
[mol (L of org phase)-1
]
BT
DHBT
ETP
EB
cat:base:BT
rate × 10c
1
2
3
3.5
3.5
3.5
3.5
3.5
3.5
3.5
2.1
4.5
5.6
5
10
20
30
30
50
70
30
30
30
61
33
22
8
39
66
77
89
94
82
68
94
84
80
<1
1
<1
1
2
3
3
1
1:130:100
1:130:100
1:130:100
1:130:100
1:130:100
1:130:100
1:130:100
1:130:60
2.7
4.6
5.4
6.2
6.6
5.7
4.8
3.9
7.6
8.9
1
2
4
5d
6
4
13
25
4
14
18
2
4
1
1
1
7
8
9
1
1
1:130:130
1:160:160
10e
a Reaction conditions: catalyst, 35 mg, 0.035 mmol, 3.52 × 10-3 mol (L of aq phase)-1; NaOH, 180 mg, 4.5 mmol, 4.5 × 10-1 mol (L of aq
phase)-1; solvent, 20 mL, MeOH-H2O/n-heptane (1:1:2, v/v/v); temperature, 160 °C; time, 5 h. b Key: benzo[b]thiophene (BT), 2,3-
dihydrobenzo[b]thiophene (DHBT), 2-ethylthiophenol (ETP), ethylbenzene (EB). c Average rate expressed as (mmol of ETP) h-1 d Reaction carried
.
out in the presence of excess elemental mercury (Hg/Rh ) 2000). e NaOH, 220 mg, 5.5 mmol, 5.5 × 10-1 mol (L of aq phase)-1
.
consistent with increased decomposition. At 200 °C, long
reaction times result also in the conversion of the 2-ethyl-
thiophenolate product to the disulfide METP (11% after 17 h)
that accumulates in the hydrocarbon phase. This compound is
formed by a noncatalytic reaction between the 2-ethylthio-
phenolate anion and MeOH whose rate increases with both the
temperature and the MeOH concentration (Scheme 2).21 Con-
sistently, the analogous reaction in pure methanol (entry 3) gives
a greater production of METP as compared to the reactions
under biphasic conditions or in a MeOH-H2O mixture (entry
4).
tration, and base concentration) were changed systematically
to obtain kinetic information. The results of this investigation
are summarized in Tables 4 and 5 and in Figure 2.
As shown in Tables 4 (entries 1-4) and 5 (entries 4 and
8-10), the hydrogenolysis reaction is apparently first-order in
both the catalyst and substrate concentration (Figure 2a,b). The
average rate (expressed as (mol of ETP) h-1) also increases
linearly with the base concentration as long as the base:BT ratio
is lower or equal to 0.6 (Table 4, entries 6-9) (Figure 2c). For
higher concentrations of base (entries 1 and 5), the rate tends
to flatten.
To avoid any decomposition of the catalyst as well as to
minimize systematic errors which may be connected with the
heating and cooling processes (ca. 40 min to reach the working
temperature and ca. 1 h to cool the reactor), all the subsequent
biphasic reactions were carried out under the following condi-
tions: temperature 160 °C, reaction time 5 h, solvent 1:1:2
(v/v/v) MeOH-H2O/n-heptane (20 mL). The other reaction
parameters (pressure of H2, catalyst concentration, BT concen-
The influence of the H2 pressure on the hydrogenolysis rate
is rather complicated (Table 5). In the range from 5 to 30 bar
(entries 1-4), the rate (Figure 2d) increases with the H2 pressure.
Slowing occurs at ca. 10 bar, however. The highest conversion
is observed at 30 bar. Above this pressure, the rate significantly
decreases (ca. 23% at 70 bar) with no apparent change of
selectivity (entries 6 and 7).
HPNMR Studies. When a MeOH-d4 solution of 1 is
pressurized with 30 bar of H2 in a 10 mm HPNMR tube in the
presence of a 10-fold excess of both BT and KOBut, no
(21) March, J. AdVanced Organic Chemistry, 4th ed.; Wiley-Inter-
science: New York, 1992; pp 406-411.

Liquid-Biphase Hydrogenolysis of Benzo[b]thiophene
J. Am. Chem. Soc., Vol. 119, No. 21, 1997 4951
Figure 2. Hydrogenolysis reactions of BT catalyzed by 1 at 160 °C. Dependence of the rate on the catalyst concentration (a), substrate concentration
(b), base concentration (c), and hydrogen pressure (d).
[(sulphos)Rh(η3-S(C6H4)CHdCH2)]- (2) (31P NMR AMQX
pattern with δ(PA) 31.3, δ(PM) -0.8, and δ(PQ) -9.8). The
concomitant action of H2 and BT finally gives the 2-ethyl-
thiophenolate complex [(sulphos)Rh(H)2(o-S(C6H4)C2H5)]-
(3) (31P NMR AM2X pattern with δ(PA) 37.2 and δ(PM) 1.8).
Traces of phosphine oxide (31.9 ppm) and of a fourth Rh species
of undefined composition (A3X pattern at 15.0 ppm) and with
no active role in the conversion of BT (see below) are also
detected.
Figure 3. 31P{1H} HPNMR study (sapphire tube, MeOH-d4, 81.01
At 60 °C for 1.5 h, no catalytic transformation of BT occurs
MHz) of the catalytic hydrogenation of BT in the presence of 1 and
as shown by 1H NMR spectroscopy. Conversely, BT is
catalytically hydrogenated to ETP (in the form of potassium
KOBut (30 bar of H2, BT:KOBut:1 ) 10:10:1). Spectrum at 20 °C (a),
after the NMR probe was heated to 60 °C for 1.5 h (b) and to 120 °C
2-ethylthiophenolate salt) and DHBT as the temperature is
for 2 h (c), and after the NMR probe was cooled to 20 °C (d). (*)
increased to 120 °C. After 2 h at this temperature, a GC/MS
Phosphine oxide. (O) Catalytically inactive, undefined species.
analysis gave 47% conversion of BT to hydrogenolysis and
transformation of either 1 or BT is observed (on the NMR time
scale) before a reaction temperature of ca. 60 °C is reached
(Figure 3, trace a).
At 60 °C (trace b), all 1 is rapidly hydrogenated (formation
of cyclooctene and cyclooctane) and the resulting metal fragment
can react with either H2 or BT or both. The reaction with H2
gives a rhodium(III) trihydride product, [(sulphos)RhH3]- (4)
(31P NMR A3X pattern at 22.8 ppm), while the reaction with
BT leads to the formation of the C-S insertion product
hydrogenation products. At 120 °C under catalytic conditions,
the only species visible on the NMR time scale are the trihydride
4 and the 2-ethylthiophenolate dihydride 3, the latter product
being formed in larger concentration than that of 4 (trace c).
After the HPNMR reaction was quenched by decreasing the
temperature of the probe head to 20 °C, a 31P{1H} NMR
spectrum was acquired. The spectral resolution at 20 °C is
sufficiently good to allow all coupling constants and the product
ratios (2:3:4 ) 33:48:19) to be determined (trace d). In

4952 J. Am. Chem. Soc., Vol. 119, No. 21, 1997
Bianchini et al.
triphos ) MeC(CH2PPh2)3),9c,d,13,22 the base is needed to
promote the heterolytic splitting of H2 and ultimately generate
in situ the [(sulphos)RhH]- fragment. The occurrence of a
heterolytic splitting path accounts for the formation of a Rh
complex with an odd number of hydride ligands (i.e., 4) in a
reaction that involves the addition of H2 to a Rh precursor devoid
of a hydride ligand (i.e., 1). Accordingly, the strong base may
be assigned a role as cocatalyst in the hydrogenolysis of BT
assisted by 1. In the absence of a base, 1 still reacts with H2
(30 bar, 60 °C, HPNMR experiment) to give cyclooctane and
a number of hydrido rhodium complexes which defied first-
order analysis.
Although most of the sulphos products spectroscopically
detected during the HPNMR experiments were not isolated in
the solid state, their unambiguous identification is possible
through a comparison of their NMR spectra with those of
analogous Rh complexes with the triphos ligand. This differs
from sulphos only in the group bonded to the bridgehead carbon
atom (CH3 vs CH2(C6H4)SO3-). Indeed, it has already been
found that the substitution of sulphos for triphos leads to
insignificant variations of the 31P NMR characteristics of
corresponding compounds.16 Thus, from a comparison of the
NMR spectra of 4, 2, and 3 with those of authentic samples of
[(triphos)Rh(H)3] (4*),19 [(triphos)Rh(η3-S(C6H4)CHdCH2)]
(2*),18 and [(triphos)Rh(H)2(o-S(C6H4)C2H5)] (3*),9c respec-
tively, one may realize that, two by two, these compounds are
essentially identical. On the other hand, this is quite reasonable
as one is reminded how the triphos complexes are prepared:
2* is synthesized by the thermolysis of the trihydride 4* in the
presence of BT,18 and reacts with H2 to give 3*.9c Moreover,
both 2* and 3* have been found to play a role in the
homogeneous hydrogenolysis of BT to ETP in THF: the former
as a catalyst precursor, the latter as the major Rh complex during
the catalysis and the Rh species that participates in the rate-
determining step (Vide infra).9c,13
Figure 4. 31P{1H} HPNMR study (sapphire tube, MeOH-d4, 81.01
MHz). Reaction of a 1:10 mixture of 1 and KOBut with 30 bar of H2
[spectrum at 20 °C (a); spectrum recorded at 20 °C after 3 h of heating
to 60 °C (b)]. Addition of a 10-fold excess of BT [spectrum recorded
at 20 °C after 1 h of heating at 100 °C (c)]. Reaction of the resulting
mixture with 30 bar of H2 [spectrum recorded at 120 °C after 40 min
(d); spectrum recorded after the NMR probe was cooled to 20 °C (e)].
(*) Phosphine oxide; (O) Catalytically inactive, undefined species.
comparison with the spectrum acquired at 120 °C (trace c), the
spectrum at 20 °C shows that the dihydride complex 3 is still
the major species but a large proportion of the 2-vinylthio-
phenolate product is also present. This means that 2 is formed
during all the cooling time of the reaction mixture, while the
reaction with H2 to give 3 does not occur below ca. 60 °C. As
already reported, this result is consistent with a lower energy
barrier to C-S insertion as compared to the hydrogenation of
the C-S inserted product.9c,d,13
A modeling study of independent reactions carried out in
HPNMR tubes has provided conclusive evidence for the
chemical relationships among 1, 4, 2, and 3, and between these
compounds and the catalytic hydrogenolysis of BT. The results
of this study are illustrated in Figure 4.
Discussion
Hydrogenation and Hydrogenolysis of BT. In combination
with a strong Brønsted base, the zwitterionic complex 1 reacts
with H2 to form an efficient catalyst for the hydrogenolysis of
BT to ETP. The reactions can be carried out in pure MeOH or
in biphasic systems comprising n-heptane as the organic phase
and either MeOH or MeOH-H2O as the polar phase. The
catalyst system is thermally robust and can easily be recycled
from the liquid-biphase reactions by phase separation. Small
quantities of DHBT (1-4%) are also produced in the course of
the reactions. As already reported for other related hydro-
genolysis reactions of BT in the homogeneous phase,9c,13 the
production of DHBT is due to a side catalysis cycle (see below).
Moreover, as shown by independent catalytic reactions, DHBT
does not react with H2 in the presence of the 1/NaOH system;
thus, any role of DHBT in the C-S bond scission reaction can
be ruled out.3-5
MeOH is a necessary cosolvent in the biphase reactions
because of the too low solubility of the catalyst precursor in
pure H2O. Since MeOH is a good solvent for BT, one may
not exclude that the reactions in MeOH-H2O/n-heptane may
occur in the polar phase instead of taking place at the phase
boundary as suggested for other ternary systems, e.g., the
hydrogenation of oct-1-ene in water/alcohol/oct-1-ene catalyzed
by Rh/TPPMS.23 On the other hand, 1, like other sulphos
In the presence of a strong base, 1 in MeOH-d4 (trace a) reacts
with H2 (30 bar) already at 60 °C to give the trihydride 4 (trace
b). After the HPNMR tube was depressurized and all H2 was
replaced with N2, a 10-fold excess of BT was introduced and
the spectrometer probe head was heated to 100 °C for 1 h. As
a result, the C-S bond cleavage of BT occurred with quantita-
tive conversion of 4 to 2 (trace c) (following this procedure, 2
has been isolated as potassium salt K2). The tube was
repressurized at room temperature with 30 bar of H2 and then
heated in the probe head. A reaction already occurred at 60
°C with transformation of the C-S insertion product 2 into both
the 2-ethylthiophenolate dihydride 3 and the trihydride 4. After
40 min at 120 °C (trace d), 3 and 4 were formed in a ca. 3 to
1 ratio. Consistent with the HPNMR study shown in Figure 3,
when the probe head was slowly cooled to room temperature,
the 2-vinylthiophenolate complex 2 reappeared (trace e).
The formation of some phosphine oxide in both HPNMR
experiments is most likely due to the unavoidable introduction
of traces of oxygen from air into the sapphire tube during the
charging and pressurizing operations. Likewise, one may not
exclude that the byproduct at 15.0 ppm is formed by a side
reaction with O2. Whatever the origin of this compound may
be, it is formed even in the absence of BT and reacts with neither
BT nor its hydrogenation products. Any role of this species in
the catalysis can thus be disregarded.
(22) Bianchini, C.; Meli, A.; Peruzzini, M.; Vizza, F.; Moneti, S.; Herrera,
V; Sa´nchez-Delgado, R. A. J. Am. Chem. Soc. 1994, 116, 4370.
(23) Hablot, I.; Jenck, J.; Casamatta, G.; Delmas, H. Chem. Eng. Sci.
1992, 47, 2689.
The formation of the trihydride 4 by treatment of 1 with H2
does require the presence of a strong Brønsted base. As
previously shown for the “(triphos)M” moieties (M ) Rh, Ir;

Liquid-Biphase Hydrogenolysis of Benzo[b]thiophene
J. Am. Chem. Soc., Vol. 119, No. 21, 1997 4953
organometallic complexes, exhibits inherent surfactant properties
that may lead to highly efficient emulsions.14,16 Indeed, we have
found that (i) irrespective of the phase system, comparable
reactions proceed with essentially identical rate and selectivity,
(ii) the addition of a surfactant does not increase the hydro-
genolysis rate,14,24 but a decrease eventually occurs, and (iii)
the reactions do not need very high stirring rates to ensure rapid
mass transfer. The hydrogenolysis rate of BT in MeOH-H2O/
n-heptane is actually independent of the stirring rate in the 650-
1800 rpm range. Accordingly, one can reasonably assume that
the reaction rates are determined by the chemical aspects
governing the catalytic process and not by the diffusion of the
substrate between the aqueous and organic phases.
Scheme 3
Scheme 4
The exclusion of physical phenomena in determining the rate
of conversion of BT is also suggested by the rate studies
illustrated in Figure 2. As indicated in this Figure (a and b),
the hydrogenolysis reaction is first-order with respect to the
concentrations of the catalyst precursor and substrate. A linear
correlation between the rate and the base concentration is also
observed at initial NaOH concentrations lower than 0.224 M
(NaOH to BT ratio of 0.6) (Figure 2c). At higher base
concentrations, the rate almost levels off, most probably because
of the too high ionic strength of the polar phase which may
disfavor the mixing of the phases as well as the diffusion of
BT.14,25 Less intuitive is the dependence of the rate on the H2
pressure illustrated in Figure 2d. The rate initially exhibits an
increase, ultimately reaching a maximum at 30 bar, after which
the rate decreases. A reasonable interpretation for this experi-
mental observable may be forwarded in light of the HPNMR
studies under catalytic conditions. These show that the propor-
tion of the active catalyst, i.e., the 16e- fragment [(sulphos)RhH]-
(see below), sequestrated as the trihydride 4, increases with
pressure.
In the absence of a strong Brønsted base, 1 still reacts with
H2 to form a catalyst for the hydrogenation of BT. Irrespective
of the phase variation, no opening of the thiophene ring occurs,
however, whereas the C2-C3 double bond is selectively reduced
to give DHBT. Due to the poor catalytic activity as well as
the severe experimental conditions to achieve an appreciable
production of DHBT, this hydrogenation reaction was not
studied by HPNMR spectroscopy. A 31P{1H} NMR spectrum
was acquired on a MeOH-d4 solution of 1 pressurized with 30
bar of H2. At 60 °C, all 1 disappeared to give several Rh
complexes, none of which was identified by comparison with
known triphos or sulphos complexes. Interestingly, the addition
of KOBut to this mixture transformed part of the products into
the trihydride 4. This experiment further confirms that the
strong base is necessary for the hydrogenolysis reaction to
occur: the base deprotonates eventual rhodium(III) polyhydrido
species (either classical or nonclassical) to give the rhodium(I)
monohydrido fragment which brings about the C-S insertion
step.9d,13,24,26 On the other hand, the rate study illustrated in
Figure 2c clearly shows that the role of the strong base cannot
exclusively be that of generating the catalytically active species.
It may be anticipated here that the base is of fundamental
importance to facilitate the reductive elimination of the hydro-
genolysis product from the metal center as commonly occurs
in a variety of base-assisted reactions.
acetone (thiophene ) thiophene, benzo[b]thiophene, dibenzo-
[b,d]thiophene, and dinaphtho[2,1-b:1′,2′-d]thiophene).9c,d,13 Ir-
respective of the thiophene, a general catalysis cycle has been
proposed in which η1-S coordination of the substrate is followed
by insertion of Rh into a C-S bond. The resulting thiacycle is
stepwise hydrogenated to give a dihydride thiolate complex that
ultimately eliminates the thiol product, thus regenerating the
[(triphos)RhH] catalyst.
Detailed mechanistic accounts for both the C-S insertion step
and the hydrogenation of the resulting product have been given
elsewhere.9c,d,13,18,27 Here, it may be useful to recall briefly that
the opening of BT proceeds by the regioselective insertion of
the 16e- fragment [(triphos)RhH] into the C2-S bond to give
a metallathiacycle hydride intermediate. This rearranges to the
2-vinylthiophenolate complex 2* by migration of the terminal
hydride to the R-carbon atom of the vinyl moiety of the thiacycle
(Scheme 3).18,27 Both the C-S bond scission and the hydride
migration reaction are low-energy processes occurring already
at room temperature. Conversely, the hydrogenation of the C-S
insertion product 2* to the thiolate complex 3* requires a high
pressure of H2 (>15 bar) and a temperature higher than 60 °C.9c
As shown by the HPNMR studies (Figures 3 and 4), an
analogous sequence of events occurs when sulphos is substituted
for triphos. Accordingly, the catalysis cycle proposed for the
hydrogenolysis of BT with [(triphos)RhH] can safely be
extended to the reactions performed with [(sulphos)RhH]-
(Scheme 4).9c,13 A catalytic cycle of this type is also accounted
for by the rate studies illustrated in Figure 2.
The reductive elimination of the thiol from the dihydride
thiolate intermediate has been proposed to be the rate-determin-
ing step in the homogeneous hydrogenolysis of BT catalyzed
by [(triphos)RhH] in aprotic organic solvents.9c,13 This as-
sumption can also be made for the reactions in polar media
herein described on the basis of the HPNMR experiments as
well as the rate studies, particularly the linear relationship
between the hydrogenolysis rate and the base concentration.
A few further comments are to be noted. The biphase
reactions are very selective, the production of DHBT being
generally less than 2%. As already reported for triphos
catalysis,9c the hydrogenation of BT to DHBT with monohydrido
rhodium(I) species can proceed through a secondary cycle (right-
Mechanistic Conclusions. In earlier work, we have shown
that the 16e- fragment [(triphos)RhH] is an efficient catalyst
for the homogeneous hydrogenolysis of different thiophenes to
the corresponding thiols in organic solvents such as THF or
(24) Buriak, J. M.; Osborn, J. A. Organometallics 1996, 15, 3161 and
references therein.
(25) Ding, H.; Hanson, B. E. J. Chem. Soc., Chem. Commun. 1994, 2747.
(26) Schrock, R. R.; Osborn, J. A. J. Am. Chem. Soc. 1971, 93, 2134.
(27) Bianchini, C.; Jime´nez, M. V.; Meli, A.; Moneti, S.; Vizza, F. J.
Organomet. Chem. 1995, 504, 27.

4954 J. Am. Chem. Soc., Vol. 119, No. 21, 1997
Bianchini et al.
Scheme 5
Conclusions
The data presented here define aqueous-biphase catalysis as
a viable technique for the purification of petroleum distillates
from residual thiophenic contaminants. Indeed, it has been
shown that relevant reactions in the HDS of thiophenes such
as the hydrogenolysis to thiols and the hydrogenation to
thioethers can effectively be carried out in a liquid-biphase
system with a metal complex exclusively soluble in the non-
hydrocarbon phase and thus easily recycled. The introduction
of the aqueous biphase technique to industrial HDS will require
an enormous amount of research work to be carried out however.
In particular, it will be necessary to design a class of water-
soluble catalysts containing relatively inexpensive metals (Co,
Ru) which are capable of tolerating the great thermal and
chemical stress of the hydrotreatment and possibly endowed
with inherent emulsifying attributes.
hand side of Scheme 4) in which the C2-C3 double bond of
BT is hydrogenated through the general mechanism of olefin
hydrogenation. In particular, the incorporation of deuterium
in the unreacted BT seen in the HPNMR experiment in MeOH-
d4 is consistent with an equilibrium between the η2-C,C-benzo-
[b]thiophene hydride complex and its hydride-migration dihy-
drobenzothienyl product. The large prevalence of hydrogenolysis
over hydrogenation can be attributed to steric effects9c as the
η1-S bonding mode of BT, precursor to C-S insertion,5c,28 is
less sterically demanding than the η2-C,C one, precursor to
double bond hydrogenation.29
The absence of a strong base in the catalytic mixture leads
to a completely different activity and selectivity. A diverse
mechanism is thus operative, which might well involve the
standard sequence of steps already proposed by Fish30 and
Sa´nchez-Delgado31 for the selective hydrogenation of BT to
DHBT catalyzed by polyhydrido rhodium(III) complexes in the
homogeneous phase (Scheme 5).
Acknowledgment. Thanks are due to the Progetto Strategico
“Tecnologie Chimiche Innovative”, CNR, Rome, Italy, for
financial support. V.S. (present address Institut fu¨r Anorga-
nische und Angewandte Chemie, Universita¨t Hamburg, Martin-
Luther-King-Platz 6, 20149 Hamburg, Germany) also thanks
the EC for Postdoctoral Grant CI1*.CT93-0329.
JA9641517
(30) (a) Fish, R. H.; Tan, J. L.; Thormodsen, A. D. J. Org. Chem. 1984,
49, 4500. (b) Fish, R. H.; Tan, J. L.; Thormodsen, A. D. Organometallics
1985, 4, 1743. (c) Fish, R. H.; Baralt, E.; Smith, S. J. Organometallics
1991, 10, 54. (d) Baralt, E.; Smith, S. J.; Hurwitz, I.; Horva´th, I. T.; Fish,
R. H. J. Am. Chem. Soc. 1992, 114, 5187.
(31) (a) Sa´nchez-Delgado, R. A.; Gonza´lez, E. Polyhedron 1989, 8, 1431.
(b) Sa´nchez-Delgado, R. A. In AdVances in Catalyst Design; Graziani, M.,
Rao, C. N. R., Eds.; World Scientific Publishing Co.: Singapore, 1991; p
214. (c) Sa´nchez-Delgado, R. A.; Herrera, V.; Rinco´n, L.; Andriollo, A.;
Mart´ın, G. Organometallics 1994, 13, 553.
(28) Dong, L.; Duckett, S. B.; Ohman, K. F.; Jones, W. D. J. Am. Chem.
Soc. 1992, 114, 151.
(29) (a) Choi, M.-G.; Robertson, M. J.; Angelici, R. J. J. Am. Chem.
Soc. 1991, 113, 4005. (b) Choi, M.-G.; Angelici, R. J. Organometallics
1992, 11, 3328. (c) Robertson, M. J.; Day, C. L.; Jacobson, R. A.; Angelici,
R. J. Organometallics 1994, 13, 179.