Hydrogenation of quinoline by rhodium catalysts modified with the tripodal polyphosphine ligand MeC(CH2PPH2)3

Article abstract of DOI:10.1002/1522-2675(20011017)84:10<2895::AID-HLCA2895>3.0.CO;2-0

As part of our modelling studies of the hydrodenitrogenation of N-heterocycles contained in raw oil materials, we investigated the selective hydrogenation of quinoline to 1,2,3,4-tetrahydroquinoline by rhodium catalysts modified with the tripodal polyphos

Full text of DOI:10.1002/1522-2675(20011017)84:10<2895::AID-HLCA2895>3.0.CO;2-0

Helvetica Chimica Acta ± Vol. 84 (2001)
2895
Hydrogenation of Quinoline by Rhodium Catalysts Modified with the
Tripodal Polyphosphine Ligand MeC(CH2PPh2)3
by Claudio Bianchini*, Pierluigi Barbaro, Michela Macchi, Andrea Meli, and Francesco Vizza
Istituto per lo Studio della Stereochimica ed Energetica dei Composti di Coordinazione, ISSECC-CNR,
Via J. Nardi 39, I-50132 Firenze (Tel.: ‡ 39-055-245-990; fax: ‡39-055-247-8366; e-mail: bianchin@fi.cnr.it)
Dedicated to the late Professor Luigi M. Venanzi, a friend, a master, and a pioneer in the organometallic
chemistry of polydentate ligands
As part of our modelling studies of the hydrodenitrogenation of N-heterocycles contained in raw oil
materials, we investigated the selective hydrogenation of quinoline to 1,2,3,4-tetrahydroquinoline by rhodium
catalysts modified with the tripodal polyphosphane ligand MeC(CH2PPh2)3. Experiments in standard
autoclaves and in high-pressure sapphire NMR tubes, kinetic and isotope labelling studies, and independent
reactions with isolated compounds have contributed to the elucidation of the catalytic mechanism as well as
identification of the electronic requisites of the metal catalyst for selective and efficient hydrogenation.
1. Introduction. ± Heteroatoms are currently removed from fossil fuels under
hydro-treating conditions in the presence of heterogeneous catalysts that must be
efficacious for four main chemical processes: hydrodesulfurization (HDS), hydro-
denitrogenation (HDN), hydrodeoxygenation (HDO), and hydrodemetallation
(HDM) [1][2]. In actual refinery reactors, HDN (Eqn. 1) occurs simultaneously to
all the other hydro-treating processes, but the reaction parameters are optimized for
the HDS reaction (Eqn. 2), and, therefore, the experimental conditions and the
catalysts employed are not specifically optimized for HDN, which requires longer
reaction times as well as higher temperature and H2 pressure to occur as efficiently as
HDS [1][3].
CaHbN ‡ c H2 ! NH3 ‡ CaHd Hydrodenitrogenation (HDN)
CaHbS ‡ c H2 ! H2S ‡ CaHd Hydrodesulfurization (HDS)
(1)
(2)
Nitrogen in petroleum and coal is contained in various organic compounds,
including polyaromatic heterocycles, aliphatic and aromatic amines, and nitriles. The
latter are in low quantity, and their degradation is efficiently performed under hydro-
treating conditions with application of commercial catalysts, while the aromatic
heterocycles, e. g., quinolines, pyridines, pyrroles, and indoles, are much more difficult
to degrade even than fused-ring thiophenes [1][3].
A relatively detailed mechanistic picture of the HDN of quinoline (Q) has been
obtained from the combined efforts of heterogeneous and homogeneous studies. The
former have been useful in rationalizing product distributions, kinetics, and selectiv-
ities, whereas homogeneous modelling studies have contributed to the elucidation of

2896
Helvetica Chimica Acta ± Vol. 84 (2001)
bonding modes and reactivity of coordinated Q, the elementary steps of its metal-
catalyzed hydrogenation, and the insertion of metal centers into either C N bond
[1][3 ± 15].
Quinoline may interact with the metal sites in two principal modes: end-on through
the N-atom and side-on through either ring or the CˆN bond [3]. However, in the
absence of steric congestion provided by a-substituents, the N-atom is used
preferentially for interaction with the catalyst.
The principal reaction pathways [1][3][4] proposed for the HDN of Q are shown in
Scheme 1. The most efficient reaction would proceed via the step sequence a) ! b) !
c), which does not involve the hydrogenation of the carbocyclic ring and, therefore,
consumes less H2 than any other path, and yields fuels with a higher octane rating due
the residual aromaticity in the denitrogenated products. Most of Q, however, undergoes
HDN over conventional catalysts via the alternative sequence of steps a) ! d) ! e) !
f), which involves the hydrogenation of both rings and, therefore, consumes twice as
much H2 to give either propylcyclohexane or ethylcyclohexane. The hydrogenation of
the heterocyclic ring is necessary to reduce the high energy required to cleave the C
N
bond that, in a typical N-heterocycle, is close to that of a CˆN bond [3a]. Since the
most efficacious HDN catalysts are also good hydrogenation catalysts of the
heterocyclic ring [1][3], a deep understanding of the elementary steps involved in
the hydrogenation mechanism of Q is of utmost importance for designing improved
HDN catalysts, possibly to drive the reaction through the pathway a) ! b) ! c). To this
purpose, homogeneous modelling studies with soluble metal catalysts might be very
helpful, especially to obtain specific information on the optimal electronic and steric
properties of the metal promoters and on the most appropriate reaction conditions for
selective hydrogenation.
Scheme 1
a)
b)
c)
+ NH3
N
N
N
H
NH2
d)
e)
f)
+ NH3
N
H
NH2
NH2
The first example of hydrogenation of Q to 1,2,3,4-tetrahydroquinoline (THQ) with
a soluble metal catalyst was reported by Jardine and McQuillin in 1970 [16]. Since then,
various transition metals (Mn, Fe, Co, Rh, Ru, Os) have been found capable of forming
catalysts for the selective conversion of Q to THQ in different phase-variation systems
that range from being truly homogeneous [4][7a,i,k,l][9a ± e,g], to aqueous-biphasic
[4][17] to heterogeneous single-site [4][18] (Scheme 2). In contrast to the relative
abundance of catalysts, very few mechanistic and kinetic studies of the selective
hydrogenation of Q have been reported so far [4][7a,i][9c,e]. Accordingly, the relative

Helvetica Chimica Acta ± Vol. 84 (2001)
2897
kinetic relevance of each catalytic step and the optimal catalyst composition to give
high productivity and selectivity are issues that need further investigation.
Scheme 2
H2, Cat
N
N
H
As part of ongoing studies of HDS catalysis by phosphane-modified late-transition-
metal catalysts [19], we have recently discovered that the tripodal tridentate ligand
MeC(CH2PPh2)3 (triphos) forms, in combination with Rh and Ru, either 16e or 14e
fragments of the formulas [M(H)(triphos)]n (M ˆ Rh, n ˆ 0; M ˆ Ru, n ˆ 1, ‡1) and
‡
[Rh(H)2(triphos)] . Irrespective of the metal, the RhI and Ru0 16e fragments
promote exclusively the hydrogenolysis reaction of thiophenes to thiols (C S insertion
followed by hydrogenation), while the 14e RuII complex is a selective catalyst for
hydrogenation to thioethers. No reaction was observed between the 16e RhIII
‡
fragment [Rh(H)2(triphos)] and thiophenes, which was attributed to the low
propensity of RhIII to bind S-heterocycles [20][21].
n
+
P
M
H
P
H
H
P
P
Rh
P
P
M = Rh, n = 0
M = Ru, n = –1, +1
Since HDS and HDN catalysis can occur simultaneously over the surface of
heterogeneous catalysts, we decided to investigate systematically the hydrogenation of
N-heterocycles with the same triphos precursors employed for the hydrogenation and/
or hydrogenolysis of thiophenes. In this paper, we show that the selective hydro-
genation of Q to THQ can be achieved effectively in homogeneous phase with catalyst
‡
precursors that generate the fragment [Rh(H)2(triphos)] . The results obtained
confirm that hydro-treating catalysts must contain metal sites with variable coordina-
tion numbers and variable oxidation states to be efficient for both HDS and HDN.
2. Results. ± 2.1. The Catalyst Precursors and Their Transformation in Hydrogen
Atmosphere. 2.1.1. Hydrogenation in Tetrahydrofuran (THF). From a formal view-
point, both [Rh(DMAD)(triphos)]PF6 (1) (DMAD ˆ dimethyl acetylenedicarboxy-
late (dimethyl but-2-ynedioate)) [21d][22a] and [Rh(S(C6H4)CHˆCH2-kC1',2',kS)(tri-
phos)] (2) [21j] contain RhI metal ions that are oxidized to RhIII upon oxidative
addition of H2 (Scheme 3) [21][22]. In THF, both 1 and 2 behave as precursors to the
‡
16e fragments [Rh(H)2(triphos)] and [RhH(triphos)], respectively, which differ
from each other in metal oxidation state and the number of terminal hydride ligands
[23]. The RhI fragment can be trapped by nucleophiles to give stable five-coordinate
complexes [24] or by electrophiles bearing C H, C S, and H H bonds to give
octahedral RhIII insertion products [21 ± 24].

2898
Helvetica Chimica Acta ± Vol. 84 (2001)
Scheme 3
2+
+
P
CO2Me
CO2Me
P
P
Rh
H
H
– DMSU
+
+
H2
H
H
P
> 40°
P
P
H
H
4
P
Rh
Rh
5
Rh
P
P
P
P
THF, r.t.
O
P
1
3
DMSU = dimethyl succinate
P
Rh
H
P
P
H2
THF, > 60°
H
P
P
P
P
Rh
S
Rh
P
H
P
– C6H4SH
S
2
‡
The plain transformation of [RhH(triphos)] into [Rh(H)2(triphos)] , which can be
‡
stabilized by THF, to give [Rh(H)2(THF-kO)(triphos)] (3) [25] can readily be
achieved by reaction with a Brùnsted acid. High-pressure NMR (HP-NMR) spectro-
scopy showed 3 to be thermally unstable in THF, where it is transformed into the
binuclear compound [{Rh(H)}2(triphos-k3P)2(m-H)2](PF6)2 (4) [23][25] and, to a much
lesser extent, into a new hydride-containing triphos-Rh complex (5), whose structure is
still obscure. In view of the propensity of Rh to form, in combination with triphos,
various binuclear polyhydride complexes [23][25], it is very likely that 5 is also a
binuclear species.
The hydrogenation of 2 to give free thiol C6H4SH and transient [RhH(triphos)] has
already been reported [21a,f,g].
2.1.2. Hydrogenation in THF in the Presence of Quinoline. A sequence of 31P{1H}-
HP-NMR spectra relative to the hydrogenation of 1 in a (D8)THF solution of Q
(220 mm) is reported in Fig. 1. No transformation of the precursor took place until the
tube was pressurized to 30 bar with H2.
Within the time necessary to acquire the first spectrum at room temperature, the
starting complex was completely converted to a mixture of Rh compounds, which have
been unambiguously identified by either comparison with authentic specimens or a
detailed spectroscopic analysis (Fig 1,b) (Scheme 4). The major products, which are
featured by NMR signals in the chemical-shift regions from 31 ± 35 ppm (coordinated
phosphine atoms) and from
26 ± 28 ppm (free phosphine atoms), constitute a
‡
family of geometric isomers of the general formula [Rh(Q-kN)2(triphos-k2P)] (6a ±
c). Occasionally, the formation of 6a ± c was accompanied by that of minor amounts
‡
of their phosphane oxide derivatives [Rh(Q-kN)2(triphosˆO-k2P)] (6x ± z)
(triphosˆO ˆ (Ph2PCH2)2C(Me)CH2PPh2O) (vide infra). The RhIII dihydride
[Rh(H)2(Q-kN)(triphos-k3P)]PF6 (7) and the known RhI monohydride
[RhH(DMFU-k2C2,3)(triphos-k3P)] (8) (DMFU ˆ dimethyl fumarate (dimethyl but-
2-enedioate)) [22b,c], were also formed in comparable amounts.
With time, the concentration of the bis-Q complexes increased at the expense of
that of 7 and became largely predominant only after the NMR tube was heated at 408
for 15 min (Fig. 1,c). In the meantime, the resonance due to 8 disappeared, and the
hydride-containing Rh complex 5 began to form [31P{1H}-NMR: d 27.1 (m, PA); 25.2 (m,
PM); 9.1 (m, PQ). 1H-NMR: d 8.55 (dm, J(H,P) ˆ 153, Rh H)]. A parallel 1H-NMR

Helvetica Chimica Acta ± Vol. 84 (2001)
2899
Fig. 1. 31P{1H}-HP-NMR Study (sapphire tube, (D8)THF, 81.01 MHz) of the catalytic hydrogenation of Q
(220 mm) in the presence of 1, (Q/1 40 :1). a) At r.t. and ambient pressure b) pressurized to 30 bar with H2 at r.t.,
c) after heating to 408 for 15 min, and d) after heating to 408 for 120 min.
Scheme 4
P
P
P
CO2Me
CO2Me
+
CO2Me
H2
P
+
+
P
+
Rh
P
H
H
P
P
N
P
Rh
+
+
Rh
THF, r.t.
Rh
P
N
P
P
N
H
1
N
(excess)
MeO2C
8
6a-c
7
investigation showed that the selective reduction of Q to THQ occurred only after
heating to 408. After 2 h heating at 408, ca. 40% of the initial Q was reduced to THQ,
and 31P{1H}-NMR spectroscopy showed the presence of 6a ± c (eventually accompanied
by minor amounts of 6x ± z) and of 5 (Fig. 1,d). The latter product was obtained also
by hydrogenation of 8 in THF in the absence of Q, and, like 8, proved to be inactive for
the hydrogenation of Q (vide infra). For this reason, the elucidation of its structure was
not considered a primary objective. Moreover, its formation under real autoclave
conditions was negligible.
The reactions illustrated in Scheme 4 are not difficult to rationalize as 1 is a
‡
precursor to the fragment [Rh(H)2(triphos)] (Scheme 3), which may actually fix
either one or two (via H2 elimination [21]) molecules of Q to yield 7 and 6a ± c,
respectively. Even the formation of the fumarate-hydride complex 8 is rather
straightforward as it is known that hydrogenation of 1 in THF is a stepwise process
‡
involving the intermediacy of the fumarate complex [Rh(DMFU-k2C2,3)(triphos)]

2900
Helvetica Chimica Acta ± Vol. 84 (2001)
[22a][23a]. The presence of a terminal hydride in 8 is reasonably due to the occurrence
of heterolytic splitting of H2 in the basic environment of the reaction mixture
containing a large excess of Q. For d6 metal complexes with triphos the occurrence of
H2 heterolytic splitting in basic media is a routine reaction, indeed [20c][21b,d,e]. In
the case at hand, a recent paper has shown that the hydrogenation of 1 in THF gives the
trihydride [Rh(H)3(triphos)] (9) in the presence of a base [21e].
In an attempt to mimick analogous autoclave reactions (see below), the hydro-
genation of 1 was studied by HP-NMR spectroscopy at lower concentrations of Q.
With an initial 70 mm concentration of Q, the following observations were made: i)
The dihydride complex 7 was the first and largely prevalent Rh product formed at room
temperature. ii) Appreciable formation of the bis-Q complexes occurred only when the
tube was heated to 408. iii) The reduction of Q started already at room temperature and
was complete after 1 h at 408. iv) The binuclear complex 5 became the predominant Rh
product when all Q had been converted to THQ.
When the concentration of Q was reduced further to 30 mm, the conversion of Q to
THQ occurred already at room temperature and was much faster in comparison to the
reaction with 70 mm Q. No formation of the bis-Q derivatives was detected by NMR at
any stage of the conversion of Q to THQ, which required only 30 min to be complete at
room temperature. Under these conditions, 5 and 7 were the predominant products,
while the formation of 8 was rather marginal.
In summary, the HP-NMR studies have shown that the hydrogenation rate of Q has
an inverse concentration dependence in the presence of 1. The two major products
visible on the NMR time scale in catalytic conditions contain Q-kN ligands. The
catalytically inactive RhI compound 8 is formed in an appreciable amount only when
the hydrogenation of Q does not occur or is very slow. Above 408, 8 apparently
degrades to 5. In no case was either 1,2-dihydroquinoline (1,2-DHQ) or 1,4-
dihydroquinoline (1,4-DHQ) detected as an intermediate in the reduction of Q to
THQ.
The hydrogenation of the 2-vinylthiophenolate complex 2 in the presence of Q has
been similarly studied by HP-NMR spectroscopy. However, no useful information was
obtained: in particular, neither reduction of Q nor formation of a Q complex was
observed, and the starting Rh complex underwent degradation to several compounds as
occurs in THF [21a,f,g].
2.1.3. Hydrogenation of the Catalyst Precursors in THF in the Presence of Triflic
Acid. Since the rate of hydrogenation of Q has been found to increase remarkably in
the presence of a Brùnsted acid (see below), the hydrogenation of the Rh pre-
cursor 1 was studied by HP-NMR spectroscopy in (D8)THF containing CF3SO3H
(TfOH).
A selection of 31P{1H}-HP-NMR spectra is reported in Fig. 2.
The precursor 1 was stable in the presence of 3 equiv. of TfOH under N2 (Fig. 2,a),
whereas it reacted already at room temperature when the tube was pressurized with
30 bar H2 (Fig. 2,b). After 1 h at 408, all of 1 had disappeared (Fig. 2,c). Formed in its
place was a mixture of two fluxional products featured by broad 31P{1H}-NMR
resonances centered at 42 and 13 ppm as well as a doublet of quartets at 6.29 ppm
1
in the H-NMR spectrum (J(HRh) ˆ J(HP)cis ˆ 7.5 Hz ; J(HP)trans ˆ 202.6 Hz). The
temperature of the probe-head was then decreased to 20, 0, and
308. The

Helvetica Chimica Acta ± Vol. 84 (2001)
2901
Fig. 2. 31P{1H}-HP-NMR Study (sapphire tube, (D8)THF, 81.01 MHz) of the hydrogenation of 1 in the presence
of TfOH (TfOH/1 3 :1). a) At r.t. and ambient pressure, b) under 30 bar H2 for 20 min at r.t., c) after heating
60 min at 408, d) after cooling to r.t., e) to 08, and f) to 308.
corresponding 31P{1H}-NMR spectra are shown in Fig. 2,d, e, and f, respectively. At
308, the dynamic process affecting the triphos-containing species was completely
frozen out, and two Rh complexes in a 1:2 ratio (11 and 12) were clearly visible, which
have been unequivocally identified through their independent synthesis as outlined in
Scheme 5.
Scheme 5
P
P
Rh
H
CF3SO3H
– H2
H
H
H
H
P
P
Rh
P
P
OSO2CF3
10
9
CF3SO3H
– H2
P
P
+
THF
THF
H
OSO2CF3
P
P
P
P
Rh
Rh
-
CF3SO3
H
OSO2CF3
OSO2CF3
12
11

2902
Helvetica Chimica Acta ± Vol. 84 (2001)
The products that were actually formed by hydrogenation of 1 in THF in the
presence of a threefold excess of TfOH are [RhH(OSO2CF3)2(triphos)] (11) and
[RhH(OSO2CF3)(THF)(triphos)] (12). The dihydride derivative [Rh(H)2(O-
SO2CF3)(triphos)] (10) was intercepted in the stoichiometric protonation of the
trihydride 9 used in the independent synthesis of 11 and 12. These two complexes are in
rapid equilibrium in room temperature THF solution (Scheme 5). The bis-triflate
complex 11 was isolated as pale yellow microcrystals by reacting 9 with three
equivalents of TFOH in THF, followed by addition of heptane. The spectroscopic
characterization of the new compounds 10, 11, and 12 was straightforwardly obtained
by routine NMR methods. In the slow-exchange regime (Fig. 2,f), 11 and 12 exhibit
canonical 31P{1H}-NMR AM2X and AMQX patterns, respectively, with hydride
resonances at
6.23 ppm for 11 (dq, J(H,Ptrans) ˆ 196.2, J(H,Pcis) ˆ J(H,Rh) ˆ
7.8 Hz) and at 6.15 ppm for 12 (dq, J(H,Ptrans) ˆ 195.3, J(H,Pcis) ˆ J(H,Rh) ˆ 7.4 Hz ) .
The substitution of TsOH for triflic acid in the reaction with the trihydride 9 yielded
[RhH{OSO2(C6H4)p-Me}2(triphos)] analogous to 11, while the use of HBF4 ´ OEt2 led
exclusively to the formation of the binuclear tetrahydride 4.
Interestingly, the addition of Q to a (D8)THF solution of 11 resulted in the
complete transformation of the starting Rh complex into 6a ± c (Scheme 6). Consistent
with the results reported in Scheme 6, the hydrogenation of 1 in a (D8)THF solution
containing also 40 equiv. of Q and 10 equiv. of CF3SO3H gave selectively 6a ± c,
eventually contaminated by the oxidation product 6x ± z (HPNMR and autoclave
experiments). Half of the Q was selectively converted to THQ in ca. 60 min at 408
(Fig. 3). In the presence of an excess of strong protic acids like TfOH or TsOH, the
hydrogenation of 2 with or without added Q, gave identical results to analogous
hydrogenation reactions of 1, which is consistent with the similar catalytic activity
exhibited by the two precursors (see below).
Scheme 6
P
P
r.t.
+
Rh
P
P
N
OSO2CF3
H
P
+
Rh
– TfOH
P
N
N
OSO2CF3
(excess)
11
6a–c
2.2. Synthesis and Characterization of the Mono-Q and Bis-Q Complexes 7 and 6a ±
c/6x ± z. While the authentication of 7 was readily accomplished by comparing its
spectroscopic properties with those of a pure sample independently prepared by
treatment of a THF solution of [RhCl(H)2(triphos)] [22a] with TlPF6 in the presence
of Q, the characterization of 6a ± c, particularly when contaminated by its phosphine
oxide derivatives 6x ± z, required a long and accurate NMR analysis. Details of this
study are given below as they represent a paradigmatic case of the potential of
multidimensional NMR spectroscopy for the elucidation of complicated structures and/
or mixtures of products in solution.
Besides in situ HP-NMR and batch catalytic experiments, the mixture of geo-
metrical isomers denoted as 6a ± c was also synthesized by independent methods

Helvetica Chimica Acta ± Vol. 84 (2001)
2903
Fig. 3. 31P{1H}-HP-NMR Study (sapphire tube, (D8)THF, 81.01 MHz) of the catalytic hydrogenation of Q in the
presence of 1 and TfOH (Q/TfOH/1 40 :10 :1). a) at r.t. and ambient pressure, b) pressurized to 30 bar with H2
for 15 min at r.t., and c) after heating to 408 for 60 min.
involving the plain displacement of ligands from appropriate RhI-triphos precursors.
Two of these methods are reported in Scheme 7 and described in the Exper. Part. Slow
transformation of 6a ± c into 6x ± z was observed when a controlled amount of air was
allowed to react with a solution of 6a ± c contained in either a NMR tube or an
autoclave. In no case, however, was it possible to isolate either 6a ± c or 6x ± z in the
solid state, an excess of Q being necessary for their stabilization.
Scheme 7
TlPF6
Q (excess)
r.t.
P
P
H2
H2
P
4 + 5
7
Rh
Cl
–TlCl
P
+
Rh
P
P
N
P
O
N
+
Rh
P
P
N
P
+
Q (excess)
air
r.t.
P
Rh
N
6a-c
P
–1,3-COD
100°
6x–z
Consistent with the HP-NMR studies, the hydrogenation (30 bar H2) of a (D8)THF
solution of 6a ± c/6x ± z obtained by either method shown in Scheme 7, gave the
dihydride 7, and then a mixture of 4 and 5 when all Q was consumed.
A high-resolution 1H- and 31P{1H}-NMR study has been performed on a (D8)THF
solution of 6a ± c/6x ± z prepared in situ by reaction of 11 with excess Q.

2904
Helvetica Chimica Acta ± Vol. 84 (2001)
The 1D 31P{1H}- and 2D 31P{1H}-COSY NMR spectra show the sample to contain a
mixture of six Rh compounds denoted as 6a ± c, and 6x ± z. A small amount of free
triphos oxide was also detected. The molar ratio of the six complexes in the sample
studied was 6a/6b/6c/6x/6y/6z  6 :4 :2 :3 :2 :11). After the solution was exposed to air,
the isomeric product ratio within each series 6a ± b or 6x ± z did not vary, but the overall
concentration of 6x ± z increased at the expense of 6a ± c. 31P{1H}-NMR Data for all
compounds are reported in Table 1.
Table 1. 31P{1H}-NMR Data of Complexes 6 a)
Compound 6a
6b
6c
6x
6y
PY1 (2 P) ˆ 31.71 PZ1 (2 P) ˆ 33.42
PY2 ˆ 27.93 PZ2 ˆ 26.59
6z
Chemical PA1 ˆ 34.93
PB1 (2 P) ˆ 32.80 PC1 (2 P) ˆ 34.06 PX1 ˆ 33.80
shifts
PA2 ˆ 33.98
PB2
ˆ
26.01
PC2
ˆ
27.19
PX2 ˆ 31.91
PX3 ˆ 28.00
PA3
ˆ
26.28
ˆ 65
Coupling
constants 4JP
2JP
4JP
1JP
ˆ 3.0
4JP
1JP
ˆ 2.0
2JP
4JP
ˆ 64
ˆ 4JP
ˆ 3.5
4JP
1JP
ˆ 3.0
4JP
1JP
ˆ 2.0
A1,PA2
B1,PB2
B1,Rh
C1,PC2
C1,Rh
X1,PX2
Y1,PY2
Y1,Rh
Z1,PZ2
Z1,Rh
ˆ 4JP
ˆ 163
ˆ 162
ˆ 162
ˆ 162
A1,PA3
A2,PA3
X1,PX3
X2,PX3
ˆ 3.5
1JP
ˆ 163
1JP
ˆ 162
A1,Rh
X1,Rh
a) 161.98 MHz, (D8)THF, 294 K. Chemical shifts in ppm, coupling constants in Hz.
1
The six Rh complexes have been identified and their H-NMR characteristics
unequivocally assigned by a combination of 1D and 2D NMR techniques (see Exper.
1
Part). Selected H-NMR data are reported in Table 2. The compounds constitute a
family of geometric isomers (syn-syn, anti-anti, syn-anti) having square-planar
geometry and the general formula [cis-MeC(CH2P(J)Ph2)(CH2PPh2-kP)2Rh(Q-
‡
kN)2] , where J represents a pair of electrons for 6a ± c and an O-atom for 6x ± z
(Fig. 4). The dissociation of a phosphane arm of triphos in RhI complexes (arm-off
mechanism) with formation of square-planar compounds has several precedents in
either catalytic intermediates (e. g., in the hydroformylation of olefins catalyzed by
[RhH(CO)(triphos)] [22a]) or stable species (e.g., [Rh(triphos-k2P)(h6-C6H5BPh3)]
[26]).
In all compounds 6a ± c and 6x ± z, the Q molecules have a N metal bond as shown
by the chemical-shift difference of corresponding proton resonances in coordinated and
free Q2), and by the coupling of the H C(2) protons to the P-atoms in the trans
1
position with respect to the corresponding Q ligand (Table 1). Sections of the H-
1
TOCSY and H-31P correlation spectra are reported in Fig. 5. No NMR resonance
attributable to protons of either THQ or p-Q was detected.
In each complex, two P-atoms are bonded to the metal (1J(P,Rh) ca. 163 Hz) in cis
fashion with a 2J(P,P) coupling constant of ca. 64 Hz [27][28]. The third P-atom is not
1
Calculated from 31P{1H}- and H-NMR integration.
1
2
)
)
As compared to the corresponding resonances in free Q, the signals of the H C(2), H C(3), and H C(5)
nuclei in p-coordinated Q show a downfield shift, whereas the H C(5), H C(6), H C(7), and H C(8)
protons show an upfield shift. In N-coordinated Q, all the ring protons show a downfield shift. Typical
values for H C(2) and H C(8) are ca. ‡ 0.5 and ‡ 0.7 ppm, resp. [7b]. In 6a ± c/6x ± z, a univocal trend of
the chemical-shift variation of the H-atoms in coordinated and free Q is not easily distinguishable, likely
due to chemical shielding effects. However, a diagnostic downfield shift can be given for of the H C(8)
protons (‡1.9 ± ‡ 0.9 ppm).

Helvetica Chimica Acta ± Vol. 84 (2001)
2905
J
J
J
P2
P2
P3
'
8
8
8
H
H
H
H
N
H
N
H
N
P1
P1
H
P1
P1
H
P1
H
P2
2
Rh
Rh
2
2
Rh
N
N
N
H
'
2
2
2
H
H
8
8
8
cis - syn - syn
cis - anti - anti
cis - syn - anti
6b, 6y
6c, 6z
6a, 6x
J = 2e- for 6a, b, c J = O for 6x, y, z
6b
6y
6a
6x
6c
6z
Fig. 4. Schematic view of the bis-Q complexes 6a ± c, 6x ± z in the coordination plane of the metal. The dashed
lines represent the symmetry planes, while the dashed arrows represent the significant NOEs.
a
Table 2. Selected 1H-NMR Chemical Shifts [ppm] and Coupling Constants [Hz] of Complexes 6 )
Compound
Nucleus 6a
6b
6c
6x
6y
6z
b
c
b
b
b
c
b
b
H
C(2) 9.48 (dt)
)
8.89 (dt)
)
10.18 (d, 2 H)
)
9.73 (d, 2 H)
J(H,H) ˆ 5.0
)
9.54 (dt)
)
8.70 (dt)
)
10.19 (d, 2 H)
)
9.57 (d, 2 H)
J(H,H) ˆ 5.0
)
J(H,H) ˆ 5.0 J(H,H) ˆ 5.0 J(H,H) ˆ 6.0
J (H P) ˆ 2.5 J(H,P) ˆ 2.5
J(H,H) ˆ 5.0 J(H,H) ˆ 5.0 J(H,H) ˆ 6.0
J(H,P) ˆ 2.5 J(H,P) ˆ 2.5
d
d
d
d
d
d
d
d
H
H
H
C(4) 7.75
C(6) 7.01
)
)
7.71
6.66
)
)
7.80 d) (2 H)
7.33 d) (2 H)
9.02 (d, 2 H)
7.77 d) (2 H) 7.72
7.10 d) (2 H) 7.03
)
)
7.68
6.55
)
)
7.78 d) (2 H)
7.32 d) (2 H)
9.04 (d, 2 H)
7.75 d) (2 H)
7.01 d) (2 H)
9.40 (d, 2 H)
J(H,H) ˆ 8.5
e
e
C(8) 9.64 (d)
)
9.90 (d)
9.34 (d, 2 H)
J(H,H) ˆ 8.0
9.64 (d)
)
9.97 (d)
J(H,H) ˆ 8.5 J(H,H) ˆ 8.5 J(H,H) ˆ 8.0
J(H,H) ˆ 8.5 J(H,H) ˆ 8.5 J(H,H) ˆ 8.0
d
d
d
d
d
d
d
d
d
d
d
d
H
H
H
C(7) 7.84
C(5) 7.32
C(6) 7.26
)
)
)
8.08
7.48
7.36
)
)
)
7.30 d) (2 H)
7.25 d) (2 H)
7.04 d) (2 H)
7.52 d) (2 H) 7.81
7.38 d) (2 H) 7.36
7.14 d) (2 H) 7.24
)
)
)
8.12
7.50
7.38
)
)
)
7.29 d)e) (2 H) 7.59 d) (2 H)
7.29 d)e) (2 H) 7.40 d) (2 H)
7.07 d) (2 H)
7.20 d) (2 H)
a) 500.132 MHz, (D8)THF, 294 K. b) A coupling to P1 is apparent in the 1H-31P correlation. c ) A coupling to P2 is apparent in
d
the 1H-31P correlation. ) Partially masked by other proton resonances. Detected from the H-TOCSY spectrum. e) Over-
1
lapped resonances.
coordinated in 6a ± c (d 26.0/ 27.2) and oxidized to phosphine oxide in 6x ± z (d 26.6/
28.0). In complexes 6b, 6c, 6y, and 6z, the two coordinated arms of triphos and the two
1
Q ligands appear to be chemically equivalent in the 31P{1H}- and H-NMR spectra,
which indicates that, in THF solution at 218, these complexes adopt a time-averaged
preferred conformation in which a symmetry plane perpendicular to the coordination
plane of the metal bisects the P(1) P(1) and N N axes. This arrangement and the s
plane are schematically represented in Fig. 4. In contrast, 1D-NMR spectra show no
symmetry elements for the complexes 6a and 6x. Accordingly, the syn-anti structure can
unequivocally be assigned to 6a and 6x and, thus, each complex exists as a couple of

2906
Helvetica Chimica Acta ± Vol. 84 (2001)
Fig. 5. Section of the 1H-TOCSY spectrum (upper) and 2D 1H-31P correlation (lower) of the 6a ± c/6x ± z mixture
(500.13 MHz, (D8)THF, 218)
atropisomers. The two diastereotopic Q ligands in 6a and 6x show strong NOE
interactions between the H C(2) and H C(8) protons belonging to different Q
molecules (Fig. 4). A representative section of the 1H-NOESY spectrum of 6a ±c/6x±z
is reported in Fig. 6.
A skew conformation for the six-membered triphos-Rh chelate rings in these
complexes is ruled out by the presence of the reflection plane, whilst boat or chair
conformations are equally plausible [29][30]. Accordingly, irrespective of the
conformation, the phenylphosphino rings give rise to two couples of pseudo-equatorial
and pseudo-axial groups with respect to the coordination plane of Rh. The arrangement
of the phenylphosphino groups is shown below.
The ortho-protons of each phenylphosphino group have been assigned on the basis
1
of the 2D H-31P-NMR correlations. The ortho protons of the pseudo-equatorial Ph
moieties are distinguished from the pseudo-axial ortho Ph protons by their higher-field

Helvetica Chimica Acta ± Vol. 84 (2001)
2907
1
Fig. 6. Section of the H-NOESY spectrum of the 6a ± c/6x ± z mixture (500.13 MHz, (D8)THF, 218, tm 0.65 s).
Positive-phased (exchange) cross-peaks are represented by filled circles; negative phased (NOE) cross-peaks
are represented by open circles.
ax
P
ax
P
ax
P
ax
P
eq
eq
eq
eq
Rh
Rh
N
N
N
N
1
1
chemical shifts3). H-NOESY and H-ROESY spectra of the syn-syn isomers show
strong NOE interactions between the ortho protons of the axial Ph groups and the
H
C(8) protons of Q, which confirms that these protons lie on the same side of the
coordination plane of Rh (the adopted nomenclature refers to this arrangement). This
kind of interaction has not been seen for the H C(2) protons. The situation is just
opposite for the anti-anti isomers, so that one can easily identify 6b and 6y as the anti-
anti isomers, and 6c and 6z as the syn-syn isomers. Following this reasoning, the
H
C(8') and H C(2') protons (Figs. 4 and 5) may be identified as those belonging to
the syn Q ligand in 6a and 6x.
3
)
This assignment has been made on the basis of the stronger shielding effect exerted by the Q ligands on the
pseudo-equatorial phenyls in comparison to the pseudo-axial Ph protons, as they lie ꢀoutꢁ of the
coordination plane of the metal. Similar behavior has been reported for comparable complexes, see, e.g.,
[30 ± 33]. The 1H-NMR chemical shifts of the s-phenylphoshino protons in 6a ± c/6x ± z are available upon
request from the authors.

2908
Helvetica Chimica Acta ± Vol. 84 (2001)
The geometric isomerism in these square-planar bis-Q complexes is apparently due
to hindered rotation of the Q ligands about the Rh N axis because of their mutual
interaction and the presence of the phenylphosphino ligands. The 6a ± c isomers (as well
as 6x ± z) exhibit a slow-exchange-motion regime, as demonstrated by the 1H-ROESY
1
and H-NOESY spectra (Fig. 6) [34][35]. Consistent with the interconversion of the
isomers, the proton-exchange pattern observed for the 6a, 6b, and 6c isomers can be
illustrated as in Scheme 8 (obviously, a similar pattern can be proposed for 6x ± z).
Scheme 8
H–C(2)a
H–C(2)c
H–C(8)c
H–C(2)b
H–C(8)b
H–C(2')a
H–C(8)a
H–C(8')a
No direct-exchange peak relating the Q protons of 6b to those of 6c (nor those of
6y to those of 6z) was observed. This means that the conversion of one isomer to the
other takes place exclusively upon rotation of only one Q ligand at a time, i.e., no
simultaneous rotation of the two coordinated Qs about the Rh N bonds occurs.
The observed concentrations of the various isomers in solution (syn-anti > anti-
anti > syn-syn) can be rationalized in terms of their relative stabilities. The intra-
molecular steric strain that differentiates the three isomers is occasioned by two types
of repulsive interactions involving either the axial phenyls and the syn quinoline ligands
or the two coordinated quinolines. In this picture, the syn-syn isomers are the least
stable, as they can undergo both interactions. In the anti-anti isomers, the repulsion with
the axial Ph groups is minimized, but the Q Q interactions are still present. The latter
interactions are evidently the most destabilizing ones as the isomer concentration
decreases in the order syn-anti > anti-anti.
2.3. Catalytic Hydrogenation of Q. Complex 1 is an efficient and selective
homogeneous catalyst for the hydrogenation of Q to THQ. As shown in Table 3, a
1
turn-over frequency (TOF) of 98 (mol product (mol cat  h) ) was obtained already
at 608 (Entry 2), while at 408, the TOF decreased to 43 (Entry 1). Substitution of 7 for
1 gave a much higher conversion at 408 (Entry 3) and a comparable result at 608
(Entry 4). To produce high conversion even at 408 with 1, some triflic acid was added to
the mixture (Entry 5). However, the addition of too much acid, as in Entries 6 and 7,
was found to inhibit the hydrogenation activity due to two possible negative effects: i)
excess triflate anions can compete with Q for coordination to Rh (thus disfavoring the
‡
conversion of 10 to 7), and ii) Q can be protonated at the N-atom to give QH that, as
‡
shown in Entry 8, is not easily hydrogenated. In actuality, QH has no way of binding
the metal center through the N-atom, which is crucial for hydrogenation [3], and the
small activity observed is apparently due to the equilibrium concentration of
nonprotonated Q. The beneficial effect of the acid on the catalytic activity of 1 at
408 is ascribed to the conversion of inactive RhI by-products that may form in the
catalytic mixture. As a matter of fact, the presence of an excess of TfOH makes the RhI

Helvetica Chimica Acta ± Vol. 84 (2001)
2909
a
Table 3. Hydrogenation of Q to THQ Catalyzed by Rhodium-triphos Complexes
)
Entry
Catalyst
TfOH [equiv.]
T [8]
Conversion [%]
1
2
3
4
5
6
7
8
1
1
7
7
1
1
1
1
2
2
2
8
8
0
0
0
40
60
40
60
40
40
40
40
60
60
60
60
60
43
98
94
98
98
83
30
28
< 1
77
98
0
0
5 ± 20
40
100
100 QH (OTf)
0
5
20
0
20
b
‡
)
9
10
11
12
13
95
a) Experimental conditions: [Cat] 0.022 mmol, Q/Cat ratio 100 : 1, THF 30 ml, pH2 30 bar, 1 h. b) Quinolinium
triflate.
precursors 2 and 8 as efficient as 1 in the hydrogenation of Q (Entries
9 ± 13).
NMR Analysis of the reaction mixtures collected at the end of selected catalytic
runs showed the formation of the bis-Q adducts 6a ± c, while the presence of 5 was, in
all cases, almost negligible. Occasionally, the formation of the phosphine oxide 6x ± z
was also observed. However, when air was purposefully introduced into the autoclave
prior to heating, the catalytic activity was almost totally inhibited.
2.4. Kinetics Studies of the Hydrogenation of Q in the Absence of Acid. To determine
the dependence of the hydrogenation rate on the concentrations of Rh, H2, and Q, a
series of reactions were carried out with different initial concentrations of catalyst and
reagents. The reactor contents were analyzed by GC every 10 min at constant
temperature (608) and internal pressure (Table 4). The H2 concentrations in the liquid
phase have been calculated on the basis of solubility data reported in the literature
[36].
a
Table 4. Kinetic Data for the Hydrogenation of Q with 1 as the Catalyst Precursor
)
1
[Rh] ´ 104 [m]
pH2 (psi)
[H2] ´ 102 [m]
[Q] ´ 102 [m]
k ´ 103 (min
)
7.33
7.33
7.33
7.33
7.33
7.33
7.33
7.33
3.67
5.50
11.00
115
115
115
115
115
58
230
460
115
115
115
2.99
2.99
2.99
2.99
2.99
1.39
6.10
12.83
2.99
2.99
2.99
1.83
3.67
5.50
7.33
9.17
7.33
7.33
7.33
7.33
7.33
7.33
15.47
7.71
5.89
3.92
2.84
1.92
9.11
19.21
1.89
2.96
6.14
a) Experimental conditions: 30 ml THF, 608, 1 h.

2910
Helvetica Chimica Acta ± Vol. 84 (2001)
A simple rate equation that accounts for the hydrogenation of Q with the catalyst
precursor 1 is:
d[Q]/dt ˆ d[THQ]/dt ˆ k[Rh]m[Q]n[H2]p
(3)
Assuming that the Rh and H2 concentrations do not vary over the time of each
reaction (this is correct for experiments in which the amount of dissolved H2 is much
larger than that consumed in the reduction of the substrate), the above equation can be
simplified as follows:
d[Q]/dt ˆ k[Q]n with k ˆ k[Rh]m[H2]p
(4)
In an attempt to determine the dependence of the hydrogenation rate on [Q], the
observed variations of [Q] vs. time in each experiment were plotted according to
models of second-order (1/[Q] 1/[Q]o), first-order (log([Q]o/[Q])), or zero-order
([Q]o/[Q]) dependence. For all the experiments reported in Table 4, the plots log([Q]o/
[Q]) vs. time gave straight lines with correlation coefficients as high as 99.9%.
Accordingly, it was concluded that, in the range of [Q]o reported in Table 4, the
reactions were first-order with respect to substrate concentration, and k was calculated
for each experiment from the slope of the straight line obtained. As an example, Fig. 7
shows a plot of log([Q]o/[Q]) vs. time for the reaction performed with the following
4
2
2
parameters: [Rh] ˆ 7.3 ´ 10 m, [H2] ˆ 2.99 ´ 10 m, and [Q]o ˆ 7.3 ´ 10 m. The first-
order dependence on [Q] was maintained within the time investigated during which
42% of Q was converted to THQ. A treatment of the experimental data in terms of
second-order dependence gave a better fit only for reactions carried out with [Q]o < 3 ´
10 2 m, while, for [Q]o > 7.3 ´ 10 2 m, straight lines with correlation coefficients close to
100% were obtained when zero-order dependence on substrate concentration was
assumed.
4
Fig. 7. Plot of log ([Q]o/[Q]) vs. reaction time for the hydrogenation of Q catalyzed by 1 ([Rh] ˆ 7.3 ´ 10 m,
[H2] ˆ 2.99 ´ 10 2 m, [Q]o ˆ 7.3 ´ 10 2 m, 608)
When, however, k was calculated for a series of reactions carried out with different
[Q]o it was found that the values, instead of being equivalent within the experimental
error, decreased exponentially with increasing [Q]o. The observed dependence of k on

Helvetica Chimica Acta ± Vol. 84 (2001)
2911
[Q]o clearly indicates that the hydrogenation of Q with 1 is a reaction whose rate law
changes with the initial concentration of the substrate, in particular the rate decreases
as [Q]o increases, which is consistent with the HP-NMR experiments. A typical reaction
of this type is the hydroformylation of olefins catalyzed by [RhH(CO)(PPh3)3] [37].
Much simpler reaction orders have been determined for [H2] and [Rh]. Experi-
ments carried out at H2 pressures of 4, 8, 15, and 30 bar were unequivocally consistent
with first-order dependence on [H2] (Fig. 8). Analogously, experiments with [Rh]o ˆ
4
4
4
3
3.67 ´ 10 , 5.50 ´ 10 , 7.33 ´ 10 , and 1.10 ´ 10 m showed first-order dependence on
[Rh] (Fig. 9).
Fig. 8. Dependence of the rate of hydrogenation of Q to THQ catalyzed by 1 on [H2]. Conditions are as in
Table 4.
Fig. 9. Dependence of the rate of hydrogenation of Q to THQ catalyzed by 1 on catalyst concentration.
Conditions are as in Table 4.
2.5. Kinetic Studies of the Hydrogenation of Q in the Presence of Triflic Acid. The
1
dependences of the hydrogenation rate expressed as mmol THQ h with respect to
[Rh] and [H2] are reported in Figs. 10 and 11. The corresponding experimental data
may be obtained as supplementary material from the authors.
While a first-order dependence can readily be assumed for both [Rh] and [H2], the
‡
rate dependence on [Q] and [H ] is very complicated due to the occurrence in the
‡
‡
4.9
reaction mixtures of acid-base equilibria involving Q (ka ˆ [Q][H ]/[QH ] ˆ 10
)
‡
‡
5
and THQ (ka ˆ [THQ][H ]/[THQH ] ˆ 10 ). From Figs. 12 and 13, which show the

2912
Helvetica Chimica Acta ± Vol. 84 (2001)
2
Fig. 10. Dependence of the rate of hydrogenation of Q (7.33 ´ 10 m) to THQ catalyzed by 1 in the presence of
TfOH (1.47 ´ 10 2 m) on catalyst concentration
Fig. 11. Dependence of the rate of hydrogenation of Q (7.33 ´ 10 2 m) to THQ catalyzed by 1 (7.33 ´ 10 4 m) in the
presence of TfOH (1.47 ´ 10 2 m) on H2 pressure
4
Fig. 12. Dependence of the rate of hydrogenation of Q to THQ catalyzed by 1 (7.33 ´ 10 m) in the presence of
TfOH (1.47 ´ 10 2 m) on substrate concentration

Helvetica Chimica Acta ± Vol. 84 (2001)
2913
Fig. 13. Dependence of the rate of hydrogenation of Q (7.33 ´ 10 2 m) to THQ catalyzed by 1 (7.33 ´ 10 4 m) in the
‡
presence of TfOH on H concentration
‡
variation in rate as a function of [Q] and [H ], it may be concluded that low rates are
‡
‡
2
expected when either [H ] or [Q] are low. For example, the rate at [H ]in ˆ 1.47 ´ 10
m, increased by an order of magnitude by increasing [Q] from 1.47 ´ 10 2 m to 3.67 ´ 10
m, then the rate remained constant up to [Q]o ˆ 8 ´ 10 m. With a further increase in
2
2
[Q], the rate decreased.
2.6. Deuterium Gas Experiments. Catalytic deuteration of Q was allowed to proceed
to 40% conversion to THQ under D2 (15 bar) in the presence of the catalyst precursor 1
at 608 (100 :1 substrate to catalyst ratio). Analysis by GC/MS and 1H-, 2H-, and 13C{1H}-
NMR of the reaction mixture showed that the formation of deuterated THQ was
accompanied by that of D-enriched Q (Scheme 11). In particular, THQ was found to
contain 1.1 D-isotope at position 2 and 1 D-isotope each at positions 3 and 4. The
enrichment of deuterium at the 2-position in THQ is attributable to reversible H/D
exchange mediated by the metal before D transfer (see Scheme 9). According to
previous reports [7a,i,k], the observed line-width value of 15 Hz for the multiplets due
to both H C(3) and H C(4), might indicate the occurrence of cis-deuteration. The N-
atom was found to be nondeuterated as a result of rapid H/D exchange of the
deuterium in this position due to the presence of protic impurities during workup
[7a,i,k]. As regards the residual Q, 20% deuterium enrichment occurred at the
2-position as indicated by 1H-NMR integration and confirmed by the GC/MS analysis
showing a m/z 130 ion in ca. 20% relative abundance.
Scheme 9
D H
D2
(15 bar)
H
D
Cat 1
H(D)
H(D)
N
N
H
D

2914
Helvetica Chimica Acta ± Vol. 84 (2001)
2.7. Catalytic Hydrogenation of 1,2-Dihydroquinoline (1,2-DHQ). To obtain further
information on the mechanism of hydrogenation of Q by 1, experiments of competitive
reduction of Q and 1,2-DHQ (Scheme 10) have been carried out in sapphire HP-NMR
tubes. Details of the procedure employed are provided in the Exper. Part.
Scheme 10
H2, Cat 1
THF
N
N
H
H
Under experimental conditions fully comparable to those used in the autoclave
4
reactions ([Rh] ˆ 7.33 ´ 10 m; [Q]o ˆ 7.33 ´ 10 1 m; [1,2-DHQ]o ˆ 7.33 ´ 10 1 m), it was
found that 1,2-DHQ is hydrogenated to THQ ca. eight times faster than Q.
3. Discussion. - 3.1. The Mechanism of Hydrogenation of Q to THQ Catalyzed by 1.
In Scheme 11 is reported the mechanism we propose for the selective hydrogenation of
Q to THQ in the presence of 1 as univocally indicated by D2 experiments, in situ HP-
NMR spectra, kinetic studies, identification of catalytically relevant intermediates, and
reactions with independent Rh compounds or partially reduced N-heterocycles. The
proposed mechanism can be extended also to the reactions catalyzed by precursor 2 in
the presence of a slight excess of protic acid, as the main role of the latter is exclusively
‡
to convert catalytically inactive [RhH(triphos)] to active [Rh(H)2(triphos)] .
Prior to any catalytic step, the precursor 1 is hydrogenated to give dimethyl
‡
succinate and the 14e fragment [Rh(triphos)] . This electronically unsaturated system
coordinates a Q molecule to attain the 16e configuration that enables Rh to bring
about the oxidative addition of H2 (Steps a ± b). The RhIII dihydride 7 has been actually
intercepted under standard catalytic conditions (608, [Q] ꢀ 70 mm). In the presence of
a larger excess of Q, the coordination of a second Q molecule competes with H2
oxidative addition and the bis-Q complexes 6a ± c may form (Step c). When this
happens, the hydrogenation rate is slowed as the number of catalytically active Rh-
atoms decreases.
Under the experimental conditions employed, 7 largely predominates over 6a ± c,
and the hydrogenation of the C(2)ˆN bond of Q is fast. It has been amply
demonstrated that a change in the Q bonding mode from kN to (1,2-h) is required to
accomplish the conversion of Q to 1,2-DHQ (Step d) [3][7 ± 9]. In principle, the
hydrogenation of Q may also proceed through the intermediate 1,4-DHQ [9e]. In our
case, this alternative reaction path has been ruled out by the lack of deuterium
incorporation in the 4-position of the residual Q as well as the presence of one
deuterium label only in the 4-position of THQ.
The reduction of the C(2)ˆN bond would occur by transfer of hydride from Rh to
C(2), which, according to charge density studies [9c], is the most electrophilic site in Q
(Step e). A reductive elimination step involving the residual hydride ligand and the
amide N-atom ultimately generates 1,2-DHQ (Step f). This overall process is reversible
in nature as shown by the deuterium enrichment at C(2) of both unreacted Q and THQ,
as well as by HD formation. The 1,2-DHQ formed is a cyclic olefin and, as such, can be

Helvetica Chimica Acta ± Vol. 84 (2001)
2915
Scheme 11
N
H
H
H
P
[Rh]
+
[Rh]
=
+
P
P
Rh
[Rh]
N
N
[Rh]
N
H2
b)
H
a)
H
[Rh]
j)
N
P
c)
H
N
H
P
d)
H
P
H
H
[Rh]
H
[Rh]
Rh
+
N
H
H
N
N
N
H
i)
e)
H
[Rh]
N
H
[Rh]
N
H
H
f)
H
H
h)
[Rh]
N
[Rh]
N
H
H2
H
H
g)
H
‡
hydrogenated by the [Rh(triphos)] system through a classical olefin hydrogenation
mechanism (Steps h ± j) [22]. No further deuterium enrichment, aside from the
expected D-atoms at C(3) and C(4), was found, and, therefore, it can be concluded that
the reduction of the C(3)ˆC(4) bond of 1,2-DHQ is irreversible in nature.
In the absence of a quantitative treatment of the kinetic data, the rate-limiting step
can be identified only through a perusal of indirect experimental evidence. The first
substantial indication that the rate-limiting step had to be sought in the conversion of Q
to 1,2-DHQ was provided by the much faster rate of the irreversible hydrogenation of
1,2-DHQ to THQ in comparison with the overall transformation of Q into THQ. In
standard catalytic conditions ([Q] ꢀ 70 mm), the dihydride 7 is the largely predominant
Rh complex (at least on the NMR time scale), and the rate is first-order in [Rh],
[H2], and [Q]. From a mechanistic viewpoint, this means that all the reactions in-
volving the Rh complex, H2, and Q should take place reversibly before or within the
rate-determining step. The reversible migration of the hydride from Rh to C(2)
(Step e), which also disrupts the aromaticity of the substrate, possesses all of these
features and thus, in our opinion, is the best candidate for the rate-determining
step.
The mechanism of reduction of Q to THQ by a homogeneous metal catalyst has
been studied in a few cases only, which reflects the inherent difficulty of disrupting the
aromaticity of N-heterocycles [7a][9e,c].

2916
Helvetica Chimica Acta ± Vol. 84 (2001)
The mechanism proposed in Scheme 11 shares some analogies with that reported by
Fish and co-workers for the selective hydrogenation of Q to THQ in CH2Cl2 catalyzed
by the RhIII precursor [Rh(MeCN)3Cp*]2‡ (408, 33 bar H2) [7a]. Common features are
the critical role of N-atom coordination for selective heterocycle reduction, the cis-
deuteration of the C(2) C(3) bond of Q, and the reversibility of the C(2)ˆN bond
hydrogenation, which, in both systems, has been named the rate-determining step.
Unlike the triphos-based catalyst, the Cp* system shows a reversible C(3)ˆC(4) bond
reduction as well as Rh complexation to the carbocyclic ring resulting in C(6) H and
C(8) H bond activation, which reflects the more sterically congested nature of triphos
as compared to Cp*.
4. Conclusions. ± In this and previous papers, we have investigated the hydro-
genation of N- and S-heterocycles in the presence of homogeneous catalysts containing
the promoter metals Rh and Ru supported by the tripodal ligand triphos. It has been
found that the highly energetic Cs-symmetric 16e RhI and Ru0 fragments [RhH(tri-
phos)] and [RuH(triphos)] insert into C S bond of thiophenes, leading ultimately to
catalytic hydrogenolysis to thiols [20c][21b,d,g] while they are inactive towards N-
heterocycles. The much higher bond energy of C N bonds as compared to C S bonds
1
(by 3 ± 9 kcal mol ) certainly contributes to make C N insertion much more difficult
to accomplish than C S insertion. On the other hand, the Rh and Ru electron-rich
systems are also unable to catalyze the plain hydrogenation of N-heterocycles to cyclic
amines. In our opinion, the soft character of the metal center in [RuH(triphos)] and
[RhH(triphos)] disfavors N-atom coordination that is crucial for the hydrogenation of
N-heterocycles [3][7]. In contrast, the soft character of sulfur allows for the S-atom
coordination of thiophenes, hence for C S insertion, but disfavors the (2,3-h)
coordination that is necessary for the selective hydrogenation of the heterocycle
[20][21]. Consistent with this picture, catalysts containing hard metal ions, as in the
‡
RhIII fragment [Rh(H)2(triphos)] and in the recently investigated RuII fragment
‡
[RuH(triphos)] [20][21], are efficacious for the selective hydrogenation of both N-
heterocycles and S-heterocycles due to their propensity to bind the former via N- and
the latter via two C-atoms. The overall reactivity pattern for the prototypical substrates
BT and Q is summarized in Scheme 12.
In conclusion, besides providing further mechanistic insight into the selective
hydrogenation of the heterocyclic ring in Q, this study may contribute to ration-
alizations of why efficacious hydrotreating catalysts contain component and promoter
metal sites in variable coordination numbers and oxidation states [1][3][19]. Such
variability in chemical structure is actually needed to ensure the simultaneous reductive
degradation of as many different types of heteroatoms as possible. A clear-cut
rationalization of the relationship between the nature of the metal site and the
reactivity of a specific heteroaromatic substrate is lacking in the relevant literature, and
most of current improvements in catalyst composition are still achieved via trial-and-
error procedures. Homogeneous model studies, like the present one, may be very useful
to elucidate the structure/reactivity relationships by exploitation of the relative ease by
which the molecular architecture of homogeneous catalysts can be tailored to meet the
optimal steric and electronic requisites for either hydrogenation or hydrogenolysis of a
specific substrate.

Helvetica Chimica Acta ± Vol. 84 (2001)
2917
Scheme 12
H2
n
n
Q
BT
no C–N insertion,
no stable coordination
[M]
H
[M]
hydrogenolysis
S
M = Rh, n = 0; Ru, n = –1
S
+
+
BT
+
Q
+
+
S
[Ru]
[Ru]
H
[Ru]
H
[Ru]
[Ru]
N
N
H
H
H
H2
H2
hydrogenation
hydrogenation
+
+
H
H
BT
+
Q
H
[Rh]
[Rh]
no stable bonding interaction
P
[Rh]
H
H
H
N
N
[M] =
P
M
P
H2
hydrogenation
Thanks are due to MURST (Italy) for cofinancing this research (legge 95/95).
Experimental Part
General. All reactions and manipulations, except as stated otherwise, were routinely performed under N2
with standard Schlenk techniques. Reactions under a controlled pressure of H2 were performed with a stainless
steel Parr 4565 reactor (100 ml) equipped with a Parr 4842 temperature and pressure controller. The reactor was
connected to gas reservoir to maintain a constant pressure over the catalytic reactions. The Rh complexes
[Rh(DMAD)(triphos)]PF6 [22a], [Rh(S(C6H4)CHˆCH2-k2C1',2',kS)](triphos)] [21j], [RhH3(triphos)] [24b],
[RhCl(C2H4-k2C)(triphos)] [22a], [Rh(COD)(triphos)]PF6 [38], [RhH(DMFU-k2C2,3)(triphos)] [22b,c], and
[RhCl(H)2(triphos)] [22a] were prepared as previously described. The isolated metal complexes were collected
on sintered-glass frits and washed with appropriate solvents before being dried under a stream of N2. Q (98%,
Aldrich) was purified by distillation under N2. TfOH (98%), and TsOH ´ H2O (98%) were purchased from
Aldrich and used without further purification. 1,2-DHQ was prepared by reduction of Q with LiAlH4 in diethyl
ether according to [39]: [1H-NMR ((D8)THF, 208, 200.13 MHz): 4.28 (m, CH2); 4.82 (br., NH); 5.65 (dtd,
J(CHCH') ˆ 9.8, J(CHCH2) ˆ 3.8, J(CHNH) ˆ 1.6, CH); 6.30 (dt, J(CH'CH) ˆ 9.8, J(CH'CH2) ˆ 1.5, CH');
6.3 ± 7.0 (4 arom. H)]. The samples of 1,2-DHQ were invariably contaminated with some 1,4-DHQ (5 ± 10%).
THF and (D8)THF were purified by distillation from LiAlH4 under N2. Deuterated solvents for NMR
measurements were dried over molecular sieves. All of the other reagents and chemicals were reagent grade and
used as received from commercial suppliers.
GC Analyses: Shimadzu GC-14A gas chromatograph equipped with a flame ionization detector and a 30 m
(0.25 mm i.d., 0.25 mm film thickness) SPB-1 Supelco fused silica capillary column. IR Spectra: Perkin-Elmer
1600 series FT-IR spectrophotometer; samples as Nujol mulls between KBr plates, n in cm 1. Routine 1H-
(200.13 MHz), 13C{1H}- (50.32 MHz), and 31P{1H}-NMR (81.01 MHz) spectra were obtained on a Bruker ACP-
200 spectrometer. All chemical shifts (d) are reported in ppm relative to Me4Si referenced to the chemical shifts
of residual solvent resonances (1H, 13C) or 85% H3PO4 (31P), with downfield values reported as positive. 31P{1H}-
and 1H-NMR experiments on compounds 6a ± c were recorded on either a Bruker Avance DRX-500
spectrometer at 202.47 and 500.13 MHz, respectively, or a Bruker Avance DRX-400 spectrometer at 161.98 and
400.132 MHz, respectively. Both instruments were equipped with variable-temp.-control units accurate to Æ
0.18. The assignments of the signals were based on 1D spectra, 2D 31P{1H}-COSY, 1H-TOCSY, 1H-NOESY, 1H-

2918
Helvetica Chimica Acta ± Vol. 84 (2001)
ROESY and proton-detected 1H-31P correlations performed on nonspinning samples. 2D NMR Spectra were
recorded with pulse sequences suitable for phase-sensitive representations with TPPI. The 31P{1H}-COSY
spectrum was acquired at 161.98 MHz, whereas all the proton-detected 2D experiments were acquired at
500.13 MHz. The coupling constants were obtained from the 1D spectra and 1H{31P} heteronuclear decoupling
experiments. Standard pulse sequences were used for the 1H-TOCSY [40], 1H-NOESY [41], 1H-ROESY [42],
and 1H-31P correlations [43]. In a typical experiment, 1024 increments of size 2K (with 16 scans each) covering
the full range in both dimensions (ca. 6000 Hz for 1H and ca. 13000 Hz for 31P) were acquired with a relaxation
delay of 2.0 sec. ATOCSY mixing time of 170 ms was used, whereas the NOESY mixing time and ROESY spin-
lock time were 650 ms and 500 ms, resp. 1H-31P correlations optimized for the detection of J(H,P) couplings of 5
and 15 Hz were recorded. The 31P{1H}-COSY [44] spectrum was obtained with 1H decoupling during
acquisition: 1024 increments of size 2K (with 64 scans each) covering the full range in both dimensions (ca.
12000 Hz) were collected with a relaxation delay of 1.0 sec. For HP-NMR experiments, a 10-mm sapphire NMR
tube (Saphikon, Milford, NH) was used, and the Ti high-pressure charging head was constructed at the ISSECC-
CNR (Firenze, Italy) [45]. Caution: Since high gas pressures are involved, safety precautions must be taken at
all stages of studies involving high-pressure NMR tubes. GC/MS Analyses: Shimadzu QP-5000; column
identical to that used for GC analyses; m/z (%). Elemental analyses (C, H, N) were performed with a Carlo-
Erba Model 1106 elemental analyser. Atomic-absorption analyses were performed with a Perkin-Elmer 5000
instrument.
Reaction of 1 with H2 in (D8)THF. A 10 mm sapphire tube was charged with a (D8)THF (2 ml) soln. of 1
(22.3 mg, 0.022 mmol) under N2. The tube was pressurized with H2 to 30 bar at r.t. and then placed into a NMR
probe at 208. The reaction was followed by variable-temp. 31P{1H}- and 1H-HP-NMR spectroscopy. Selective
hydrogenation occurred already at r.t., yielding DMSU (1H-NMR: 3.72 (s, 2 MeO); 2.67 (s, 2 CH2)) and the
known THF adduct 3 [25]. Increasing the temp. to 408 led to the quantitative conversion of 1 to the binuclear
compound 4 and, to a lesser extent, to the hydride containing triphos-Rh complex 5 (31P{1H}-NMR: 27.1 (m, PA);
1
25.2 (m, PM); 9.1 (m, PQ). H-NMR: 8.55 (dm, J(H,P) ˆ 153, Rh H)].
Catalytic Hydrogenation of Q (ca. 200 mm) with the Catalyst Precursor 1. A 10 mm sapphire tube was
charged with a (D8)THF (2 ml) solu. of 1 (11 mg, 0.011 mmol) and a 40-fold excess of Q (25 ml, 0.44 mmol)
under N2. The tube was pressurized with H2 to 30 bar at r.t. and then placed into a NMR probe at 208. The
reaction was followed by variable-temp. 31P{1H}- and 1H-HP-NMR spectroscopy. A sequence of selected
31P{1H}-NMR spectra is reported in Fig. 1. The details of this investigation are discussed in a previous section.
Transformation of 1 occurred at r.t. only in the presence of H2 yielding the bis-Q complexes 6a ± c, as major
products, accompanied by the dihydride 7 and the known fumarate-monohydride 8 [22b,c]. Occasionally, minor
amounts of the phosphine oxide derivatives 6x ± z were detected in soln. Compounds 6a ± c, 6x ± z, and 7 were
identified by comparison of their 31P{1H} and 1H-NMR spectra with those of authentic specimens (see below).
Selective reduction of Q to THQ occurred only after heating to 408.
Catalytic Hydrogenation of Q (ca. 70 mm) with 1. A 10 mm sapphire tube was charged with a (D8)THF
(2 ml) soln. of 1 (11 mg, 0.011 mmol) and a ca. 13-fold excess of Q (17 ml, 0.15 mmol) under N2. The tube was
pressurized with H2 to 30 bar at r.t. and then placed into a NMR probe at 208. The reaction was followed by
1
variable-temp. 31P{1H}- and H-HPNMR spectroscopy. The spectra, immediately recorded at 208, showed the
conversion of 1 to the dihydride complex 7 and to the fumarate derivative 8, as a minor species. Hydrogenation
of Q to THQ already occurred at this temp. Increasing the temp. to 408 led to the conversion of 7 and 8 to the
bis-Q derivatives 6a ± c/6x ± z and also accelerated the hydrogenation of Q, which was complete in ca. 1 h. When
all Q was hydrogenated, the hydride species 5 became the predominant Rh compound in soln.
Catalytic Hydrogenation of Q (ca. 30 mm) with 1. A 10 mm sapphire tube was charged with a (D8)THF
(2 ml) soln. of 1 (11 mg, 0.011 mmol) and a ca. 5-fold excess of Q (7 ml, 0.06 mmol) under N2. The tube was
pressurized with H2 to 30 bar at r.t. and then placed into a NMR probe at 208. The reaction was followed by
variable-temp. 31P{1H}- and 1H-HP-NMR spectroscopy. The hydrogenation of Q to THQ was almost complete
within 30 min. In the meantime, 1 was rapidly converted to 7 and, to a much lesser extent, to 5. When all Q was
hydrogenated, only the hydride 5 was found in soln. During the course of the experiment, no formation of the
bis-Q complexes was observed.
In Situ NMR Study of the Catalytic Hydrogenation of 1,2-DHQ with the Catalyst Precursor 1 at the Same
Catalyst and Substrate Concentrations as in the Autoclave Experiments. In a typical NMR experiment, a 10 mm
sapphire tube was charged with a (D8)THF (2 ml) soln. of 1 (1.5 mg, 1.45 ´ 10 3 mmol) and a 100-fold excess of
1,2-DHQ (17 ml, 1.45 mmol) under N2 at r.t.. The tube was pressurized with H2 to 30 bar and then placed into a
NMR probe at r.t. (ca. 10 min after pressurization). The temp. of the probe was then increased to 608. As soon as
the temp. stabilized, 1H-NMR spectra were acquired every 5 min. Under these experimental conditions, all 1,2-

Helvetica Chimica Acta ± Vol. 84 (2001)
2919
DHQ disappeared to give a 3 :1 mixture of THQ and Q within 10 min as a consequence of both
disproportionation into Q and THQ, and hydrogenation to THQ. In an analogous experiment performed
with Q as the starting substrate, 10% Q was hydrogenated in 10 min. Finally, when the NMR tube was charged
with 100 equiv. of both Q and 1,2-DHQ, transformation of ca. 90% 1,2-DHQ occurred in 10 min. In all the
experiments, the small quantity of 1,4-DHQ disappeared under H2 already at r.t.
Attempted Catalytic Hydrogenation of Q with the Catalyst Precursor 2. A 10 mm sapphire tube was charged
with a (D8)THF (2 ml) soln. of 2 (12 mg, 0.014 mmol) and Q (25 ml, 0.44 mmol) under N2. The tube was
pressurized with H2 to 30 bar at r.t. and then placed into a NMR probe at r.t.. The reaction was followed by
variable-temp. 31P{1H}- and 1H-HPNMR spectroscopy. No reaction occurred at temp. lower than 608. At higher
temp., the initially formed (o-ethylthiophenolate) dihydride complex [Rh(H)2(o-S(C6H4)C2H5)(triphos)] [21g]
underwent decomposition to several unknown species, as found in analogous experiments in performed THF
alone [21a,f,g], without hydrogenation of Q.
Synthesis of 6a ± c. NMR Experiments. a) A 5 mm NMR tube was charged under N2 at r.t. with a mixture of
30 mg of [RhCl(h2-C2H4)(triphos)] (0.04 mmol) and 13 mg of TlPF6 (0.04 mmol), and then with a (D8)THF
(0.9 ml) soln. of Q (18 ml, 0.16 mmol). The 31P{1H}- and 1H-NMR spectra of this sample showed the quantitative
formation of 6a ± c. Occasionally, minor amounts of the phosphine oxide derivatives 6x ± z formed in soln.
b) A 5 mm NMR tube was charged under N2 at r.t. first with 25 mg of [Rh(COD)(triphos)]PF6
(0.026 mmol) and then with a CD3OD (0.9 ml) soln. of Q (120 ml, 1 mmol). The tube was flame-sealed and
then placed into a NMR probe at r.t.. The reaction was followed by variable-temp. 31P{1H}- and 1H-NMR
spectroscopy. The starting Rh complex was found to convert to 6a ± c only when the probe head was heated to
1008; complete conversion occurred in ca. 1 h.
Oxidation of 6a ± c to 6x ± z in Air. Either air or O2 (0.5 ml) was added by syringe to a soln. of 6a ± c
(0.022 mmol) in (D8)THF (0.9 ml) prepared as described above. The formation of 6x ± z was almost
instantaneous and the complete transformation of 6a ± c occurred in ca. 30 min at r.t..
Reaction of 6a ± c with H2. A 10 mm sapphire tube was charged with a (D8)THF (2 ml) soln. of 6a ± c
(0.022 mmol), pressurized with H2 to 30 bar and then placed into a NMR probe at r.t. The reaction was followed
1
by variable-temp. 31P{1H}- and H-HP-NMR spectroscopy. At 408, 6a ± c were converted to 7 and THQ. With
time, 7 decomposed to give a mixture of 4 and 5.
Synthesis of 7. To a soln. of [RhCl(H)2(triphos)] (0.15 mg, 0.2 mmol) and Q (25 ml, 0.2 mmol) in 20 ml of
THF was added 70 mg of TlPF6 (0.2 mmol) under vigorous stirring. The mixture was allowed to stir for 1 h at r.t.
After TlCl was removed by filtration, the resultant soln. was concentrated to half the volume under vacuum.
Portionwise addition of heptane (40 ml) led to the precipitation of 7 as a pale yellow solid, which was filtered
and washed with pentane, yield 70%. 31P{1H}-NMR ((D8)THF, 208, 81.01 MHz): AM2X spin system, 42.2 (dt,
J(PM,PA) ˆ 23.4, J(PM,Rh) ˆ 77.9, PM); 13.5 (dd, J(PA,Rh) ˆ 111.3, PA). 1H-NMR ((D8)THF, 208, 200.13 MHz):
2
6.62 (second-order dm, AA'XX'YZ spin system, j J(H,PM) ‡ 2J(H,PM') jˆ 160.3, 2J(H,PA) ˆ 12.1, 1J(H,Rh) ˆ
8.5, 2 Rh H). Anal. calc. for C50H48F6NP4Rh: C 59.8, H 4.8, N 1.4; found: C 59.7, H 4.9, N 1.4.
Reaction of 1 with H2 in (D8)THF in the Presence of TfOH. A 10 mm sapphire tube was charged with a
(D8)THF (2 ml) soln. of 1 (22.3 mg, 0.022 mmol) and TfOH (6 ml, 0.07 mmol) under N2. The tube was
pressurized with H2 to 30 bar at r.t. and then placed into a NMR probe at 208. The reaction was followed by
variable-temp. 31P{1H}- and 1H-HP-NMR spectroscopy. A sequence of selected 31P{1H}-NMR spectra is
reported in Fig. 2. Compound 1 was found to be converted to two rapidly interchanging monohydride complexes
11/12, ca. 1:2. Details of this study are discussed in a previous section.
Data for 11: 1H-NMR (200.13 , 208):
308):
6.29 (dq, J(H,Ptrans) ˆ 202.6, J(H,Pcis) ˆ J(H,Rh) ˆ 7.5, Rh H)
(
6.23 (dq, J(H,Ptrans) ˆ 196.2, J(H,Pcis) ˆ J(H,Rh) ˆ 7.8, Rh H). 31P{1H}-NMR (81.01 MHz, 208):
AM2X spin system; 42.1 (br., PM); 12.3 (dt, J(PA,PM) ˆ 18.5, J(PA,Rh) ˆ 69.8, PA) ( 308): AM2X spin system;
40.4 (dd, J(PM,PA) ˆ 20.1, J(PM,Rh) ˆ 122.6, PM);
(200.13 MHz, 208);
11.5 (dt, J(PA,Rh) ˆ 70.3, PA). Data for 12: 1H-NMR
6.29 (dq, J(H,Ptrans) ˆ 202.6, J(H,Pcis) ˆ J(H,Rh) ˆ 7.5, Rh H); ( 308):
6.15 (dq,
J(H,Ptrans) ˆ 195.3, J(H,Pcis) ˆ J(H,Rh) ˆ 7.4, Rh H). 31P{1H}-NMR (81.01 MHz, 208): AM2X spin system; 42.1
(br., PM); 12.3 (dt, J(PA,PM) ˆ 18.5, J(PA,Rh) ˆ 69.8, PA); ( 308): AMQX spin system; 47.9 (ddd, J(PA,PM) ˆ
41.4, J(PA,PQ) ˆ 15.0, J(PA,Rh) ˆ 130.1, PA); 35.4 (ddd, J(PM,PQ) ˆ 23.6, J(PMRh) ˆ 121.5, PM);
J(PQ,Rh) ˆ 72.0, PQ).
12.8 (dt,
Reaction of 9 with TfOH. a) NMR at 208. A Teflon-capped resealable 5 mm NMR tube was charged under
N2 at r.t. with a solid sample of 9 (30 mg, 0.04 mmol). A soln. of 1 equiv. of TfOH (7.5 ml, 0.04 mmol) in
(D8)THF (0.9 ml) was then introduced by syringe. The solid immediately dissolved, with gas evolution. 31P{1H}-
and 1H-NMR spectra taken immediately after dissolution showed the formation of H2 and the disappearance of
9. Formed in its place were the dihydride complexes 3 as triflate salt and 10, together with the monohydrides 11

2920
Helvetica Chimica Acta ± Vol. 84 (2001)
and 12 as minor species. Further addition of 1 equiv. of TfOH caused the conversion of the dihydride complexes
to a mixture of 11 and 12. Substitution of TsOH for TfOH yielded the monohydride complex [RhH(O-
SO2(C6H4)p-Me)2(triphos)] analogous to 11, while the use of HBF4 ´ OEt2 led to the formation of dihydride
intermediates, such as 3 as tetrafluoroborate salt, that decomposed to 4 BF4.
2
Data for 10. 1H-NMR (200.13 MHz):
6.52 (second-order dm, AA'XX'YZ spin system, j J(H,PM) ‡
2
2J(H,PM') jˆ 179.6, J(H,PA) ˆ 1J(H,Rh) ˆ 9.2, 2 Rh H). 31P{1H}-NMR (81.01 MHz): AM2X spin system; 56.9
(dt, J(PA,PM) ˆ 22.5, J(PA,Rh) ˆ 150.2, PA); 2.1 (dd, J(PM,Rh) ˆ 79.6, PM).
Data for [RhH(OSO2(C6H4)p-Me)2(triphos)]: 1H-NMR (200.13 MHz):
6.25 (dq, J(H,Ptrans) ˆ 201.6,
J(H,Pcis) ˆ J(H,Rh) ˆ 7.5, Rh H). 31P{1H}-NMR (81.01 MHz): AM2X spin system; 43.2 (br., PM); 14.1 (dt,
PA).
b) NMR Experiment at 708. Neat TfOH (11 ml, 0.12 mmol) was added by syringe to a suspension of 9
(30 mg, 0.04 mmol) in (D8)THF (0.9 ml) contained in a Teflon-capped resealable 5-mm NMR tube under N2 at
708. The tube was rapidly transferred into a NMR probe precooled to 708. 31P{1H}- and 1H-NMR spectra,
immediately recorded at this temp., showed the quantitative conversion of 9 to the known nonclassical
dihydrogen complex [Rh(H)2(h2-H2)(triphos)](OSO2CF3) [25] and, to a much lesser extent, to the THF adduct
3 as triflate salt. Increasing the temp. to 308 led to the complete conversion of the dihydrogen complex to 3
OTf and, as minor species, 11 and 12. Within 60 min at this temp., the monohydride complexes 11 and 12 (1:2
ratio) became the only Rh species to be detected in soln.
Synthesis of 11. Into a stirred suspension of 9 (250 mg, 0.34 mmol) in THF (20 ml) was added a soln. of
TfOH (125 ml, 1.37 mmol) in THF (10 ml). The starting complex rapidly dissolved with H2 evolution. After
30 min, the resultant yellow-brown soln. was conc. to ca. 5 ml in vacuo. Portionwise addition of heptane (5 Â
3 ml) led to the precipitation of 11 as a microcrystalline pale yellow solid, which was filtered off and washed with
pentane. Yield 75%. IR: 2026 (Rh H). Anal. calc. for C43H40F6O6P3RhS2: C 50.3, H 3.9, Rh 10.0; found: C 50.5,
H 4.0, Rh 9.8.
Reaction of 11 with Q. Into a 5-mm NMR tube containing a (D8)THF (0.9 ml) soln. of 11 (22.6 mg,
0.022 mmol), 104 ml of Q (0.88 mmol) was added with a syringe at r.t. under N2. The 31P{1H}- and 1H-NMR
spectra of this sample showed quant. formation of 6a ± c as triflate salts, somewhat contaminated with 6x ± z.
Catalytic Hydrogenation of Q with the Catalyst Precursor 1 in the Presence of TfOH. A 10-mm sapphire tube
was charged with a (D8)THF (2 ml) soln. of 1 (11 mg, 0.011 mmol), TfOH (10 ml, 0.11 mmol), and a 40-fold
excess of Q (25 ml, 0.44 mmol) under N2. The tube was pressurized with H2 to 30 bar at r.t. and then placed into a
NMR probe at 208. The reaction was followed by variable-temp. 31P{1H}- and 1H-HP-NMR spectroscopy. A
sequence of selected 31P{1H}-NMR spectra is reported in Fig. 3. Transformation of 1 occurred at r.t. only in the
presence of H2 to give slowly the bis-Q complexes 6a ± c (Fig. 3,b). Increasing the temp. to 408 favored the
conversion process. Within 1 h, all of 1 had disappeared (Fig. 3,c), and 50% Q had been selectively
hydrogenated to THQ.
Catalytic Hydrogenation of Q. Autoclave Experiments. The reaction conditions and the results of these
experiments have been collected in Table 3. In a typical experiment, a 30 ml soln. of THF containing 0.022 mmol
of catalyst precursor, 2.2 mmol substrate, and the required amount of TfOH was introduced by suction into an
autoclave. After pressurizing with H2 to the desired pressure at r.t., the mixture was heated to the appropriate
temp. and then immediately stirred (750 rpm). During the reaction, the pressure level was kept constant with a
continuous feed of H2 from a high-pressure gas reservoir. After the desired time, the reaction was stopped by
cooling the autoclave to r.t., and the pressure was then released. The product composition was determined by
GC analysis of the crude reaction mixture. In the runs performed in the presence of acid cocatalysts, the added
amounts of acid were neutralized with THF/H2O solutions of NaOH before carrying out the GC analysis. In
selected runs, the final soln. was concentrated to dryness in vacuo and the residue was dissolved in (D8)THF and
studied by 1H- and 31P{1H}-NMR spectroscopy.
Autoclave Kinetics Experiments. The catalytic runs were performed according to a procedure analogous to
that reported above. The reaction conditions and the results of these experiments have been collected in Table 4.
To measure the substrate concentration as a function of time, the reactions were sampled every 10 min by
withdrawing a few ml of soln. through a liquid outlet valve connected with a dip needle-pipe. The samples were
maintained at 208 for GC analysis. Each run was repeated at least three times to ensure reproducibility of the
results. Concentrations of dissolved H2 were calculated on the basis of published solubility data [36].
Catalytic Deuteration of Q with the Catalyst Precursor 1. A soln. of 1 (22.3 mg, 0.022 mmol) and Q (266 ml,
2.2 mmol) in THF (30 ml) was placed into the Parr autoclave reactor under N2. The system was charged with
15 bar of D2 and heated to 608 with stirring. After 1 h, the reactor was cooled to r.t. and the excess D2 was vented.
The contents of the reactor were transferred into a Schlenk-type flask. GC Analysis showed 39% conversion of

Helvetica Chimica Acta ± Vol. 84 (2001)
2921
Q to THQ. The solvent was then evaporated under reduced pressure and the residue was redissolved in pentane
and passed through a bed of silica (pentane) to remove the catalyst. The soln. was then concentrated to dryness
and the product was found to contain D-enriched Q and (D3)THQ by 1H-, 2H-, and 13C{1H}-NMR spectroscopy
and GC/MS.
Residual Q. NMR Spectroscopy showed a 20% deuterium enrichment at the 2-position in the residual Q,
which is in agreement with the relative abundance of the isotopic species Q and (D1)Q (81 and 19%,
respectively) determined by GC/MS. 1H-NMR (CD2Cl2, 500.13 MHz): 8.99 (dd, J(2,3) ˆ 4.2, J(2,4) ˆ 1.6, 0.8 H,
H
C(2)); 8.22 (dd, J(4,3) ˆ 8.3, J(4,2) ˆ 1.6, H C(4)); 8.10 (dd, J(8,7) ˆ 8.1, J(8,6) ˆ 1.2, H C(8)); 7.90 (dd,
J(5,6) ˆ 8.1, J(5,7) ˆ 1.2, H C(5)); 7.80 (td, J(7,6) ˆ J(7,8) ˆ 8.2, J(7,5) ˆ 1.4, 0.9 H C(7)); 7.63 (td, J(6,5) ˆ
J(6,7) ˆ 8.1, J(6,8) ˆ 1.1,
H
C(6)); 7.45 (dd, J(3,2) ˆ 4.2, J(3,4) ˆ 8.3,
H
C(3)). 2H-NMR (CH2Cl2 ,
76.77 MHz): 8.99 (br. s, 2H C(2)). 13C{1H}-NMR (CD2Cl2, 125.77 MHz): 151.11 (s and t, J(CD) ˆ 24.0,
C(2)); 149.06 (s, C(8a)); 136.53 (s, C(4)); 130.07 (s, C(8)); 130.04 (s, C(7)); 128.97 (s, C(5a)); 128.52 (s, C(5));
127.28 (s, C(6)); 121.81 (s, C(3)). GC/MS (EI): 129 (100), 130 (33), 131 (2).
(D3)THQ. NMR Spectroscopy showed the presence of 1.1 D-atoms at C(2) and 1 D-atom each at C(3) and
C(4). Consistently, GC/MS analysis gave m/z values corresponding to the isotopic species (D3)THQ (89%),
(D4)THQ (11%). The N-atom was found to be undeuterated because of rapid H/D exchange due to contact
with moisture during workup. 1H-NMR (CD2Cl2, 500.13 MHz): 7.02 (m, H C(5), H C(7)); 6.65 (td, J(6,5) ˆ
J(6,7) ˆ 7.3, J(6,8) ˆ 1.2, H C(6)); 6.52 (dd, J(8,7) ˆ 7.8, J(8,6) ˆ 1.2, H C(8)); 3.31 (br. m, 0.9 H C(2)); 2.79
(br. m, H C(4)); 1.94 (br. m, H C(3)). 2H-NMR (CH2Cl2, 76.77 MHz): 3.31 (br. s, 1.1 2H C(2)); 2.79 (br. m,
2H C(4)); 1.94 (br. s, 2H C(3)). 13C{1H}-NMR (CD2Cl2, 125.77 MHz): 145.76 (s, C(8a)); 130.01 (s, C(5));
127.13 (s, C(7)); 121.89 (s, C(5a)); 117.19 (s, C(6)); 114.58 (s, C(8)); 42.10 (t, J(C(2),D) ˆ 20.9, C(2)); 27.19 (t,
J(C(4),D) ˆ 19.6, C(4)); 22.33 (t, J(C(3),D) ˆ 19.7, C(3)). GC/MS (EI): 136 (100), 137 (22), 138 (3).
REFERENCES
[1] a) H. Topsùe, B. S. Clausen, F. E. Massoth, ꢀ Hydrotreating Catalysisꢁ, Springer-Verlag, Berlin Heidelberg,
1996; b) J. Scherzer, A. J. Gruia, ꢀ Hydrocracking Science and Technologyꢁ, Marcel Dekker, New York,
1996; c) C. N. Satterfield, ꢀ Heterogeneous Catalysis in Industrial Practiceꢁ, McGraw-Hill, New York, 1991.
[2] M. L. Occelli, R. Chianelli, ꢀ Hydrotreating Technology for Pollution Controlꢁ, Marcel Dekker, New York,
1996.
[3] a) M. J. Girgis, B. C. Gates, Ind. Eng. Chem. Res. 1991, 30, 2031; b) T. C. Ho, Catal. Rev.- Sci. Eng. 1988, 30,
117; c) R. M. Laine, Catal. Rev.- Sci. Eng. 1983, 25, 459; d) J. R. Katzer, R. Sivasubramanian, Catal. Rev. -
Sci. Eng. 1979, 20, 155; e) B. C. Gates, J. R. Katzer, G. C. A Schuit, ꢀChemistry of Catalytic Processesꢁ,
McGraw-Hill, New York, 1979, Chap. 6.
[4] C. Bianchini, A. Meli, F. Vizza, Eur. J. Inorg. Chem. 2001, 43.
[5] R. Prins, M. Jian, M. Flechsenhar, Polyhedron 1997, 16, 3235.
[6] a) K. J. Weller, P. A. Fox, S. D. Gray, D. E. Wigley, Polyhedron 1997, 16, 3139; b) K. J. Weller, S. D. Gray, P.
Briggs, D. E. Wigley, Organometallics 1995, 14, 5588; c) S. D. Gray, K. J. Weller, M. A. Bruck, P. Briggs,
D. E. Wigley, J. Am. Chem. Soc. 1995, 117, 10678; d) S. D. Gray, D. P. Smith, M. A. Bruck, P. Briggs, D. E.
Wigley, J. Am. Chem. Soc. 1992, 114, 5462.
Â
[7] a) E. Baralt, S. J. Smith, J. Hurwitz, I. T. Horvath, R. H. Fish, J. Am. Chem. Soc. 1992, 114, 5187; b) R. H.
Fish, H-S. Kim, R. H. Fong, Organometallics 1991, 10, 770; c) R. H. Fish, R. H. Fong, A. Than, E. Baralt,
Organometallics 1991, 10, 1209; d) R. H. Fish, E. Baralt, H-S. Kim, Organometallics 1991, 10, 1965; e) R. H.
Fish, J. N. Michaels, R. S. Moore, H. Heinemann, J. Catal. 1990, 123, 74; f) R. H. Fish, H-S. Kim, R. H. Fong,
Organometallics 1989, 8, 1375; g) R. H. Fish, H-S. Kim, J. E. Babin, R. D. Adams, Organometallics 1988, 7,
2250; h) R. H. Fish, T.-J. Kim, J. L. Stewart, J. H. Bushweller, R. K. Rosen, J. W. Dupon, Organometallics
1986, 5, 2193; i) R. H. Fish, J. L. Tan, A. Thormodsen, Organometallics 1985, 4, 1743; j) R. H. Fish, A. D.
Thormodsen, H. Heinemann, J. Mol. Catal. 1985, 31, 191; k) R. H. Fish, J. L. Tan, A. Thormodsen, J. Org.
Chem. 1984, 49, 4500; l) R. H. Fish, A. Thormodsen, G. A. D. Cremer, J. Am. Chem. Soc. 1982, 104, 5234.
[8] a) R. M. Laine, New J. Chem. 1987, 11, 543; b) A. Eisenstadt, C. M. Giandomenico, M. F. Frederick, R. M.
Laine, Organometallics 1985, 4, 2033; c) R. M. Laine, J. Mol. Catal. 1983, 21, 119; d) R. M. Laine, D. W.
Thomas, L. W. Cary, J. Org. Chem. 1979, 44, 4964.
Â
[9] a) Y. Alvarado, M. Busolo, F. Lopez-Linares, J. Mol. Catal. A: Chem. 1999, 142, 163; b) M. Rosales, J.
Navarro, L. Sanchez, A. Gonzales, Y. Alvarado, R. Rubio, C. De la Cruz, T. Rajmankina, Transition Met.
Â
Chem. 1996, 21, 11; c) M. Rosales, Y. Alvarado, M. Boves, R. Rubio, H. Soscun, R. Sanchez-Delgado,

2922
Helvetica Chimica Acta ± Vol. 84 (2001)
Â
Transition Met. Chem. 1995, 20, 246; d) R. A. Sanchez-Delgado, J. Mol. Catal. 1994, 86, 287; e) R. A.
Â
Â
Sanchez-Delgado, D. Rondon, A. Andriollo, V. Herrera, G. Martin, B. Chaudret, Organometallics 1993, 12,
Â
Â
Â
4291; f) B. Chaudret, F. A. Jalon, M. Perez-Manrique, F. Lahoz, F. J. Plou, R. Sanchez-Delgado, New J.
Â
Chem. 1990, 14, 331; g) R. A. Sanchez-Delgado, E. Gonzalez, Polyhedron 1989, 8, 1431.
[10] a) A. J. Deeming, M. J. Stchedroff, C. Whitaker, A. J. Arce, Y. De Sanctis, J. W. Steed, J. Chem. Soc., Dalton
Trans. 1999, 3289; b) A. J. Arce, R. Machado, Y. De Sanctis, M. V. Capparelli, R. Atencio, J. Manzur, A. J.
Deeming, Organometallics 1997, 16, 1735; c) A. J. Arce, J. Manzur, M. Marquez, Y. De Sanctis, A. J.
Deeming, J. Organomet. Chem. 1991, 412, 177; d) A. J. Deeming, A. J. Arce, Y. De Sanctis, M. W. Day, K. I.
Hardcastle, Organometallics 1989, 8, 1408; e) K. I. Hardcastle, H. Minassian, A. J. Arce, Y. De Sanctis, A. J.
Deeming, J. Organomet. Chem. 1989, 368, 119.
[11] a) M. Rakowski DuBois, L. D. Vasquez, R. F. Ciancianelli, B. C. Noll, Organometallics 2000, 19, 3507; b) M.
Rakowski DuBois, Coord. Chem. Rev. 1998, 174, 191; c) L. Vasquez, B. C. Noll, M. Rakowski DuBois,
Organometallics 1998, 17, 976; d) S. Chen, B. C. Noll, L. Peslherbe, M. Rakowski DuBois, Organometallics
1997, 16, 1089; e) S. Chen, L. Vasquez, B. C. Noll, M. Rakowski DuBois, Organometallics 1997, 16, 1757;
f) M. Rakowski DuBois, K. G. Parker, C. Ohman, B. C. Noll, Organometallics 1997, 16, 2325; g) S. Chen, V.
Carperos, B. C. Noll, R. J. Swope, M. Rakowski DuBois, Organometallics 1995, 14, 1221; h) F. Kvietok, V.
Allured, V. Carperos, M. Rakowski DuBois, Organometallics 1994, 13, 60.
[12] a) T. S. Kleckley, J. L. Bennett, P. T. Wolczanski, E. B. Lobkowsky, J. Am. Chem. Soc. 1997, 119, 247; b) K. J.
Covert, D. R. Neithamer, M. C. Zonnevylle, R. E. LaPointe, C. Schaller, P. T. Wolczanski, Inorg. Chem.
1991, 30, 2494; c) D. R. Neithamer, L. Parkanyi, J. F. Mitchell, P. T. Wolczanski, J. Am. Chem. Soc. 1988,
110, 4421.
[13] a) D. P. Smith, J. R. Strickler, S. D. Gray, M. A. Bruck, R. S. Holmes, D. E. Wigley, Organometallics 1992,
11, 1275; b) J. R. Strickler, M. A. Bruck, D. E. Wigley., J. Am. Chem. Soc. 1990, 112, 2814.
[14] a) D. A. Vicic, W. D. Jones, Organometallics 1999, 18, 134; b) D. A. Vicic, W. D. Jones, Organometallics
1997, 16, 1912; c) W. D. Jones, L. Dong, A. W. Myers, Organometallics 1995, 14, 855; d) W. D. Jones, R. M.
Chin, R. M., J. Am. Chem. Soc. 1994, 116, 19826; e) G. C. Hsu, W. P. Kosar, W. D. Jones, Organometallics
1994, 13, 385.
[15] N. Nelson, R. B. Levy, J. Catal. 1979, 58, 485.
[16] I. Jardine, F. J. McQuillin, J. Chem. Soc. D 1970, 626.
Â
Â
Â
[17] a) D. E. Paez, A. Andriollo, R. A. Sanchez-Delgado, N. Valencia, R. E. Galiasso, F. Lopez-Linares, U.S.
Â
Â
Pat. 5.958.223 to INTEVEP S.A., 1999; b) D. E. Paez, A. Andriollo, R. A. Sanchez-Delgado, N. Valencia,
Â
Â
F. Lopez-Linares, R. E. Galiasso, U.S. Pat. 5.753.584 to INTEVEP S.A., 1998; c) D. E. Paez, A. Andriollo,
Â
Â
F. Lopez-Linares, R. E. Galiasso, J. A. Revete, R. A. Sanchez-Delgado, A. Fuentes, Am. Chem. Soc. Div.
Fuel Chem. Symp. Prepr. 1998, 43, 563.
[18] C. Bianchini, M. Frediani, G. Mantovani, F. Vizza, Organometallics 2001, 20, 2660.
[19] a) C. Bianchini, A. Meli in ꢀTransition Metal Sulphides ± Chemistry and Catalysisꢁ, Eds. T. Weber, R. Prins,
R. A. van Santen, Kluwer, Dordrecht, 1998, p. 129; b) C. Bianchini, A. Meli, Acc. Chem. Res. 1998, 31, 109;
c) C. Bianchini, A. Meli in ꢀAqueous-Phase Organometallic Catalysis ± Concepts and Applicationsꢁ, Eds.:
B. Cornils, W. A. Herrmann, VCH, Weinheim, 1998, p. 477; d) Bianchini, A. Meli in ꢀApplied
Homogeneous Catalysis with Organometallic Compoundsꢁ, Eds.: B. Cornils, W. A. Herrmann, VCH,
Weinheim, 1996, Vol. 2, p. 969; e) C. Bianchini, A. Meli, J. Chem. Soc., Dalton Trans. 1996, 801.
Â
[20] a) C. Bianchini, A. Meli, S. Moneti, W. Oberhauser, F. Vizza, V. Herrera, A. Fuentes, R. A. Sanchez-
Delgado, J. Am. Chem. Soc. 1999, 121, 7071; b) C. Bianchini, D. Masi, A. Meli, M. Peruzzini, F. Vizza, F.
Zanobini, Organometallics 1998, 17, 2495; c) C. Bianchini, A. Meli, S. Moneti, F. Vizza, Organometallics
1998, 17, 2636.
Â
[21] a) C. Bianchini, M. V. Jimenez, A. Meli, S. Moneti, V. Patinec, F. Vizza, Organometallics 1997, 16, 5696;
b) C. Bianchini, A. Meli, V. Patinec, V. Sernau, F. Vizza, J. Am. Chem. Soc. 1997, 119, 4945; c) C. Bianchini,
A. Meli, W. Pohl, F. Vizza, G. Barbarella, Organometallics 1997, 16, 1517; d) C. Bianchini, J. A. Casares, A.
Â
Meli, V. Sernau, F. Vizza, R. A. Sanchez-Delgado, Polyhedron 1997, 16, 3099; e) C. Bianchini, D. Fabbri, S.
Â
Gladiali, A. Meli, W. Pohl, F. Vizza, Organometallics 1996, 15, 4604; f) C. Bianchini, M. V. Jimenez, C.
Mealli, A. Meli, S. Moneti, V. Patinec, F. Vizza, Angew. Chem., Int. Ed. 1996, 35, 1706; g) C. Bianchini, V.
Â
Â
Herrera, M. V. Jimenez, A. Meli, R. A. Sanchez-Delgado, F. Vizza, J. Am. Chem. Soc. 1995, 117, 8567; h) C.
Â
Â
Bianchini, M. V. Jimenez, A. Meli, F. Vizza, Organometallics 1995, 14, 3196; i) C. Bianchini, M. V. Jimenez,
Â
A. Meli, F. Vizza, Organometallics 1995, 14, 4858; j) C. Bianchini, P. Frediani, V. Herrera, M. V. Jimenez, A.
Â
Meli, L. Rincon, R. A. Sanchez-Delgado, F. Vizza, J. Am. Chem. Soc. 1995, 117, 4333.

Helvetica Chimica Acta ± Vol. 84 (2001)
2923
[22] a) C. Bianchini, A. Meli, M. Peruzzini, F. Vizza, Organometallics 1990, 9, 226; b) C. Bianchini, A. Meli, M.
Peruzzini, F. Vizza, Organometallics 1990, 9, 2283; c) C. Bianchini, F. Laschi, A. Meli, M. Peruzzini, P.
Zanello, P. Frediani, Organometallics 1988, 7, 2575.
[23] a) C. Bianchini, A. Meli, F. Laschi, J. A. Ramirez, P. Zanello, A. Vacca, Inorg. Chem. 1988, 27, 4429; b) C.
Bianchini, A. Meli, P. Zanello, J. Chem. Soc., Chem. Commun. 1986, 628; c) C. Bianchini, C. Mealli, A.
Meli, M. Sabat, J. Chem. Soc., Chem. Commun. 1986, 777.
[24] a) G. Kiss, I. T. Horvath, Organometallics 1991, 10, 3798; b) J. Ott, L. M. Venanzi, C. A. Ghilardi, S.
Midollini, A. Orlandini, J. Organomet. Chem. 1985, 291, 89.
[25] V. I. Bakhmutov, C. Bianchini, M. Peruzzini, F. Vizza, E. V. Vorontsov, Inorg. Chem. 2000, 39, 1655.
[26] E. G. Thaler, K. G. Caulton, Organometallics 1990, 9, 1871.
[27] A. L. Crumbliss, R. J. Topping, in ꢀPhosphorus-31 NMR Spectroscopy in Stereochemical Analysisꢁ, Eds.
J. G. Verkade, L. D. Quin, VCH, Weinheim, Germany, 1987, Chapter 15, p. 531.
[28] P. S. Pregosin, R. W. Kunz, ꢀ31P and 13C NMR of Transition Metal Phosphine Complexesꢁ, Ed. P. Diehl, E.
Fluck, R. Kosfeld, Springer-Verlag, Berlin, 1979.
Â
[29] P. A. MacNeil, N. K. Roberts, B. Bosnich, J. Am. Chem. Soc. 1981, 103, 2273; J. Bakos, I. Toth, B. Heil, G.
Â
Â
Szalontai, L. Parkanyi, V. Fülöp, J. Organomet. Chem. 1989, 370, 263.
[30] C. Bianchini, H. M. Lee, P. Barbaro, A. Meli, S. Moneti, F. Vizza, New J. Chem. 1999, 23, 929.
[31] W. McFarlane, J. D. Swarbrick, J. L. Bookham, J. Chem. Soc., Dalton Trans. 1998, 3233.
[32] P. S. Pregosin, R. Salzmann, A. Togni, Organometallics 1995, 14, 842.
[33] H. Berger, R. Nesper, P. S. Pregosin, H. Rügger, M. Wörle, Helv. Chim. Acta 1993, 76, 1520.
[34] A. Bax, ꢀTwo-Dimensional Nuclear Magnetic Resonance in Liquidsꢁ; Delft University Press, Dordrecht,
Holland, 1982.
[35] I. Goljer, P. H. Bolton, in ꢀTwo-Dimensional NMR Spectroscopy, Application for Chemists and
Biochemistsꢁ, Ed. W. R. Croasmun, R. M. K. Carlson, VCH, New York, 1994.
[36] E. Brunner, J. Chem. Eng. Data 1985, 30, 269; C. L. Young, ꢀSolubility Data Series IUPACꢁ, Pergamon
Press, Oxford, 1981, Vol. 516, p 219; R. Battino, H. L. Clever, Chem Rev. 1966, 66, 395.
[37] C. Bianchini, H. M. Lee, A. Meli, F. Vizza, Organometallics 2000, 19, 849, and refs. therein.
[38] F. Bachechi, J. Ott, L. M. Venanzi, Acta Cryst. 1989, C45, 724.
[39] K. W. Rosenmund, F. Zymalkowski, Chem. Ber. 1953, 86, 37.
[40] A. Bax, D. G. Davis, J. Magn. Reson. 1985, 65, 355.
[41] V. Sklener, H. Miyashiro, G. Zon, H. T. Miles, A. Bax, FEBS Lett. 1986, 208, 94; J. Jeener, B. H. Meier, P.
Bachmann, R. R. Ernst, J. Chem. Phys. 1979, 71, 4546.
[42] A. Bax, D. G. Davis, J. Magn. Reson. 1985, 63, 207.
[43] A. Bax, R. H. Griffey, B. L. Hawkins, J. Magn. Reson. 1983, 55, 301.
[44] W. P. Aue, E. Bartholdi, R. R. Ernst, J. Magn. Reson. 1976, 64, 2229.
[45] C. Bianchini, A. Meli, A. Traversi, Ital. Pat. FI A000025, 1997.
Received March 30, 2001