E.A. Jaseer et al. / Inorganic Chemistry Communications 40 (2014) 78–81
79
influence on the polymer properties (entry 2, pyrrolidine used as
amine). The addition of bulky phosphines (PCy3, P(o-OMeC6H4)3) to
the in situ preformed [(Q)Rh(COD)] complex also rendered polymeric
PPA under the same conditions (entries 4 and 5, respectively), although
in less yields than those obtained with the phosphine-free ternary cata-
lytic systems. Changing of the diene in the methoxo starting dinuclear
complexes [{Rh(μ-OMe)(diolefin)}2] from COD to other known diole-
fins with higher π-acidity such as NBD or TFB did have a pronounced ef-
fect on the outcome of the catalytic reactions. As a matter of fact the
polymerization reactions proceeded much faster with the NBD and
TFB-containing complexes than those with the system containing
COD, as confirmed by direct comparison of the kinetic experiments per-
formed with these precursors by monitoring PA consumption (see
Figure S1). These observations, in line with some reports from Masuda
and coworkers on some related cationic rhodium complexes [13,14], in-
directly point out that the diene ligand should be present in the active
catalytic rhodium species, since it affects both the efficiency of the cata-
lytic process and the nature of the PPAs obtained (Mw and Mn). More
precisely, the NBD- and TFB-containing catalytic systems increase the
efficiency indexes of the reactions (Ieff; entries 6 and 7, respectively),
which in general are modest and comparable to those obtained with
other reported rhodium catalytic systems based on chelate ligands
[15]. Interestingly, the polydispersity ratios observed do not seem to de-
pend on the nature of the diene, and were found to be in general quite
good (1.3–2.0). The analysis of the orange polymers by NMR spectrosco-
py showed them to be stereoregular (see Figure S2); in all cases they
showed a significant sharp singlet at 5.84 ppm in their 1H NMR spectra
and peaks at 131.9 and 139.2 ppm in the 13C{1H} NMR spectra corre-
sponding to the “CH_C” functional groups.
hydroamination [6b], cycloaddition [6c] and dimerization [6d] that
showed an unexpected activity of these quinolato rhodium complexes.
Therefore we carried out our own study focused on assessing
the catalytic performance of the in situ ternary systems composed of
[{Rh(μ-OMe)(COD)}2]/HQ/PR3 (ratio 1:2:4) in this catalytic transforma-
tion that accounts for an 100% atom economy and affords enamines,
which are themselves a valuable source to be used for further trans-
formation to other N-containing compounds. NMR experiments
showed that replacement of the diene takes place fast and renders
the bis(phosphine) complexes of the type [Rh(Q)(PR3)2] quantitatively
(see Supporting Information). Table 2 shows the available catalytic data
obtained from the testing of HQ and 5-chloro-8-hydroxyquinoline (5-
Cl-HQ) derivatives as bidentate ligands for the rhodium-catalyzed
hydroamination of PA with piperidine.
Table 2
Catalytic data in intermolecular hydroamination of PA with piperidine.a
Entry
Catalytic system
T
(°C)
Time
(h)
Conv.
(%)
TOF (h−1
(%)b
)
1
2
3
4
5
6
7
8
9
[{Rh(μ-OMe)(COD)}2]/HQ/PR3
[{Rh(μ-NH2)(COD)}2]/HQ/PR3
[{Rh(μ-OMe)(COD)}2]/HQ/PR3
[{Rh(μ-OMe)(COD)}2]/HQ/PR3
[{Rh(μ-OMe)(COD)}2]/HQ/PR3
[{Rh(μ-OMe)(COD)}2]/HQ/PR3
[{Rh(μ-OMe)(COD)}2]/HQ/PR3
[{Rh(μ-OMe)(COD)}2]/HQ/PR3
[{Rh(μ-OMe)(COD)}2]/5-Cl-HQ/PR3
[{Rh(μ-OMe)(COD)}2]/5-Cl-HQ/PR3
RT
RT
RT
40
RT
RT
RT
RT
RT
40
24
24
24
24
24
24
48
48
24
8
100
100
92c
100
33d
63e
43f
1.3 (51%)
1.2 (47%)
1.3 (52%)
4.4 (44%)
0.1 (33%)
0.4 (46%)
0.2 (41%)
0.3 (28%)
2.3 (35%)
9 (45%)
28g
100
100
Table 1
10
Catalytic data in stereoregular in Rh-catalyzed polymerization of PA
a
Conditions: 0.05 mmol (10 mol%) [Rh], 0.1 mmol HQ (or 5-Cl-HQ), 1 mmol PA,
3 mmol piperidine, 0.2 mmol P(p-OMeC6H4)3, toluene (2 mL), mesitylene (1.8 mmol)
as internal standard.
.
b
TOF = Mol PA/mol Rh/h, calculated at the indicated PA consumption (%).
1 mmol Cs2CO3 added.
0.2 mmol PPh3.
0.2 mmol P(p-FC6H4)3.
THF used as solvent.
2% mol [Rh].
c
d
a
Entry Catalytic system
Mn
Mw
Mw/Mn Conv. (%) Ieff
e
1
2
3
4
5
6
7
[{Rh(OMe)(COD)}2]/HQ 79,600 123,000 1.55
[{Rh(OMe)(COD)}2]/HQ 96,800 170,000 1.76
81
90b
88
1.02
0.93
1.05
6.56
3.72
1.26
1.56
f
g
[{Rh(NH2)(COD)}2]/HQ
84,150 147,800 1.76
[{Rh(OMe)(COD)}2]/HQ 15,250
[{Rh(OMe)(COD)}2]/HQ 26,900
38,960 2.55
71,100 2.64
39c
86d
100e
100f
The reactions proceeded at room temperature and in all cases they
were regioselective rendering the E-enamine showed above, the prod-
uct of an anti-Markovnikov amine addition. Several trends were
established upon changing key catalytic parameters in the model reac-
tion. The need of using a loading of 10% [Rh] of catalyst became evident
from the first screening experiments in order to achieve quantitative
conversions to the enamine (compare entries 1 and 8). On the other
hand, the use of PPh3 or P(p-FC6H4)3 as external ligands in these reac-
tions gave only moderate yields of the enamine after 24 h at room tem-
perature (entries 5 and 6, respectively). The catalytic performance was
found to be substantially increased by shifting to the more basic P(p-
OMeC6H4)3, which was then chosen as the added phosphine in the ca-
talyses. Again, the use of the amido precursor [{Rh(μ-NH2)(COD)}2] or
the methoxo-bridged analog in combination with HQ/P(p-OMeC6H4)3
in these catalytic processes did not change the activity of the resulting
catalytic species in terms of yields of enamine (entries 1 and 2), which
indirectly proves that the bridging ligands (NH2 and OMe) are the key
for deprotonating the HQ derivatives and subsequently coordinating
them to rhodium. On the other hand, the use of an external weak base
(such as Cs2CO3) does not really affect the catalytic performance, al-
though slight less yields of the enamine are obtained (entry 3). In this
line the shifting to more polar THF as solvent in the catalytic reactions
was not beneficial in terms of yield, since a drop to 43% was achieved
(compare entries 1 and 7). A subtle change in the HQ skeleton does
have a noticeable positive effect in catalysis; more specifically, the
shifting from HQ to 5-Cl-HQ provides quantitative conversions to the
[{Rh(OMe)(NBD)}2]/HQ 79,600 167,000 2.10
[{Rh(OMe)(TFB)}2]/HQ 64,000 83,120 1.30
Conditions: RT, in toluene, 24 h, 10% mol [Rh], piperidine (3 mmol), except in entry 2,
where pyrrolidine was used.
a
Initiation efficiency, Ieff = Mtheor / Mn × 100; where Mtheor = [PA]0 / [Rh] ×
MWphenylacetylene × polymer yield [18].
b
Pyrrolidine.
Two molar-equiv. of PCy3 added.
Two molar-equiv. of P(o-OMeC6H4)3 added.
Reaction time 3.5 h.
c
d
e
f
Reaction time 0.5 h.
As mentioned earlier, the addition of steric-demanding phosphines
to the ternary catalytic mixture composed of [{Rh(μ-OMe)(COD)}2]/
HQ/amine did not have apparently any effect on the polymer structures.
The yields of these polymerization reactions were however relatively
low, particularly in the case of the sterically demanding P(o-
OMeC6H4)3 phosphine, in which the o-methoxo substituent precludes
an easy coordination to the metal center. The situation changes drasti-
cally when strong-coordinating phosphines such as the ubiquitous
PPh3 or P(p-OMeC6H4)3 are present in the formation of the catalytic pre-
cursors: under given conditions, piperidine adds to PA affording the cor-
responding enamine through an intermolecular hydroamination
catalytic process and no polymerization was observed. In this line, re-
cent reports by Kakiuchi and coworkers showed that the isolated
[Rh(Q)(PR3)2] mononuclear complexes were active species in catalytic
transformations of terminal alkynes such as hydroalkoxylation [6a],