1
4
K.C.B. Oliveira et al. / Applied Catalysis A: General 497 (2015) 10–16
Table 3
a
Hydroaminomethylation of estragole (1) with di-n-butylamine: ligand effect .
Entry
Ligand
Tolman’s parameters
Conversion (%)
Product distribution (%)c
Regioselectivity(%)c,d
CO (cm 1
−
)
ꢁ
◦
2
Aldehydes
Enamines
Amines
Others
˛
ˇ
ꢂ
ꢀ
16
17
18
19
20
21
22
None
TBPP
P(OPh)3
TPP
PPP
DBP
100
100
100
100
100
100
100
10
0
6
1
1
0
0
0
23
24
13
47
0
0
0
0
1
0
1
79
87
85
70
67
80
49
11
13
9
6
7
2
10
5
1
1
31
32
34
28
25
41
41
67
58
61
71
74
57
56
2086.1b
175b
b
b
2085.3
128
e
160f
2071.9
e
f
2070.4
160
g
151f
∼2070
7
3
0
0
2
3
2068.9b
145
b
PPh3
a
Conditions: 1 (10 mmol); di-n-butylamine (10 mmol); [Rh(cod)(-OMe)]2 (5.0 × 10−3 mmol), ligand (if any, 5.0 × 10−2 mmol) toluene (30 mL), 40 bar (CO:H2 = 1:3), 80 C,
◦
2
4 h. For products, the value “zero” means not observed or less than 0.5%.
b
From ref. [52].
Determined by GC.
c
d
˛
= (5 + 8 + 11); ˇ = (4 + 7 + 10); ꢂ = (3 + 6 + 9).
e
f
From ref. [53].
Measured employing the method of ref. [52] for ligands with different substituents.
Estimated value from Ref. [54].
g
tris(2-t-butylphenyl)phosphite (TBPP), was added in the study to
compare steric effect also in the class of phosphites.
The reaction led to three isomeric amines (9–11) as final
products (Scheme 2). Depending on the reaction conditions and
time, the intermediate products, aldehydes (3–5) and enamines
remain in the reaction solution even after 24 h, indicating that the
system is not efficient for the reductive amination (amine conden-
sation + enamine hydrogenation). Fuentes et al. [44] demonstrated
that the less electro-donating tris(3,4,5-trifluorophenyl)phosphine
is better than PPh3 and other stronger electron donor phosphines
to promote the enamine hydrogenation step, contrasting with the
fact that alkene hydrogenation is generally accelerated by electron
rich ligands. Indeed, the phospholes TPP and PPP, which are less
(
6–8), were observed in the reaction solution. Aldol condensation
products, hydrogenation product (4-propylanisole), alcohols, and
several unidentified products were also observed as minor prod-
ucts in some experiments and were considered together as “others”
in the product distribution shown in Table 3.
electron donating than PPh , (entries 19 and 20) proved to be effi-
3
cient in avoiding the double bond isomerization of the substrate
and are more efficient ancillaries than PPh3 for the reductive ami-
nation, although a fair amount of aldehydes (c.a. 24%) remain in
solution even after 24 h. The phosphole DBP (Entry 21) showed to
promote more efficiently the reductive amination than TPP or PPP
and, although some double bond isomerization of the substrate is
still observed, no other side product is formed, what makes DBP a
good starting point to design more efficient ancillaries for HAM.
The regioselectivity in Table 3 is reported as the sum of prod-
ucts derived from the aldehydes with the formyl group on the
carbons ␣, , and ␥ in relation to the aryl ring. This regioselectivity
does not correlate straightforwardly with the regioselectivity of the
migratory insertion of the hydride in the hydroformylation cycle. As
mentioned before, the double-bond isomer 2 also results from the
branched metal-alkyl intermediate. Furthermore, in the systems
with phosphites, alkene 2 is also converted into the HAM products
and part of the products is transformed in secondary unidentified
side products. For the more electron donor phospholes and PPh3,
the contribution of the isomerization is not high, so that the regios-
electivities correlate quite well with those presented in Table 2 for
In Table 3 the results for the HAM of 1 employing the ligands
depicted in Fig. 1 are presented. It is well known from the literature
that high ligand/Rh molar ratios are detrimental for the enamine
hydrogenation step in HAM if the ligand is too electron rich. On the
other hand, a too low ligand/Rh ratio would allow a significant con-
tribution of non promoted rhodium species (i.e. [Rh(CO) H]). Thus,
3
we decided to keep a ligand/Rh atomic ratio of 5, the temperature
◦
of 80 C and a rather long reaction time (24 h) to favor the yield of
amines in all HAM experiments.
The results are presented, as in Table 1, in the crescent order of
the electron donor ability of the ligand (descendent order for ꢀCO).
The non-promoted system (entry 16) and the systems with phos-
phites (entries 17 and 18), which are the less electron-donating
ligands than the others included in Table 3, are more efficient to
promote the HAM as a whole, since no aldehydes or enamines
are observed after 24 h. However, these systems also promote the
double bond isomerization to give internal alkene 2 (ca. 10%).
For the non-promoted systems (entry 16), alkene 2 is only spar-
ingly converted to hydroformylation products. In the presence of
phosphites, especially the bulky phosphite TBPP, the conversion of
◦
the runs performed at 70 C with the P/Rh atomic ratio of 5.
2
1
into the carbonylation products occurs as the products 5, 8 and
1 (alpha products), which can only be formed from 2, appear in
Also in apparent contradiction with the steric hin-
drance/regioselectivity correlation is the regioselectivity of
the Rh/TBPP system: the cone angle of TBPP is 175 , but the l/br
◦
considerable amounts. TBPP (entry 17) is the most efficient ligand
in the series to promote the HAM of estragole with di-n-butylamine
giving amines in 87% yield; however, a significant amount of side
products (13%) is also observed. In the HAM sequence, water is
formed as an intrinsic co-product and the employment of phos-
phites as ancillaries was taken with caution due to its instability
in the presence of water. In a recent paper, Tricas et al. [57] have
demonstrated that some phosphites, including TBPP, are quite sta-
ble in the presence of water, even under acidic conditions. Indeed,
phosphites have proven to be useful in HAM reactions [11] but, if
industrial application is aimed, a longer-term stability of the ligand
has to be demonstrated.
ratio is comparable to the system with P(OPh) , with cone angle
3
◦
of 128 . In this case, it should be considered that due to a high
ligand steric bulkiness, mostly monoligand species with TBPP (i.e.
[Rh(H)(CO) TBPP]) prevail even at P/Rh = 20 [57]. Bearing only one
2
ligand, the metal center in these species is not as steric hindered
as the metal center bearing two phosphorus ligands.
To explore in more details the ligand performance, the kinetic
curves for the reaction with selected ligands are presented in Fig. 2.
It is noteworthy that the hydroformylation step in the non-
promoted system (Fig. 2a) is slower than that in the promoted
system. The substrate conversion was complete in nearly half
an hour in the presence of any phosphorous ligand, whereas 2 h
were necessary with the non-promoted system. Thus, an average
The ligand PPh3 is cheap, quite stable and efficient to promote
the hydroformylation step as shown in Tables 1 and 2, but it fails
in promoting efficiently the HAM. In entry 22, 47% of aldehydes
−
1
turnover frequency (TOF) of at least 2000 h
is reached for all
Oliveira, Kelley C.B.
