6
D.J. Segobia et al. / Journal of Catalysis xxx (xxxx) xxx
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.8
0.6
0.4
0.2
0.0
(A)
(B)
(C)
0
100
200
Time (min)
300
400
500
0
100
200
Time (min)
300
400
500
0
200
300
100Time (min)
Fig. 5. Asymmetric secondary amines (ASA) synthesis from butyronitrile and secondary alcohols: (A) 2-propanol, (B) 2-butanol, (C) 2-octanol. Butyronitrile conversion (XBN
)
and yields (Yi): ASA, Dibutylamine (DBA), Butylamine (BA), Butylidene-butylamine (BBA), Asymmetric secondary imine (ASI). Reaction conditions as in Fig. 4.
(r0ASA = 5.5 mmol/g h). The XBN vs t and Yi vs t curves obtained for
BN/2-butanol (Fig. 5B) and BN/2-octanol (Fig. 5C) reactions were
qualitatively similar to those observed for BN/2-propanol in
Fig. 5A. The ASA yields obtained for BN/2-butanol (85%) and
BN/2-octanol (78%) were comparable to that of BN/2-propanol
(85%), but the ASA formation rate depended on the alcohol size, fol-
lowing r0ASA the order 2-propanol > 2-butanol > 2-octanol. Overall,
the results of Fig. 4, Fig. 5 and Table 2 show that the synthesis of
ASA is more selective and efficient when BN reacts with a sec-
ondary alcohol than with the primary alcohol of the same carbon
atom number. Indeed, the ASA yields obtained here for BN/sec-
ondary alcohol reactions varied between 78% and 85% whereas
those determined for BN/primary alcohols were in the 49–58%
range. Furthermore, it is significant noting that the ASA formation
rate was clearly higher when using secondary alcohols. In fact,
from the values of Table 2, we observe that the r0ASA ratios for sec-
ondary/primary alcohols of the same carbon atom number,
and show that the YASA value obtained at the end of the runs
decreased when C0BN was increased for both alcohols. Regarding
the reaction kinetics, it is observed in Table 3 that r0BN and rA0SA con-
tinuously decreased by increasing C0BN. In order to quantify the
effect of BN concentration on the initial rates of BN conversion
and ASA formation, the experimental data were interpreted by
considering a power-law rate expression (Eq. (1))
ꢀ
ꢁ ꢀ
ꢁ
a
b
r0i ¼ k C0BN
Cj0
ð1Þ
where i is BN or ASA, and j is 2-propanol or 2-butanol. Reaction
orders
a
were determined graphically from the r0i vs CB0N logarithmic
plots that are given in Supporting Information (Fig. SI1) for both
secondary alcohols. In all the cases, a negative order with respect
to BN was determined. The values of
a
in 2-propanol were ꢂ0.10
and ꢂ0.23 for r0BN and rA0SA, respectively, whereas in 2-butanol the
corresponding orders in BN were ꢂ0.15 and ꢂ0.66. These results
show that the alkylation of BN with secondary alcohols on Co/
SiO2 is negative order in BN, which suggests that the strong adsorp-
tion of BN-derived nitrene intermediates on the metal would
increasingly block the accessible surface sites for the alcohol
À
Á
À
Á
s
r0ASA = r0ASA p, was 2.3 for 2-propanol/n-propanol and 2-butanol/
iso-butanol, and 4.9 for 2-octanol/n-octanol. These latter results
probably reflect the fact that aliphatic secondary alcohols are bet-
ter hydrogen donors than the aliphatic primary alcohols with the
same carbon atom number because dialkyl ketones have higher
reduction potential than aldehydes [39,40]. Dehydrogenation of
secondary alcohols on the metal surface of the catalyst forms
adsorbed hydride and ketone-type species that react with
adsorbed BA releasing a water molecule to give by condensation
ASI which is finally hydrogenated to ASA by the surface hydride
species. A higher coverage of the Co surface by ketone-type species
than by aldehydic-type species would explain therefore the higher
r0ASA values obtained in secondary alcohols as compared to primary
alcohols. Consistently, the BA concentrations observed in BN/sec-
ondary alcohol reactions were clearly lower than those determined
in the corresponding BN/primary alcohol reactions, which proba-
bly reflects that adsorbed BA species are more rapidly converted
to the asymmetric secondary imine in secondary alcohols because
of the higher concentration of surface carbonyl-type species
formed from secondary alcohols. In summary, our results of
Fig. 5 show that Co/SiO2 promotes efficiently the synthesis of
ASA by BN alkylation with secondary alcohols (2-propanol, 2-
butanol and 2-octanol) yielding selectively (ASA + DBA) mixtures
that contain 78–85% of ASA. It is important noting that the boiling
point of DBA at 1.013 bar (432.6 K) is significantly higher than that
of isopropylbutylamine (401.1 K), the ASA obtained in 2-propanol/
BN reaction, and of N-sec-butyl-1butanamine (424.7 K) the ASA
produced in 2-butanol/BN reaction, which allows to obtain pure
ASA from the (DBA + ASA) mixtures by distillation.
adsorption as C0BN is augmented, thereby decreasing the ASA forma-
tion rate.
The apparent activation energies (Ea) of r0BN and rA0SA for BN/2-
propanol reaction on Co/SiO2 were obtained via an Arrhenius-
type function, by plotting ln r0i as a function of 1/T (Fig. SI2 in
Supporting Information). The experiments were carried out in
the 383–413 K range. From the slopes of the resulting linear plots
we determined Ea values of 26.1 kJ/mol and 30.6 kJ/mol for r0BN and
r0ASA, respectively. The higher Ea value obtained for r0ASA is consistent
with the results presented in Table 2 showing that, at a given CB0N
value, r0ASA was significantly lower than rB0N for all the BN/alcohol
reactions investigated in this work.
Finally, it is worth noting that the conventional synthesis of ASA
by N-alkylation of primary amines with alcohols has been widely
studied in absence of hydrogen via a borrowing-hydrogen or
hydrogen-autotransfer strategy [9,41,42,43], in which the hydro-
gen required for the final reaction step (hydrogenation of ASI to
ASA) is furnished by hydrogen transfer from the alcohol molecule.
In order to compare the synthesis of ASA from N-alkylation of BN
or BA with a secondary alcohol we carried out the BA/2-propanol
reaction on Co/SiO2 under the same conditions used for BN/2-
propanol reaction in Fig. 5A, but in absence of hydrogen. In Fig. 6
we show the results obtained for the N-alkylation of BA with 2-
propanol; the reaction was completed after 550 min and formed
a product mixture containing 89% of ASA and 9% DBA. The ASA for-
mation rate determined from the YASA vs t curve of Fig. 6 was
r0ASA = 2.4 mmol/g h, significantly lower than that obtained for
BN/2-propanol reaction in Fig. 5A (12.9 mmol/g h). In summary,
The effect of BN concentration on the reaction kinetics and
yields was investigated for 2-propanol and 2-butanol. The initial
BN concentration was changed while maintaining unmodified the
rest of the operating conditions. Results are presented in Table 3
Please cite this article as: D. J. Segobia, A. F. Trasarti and C. R. Apesteguía, Selective one-pot synthesis of asymmetric secondary amines via N-alkylation of