Full Papers
tral or endothermic adsorption energy, so that it can easily
detach from the surface, suppressing over-hydrogenation. C2H4
binds to Ni(111) via one carbon to the nickel top site and the
other carbon to the NiÀ Ni bridge site[9] with a binding energy of
À 1.14 eV (Figure S1). This binding geometry is not affected by
the surface composition since the adsorption energy on Ni3P
(0001) and Ni3P2(0001) is À 1.13, and À 1.26 eV, respectively
(Figures S2 and S3). On Ni4P3(0001), the adsorption of C2H4 to a
same conditions, and call for a molecular-level understanding
of this behavior. In line with the observations for 1-hexyne, Ni
and Ni/SiO2À Al2O3 exhibit lower and similar reaction rates to
NixPy, respectively. Also, at 343 K and 20 bar, the selectivity to 1-
hexene on Ni and Ni/SiO2À Al2O3 is 60% and 10%, respectively.
To verify the robustness on stream, a long-term test was
conducted over the catalysts. As shown in Figure 4, both Ni2P
and Ni5P4 displayed an induction period of 2 h, likely due to the
reduction of the oxidic layer covering the surface,[16] before
reaching stable performance for 28 h on stream, confirming the
suitability of these systems for continuous-flow applications.
Analysis of the used catalysts by X-ray diffraction and X-ray
fluorescence spectrometry confirmed the preservation of the
crystalline structure and the absence of leaching of either Ni or
P.
The recent kinetic study over palladium sulfides evidenced the
role of the p-block element (i.e., sulfur) in stabilizing the activated
H atoms, imparting a bifunctional mechanism that originates
higher activity than unmodified palladium nanoparticles, where H2
and the alkyne compete for the same site.[13] Prompted by these
observations, the mechanism was explored by conducting kinetic
tests to determine the reaction order with respect to hydrogen
and the alkyne over Ni2P, Ni5P4, and Ni/SiO2À Al2O3. In line with
previous investigations,[13] the order with respect to molecular
hydrogen is positive (ca. 1.2) in all cases, while the order to the
alkyne is close to 0 for nickel phosphide (À 0.04 and À 0.20 for Ni2P
and Ni5P4, respectively, Figure S6), and À 1 over Ni/SiO2À Al2O3,
suggesting in the latter situation a competition between the
alkyne and hydrogen on the surface of the unmodified nickel
nanoparticles. Also, this indicates a role of the heteroatom in the
reaction mechanism as now a two-site model is required to
describe the kinetic behavior.
1
2
3
4
5
6
7
8
9
phosphorus site, forming PÀ CH2À CH2À P, is even stronger (Eads
=
À 2.23 eV) and more stable than on the Ni3 site of the same
surface (Eads =À 0.98 eV). This shows that having trimeric Ni
ensembles spaced by phosphorus atoms does not necessarily
poison nickel, but it provides new P-containing sites on the
surfaces with stronger adsorption features.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
Hydrogenation of Alkynes
The nickel phosphide catalysts were evaluated in the continu-
ous-flow three-phase semi-hydrogenation of 1-hexyne, a sub-
strate widely studied to probe the selective character of alkyne
hydrogenation catalysts, and benchmarked against the nickel
precursor. Figure 3a maps the influence of the reaction temper-
ature and pressure on the rate and selectivity to 1-hexene.
While both Ni2P and Ni5P4 display very similar performance
under every condition evaluated, for instance, at 303 K and
10 bar the rate of 1-hexyne conversion is 1.3 and 1.4 hÀ 1, on
Ni2P and Ni5P4, respectively, Ni present much lower reaction
rate, 0.4 hÀ 1. Remarkably, nickel phosphides are almost fully
selective (higher than 95%) to the desired alkene even at high
pressure (P=20 bar) and temperature (T=343 K), while Ni
displays ca. 70% selectivity at a much lower conversion level
(Figure S5). This points to the robustness of nickel phosphides
to the formation of β-hydrides, which bare nickel is known to
suffer from, and to the beneficial effect of introducing
phosphorus into the nickel lattice to increase both the activity
and the selectivity. To assess the selectivity patterns of a Ni-
based catalyst at higher conversion levels, a commercial
60 wt.% Ni/SiO2À Al2O3 catalyst was evaluated in 1-hexyne
hydrogenation. At 343 K and 20 bar the rate of 1-hexyne
conversion on Ni/SiO2À Al2O3 is 3.5 hÀ 1 and the selectivity to 1-
hexene of 18%, with n-hexane as the major side product and
traces of isomers of the double bond (ca. 5%). These results
indicate that supported nickel-based catalyst can match and
surpass the activity of bulk nickel phosphides, but at the
expenses of a very low selectivity to the semi-hydrogenated
product.
It is noteworthy that in comparison with the nickel phos-
phides, a nanostructured Pd3S catalyst, the highest-performing Pd-
based catalyst identified to date, displayed a two orders of
magnitude higher reaction rate under every condition evaluated.
This indicates that phosphorus is not able to sufficiently promote
the performance of nickel to match that of palladium.
Reaction Mechanism
To understand the structure sensitive or insensitive activity profile
as a function of the substrate, molecular solvation corrected
reaction profiles for the hydrogenation of C2H2 to C2H6, comprising
four hydrogenation steps (CHCH*+4H*!CHCH2*+3H*!
CH2CH2*+2H*!CH2CH3*+H*!CH3CH3*), were computed (Fig-
ure 5). On Ni(111), the reaction barriers (energies) are 1.02 (0.61),
0.60 (À 0.02), 0.55 (0.39), 1.17 (À 0.07) eV. For Ni2P, Ni3P2(0001) was
selected to investigate the reaction due to its higher stability.[34]
The corresponding values are: 0.93 (0.20), 1.00 (À 0.04), 1.05 (0.49),
and 1.09 (À 0.05) eV. The situation on Ni3P4(0001) is slightly
different. Since C2H2 can be bound with two different adsorption
structures, PÀ CHÀ CHÀ P and PÀ CHÀ CHÀ Ni (vide supra), with the
former showing stronger adsorption that might potentially poison
the surface, both were taken as starting points to investigate the
reaction. From P-CH-CH-P the energies of the reaction profile are
The catalysts were additionally evaluated in the continuous-
flow three-phase semi-hydrogenation of 2-methyl-3-butyn-2-ol
at various temperatures and pressures. Conversely from the
case of 1-hexyne, where Ni2P and Ni5P4 display similar activity
profiles, significant differences are observed in the hydro-
genation of the alkynol (Figure 3b). In particular, at 303 K and
5 bar the rate of 2-methyl-3-butyn-2-ol conversion is 1.6 and
1.1 hÀ 1, while the selectivity to 2-methyl-3-buten-2-ol is 92 and
90% on Ni2P and Ni5P4, respectively. These results point to the
higher activity of Ni2P than Ni5P4 when compared under the
ChemCatChem 2018, 10, 1–9
4
© 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!