A.W. Augustyniak and A.M. Trzeciak
Inorganica Chimica Acta 518 (2021) 120255
Fig. 6. High-resolution XPS spectra for Pd 3d region and XRD for Pd-dmpzc catalyst: fresh (A-B) and after the catalytic reaction (C-D).
phenylacetylene increased to 100%; however, the selectivity to styrene
decreased to 70%. Thus, better results were obtained using the catalyst
containing dmpzc ligand from the beginning of the reaction than with
the ligand added in situ.
reaction. For comparison, without Hg(0) 19% of phenylacetylene reac-
ted in 15 min.
Subsequently, we evaluated the addition of thiourea for our system.
This poisoning method with application of thiourea was developed by
Elsevier et al. [41]; however, instead of a tetramethylthiourea (TMTU)
we used thiourea. The addition of 70-fold excess of thiourea at the
beginning of the reaction enables only 7% of conversion of phenyl-
acetylene, similarly as in the experiment with Hg(0). It was also checked
whether a smaller amount of thiourea would affect the reaction. After
adding 4.8-fold thiourea excess, the conversion of phenylacetylene was
12% during the 30 min of reaction. It confirmed that thiourea effectively
poisoned our catalyst. Inhibiting effect observed in both poisoning tests
indicated the contribution of Pd0 NPs in the catalytic process. On the
other hand, in the presence of poisons, some amount of product was
formed, and therefore a contribution of molecular catalysts cannot be
excluded. Moreover, identification of different palladium species,
namely Pd0 and Pd2+, in the post-reaction sample of catalyst allowed us
to consider two alternative reaction pathways of hydrogenation with Pd-
dmpzc precursor. In the first one the reaction occurred on the surface of
Pd0 NPs, whereas in the second one mononuclear Pd species were
involved (Fig. 8). In both cases the active hydrido intermediates were
formed in reaction with BHꢀ4 and simultaneous production of BH3(OH)ꢀ .
Further hydrolysis led to B(OH)4ꢀ , which was identified by the signal at
943 cmꢀ 1 in the IR spectrum (Fig. S3) [42]. Moreover, the XPS spectrum
presented the signal at BE 192.5 eV, which can be attributed to the B
bond (Fig. S4) [43]. In both pathways the N-donor ligand (H2dmpzc or
dmpzcꢀ ) played an important role creating the Pd environment by
electronic and steric interactions with palladium. This assumption was
supported by XPS data measured for the reused catalyst in the N-region
(Fig. S6b). The binding energy of 400.2 eV is a little different compared
to the fresh catalyst (399.6 eV). This may indicate that N-donor ligand is
still coordinated to palladium.
3.2. Catalyst transformation and mechanistic study
Characterisation of the catalyst after the reaction was done using
XPS, XRD (Fig. 6), TEM, and SEM (Fig. 7) techniques.
The oxidation state Pd2+ in the fresh Pd-dmpzc catalyst was specified
by the presence of two prominent peaks located at 337.1 and 342.4 eV
for Pd 3d5/2 and 3d3/2 regions in the XPS spectrum (Fig. 6A). The Pd 3d
spectrum of the sample recovered after hydrogenation (Fig. 6C) was
deconvoluted into several peaks. Upon fitting analysis, the peaks with
binding energies of 335.0 and 340.4 eV were assigned to Pd0 (43.6%)
[38]. The binding energy 336.0 eV can correspond to PdO on the surface
(17.9%) [39]. The other peaks at 338.5 and 343.7 eV indicated the
presence of oxidised palladium species (25%). Furthermore, the binding
energies 337.3 and 342.6 eV may be assigned to the pristine Pd-dmpzc
complex (13.5%).
XRD analysis of the reused catalyst confirmed the formation of Pd0
NPs and decomposition of the crystal structure of Pd-dmpzc during the
reaction (Fig. 6D). The changes of the catalyst morphology were further
observed by SEM and TEM analyses. The presence of Pd NPs was evi-
denced by TEM microscopy. In the SAED pattern (Fig. 7 D) the diffrac-
tion rings matched with (111), (200), (220), and (311)
crystallographic planes of Pd nanocrystals were identified [40].
In light of the above XPS, XPRD, and TEM results, one can assume
that Pd0 NPs formed under hydrogenation conditions present in the
catalytically active form. To verify this hypothesis, a mercury poisoning
test was carried out [34]. Two experiments were carried out with a 500-
fold excess of Hg(0) to the catalyst. When the Hg(0) was added at the
beginning of the reaction, very low conversion (6%) of phenylacetylene
was obtained. Under the same conditions, but without the addition of Hg
(0), the conversion was 57%. In the second experiment the Hg(0) was
added after 10 min of the reaction. In this case the conversion of phe-
nylacetylene was 13% and it did not increase during the 30-minute
It should also be mentioned that the reusability test of the catalyst
demonstrated that after the second cycle the catalyst is still active and
selective (see Fig. S5). The leaching of Pd in the solvent estimated by ICP
analysis was at the limit of detection (Table S3).
5