Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
J.G. Tillou and A.K. Vannucci
Journal of Organometallic Chemistry 944 (2021) 121848
Table 2
increase or decrease in observed catalytic activity, indicating that
silver is just a spectator ion in the reaction.
Catalytic HDO of benzyl alcohol using 2 under various temperature
and pressure conditions.
Detailed examination of the Pd present in 2 showed the binding
energy of the Pd 3d electron was 335.25 eV (Fig. 3b) which corre-
sponds well with metallic Pd.[26] Furthermore, the Ag 3d binding
energy, shown in Fig. 3c was determined to be 367.84 eV which is
close to the known binding energy of AgF (367.6 eV).[27] Lastly, the
carbon 1s spectrum (Fig. 3d) was fit with three peaks which could
indicate the presence of C-C, C-N, and C-O bonds, as are present in
the original ligand structure. Therefore, the solid catalyst 2 is com-
posed of metallic palladium in a matrix of decomposed ligand. This
amorphous solid particle surprisingly exhibits complete selectivity
for HDO over ring hydrogenation, which indicates the metallic Pd
in the particles does not form large metallic particles.[13]
Entry
Temperature (°C)
H2 pressure (bar)
% Conversiona
7
20
5
0
8
100
150
200
100
200
5
0
9
5
1
10
11
12
5
99
5
20
20
96
Conditions: 0.1 M benzyl alcohol in 50 mL of methanol with 50 mg
of catalyst 2 per reaction.
a
Conversion of benzyl alcohol to toluene. All reactions exhibited
complete selectivity as determined by GC-MS analysis.
To further characterize 2, powder X-ray diffraction (PXRD) anal-
ysis was performed. The black line pattern in Fig. 4 shows the
XRD spectrum of particulate 2 obtained directly from decom-
posed molecular catalyst. The spectrum indicates that the palla-
dium forms crystalline material with a face-centered cubic struc-
ture. Fitting of the peaks at the 220 position between 60–75° 2θ
shows an average palladium particle size between 3.5 and 5.5 nm.
Taken in conjunction with the XPS data in Fig. 3, the molecular
catalyst decomposes into 2, which consists of a carbon, oxygen, ni-
trogen matrix that contains metallic palladium nanoparticles.
The recyclability of catalyst 2 was then tested to gain an un-
derstanding of the stability of this unique structure. Table 2 shows
that catalyst 2 is active for HDO of benzyl alcohol at 200 °C. How-
ever, upon successive reactions when catalyst 2 is filtered from so-
lution and placed in a new reaction solution, the activity of the
catalyst dies off and becomes completely inactive by the third re-
action. The recycled catalyst 2 sample was then analyzed to un-
derstand the loss in activity. The red pattern in the PXRD data in
Fig. 4 shows the spectrum of recycled 2. As can be seen in the red
spectrum, the palladium appears to lose crystallinity as evidenced
by the broadening of the peaks compared to the black spectrum.
This result could indicate a reorganization of the carbon matrix
around the Pd particles resulting in an essential coking of the Pd
catalyst and deactivation.
2.2. Solid palladium precipitate activity
The results shown in Table 1, taken alone, could lead to an im-
proper hypothesis about the molecular catalytic mechanism if po-
tential decomposition of the catalyst is not considered. For all re-
actions in Table 1 performed at temperature ≥ 100 °C, solid partic-
ulate formation was observed in the reaction vessel. These partic-
ulates were then filtered and dried and tested for catalytic activity.
The results in Table 2 show that at temperatures below 200 °C, the
solid particulates (catalyst 2) that formed during the homogeneous
reactions do not exhibit appreciable catalytic activity for the hy-
drodeoxygenation of benzyl alcohol. Increasing the H2 pressure in
the reaction cell at 100 °C also does not lead to the catalytic con-
version of benzyl alcohol. Increasing the reaction temperature to
200 °C, however, led to a marked increase in benzyl alcohol con-
version. Near quantitative hydrodeoxygenation of benzyl alcohol is
achieved by 2 with complete selectivity for HDO over ring hydro-
genation. This result indicates that the decomposition product, 2,
of the molecular palladium catalyst 1 is catalytically active for ben-
zyl alcohol HDO at temperatures ≥ 200 °C and helps explain the
results presented in Table 1.
From combining the results of Tables 1 and 2 a more complete
picture of benzyl alcohol HDO catalysts starting from 1 can be
formed. Catalyst 1 exhibits moderate catalytic activity for the HDO
of benzyl alcohol at temperature as low as 20 °C. Above room tem-
perature and at pressures of at least 5 bar H2, homogenous, molec-
ular catalyst 1 decomposes to solid, heterogeneous catalyst 2. The
newly formed catalyst 2, however, does not exhibit catalytic ac-
tivity at temperatures below 200 °C. Furthermore, increased pres-
sures of H2 appear to increase the rate of catalyst 1 decomposition
as evidenced by entries 5 and 6 in Table 1. Therefore, to achieve
optimal hydrodeoxygenation of benzyl alcohol with this catalytic
system, temperatures at or above 200 °C with H2 pressures at or
below 5 bar are ideal.
2.3. Leached palladium activity
To examine whether leached palladium, either from the molec-
ular catalyst 1 or the particulate catalyst 2, was also catalytically
active post reaction solutions were tested for HDO activity. First,
starting from molecular catalyst 1 a reaction was performed iden-
tical to the conditions described in Table 1 Entry 2. After the reac-
tion, any solid particulates were filtered from the reaction solution
and the solution was placed back into the reaction vessel with ad-
ditional benzyl alcohol. A second attempted reaction was then per-
formed on the solution without the solid particles, and the solu-
tion no longer exhibited catalytic activity. This result suggests that
catalytically active 1 had decomposed to a heterogeneous particle
and no longer existed as a homogeneous solution-based catalyst or
as a leached palladium catalyst.
With catalyst 2 being active and selective for HDO of benzyl
alcohol it became important to characterize 2. An XPS analysis of
2 was performed to determine its elemental composition. Survey
scans on three samples were performed and shown in Fig. 3a. Cat-
alyst 2 is composed of Pd, C, O, and N in addition to P and F from
the counter anions. This result indicates that the molecular catalyst
completely decomposes, likely through ligand degradation. The re-
sulting solid particles are a conglomeration of the individual atoms
that composed catalyst 1. In addition to the expected elements, a
small weight percent of silver was also detected in the particles.
This silver may arise from the AgPF6 salt used for chloride anion
exchange in the synthesis of catalyst 1 (see experimental section).
To examine if Ag may play any role in catalysis, we performed cat-
alytic reactions with the addition of roughly 0.7 weight percent
silver added to the reaction solution. Under three separate trials,
the addition of silver to the reaction mixture did not result in an
Next, the reaction solution from testing particulate catalyst 2
was tested for activity. Following a reaction that was performed
under the condition of Table 2 entry 10, additional benzyl alcohol
was added to the reaction solution and a second reaction was per-
formed. In this case, HDO conversion of benzyl alcohol to toluene
was observed using GC-MS analysis. In fact, three consecutive re-
actions were performed at 200 °C in which the reaction vessel was
allowed to cool, and then additional benzyl alcohol was added to
the solution before another reaction was performed. In each case,
complete conversion of benzyl alcohol to toluene was observed in
four hours. This result indicates that Pd leaches from the particu-
late catalyst 2 and forms active catalyst in solution. This leaching-
3