Titanium nitride-nickel nanocomposite as heterogeneous catalyst for the hydrogenolysis of aryl ethers

Article abstract of DOI:10.1021/ja4119412

Lignin from biomass can become a sustainable source of aromatic compounds. Its depolymerization can be accomplished through hydrogenolysis, although the development of catalysts based on cheap and abundant metals is lacking. Herein, a sustainable composite based on titanium nitride and nickel is synthesized and employed as catalyst for the hydrogenolysis of aryl ethers as models for lignin. The catalytic activity of the new material during hydrogenation reactions is proven to be superior to that of either component alone. In particular, different aryl ethers could be efficiently converted under relatively mild conditions into aromatic compounds and cycloalkanes within minutes.

Full text of DOI:10.1021/ja4119412

Communication
pubs.acs.org/JACS
Titanium Nitride-Nickel Nanocomposite as Heterogeneous Catalyst
for the Hydrogenolysis of Aryl Ethers
Valerio Molinari, Cristina Giordano, Markus Antonietti, and Davide Esposito*
Max-Planck-Institute of Colloids and Interfaces, 14424 Potsdam, Germany
S
* Supporting Information
advantages of lower costs and higher availability. Catalytic
applications of this class of metallic ceramics for oxidation,19
ABSTRACT: Lignin from biomass can become a
reduction,20 and alkylation21 reactions have been recently
sustainable source of aromatic compounds. Its depolyme-
rization can be accomplished through hydrogenolysis,
although the development of catalysts based on cheap and
abundant metals is lacking. Herein, a sustainable
composite based on titanium nitride and nickel is
synthesized and employed as catalyst for the hydro-
genolysis of aryl ethers as models for lignin. The catalytic
activity of the new material during hydrogenation reactions
is proven to be superior to that of either component alone.
In particular, different aryl ethers could be efficiently
converted under relatively mild conditions into aromatic
compounds and cycloalkanes within minutes.
reported. Interestingly, doping with Ni has been used as an
approach to enhance the activity of tungsten carbides during
the hydrogenation of cellulose to ethylene glycol.20 More
recently, tungsten carbides have been reported as efficient
supports for the preparation of platinum-based electro-
catalysts.22 The role of semiconducting supports in heteroge-
neous catalysts and their ability to modify the electron density
of the supported metal (Mott−Schottky effect) thereby
modifying their activities has been recently highlighted.23 In
this work, titanium nitride (TiN) was used for the first time as
support to enhance the catalytic activity of Ni. TiN is a
biocompatible material24,25 with excellent physical and chemical
properties including thermal stability and acid resistance.26,27
Its hybridization with Ni through a facile and scalable synthesis
provided a noble metal-free nanocomposite characterized by
high catalytic activity for the hydrogenolysis of aryl ethers,
model molecules of lignin structure. A comparison of this
material with pure TiN or Ni particles during hydrogenolysis
was performed to elucidate the synergistic effect between the
two components in the nanocomposite. We started with the
synthesis of TiN nanoparticles by a reported urea route,28
which involves the calcination under nitrogen flux of a titanium-
urea gel-like precursor at 750 °C. The resulting powders
presented a bronze color typical of TiN (Figure S1), whose
formation was confirmed by XRD analysis (Figure 1). The
addition of Ni (Ni/TiN ∼0.5 molar ratio) was performed in a
second step, by impregnation of TiN with an ethanolic solution
of Ni acetate tetrahydrate (NiAc), followed by solvent
evaporation to give a bronze homogeneous powder (NiAc@
TiN). Infrared analysis on NiAc@TiN showed an intense shift
in the peak of the carbonyl group of the acetate compared to
pure NiAc (Figure S2), diagnostic of an interaction with TiN.
This powder was then calcined under nitrogen at different
temperatures to obtain the composite material TiN-Ni. Figure
1 shows the XRD patterns of pure TiN-Ni nanocomposite
obtained at 500 °C, together with the partially oxidized TiN-Ni
obtained at 600 and 700 °C.
ignin is the second most abundant biopolymer on earth
1
Lafter cellulose.
The development of strategies to
depolymerize this complex molecule is gaining interest due to
the possibility of obtaining aromatic platform chemicals from
nonfossil resources.2 However, this goal remains challenging
due to the presence in the lignin structure of several ether and
carbon−carbon bonds that are characterized by high energies
and very different reactivities.3 For instance, diaryl ethers are
usually recalcitrant to react even under the severe conditions
commonly applied during the hydrothermal, organosolv, or
hydrolytic treatment of lignin.4 To date, hydrogenolysis has
been proposed to depolymerize lignin,5 but the use of harsh
conditions in combination with expensive and nonsustainable
catalysts based on Ru, Rh, Pt, Pd was required.5−8 Furthermore,
the poor selectivity of these strategies leads often to the
formation of mixtures of cycloalkanes and bio-oils that are
difficult to process, while the isolation of pure aromatic
compounds remains challenging. Nickel (Ni)-based catalysts
are highly effective for hydrogenolysis reactions, and several
research groups have been recently focused on investigating
their properties encouraged by the availability and competitive
price of Ni compared to other transition metals.9 The group of
Hartwig10,11 recently reported an exciting protocol for the
cleavage of aryl ethers under mild conditions in xylene based on
a homogeneous Ni catalyst. Following this seminal work and
considering the easier applicability of heterogeneous catalysts
for industrial applications, other groups have investigated the
possible use of heterogeneous Ni systems, including Raney
Ni12,13 and silica supported Ni.14−16
The oxidation of TiN at higher temperatures can be
explained due to the presence of adsorbed oxygen (Figure
S3, Table S1), which at elevated temperatures can promptly
react with TiN. The oxophilicity of titanium is indeed well-
known. Interestingly, pure TiN prepared via the urea route at
Among alternatives to noble metals, nitrides and carbides of
early transition metals have been investigated as they exhibit
similar properties to those of Pt-group metals,17,18 but with the
Received: November 22, 2013
Published: January 17, 2014
© 2014 American Chemical Society
1758
dx.doi.org/10.1021/ja4119412 | J. Am. Chem. Soc. 2014, 136, 1758−1761

Journal of the American Chemical Society
Communication
of morphology from small nanoparticles to an intergrown
aggregate structure led to a decreased surface area of 116 m2
g−1 for TiN-Ni compared to the initial 212 m2 g−1 in TiN. This
could be due to the higher density of the nanocomposites
compared to pure TiN, caused by the presence of 50% Ni
whose density is higher than the one of TiN (ρNi = 8.9 g cm−3,
ρTiN = 5.2 g cm−3). On the other hand, these values may also
reflect the increased agglomeration of the particles following Ni
addition. Noteworthy, pure Ni particles showed a relatively
lower surface area of ∼30 m2 g−1 and were characterized by a
lower homogeneity, as deduced by SEM analysis (Figure S10).
In order to evaluate the catalytic activity of the obtained
materials, the reduction of nitrobenzene to aniline in the
presence of hydrogen has been studied as simple model
reaction. All the catalytic tests were performed using a fixed bed
flow reactor (H-Cube Pro) equipped with a packed cartridge of
Ni, TiN, or TiN-Ni (500 °C) catalyst and provided with a gas
(H2) and a liquid feed. We commenced testing TiN for the
hydrogenation reaction of nitrobenzene to aniline with
temperatures ranging between 25 and 100 °C, but no
conversion was observed even at high temperatures (Table
S4, entry 1). Ni particles showed in turn moderate activity at 50
°C (entry 2), while good (80%) to high (>99%) conversions
could be achieved at 75 and 100 °C, respectively (entries 3, 4).
Interestingly, the improved activity of the TiN-Ni could be
readily appreciated as full conversion was obtained at 25 °C
(entry 5). These interesting preliminary results characterized Ni
and TiN-Ni as promising hydrogenation catalysts.
Therefore, we went on testing the more challenging
hydrogenolysis of aryl ethers. We focused our attention on
model molecules that contain the most common aryl ether
linkages found in lignin (Tables 1 and 2). Initially, we studied
the hydrolysis of benzyl-phenyl ether 1 as a model for the α-O-
4 linkage in lignin. In analogy to the reduction of nitrobenzene,
hydrogenolysis of 1 in the presence of TiN as catalyst did not
show detectable conversion (Table 1, entry 1). In contrast, Ni
nanoparticles were effective in promoting this hydrogenolysis,
and despite the low activity up to 125 °C, 90% conversion
could be obtained at 150 °C (entries 3 and 4). Nevertheless,
TiN-Ni composite showed a superior activity affording full
conversion already at 125 °C (entry 6). GC analysis revealed
the presence of phenol and toluene as the sole reaction
products.
Figure 1. XRD patterns of TiN, Ni, and TiN-Ni at different
temperatures. *: metallic Ni; †: TiO2; •: TiN.
higher temperature (up to 800 °C) has never shown TiOx as
side products.28 A shift (∼ 0.5 θ) to higher angles in the TiN
peak is observed in Figure 1 after Ni addition. A similar effect is
observed in the patterns of Ni-titanates compounds.29 Never-
theless, the presence of Ni-titanate and other possible side
products like TiC was excluded by XRD analysis (Figure S6).
To prove that the shift in the diffraction pattern of TiN is not
caused by the secondary thermal treatment, fresh TiN was
reheated at 500 °C without addition of Ni, and the resulting
XRD pattern (Figure S7) remained unchanged. Interestingly,
all attempts to prepare TiN-Ni in a one-pot procedure, by
direct mixing of Ni precursors and the Ti@urea gel, afforded
only TiO2-Ni mixed phases (Figures S4, S5). In order to
compare and understand the change in morphology and
reactivity of the TiN-Ni nanocomposite, pure metallic Ni
particles were also synthesized by carbothermal reduction of
NiAc. The TEM images show small and monodisperse primary
TiN nanoparticles (Figure 2A) and the TiN-Ni composite
material (Figure 2B) obtained after Ni addition (see also Figure
S8).
The cleavage of the alkyl aryl ether bond in phenylethyl
phenyl ether 2 as a model for β-O-4 linkages in lignin occurred
under more severe conditions, in line with its higher
dissociation energy compared to 1.15 Once again, TiN showed
no activity. Ni particles were also unable to promote conversion
up to 150 °C. However, TiN-Ni catalyst presented a conversion
of 65% at 150 °C with a flow rate of 0.5 mL min−1, while full
conversion could be achieved using a lower volumetric flow
(0.3 mL min−1) as can be seen in entries 9 and 10. In this case,
chromatographic analysis revealed the formation of 5 and 7 as
the sole products. Interestingly, 7 was not detected during the
hydrogenolysis of 1 at 125 °C, while its formation becomes
considerable at 150 °C, probably reflecting a faster hydro-
genation rate for the phenol at this temperature.
Figure 2. TEM pictures of (A) TiN and (B) TiN-Ni prepared at 500
°C.
As it can be deduced, the morphology of TiN nanoparticles
was only slightly affected by the presence of Ni. The
simultaneous presence of Ni and titanium in the final
nanoparticles was confirmed by EDX investigation (see Figure
S7). Interestingly, while titanium was always found together
with Ni, few titanium-free areas could also be detected.
Nitrogen sorption analysis was performed to estimate the
surface area of Ni, TiN, and TiN-Ni (Figure S11). The change
Accordingly, control experiments on phenol 3 at 150 °C and
0.3 mL min−1 resulted in a conversion of 55% into 7.
The 4-O-5 bond is one of the strongest ether bonds in lignin,
and its hydrogenolysis was modeled using diphenyl ether 8
(Table 2). Unlike Ni and TiN (Table 2, entries 1, 2), TiN-Ni
can successfully cleave diphenyl ether, and the best results
1759
dx.doi.org/10.1021/ja4119412 | J. Am. Chem. Soc. 2014, 136, 1758−1761

Journal of the American Chemical Society
Communication
Table 1. Effect of Ni, TiN, and TiN-Ni Catalysts on Hydrogenolysis of Different Lignin Model Molecules
c
selectivity (%)
c
productivity (mmol h−1 gcat
)
−1
entry
cat.
substrate
T (°C)
conversion (%)
4
5
6
7
a
1
TiN
Ni
1
1
1
1
1
1
2
2
2
2
3
150
100
125
150
100
125
150
150
150
150
150
0
0
−
−
−
−
−
−
−
−
−
−
57
54
−
−
−
−
−
−
−
−
−
−
−
−
−
a
2
a
3
Ni
18
90
67
>99
0
51
57
53
54
−
−
−
−
−
49
43
47
46
−
−
−
−
−
0.225
1.282
0.666
1.013
−
a
4
Ni
a
5
TiN-Ni
TiN-Ni
TiN
a
6
a
7
b
8
Ni
0
−
a
9
TiN-Ni
TiN-Ni
TiN-Ni
65
>99
55
43
46
100
0.695
0.608
0.620
b
10
b
11
a
b
c
Conditions: 0.05 M solution of starting material in EtOH, 12 bar, 0.5 mL min−1. 0.3 mL min−1. Determined by GC-MS analysis.
Table 2. Effect of Ni, TiN, and TiN-Ni Catalysts on Hydrogenolysis of Different Diaryl Ethers
c
selectivity (%)
c
productivity (mmol h−1 gcat
)
−1
entry
cat.
substrate
T (°C)
conversion (%)
10
11
12
13
14
15
16
a
1
TiN
Ni
8
8
8
8
9
150
150
125
150
150
0
−
−
42
5
−
−
−
−
−
−
−
−
−
−
−
35
46
49
−
−
−
−
−
−
−
−
−
25
−
−
23
49
4
−
−
0.223
a
2
<5
68
99
65
a
3
TiN-Ni
TiN-Ni
TiN-Ni
b
4
0.259
bd
,
5
−
22
b
0.179
a
c
d
Conditions: 0.025 M of starting material in EtOH, 12 bar, 0.5 mL min−1. 0.3 mL min−1. Determined by GC-MS analysis. Uncorrected GC data.
could be obtained at 150 °C and 0.3 mL min−1 volumetric flow,
with a feed concentration of 25 mM (entry 4). Interestingly,
during hydrogenolysis of compound 8 we observed the
formation of 10 at low conversion (entry 3). In contrast,
traces of 10 were observed at 150 °C (entry 4), leading to the
conclusion that, at this temperature, dearomatization does not
compete with C−O cleavage. A comparison of the selectivity
for the hydrogenolysis of compound 8 at full conversion proved
TiN-Ni highly competitive with respect to representative
commercial hydrogenation catalysts, including Pd/C and
Raney Ni (Table S5). To investigate a possible selectivity in
the hydrogenolysis of substituted 4-O-5 linkages, compound 9
was also hydrogenated using the conditions optimized for
compound 8 (entry 5). According to GC analysis, the mixture
of products is mostly composed of benzene and 1,4
cyclohexanediol revealing a preference of the catalyst to
promote the cleavage of the C−O bond associated with the
most electron-deficient phenyl unit.
its stability vs time on stream. The efficiency of the particles
was tested up to ∼40 h of reaction at 150 °C and remained
unaffected. Also structural features remained unchanged as
confirmed by XRD (Figure S12) performed after catalyst
utilization. A preliminary and qualitative hydrogenolysis test on
commercial lignin samples under nonoptimized conditions
showed the formation of an array of low molecular weight
aromatic and aliphatic products (Figure S14) anticipating the
possible application of the catalysts for depolymerization of real
lignin samples. Noteworthy, the use of this new nanocomposite
in combination with a fixed bed reactor resulted particularly
effective in achieving high productivities. Considering a void
volume of 0.5−0.8 mL for the catalyst bed and a flow rate of 0.3
mL min−1, high conversions could be obtained with very short
residence times (1.5−2.6 min) with respect to previously
published data.11,15 Although additional insights will be
required in order to understand the improved activity of the
new composites with respect to pure Ni, a speculative
interpretation of the physochemical effects operating in the
catalysts can be done on the basis of the observed results. The
Having proved the applicability of the catalyst to the cleavage
of different aryl ethers, we performed a qualitative evaluation of
1760
dx.doi.org/10.1021/ja4119412 | J. Am. Chem. Soc. 2014, 136, 1758−1761

Journal of the American Chemical Society
Communication
partial oxidation of TiN in the nanohybrid at elevated
temperatures (Figure 1), together with the oxidative stability
of Ni, suggests an electron transfer from the Ni to the TiN. As
both component are metals and structurally tightly bonded,
electrons will flow from the less noble (Ni) to the more noble
(TiN) side of the composite in order to equilibrate the Fermi
levels.23 The resulting electron-deficient Ni, for instance, might
more effectively coordinate the oxygen of ether linkages and
activate them for hydrogenation, while accelerating the hydride
transfer to the arene during reductive elimination. Nevertheless,
a classical binary catalysis scheme in which ether linkages are
activated by available Ti-coordination sites on the TiN
substructure and hydrogen is activated by Ni cannot be
excluded. Finally, a possible templating effect in which TiN
could favor the formation of Ni particles with optimal
morphology and site density remains to be elucidated. A
more detailed reaction mechanism will be provided on the basis
of additional studies.
In conclusion, we introduced a new TiN-Ni nanocomposite
composed of spherical intergrown core−shell nanoparticles of
∼10 nm in diameter, as estimated both via XRD and HR-TEM.
This novel material was synthesized on a multigram scale and
proved to be stable at room temperature both in air and in
solvent for several months. The close contact between TiN and
Ni was observed both via peak position and intensity changes in
the corresponding XRD pattern. Furthermore, a different
chemical behavior and catalytic activity was observed with
respect to pure Ni and TiN. While simple Ni particles were
found to be less efficient for the reductive splitting or
hydrogenolysis of lignin model molecules, a superior catalytic
activity of the TiN-Ni nanocomposite was shown, even with the
very recalcitrant diphenyl ether linkages.
REFERENCES
■
(1) Boerjan, W.; Ralph, J.; Baucher, M. Annu. Rev. Plant Biol. 2003,
54, 519.
(2) Marshall, A.-L.; Alaimo, P. J. Chem.Eur. J. 2010, 16, 4970.
(3) Wang, X.; Rinaldi, R. ChemSusChem 2012, 5, 1455.
(4) Laskar, D. D.; Yang, B.; Wang, H.; Lee, J. Biofuels, Bioprod.
Biorefin. 2013, 7, 602.
(5) Zakzeski, J.; Bruijnincx, P. C. A.; Jongerius, A. L.; Weckhuysen, B.
M. Chem. Rev. 2010, 110, 3552.
(6) Michael Robinson, J.; E. Burgess, C.; A. Bently, M.; D. Brasher,
C.; O. Horne, B.; M. Lillard, D.; M. Macias, J.; D. Mandal, H.; C. Mills,
S.; D. O’Hara, K.; T. Pon, J.; F. Raigoza, A.; H. Sanchez, E.; Villarreal,
S. J. Biomass Bioenergy 2004, 26, 473.
(7) Pepper, J. M.; Hibbert, H. J. Am. Chem. Soc. 1948, 70, 67.
(8) Pepper, J. M.; Lee, Y. W. Can. J. Chem. 1969, 47, 723.
(9) Alonso, F.; Riente, P.; Yus, M. Acc. Chem. Res. 2011, 44, 379.
(10) Sergeev, A. G.; Hartwig, J. F. Science 2011, 332, 439.
(11) Sergeev, A. G.; Webb, J. D.; Hartwig, J. F. J. Am. Chem. Soc.
2012, 134, 20226.
(12) Wang, X. Y.; Rinaldi, R. Energy Environ. Sci. 2012, 5, 8244.
(13) Wang, X.; Rinaldi, R. Angew. Chem., Int. Ed. 2013, 52, 11499.
(14) Toledano, A.; Serrano, L.; Balu, A. M.; Luque, R.; Pineda, A.;
Labidi, J. ChemSusChem 2013, 6, 529.
(15) He, J.; Zhao, C.; Lercher, J. A. J. Am. Chem. Soc. 2012, 134,
20768.
(16) Toledano, A.; Serrano, L.; Labidi, J.; Pineda, A.; Balu, A. M.;
Luque, R. ChemCatChem 2013, 5, 977.
(17) Levy, R. B.; Boudart, M. Science 1973, 181, 547.
(18) Hwu, H. H.; Chen, J. G. Chem. Rev. 2004, 105, 185.
(19) Villa, A.; Campisi, S.; Giordano, C.; Otte, K.; Prati, L. ACS
Catal. 2012, 2, 1377.
(20) Ji, N.; Zhang, T.; Zheng, M.; Wang, A.; Wang, H.; Wang, X.;
Chen, J. G. Angew. Chem., Int. Ed. 2008, 47, 8510.
(21) Yao, W.; Makowski, P.; Giordano, C.; Goettmann, F. Chem.
Eur. J. 2009, 15, 11999.
The nanocomposite worked as an efficient catalyst in a fixed
bed reactor, and using ethanol as an environmentally benign
solvent afforded high conversions in very short reaction times.
These promising results open the way to new competitive
lignin refining strategies based on relatively cheap and abundant
materials like Ti and Ni. A detailed investigation of the reaction
mechanism including the influence of other solvents will be the
subject of future studies. Finally, this work anticipates the
development of an extended family of new catalysts, whose
activity is based on the interaction of early transition-metal
nitrides with Ni, which can find application in different fields of
catalysis.
(22) Esposito, D. V.; Hunt, S. T.; Kimmel, Y. C.; Chen, J. G. J. Am.
Chem. Soc. 2012, 134, 3025.
(23) Li, X.-H.; Antonietti, M. Chem. Soc. Rev. 2013, 42, 6593.
(24) Hammerle, H.; Kobuch, K.; Kohler, K.; Nisch, W.; Sachs, H.;
̈
Stelzle, M. Biomaterials 2002, 23, 797.
(25) Birkholz, M.; Ehwald, K. E.; Wolansky, D.; Costina, I.;
Baristiran-Kaynak, C.; Frohlich, M.; Beyer, H.; Kapp, A.; Lisdat, F.
̈
Surf. Coat. Technol. 2010, 204, 2055.
(26) Heide, N.; Schultze, J. W. Nucl. Instrum. Methods Phys. Res., Sect.
B 1993, 80−81 (Part1), 467.
(27) Marlo, M.; Milman, V. Phys. Rev. B 2000, 62, 2899.
(28) Giordano, C.; Erpen, C.; Yao, W.; Milke, B.; Antonietti, M.
Chem. Mater. 2009, 21, 5136.
(29) pattern of Ni-titanates can be found under http://www.icdd.
com.
ASSOCIATED CONTENT
■
S
* Supporting Information
Additional experimental and characterization details. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
Davide.Esposito@mpikg.mpg.de
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We gratefully thank Dr. Guylhaine Clavel for TEM measure-
ments and fruitful discussion and the Max Planck-Fraunhofer
cooperation scheme for financial support.
1761
dx.doi.org/10.1021/ja4119412 | J. Am. Chem. Soc. 2014, 136, 1758−1761