Hydrogenation of Benzonitrile over Supported Pd Catalysts: Kinetic and Mechanistic Insight

Article abstract of DOI:10.1021/acs.oprd.9b00058

The liquid phase hydrogenation of benzonitrile over a 5 wt % Pd/C catalyst using a stirred autoclave is investigated. The reaction conforms to a consecutive reaction sequence: first benzonitrile is hydrogenated to produce benzylamine, which subsequently u

Full text of DOI:10.1021/acs.oprd.9b00058

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE
Full Paper
The hydrogenation of benzonitrile over supported
Pd catalysts: kinetic and mechanistic insight
Mairi McAllister, Cedric Boulho, Lauren Gilpin, Liam McMillan, Colin Brennan, and David Lennon
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.9b00058 • Publication Date (Web): 27 Mar 2019
Downloaded from http://pubs.acs.org on March 28, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
The hydrogenation of benzonitrile over supported Pd
catalysts: kinetic and mechanistic insight
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
a
a
a
Mairi I. McAllister , Cédric Boulho , Lauren F. Gilpin , Liam McMillan , Colin
b
a
Brennan and David Lennon *
a. School of Chemistry, Joseph Black Building, University of Glasgow, Glasgow G12 8QQ,
UK.
b. Syngenta, Jeallot’s Hill International Research Centre, Berkshire RG42 6EY, UK.
Corresponding author:-
Professor David Lennon CChem, FRSC
School of Chemistry
Joseph Black Building
The University of Glasgow
Glasgow, G12 8QQ.
U.K.
Email:
David.Lennon@glasgow.ac.uk
Telephone: (+44)-(0)-141-330-4372
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 35
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Abstract
The liquid phase hydrogenation of benzonitrile over a 5wt% Pd/C catalyst using a stirred
autoclave is investigated. The reaction conforms to a consecutive reaction sequence: firstly
benzonitrile is hydrogenated to produce benzylamine, which subsequently undergoes a
hydrogenolysis step to form toluene. Benzonitrile hydrogenation obeys first order kinetics with
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
−
1
an activation energy of 27.6 kJ mol . In contrast, the benzylamine hydrogenolysis stage obeys
−1
zero order kinetics and exhibits an activation energy of 80.1 kJ mol . A 1wt% Pd/Al O catalyst
2
3
is additionally examined, which is also seen to support hydrogenolysis activity alongside the
hydrogenation pathway. Gas phase transmission infrared spectroscopic measurements of the
hydrogenation of benzonitrile and benzylamine over the 1wt% Pd/Al O catalyst utilising
2
3
hydrogen and deuterium are undertaken, which enable reaction schemes incorporating
adsorption geometries of intermediate adsorption complexes to be proposed.
Keywords
Nitrile hydrogenation; carbon-supported palladium; hydrogenolysis; benzonitrile; benzylamine.
ACS Paragon Plus Environment
2

Page 3 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
1
.
Introduction
Primary aromatic amines represent an important class of chemicals with widespread
1
application in many areas of the chemical industry . Several different routes to these
compounds are available, including the heterogeneously catalysed hydrogenation of nitriles 2.
A particular example of this would be the industrial production of hexamethylenediamine, a
precursor of nylon-6,6, via the heterogeneously catalysed hydrogenation of adiponitrile 3.
Typically, skeletal metal catalysts such as Raney Co or Raney Ni,4,5 have been favoured for
hydrogenations of fine chemicals due to a high resultant selectivity. Nevertheless, such
catalysts are often found to be too active in a fine chemicals environment, where additional
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
functionality may be compromised by over-hydrogenation . As a result, use of supported Pd
catalysts has become a more favorable choice to facilitate the transformation of the nitrile
functionality.
The hydrogenation of a range of aromatic nitriles over a commercially available 5wt% carbon
supported Pd catalyst has recently been reported by McMillan and co-workers 7,8. One of the
substrates considered, benzonitrile, was notable due to the occurrence of a hydrogenolytic
pathway alongside hydrogenation of the nitrile moiety that resulted in the selective production
of toluene. As the simplest aromatic nitrile, benzonitrile provides a good bridging point
between aliphatic nitriles, already extensively researched,9–11 and more complex, substituted
species that have greater industrial relevance in several sectors of the chemical manufacturing
industry such as agri-chemicals, pharmaceuticals and fine chemicals. A number of other
researchers have also reported on this topic, and related reaction systems, as outlined below.
On examining the liquid phase hydrogenation of benzonitrile over a Pd/C catalyst, in 2005
Hegedűs and Máthé reported complete nitrile conversion and high benzylamine selectivity
12
(95%) . The high selectivity towards the primary amine was obtained under mild conditions
(303 K, 6 bar H ) in the presence of an acid additive. Nonetheless, the presence of a
2
1
2
hydrogenolysis pathway to yield toluene was also acknowledged . Similarly, the presence of a
ACS Paragon Plus Environment
3

Organic Process Research & Development
Page 4 of 35
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
hydrogenolysis step as a side reaction when hydrogenating nitriles was also reported by Yap
and co-workers 13.14
.
Historically, this heteroatom cleavage has not been a predominant feature of the
heterogeneous catalysis literature, with the phenomenon thought to be the dominant process
of the reaction only under ‘harsh’ reaction conditions 14,15. Despite this, selective production of
the hydrogenolysis product under mild conditions has been reported; for example, by
Maschmeyer et al.,14 through the study of benzonitrile over a Pd/C catalyst. Given the
industrial importance of primary amines, the occurrence of concurrent or consecutive
hydrogenolysis pathways alongside hydrogenation reactions constitutes a topic worthy of
further investigation.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Interestingly, the hydrogenation of 4-hydroxybenzyl cyanide to form tyramine studied by
McAllister and co-workers did not yield the hydrogenolysis product (4-hydroxyethylbenzene) 8.
This outcome was rationalized with reference to work undertaken by Maschmeyer et al.,13
which proposed that it is the positioning of the aromatic group with respect to the heteroatom
that dictates the viability of a post-hydrogenation hydrogenolysis stage for aromatic nitriles. If
the α-carbon, with respect to the heteroatom, is adjacent to the aromatic ring then stabilization
of the carbon by the aromatic ring may occur making cleavage under mild conditions favorable.
Thus, as is the case for tyramine, when the α-carbon is not adjacent to the aromatic ring, the
phenomenon cannot occur and hence heteroatom cleavage is not observed 8.
Examining the hydrogenolysis process further, Maschmeyer and co-workers found that on
substrates with aromatic groups adjacent to the nitrile functionality, a greater degree of
hydrogenolysis could be achieved by using a high catalyst loading. Subsequent kinetic
investigation by this group identified that the hydrogenation of benzonitrile was first order but,
conversely, the hydrogenolysis of benzylamine was found to be zero-order. Zero-order kinetics
are often observed when the catalyst active sites are fully saturated with the substrate due to
ACS Paragon Plus Environment
4

Page 5 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
favorable substrate-surface interactions. This finding was hence attributed to strong
benzylamine adsorption to the catalyst surface 14.
Bakker et al. further highlight the importance of this hydrogenolytic pathway in their study of
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the hydrogenation of benzonitrile over a 5wt% alumina supported Pd catalyst, where toluene
1
6
(C H CH ) was observed as a by-product . A role for a Pd-H phase at the catalyst surface that
6
5
3
is populated at hydrogen pressures in excess of 10 bar was thought to be responsible for
preventing the hydrogenolysis reaction, whilst increasing the formation of the coupling product
1
6
dibenzylamine . It is suggested that the toluene can be formed by two distinct processes: (i)
hydrogenolysis of benzylamine and (ii) direct hydrogenolysis of benzonitrile 16.
The liquid phase, variant of this reaction is predominant in the literature with a number of
factors identified as important in achieving high selectivity; these include catalyst choice,
reaction conditions, the starting material, and hydrogen availability 17,18. In a move towards the
selective production of benzylamine, more exotic experimental conditions have been
examined. For example, Chatterjee et al. report, in the absence of additives, both high
1
9
conversion (90%) and selectivity (91%) in supercritical CO over a Pd/MCM-41 catalyst . Whilst
2
this methodology shows promise for ‘clean’ chemical processes the energy demands associated
with supercritical CO , at present, lowers the overall efficiency of the process.
2
As an extension, however, continuous hydrogenation of benzonitrile in the gas phase is
comparatively less well studied 20–22. Such studies note that residence time also plays a key role
2
0
in achieving high selectivity to the desired product . Moreover, a shorter contact time in a
continuous gas phase operation can prevent condensation of the imine with benzylamine to
yield unwanted by-products 20,22. Keane and co-workers report that, whilst toluene was the
predominant product of benzonitrile hydrogenation over Pd/C, it was benzylamine that was
favoured over Pd/Al O . Despite this, hydrogenolysis to toluene was observed at the higher
2
3
2
contact times required for sufficient conversion to be achieved. The use of multiple Pd/Al O3
2
2
beds in series was thus used as a successful means of retaining full benzylamine selectivity .
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 35
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Use of a Pd/Al O catalyst in the hydrogenation of benzonitrile in a fixed-bed reactor has also
2
3
2
3
been employed by Zhang et al. . Deactivation studies on this system indicated that the loss in
activity observed at extended run times could be attributed to carbonaceous compounds
deposited on the surface of the catalyst. The irreversibility of the bonding of these species to
the surface causes loss of catalytic activity as a consequence of active site obstruction.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
With reference to the aforementioned literature, it is clear that there are complications,
particularly under liquid phase batch conditions, arising in the hydrogenation of benzonitrile.
The batch variant of this process is commonly plagued by unwanted side reactions such as
7
hydrogenolysis to yield toluene, compromising primary amine yield. Further insight into the
interconnectivity of the hydrogenation and hydrogenolysis pathways associated with
benzonitrile hydrogenation over heterogeneous Pd catalysts in the liquid phase is therefore
desirable.
Against this background, further investigation into the hydrogenation of benzonitrile over the
same 5wt% Pd/C reference catalyst used in previous investigations of aromatic nitrile
hydrogenations by McMillan and MCAllister 7,8 is desirable. The results section of the paper
comprises four parts. Section 1 examines the kinetics and energetics of the liquid phase
process over the 5wt% Pd/C reference catalyst in order to ensure this catalyst’s connectivity to
recent literature studies. With those benchmark measurements undertaken and the
applicability of the autoclave measurements established, Section 2 undertakes a new analysis
of the kinetic data in order to establish the conformity of the hydrogenation/hydrogenolysis
stages to a consecutive process. Following this, Section 3 briefly examines of the effectiveness
of (i) Adams catalyst (PtO ) and (ii) a 1wt% Pd/Al O catalyst for benzonitrile hydrogenation in
2
2 3
the liquid phase. Finally, Section 4 addresses mechanistic issues utilizing the 1wt% Pd/Al O
2
3
catalyst applied to the gas phase hydrogenation of benzonitrile and benzylamine, with an
adapted IR gas cell acting as a batch reactor that can be operated with either hydrogen or
deuterium as the reductant. Identification of deuterium incorporation within the benzylamine
ACS Paragon Plus Environment
6

Page 7 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
and toluene products provides new information on the form of the adsorption complexes
linked to the hydrogenation and hydrogenolysis steps. Consequently, reaction schemes are
proposed to describe (i) the deuteration of benzonitrile resulting in the formation of toluene-d3
and (ii) the direct deuterolysis of benzylamine resulting in formation of toluene-d1.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
.
Experimental
2
.1
Liquid phase reactions
The liquid phase hydrogenation reactions were carried out in a 500 mL stirred autoclave (Büchi
Glas Uster). An automated gas flow controller (BPC 1202) allowed the delivery of inert (N2,
BOC, 99.999% purity) and active (H , BOC, ≥99.995% purity) gases directly to the reactor via a
2
gas reservoir. The reaction vessel was heated by silicon oil passed around the double jacketed
reactor via a heating circulator (Julabo F25). The hydrogen uptake provided a direct indication
of reaction rate and hydrogen consumption.
The reactor was first charged with the catalyst (500 mg), and solvent (300 mL propan-2-ol;
Riedel-de Haën, 99.5% purity) prior to being purged with inert gas (N ). Reduction of the
2
catalyst under 4 barg of hydrogen at 300 r.p.m. was carried out for 1 hour whilst the
catalyst/solvent mixture was heated to reaction temperature (333 K). In a separate vessel, ca.
1
7 mmol of the reagent (benzonitrile; Sigma-Aldrich, 99.9% purity or benzylamine; Sigma-
Aldrich, 99% purity) was dissolved in 50 mL of solvent and degassed under a constant flow of
helium (He, BOC, 99.999% purity). Upon addition of the solubilized and degassed reagent,
stirring was stopped, the reactor sealed and the hydrogen pressure raised. Once reaction
pressure (4 barg) was obtained, the reaction mixture was stirred at 800 r.p.m. and samples
taken periodically via an outlet valve throughout the course of the reaction. An agitation rate
of 800 r.p.m. corresponded to the plateau region of a plot of benzonitrile consumption rate
versus agitation speed, indicative of operation under a chemical/kinetic regime. No auxiliary
reagents, e.g. acid, were used in any of the test reactions. All reactions were performed at least
in duplicate, with representative profiles presented here.
ACS Paragon Plus Environment
7

Organic Process Research & Development
Page 8 of 35
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Liquid samples were collected periodically and filtered (Minisart 0.2 µm single use sterile filter)
to remove any catalyst residue with analysis being undertaken off-line. Gas-liquid
chromatograph (GLC) analysis was carried out on a Perkin-Elmer 8500 Series Gas
Chromatograph fitted with a Varian Chrompack CP-Sil 8 CB column (30 m x 0.20 mm ID, 0.33
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
µm film). Initial concentration of starting material (A ) values were calculated based on the
o
amount of pre-weighed starting material added to the reactor prior to commencement of
reaction. Errors in GLC output and hydrogen consumption were determined by repeat analysis
of a known stock of standards and repeat reactions respectively, with both being found to be
below 5%.
For the purposes of this study three catalysts have been used. The primary catalyst utilized for
the liquid phase investigation of benzonitrile hydrogenation was a commercial grade 5wt%
Pd/C (Sigma Aldrich, code number: 205680). This sample was intended to represent a generic
hydrogenation catalyst similar to that typically encountered in a fine chemicals setting. The
number of Pd surface atoms per gram of catalyst, calculated from CO adsorption isotherms,
2
0
was found to be 1.11 x10 corresponding to a metal dispersion of 39% and a mean particle size
2
-1
of 2.8 nm. The catalyst exhibited a BET surface area of 774  19 m g , with the error
representing the standard deviation of three replicate measurements. Further characterization
of this catalyst is reported elsewhere 7.
Two other catalysts were examined to a lesser degree. Firstly, due to Adams catalyst (PtO2)
being reported as a generic hydrogenation catalyst effective in preserving the amine
functionality in nitro group reduction reactions,24–28 a commercial grade PtO (Sigma Aldrich,
2
code number: 206032) was briefly investigated. Secondly, an in-house 1wt% Pd/Al O catalyst
2
3
previously used in IR analysis of gas phase catalytic reactions was additionally explored. The
wt% Pd/Al O catalyst was prepared via a wet impregnation process that is described
1
2
3
elsewhere 29,30. Carbon monoxide adsorption isotherm measurements established a value of
1
8
7.6 x10 surface palladium atoms per gram of catalyst, corresponding to a metal dispersion of
ACS Paragon Plus Environment
8

Page 9 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
2
-
1
3.4% and a mean particle size of 8.3 nm. The catalyst exhibited a BET surface area of 196 m g
1 30.
2.2
Vapour phase infrared spectroscopic measurements
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The 1wt% Pd/Al O catalyst was selected for use in the gas phase IR experiments. The lower Pd
2
3
loading compared to the reference Pd/C catalyst facilitated a slower benzonitrile conversion,
thereby enabling the acquisition of sufficient IR spectra with acceptable signal to noise ratio to
define a reaction profile over the full reaction coordinate (i.e. benzonitrile  benzylamine 
toluene). A methodology used previously was adopted 30,31. Briefly, the apparatus consisted of
a modified heated gas cell (Graseby-Specac 5660) housed within a Nicolet-Avatar 360 FTIR
spectrometer 30,31. The IR cell was connected to the spectrometer’s internal gas purge facility
via rubber bellows, thereby ensuring that there were minimal complications to the resultant IR
spectra as a result of contributions from atmospheric water or carbon dioxide signals. Brooks
mass flow controllers allowed delivery of inert (He, BOC, 99.999 % purity) and active (H , BOC, ≥
2
99.995% purity and D , BOC, > 99.8% purity) gases to be delivered directly to the reactor.
2
Approximately 500 mg of the 1wt% Pd/Al O catalyst was placed in a glass boat and loaded into
2
3
the gas cell outside of the path of the IR beam, ensuring only gaseous species were analyzed
during the reaction. The catalyst was reduced under a flow of 10% H /90% He at a rate of 20
2
−1
mL min whilst being heated to 353 K. The hydrogen composition was subsequently increased
to yield an equimolar mixture of hydrogen and helium at a constant flow rate. After 10 minutes
the reactor exit valve was closed allowing a pressure of 1.2 bar to build. At this pressure, the
reactor entrance valve was closed, sealing the reactor. A liquid chromatography syringe
(
Hamilton Bonaduz) was used to inject a 10 µL aliquot of nitrile reagent into the reactor via a
septum.
In situ measurements of the gas phase species were recorded as a function of time, with t = 0
being defined as the time the injection of the liquid phase reagent into the reactor was
−1
complete. Infrared spectra were recorded at a resolution of 4 cm by the co-addition of eight
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 35
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
scans. This procedure resulted in a total acquisition time of approximately 10 seconds. The
reacting gases obey the Beer-Lambert law, thereby permitting the reliable calibration of
reagents by the integration of infrared intensity of particular vibrational features. Due to
overlap of diagnostic aromatic and aliphatic CH stretching bands with those of toluene,
quantification of benzylamine could not easily be determined. Nevertheless, it was possible for
semi-quantitative reaction profiles to be constructed. All measurements were performed at
least in duplicate, with representative datasets presented here.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3.
Results
3
3
.1
Kinetic studies (liquid phase, 5wt% Pd/C)
.1.1 Benzonitrile hydrogenation: reaction order and energetics
McMillan and co-workers previously reported on benzonitrile hydrogenation over the same
7
5
wt% Pd/C catalyst under consideration here . On cessation of hydrogen uptake, indicating
reaction completion, the nitrile species was fully converted but it was the hydrogenolysis
product toluene, formed following the hydrogenation step, which was produced with 100%
7
selectivity . Interestingly, there were no coupling reactions to form the corresponding higher
amines in this instance. This finding is attributed to the catalyst choice. Supported Pd is found
to be an effective catalyst for both hydrogenation and hydrogenolysis reactions. As a
consequence of this, the concentration of both the imine intermediate and benzyl amine will be
low in the system due to the occurrence of hydrogenation and hydrogenolysis of these
molecules respectively. This then means that both of the species required for coupling to form
the secondary amine (dibenzylamine) are not prevalent in the reaction mixture, thus providing
a rational for the absence of dibenzylamine formation in a system without additives. As an
7
extension of McMillan’s studies, the rate law and energetics of both hydrogenation and
hydrogenolysis processes are explored here.
ACS Paragon Plus Environment
1
0

Page 11 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. First order data fit (ln(A /A ) versus time) for benzonitrile hydrogenation over
o
t
Pd/C at 333 K, 4 barg, 800 r.p.m. and ca. 17 mmol benzonitrile. A and A represent initial
o
t
concentration and concentration at time t respectively.
Analysis of the benzonitrile concentration versus time dataset at 333 K, where the experimental
data is evaluated for correspondence to first order kinetics, is presented in Figure 1. A highly
correlated linear plot is obtained, indicating the reaction conforms to a first order regime. An
−2
−1
associated rate coefficient of 5.9 x10 min is obtained. This outcome is in agreement with
Maschmeyer et al. who also reported that benzonitrile hydrogenation over a Pd catalyst
1
4
followed first order reaction kinetics .
ACS Paragon Plus Environment
1
1

Organic Process Research & Development
Page 12 of 35
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Arrhenius plot for the hydrogenation of benzonitrile (black squares) and the
hydrogenolysis of benzylamine (red circles) over a temperature range of 303–333 K, 4 barg, 800
r.p.m. using 500 mg 5wt% Pd/C catalyst and ca.17 mmol benzonitrile/benzylamine.
Reaction temperature, varied within the range 303–333 K, was found to have a significant
effect on the initial rate of reaction. A series of first order plots used for the calculation of rate
coefficients, as observed in Figure 1, were constructed in triplicate for each measured
temperature in the hydrogenation of benzonitrile. A representative Arrhenius plot is presented
−1
in Figure 2 and yields an activation energy of 27.6 kJ mol with a standard deviation on the
−1
three measurements of 1.8 kJ mol .
−1
An observed activation energy of 20+ kJ mol typically represents a heterogeneously catalyzed
reaction under kinetic control 32,33. Nevertheless, at 27.6 kJ mol the activation energy of
benzonitrile hydrogenation is closer to the value for diffusion control than would be expected
−1
1
6
for a well stirred reaction vessel of this nature . This outcome is therefore suggestive of the
presence of a diffusional constraint within the reaction system. Furthermore, a similar study
ACS Paragon Plus Environment
1
2

Page 13 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
conducted by Hegedüs et al. on the hydrogenation of benzonitrile over a 10% Pd/C catalyst
12
reports two temperature controlled regimes for the activation energy of this reaction . For
−1
temperatures below 323 K an activation of 49.8 kJ mol was quoted indicating the absence of
diffusional constraints. Conversely, however, diffusional constrictions were observed above
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
−1
3
23 K (5.2 kJ mol ). Somewhat interestingly the activation energy reported in this study, 27.6
−1
kJ mol , falls somewhere in the middle of these two values perhaps providing an indication
that operation is within a transition regime 12.
Working on a similar reaction system, Maschmeyer et al. tested for the presence of mass
transfer limitations by altering agitation speed and monitoring the effect on the rate of reaction
1
4
for both benzonitrile hydrogenation and benzylamine hydrogenolysis . Whilst kinetic control
was established for the latter reaction at operational agitation, this was not found to be the
14
case for benzonitrile hydrogenation, with diffusion limitations being apparent . These findings
indicate the presence of gas-to-liquid mass transport issues which have been shown to be
problematic for Pd/C catalysts 17,21,24
.
Figure 3. Effect of catalyst mass on the initial rate of benzonitrile hydrogenation at 333 K, 4
barg, 800 r.p.m. over a catalyst mass range of 0.1–1 g using ca. 17 mmol benzonitrile.
ACS Paragon Plus Environment
1
3

Organic Process Research & Development
Page 14 of 35
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
For the reaction reported here, however, it was found that increasing autoclave agitation speed
did not result in an increase in the benzonitrile hydrogenation rate (Section 2.1), indicating an
absence of a specific gas-to-liquid mass transfer restriction. Moreover, the initial rate was
found to increase linearly with catalyst mass, Figure 3; further reinforcing the absence of gas-
to-liquid transport constraints 17,21,34. Nevertheless, internal pore diffusion within the carbon
support cannot be definitively excluded as a means of providing a mass transfer limitation
within the reaction system.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
.1.2 Benzylamine hydrogenolysis (liquid phase): reaction order and energetics
Figure 4. Reaction profile for the liquid phase hydrogenolysis of benzylamine over 500 mg 5wt%
Pd/C, at 333 K, 4.0 barg, 800 r.p.m. and ca. 16 mmoles of benzylamine. The dashed line
represents the incident concentration of benzylamine.
The second stage of the consecutive reaction, benzylamine hydrogenolysis, was likewise
considered. The profile for this reaction over the 5wt% Pd/C catalyst can be visualised in Figure
4
. Full benzylamine conversion was achieved within 100 minutes, selectively yielding the
ACS Paragon Plus Environment
1
4

Page 15 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
hydrogenolysis product, toluene, and corresponding to a turnover number of 174. On
inspection of the reaction profile, toluene formation appears to follow zero order kinetics, as
evidenced by the near linear increase in toluene production with respect to time. The near
quantitative transformation of benzylamine to toluene indicates hydrogenolysis to be a facile
process under the conditions reported. The hydrogen uptake curve (blue line) closely follows
toluene production. A noticeable mass imbalance is initially observed but, following a similar
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
trend to that seen for benzonitrile hydrogenation, this imbalance is nearly completely
recovered at extended reaction times. At the time corresponding to the maximum mass
imbalance (ca. 15 min) the ratio of adsorbed species to surface palladium is equal to 34.6. It is
thus believed that the origin of the initial mass imbalance is due to adsorption of the amine
onto both the Pd surface and the carbon support, causing a momentary retention of this
species by the catalyst. As the amine undergoes hydrogenolysis toluene is released into the
liquid phase allowing the mass balance to be recovered by reaction completion.
Figure 5. A zero order plot (A −A ) versus time for benzylamine hydrogenolysis over 500
o
t
mg Pd/C catalyst at 333 K, 4 barg, 800 r.p.m. and ca. 16 mmol benzylamine. A and A represent
o
t
initial concentration and concentration at time t respectively. The linear fit does not include the
first data point.
ACS Paragon Plus Environment
1
5

Organic Process Research & Development
Page 16 of 35
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 5 assesses the suitability of the benzylamine hydrogenolysis reaction to zero order
kinetics. The high linearity of the data over three half-lives indeed indicates a conformation of
the hydrogenolysis of benzylamine to zero order kinetics, with a resultant rate coefficient of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
−4
−1
−1
5
.74 x10 mol L min . Therefore, the direct hydrogenolysis of benzylamine can be expressed
as rate=−d[BA]/dt=k where k =k[BA][H ][catalyst]. Here, k represents the observed rate for
2
2
2
2
the transformation of benzylamine to toluene with [BA], [H ] and [catalyst] denoting the
2
concentration of benzylamine, hydrogen and catalyst respectively. As with the hydrogenation
of benzonitrile both [H ] and [catalyst] are constant. However, in addition to this, [BA] can also
2
be considered as a constant due to saturated binding to the surface. This outcome compares
1
4
favourably with previous reports on this reaction system by Maschmeyer and co-workers .
The Arrhenius plot for benzylamine consumption over a temperature range of 303–333 K is
−1
shown in Figure 2. From this plot an activation energy of 80.1 ( 6.4) kJ mol is obtained. The
−1
elevated activation energy for this process (cf. E = 27.6 (± 1.8) kJ mol for benzonitrile
a
hydrogenation, Section 3.1.1) indicates the establishment of distinct kinetic control in this
instance. It is noted that the hydrogenolysis of benzylamine is a single step reaction requiring
only one equivalent of hydrogen. In contrast, the hydrogenation of benzonitrile requires two
equivalents of hydrogen to support a relatively fast reaction that additionally includes a highly
reactive imine intermediate. In combination with the reduced hydrogen demand beyond 100%
conversion, the slower rate of the hydrogenolysis means that the possibility of mass transport
restrictions are less prevalent for this particular reaction step.
3
.2
Kinetics of consecutive reactions (liquid phase, 5 wt% Pd/C): hydrogenation (first order)
followed by hydrogenolysis (zero order)
It has previously been demonstrated that there are two steps, thought to be consecutive in
nature, involved in the reduction of benzonitrile: hydrogenation of the nitrile, followed by
7
hydrogenolysis of the amine functionality . To further investigate the kinetics of the system,
the experimentally gained data may be compared with a data fit for kinetic expressions
ACS Paragon Plus Environment
1
6

Page 17 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
governing a consecutive reaction with a change from first to zero order 35. As such, the
following sequence may be considered:
k1
k2
Benzonitrile
Benzylamine
Toluene
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(1),
where k and k represent the rate coefficients for the hydrogenation of benzonitrile and
1
2
hydrogenolysis of benzylamine respectively. Let [BN], [BA] and [TOL] represent the respective
liquid phase concentrations of benzonitrile, benzylamine and toluene. Manipulation of the
differential forms of the rates of change for both reactant and product gives:
―푘1푡
[퐵푁] = [퐵푁] 푒
(2)
푡
0
―
푘1푡
[
퐵퐴] = [퐵푁] ― [퐵푁] 푒
― 푘2푡
(3)
푡
0
0
Subsequent use of the mass balance relationship, where the sum concentration of all reactants
and products must equal the initial benzonitrile concentration (A ), infers that,
0
[퐵푁] + [퐵퐴] + [푇푂퐿] = [퐵푁]
(4).
푡
푡
푡
0
Therefore,
[
푇푂퐿] = 푘 푡
(5)
푡 2
Equation 5, describing toluene change over time must, however, be considered with care as it
suggests that the concentration of this species in the system increases without limit. In reality,
as visualised in the reaction profile (Figure 4), this is impossible due to the quantity of the
product being a finite parameter. Nevertheless, Figure 6 shows that the experimentally
ACS Paragon Plus Environment
1
7

Organic Process Research & Development
Page 18 of 35
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
obtained data, denoted (obs, closed shapes), fits well with the modelled kinetics (calc, open
shapes).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. A comparison of the reaction profiles for benzonitrile hydrogenation by experimental
means (closed shapes) and by calculation of the kinetic expressions governing a consecutive 1st
th
to 0 order process (open shapes), where [BN], [BA], and [TOL] represent the concentrations of
benzonitrile, benzylamine and toluene respectively.
From Figure 6 it can be seen, as postulated, that the modeled data for toluene formation
follows an approximately linear increase at extended times, whereupon it begins to deviate
from the experimentally observed profile which begins to plateau. Despite this anomaly that is
explained above, the only difference between the experimental and calculated data sets relates
to benzylamine concentration. The maximum concentration of this species observed
experimentally is consistently low when compared to the modeled kinetic data. This is thought
to be due to benzylamine retention by the Pd/C catalyst, as indicated by the momentary mass
imbalance at the beginning of the reaction (Figure 4). Nonetheless, a complete mass balance
was returned on reaction completion, indicating that product formation was not impeded by
this process. Discrepancies between observed and calculated values can also be attributed to
ACS Paragon Plus Environment
1
8

Page 19 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
the fact that the model takes no account of the imine formation step, vital for amine
production, as these species were not detectable in the liquid phase during these experiments.
3.3
PtO and 1% Pd/Al O catalysts (liquid phase)
2
2 3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Catalyst choice was briefly explored with the aim of further understanding if the mechanistic
insights garnered from the 5wt% Pd/C catalysed reaction were unique to this catalyst. Two
alternative precious metal catalysts, similar to that already used, were hence selected for
investigation. Firstly, benzonitrile was hydrogenated over an in-house 1wt% alumina supported
1
8
−1
catalyst. From CO chemisorption isotherm data (7.6 x10 Pd(s) g
) this Pd/Al O catalyst
(cat)
2 3
was found to exhibit a metal dispersion of 13.4% corresponding to a mean particle size of 8.3
3
0
nm . The dispersion quoted for this catalyst is significantly lower than that of the Pd/C
7
reference catalyst (dispersion = 39%, mean particle size = 2.8 nm), providing the opportunity to
assess whether the larger Pd crystallites of the alumina supported catalyst and/or the support
material will have an effect on the sequential hydrogenation and hydrogenolysis reactions
under consideration.
Figure 7. Reaction profile for the liquid phase hydrogenolysis of benzonitrile over 500 mg 1wt%
Pd/Al O , at 333 K, 4.0 barg, 800 r.p.m. and ca. 17 mmoles of benzonitrile.
2
3
ACS Paragon Plus Environment
1
9

Organic Process Research & Development
Page 20 of 35
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
It is recognised that the rate of a reaction and the product distribution can vary with metal
3
5
particle size . In line with the reduced number of surface Pd atoms, Figure 7 shows the 1wt%
Pd/Al O catalyst to have a reduced activity when compared to the 5wt% Pd/C reference
2
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
−1
catalyst. Indeed, the use of the 1% Pd/Al O catalyst afforded a reduced initial TOF (0.113 s )
2
3
−1
when compared to the reference catalyst (initial TOF = 0.234 s ). Moreover, it was found that
the increased particle size of the 1wt% Pd/Al O catalyst did not significantly alter the product
2
3
distribution, only the rate of reaction. At 50% conversion of benzonitrile, selectivity to
benzylamine was 47% for the alumina supported catalyst (cf. 40% selectivity under the same
criteria was reported for the 5wt% Pd/C catalyst). Additionally considered was platinum
dioxide (PtO ), commonly known as Adams catalyst, due to its reported preservation of the
2
amine functionality in the reduction of nitro compounds 36,37. Again, a similar profile was
obtained with 38% benzylamine selectivity at 50% benzonitrile conversion achieved over this
catalyst. Thus, accessibility of the hydrogenolysis pathway is attainable using both Pd and Pt
based catalysts.
3
.4
Gas phase infrared spectroscopic studies
CN
H2
H2
H2
NH2
NH
-
NH3
Benzonitrile
Benzylimine
Benzylamine
Toluene
Scheme 1. Reaction scheme for the hydrogenation of benzonitrile 7.
With reference to Scheme 1, the production of toluene is expected to coincide with ammonia
formation. However, the latter entity is undetectable in the liquid phase by the available
analytical procedures. Given that coincident detection of ammonia alongside toluene
formation would provide supporting evidence for a hydrogenolysis stage, the application of gas
phase infrared spectroscopic measurements was thought to be informative in terms of
ACS Paragon Plus Environment
2
0

Page 21 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
confirming the classification of a hydrogenolysis stage. It is noted that, to the best knowledge
of the authors, to-date and due to analytical constraints, ammonia formation has not been
directly determined during the hydrogenation of benzonitrile.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Previous work from this group has established the usefulness of using IR spectroscopy to
examine the composition of the gas phase present over a supported metal catalyst operating
within a IR gas cell acting as a batch reactor 30,31. A similar approach was adopted here to
explore the benzylamine hydrogenation reaction, with hydrogenation and deuteration
reactions examined over the 1wt% Pd/Al O catalyst. This material was selected in preference
2
3
to the Pd/C catalyst as its relatively impeded rate of reaction (Section 3.4) enabled sufficient
spectra of suitable signal to noise ratio to be recorded in order to define a complete reaction
profile within a time period (ca. 30 min) for which the integrity of the cell could be maintained.
Diagnostic Band / cm−1
Compound
Benzonitrile
Toluene
CN stretch at 2238
CH bend at 729
NH wag at 950
Ammonia
Table 1. The diagnostic bands used to distinguish each compound during the course of the
reaction. Due to overlapping peaks, a diagnostic band for benzylamine was not assigned 39.
The operational temperature of the gas phase IR measurements of 353 K was an experimental
compromise. At temperatures in excess of this, over-hydrogenation of benzonitrile to yield
fully saturated ring structures occurred, whilst at lower temperatures there was insufficient
volatilisation, as evidenced by loss of reactant concentration in the vapour phase. Diagnostic
39
bands used for the identification of each compound are presented in Table 1 . Unfortunately,
due to peak overlap, a diagnostic band was not accessible for benzylamine, meaning that an
accurate concentration for this compound could not be obtained.
ACS Paragon Plus Environment
2
1

Organic Process Research & Development
Page 22 of 35
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
.4.1 Hydrogenation of benzonitrile
-1
Figure 8 presents the infrared spectrum in the range 2175-2290 cm as a function of reaction
time for benzonitrile hydrogenation over the 1 wt% Pd/Al O catalyst at 353 K. At short
2
3
-1
reaction times, an intense band centered at 2240 cm is evident that is assigned to the CN
stretching mode of benzylamine 39. The band progressively diminishes in intensity for
increasing reaction times up to  25 minutes, when the feature is no longer present, indicating
benzonitrile consumption to be complete about that time. Figure 9 extends the spectral range
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-1
presented to 400-4000 cm but at a reaction time of 30 minutes, i.e. corresponding to reaction
completion.
-1
Figure 8. FTIR spectra (2175-2290 cm ; t = 5 – 30 min) for the gas phase hydrogenation of
−1
benzonitrile over 1% Pd/Al O at 353 K showing the CN stretch at 2238 cm .
2
3
ACS Paragon Plus Environment
2
2

Page 23 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-1
Figure 9. FTIR spectrum (400-4000 cm ) recorded at t = 30 minutes in the gas phase
hydrogenation of benzonitrile over Pd/Al O at 353 K.
2
3
Figure 9 shows full conversion of the nitrile moiety, as evidenced by absence of the CN stretch
at 2238 cm−1 39. Toluene was produced, confirmed by the presence of the diagnostic CH bend
−1
at 729 cm , alongside ammonia, demonstrated by its characteristic bands at 950, 1631 and
3444 cm−1 39. The production of these compounds from the benzonitrile hydrogenation
reaction confirms the occurrence of the hydrogenolysis reaction.
ACS Paragon Plus Environment
2
3

Organic Process Research & Development
Page 24 of 35
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 10. Reaction profile for the gas phase hydrogenation of ca. 0.09 mmoles benzonitrile over
5
00 mg 1wt% Pd/Al O at 353 K and 1 bar. The benzonitrile and toluene are presented as moles
2
3
(left y axis), whereas peak height is used to indicate ammonia concentration (right y axis).
Figure 10 shows the resultant reaction profile for the hydrogenation of benzonitrile and
provides further confirmation that toluene production was indeed coupled with the formation
of ammonia. Difficulties were encountered in the determination of an initial concentration for
benzonitrile, as a consequence of the hydrogenation rate being faster than the rate of
volatilisation at 353 K. Nevertheless, since the quantity of reagent added was known, a
calculated A value is included in the figure. The predicted intensity profile over the first 5
o
minutes of reaction is consistent with a first order decay of benzonitrile as a function of time
(Section 3.1.1). When the hydrogenolysis of benzylamine was considered, it was again
observed that ammonia production coincided with toluene formation. Thus, a hydrogenolysis
reaction can be unambiguously associated with the loss of selectivity of benzylamine in the
hydrogenation of benzonitrile.
3
.4.2 Deuteration of benzonitrile
ACS Paragon Plus Environment
2
4

Page 25 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-1
Figure 11. FTIR spectrum (400-4000 cm ), recorded at t = 30 minutes, for the gas phase
deuterium hydrogenation of benzonitrile over 1wt% Pd/Al O at 353 K.
2
3
From the final IR spectrum recorded at reaction completion for the deuteration of benzonitrile
Figure 11), it can be seen that both toluene-d and ND were formed. Of particular note is the
(
3
3
−1
shifting of the band assigned to the CD aliphatic stretch of toluene to around 2170 cm (cf. the
−1
corresponding CH stretch observed at 2912 cm in Figure 9). Moreover, absence of the CH
stretch indicates that toluene-d has been exclusively produced. Shifting of the associated
3
ammonia peaks to lower frequencies has also been observed, accounted for by the deuteration
of this species.
The proposed surface chemistry involved in the formation of fully deuterated toluene and
ammonia can be visualised in Scheme 2. With reference to Scheme 1, Scheme 2 shows that
deuterated benzylamine (C H CD ND ) is formed as an intermediate species via the
6
5
2
2
corresponding imine. Additionally observed was the preservation of aromaticity under the
designated reaction conditions, as evidenced by the maintenance of the CH aryl stretch at
ACS Paragon Plus Environment
2
5

Organic Process Research & Development
Page 26 of 35
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
−1
approximately 3080 cm . This is therefore indicative of H/D exchange occurring solely at the
unsaturated centers external to the ring.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Pd
Pd
Pd
D
D
Pd
Pd
Pd
D
D
D2
Pd/Al O3
2
ND
C
CN
C
N
D
D2
D2
Pd
Pd
Pd
D
Pd
Pd
Pd
D
D
Pd
Pd
Pd D
ND3
D
D
ND2
ND2
CD3
C
D
C
D
D
D
-ND3
CD3
Scheme 2. Proposed reaction scheme for the deuteration of benzonitrile over an alumina
supported Pd catalyst resulting in the formation of toluene-d3.
Scheme 2 involves a number of distinct transformations. The benzonitrile first binds through its
nitrile functionality, and is reduced stepwise, to benzylamine via benzylimine. The bound
benzylamine can then interact with the surface via its aromatic ring where hydrogenolysis can
occur 40–43. The imine unit is then reduced to produce a surface anchored -CD -ND substituent
2
2
group. Further deuterium addition leads to coincident toluene-d and ND formation. It is
3
3
expected that formation of these products leads to a concerted desorption process.
3
.4.3 Deuterium hydrogenolysis of benzylamine
ACS Paragon Plus Environment
2
6

Page 27 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-1
Figure 12. FTIR spectrum (400-4000 cm ), recorded at t = 30 minutes, for the gas phase
deuterium hydrogenolysis of benzylamine over Pd/Al O at 353 K. Of note is the presence of
2
3
both the aliphatic CH and CD stretches in the spectrum.
Investigation of the deuterolysis of benzylamine, Figure 12, shows that upon reaction
completion it is toluene-d (C H CH D) which is observed, and not, toluene-d (C H CD ), as was
1
6
5
2
3
6
5
3
detected for benzonitrile deuteration (Figure 11, Scheme 2). The production of a non-fully
deuterated form of toluene is evidenced, with reference to Figure 12, through the presence of
−1
−1
both aliphatic CH (2912 cm ) and CD (2141 cm ) stretches. A proposed reaction scheme for
the formation of toluene-d as a surface process in the deuterolysis reaction is presented in
1
Scheme 3.
ACS Paragon Plus Environment
2
7

Organic Process Research & Development
Page 28 of 35
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Pd
Pd
Pd
D
D
Pd
Pd
Pd
D
D
Pd/Al O
NH2
2
3
NH2
NH2
D2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
D
H
H
Pd
Pd
Pd
Pd
Pd
Pd
D
D
-
NH D
D
H
H
2
NH2
Scheme 3. Proposed reaction scheme for the direct deuterolysis of benzylamine resulting in
formation of toluene-d1.
For the adsorption of benzylamine, the amine unit is thought to bind via either the nitrogen
lone pair, or, the aromatic ring to the Pd surface. However, in order for hydrogenolysis to
occur, it must be the aromatic ring which is bound to the palladium 40–43. Subsequent C-N
cleavage leads to an allylpalladium species, which is further deuterated to yield toluene-d and
1
deuterated ammonia.
It was assumed that, alongside toluene-d NH D would be observed. However, only the fully
1
2
deuterated form of ammonia (ND ) was detected experimentally. The simultaneous formation
3
of partially deuterated toluene and fully deuterated ammonia provides an indication that these
processes are decoupled from each other. It is therefore suggested that ND must be produced
3
as the result of a hydrogen exchange process: NH + xD
 ND + H (with x indicating an
3
2
3
2
excess of D ). Exclusive formation of this species indicates that this reaction is fast, Pd
2
(g)
catalysed, and in this case, effectively makes the ammonia production unselective.
Theoretically it should be possible to use ammonia isotopomers, distinguishable by IR
4
4
spectroscopy, to probe the adsorption geometry linked with the benzylamine hydrogenation
process. For this particular application, however, this information was not accessible as a
consequence of the rapid nature of the H/D exchange under the designated reaction
conditions.
ACS Paragon Plus Environment
2
8

Page 29 of 35
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
The other product of the reaction, toluene-d , was identified by the presence of both CH and
1
CD aliphatic stretches. Nevertheless, care is required in the interpretation of these results. The
presence of both CH and CD stretches in the IR spectra infers only that toluene-d is not
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
exclusively formed. As a consequence, the presence of both toluene-d and toluene-d
2
3
alongside toluene-d cannot be excluded. As with the rapid H/D exchange observed in
1
ammonia formation, it is possible that the bound toluene-d may also undergo H/D exchange
1
resulting in formation of toluene-d and toluene-d . Due to detection of a toluene CH bond it
2
3
can be deduced that H/D exchange for this species is not fast. Thus, whilst deuterium
incorporation to form toluene-d and toluene-d cannot be ruled out, nevertheless, it is shown
2
3
to be a slow process when compared to the ammonia exchange.
As a consequence of the gas phase reaction conditions there is also the potential for an
additional method of deuterium enrichment. Scheme 1 shows that, as with all nitrile
hydrogenations, amine formation is preceded by an imine intermediate. Taking into account
the reversible nature of this process, Maschmeyer et al. proposed that this could provide a
route ‘back’ from the amine to the imine, a process that would allow further deuteration of the
1
3
toluene to be achieved . The implication is thus that for benzylamine the direct deuterolysis
to yield toluene-d is now in competition with the ‘back’ reactions which allow toluene-d and
1
2
toluene-d to be produced.
3
4
.0
Conclusions
The heterogeneously catalysed liquid phase hydrogenation of benzonitrile has been considered.
Analysis of the hydrogenation and hydrogenolysis pathways which resulted in formation of
toluene allowed the interconnectivity of the two reactions to be assessed. When separately
considered, benzonitrile hydrogenation and benzylamine hydrogenolysis were found to have
differing kinetics and energetics. The hydrogenation of benzonitrile followed first order
−1
reaction kinetics with a corresponding activation energy of 27.6 kJ mol . Contrastingly, the
−1
hydrogenolysis of benzylamine not only had a much higher activation energy (80.1 kJ mol ),
ACS Paragon Plus Environment
2
9