The production of tyramine: Via the selective hydrogenation of 4-hydroxybenzyl cyanide over a carbon-supported palladium catalyst

Article abstract of DOI:10.1039/c8ra05654d

The selective production of primary amines is a problem that plagues heterogeneously catalysed nitrile hydrogenation reactions. Whilst the target amine tyramine (HOC₆H₄CH₂CH₂NH₂) is biochemically available through the action of enzymes, synthetic routes to this species are not widely reported. Here, a heterogeneously catalysed method is proposed that utilises a Pd/C catalyst to effect the selective hydrogenation of 4-hydroxybenzyl cyanide within a three-phase reactor. The aforementioned selectivity issues are overcome by adjustment of various experimental parameters (hydrogen supply, agitation rate, temperature, use of an auxiliary agent) that result in improved catalytic performance, such that the desired tyramine salt (tyramine hydrogen sulphate) can be produced in quantitative yield. Accordingly, through consideration of the interconnectivity of hydrogenation and hydrogenolysis processes, a selective synthetic strategy is achieved with the findings suitable for extension to other substrates of this nature.

Full text of DOI:10.1039/c8ra05654d

RSC Advances
View Article Online
View Journal | View Issue
PAPER
The production of tyramine via the selective
hydrogenation of 4-hydroxybenzyl cyanide over
Cite this: RSC Adv., 2018, 8, 29392
a carbon-supported palladium catalyst†
Mairi I. McAllister,a Cedric Boulho,a Liam McMillan,a Lauren F. Gilpin,a
´
a
Sandra Wiedbrauk,a Colin Brennanb and David Lennon
*
The selective production of primary amines is a problem that plagues heterogeneously catalysed nitrile
hydrogenation reactions. Whilst the target amine tyramine (HOC6H4CH2CH2NH2) is biochemically
available through the action of enzymes, synthetic routes to this species are not widely reported. Here,
a heterogeneously catalysed method is proposed that utilises a Pd/C catalyst to effect the selective
hydrogenation of 4-hydroxybenzyl cyanide within a three-phase reactor. The aforementioned selectivity
issues are overcome by adjustment of various experimental parameters (hydrogen supply, agitation rate,
temperature, use of an auxiliary agent) that result in improved catalytic performance, such that the
desired tyramine salt (tyramine hydrogen sulphate) can be produced in quantitative yield. Accordingly,
through consideration of the interconnectivity of hydrogenation and hydrogenolysis processes,
a selective synthetic strategy is achieved with the findings suitable for extension to other substrates of
this nature.
Received 2nd July 2018
Accepted 8th August 2018
DOI: 10.1039/c8ra05654d
rsc.li/rsc-advances
with ammonia and the reductive amination of oxo compounds,3
as well as via the heterogeneously catalysed hydrogenation of
nitriles.4–7 It is the later option which is explored herewith. In
1. Introduction
Primary amines are chemically signicant reagents nding
application in many industries including pharmaceuticals,
plastics and agri-chemicals.1 Tyramine (HOC6H4CH2CH2NH2)
is such a species and is a molecule which presents considerable
interest due to its biological activity in the human body, where it
may be accessed via foods as diverse as cheese, red wine,
sausages and chocolate.2 In addition to involvement in a range
of biochemical pathways within mammalians, trace amines
such as tyramine and its analogues are additionally thought to
inuence aspects of brain chemistry. Biochemically, tyramine
may be produced by the action of the enzyme tyrosine hydrox-
ylase to form tyrosine (HOC6H5CH2CH(CO2H)(NH2)) that,
subsequently, undergoes decarboxylation via the action of the
enzyme tyrosine decarboxylase to form tyramine. The matter of
the pharmacology and therapeutics of trace amines has been
comprehensively reviewed by Broadley.2
the case of nitrile hydrogenations, conversion of the nitrile
functionality is a facile process,8 nonetheless, achieving high
selectivity towards the primary amine is signicantly more
challenging due to the presence of a highly reactive imine
intermediate. Amongst other side reactions, condensation
reactions between this species and the amine product, in
a process initially proposed by Braun et al.,9 can lead to the
formation of undesired side products such as secondary and
tertiary amines and, consequently, a reduced selectivity to the
desired product.5,10,11
Whilst previous investigations into the selective hydrogena-
tion of aromatic nitriles in the liquid phase have highlighted
the use of skeletal metal catalysts such as RANEY® Cobalt,12
application of supported Pd catalysts has increasingly featured
in recent years, not least because Pd is thought to minimise
reduction of substituent groups, other than the nitrile func-
tionality, when present. This attribute is particularly pertinent
The value and biochemical availability of tyramine has been
demonstrated, and thus, development of a chemical synthetic
strategy is desirable. There are a number of different ways in
which primary amines can be synthesised including alkylation
to ne chemical applications. Notable examples being the work
5,8
of Hegedus et al., Bakker et al.,13 and Maschmeyer et al.,14,15
}
with the hydrogenation of benzonitrile featuring prominently.
More recently, Dai and co-workers have examined the applica-
tion of supported Pd catalysts for the reduction of benzonitrile
to benzylamine; considering a role for the support material,
reactor conguration and catalyst deactivation in catalytic
performance.16–18 Keane and co-workers have investigated the
gas phase variant of this reaction.19 Illustrating imaginative
aSchool of Chemistry, University of Glasgow, Joseph Black Building, Glasgow, G12
8QQ, UK. E-mail: David.Lennon@glasgow.ac.uk; Tel: +44-141-330-4372
bSyngenta, Jeallot's Hill International Research Centre, Bracknell, Berkshire, RG42
6EY, UK
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c8ra05654d
29392 | RSC Adv., 2018, 8, 29392–29399
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
Scheme 1 Reaction scheme corresponding to the hydrogenation of 4-hydroxybenzyl cyanide (ca. 17 mmol including ca. 34 mmol H2SO4) in
MeOH (350 mL) over 0.50 g 5% Pd/C. The desired pathway is highlighted in black with unwanted routes, hydrogenolysis and coupling reactions,
represented in blue and red respectively.
exploitation of this technology, Saito and co-workers have relevance to certain agri-chemical production chains where
applied a supported Pd catalyst operating in a continuous ow selective nitrile hydrogenation is oen required in the presence
regime to employ a selective nitrile reduction stage to synthesize of other ring substituents. To a degree, the inclusion of the
the antidepressant drug venlafaxine.20
In 2016 McMillan and co-workers reported on the hydroge- requirement.
hydroxyl group para to the nitrile chain addresses this
nation of aromatic nitrile compounds over a carbon-supported
Chemically, following on from the early work of Buck using
Pd catalyst.21 Hydrogenation of benzonitrile to form benzyl- a platinum oxide catalyst,22 there are surprisingly few accounts
amine was readily achieved but this was subsequently followed for the synthesis of this trace amine.23–25 Thus, this work seeks
by a hydrogenolysis reaction to produce toluene. Co-adsorption to use a potential synthesis route to a biologically relevant
studies of benzonitrile and benzylamine provided evidence for primary amine to additionally explore the factors that inuence
site-selective chemistry over this particular Pd/C catalyst, with hydrogenation and hydrogenolysis steps over the same carbon-
specic sites associated with the hydrogenation and hydro- supported Pd catalyst used by McMillan et al.21 In this way, the
genolysis pathways.
Connected to the background outlined above, an important correlation between molecular structure and selectivity issues
parameter for reduction reactions incorporated within ne connected with the hydrogenation of substituted
chemical synthesis regimes is to be able to better understand phenethylamines.
factors that contribute to the presence (or not) of hydrogenation With reference to Scheme 1, the routes to the above-
selected reaction system will provide new information on the
and hydrogenolysis pathways. The reagent selected for the mentioned selectivity issues associated with nitrile hydrogena-
studies reported here is 4-hydroxybenzyl cyanide (HOC6H5- tions are outlined alongside the desired pathway. Notably, the
CH2CN), which was selected as a reagent for two reasons. potential hydrogenolysis, as observed for the benzonitrile
Firstly, selective hydrogenation of 4-hydroxybenzyl cyanide will system reported by McMillan,21 is shown. In this case, 4-
yield 4-(2-aminoethyl)phenol (HOC6H5CH2CH2NH2), also hydroxyethylbenzene is expected upon loss of the amine func-
known as tyramine and previously identied as a biologically tionality. In addition, highlighted in red and commonly
valuable species. Secondly, this molecule is also found to be of observed in nitrile reductions, is the coupling reaction of the
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 29392–29399 | 29393

View Article Online
RSC Advances
Paper
highly reactive imine intermediate, 4-hydroxybenzyl imine, with
The reactor was charged with catalyst and solvent. The
the amine products.5,8 Di-hydroxyphenylamine is formed by the mixture was then heated and reduced under hydrogen for 1
coupling of this imine with tyramine in a reaction that is hour prior to substrate addition via a syringe. The reaction
accompanied by the elimination of ammonia. It is then possible mixture was sampled at regular intervals and analysed by HPLC
for the 4-hydroxybenzyl imine to couple with this secondary (HP series 1100 HPLC tted with a BDS Hypersil C18 column).
amine in a similar process to yield the tertiary amine tri- In the resultant reaction proles t ¼ 0 is considered as the point
hydroxyphenylamine. The production of secondary and where there is no reagent consumption or product formation.
tertiary amines is a signicant compromise to primary amine Experiments connected with high product yields were addi-
1
selectivity and highlights the essential nature of an adequate tionally analysed by H NMR spectroscopy using a Bruker AVI
hydrogen supply which is shown to be essential in the 400 MHz spectrometer. Inclusion of a known concentration of
production of tyramine (see Section 3.2).
ethylene glycol in the NMR tube was used as an internal stan-
Care was taken to explore the various experimental param- dard, which allowed quantication of both reagent and
eters that could potentially impede product formation. Conse- product. Deuterium oxide, selected for its effective peak reso-
quently, invoking an experimental protocol that reduces the lution in this system, was utilised as the deuterated solvent for
production of secondary and tertiary amines is essential in NMR analysis.
order to enhance selectivity towards the primary amine. Several
authors report the use of auxillary agents to enhance primary
amine yields. Amongst other methods, both acids (e.g. HCl),26–28
and bases (e.g. NH3),29 have been used for such a purpose.
On the assumption that protonation of the amine will reduce
its reactivity as a nucleophile, and thus, prevent any coupling
reactions, inclusion of an acid additive was implemented.8 In this
instance, sulphuric acid was selected; an excess of a diprotic acid
should ensure protonation of the amine functionality throughout
the full reaction coordinate and, moreover, the anion (HSO4ꢀ) is
not expected to cause metal complexation problems as might be
the case if, for example, HCl was used. Furthermore, the
protonation of the amine to form a salt should keep this species
in the liquid phase, and in doing so, prevent its coordination to
the active sites of the catalyst which could be of detriment to the
conversion of the nitrile/imine.8 The strong chemisorption of the
nitrogen lone pair to the Pd surface is therefore thwarted by the
acid's presence and, as such, poisoning of the catalyst should be
prevented. In order to achieve the selective production of the
primary amine it is essential that the previously discussed
hydrogenolysis and coupling reactions are eliminated from the
system. These matters are the focus of this communication.
3. Results and discussion
3.1 Hydrogenation versus hydrogenolysis
Considering rst the hydrogenolysis pathway, found to be
accessible over this catalyst for the benzonitrile system, but also
a phenomenon observed by other authors.5,8,13,14,21 Indeed, for
benzonitrile hydrogenation over the same Pd/C catalyst, used in
this instance for 4-hydroxybenzyl cyanide, hydrogenolysis was
a dominant pathway to such a degree that, ultimately, toluene
was produced with complete selectivity.21 Despite the incidence
of hydrogenolytic cleavage of the amine group described in the
case of benzonitrile, no 4-hydroxethylbenzene was detected in
the liquid phase for the hydrogenation of 4-hydroxybenzyl
cyanide (see Section 3.2).
Given that this is exactly the same Pd/C catalyst and experi-
mental apparatus as used by McMillan et al.,21 the inability of
this catalyst to support the proposed hydrogenolysis stage may
initially be considered a surprising result. For the benzonitrile
system the documented reactivity was interpreted within a 3-
site model for the catalyst. Namely, one site was assigned to the
hydrogenation of the nitrile/imine, whilst another facilitated
the hydrogenolysis, a nal site was then allocated for the
dissociative adsorption of dihydrogen (which may also encom-
pass the other sites). This model was supported by kinetic and
co-adsorption studies. Thus, given that a hydrogenolysis func-
tion is known to occur on this catalyst, why is the hydro-
genolysis stage inaccessible under any of the presented reaction
conditions?
2. Experimental
Reactions were carried out in the liquid phase in a continuously
¨
stirred batch slurry reactor (500 mL, Buchi bmd) using meth-
anol (Sigma Aldrich, $99.7%) as a solvent. An automated gas
ow controller (BPC 1202) allowed the delivery of inert (N2,
BOC, 99.999% purity) and active (H2, BOC, $99.995% purity)
gases to be delivered directly to the reactor via a gas reservoir.
Reactions were heated by silicon oil passed around the reactor
via a heating circulator (Julabo F25). 4-Hydroxybenzyl cyanide
(TCI, purity > 99%) was hydrogenated over a 5 wt% Pd/C catalyst
(Sigma Aldrich, code number 205680) in the presence of a sul-
phuric acid additive (Sigma Aldrich, 99.999% purity). Charac-
terisation of this catalyst, which was also used in the previously
mentioned benzonitrile studies, has been undertaken else-
where.21 Briey, CO chemisorption determined the number of
surface metal sites to be 1.11 ꢁ 1020 Pd(s) atoms per g(cat), cor-
responding to a Pd dispersion of 39%. Transmission electron
microscopy revealed a mean Pd particle size of 2.5 nm.21
Fig. 1 presents the molecular structure of tyramine hydrogen
sulfate and identies the aliphatic carbon atoms as a and
Fig. 1 The molecular structure of the tyramine hydrogen sulfate salt.
The a and b methylene units of the substituent arm are referenced
relative to the amine functionality.
29394 | RSC Adv., 2018, 8, 29392–29399
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
b positions relative to the amine functionality. It is suggested
that it is the positioning of the a-carbon relative to the aromatic
ring which dictates whether a hydrogenolysis pathway is sup-
ported over this catalyst. Indeed, Maschmeyer et al. similarly
propose that it is the presence of an aromatic group adjacent to
the a-carbon that facilitates hydrogenolysis.14 It is thus advo-
cated that the ammonium leaving group needs to be connected
to the aromatic ring via a single methylene group in order to
facilitate the hydrogenolysis step. In the case of the tyramine
hydrogen sulfate salt, the additional methylene linker in the
substituent arm does not meet this molecular requirement,
endorsing the absence of 4-hydroxyethylbenzene in the reaction
mixture. This scenario implies that the site-selective
hydrogenation/hydrogenolysis chemistry invoked by McMillan
et al.21 is not a dominant process. Further, computational
calculations conducted by Kieboom et al. on the hydrogenolysis
of benzyl alcohol derivatives over Pd/C suggest that the aromatic
ring is p-bonded to the catalyst surface. Here it is postulated
that for hydrogenolysis to occur it is essential that the transition
state must facilitate overlap between the electron decient p-
orbital of the a-carbon and the p-orbitals of the benzene ring.
The necessity for this orbital overlap rationalises, on account of
geometry, why an additional methylene linker, when compared
to benzonitrile, between the aromatic group and the hetero-
atom is not conducive to a hydrogenolytic process.30
Fig. 2 Influence of agitation rate on the hydrogenation of 4-hydrox-
ybenzyl cyanide (ca. 17 mmol) over 0.50 g of a 5% Pd/C catalyst in the
presence of 2 molar equivalents of acid. Reaction carried out at 60 ꢂC
and at 5 bar g pressure.
supply is pivotal to tyramine formation, it is crucial that stirring
is sufficiently fast as to avoid operation within a domain of
constrained hydrogen supply to the Pd surface.
Fig. 3 presents the reaction prole for an acidic mixture
operating at a stirring speed of 1050 rpm. Complete conversion
of the starting material was achieved in approximately 5
minutes, signicantly faster that the reaction in the absence of
acid (cf. 93% conversion over 6 hours, not shown). The imine
intermediate was rapidly consumed affording the primary
amine tyramine with 92% selectivity. A complete mass balance
was achieved by reaction completion with the remainder made
up by the secondary amine formed via the aforementioned
coupling reaction. It should be noted that Fig. 3 shows
a concentration prole for the secondary amine. This species
was not commercially available meaning that calibration by
HPLC, as was employed for 4-hydroxybenzyl cyanide and
3.2 Towards the selective production of tyramine
Section 3.1 states that a hydrogenolytic pathway is not opera-
tional for this substrate and, consequently, can be removed
from consideration. Unfortunately, coupling reactions were
found to complicate the reaction prole with an acid additive
found to be essential for production of the primary amine.
Inclusion of acid allowed complete removal of a tertiary amine
component indicating that the additive was effective in
hindering coupling reactions. Nonetheless, the route to the
secondary amine remained accessible, restricting complete
selectivity to the primary amine. The identication of the
secondary and tertiary amines was achieved by LC-MS analysis
and can be found in the ESI, S1.† Hydrogen availability in liquid
phase nitrile hydrogenation reactions is a further critical
experimental parameter that can provide a means for altering
the selectivity of a reaction.6 Therefore, in order to evaluate the
possibility of mass transport restrictions in the reaction of
interest, the effect of agitation rate on the rate of consumption
of the starting material was considered,31 Fig. 2.
Fig. 2 shows the expected ‘plateau’ dependence of reagent
consumption as a function of agitation rate at high mixing
speeds. Over the range 650–1050 rpm, increasing stirring
speeds leads to no real increase in the 4-hydroxybenzyl cyanide
hydrogenation rate, indicative of the reaction being under
kinetic control and free of mass transport restrictions. However,
below ca. 650 rpm the reaction is exhibiting characteristics of
mass transport limitations (reaction rate f stirring speed) and
is operating under diffusion control; presumably as a result of
inefficient mass transport at the gas/liquid interface. Given this
Fig. 3 Reaction profile for 4-hydroxybenzyl cyanide hydrogenation
under acidic conditions at an agitation rate of 1050 rpm [5 bar g, 60 ꢂC,
0.50 g 5% Pd/C, ca. 17 mmol of nitrile, ca. 34 mmol H2SO4, 350 mL
knowledge, and the inference from Scheme 1 that hydrogen MeOH].
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 29392–29399 | 29395

View Article Online
RSC Advances
Paper
1
tyramine, could not be achieved. Therefore, H NMR spectros- appears to be apparent. Namely, increasing the operational
copy was used as a means of calculating a response factor based pressure from 3 to 5 bar g increases tyramine hydrogen sulphate
on the relative integration of peak areas allowing a concentra- selectivity on completion of reaction from ꢃ81% to ꢃ94%. A
tion to be obtained. As a consequence of the highly reactive further increase in pressure to 7 bar g, however, was found to be
nature of imines, however, it was not possible to determine the approximately equal to the 5 bar g reaction, indicating that no
response factor for the 4-hydroxybenzyl imine intermediate. As further improvements could be made to selectivity via pressure
such, imine peak area was plotted seperately on the second y- within the limitations of the experimental set-up.
axis. This is also the case for all future gures involving the
Nevertheless, the necessity to improve selectivity towards the
imine species. The unsuccessful quantication of the imine primary amine remained. Thus, in order to further improve
results in an apparent missing mass at the beginning of the selectivity to the primary amine, the role of temperature was
reaction. Nonetheless, a complete mass balance is returned by explored. Intriguingly, the literature provides conicting opin-
reaction completion, indicating the absence of undesirable side ions on the effect of temperature on nitrile hydrogenation
reactions (see ESI, S2†).
reactions. Augustine, for example, suggests that high tempera-
Despite the high selectivity towards tyramine achieved in tures favour formation of primary amines as a consequence of
Fig. 3, tuning of the reaction conditions and parameters was desorption of the imine intermediate from the surface being
essential for this outcome to be realised. A brief summary of the minimised under these conditions.32 In contrast, Cerveny
optimisation process and the subsequent effect on selectivity is indicates that it is secondary amine formation that is favoured
presented in Table 1. The most notable improvement was by elevated temperatures.33 Clearly, temperature also links to
detected upon inclusion of an acid additive. The addition of this the hydrogen availability issues central to tyramine formation.
species to the reaction mixture effectively removed the tertiary Arrhenius theory tells us that increasing temperature can
amine from the system and drastically enhanced tyramine increase reaction rates but, given the breadth of pathways
production. Further, at low stirring speeds, found to be outside accessible to this reaction system presented in Scheme 1, it is
the kinetic regime for this system (Fig. 2), the hydrogen supply desirable to disfavour the rate for the detrimental coupling
was found to be limited resulting in an unoptimised selectivity. reactions. Further interwoven into the complexity of the reac-
This can be evidenced by the marked improvement to tyramine tion matrix is the matter of hydrogen solubility in the process
production (82% / 92%) upon elevation of agitation rate. solvent: hydrogen solubility in methanol is enhanced at
Thus, the combination of employing an acid as an auxiliary elevated temperatures.31 It is therefore necessary to identify an
agent plus operation within a kinetic regime effectively free optimal temperature that will facilitate effective solubility of
from gas/liquid mass transport limitations has led to signicant hydrogen in methanol, whilst additionally maintaining the
improvements in the returnable yield of the primary amine salt selectivity of the reaction.
(tyramine hydrogen sulfate). These alterations illustrate the
intricate nature of the process involved in achieving the with Fig. 4 showing the reaction prole for 4-hydroxybenzyl
completely selective production of a primary amine.
cyanide hydrogenation under acidic conditions at 30 ꢂC. The
Temperature alterations on the system were implemented,
It should be noted that the reaction parameters presented decrease in temperature had a signicant effect on the product
here do not represent an exhaustive list of factors inuential distribution. As before, the nitrile unit is rapidly consumed with
over hydrogen supply. Additionally considered were the effects complete conversion to form the tyramine salt via the 4-
of hydrogen pressure in the headspace of the autoclave. hydroxybenzyl imine intermediate. However, no secondary or
Increasing hydrogen pressure resulted in faster hydrogenation tertiary amine products are identied chromatographically in
rates; notably complete conversion of 4-hydroxybenzyl cyanide
was attained at 10, 5 and 2.5 minutes at respective over-
pressures of 3, 5, and 7 bar g. Outwith the error bars associated
with the analytical measurement, a sensitivity of selectivity
Table 1 Selectivity changes observed as a result of various parameter
alterations. All values are calculated at 60 minutes into the reaction. In
all instances the following were constant: [5 bar g, 0.50 g 5% Pd/C, ca.
17 mmol of nitrile, ca. 34 mmol H2SO4, 350 mL MeOH]. For all reac-
tions involving acid it is the salt of the amine that is detected. *Only
93% conversion achieved over 6 hours
Conditions
Selectivity
Acid
S/rpm
T/ꢂC
1ꢂ amine/%
2ꢂ amine/%
3ꢂ amine/%
*No
Yes
Yes
Yes
450
450
1050
1050
60
60
60
30
14
82
92
44
18
8
30
0
0
Fig. 4 Reaction profile for 4-hydroxybenzyl cyanide hydrogenation
under acidic conditions carried out at a temperature of 30 ꢂC [5 bar g,
1050 rpm, 0.50 g 5% Pd/C, ca. 17 mmol of nitrile, ca. 34 mmol H2SO4,
350 mL MeOH].
100
0
0
29396 | RSC Adv., 2018, 8, 29392–29399
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
this case, indicating that the primary amine has been selectivity
produced. Thus, Fig. 4 represents a totally selective conversion
of 4-hydroxybenzyl cyanide in an acidic medium to form tyra-
mine hydrogen sulphate in the liquid phase.
With reference to Scheme 1, the selectivity of the reaction (S)
can be expressed in terms of the kinetics of primary amine
formation versus the coupling reactions to form the secondary
and tertiary reactions as S ¼ k2 [H2]/k3[primary amine].
Lowering the reaction temperature will affect both of the rate
constants presented here and the improved selectivity evident
in Fig. 4 is reective of the difference in the activation energies
of these competing reactions. A lower activation energy for the
hydrogenation reaction to the primary amine than for the
coupling reactions could provide reasoning for the improved
selectivity at lower temperatures. Another explanation, however,
may be formulated. Namely, that if the hydrogen supply in the
system has been optimised to avoid gas/liquid mass transport
limitations, thus allowing a xed concentration of hydrogen in
solution, it is also possible that a liquid/surface mass transport
limitation could occur. In this particular case the hydrogen
availability for the high temperature reaction is lower and is
detrimental for selectivity. Further work is required for these
two hypotheses to be differentiated.
The reaction prole exhibits the form of a consecutive
reaction. 4-Hydroxybenzyl cyanide conversion is rapid, being
complete within a reaction time of ꢃ8 minutes; the complete
conversion of 4-hydroxybenzyl cyanide corresponds to a turn-
over number of 184. Hydrogenation of the hydroxybenzyl imine
to produce tyramine is clearly a much slower process. The
product is formed on commencement of reaction and
progressively increases thereaer; the reaction is complete by
ꢃ40 minutes. This sequence indicates the imine hydrogenation
step to be rate limiting under the designated reaction
conditions.
Fig. 5 presents the mass balance plot for the reaction prole
presented in Fig. 4. Upon commencement of reaction there is
a clear mass imbalance over the period 0–20 minutes. With
reference to Fig. 4, this corresponds to the period in which
chromatography detects the presence of the 4-hydroxybenzyl
iminium hydrogen sulphate in the liquid phase. From approx-
imately 12 minutes the mass balance loss progressively reduces,
so that by approximately 40 minutes a closed mass balance is
obtained. Collectively Fig. 4 and 5 indicate that during the
earlier stages of the reaction coordinate the 4-hydroxybenzyl
imine is responsible for the mass imbalance.
An additional point worthwhile noting relates to the possi-
bility of homogeneous reactions occurring within this reaction
system. The reaction prole depicted in Fig. 4 shows a total
absence of secondary and tertiary amine by-products aer 60
minutes reaction time. However, when sample vials containing
the colourless analyte were le for an extended period of time
(overnight) under ambient conditions, certain samples exhibi-
ted a distinct colour change. Specically, those samples that
contain both the 4-hydroxybenzyl iminium hydrogen sulfate
and the tyramine hydrogen sulfate (i.e. samples corresponding
to the reaction period 5–15 minutes) developed a brown col-
ouration whilst, in marked contrast, samples connected with
Fig. 5 Mass balance plot for the reaction profile presented in Fig. 4 (4-
hydroxybenzylcyanide hydrogenation, 30 ꢂC, 5 bar g, 1050 rpm, 0.50 g
5% Pd/C, ca. 17 mmol of nitrile, ca. 34 mmol H2SO4, 350 mL MeOH).
The solid black line defines the experimental mass balance (derived
from quantification of hydroxybenzyl cyanide and tyramine, Fig. 4),
whereas the dotted green line represents the theoretical mass balance.
reaction times where the imine was absent (i.e. 20–60 minutes)
remained colourless. The colouration is assumed to reect
homogeneous chemistry that results in the formation of higher
molecular weight conjugated molecules from
a reaction
between solvated 4-hydroxybenzyl iminium hydrogen sulfate
and the tyramine hydrogen sulfate entities. This action high-
lights the reactive nature of the imine species. Nonetheless,
given the relatively short reaction time of the heterogeneous
chemistry illustrated in Fig. 4 (#60 minutes) compared to the
long period of the homogeneous process ($12 hours), the
observed homogeneous chemistry is not a major concern for the
heterogeneously catalysed selective production of tyramine
under consideration here.
The mass imbalance detected by HPLC, despite being tran-
sient and its presence rationalised, is less than ideal. Never-
theless, whilst the primary analytical technique used to examine
this molecular system was liquid chromatography, proton NMR
spectroscopy was also found to be a suitable probe to monitor
the progress of the reaction. Fig. 6 therefore shows a represen-
tative 1H NMR spectrum at reaction completion (t ¼ 60
minutes) corresponding to the optimised reaction conditions
utilised to afford Fig. 4. Spectra taken throughout the reaction
coordinate along with complete spectral assignment and the
corresponding reaction prole can be found in the ESI (S3†).
Analysis of the spectra presented in Fig. 6 reveals that, with
the exception of features connected with the NMR solvent (D2O)
and the internal standard (ethylene glycol), only peaks associ-
ated with tyramine hydrogen sulphate are present.34 No other
features were detected. Thus, at reaction completion Fig. 6
endorses the outcome of Fig. 4; namely, the selective formation
of the primary aromatic amine hydrogen sulfate salt from the
catalytic hydrogenation of hydroxybenzyl cyanide.
Whilst the completely selective production of tyramine
represents a satisfactory outcome, the batch conditions utilised
do not sufficiently provide an evaluation of catalyst recyclability
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 29392–29399 | 29397

View Article Online
RSC Advances
Paper
Fig. 6 Resultant 1H NMR spectrum at reaction completion (t ¼ 60 minutes) for the hydrogenation of 4-hydroxybenzyl cyanide to tyramine
hydrogen sulfate at 5 bar g pressure, 30 ꢂC and an agitation speed of 1050 rpm. The D2O peak of the NMR solvent is present at d ꢃ 4.9 ppm, whilst
the ethylene glycol reference peak is present at d ¼ 3.65 ppm.
and durability towards poisoning. This issue is acknowledged to
be one of specic relevance to this and related substrates and,
References
consequently, catalyst deactivation investigations represent
work in progress.
1 R. Kadyrov and T. H. Riermeier, Angew. Chem., Int. Ed., 2003,
42, 5472–5474.
2 K. J. Broadley, Pharmacol. Ther., 2010, 125, 363–375.
3 R. A. Sheldon and H. van Bekkum, Fine Chemicals Through
Heterogeneous Catalysis, Wiley-VCH, Weinheim, 2001.
4 M. Bowker, The Basis and Application of Heterogeneous
Catalysis, Oxford Chemistry Primers, OUP, Oxford, 1998.
4. Conclusions
A Pd/C catalyst has been used to effect the hydrogenation of 4-
hydroxybenzyl cyanide within a three-phase reactor to yield the
primary amine tyramine. Alongside the desired pathway, addi-
tional routes, resulting in unwanted by-products, were identi-
ed. The absence of a hydrogenolytic cleavage rationalised by
the unfavourable molecular framework of this species for
hydrogenolysis preserved the valuable amine functionality. It
was additionally found that coupling products, impairing the
selectivity, could be effectively eliminated by tuning various
experimental parameters leading to improved catalytic perfor-
mance, such that the desired tyramine salt could be produced
in quantitative yield.
´
}
5 L. Hegedus and T. Mathe, Appl. Catal., A, 2005, 296, 209–215.
6 H.-U. Blaser, Catal. Today, 2000, 60, 161–165.
7 C. de Bellefon and P. Fouilloux, Catal. Rev.: Sci. Eng., 1994,
36(3), 459–506.
´
´ ´
}
8 L. Hegedus, T. Mathe and T. Karpati, Appl. Catal., A, 2008,
349, 40–45.
9 J. Von Braun, G. Blessing and F. Zobel, Ber. Dtsch. Chem. Ges.
B, 1923, 56, 1988–2001.
10 P. Kukula, M. Studer and H.-U. Blaser, Adv. Synth. Catal.,
2004, 346, 1487–1493.
11 H. Chen, M. Xie and J. Shen, Catal. Lett., 2010, 135, 246–255.
¨
¨
12 P. Scharringer, T. E. Muller, A. Jentys and J. A. Lercher, J.
Catal., 2009, 263(1), 34–41.
Conflicts of interest
13 J. J. W. Bakker, A. G. van der Neut, M. T. Kreutzer,
J. A. Moulijn and F. Kapteijn, J. Catal., 2010, 274, 176–191.
14 A. J. Yap, B. Chan, A. K. L. Yuen, A. J. Ward, A. F. Masters and
T. Maschmeyer, ChemCatChem, 2011, 3, 1496–1501.
15 A. J. Yap, A. F. Masters and T. Maschmeyer, ChemCatChem,
2012, 4, 4839–4849.
There are no conicts of interest.
Acknowledgements
16 C. Dai, Y. Li, C. Ning, W. Zhang, X. Wang and C. Zhang, Appl.
Catal., A, 2017, 545, 97–103.
17 C. Dai, F. Liu, W. Zhang, Y. Li, C. Ning, X. Wang and
C. Zhang, Appl. Catal., A, 2017, 538, 199–206.
Syngenta and WestChem are thanked for provision of student-
ships (LM, LFG and MMcA) and project support. The EPSRC are
thanked for support via awards from a Knowledge Transfer
Account (EP/H500138/1) and an Impact Acceleration Account
(EP/K503903/1).
29398 | RSC Adv., 2018, 8, 29392–29399
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
18 C. Dai, S. Zhu, X. Wang, C. Zhang, W. Zhang, Y. Li and 27 T. S. Work, J. Chem. Soc., 1942, 79, 426–429.
C. Ning, New J. Chem., 2017, 41, 3758–3765.
19 Y. Hao, M. Li, F. Cardenas-Lizana and M. A. Keane, Catal. 29 H. Greeneld, Ind. Eng. Chem. Prod. Res. Dev., 1967, 6, 142–
Lett., 2016, 146(1), 109–116. 144.
20 Y. Saito, H. Ishitani, M. Ueno and S. Kobayashi, 30 A. P. G. Kieboom, J. F. de Kreuk and H. van Bekkum, J. Catal.,
ChemistryOpen, 2017, 6, 211–215. 1971, 20, 58–66.
21 L. McMillan, L. F. Gilpin, J. Baker, C. Brennan, A. Hall, 31 P. G. T. Fogg and W. Gerrard, Solubility of Gases in Liquids,
28 M. Freifelder and Y. H. Ng, J. Pharm. Sci., 1965, 54, 1204.
D. T. Lundie and D. Lennon, J. Mol. Catal. A: Chem., 2016,
411, 239–246.
22 J. S. Buck, J. Am. Chem. Soc., 1933, 55(8), 3388–3390.
John Wiley and Son, New York, 1991.
32 R. L. Augustine, Heterogeneous Catalysis for the Synthetic
Chemist, Marcel Dekker, Inc., New York, 1996.
23 M. Kohno and S. S. S.-I. Murahashi, Bull. Chem. Soc. Jpn., 33 L. Cerveny, Catalytic Hydrogenation, Elsevier, New York,
1990, 63, 1252–1254.
1986.
24 E. Ouellet and D. Poirier, Synlett, 2011, 14, 2025–2028.
34 G. R. Fulmer, A. J. M. Miller, N. H. Sherden, H. E. Gottlieb,
A. Nudelman, B. M. Stoltz, J. E. Bercaw and K. I. Goldberg,
Organometallics, 2010, 29, 2176–2179.
¨
25 M. Vilches-Herrera, S. Werkmeister, K. Junge, A. Borner and
M. Beller, Catal. Sci. Technol., 2014, 4, 629–632.
26 W. H. Hartung, J. Am. Chem. Soc., 1928, 50, 3370–3374.
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 29392–29399 | 29399