The reductive deaminative conversion of nitriles to alcohols using: Para -formaldehyde in aqueous solution

Article abstract of DOI:10.1039/c9cy01484e

We report herein, for the first time, the application of para-formaldehyde (pFA) to the reductive deamination of both aliphatic and aromatic nitriles in aqueous solution under transfer hydrogenation conditions. A broad range of primary alcohols have been synthesized selectively with very good to excellent yields under the optimized conditions. The study disclosed that the air-stable, inexpensive and commercially available catalyst [Ru(p-cymene)Cl₂]₂ acts as the catalyst precursor in this reaction, converting to other more active catalytic species in the presence of pFA, resulting in its degradation to CO₂ and H₂. Nitriles are also showed to play a dual role in this transformation, both as a substrate and as a ligand, where the dimeric catalyst structures convert to monomeric ones upon the coordination of nitrile molecules.

Full text of DOI:10.1039/c9cy01484e

Catalysis
Science &
Technology
View Article Online
View Journal | View Issue
PAPER
The reductive deaminative conversion of nitriles
to alcohols using para-formaldehyde in aqueous
solution†
Cite this: Catal. Sci. Technol., 2019,
9, 6092
ab
Ghazal Tavakolia and Martin H. G. Prechtl
*
We report herein, for the first time, the application of para-formaldehyde (pFA) to the reductive
deamination of both aliphatic and aromatic nitriles in aqueous solution under transfer hydrogenation
conditions. A broad range of primary alcohols have been synthesized selectively with very good to
excellent yields under the optimized conditions. The study disclosed that the air-stable, inexpensive and
commercially available catalyst [RuIJp-cymene)Cl2]2 acts as the catalyst precursor in this reaction,
converting to other more active catalytic species in the presence of pFA, resulting in its degradation to
CO2 and H2. Nitriles are also showed to play a dual role in this transformation, both as a substrate and as a
ligand, where the dimeric catalyst structures convert to monomeric ones upon the coordination of nitrile
molecules.
Received 26th July 2019,
Accepted 16th September 2019
DOI: 10.1039/c9cy01484e
rsc.li/catalysis
frequently used catalysts in this transformation. Also the use of
abundant metals such as Fe,6–8 Co,9 and Mn (ref. 10) PNP and
Introduction
Hydrogenation is among the most important chemical
reactions, at all stages, from laboratory to industry, in order for
the synthesis of high value products from unsaturated organic
molecules. Traditional methods including the use of main
group metal hydrides such as LiAlH4 and NaBH4 have received
less attention recently as their application is accompanied by
the use of harsh reaction conditions, low yields and poor
selectivity.1,2 The use of pressurized hydrogen is also not
favoured due to the safety problems and environmental
concerns as well as the expense and inconvenience of working
with gaseous H2, especially from the industrial point of view.
Transfer hydrogenation (TH), on the other hand, has emerged
as a powerful alternative with lots of advantages including
milder reaction conditions, easier operational setup, more
selectivity and reduced flammability compared to direct
hydrogenation using pressurized hydrogen gas.3–5 With the aid
of TH, various easily available and safe to handle hydrogen
donating molecules can be used as the hydrogen source
instead of molecular dihydrogen in the presence of a transition
metal-based catalyst to activate both substrate and the
hydrogen source in this reaction. Organometallic complexes of
ruthenium as well as Ru nanoparticles are among the most
NNP pincers and phosphino complexes11 has been reported
more recently. However, most of these catalytic systems are still
rather expensive, owing to the ligand backbone which
surpasses the costs of the central metal. Moreover, they are air
and moisture sensitive and often not commercially available
and therefore it requires a synthesis under inert conditions. All
these factors make the whole system more complicated and
less sustainable. Therefore, considering the increasing interest
in developing more sustainable synthetic methods, finding a
catalytic system which does not suffer from these drawbacks
and can work under milder reaction conditions means that TH
reactions are very beneficial and of great importance.
As previously mentioned, different organic and inorganic
molecules have been employed as the hydrogen resource in
TH and so far these have included alcohols,12–16 boranes,17–19
Hantzsch ester,20 and formates,21,22 among others.23,24
Recently, this group has reported the selective and mild
generation of H2 from formaldehyde in aqueous solution
which can also be generated in situ from para-formaldehyde
(pFA) or methanol in the presence of a [RuIJp-cymene)Cl2]2
dimer; a reaction which produces CO2 as the only by-product
(Scheme 1).16,25–28 It has also been shown that methanol can
be used as a H2 source for TH in a chemoenzymatic
a Department of Chemistry, University of Cologne, Greinstr. 6, 50939 Köln,
Germany. E-mail: martin.prechtl@uni-koeln.de
b Department of Science and Environment, Roskilde University, 4000 Roskilde,
Denmark. E-mail: prechtl@ruc.dk
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c9cy01484e
Scheme 1 The sequence of the catalysed H2 generation steps of the
pFA/water system.28
6092 | Catal. Sci. Technol., 2019, 9, 6092–6101
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
conversion for the reduction of alkenes which cannot be
reduced by enzymes.16
However, they needed to use TPPTS to stabilize and solubilize
their catalyst which is an expensive ligand and makes the
protocol less attractive. In addition, the application of 50 bar
H2 and 120 °C in this work can be considered as harsh
reaction conditions. Most recently, Hashmi and co-workers
also showed successful conversion of nitriles to primary
alcohols in the presence of RuHClIJCO)IJPPh3)3 as catalyst
(Scheme 2b).34
Again, despite obtaining good selectively, the reaction
conditions are still quite harsh (10 bar H2 in a 1 : 1 mixture of
1,4-dioxane : H2O at 140 °C for 18 h). Also, these reactions
need to be done under an inert atmosphere with use of a
glovebox or Schlenk line to handle the chemicals under an
argon atmosphere which increases the complexity of the
reaction and decreases the chance for use in industrial
applications. Therefore, selective conversion of aliphatic and
aromatic nitriles to primary alcohols under mild, feasible
and safe conditions, using an air-stable, inexpensive and
commercially available catalyst is still the missing piece in
the reductive deamination puzzle. Notably, the reverse
reaction, the conversion of alcohols to nitriles is also possible
with Ni-catalysts and ammonia.38,39
Formaldehyde is an inexpensive and readily available C1
building block which is produced annually in the range of 30
million tons,25 because many industries use formaldehyde
for the production of daily commodities.26 It also has great
potential for use as a liquid organic hydrogen carrier (LOHC)
molecule for a future hydrogen economy based on C1-
molecules.25 Although the hydrogen content of aqueous
formaldehyde is 8.4 wt% because the equilibrium favours
methanediol, which is lower than in methanol (12.5 wt%),
and from the thermodynamic point of view, H2 production
from aqueous formaldehyde is more attractive because
methanol reforming is endergonic and formaldehyde
reforming is exergonic.25,29,30
In continuation of previous research where it was
demonstrated that on the one hand aqueous formaldehyde
formed in situ can be used for TH of alkenes,16 and on the
other hand nitriles had been converted selectively to primary
amines and secondary imines,31 now the challenge to convert
selectively, under TH conditions, nitriles into alcohols must
be tackled. Examples for this deaminative transformation of
nitriles are rare and usually the reduction has a poor
selectivity giving a mixture of, for example, amine, imine,
aldehyde and alcohol or the reaction stops at the aldehyde
stage, or it requires harsh reaction conditions (Scheme 2)
and sensitive/expensive catalysts.32–37
Herein, considering all the previously mentioned
limitations and to avoid using pressurized hydrogen, selective
reductive deamination of both aliphatic and aromatic nitriles
under mild conditions using pFA C1 source of TH in an
aqueous system as shown in Scheme 2c, is reported first. The
commercially available, air-stable and inexpensive RuIJp-
cymene)Cl2 dimer has been proved to be an active and
appropriate catalyst for this transformation.
In 2004, Li and co-workers reported the first selective
conversion of nitriles to alcohols using a Ru-TPPTS catalyst
(TPPTS = [PIJm-C6H4SO3Na)3]) prepared in situ (Scheme 2a).35
Results and discussion
Optimization of reaction conditions
As reduction of nitriles is normally a non-selective reaction
which results in the formation of different by-products, from
benzamide to imine and amines, it was necessary to find the
best conditions which lead to the selective formation of the
desired product. To this end, benzonitrile was selected as the
benchmark substrate. As is already known from previous
work, in order to produce hydrogen from pFA, the presence
of water and catalyst in the reaction medium is necessary.
Therefore, a combination of 1 mol% of RuIJp-cymene)Cl2
dimer as catalyst and water as solvent at 120 °C was
employed in the first attempt, and for the ratio of pFA to
benzonitrile 10 : 1was selected to be. It has been reported that
benzonitrile can be hydrated in the presence of water and
Ru-based catalysts to produce benzamide as the only
product,40 moreover it is known that the hydration of nitriles
to amides can also be performed by base catalysis (NaOH) or
acid catalysis.41 The challenge was to prevent the formation
of this side product. As was expected, a mixture of benzyl
alcohol to benzamide with the ratio of 1 : 6.8, was obtained
under these conditions and no formation of amine or imine
was detected in this case (Table 1, entry 1). To prevent the
formation of benzamide, various organic solvents were added
Scheme 2 A comparison of previously reported selective conversions
of nitriles to primary alcohols with the present work.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6092–6101 | 6093

View Article Online
Paper
Catalysis Science & Technology
Table 1 Solvent screening
Table 2 Base and ligand screening
1 : 2a (ratio)
Entry
Additive (base or ligand)
1 : 2a (ratio)
Entry
Solvent (ratio)
1
2
3
4
5
6
7
KOtBu
K2CO3
NH4OH
2,2′-Bipyridine
1,10-Phenanthroline
PPh3
1 : 21
1 : 6.4
1 : 11
55 : 45
51.7 : 48.3
48.6 : 52.4
57.5 : 42.5
1
2
3
4
5
6
7
8
H2O
13 : 87
43 : 57
52 : 48
51 : 49
56.4 : 43.6
45 : 55
57.2 : 42.8
38.5 : 61.5
H2O : EtOH (1 : 1)
H2O : i-PrOH (1 : 1)
H2O : 1,4-dioxane (1 : 1)
H2O : toluene (1 : 1)
H2O : MeOH (1 : 1)
H2O : toluene (1.5 : 2.5)
H2O : toluene (2.5 : 1.5)
dppe
Reaction conditions: benzonitrile (1 mmol), pFA, [Ru]
= [RuIJp-
a
cymene)Cl2]2, additive, solvent (4 mL in total), 120 °C, 10 h. NMR
yield (hexamethyldisilane was used as an internal standard); ratio
given at full conversion.
Reaction conditions: benzonitrile (1 mmol), pFA, [Ru]
= [RuIJp-
a
cymene)Cl2]2, solvent (4 mL in total), 120 °C, 10 h. NMR yield
(hexamethyldisilane was used as an internal standard); ratio given at
full conversion.
The use of triphenylphosphine increased the yield of
benzamide production whereas a slightly higher yield of
benzyl alcohol was obtained using dppe. However, as the
effects of all these ligands were negligible, and in order to
have better atom economy and less waste production as well
as to keep the reaction conditions as simple as possible, it
was decided that the reactions would be performed in the
absence of any additional ligand, because the reaction with
dppe (Table 2, entry 7) gave similar results in comparison to
the reaction without dppe (Table 1, entry 5).
Following these reactions, the temperature sensitivity and
its effect on the product selectivity was investigated.
Therefore, the benchmark reaction was done at different
temperatures ranging from RT to 130 °C (Table 3, entries 1–
6). At low temperatures (Table 3; entries 1–2) the nitrile
hydration to benzamide was much faster than the
hydrogenation generation, and TH was required for the
to the system to change the reaction medium. Addition of a
second solvent improved the formation of benzyl alcohol to a
large extent, while there was no sign of the formation of
either amine nor imine as the other possible by-products
(Table 1, entries 2–6). This is already a big step towards the
selective synthesis of benzyl alcohol from benzonitrile as in
most of the previously reported conditions, the formation of
a mixture of all possible by-products had been unavoidable.
This confirms that deaminative hydrogenation of nitriles
using H2 generated in situ from pFA was more selective
compared to the use of pressurized H2. According to the
obtained results, although a mixture of toluene and water
resulted in the highest ratio of benzyl alcohol to benzamide,
other organic solvents, even methanol and ethanol, could
also improve the selectivity (Table 1, entries 2, 6).
Observing the significant effect of the reaction medium
on this transformation, the ratio of the solvents was changed
in the next step. The best results were obtained using a ratio
of 1 : 1 from toluene and water which was selected as the
optimum solvent system (Table 1, entries 5, 7–8). It should
be also mentioned that for reaction times of 30, 60 and 120
min product formation could not be observed which
indicated that there was an induction period to form the
active hydrogenation catalyst species.
product formation. Strikingly,
a
high selectivity was
determined for benzyl alcohol (Table 3; entries 3–4) for
temperatures as low as 80 °C and 90 °C. Higher temperatures
were found to be favorable for the hydration of benzonitrile
to benzamide under our conditions such that no benzyl
Table 3 Temperature optimization
Different bases were also added to the reaction to
investigate the effect of pH on this transformation (Table 2,
entries 1–3). Based on the results, benzamide was the major
product in the presence of any organic or inorganic base in
the medium. This observation was also not surprising as the
conversion of nitrile to amide can be facilitated in basic
media.41
2,2′-Bipyridine, 1,10-phenanthroline, triphenylphosphine
and 1,2-bisIJdiphenylphosphino)ethane (dppe) were added
next, to the reaction mixture to attempt to see if they improve
the catalytic activity of the Ru catalyst (Table 2, entries 4–7).
As a result, the nitrogen-based ligands did not change the
outcome of the reaction although when using 1,10-
phenanthroline, the reaction did not complete within 12 h.
Entry
Temp. [°C]
1 : 2a (ratio)
1
2
3
4
5
6
25
60
80
90
110
130
0 : 100
<5 : 95
87.5 : 12.5
100 : 0
61.2 : 38.8
0 : 100
Reaction conditions: benzonitrile (1 mmol), pFA, [Ru]
= [RuIJp-
a
cymene)Cl2]2, solvent (4 mL in total), 25–130 °C, 10 h. NMR yield
(hexamethyldisilane was used as an internal standard); ratio given at
full conversion.
6094 | Catal. Sci. Technol., 2019, 9, 6092–6101
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
Table 4 Loading optimization of pFA
Table 5 Catalyst loading and control experiments
Entry
pFA : substrate ratio
1 : 2a (ratio)
Entry
Cat. (mol%)
1 : 2a (ratio)
1
2
3
4
5
6
1 : 1
3 : 1
5 : 1
7 : 1
8 : 1
9 : 1
0 : 100
0 : 100
58.3 : 41.7
86 : 14
92 : 8
1
2
3
4
5
6
1
0.2
0.5
100 : 0
54 : 46b
83.9 : 16.1
0.0 (without cat.)
1 (without water)
1 (without pFA)
No conversion
No conversion
No conversion
100 : 0
Reaction conditions: benzonitrile (1 mmol), pFA, [Ru]
=
[RuIJp-
Reaction conditions: benzonitrile (1 mmol), pFA, [Ru] = [RuIJp-
a
a
cymene)Cl2]2, solvent (4 mL in total), 90 °C, 10 h. NMR yield
(hexamethyldisilane was used as an internal standard); ratio given at
full conversion.
cymene)Cl2]2, solvent (4 mL in total), 90 °C, 10 h. NMR yield
(hexamethyldisilane was used as an internal standard); ratio given at
b
full conversion. After 12 h reaction time.
alcohol was detected by increasing temperature up to 130 °C
at all (Table 3; entry 6). The highest ratio of benzyl alcohol to
benzamide was observed at 90 °C which was selected as the
optimum temperature (Table 3; entry 4).
experiments were performed to make sure that the reaction
cannot proceed in the absence of the catalyst, water and pFA
(Table 5, entries 4–6). The results confirm the necessity of
these reagents for the formation of the alcohol derived from
a nitrile. Moreover, ruthenium nanoparticles and other metal
precursors were also tested in this reaction, but no activity
was observed. Activity was only observed for the dimeric
ruthenium arene complexes [RuIJbenzene)Cl2]2, [RuIJp-
cymene)I2]2 and [RuIJp-cymene)Cl2]2 (Table S1, ESI†).
Moreover, the effect of the reduction of the ratio of pFA to
the substrate was examined in order to reach the highest
possible atom economy in this reaction (Table 4, entries 1–6).
As a result, at ratios lower than 5 : 1 of pFA to benzonitrile,
no benzyl alcohol can be formed at all. This observation
originates from the fact that at lower pFA to benzonitrile
ratios, the concentration of dissolved H2 was not high
enough for the deaminative hydrogenation pathway to
compete with the hydration of benzonitrile and the
formation of benzamide. The very limited solubility of H2 in
water and organic solvents was a crucial parameter to
realizing significant H2 concentrations in solution.42–47
Therefore, at least a H2 saturated solution was needed to
overcome the obstacle with catalytic hydration due to the lack
of hydrogen. However, when the ratio of 7 : 1 was reached,
86% of benzyl alcohol was produced compared to 14% of
benzamide. This ratio was improved further, by employing
8 : 1 and 9 : 1 ratios of pFA to benzonitrile so that the latter
resulted in full conversion of the substrate to benzyl alcohol
without producing any benzamide. Therefore, a ratio of 9 : 1
of pFA to the substrate was chosen as optimum for the
reaction.
The catalyst loading was also evaluated further (Table 5,
entries 1–3). The results showed that the reaction can be
done even in the presence of lower amounts of RuIJp-
cymene)Cl2 dimer although full conversion of the starting
material was not achieved within 12 h using 0.2 mol% of it.
Furthermore, using 0.5 and 0.2 mol% of the catalyst, the
selectivity of benzyl alcohol over benzamide was reduced.
The best result was obtained using 1 mol% of the catalyst.
Higher catalyst loadings were not tested under these reaction
conditions as 1 mol% of the Ru complex already resulted in
full conversion of benzonitrile producing benzyl alcohol as
the only product of the reaction (Table 5, entry 1). Under
these optimized reaction conditions, three control
It is worth noting that this work has many advantages over
the previous reports on this topic. In none of the previous
reactions, were imine or amine detected as the by-product of
the reaction even if the reaction time was shortened. The
reaction was going through the reduction of benzonitrile via
its corresponding imine which readily converted in
equilibrium via the hemi-aminal to benzaldehyde, in aqueous
media. Thereafter, the latter would be easily, and rapidly,
reduced to benzyl alcohol under the reaction conditions.
However, as this was the last step, i.e., the reduction of
benzaldehyde to benzyl alcohol was happening very fast, and
no benzaldehyde can be detected. The other advantage of this
reaction was that all of the reactions were not sensitive to air,
thus neither precautions such as degassed oxygen-free
conditions nor autoclaves for hydrogen gas are required
because pFA acts as a hydrogen source in mild conditions.
Compared to previous protocols, a higher selectivity can be
achieved at a lower temperature with a practical and robust
protocol. Furthermore, CO2 is the only by-product of the
degradation of pFA. No ligand, base, or any other special
additive is required under these conditions which reduces
the amount of waste and increases the atom economy of the
reaction. In addition, the catalyst is commercially available
and inexpensive.
Substrate scope
The scope of deaminative hydrogenation of a wide variety of
aliphatic and aromatic nitriles was investigated using pFA as
the hydrogen source. Aliphatic nitriles which are known to be
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6092–6101 | 6095

View Article Online
Paper
Catalysis Science & Technology
less active than aromatic ones resulted in their
corresponding alcohols successfully without producing any
amine or imine as the side product. The conversions were
quantitative in most cases during the defined reaction times,
and very high yields were obtained. The reaction conditions
showed high tolerance for various functional groups
(Table 6). In the case of aromatic nitriles, very good to
excellent yields were obtained with short reaction times.
Aromatic compounds bearing both electron donating and
electron withdrawing substituents produced the desired
products selectively. Using 4-cyanobenzoic acid as starting
material, the nitrile group was converted to its alcohol
whereas the carboxylic moiety remained intact (Table 6, entry
5). No coupling product was observed in the reaction of
4-bromo- and 4-iodobenzonitrile under these conditions
(Table 6, entries 6, 7). 2-Cyanonaphthalene could also result
in the desired primary alcohol with high yield (Table 6, entry
8). However, by increasing the steric hindrance around the
cyano group, the rate of the reaction can be reduced such
that full conversion of the starting material was not achieved
during 16 h (Table 6, entries 9, 10). Phthalide was detected
as the major product of the deaminative hydrogenation of
phthalonitrile with 66% yield where 26% of the starting
material has been converted to 2-cyanobenzyl alcohol
(Table 6, entry 11). In this case, after the reductive
deamination of one of the nitrile groups resulting in the
formation of 2-cyanobenzyl alcohol, the OH group acts as a
nucleophile and attacks the neighboring cyano group which
after hydrolysis, produces phthalide. Increasing the reaction
time in this case will probably cause the rest of the
2-cyanobenzylalcohol (26%) to be converted to phthalide as
well. The deaminative hydrogenation of the nitrile group in
cinnamonitrile was accompanied with the reduction of the
double bond giving 3-phenylpropanol as the product (Table 2,
entry 12). Attempts to produce 3-pyridinemethanol from
3-pyridinecarbonitrile, however, were not as successful as
expected and isonicotinamide was formed as the main
product under the reaction conditions (Table 6, entry 13).
This observation is in accordance with the results of
Hashmi's work34 where they reported the same trend when
employing 3-pyridinecarbonitrile as the starting material. This
result can be attributed to the fact that the pyridine ring has
a stronger electron withdrawing effect on the cyano group
increasing the rate of the competing hydrolysis reaction by
facilitating the attack of water molecules to the more
electrophilic Csp center and production of nicotinamide. It
was also possible that 3-pyridinecarbonitrile was coordinated
to the ruthenium as a ligand and poisoned the catalyst.
quantitatively with good yields and selectivity (Table 6,
entries 14–17). Linear aliphatic nitriles have also worked as
well as the cyclic ones (Table 2, entries 18, 19), however,
adiponitrile as
a bifunctional nitrile required a longer
reaction time to be converted completely to 1,6-hexanediol
under the same conditions (Table 6, entry 20).
Production of deuterated alcohols from nitriles
Following these results, this method seems suitable for the
production of deuterated alcohols by employing deuterated
para-formaldehyde (D2-pFA) in D2O as D2 source for the
conversion of nitriles to α-deuterated alcohols under similar
conditions (Scheme 3). The deuterated product A was formed
quantitatively within 24 hours, as confirmed by 1H-NMR,
disappearance of CH2OH and CH2OH signals, and the
isotope shift of the GC-MS analysis. Within 16 h the degree
of deuteration reached 93% for the α,α-dideuterated alcohol
and 6.5% of the α-monodeuterated alcohol. By running the
reductive deamination with non-deuterated pFA in D2O, a
mixture of α,α-dideuterated alcohol (42.5%), non-deuterated
alcohol (32.5%) and α-monodeuterated alcohol was obtained.
Therefore, the procedure presented with D2-pFA and D2O has
the potential to be used for the production of α,α-
dideuterated alcohols from nitriles in general.
Mechanistic studies
Following the previous studies on hydrogen generation from
aqueous formaldehyde and pFA,27,28 under similar reaction
conditions in terms of the pH of the aqueous phase of the
system (pH
= 5.1), the hydrogen generation in the
deaminative reduction followed the same pathway and
neither dismutase27 nor the Cannizarro reaction played a role
in the degradation of pFA, while the latter one only occurred
in strongly basic media (pH > 10). Indeed no methanol was
detected and intermediate formed formic acid decomposed
solely to CO2 and H2 during the formaldehyde reforming.28
For the identification of the active species the deaminative
reduction was conducted with H2 instead of pFA with the
same catalyst precursor. Interestingly, no benzyl alcohol was
formed which implied that pFA was required to form the
catalytic active species as found in the formaldehyde
reforming protocol.28 To underline this assumption the
reaction was performed in a two-chamber reaction vessel set-
up connected via the gas phase (ESI), where one flask
contained the catalyst precursor and pFA in a water–toluene
mixture for the H2 production and the second flask
contained the catalyst precursor and benzonitrile in a water–
toluene mixture. Again, no benzyl alcohol was detected in
this type of decoupled TH, instead the benzonitrile was fully
converted into benzamide.
After screening the aryl nitriles, reactivity of aliphatic
nitriles as less reactive substrates were investigated under the
optimized conditions. The corresponding primary alcohols
could be produced from the aliphatic nitriles not only with
good yields but also with higher selectively compared to the
aromatic ones. Aliphatic cyclic nitriles such as benzyl
cyanide, cyclohexane carbonitrile and 1-adamantane
carbonitrile were converted to the desired products
Additionally, to obtain further insights the benchmark
reaction with benzonitrile as substrate was analysed by ESI-
MS to identify complex species carrying different ligands
incorporated by ligand exchange during the reaction.
Moreover, a complementary reaction was conducted without
6096 | Catal. Sci. Technol., 2019, 9, 6092–6101
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
Table 6 Substrate scope
Entry
1
Substrate
Product
Yielda [%]
93
2
3
93
92
4
5
6
7
93
89
88
87
8
85
83
9a
10b
11b
71
66
12
89
21
13c
14
15
16
91
87
90
17
18
77
88
19
20b
86
83
a
Reaction conditions: nitrile (1 mmol), pFA (9 mmol), [Ru] = [RuIJp-cymene)Cl2]2 (1 mol%), H2O : toluene (2 mL : 2 mL), 90 °C, 16 h. Isolated
yield. b Reaction time was increased to 24 h. c Nicotinamide has been formed as the main product with 74% yield.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6092–6101 | 6097

View Article Online
Paper
Catalysis Science & Technology
Scheme 3 Production of deuterated primary alcohols using D2-pFA in
D2O : toluene.
pFA. An aliquot was taken after 4 h from each reaction. The
ESI-MS data of the deaminative reduction showed that the
decomposition of pFA via aqueous formaldehyde to H2 and
CO2 in the presence of [RuIJp-cymene)Cl2]2 1 resulted in the
formation of the dimeric species [(RuIJp-cymene))2μ-HIJμ-
HCO2)μ-Cl]+ 3 (m/z: 553.002 amu) as previously reported,28
whereas the short-living methanediol complex 2 cannot be
detected. This indicated that the pFA decomposition also
occurs in the presence of nitrile in toluene. In addition,
[(RuIJp-cymene))2IJμ-H)2IJμ-HCO2)]+ 4 (m/z: 518.042 amu) was
detected, which was formed by ligand exchange. Also, [(RuIJp-
cymene))2IJμ-H)2μ-Cl]+ 5 (m/z: 509.012 amu) which was formed
upon decarboxylation of the formiato ligand and [(RuIJp-
cymene))2IJμ-H)3]+ 6 (m/z: 475.051 amu) which was formed via
heterolytic cleavage of H2 and the substitution of the halide
Scheme 5 The structures of some of the unsaturated species 7–9 and
the intermediate/product complex 10 detected by ESI-MS from
aliquots of the reaction mixture.
ligand (Scheme 4). The ESI MS spectra did not even show
traces of [RuIJp-cymene)Cl2]2 which means that it had been
converted completely to the other more reactive species
under these conditions. These observations were in
agreement with the results of previous work on the
degradation of (para)formaldehyde in the presence of the
same catalyst but in the absence of nitrile.26–28
After confirming the pFA degradation pathway under these
conditions, the ESI-MS data was analysed further to get
insights into the reductive deamination of benzonitrile.
Different pathways were possible for such transformation
Scheme 6 Two possible pathways (a and b) for the formation of the
Scheme
4
The formation of dimeric complexes during pFA
hemi-aminal intermediate 12 as a potential intermediate for the
degradation in the reductive deamination of nitrile.
formation of benzyl alcohol.
6098 | Catal. Sci. Technol., 2019, 9, 6092–6101
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
which can be divided in to two categories: (I) the one with the
dimeric catalytic species, and, (II) the one involving the
monomeric ruthenium species. The data showed that no
dimeric ruthenium species containing different combinations
of the catalyst and benzonitrile and pFA in the spectrum can
be detected. Likewise, no complex species carrying ligands
such as benzamide or benzaldehyde were formed from
benzonitrile. In addition, under these conditions no ligand
exchange between p-cymene and toluene could be observed
for the dimeric ruthenium species. Taking the previously
reported observations into account, the direct involvement of
the dimeric complexes 2–6 for the deaminative reduction of
nitriles was excluded. Therefore, the reaction mechanism
most likely involved monomeric ruthenium species. The
comparison of the ESI MS spectra of the benchmark reaction
with the pFA-free blank reaction, enabled easier identification
of the ruthenium species carrying benzonitrile as ligand.
As has been mentioned previously, benzonitrile, per se,
can coordinate to the ruthenium center via its nitrogen loan
pair or its triple bond. In 2010, Sabo-Etienne investigated the
Scheme 7 The proposed reaction pathway for the deaminative reduction of nitriles.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6092–6101 | 6099

View Article Online
Paper
Catalysis Science & Technology
mechanism of the hydrogenation of benzonitrile in the
presence of RuH2IJσ-H2)2IJPCyp3)2 as catalyst to produce
amine. This showed that under their conditions, benzonitrile
can be coordinated to the ruthenium center via ortho-
directed C–H activation. The formation of such an
intermediate enabled the imine intermediate to be trapped
and be reduced to amine before it was hydrolysed to
benzaldehyde. Therefore, considering that in this case benzyl
alcohol is the only product, such species were not expected.
This assumption was verified by the absence of any possible
species of this kind in the ESI-MS spectra. It is also worth
noting that none of the expected primary, secondary and
tertiary amines resulting from the simple reduction of
benzonitrile31 existed in the reaction mixture according to
the spectra. Interestingly, signals were found in the ESI-MS
which could be assigned to the exact masses of the
monomeric ruthenium species 7–9 (Scheme 5) carrying
benzonitrile or benzamide (derived from the corresponding
hemi-aminal), whereas the absence of all the potential
related dimeric species in the spectra, confirmed the
assumption of the involvement of the monomeric ruthenium
intermediate in the course of the deaminative reduction
(Scheme 6). The exact formation of these species 7–9 was
unknown, although a cascade of ligand exchange reactions in
the presence of an excess of nitrile is most likely to be an
important issue. However, highly-unsaturated species were
usually formed under MS-conditions due to the loss of
ligands during the measurement,28 therefore in solution the
vacant coordination sites were occupied by neutral molecules
from the reaction mixture (e.g., H2O, substrates,
intermediates and products). Strikingly, a ruthenium species
10 (Scheme 5) was also detected carrying three intermediates
as ligands: hemi-aminal, benzaldehyde and benzyl alcohol
which gave further indications that these monomeric species
were involved in the nitrile hydrogenation.
This lead us two possible reaction pathways
(Scheme 6a and b). In the first one, benzonitrile could be
reduced to the imine intermediate 11 under the reaction
conditions which would be converted to the hemi-aminal 12
immediately due to the presence of plenty of water molecules
in the medium.34 In the other pathway, benzonitrile could be
hydrolysed initially to benzamide 13 which could then be
subsequently reduced to 12 in the presence of the catalyst and
the hydrogen formed in situ.48 Releasing NH3 from 12 formed
benzaldehyde 14 in both pathways which then, could be easily
reduced to benzyl alcohol 15 under the reaction conditions.
To distinguish between the pathways in Scheme 6, an
experiment was performed using benzamide as substrate
instead of benzonitrile. If benzyl alcohol can be produced
from benzamide as well, it could be concluded that the
second pathway in Scheme 6 was valid in this case. However,
no benzyl alcohol was detected after 24 h under these
conditions and only starting material could be observed in
the NMR spectrum. Therefore, this pathway can be excluded.
Based on these evidence and the previous studies in the
field,34,48–51 the following reaction pathway can be proposed
for the reductive deamination of nitriles using pFA in
aqueous media (Scheme 7). Initially, [RuIJp-cymene)Cl2]2
1
reacted in two different pathways: (I) the dimeric complexes
3 and 4 were formed by reforming pFA in water,28 which then
produced (catalytically) H2, CO2 and formic acid over the
course of the reaction (Scheme 7: hydrogen production cycle).
(II) Complex 1 was converted into complex 8 which readily
reacted with the hydrogen produced in situ, then yielded the
benzylimine 11 (Scheme 7: nitrile hydrogenation cycle) which
was converted to benzaldehyde 14 via the hemi-aminal 12 in
equilibrium followed by NH3 extrusion because of the large
excess of water. The hemi-aminal 12 could also react with
one of the ruthenium species present in the medium to form
species 9. Benzaldehyde 14 could then react with complex 8
in a Noyori-type HT reaction via the transition state 17 and
then yielding benzyl alcohol and the catalytically active
species 9 was regenerated (Scheme 7: aldehyde reduction
cycle). In particular the intermediates 8–10 (Scheme 5)
support the proposed catalytic nitrile hydrogenation and
deamination as depicted in Scheme 7.
Conclusions
In summary, the reductive deamination of different kinds of
aliphatic and aromatic nitriles was reported for the first time
using para-formaldehyde as a hydrogen donor under mild and
practical reaction conditions because no air or moisture
sensitive chemicals or special glassware are required. The
reaction has a high selectivity for the production of primary
alcohols in the presence of the ruthenium dichloroIJp-
cymene)rutheniumIJII) dimer used as an air-stable and
commercially available catalyst and does not result in the
formation of any amine or other possible side products with
this protocol. The analysis of the ESI-MS data also revealed the
nature of the catalytically active species in this transformation.
Based on these results, the reaction mechanism contains three
different catalytic cycles: (I) a hydrogen generation cycle, (II) a
nitrile hydrogenation cycle, and, (III) an aldehyde
hydrogenation cycle. In the degradation of pFA to CO2 and H2,
the dimeric ruthenium complexes are involved, while the
involvement of benzonitrile in the following catalytic cycles
results in the formation of N,O-coordinated monomeric species,
which act as the active catalyst in the reductive deamination of
nitriles in the following cycles. It is worth noting that the
ruthenium nanoparticles showed no activity under these
reaction conditions.
Based on these data, as well as other experimental
observations, a plausible mechanism was also proposed for
this reaction, although more mechanistic studies are
required to reveal further aspects about the reaction pathway.
The authors believe that this work can open-up a new avenue
in the field of transfer hydrogenation.
Conflicts of interest
There are no conflicts to declare.
6100 | Catal. Sci. Technol., 2019, 9, 6092–6101
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
Acknowledgements
21 M. Vilches-Herrera, S. Werkmeister, K. Junge, A. Börner and
M. Beller, Catal. Sci. Technol., 2014, 4, 629–632.
22 S. Manna and A. P. Antonchick, ChemSusChem, 2019, 12(13),
3094–3098, Special Issue: Sustainable Organic Synthesis.
23 D. Wang and D. Astruc, Chem. Rev., 2015, 115, 6621–6686.
24 G. Brieger and T. J. Nestrick, Chem. Rev., 1974, 74, 567–580.
25 L. E. Heim, H. Konnerth and M. H. G. Prechtl, Green Chem.,
2017, 19, 2347–2355.
26 L. E. Heim, S. Vallazza, D. van der Waals and M. H. G.
Prechtl, Green Chem., 2016, 18, 1469–1474.
27 D. van der Waals, L. E. Heim, S. Vallazza, C. Gedig, J. Deska
and M. H. G. Prechtl, Chem. – Eur. J., 2016, 22, 11568–11573.
28 L. E. Heim, N. E. Schlörer, J.-H. Choi and M. H. G. Prechtl,
Nat. Commun., 2014, 5, 3621.
29 M. Trincado, V. Sinha, R. E. Rodriguez-Lugo, B. Pribanic, B.
de Bruin and H. Grützmacher, Nat. Commun., 2017, 8, 14990.
30 M. Trincado, H. Grutzmacher and M. H. G. Prechtl, Phys.
Sci. Rev., 2018, 3(5), DOI: 10.1515/psr-2017-0013.
31 J. H. Choi and M. H. G. Prechtl, ChemCatChem, 2015, 7, 1023–1028.
32 N. M. Yoon, J. H. Ahn, D. K. An and Y. S. Shon, J. Org.
Chem., 1993, 58, 1941–1944.
The authors gratefully acknowledge financial support
provided by Ministerium für Innovation, Wissenschaft und
Forschung (NRW-returnee award 2009), the Heisenberg-
Program (Deutsche Forschungsgemeinschaft), the COST
Action “C–H Activation in Organic Synthesis (CHAOS)”, and
the Ernst-Haage-Prize 2014 of the Max-Planck-Institute for
Chemical Energy Conversion. G. T. acknowledges the
Alexander von Humboldt Foundation for a fellowship. Prof.
Dr. Mathias Schäfer and Michael Neihs are acknowledged for
support with the ESI-MS analysis.
Notes and references
1 R. C. Hoye, J. Chem. Educ., 1999, 76, 33.
2 R. O. Hutchins and M. K. Hutchins, in Comprehensive
OrganicSynthesis, ed. B. M. Trost and I. Fleming, Pergamon
Press, Oxford, U. K., 1991, vol. 8, pp. 25–78.
3 S. E. Clapham, A. Hadzovic and R. H. Morris, Coord. Chem.
Rev., 2004, 248, 2201–2237.
4 S. Gladiali and E. Alberico, Chem. Soc. Rev., 2006, 35, 226–236.
5 R. Noyori and S. Hashiguchi, Acc. Chem. Res., 1997, 30, 97–102.
6 C. Bornschein, S. Werkmeister, B. Wendt, H. Jiao, E.
Alberico, W. Baumann, H. Junge, K. Junge and M. Beller,
Nat. Commun., 2014, 5, 4111.
33 O. G. Backeberg and B. Staskun, J. Chem. Soc.,
1962, 3961–3963.
34 I. G. Molnár, P. Calleja, M. Ernst, A. S. K. Hashmi and T.
Schaub, ChemCatChem, 2017, 9, 4175–4178.
35 Y. P. Xie, J. Men, Y. Z. Li, H. Chen, P. M. Cheng and X. J. Li,
Catal. Commun., 2004, 5, 237–238.
36 C. Bianchini, V. Dal Santo, A. Meli, W. Oberhauser, R. Psaro
and F. Vizza, Organometallics, 2000, 19, 2433–2444.
37 S. Takemoto, H. Kawamura, Y. Yamada, T. Okada, A. Ono, E.
Yoshikawa, Y. Mizobe and M. Hidai, Organometallics,
2002, 21, 3897–3904.
38 Y. Z. Wang, S. Furukawa, Z. Zhang, L. Torrente-Murciano,
S. A. Khan and N. Yan, Catal. Sci. Technol., 2019, 9, 86–96.
39 Y. Z. Wang, S. Furukawa and N. Yan, ACS Catal., 2019, 9,
6681–6691.
7 S. Chakraborty, G. Leitus and D. Milstein, Chem. Commun.,
2016, 52, 1812–1815.
8 S. Lange, S. Elangovan, C. Cordes, A. Spannenberg, H. Jiao,
H. Junge, S. Bachmann, M. Scalone, C. Topf, K. Junge and
M. Beller, Catal. Sci. Technol., 2016, 6, 4768–4772.
9 A. Mukherjee, D. Srimani, S. Chakraborty, Y. Ben-David and
D. Milstein, J. Am. Chem. Soc., 2015, 137, 8888–8891.
10 S. Elangovan, C. Topf, S. Fischer, H. Jiao, A. Spannenberg,
W. Baumann, R. Ludwig, K. Junge and M. Beller, J. Am.
Chem. Soc., 2016, 138, 8809–8814.
11 R. Adam, C. B. Bheeter, J. R. Cabrero-Antonino, K. Junge, R.
Jackstell and M. Beller, ChemSusChem, 2017, 10, 842–846.
12 I. D. Alshakova, B. Gabidullin and G. I. Nikonov,
ChemCatChem, 2018, 10, 4860–4869.
13 J. A. Garduño and J. J. García, ACS Omega, 2017, 2, 2337–2343.
14 D. Kim, B. Kang and S. H. Hong, Org. Chem. Front., 2016, 3,
475–479.
15 V. H. Mai and G. I. Nikonov, Organometallics, 2016, 35,
943–949.
16 L. E. Heim, D. Thiel, C. Gedig, J. Deska and M. H. G.
Prechtl, Angew. Chem., Int. Ed., 2015, 54, 10308–10312.
17 A. Kaithal, B. Chatterjee and C. Gunanathan, J. Org. Chem.,
2016, 81, 11153–11161.
40 W. L. Ounkham, J. A. Weeden and B. J. Frost, Chem. – Eur. J.,
2019, 25, 10013–10020.
41 T. E. Schmid, A. Gomez-Herrera, O. Songis, D. Sneddon, A.
Revolte, F. Nahra and C. S. J. Cazin, Catal. Sci. Technol.,
2015, 5, 2865–2868.
42 U. J. Jauregui-Haza, E. J. Pardillo-Fontdevila, A. M. Wilhelm
and H. Delmas, Lat. Am. Appl. Res., 2004, 34, 71–74.
43 V. I. Baranenko and V. S. Kirov, Sov. At. Energy, 1989, 66, 30–34.
44 E. Brunner, J. Chem. Eng. Data, 1985, 30, 269–273.
45 V. V. Leskin, V. A. Larionov, G. L. Tudorovskaya and N. A.
Safonova, Khim. Promst., 1983, 407–409.
46 U. Schmidt, Tellus, 1979, 31, 68–74.
47 T. E. Crozier and S. Yamamoto, J. Chem. Eng. Data, 1974, 19,
242–244.
48 N. M. Rezayee, D. C. Samblanet and M. S. Sanford, ACS
Catal., 2016, 6, 6377–6383.
18 L. Liu, Y. Liu, Y. Ai, J. Li, J. Zhou, Z. Fan, H. Bao, R. Jiang, Z.
Hu, J. Wang, K. Jing, Y. Wang, Q. Liang and H. Sun, iScience,
2018, 8, 61–73.
19 S. Sahler, S. Sturm, M. T. Kessler and M. H. G. Prechtl,
Chem. – Eur. J., 2014, 20, 8934–8941.
49 M. Naruto and S. Saito, Nat. Commun., 2015, 6, 8140.
50 E. Baráth, Catalysts, 2018, 8, 671.
20 J. W. Yang, M. T. Hechavarria Fonseca and B. List, Angew.
Chem., Int. Ed., 2004, 43, 6660–6662.
51 J. S. M. Samec, J.-E. Bäckvall, P. G. Andersson and P. Brandt,
Chem. Soc. Rev., 2006, 35, 237–248.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol., 2019, 9, 6092–6101 | 6101