Manganese(III) Porphyrin-Catalyzed Dehydrogenation of Alcohols to form Imines, Tertiary Amines and Quinolines

Article abstract of DOI:10.1002/chem.201900737

Manganese(III) porphyrin chloride complexes have been developed for the first time as catalysts for the acceptorless dehydrogenative coupling of alcohols and amines. The reaction has been applied to the direct synthesis of imines, tertiary amines and quinolines where only hydrogen gas and/or water are formed as the by-product(s). The mechanism is believed to involve the formation of a manganese(III) alkoxide complex which degrades into the aldehyde and a manganese(III) hydride species. The latter reacts with the alcohol to form hydrogen gas and thereby regenerates the alkoxide complex.

Full text of DOI:10.1002/chem.201900737

A Journal of
Accepted Article
Title: Manganese(III) Porphyrin-Catalyzed Dehydrogenation of
Alcohols to form Imines, Tertiary Amines and Quinolines
Authors: Kobra Azizi, Sedigheh Akrami, and Robert Madsen
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201900737
Link to VoR: http://dx.doi.org/10.1002/chem.201900737
Supported by

10.1002/chem.201900737
Chemistry - A European Journal
FULL PAPER
Manganese(III) Porphyrin-Catalyzed Dehydrogenation of Alcohols
to form Imines, Tertiary Amines and Quinolines
Kobra Azizi, Sedigheh Akrami and Robert Madsen*[a]
Abstract: Manganese(III) porphyrin chloride complexes have been
developed for the first time as catalysts for the acceptorless
dehydrogenative coupling of alcohols and amines. The reaction has
been applied to the direct synthesis of imines, tertiary amines and
quinolines where only hydrogen gas and/or water are formed as the
byproduct(s). The mechanism is believed to involve the formation of
shown to proceed through the amido complex of the
corresponding manganese(III) salan species (Scheme 1).[6] This
discovery raises the question whether other manganese(III)
complexes with tetradentate ligands would also be able to
catalyze acceptorless dehydrogenations?
a
manganese(III) alkoxide complex which degrades into the
aldehyde and a manganese(III) hydride species. The latter reacts
with the alcohol to form hydrogen gas and thereby regenerates the
alkoxide complex.
Introduction
Since the turn of the millennium, significant efforts have been
devoted to the development of acceptorless dehydrogenative
reactions with alcohols.[1] In these transformations, a metal
catalyst mediates the release of hydrogen gas to form the
corresponding carbonyl compound, which can then undergo
subsequent reactions in the same pot to generate other
functional groups such as amines, imines, amides, carboxylic
acids and esters. The transformations are usually catalyzed by
complexes of precious metals where especially ruthenium and
iridium compounds have been widely employed.[1] In recent
years, however, complexes based on the Earth-abundant metals
iron, cobalt and manganese have surfaced as a new type of
catalysts for the acceptorless dehydrogenations.[2]
In fact, the first manganese complex was presented in
2016[3] and since then many manganese-based catalysts have
been developed for both the dehydrogenation[2a-d,4] and the
reverse hydrogenation reaction.[2a-d,5] With one exception,[6]
although, they are all manganese(I) complexes stabilized by CO
ligands which are not inexpensive since they are prepared from
the costly carbonyl complex Mn2(CO)10. Thus, it would be highly
desirable to identify manganese complexes in the more common
oxidation states for these dehydrogenative reactions.
Scheme 1. Manganese(III) salen-catalyzed alcohol dehydrogenation and the
mechanism.
Manganese(III) porphyrin complexes and related compounds
are well known for catalyzing oxidation reactions in the presence
of a stoichiometric oxidant including hydroxylation/halogenation
of alkanes, epoxidation of alkenes and oxidation of alcohols.[7]
The transformations involve catalytic cycles with various
manganese(III), (IV) and (V) species.[7] However, no
dehydrogenation of a functional group with the release of
hydrogen gas has been reported before with these complexes.
Herein, we describe the development of different
manganese(III) porphyrin complexes for catalyzing the
dehydrogenative coupling of alcohols and amines to form imines,
tertiary amines and quinoline. In addition, the mechanism has
been investigated indicating another pathway for the
dehydrogenation than with the manganese(III) salen complexes.
The findings underline the unique reactivity of manganese(III)
complexes with tetradentate ligands in hydrogen gas-releasing
transformations.
Very recently, we presented the first manganese(III) complex
for the acceptorless dehydrogenation of alcohols and elucidated
the mechanism (Scheme 1).[6] Notably, it is a salen complex,
which is a known catalytic motif in organic chemistry, but for
different reactions. The mechanism was investigated by
experimental and theoretical methods and the reaction was
Results and Discussion
Benzyl alcohol and cyclohexylamine were selected as the
substrates for the exploratory experiments where several
manganese(III) (tetraaza)porphyrin chloride complexes were
investigated for the dehydrogenation (Figure 1). Notably, the
common and commercially available tetraphenylporphyrin (TPP)
complex A gave a good conversion into the imine 1 in the
absence of any additional additives (Table 1, entry 1). The
[a]
Dr. K. Azizi, S. Akrami, Prof. Dr. R. Madsen
Department of Chemistry
Technical University of Denmark
2800 Kgs. Lyngby (Denmark)
E-mail: rm@kemi.dtu.dk
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/chem.201900737
Chemistry - A European Journal
FULL PAPER
influence of the substituents on the tetradentate ligand was then
examined and the yield decreased with the corresponding para-
carboxylate complex B (entry 2). Electron-donating groups, on
the other hand, had a beneficial effect on the imine formation
and a quantitative yield was obtained with both the para-
methoxy complex C and the para-(diethylamino) compound D
(entries 3 and 4). Slightly lower yields were observed with
C and D (entries 12 – 14). These results point towards complex
H as the preferred catalyst for the dehydrogenation. Lowering
the loading further to 1% gave a moderate yield of the imine and
the same was observed when the reaction was performed in
refluxing toluene (entries 15 and 16). Surprisingly, additives
such as inorganic bases, phosphines and amines had
essentially no impact on the transformation (see Table S1 in the
supporting information). Hence, an optimized protocol has been
developed where equimolar amounts of alcohol and amine are
reacted with 2% of complex H in refluxing mesitylene to afford
the corresponding imine.
tetramesityl- and octaethylporphyrin complexes
E and F,
respectively, as well as with phthalocyanine complex G (entries
5 – 7). Tert-butyl groups on the tetraazaporphyrin ring, on the
other hand, had a beneficial effect on the transformation and a
quantitative yield was obtained with complex H (entry 8). Except
for complex B, all the investigated (tetraaza)porphyrin
complexes were fully soluble under the reaction conditions.
Table 1. Optimizing manganese(III)-catalyzed alcohol dehydrogenation.[a]
Entry
X
Catalyst
t [h]
Yield [%][b]
1
5
5
5
5
5
5
5
5
3
3
3
2
2
2
1
5
A
B
C
D
E
F
60
60
60
60
60
60
60
60
48
48
48
48
48
48
48
48
90
2
55
3
100
100
85
4
5
6
94
7
G
H
C
D
H
C
D
H
H
H
83
8
100
88
9
10
11
12
13
14
15
16[c]
90
100
76
83
92
68
50
[a] Reaction conditions: BnOH (1 mmol), CyNH2 (1 mmol), catalyst (0.0X
mmol), tetradecane (0.5 mmol, internal standard), 4Å MS (150 mg),
mesitylene (4 mL), reflux. [b] Determined by GC. [c] In toluene at reflux.
Figure 1. Structure of (tetraaza)porphyrin complexes used in the investigation.
This new method was then applied to several different
alcohols and amines to explore the substrate scope of the
transformation (Table 2). First, various alcohols were reacted
with cyclohexylamine and the products were isolated by flash
chromatography. This afforded a 85% yield of 1 with the parent
benzyl alcohol while p-methyl- and p-(tert-butyl)benzyl alcohol
gave 89 and 90% yield, respectively, of 2 and 3. The electronic
Thus, the most promising results were achieved with
complexes C, D and H and several experiments were performed
to lower the loading, the reaction time and the temperature. With
3% of the catalysts and a 48 h reaction time, complex H still
gave a quantitative yield of 1 while C and D furnished around
90% yield (entries 9 – 11). With 2% of the catalysts, complex H
afforded 92% yield while lower yields were again obtained with
This article is protected by copyright. All rights reserved.

10.1002/chem.201900737
Chemistry - A European Journal
FULL PAPER
properties of the para substituent had very little influence on the
imination since both electron-donating (as methoxy and
methylthio) and electron-withdrawing groups (as halides,
trifluoromethyl and methoxycarbonyl) gave very similar yields of
the products 4 – 10. Notably, no competing dehalogenation was
observed with the p-chloro, p-bromo and p-iodo substrates. The
only exception was p-nitrobenzyl alcohol, which furnished imine
11 in a moderate yield of 62% due to a concomitant reduction of
the nitro group to give N-(p-aminobenzylidene)cyclohexylamine
as a byproduct. Ortho substituted substrates gave a slightly
lower yield than the corresponding para compounds as seen for
o-methyl, o-chloro- and o-bromobenzyl alcohol which afforded
imines 12 – 14 in 72 – 82% yield. 1-Naphthalenemethanol, on
the contrary, gave a higher 88% yield of the resulting imine 15.
[a] Reaction conditions: Alcohol (1 mmol), amine (1 mmol), H (0.02 mmol), 4Å
MS (150 mg), mesitylene (4 mL), reflux, 48 h (isolated yields).
Changing the amine to benzylamine gave 92% yield of 16
with benzyl alcohol while slightly lower yields of imines 17 – 19
were obtained with o-chloro, o-bromo- and 2,4,6-trimethylbenzyl
alcohol. In the latter three cases, homocoupling of benzylamine
to generate 16 was observed as a side reaction. The more
hindered amines 1-adamantylamine and benzhydrylamine gave
the corresponding imines 20 and 21 in 76 and 62% yield,
respectively. The transformation could also be applied to
arylamines where aniline afforded 80% yield of 22 in the reaction
with benzyl alcohol. In addition, the imination could be
performed with the aliphatic alcohol hexan-1-ol to give imine 23
in the reaction with benzylamine. Although full conversion of the
alcohol was achieved, the isolated yield was only 43%, which is
due to the instability of 23 towards hydrolysis during the
purification. When the crude reaction mixture was analyzed by
NMR, the yield of 23 was determined to be 72% by comparison
with an internal standard and after removing the catalyst.
Table 2. Imine formation from different alcohols and amines.[a]
The evolution of hydrogen gas was confirmed by performing
the reaction in Table 1, entry 14 in a closed two-chamber
system[8]
where
the
other
compartment
contained
diphenylacetylene and Pd/C in methanol. This experiment
showed reduction of diphenylacetylene to (Z)-stilbene after full
conversion into the imine and verifies the dehydrogenative
pathway of the imine formation. When the experiment in Table 1,
entry 14 was carried out with PhCD2OH instead of PhCH2OH,
the only product formed was PhCD=NCy with no hydrogen
incorporation into the benzylic position. This may indicate that
the dehydrogenation takes place by a monohydride pathway and
a metal-dihydride compound is not formed. Radical species do
not appear to be part of the mechanism since addition of 2 equiv.
of the scavengers 2,4-diphenyl-4-methylpent-1-ene and
cyclohexa-1,4-diene had no influence on the conversion into the
imine under the optimized conditions in Table 1, entry 14. The
two scavengers were not converted during the transformations
suggesting that the catalyst in unable to mediate the
hydrogenation of olefins.
Table 3. Optimizing alkylation of secondary amine into tertiary amine.[a]
Entry
X
Catalyst
Additive
Conversion of
BnNHMe [%]
Yield [%][b]
1
2
3
3
A
C
-
-
80
82
70
68
This article is protected by copyright. All rights reserved.

10.1002/chem.201900737
Chemistry - A European Journal
FULL PAPER
3
3
3
3
3
2
3
H
A
C
H
A
A
-
85
65
88
81
80
81
93
4
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
100
100
100
100
100
5
6
7
8[c]
[a] Reaction conditions: BnOH (1 mmol), BnNHMe (1 mmol), catalyst (0.0X
mmol), additive (0.2 mmol), tetradecane (0.5 mmol, internal standard), 4Å MS
(150 mg), mesitylene (4 mL), reflux, 48 h. [b] Determined by GC. [c] With 1.2
mmol of BnOH.
Dehydrogenative reactions between alcohols and amines
can also be performed with secondary amines where the
product is a tertiary amine. This transformation has been
demonstrated with a variety of alcohols and amines when
ruthenium,[9] iridium,[10] palladium[11] and iron[12] catalysts are
employed. With manganese, however, only a special example
exists with a manganese(I) catalyst where activated aromatic
compounds are aminomethylated with methanol and secondary
amines.[13] Thus, benzyl alcohol and N-methyl benzylamine were
selected as the starting substrates to investigate the alkylation
where water is formed as the only byproduct. Interestingly, the
reaction afforded the desired tertiary amine 24, but complete
conversion of the starting amine was not observed with 3% of
complexes A, C and H (Table 3, entries 1 – 3). However, when
K2CO3 was added as a basic additive, all three catalysts gave
full conversion of the starting material (entries 4 – 6). The
highest yield was observed with complex A, which seems to be
the preferred catalyst in this case. Other additives were also
investigated such as KOtBu, Cs2CO3, Ca3N2, MgSO4, NaHCO3
and pyridine, but no improvements were observed (results not
shown). Lowering the loading of A to 2% decreased the yield
(entry 7) while a slightly higher yield was obtained by increasing
the amount of the alcohol to 1.2 equiv. (entry 8). These final
conditions were selected for general use and applied to different
alcohols and amines where the products again were isolated by
flash chromatography (Table 4).
Table 4. Synthesis of tertiary amines from alcohols and secondary amines.[a]
This article is protected by copyright. All rights reserved.

10.1002/chem.201900737
Chemistry - A European Journal
FULL PAPER
[a] Reaction conditions: Alcohol (1.2 mmol), amine (1 mmol), A (0.03 mmol),
K2CO3 (0.2 mmol), 4Å MS (150 mg), mesitylene (4 mL), reflux, 48 h (isolated
yields).
this case since it only gave 2-phenylquinoline (48) in 9% yield
(entry 1) while KOH, NaOH and KOtBu afforded 21 – 32% yield
(entries 2 – 4). Combining KOH and KOtBu gave an improved
outcome (entry 5) and especially when the reaction time was
extended (entry 6). A further increase in the yield was obtained
upon adding a small amount of pyridine as an additional base
(entry 7) and with 1.5 equiv. of 1-phenylethanol the target
quinoline was produced in 76% yield (entry 8). No improvement
was observed by further increasing the amount of 1-
phenylethanol (entry 9) while the yield decreased with lower
amounts of complex A (entry 10) or with only one of the
inorganic bases (entries 11 and 12). Thus, an optimized protocol
has been established where quinolines are prepared by reacting
o-aminobenzyl alcohols with 1.5 equiv. of secondary alcohols in
the presence of 5% of complex A and a mixture of KOH, KOtBu
and pyridine.
The transformation afforded a 78% isolated yield of the
parent compound 24 while p-methoxy- and p-chlorobenzyl
alcohol gave 82 and 70% yield, respectively, of 25 and 26 in the
reaction with N-methyl benzylamine. In the same reaction, p-
nitrobenzyl alcohol provided the tertiary amine 27 in a moderate
yield of 52% due to several byproducts whereas 2,4-
dimethylbenzyl alcohol furnished the corresponding amine 28 in
67% yield. N-Ethyl benzylamine gave the same yields as the N-
methyl analogue in the reaction with benzyl and p-
methoxybenzyl alcohol to afford 29 and 30. N-Methyl p-
methylbenzylamine produced 80% yield of both 31 and 32 in the
reaction with the two alcohols while the alkylation of N-methyl m-
methylbenzylamine with benzyl alcohol resulted in 33 in 73%
yield. Benzyl and p-methoxybenzyl alcohol were also reacted
Table 5. Optimizing dehydrogenative quinoline synthesis.[a]
with
N-methyl
p-bromobenzylamine,
N-methyl
m-
chlorobenzylamine, N-(p-methylbenzyl) p-methoxybenzylamine
and N-ethyl p-(trifluoromethyl)benzylamine where the products
34 – 41 were isolated in 58 – 86% yield. Tribenzylamine (42)
was obtained in 75% yield upon reaction between dibenzylamine
and benzyl alcohol while N-allyl benzylamine was N-benzylated
to give 43 in 62% yield without any accompanying reduction of
the olefin. In the same way, piperidine and morpholine were N-
benzylated to afford 44 and 45 in 70 and 66% yield whereas the
arylamines N-ethylaniline and tetrahydroquinoline underwent the
benzylation to provide 46 and 47 in only 47 and 35% yield,
respectively. In the latter case, tetrahydroquinoline was
converted into quinoline as a side reaction while N-ethylaniline
did not undergo complete conversion in the transformation. Thus,
the nucleophilicity of the amine influences the alkylation resulting
in lower yields with less nucleophilic substrates. The aliphatic
alcohol hexan-1-ol was also investigated, but a very poor
conversion was observed in the reaction with N-methyl
benzylamine (result not shown).
Entry
X
Additive
t [h]
Yield
[%][b]
1
2
3
4
5
6
7
1
1
1
1
1
1
1
K2CO3 (2 mmol)
48
48
48
48
48
60
60
9
KOH (2 mmol)
32
27
21
37
50
62
NaOH (2 mmol)
KOtBu (2 mmol)
KOH (2 mmol), KOtBu (2 mmol)
KOH (2 mmol), KOtBu (2 mmol)
KOH (2 mmol), KOtBu (2 mmol),
Besides the dehydrogenation of tetrahydroquinoline, the
quinoline framework can also be prepared by a dehydrogenative
coupling between o-aminobenzyl alcohols and secondary
alcohols. The condensation between the corresponding carbonyl
compounds is known as the Friedländer quinoline synthesis, but
often suffers from a moderate yield due to the limited stability of
o-aminobenzaldehydes and undesired aldol condensations.[14]
pyridine (0.2 mmol)
8
1.5
2
KOH (2 mmol), KOtBu (2 mmol),
pyridine (0.2 mmol)
60
60
60
76
75
58
9
KOH (2 mmol), KOtBu (2 mmol),
pyridine (0.2 mmol)
10[c]
1.5
KOH (2 mmol), KOtBu (2 mmol),
pyridine (0.2 mmol)
These side reactions occur to
a lower degree in the
dehydrogenative quinoline synthesis from alcohols, which has
previously been catalyzed by different ruthenium and iridium
complexes (0.025 – 0.5% loading at 110 °C)[15] as well as
various copper, nickel, cobalt and manganese(I) compounds (2
– 10% loading at 90 – 150 °C).[16] All the transformations are
carried out in the absence of a hydrogen acceptor, but 0.5 – 3
equiv. of a base is required such as KOH or KOtBu.[15,16]
11
12
2
2
KOH (2 mmol)
60
60
53
43
KOtBu (2 mmol)
[a] Reaction conditions: o-Aminobenzyl alcohol (1 mmol), 1-phenylethanol (X
mmol), A (0.05 mmol), additive, tetradecane (0.5 mmol, internal standard), 4Å
MS (150 mg), mesitylene (4 mL), reflux. [b] Determined by GC. [c] With 3% of A.
To develop
a new manganese(III)-catalyzed quinoline
The substrate scope of this new procedure was then
investigated with a variety of alcohols and the products were
isolated by column chromatography (Table 6). In this way, 2-
phenylquinoline (48) was obtained in 70% yield while the
synthesis, several exploratory experiments were first performed
with equimolar amounts of o-aminobenzyl alcohol and 1-
phenylethanol in the presence of complex A (Table 5). It was
quickly realized that K2CO3 was not a sufficiently strong base in
This article is protected by copyright. All rights reserved.

10.1002/chem.201900737
Chemistry - A European Journal
FULL PAPER
corresponding 2-(p-tolyl) compound 49 was formed in 90% yield.
1-(p-Methoxyphenyl)- and 1-(p-methylthiophenyl)ethanol reacted
with o-aminobenzyl alcohol to give quinolines 50 and 51,
respectively, in 93 and 82% yield. In the same reaction, 1-(p-
fluorophenyl)-,
1-(o-chlorophenyl)-
and
1-(p-
bromophenyl)ethanol gave the related quinolines 52 – 54 in 78 –
87% yield while a slightly lower yield was obtained of 2-(p-
iodophenyl)quinoline (55) due to partial dehalogenation. Similar
reactions were carried out with 2-amino-4-chlorobenzyl alcohol
which reacted with 1-phenyl-, 1-(p-tolyl)-, 1-(p-methoxyphenyl)-
and 1-(p-fluorophenyl)ethanol to give the corresponding 2-aryl-7-
chloroquinolines 56 – 59 in 62 – 82% yield. Again, a more
modest result was obtained with 1-(p-iodophenyl)ethanol giving
7-chloro-2-(p-iodophenyl)quinoline (60) in 53% yield due to
some dehalogenation. The transformation could also be
performed with cyclohexanol and cycloheptanol to afford tricyclic
compounds 61 and 62, respectively, in 48 and 56% yield. p-
Nitro- and p-cyano-substituted 1-phenylethanol, on the other
hand, failed to produce the desired quinolines, but instead a
complex mixture was obtained (results not shown).
[a] Reaction conditions: o-Aminobenzyl alcohol (1 mmol), secondary alcohol
(1.5 mmol), A (0.05 mmol), pyridine (0.2 mmol), KOH (2 mmol), KOtBu (2
mmol), 4Å MS (150 mg), mesitylene (4 mL), reflux, 60 h (isolated yields).
To gain more information about the mechanism several
experiments were performed to investigate the fate of the
porphyrin complex during the reaction. LCMS analysis of the
mixture after the imination in Table 1, entry 1 showed only
porphyrin complex A and no compounds with a reduced ligand
could be detected, i.e. the corresponding chlorin or
bacteriochlorin complexes. Complex A could be recovered in
70% yield and used again in the same reaction although the
yield then decreased to 58%. When the experiment was carried
out with PhCD2OH and CyNH2, 1 – 4 deuterium atoms were
incorporated into complex A. This must be in the -pyrrolic
positions, which was confirmed by isolating the complex,
subjecting it to demetallation with acid and analyzing the ligand
by NMR spectroscopy. On the other hand, when the imination
with PhCD2OH and CyNH2 was carried out with complex F, no
deuterium incorporation occurred in the porphyrin ring. Thus, a
hydrogen – deuterium scrambling is possible in the -positions
of the porphyrin by using PhCD2OH as the substrate, but no
incorporation occurs in the meso positions.
Table 6. Synthesis of quinolines from o-aminobenzyl alcohols and secondary
alcohols.[a]
The main structural feature of the porphyrin ring is the
aromatic 18 -electron conjugated network. Two pyrrole double
bonds in the -positions are not part of this system and these
two bonds are susceptible to hydrogenation to afford the chlorin
and the bacteriochlorin structure.[17] On the contrary, these
hydroporphyrins are rather easily oxidized back to the parent
porphyrin macrocycle. It seems inconceivable that PhCD2OH
coordinates to manganese and transfers deuterium to the -
positions in a similar way as the reduction of the Schiff bases in
the salen complexes.[6] It also appears highly unlikely that the 18
-electron pathway is disrupted during the process since this
usually requires very strong nucleophiles such as organolithium
reagents.[18] Therefore, the most reasonable explanation for the
deuterium introduction is a partial reduction of the porphyrin ring
by an external reducing agent in the mixture. To test this
scenario complex A was treated with NaBH4 and NaBD4 in two
separate experiments in DMSO at 120 °C. In both cases, the
manganese(III) chlorin complex could be detected by LCMS
analysis of the mixture. However, after workup only the
porphyrin complex was isolated which is presumably due to
aerobic oxidation of the chlorin compound. In the experiment
with NaBD4 some incorporation of deuterium occurred into the
porphyrin framework and again it was confirmed by NMR to be
in the -pyrrolic positions. This experiment shows that a
This article is protected by copyright. All rights reserved.

10.1002/chem.201900737
Chemistry - A European Journal
FULL PAPER
reducing agent is able to react with the porphyrin ring and result
in deuterium incorporation.
Another experiment was designed by using a closed system
with two connected reaction vessels and
a balloon. The
imination with PhCH2OH, CyNH2 and complex A (Table 1, entry
1) was carried out in the first vessel while in the second vessel
(with the balloon attached) DCl in D2O was added to zinc
metal.[19] This liberated D2 in the second vessel, which was
transferred to the first vessel. Subsequent analysis of complex A
after the imination reaction revealed deuterium incorporation into
the complex while no isotope insertion occurred in the alcohol or
the imine. This again indicates that reduction of the porphyrin
takes place by a reducing agent in the mixture (such as a metal
hydride) and not directly by the alcohol.
Thus, manganese(III) porphyrin complexes with axial hydride
and alkoxide ligands are most likely formed as intermediates in
the reaction. To the best of our knowledge such complexes have
not previously been structurally characterized.[20,21] To gain some
experimental evidence for the formation of these complexes, an
experiment was carried out where complex A (1 mmol) was
mixed with BnOH (50 mmol) and NaH (10 mmol) in mesitylene.
After stirring for 30 min at room temperature LCMS analysis of
the mixture showed complete conversion of complex A into the
benzyloxy complex (TPP)MnIII(OBn). The mixture was then
slowly heated to 180 °C with a distillation head attached which
resulted in 3 mmol of a 6:1 mixture of benzyl alcohol and
benzaldehyde distilling off from the reaction. GC analysis of the
residue showed benzyl alcohol and small amounts of
Scheme 2. Proposed mechanism for acceptorless alcohol dehydrogenation.
Conclusions
In summary, manganese(III) porphyrin complexes have been
shown to catalyze the dehydrogenation of alcohols with the
release of hydrogen gas. The transformation has been applied
to the direct synthesis of imines, tertiary amines and quinolines
from alcohols and amines. The reactions are believed to
proceed through a porphyrin manganese(III) alkoxide complex
as the intermediate. Experimental studies have indicated that
this complex can degrade with the formation of the aldehyde and
most likely the corresponding hydride complex. Subsequent
reaction between the latter and the alcohol forms hydrogen gas
and regenerates the alkoxide complex. This constitutes the
benzaldehyde. This test illustrates that
a manganese(III)
alkoxide complex can undergo degradation with the formation of
the aldehyde (and most likely the corresponding manganese(III)
hydride complex).
Based on these observations we believe the most probable
mechanism involves initial formation of the alkoxide complex 63
(Scheme 2). No coordination site is available in this compound
second example of
a
manganese(III) catalyst for the
acceptorless dehydrogenation of alcohols, but contrary to the
first example with a salen complex the porphyrin complexes
react by a different mechanism.
for
a classical -hydride elimination. However, a related
rhodium(III) porphyrin complex was shown to react with
isopropanol to produce acetone and the rhodium(III) hydride
species[22] although the exact mechanism for this reaction is not
known.[23] In addition, (TPP)RhIIICl was shown to mediate the
photocatalytic conversion of secondary alcohols into ketones
and hydrogen gas.[24] Therefore, we contemplate that alkoxide
complex 63 can be converted into the aldehyde and hydride
complex 64 in refluxing mesitylene.[25] Subsequent reaction with
the alcohol regenerates complex 63 and forms hydrogen gas.
This is a different mechanistic pathway than elucidated for the
manganese(III) salen-catalyzed alcohol dehydrogenation in
refluxing toluene (Scheme 1).[6] The higher temperature needed
for the porphyrin-catalyzed protocol is most likely due to the
conversion of the alkoxide complex 63 into the hydride species
64.
Experimental Section
General procedure for imine synthesis: An oven-dried tube (20 mL)
was charged with complex H (12 mg, 0.02 mmol) and pre-activated 4Å
molecular sieves (150 mg). The tube was placed in a Radley carousel on
a hotplate, subjected to vacuum and then filled with N2 (repeated 3 times).
Freshly degassed mesitylene (4 mL) was injected into the mixture
followed by heating to 164 °C under a N2 atmosphere. Then, the alcohol
(1 mmol), the amine (1 mmol) and tetradecane (0.5 mmol, internal
standard) were added to the blue solution and the reaction was refluxed
for 48 h. The mixture was purified directly by silica gel column
chromatography (hexane/Et3N, 98:2) to obtain the desired product.
General procedure for tertiary amine synthesis: Complex A (20 mg,
0.03 mmol), K2CO3 (20 mg, 0.2 mmol) and pre-activated 4Å molecular
sieves (150 mg) were placed in an oven-dried tube (20 mL). The tube
was placed in a Radley carousel on a hotplate, subjected to vacuum and
then filled with N2 (repeated 3 times). Freshly degassed mesitylene (4 mL)
was injected into the mixture followed by heating to 164 °C under a N2
atmosphere. Then, the alcohol (1.2 mmol), the amine (1 mmol) and
tetradecane (0.5 mmol, internal standard) were added to the green
solution and the reaction was refluxed for 48 h. The mixture was purified
This article is protected by copyright. All rights reserved.

10.1002/chem.201900737
Chemistry - A European Journal
FULL PAPER
directly by silica gel column chromatography (hexane/Et3N, 98:2) to
obtain the desired product.
[8]
[9]
a) K. Azizi, R. Madsen, ChemCatChem 2018, 10, 3703‒3708; b) P.
Hermange, A. T. Lindhardt, R. H. Taaning, K. Bjerglund, D. Lupp, T.
Skrydstrup, J. Am. Chem. Soc. 2011, 133, 6061‒6071.
General procedure for quinoline synthesis: Complex A (35 mg, 0.05
mmol), 2-aminobenzyl alcohol (1 mmol), KOH (2 mmol), KOtBu (2 mmol)
and pre-activated 4Å molecular sieves (150 mg) were placed in an oven-
dried tube (20 mL). The tube was placed in a Radley carousel on a
hotplate, subjected to vacuum and then filled with N2 (repeated 3 times).
Freshly degassed mesitylene (4 mL) was injected into the mixture
followed by heating to 164 °C under a N2 atmosphere. Then, the
secondary alcohol (1.5 mmol), pyridine (0.2 mmol) and tetradecane (0.5
mmol, internal standard) were added to the green solution and the
reaction was refluxed for 60 h. The mixture was concentrated and the
residue purified by silica gel column chromatography (CH2Cl2/hexane,
1:1) to obtain the desired product.
a) R. Figliolia, S. Baldino, H. G. Nedden, A. Zanotti-Gerosa, W. Baratta,
Chem. Eur. J. 2017, 23, 14416–14419; b) A. J. A. Watson, A. C.
Maxwell, J. M. J. Williams, J. Org. Chem. 2011, 76, 2328–2331; c) A.
Tillack, D. Hollmann, K. Mevius, D. Michalik, S. Bähn, M. Beller, Eur. J.
Org. Chem. 2008, 4745–4750; d) S. Ganguly, D. M. Roundhill,
Polyhedron 1990, 9, 2517–2526.
[10] a) Q. Zou, C. Wang, J. Smith, D. Xue, J. Xiao, Chem. Eur. J. 2015, 21,
9656–9661; b) L. L. R. Lorentz-Petersen, L. U. Nordstrøm, R. Madsen,
Eur. J. Org. Chem. 2012, 6752–6759; c) N. Andrushko, V. Andrushko,
P. Roose, K. Moonen, A. Börner, ChemCatChem 2010, 2, 640–643; d)
K.-i. Fujita, Y. Enoki, R. Yamaguchi, Tetrahedron 2008, 64, 1943–1954.
[11] a) R. Mamidala, V. Mukundam, K. Dhanunjayarao, K. Venkatasubbaiah,
Tetrahedron 2017, 73, 2225–2233; b) T. T. Dang, B. Ramalingam, S. P.
Shan, A. M. Seayad, ACS Catal. 2013, 3, 2536–2540.
[12] a) T. J. Brown, M. Cumbes, L. J. Diorazio, G. J. Clarkson, M. Wills, J.
Org. Chem. 2017, 82, 10489–10503; b) T. Yan, B. L. Feringa, K. Barta,
ACS Catal. 2016, 6, 381–388.
Acknowledgements
We thank the Torkil Holm Foundation for financial support.
[13] M. Mastalir, E. Pittenauer, G. Allmaier, K. Kirchner, J. Am. Chem. Soc.
2017, 139, 8812–8815.
Keywords: alcohols • dehydrogenation • homogeneous
catalysis • manganese • synthetic methods
[14] J. Marco-Contelles, E. Pérez-Mayoral, A. Samadi, M. do Carmo
Carreiras, E. Soriano, Chem. Rev. 2009, 109, 2652–2671.
[15] a) B. Pan, B. Liu, E. Yue, Q. Liu, X. Yang, Z. Wang, W.-H. Sun, ACS
Catal. 2016, 6, 1247–1253; b) S. Ruch, T. Irrgang, R. Kempe, Chem.
Eur. J. 2014, 20, 13279–13285; c) D. Srimani, Y. Ben-David, D. Milstein,
Chem. Commun. 2013, 49, 6632–6634.
[1]
[2]
a) A. Corma, J. Navas, M. J. Sabater, Chem. Rev. 2018, 118, 1410–
1459; b) C. Gunanathan, D. Milstein, Science 2013, 341, 1229712; c) S.
Bähn, S. Imm, L. Neubert, M. Zhang, H. Neumann, M. Beller,
ChemCatChem 2011, 3, 1853–1864; d) Y. Obora, Y. Ishii, Synlett 2011,
30–51; e) G. E. Dobereiner, R. H. Crabtree, Chem. Rev. 2010, 110,
681–703.
[16] a) K. Das, A. Mondal, D. Srimani, Chem. Commun. 2018, 54, 10582–
10585; b) S. Shee, K. Ganguli, K. Jana, S. Kundu, Chem. Commun.
2018, 54, 6883–6886; c) S. Das, D. Maiti, S. De Sarkar, J. Org. Chem.
2018, 83, 2309–2316; d) D.-W. Tan, H.-X. Li, D.-L. Zhu, H.-Y. Li, D. J.
Young, J.-L. Yao, J.-P. Lang, Org. Lett. 2018, 20, 608–611; e) S. Parua,
R. Sikari, S. Sinha, S. Das, G. Chakraborty, N. D. Paul, Org. Biomol.
Chem. 2018, 16, 274–284; f) S. P. Midya, V. G. Landge, M. K. Sahoo, J.
Rana, E. Balaraman, Chem. Commun. 2018, 54, 90–93; g) M. Mastalir,
M. Glatz, E. Pittenauer, G. Allmaier, K. Kirchner, J. Am. Chem. Soc.
2016, 138, 15543–15546.
a) G. A. Filonenko, R. van Putten, E. J. M. Hensen, E. A. Pidko, Chem.
Soc. Rev. 2018, 47, 1459–1483; b) F. Kallmeier, R. Kempe, Angew.
Chem. Int. Ed. 2018, 57, 46–60; Angew. Chem. 2018, 130, 48–63; c) B.
Maji, M. K. Barman, Synthesis 2017, 49, 3377–3393; d) M. Garbe, K.
Junge, M. Beller, Eur. J. Org. Chem. 2017, 4344–4362; e) E.
Balaraman, A. Nandakumar, G. Jaiswal, M. K. Sahoo, Catal. Sci.
Technol. 2017, 7, 3177–3195.
[17] M. Taniguchi, J. S. Lindsey, Chem. Rev. 2017, 117, 344–535.
[18] a) S. Hiroto, Y. Miyake, H. Shinokubo, Chem. Rev. 2017, 117, 2910–
3043; b) M. O. Senge, Acc. Chem. Res. 2005, 38, 733–743.
[3]
[4]
A. Mukherjee, A. Nerush, G. Leitus, L. J. W. Shimon, Y. Ben David, N.
A. E. Jalapa, D. Milstein, J. Am. Chem. Soc. 2016, 138, 4298–4301.
For recent examples, see: a) U. K. Das, S. Chakraborty, Y. Diskin-
Posner, D. Milstein, Angew. Chem. Int. Ed. 2018, 57, 13444–13448;
Angew. Chem. 2018, 130, 13632–13636; b) T. Liu, L. Wang, K. Wu, Z.
Yu, ACS Catal. 2018, 8, 7201–7207; c) G. Zhang, T. Irrgang, T. Dietel,
F. Kallmeier, R. Kempe, Angew. Chem. Int. Ed. 2018, 57, 9131–9135;
Angew. Chem. 2018, 130, 9269–9273; d) M. K. Barman, S. Waiba, B.
Maji, Angew. Chem. Int. Ed. 2018, 57, 9126–9130; Angew. Chem. 2018,
130, 9264–9268.
[19] A. Modvig, T. L. Andersen, R. H. Taaning, A. T. Lindhardt, T.
Skrydstrup, J. Org. Chem. 2014, 79, 5861–5868.
[20] For X-ray structures of two alkoxide complexes of iron(III) porphyrins,
see: a) L. Cheng, N. Xu, D. R. Powell, G. B. Richter-Addo, Acta Cryst.
2010, E66, m1448; b) K. Hatano, T. Uno, Bull. Chem. Soc. Jpn. 1990,
63, 1825–1827.
[21] An iron(III) porphyrin complex has been shown to catalyze the transfer
hydrogenation of 2-methoxyacetophenone in isopropanol, but the
mechanism has not been investigated, see: S. Enthaler, B. Spilker, G.
Erre, K. Junge, M. K. Tse, M. Beller, Tetrahedron 2008, 64, 3867–3876.
[22] L. Liu, M. Yu, B. B. Wayland, X. Fu, Chem. Commun. 2010, 46, 6353–
6355.
[5]
For recent examples, see: a) A. Kaithal, M. Hölscher, W. Leitner,
Angew. Chem. Int. Ed. 2018, 57, 13449–13453; Angew. Chem. 2018,
130, 13637–13641; b) Y.-Q. Zou, S. Chakraborty, A. Nerush, D. Oren,
Y. Diskin-Posner, Y. Ben-David, D. Milstein, ACS Catal. 2018, 8, 8014–
8019; c) M. Glatz, B. Stöger, D. Himmelbauer, L. F. Veiros, K. Kirchner,
ACS Catal. 2018, 8, 4009–4016; d) D. Wei, A. Bruneau-Voisine, T.
Chauvin, V. Dorcet, T. Roisnel, D. A. Valyaev, N. Lugan, J.-B. Sortais,
Adv. Synth. Catal. 2018, 360, 676–681.
[23] For a discussion of possible mechanistic paths for this transformation,
see: J. P. Collman, R. Boulatov, Inorg. Chem. 2001, 40, 560–563.
[24] X. Li, S. Shinoda, Y. Saito, J. Mol. Catal. 1989, 49, 113–119.
[25] A dissociative -hydride abstraction mechanism has been proposed to
explain formal -hydride eliminations in alkoxide complexes where
open cis coordination sites or detachable ligands are not available, see:
a) N. A. Smythe, K. A. Grice, B. S. Williams, K. I. Goldberg,
Organometallics 2009, 28, 277–288; b) C. M. Fafard, O. V. Ozerov,
Inorg. Chim. Acta 2007, 360, 286–292; c) O. Blum, D. Milstein, J.
Organomet. Chem. 2000, 593–594, 479–484.
[6]
[7]
S. V. Samuelsen, C. Santilli, M. S. G. Ahlquist, R. Madsen, Chem. Sci.
2019, 10, 1150–1157.
a) M. J. F. Calvete, M. Piñeiro, L. D. Dias, M. M. Pereira,
ChemCatChem 2018, 10, 3615–3635; b) W. Liu, J. T. Groves, Acc.
Chem. Res. 2015, 48, 1727–1735; c) M. Costas, Coord. Chem. Rev.
2011, 255, 2912–2932; d) C.-M. Che, V. K.-Y. Lo, C.-Y. Zhou, J.-S.
Huang, Chem. Soc. Rev. 2011, 40, 1950–1975; e) H. Lu, X. P. Zhang,
Chem. Soc. Rev. 2011, 40, 1899–1909.
This article is protected by copyright. All rights reserved.

10.1002/chem.201900737
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 1:
FULL PAPER
Text for Table of Contents
Author(s), Corresponding Author(s)*
Page No. – Page No.
Title
((Insert TOC Graphic here: max.
width: 5.5 cm; max. height: 5.0 cm))
Layout 2:
FULL PAPER
K. Azizi, S. Akrami, R. Madsen*
Page No. – Page No.
Manganese(III) Porphyrin-Catalyzed
Dehydrogenation of Alcohols to form
Imines, Tertiary Amines and
Quinolines
Three times manganese(III): Manganese(III) porphyrin complexes catalyze the
dehydrogenative coupling between alcohol and amines to generate imines, tertiary
amines and quinolines. Manganese(III) alkoxide and hydride complexes are
believed to be the catalytically active species responsible for the hydrogen gas
release.
This article is protected by copyright. All rights reserved.