Unveiling the Importance of π-Stacking in Borrowing-Hydrogen Processes Catalysed by Iridium Complexes with Pyrene Tags

Article abstract of DOI:10.1002/chem.201502948

This work describes the preparation of a series of pyrene-tagged N-heterocyclic carbene complexes of iridium, and their use in two benchmark borrowing hydrogen reactions: the reduction of ketones by transfer hydrogenation and the β-alkylation of secondary alcohols with primary alcohols. The detailed study of these homogeneously catalysed reactions reveals several important implications regarding the strong influence of the pyrene tags in the catalysts. First, the catalytic activity is partially inhibited by addition of an external amount of pyrene, but only when pyrene-tagged catalysts and aromatic substrates are used. Second, the rate order of the reaction is highly dependent on the nature of the substrates and the ligand. When pyrene-tagged catalysts and aromatic substrates are used, the reaction follows a zero-order dependence on the concentration of the substrate. All other combinations afford a second-order rate in the substrates. And third, the presence or absence of the pyrene functionality in the catalyst also influences the reaction order with respect to the concentration of the catalyst. Pyrene-containing catalysts display a fractional rate order of below 1. Finally, two pyrene-tagged catalysts were supported onto reduced-graphene oxide (rGO), and used as heterogeneous catalysts. While the dimetallic catalyst was effectively recycled 12 times, the monometallic catalyst maintained its activity for only three runs.

Full text of DOI:10.1002/chem.201502948

DOI: 10.1002/chem.201502948
Full Paper
&
Organometallic Chemistry |Hot Paper|
Unveiling the Importance of p-Stacking in Borrowing-Hydrogen
Processes Catalysed by Iridium Complexes with Pyrene Tags
Sheila Ruiz-Botella and Eduardo Peris*[a]
Abstract: This work describes the preparation of a series of
pyrene-tagged N-heterocyclic carbene complexes of iridium,
and their use in two benchmark borrowing hydrogen reac-
tions: the reduction of ketones by transfer hydrogenation
and the b-alkylation of secondary alcohols with primary alco-
hols. The detailed study of these homogeneously catalysed
reactions reveals several important implications regarding
the strong influence of the pyrene tags in the catalysts. First,
the catalytic activity is partially inhibited by addition of an
external amount of pyrene, but only when pyrene-tagged
catalysts and aromatic substrates are used. Second, the rate
order of the reaction is highly dependent on the nature of
the substrates and the ligand. When pyrene-tagged catalysts
and aromatic substrates are used, the reaction follows
a zero-order dependence on the concentration of the sub-
strate. All other combinations afford a second-order rate in
the substrates. And third, the presence or absence of the
pyrene functionality in the catalyst also influences the reac-
tion order with respect to the concentration of the catalyst.
Pyrene-containing catalysts display a fractional rate order of
below 1. Finally, two pyrene-tagged catalysts were support-
ed onto reduced-graphene oxide (rGO), and used as hetero-
geneous catalysts. While the dimetallic catalyst was effective-
ly recycled 12 times, the monometallic catalyst maintained
its activity for only three runs.
Introduction
Non-covalent attractions influencing the catalytic outcomes
are typically classified as 1) intra- or interligandligand interac-
tions, 2) ligand–additive interactions, or 3) ligand–substrate in-
teractions,[1,2b] but regardless the type of interaction that
occurs, most of the times the effects on the catalytic output
refers to the selectivity of the process, rather than to the modi-
fication of the reaction rates of processes not needing selectiv-
ity improvements. For example, while it is well recognized that
p-stacking may introduce important benefits in the asymmetric
induction of some catalytic processes,[2b,3] its role in non-asym-
metric processes has been clearly underestimated.
One of the main challenges of catalyst design is to tailor the
properties of classical catalysts by facilitating reversible non-co-
valent interactions, such as hydrogen bonds, or p-stacking.
The resulting catalysts may then possess properties that differ
from those shown by analogues lacking the assembling prop-
erties.[1] Such interactions, which influence but do not form
part of the basic catalytic reaction, constitute the basis of
supramolecular catalysis.[1,2] Although the preparation of
a supramolecular catalyst must incorporate a large degree of
design, non-covalent forces are difficult to predict, and there-
fore the ‘a priori’ chances to choose a catalyst–substrate com-
bination with suitable self-assembling properties is still very
low, thus most of the supramolecular effects in catalysis are
recognised ’post-factum’. Once the ‘supramolecular catalyst’
has been prepared, the challenge is to determine whether the
catalytic activity of the catalyst is influenced by non-covalent
assembling properties, for which well-chosen control experi-
ments are required to determine both the nature of the non-
covalent interaction and their direct involvement in the rate-
and selectivity-determining steps of the catalytic reaction.
We have recently been interested in studying the effects
produced by ligand–substrate p-stacking effects in the output
of several homogeneously metal-catalysed reactions.[4] Initially,
we focused our attention on finding reactions for which the
addition of p-stacking additives should have an influence on
the rates of formation of the products,[4a,b,e] while lately we
were more interested in trying to rationalise the effects of
non-covalent interactions, by identifying and quantifying the
nature of the p-stacking interaction produced between the
ligand and the substrates and between the ligand and the p-
stacking additives.[4c,d] All of our catalysts had in common the
presence of N-heterocyclic carbene ligands (NHCs) decorated
with p-delocalised polyaromatic systems. In a parallel study,
we synthesised a series of ruthenium and palladium catalysts
with NHC ligands containing a pyrene tag,[5] which were non-
covalently immobilised onto a surface of reduced graphene
oxide (rGO) and gave heterogenised catalysts with a high
degree of recyclability. All these studies illustrated the great
potential of p-stacking interactions in both, homogeneously
[a] S. Ruiz-Botella, Prof. Dr. E. Peris
Departamento de Química Inorgµnica y Orgµnica
Universitat Jaume I
Avda. Sos Baynat, 12071 Castellón (Spain)
Fax: (+34)964387522
E-mail: eperis@uji.es
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201502948.
Chem. Eur. J. 2015, 21, 15263 – 15271
15263
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
K2CO3 and KBr allowed the formation of complexes 5–8 in
moderate to high yields (46–70%).
All complexes 5–8 gave satisfactory elemental analysis and
were characterised by NMR spectroscopy and Mass spectrome-
try. In the case of the monometallic complexes 5 and 6, the
formation of the NHC complexes were easily confirmed by
13C NMR spectroscopy, which revealed the diagnostic signal
due to the carbene carbon atoms, coincidently at 180.9 ppm.
In the case of the dimetallic complexes 7 and 8, the twofold
symmetry of the complexes is evidenced by the presence of
a single resonance due to the four equivalent protons of the
bridging phenylene ring in 7, and the presence of two dou-
blets for the two phenylene rings in 8, as seen from the
1H NMR spectra (please see the Supporting Information). The
twofold symmetry is also confirmed by the 13C NMR spectra,
which shows a single resonance for the carbene carbons at
180.8 and 180.7 ppm, for 7 and 8, respectively. These coinci-
dental chemical shifts shown for all complexes 5–8, may be in-
terpreted as a consequence of the similar electron-donating
properties provided by the NHC ligands present in their coor-
dination spheres.[6] The nuclearity of the complexes is con-
firmed by ESI-mass spectrometry, which for the monometallic
complexes shows representative peaks at m/z values of 473.3
and 671.2 (assigned to [5ÀBr]+ and [6ÀBr]+, respectively), and
for the dimetallic complexes at 1347.0 and 1464.0 (assigned to
[7ÀBr]+ and [8ÀBr+MeCN]+, respectively). By reaction of 2
with [{RhCl(COD)}2] in the presence of K2CO3 and KBr, following
the same reaction procedure depicted in Scheme 2, we ob-
tained complex 9, which is the rhodium analogue of complex
6. The 13C NMR spectrum of 9 showed the distinctive doublet
due to the carbene carbon at 183.6 ppm (1JRh,C =50 Hz).
Scheme 1. NHC-based ligands used in this work.
and heterogeneously catalysed processes, when using ligands
decorated with extended polyaromatic systems. Based on
these previous works, we decided to synthesise a new set of
catalysts, for which we could merge both the effects produced
by p-stacking interactions with additives in influencing the cat-
alytic activity of homogeneously catalysed reactions and the
non-covalent immobilisation of the catalysts onto a rGO sur-
face to afford highly recyclable heterogeneous catalysts. For
the development of this study, we decided to use the ligands
depicted in Scheme 1, which should provide an excellent basis
for comparison and constitute a good set of ligands for estab-
lishing all the control experiments required. These ligands
were coordinated to IrI, and the resulting metal complexes
were tested in typical borrowing hydrogen catalysed process-
es, such as the reduction of ketones by transfer hydrogenation
and the b-alkylation of secondary alcohols with primary alco-
hols. The heterogenisation of two selected catalysts onto rGO
is also described, together with some kinetic analysis of the
catalytic reactions under study. From the study of the kinetics
of the processes important implications regarding the role of
the catalysts and the presence of the polyaromatic tag are
concluded.
Catalytic studies
Results and Discussion
The catalytic activity of complexes 5–9 was tested in the re-
duction of ketones by transfer hydrogenation and in the b-al-
kylation of secondary alcohols with primary alcohols, two typi-
cal borrowing hydrogen processes for which rhodium and iridi-
um have proven to be very effective catalysts.[7] We were par-
ticularly interested in determining the influence of the addition
of pyrene on the catalytic output of the reactions,
and if it has any type of relation with the nature of
the catalyst and/or substrates used. For the transfer
hydrogenation process, we decided to study the re-
duction of two benchmark ketones, namely aceto-
phenone and cyclohexanone. The reactions were car-
ried out using a 0.5 mol% load of catalyst (based on
the amount of metal) at reflux (iPrOH; 828C) for one
hour in the presence of KOH (10 mol%).. The results
are shown in Table 1. Under these reaction condi-
tions, the background activity of the base was found
to be negligible (<1%).[8] One of the first conclusions
that can be extracted from these results is that all iri-
dium complexes 5–8 are significantly more active
than the rhodium complex 9 (Table 1, entries 1–4
versus entry 5). Thus, complex 9 was not used for fur-
ther catalytic studies. From the comparison of the ac-
Synthesis and characterization of complexes
The imidazolium and bisazolium salts 1–4[4e] were coordinated
to iridium by the method depicted in Scheme 2. The reaction
of either of the azolium salts with [{IrCl(COD)}2] (cod = cyclooc-
tadiene) in a mixture of THF/DMF at 758C in the presence of
Scheme 2. Preparation of compounds.
Chem. Eur. J. 2015, 21, 15263 – 15271
www.chemeurj.org
15264
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 1. Catalysts 5–9 in the reduction of acetophenone and cyclohexa-
none by transfer hydrogenation in iPrOH.[a]
Table 2. Catalysts 5–8 used in the b-alkylation of 1-phenylethanol with pri-
mary alcohols.[a]
Entry
Cat.
Ketone
Pyrene [mol]
Yield [%]
Entry
Cat.
R group
Conv. [%]
Yield [%]
1
2
3
4
5
6
7
8
5
6
7
8
9
5
6
7
8
5
6
7
8
5
6
7
8
cyclohexanone
cyclohexanone
cyclohexanone
cyclohexanone
cyclohexanone
cyclohexanone
cyclohexanone
cyclohexanone
cyclohexanone
acetophenone
acetophenone
acetophenone
acetophenone
acetophenone
acetophenone
acetophenone
acetophenone
–
–
–
–
80 (Æ1)
91 (Æ1)
91 (Æ1)
86 (Æ1)
16 (Æ1)
81 (Æ1)
89 (Æ1)
89 (Æ1)
87 (Æ1)
79 (Æ1)
30 (Æ1)
94 (Æ1)
94 (Æ1)
81 (Æ1)
16 (Æ1)
21 (Æ1)
75 (Æ1)
1
2
3
4[b]
5
6
7
8
9
10
11
12
13
14
15
16
17
5
6
7
7
8
5
6
7
8
5
6
7
8
5
6
7
8
Ph
Ph
Ph
Ph
43 (Æ1)
50 (Æ1)
99 (Æ1)
99 (Æ1)
51 (Æ1)
53 (Æ1)
43 (Æ1)
92 (Æ1)
40 (Æ1)
31 (Æ1)
32 (Æ1)
90 (Æ1)
49 (Æ1)
45 (Æ1)
43 (Æ1)
50 (Æ1)
40 (Æ1)
41 (Æ1)
50 (Æ1)
99 (Æ1)
99 (Æ1)
50 (Æ1)
52 (Æ1)
43 (Æ1)
92 (Æ1)
40 (Æ1)
31 (Æ1)
30 (Æ1)
90 (Æ1)
47 (Æ1)
40 (Æ1)
40 (Æ1)
47 (Æ1)
38 (Æ1)
–
Ph
0.05
0.05
0.05
0.05
–
–
–
–
0.05
0.05
0.05
0.05
3-Cl(C6H5)
3-Cl(C6H5)
3-Cl(C6H5)
3-Cl(C6H5)
4-Cl(C6H5)
4-Cl(C6H5)
4-Cl(C6H5)
4-Cl(C6H5)
nPr
9
10
11
12
13
14
15
16
17
nPr
nPr
nPr
[a] General reaction condition: ketone (0.5 mmol), isopropanol (2 mL),
KOH (0.05 mol) and a catalyst stirred for 1 h at 828C. Yields determined
by GC, averaging a minimum of two runs, using anisole as internal stan-
dard.
[a] General reaction condition: 1-phenylethanol (1 mmol), primary alcohol
(1 mmol), KOH (1 mmol) and a catalyst in toluene (0.5 mL) stirred at 1008C.
Yields determined by GC, averaging a minimum of two runs, using anisole
as internal standard. [b] Reaction carried out in the presence of a drop of
Hg.
tivities displayed by the iridium complexes, several interesting
consequences may be inferred. First, all iridium complexes are
very active in the reduction of cyclohexanone, but the activi-
ties shown by the dimetallic complexes 7 and 8 are significant-
ly higher than those shown by the monometallic complexes 5
and 6 in the reduction of acetophenone. Second, the addition
of pyrene has significantly different effects on the yields de-
pending on the nature of the ketone used. For instance, the
addition of pyrene (10 mol%) to the reaction in which aceto-
phenone is being reduced significantly decreases the activity
of the catalysts, but only of those bearing the pyrene tag
(Table 1, entries 11–13 versus 15–17), whereas the outcome of
the reaction is not affected in the case of the reduction of cy-
clohexanone (Table 1, entries 1–4 versus 6–9). Interestingly the
effect of the addition of pyrene to the reduction of acetophe-
none is negligible in the presence of catalyst 5 (Table 1, com-
pare entries 10 and 14). All these results are in accordance
with our previous results, which indicate that the activity of
catalysts bearing rigid polyaromatic ligands is influenced by
the p-stacking with aromatic substrates,[4a,b,e] but in this case
the effect becomes more evident, since we were able to direct-
ly compare the effects produced over one aliphatic and one ar-
omatic substrate.
nylethanol with a series of benzyl alcohols (benzyl alcohol, 3-
chlorobenzyl alcohol and 4-chlorobenzyl alcohol) and n-buta-
nol (nBuOH). The reactions were carried out using a catalyst
loading of 0.5 mol% (based on the concentration of Ir) in tolu-
ene (0.5 mL), using an equimolecular amount of both alcohols
(1 mmol) and KOH (1 mmol), at 1008C over 8 h. The results are
summarised in Table 2.
From the results shown in Table 2, it can be seen that all cat-
alysts 5–8 are very selective in the production of the final alco-
hol. Only in a few cases, traces of the a,b-unsaturated ketone
byproducts were detected (the differences between conver-
sions and yields are due to the formation of the ketone by-
products). The phenylene-bridged diiridium complex 7 is the
most active catalyst in all the reactions tested, affording almost
quantitative conversion when the two aromatic alcohols are
used. In order to determine whether Ir nanoparticles may be
formed during the process, and that the reaction may be het-
erogeneously catalysed, we also performed the reaction in the
presence of an excess Hg to selectively poison heterogeneous
forms of metallic iridium.[9] As can be seen from the compari-
son of the results (shown in Table 2;entries 3 and 4), catalyst 7
displayed virtually unperturbed activity in the presence of Hg,
strongly suggesting a genuine solution-phase catalysis.
The borrowing hydrogen methodology may be used for the
activation of two alcohols in one reaction sequence, leading to
an overall CÀC bond formation process that is often referred
as to the b-alkylation of secondary alcohols with primary alco-
hols. We were interested in evaluating the catalytic properties
of our complexes for such reactions, and in determining
whether the p-stacking abilities of the catalysts may influence
the final outputs. For this study, we decided to couple 1-phe-
In order to shed some light on the reaction mechanism of
the process, especially with regard to the interaction of the
catalysts with the substrate in the initial stages of the reaction,
we decided to perform some kinetic studies, taking complex 7
as model catalyst. We first studied the reaction profile of the b-
alkylation of 1-phenylethanol with benzyl alcohol under the re-
action conditions shown in Table 2. The time-dependent reac-
Chem. Eur. J. 2015, 21, 15263 – 15271
www.chemeurj.org
15265
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 1. Time-dependent reaction profile for the reaction between 1-phe-
nylethanol and benzyl alcohol. Reaction conditions: 1-phenylethanol
(1 mmol), benzyl alcohol (1 mmol), KOH (1 mmol) in toluene (0.5 mL) at
1008C. Yields determined by GC, using anisole as internal standard. Product:
1-phenylhexanol. The concentration of the catalyst (7) is shown for each
plot. Straight lines are drawn through the points for clarity.
tion profile of the reaction (Figure 1) followed a pseudo-zero-
order rate in the substrate, for all the reactions carried out
with catalyst loadings ranging between 0.01–5 mol%. This
type of situation, which is typically found in enzymatic cataly-
sis, normally arises when only a small portion of molecules of
the substrate are in a location in which they are able to react,
and this fraction is continually replenished from the larger
pool. In our case, we believe that this situation may arise from
the non-covalent interaction between the aromatic substrate
and the pyrene tag of the catalyst, thus saturating the catalyst,
and affording a reaction rate which is non-dependent on the
concentration of the substrate.
Figure 2. The plot at the top represents the time-dependent reaction profile
for the reaction between 1-phenylethanol and benzyl alcohol. Reaction con-
ditions: catalyst 5 (0.05 mmol), 1-phenylethanol (1 mmol), benzyl alcohol
(1 mmol), KOH (1 mmol) in toluene (0.5 mL) at 1008C. The plot at the
bottom represents 1/[1-phenylethanol] versus time.
In order to prove that the zero-order dependence with re-
spect to the substrate is due to the nature p–p interaction be-
tween the aromatic substrates and the pyrene tag of the cata-
lyst, we decided to perform two different experiments in
which this type of interaction is suppressed. First, we moni-
tored a reaction using the monometallic catalyst 5. This cata-
lyst does not bear a pyrene tag, and therefore should not
suffer from any type of non-covalent interaction with the aro-
matic substrates prior to their activation. As can be seen from
the time-dependent reaction profile shown in Figure 2 for the
reaction of 1-phenylethanol with benzyl alcohol, the reaction
follows a second-order reaction rate, as indicated by the fitting
to a straight line when plotting 1/[1-phenylethanol] versus
time. This result is consistent with the presence of two sub-
strates that react in a 1:1 stoichiometry.
Figure 3. The plot at the top represents the time dependent reaction profile
for the reaction between 1-phenylethanol and n-butanol. Reaction condi-
tions: catalyst 7 (0.05 mmol), 1-phenylethanol (1 mmol), nbutanol (1 mmol),
KOH (1 mmol) in toluene (0.5 mL) at 1008C. The plot at the bottom repre-
sents the 1/[1-phenylethanol] versus time.
Then we monitored the reaction of 1-phenylethanol with n-
butanol, using the pyrene-containing dimetallic catalyst 7. As
can be seen from the reaction profile (Figure 3), the reaction
follows a second-order reaction with respect to the substrates,
a clear indication that both alcohols need to be aromatic in
order have any influence in the kinetics of the reaction.
Further highly valuable information was obtained from the
study of the dependence of the reaction rates on the concen-
tration of the catalysts. From the results shown in Figure 1, we
obtained the corresponding observed rate constants at each
catalyst concentration ([7]=0.01–5 mol%). The dependence of
the observed rate constants on [7], followed a 0.75Æ0.06
catalyst loadings in the range of 0.05–5 mol% (see the Sup-
porting Information for graphics). In this case the plot of log(-
kobs) versus log[6], suggests a half-order dependence on the
concentration of the iridium complex (slope=0.555Æ0.008).
This half-order dependence is strongly suggestive of a mono-
mer–dimer equilibrium,[10] with the monomer being the active
catalytic species. The half-order dependence arises from the
determination of the concentration of the monomeric species,
which may be expressed as [monomer]=([dimer]/Keq)1/2, in
order, as obtained from the plot of the values of log(kobs
)
against log[7], which is shown in Figure 4. Similar kinetic stud-
ies were performed using the monometallic catalyst 6, with
Chem. Eur. J. 2015, 21, 15263 – 15271
www.chemeurj.org
15266
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Immobilization of catalysts and recycling experiments
Once we proved that non-covalent interactions between the
substrate and the ligand of the catalyst have an important in-
fluence in the kinetics of these two benchmark homogeneous-
ly catalysed borrowing hydrogen processes, we decided to see
if we could immobilize some of our pyrene-containing com-
plexes onto reduced graphene oxide (rGO), and study the recy-
clability properties of the resulting solids. Due to the inherent
capability of pyrene to afford non-covalent interactions with
graphitic surfaces,[11] some pyrene-containing metal complexes
have been supported onto such surfaces,[12] and have been
used in catalysis demonstrating interesting recyclability proper-
ties.[13] We also recently contributed to the field, by reporting
a series of ruthenium and palladium catalysts with NHC ligands
containing a pyrene tag[5] that were non-covalently immobi-
lized onto rGO, and proved to afford heterogenised catalysts
with a high degree of recyclability. The immobilisation of the
catalysts described in this work may be of special interest, be-
cause we are unaware of any heterogenised catalysts used for
the b-alkylation of alcohols with efficient recyclability proper-
ties. For our study, we decided to use catalysts 6 and 7, be-
cause the comparison of their activities should provide us in-
teresting information about the influence of the mono- or di-
metallic nature of the catalyst (complex 7 may be regarded as
exactly the combination of two molecules of 6), and about the
presence of one or two pyrene tags for the immobilization
onto the solid surface.
Figure 4. Plots of log(kobs) against log[cat]. The values of the slopes with
errors are: 1.08Æ0.07 (5), 0.555Æ0.008 (6) and 0.75Æ0.06 (7).
which Keq is the self-association constant. In our case, the self-
association may arise from the p–p stacking of the pyrene
functionalities present in complex 6, which should form non-
active and non-covalently attached dimers. In the case of the
dimetallic complex 7, for which the reaction rate order in the
catalyst is 0.75, the same self-association may be occurring,
but, in this case the complex being dimetallic, it may yield
a (tetrametallic) dimer in which one of the metals may retain
its activity, while the activity of the other is inhibited. This
would lead to an order of reaction between 0.5 and 1. We pre-
viously demonstrated the self-association capabilities of com-
plexes with the mono- and bidentate pyrene-containing li-
gands present in 6 and 7,[4e,5b] therefore the self-association
properties of their complexes seem rather plausible in solution.
In order to provide experimental support to this hypothesis,
we decided to perform a similar study using catalyst 5. Com-
plex 5 is quasi-identical to complex 6, except for the fact that
it does not have a pyrene group and, therefore, any possibili-
ties of self-association by non-covalent interactions are fully
discarded. This implies that all aspects regarding the mechanis-
tic behaviour of 5 and 6 should be identical, except for those
strictly coming from reasons involving the presence of the
pyrene tag in 6. As was done for catalysts 6 and 7, the reaction
profiles of the reaction of the b-alkylation of 1-phenylethanol
with benzyl alcohol were studied, using concentrations of 5
ranging between 0.01–5 mol%, although in this case, as previ-
ously mentioned, the reaction profiles consistently followed
a second-order dependence on the concentration of the sub-
strates. In contrast, as shown for the plot shown in Figure 2,
the reaction cleanly followed a first-order dependence on 5,
thus suggesting that the active catalytic species is monometal-
lic. This control experiment strongly supports that pyrene has
an important contribution to the reaction mechanism, most
likely due to the formation of dimers with inhibited activity in
the cases of 6 and 7. The inhibition of the activity of 6 and 7
produced upon self-association through the pyrene tags, once
more supports the idea that the interaction between pyrene
and the substrate has an important role in the overall catalytic
reaction.
Complexes 6 and 7 were supported onto reduced graphene
oxide (rGO) by mixing the molecular complexes and the rGO
material in dichloromethane in an ultrasound bath for 30 min
and stirring for 10 h (Scheme 3). This procedure afforded the
solids 6-rGO and 7-rGO. The first evidence that molecular com-
plexes 6 and 7 were anchored over the surface of rGO arises
from the disappearance of the colour of the solution. The
exact iridium content on 6-rGO and 7-rGO was determined by
digestion of the samples in hot HCl/HNO3 followed by ICP-MS
(inductively coupled plasma mass spectroscopy) analysis. The
results obtained accounted for a 2.8 wt% of 6 in 6-rGO, and
6.8 wt% of 7 in 7-rGO. The presence of the metal complexes
on the surface of the solid was also confirmed by UV/Vis meas-
urements (see the Supporting Information). The spectra of the
hybrid materials consist of the overlap of the molecular com-
plexes (6 or 7) with rGO, as seen by the pyrene UV/Vis charac-
teristic signals.[13b,14] The elemental mapping by energy-disper-
sive X-ray spectroscopic analysis (EDS) performed by means of
high-resolution transmission electron microscopy (HRTEM) of
6-rGO and 7-rGO, confirmed the homogeneous distribution of
iridium in the two hybrid materials (Figure 5 shows the STEM
and HRTEM images obtained for 6-rGO; Images for 7-rGO can
be found in the Supporting Information).
For the study of the recyclability properties of 6-rGO and 7-
rGO, we monitored the reaction between 1-phenylethanol
with benzyl alcohol. Each individual run was carried out using
a catalyst loading of 0.5 mol% (based on the amount of cata-
lyst determined by the ICP-MS analysis) in toluene (1.5 mL) at
1008C, using 1 mmol of each of the alcohols and KOH
Chem. Eur. J. 2015, 21, 15263 – 15271
www.chemeurj.org
15267
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ture dependent,[13a,b] and leach-
ing is significantly reduced when
the metal complex is decorated
with more than one pyrene
tag.[12b] These reports are in ac-
cordance with our observation
that the complex with one
pyrene tag deactivates before
the complex with two pyrene
tags, most likely due to a faster
desorption. The ICP-MS analysis
of the heterogenised catalysts 6-
rGO and 7-rGO after the recy-
cling
experiments,
revealed
a 1.5 wt% of 6 in 6-rGO (after 7
runs) and a 6.3 wt% content of
Scheme 3. Immobilisation of catalysts 6 and 7.
7 in 7-rGO (after 12 runs), thus
indicating that catalyst leaching
is very important for the case of
the material containing the complex with only one pyrene
functionality, while being negligible for the material with the
complex with two pyrene groups.
A significant feature of the results shown in Figure 6 is that
the activity shown in the first run is significantly lower than
the activity shown for the rest of the cycles. This situation is
common for 7-rGO and for the first four cycles of 6-rGO, and
may be explained as a consequence of the need of activation
of the two catalysts along the reaction course.
Aiming to gain a more detailed insight onto the recycling
properties of 7-rGO, we monitored the first three cycles of the
reaction in order to draw the corresponding time-dependent
reaction profiles (Figure 7). As can be seen from the compari-
son of the three reaction profiles, the first run shows lower re-
action rate during the first eight hours, but then accelerates
and equals the rate of runs 2 and 3, this suggests that the cat-
alyst is activated during the first hours of reaction. The kinetics
of runs 2 and 3 are superimposable, which indicates that the
deactivation of the catalyst is undetectable between subse-
quent runs. As can be seen from a closer analysis, runs 2 and 3
are indicative of zero-order kinetics. In this case, the zero-order
dependence on the concentration of the substrates is inter-
Figure 5. STEM image of 6-rGO, and EDS elemental mapping (HRTEM)
images showing the homogeneous distribution of iridium.
Figure 6. Recycling experiments of the b-alkylation of 1-phenylethanol with
benzyl alcohol. Reaction conditions: 1-phenylethanol (1 mmol), benzyl alco-
hol (1 mmol), KOH (1 mmol) and catalyst 6-rGO and 7-rGO in toluene
(1.5 mL) at 1008C. Yields determined by GC using anisole as internal stan-
dard.
(1 mmol). After completion of each run (8 h), the solid was fil-
tered, washed with CH2Cl2, dried and used in the following
run. As can be seen from the results shown in Figure 6, catalyst
7-rGO could be recycled for up to 12 runs without detectable
loss of activity, whereas the loss of activity of 6-rGO was gradu-
al, and the complete deactivation of the catalyst was observed
after 7 runs. It has been reported for other pyrene-containing
complexes grafted onto graphene that desorption is tempera-
Figure 7. Reaction profiles for three successive cycles of the b-alkylation of
1-phenylethanol with benzyl alcohol. Reaction conditions are the same as
those shown in Figure 6. Lines are drawn through the points for clarity.
Chem. Eur. J. 2015, 21, 15263 – 15271
www.chemeurj.org
15268
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
preted to indicate that the surface of the catalysts is saturated
with the substrate,[5,15] and therefore suggests that there is
some non-covalent interaction between the substrate and the
support.
times without any detectable loss of activity. The determina-
tion of the metal content in both materials before and after
being used in the catalyst experiment reveals that the catalyst
with only one pyrene tag is largely desorbed, while it is effec-
tively retained by the solid in the case of the complex with
two pyrenes. All reaction profiles show a zero–order depend-
ence on the substrates, a fact that we attributed to the satura-
tion of the graphitic surface with the aromatic substrates. To
the best of our knowledge, this is the first time that a catalyst
is effectively recycled in a reaction involving the CÀC coupling
of two alcohols by this borrowing hydrogen process.
Conclusion
This work illustrates the different properties exhibited by ho-
mogenous catalysts with pyrene functionalities compared with
the analogous systems lacking these polyaromatic groups. The
differences arise from the ability of pyrene to afford non-cova-
lent interactions with aromatic molecules, which can either be
substrates in a homogeneous catalysed reaction or in the
same catalysts to afford self-assembled systems. Our study un-
veils some important implications on the effects of non-cova-
lent interactions in homogeneously catalysed reactions, partic-
ularly in two key borrowing hydrogen processes, that is the re-
duction of ketones by transfer hydrogenation and the b-alkyla-
tion of secondary alcohols with primary alcohols. In particular,
we have reached to the following conclusions:
We strongly believe that the results shown in this article
have fundamental implications in the design of future catalysts
containing rigid polyaromatic systems, and will inspire future
researchers in the design of improved homogeneous catalysts,
in which the activities of the metal complexes may be strongly
modified by supramolecular interactions.
Experimental Section
General comments: All manipulations were carried out under ni-
trogen using standard Schlenk techniques and high vacuum. Anhy-
drous solvents were either distilled from appropriate drying agents
(SPS) or degassed prior to use by purging with dry nitrogen and
kept over molecular sieves. The azolium salts 2, 3 and 4 were ob-
tained according to the procedures reported in the literature.[4e] All
other reagents were used as received from commercial suppliers.
NMR spectra were recorded on Varian spectrometers. Electrospray
mass spectra (ESI-MS) were recorded on MicromassQuatro LC in-
strument, and nitrogen was used as a drying and nebulising gas. A
gas chromatograph GC-2010 (Shimadzu) equipped with a FID and
Technokroma (TRB-5MS, 30 m”0.25 mmx0.25 mm) column and
a gas chromatograph/mass spectrometer GCMS- QP2010 (Shimad-
1) The addition of the external p-stacking additive, pyrene, to
the catalytic reactions has an inhibiting effect only for
those reactions implicating both aromatic substrates and
pyrene-tagged catalysts; in contrast, the analogous reac-
tions are completely unaffected in the case of aliphatic sub-
strates or with catalysts lacking polyaromatic tags.
2) The reactions in which aromatic substrates and pyrene-
tagged catalysts are used follow a zero-order dependence
on the concentration of the substrates, strongly suggesting
that the catalyst is saturated with the substrate all along
the reaction course. This situation presumably arises from
the p-stacking interaction between the substrate and the
pyrene tag. All other situations (non-aromatic substrates,
catalysts without pyrene) derive into a second-order de-
pendence on the substrate.
zu) equipped with
a
Technokroma (TRB-5MS, 30 m”
0.25 mmx0.25 mm) column were used. Elemental analyses were car-
ried out in an EA 1108 CHNS-O Carlo Erba analyser. Scanning elec-
tron micrographs were taken with a field emission gun scanning
electron microscope (FEG-SEM) model JEOL 7001F, equipped with
a spectrometer energy dispersion X-ray (EDX) from Oxford instru-
ments. HRTEM and high-angle annular dark-field HAADF-STEM
images of the samples were obtained using a Jem-2100 LaB6
(JEOL) transmission electron microscope coupled with a INCA
Energy TEM 200 (Oxford) EDX operating at 200 kV. The UV/Vis
spectra were recorded between 200 and 800 nm on a Cary 300 Bio
UV/Vis Varian spectrophotometer. The determination of the metal
loading was done by ICP-MS Agilent 7500 CX. The digestion of the
samples was carried out under reflux of a mixture of concentrated
nitric and hydrochloric acids (3:1) for 12 h. The digestion of the
samples after the recycling experiments, were carried out after sev-
eral washes with milli-Q water (10”10 mL).
3) Pyrene-containing catalysts follow a fractional reaction
order <1 with respect to the concentration of the catalyst.
This result suggests self-association of the pyrene-tagged
catalysts, and that the monomers are the active catalytic
species. In contrast, the use of a similar catalyst without the
pyrene functionality derives into a first-order dependence
on the concentration of the catalyst. This result, unveils im-
portant implications about the reaction mechanism, and
clearly illustrates how the presence of a pyrene substituent
in the catalyst strongly modifies the reactivity of the cata-
lyst, compared to otherwise identical metal complexes.
The presence of the pyrene tags in two of the catalysts was
also used to prepare rGO-based materials in which the cata-
lysts were immobilized onto the graphitic surface by non-cova-
lent interactions. The use of these materials in the b-alkylation
of secondary alcohols with primary alcohols, allows to establish
important differences between the material containing one
and two pyrene fragments. While the material with only one
pyrene functionality can be effectively recycled three times,
the catalyst with two pyrene tags can be recycled up to twelve
Synthesis and characterization of new complexes
Synthesis complex 5: A mixture of 1 (93.5 mg, 0.39 mmol), [{IrCl(-
COD) }2] (133.5 mg, 0.2 mmol), K2CO3 (165.95 mg, 0.48 mmol) and
KBr (80 mg) in THF/DMF (7:2 mL) was stirred at 758C for 8 h under
N2. The resulting suspension was cooled to room temperature and
the solvent was removed under vacuum. The pure compound 5
was precipitated in a mixture of dichloromethane/hexane to give
1
a yellow solid. Yield 100 mg (46%); H NMR (300 MHz, CDCl3): d=
7.39–7.28 (m, 4H; CHph), 6.86–6.78 (m, 1H; CHimidazole), 6.69–6.61 (m,
Chem. Eur. J. 2015, 21, 15263 – 15271
www.chemeurj.org
15269
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
2
2
1H, CHimidazole), 5.75 (d, J(H,H)=14.8 Hz, 1H; CH2), 5.52 (d, J(H,H)=
14.8 Hz, 1H; CH2), 4.69 (s, 2H; CHCOD), 3.96 (s, 3H; CH3), 3.03 (m,
1H; CHCOD), 2.90 (m, 1H; CHCOD), 2.27–2.12 (m, 2H; CH2,COD), 2.12–
1.99 (m, 1H; CH2,COD), 1.82–1.68 (m, 2H; CH2,COD), 1.62 (s, 1H;
CH2,COD), 1.60–1.51 (m, 1H; CH2,COD), 1.43 ppm (m, J=8.5, 5.5 Hz,
1H; CH2,COD). 13C NMR (75 MHz, CDCl3): d=180.9 (IrCcarbene), 136.5
(Cph), 129.0(CHph), 128.4 (CHph), 128.3 (CHph), 122.3 (CHimidazole), 120.2
(CHimidazole), 105.2, 84.5 (CHCOD), 84.2 (CHCOD), 54.21 (CH3), 52.58
(CH2), 52.4 (CH2), 37.6 (CHCOD), 33.8 (CHCOD), 33.2 (CHCOD), 31.1
(CHCOD), 30.2 (CHCOD), 29.7 ppm (CHCOD); elemental analysis calcd
(%) for C19H25N2BrIrH2O (571.5515): C 40.00, H 4.70, N 5.00; found:
C 39.51, H 4.05, N 4.63; ESI-MS (cone 20 V): m/z: 473.63 [MÀBr]+.
fied by column chromatography. The pure compound 8 was
eluted with dichloromethane/ethyl acetate (8:2) and precipitated
in a mixture of dichloromethane/hexane to give a yellow solid.
1
3
Yield 87 mg (66%); H NMR (500 MHz, CDCl3): d=8.50 (d, J(H,H)=
9.2 Hz, 2H; CHpy), 8.29–8.14 (m, 7H; CHpy), 8.06 (m, 9H), 7.58 (d,
3
3J(H,H)=7.9 Hz, 4H; CHPh), 7.47 (d, J(H,H)=7.9 Hz, 4H; CHPh), 6.69–
2
6.54 (m, 4H; CHimidazole, CH2), 6.46 (s, 2H; CHimidazole), 6.17 (t, J(H,H)=
13.3 Hz, 2H; CH2), 5.87 (d, 2J(H,H)=14.9 Hz, 2H; CH2), 5.67 (d,
2J(H,H)=14.7 Hz, 2H; CH2), 4.82 (m, 4H; CHCOD), 3.21 (m, 2H;
CHCOD), 3.09 (m, 2H; CHCOD), 2.35–1.97 (m, 8H; CH2,COD), 1.71 (m,
4H; CH2,COD), 1.37–1.23 (m, 2H; CH2,COD), 0.89 ppm (m, 2H; CH2,COD);
13C NMR (126 MHz, CDCl3): d=180.8 (IrÀCcarbene), 140.4 (Pyr), 135.5
(Pyr), 131.7 (Pyr), 131.2 (Pyr), 130.8 (Pyr), 129.4 (Pyr), 128.8 (Pyr),
128.8 (Pyr), 128.5 (Pyr), 128.0 (Pyr), 127.6(Pyr), 127.5 (Pyr),
127.3(Pyr), 126.2 (Pyr), 125.6 (Pyr), 125.0 (CPh), 124.9 (CPh), 124.6
(CHPh), 123.1 (CHPh), 120.6 (CHimidazole), 120.2 (CHimidazole), 84.8 (CHCOD),
84.7 (CHCOD), 54.1 (CH2), 52.9 (CH2), 52.7 (CHCOD), 52.3 (CHCOD), 33.4
(CH2,COD), 29.7 ppm (CH2,COD); elemental analysis calcd (%) for
C70H64N4Br2Ir2CHCl3 (1622.8877): C 52.55, H 3.91, N 3.45; found: C
52.1, H 3.5, N 3.4, ESI-MS (cone 20 V): m/z: 1464.0 [(M-Br)+MeCN]+.
Synthesis of complex 6: A mixture of 2 (70.0 mg, 0.16 mmol),
[{IrCl(COD) }2] (53.73 mg, 0.08 mmol), K2CO3 (67.0 mg, 0.48 mmol)
and KBr (70 mg) in THF/DMF (7:2 mL) was stirred at 758C for 8 h
under N2. The resulting suspension was cooled to room tempera-
ture and the solvent was removed under vacuum. The pure com-
pound 6 was precipitated in a mixture of dichloromethane/hexane
to give a yellow solid. Yield 65 mg (54%); 1H NMR (300 MHz,
CDCl3): d=8.50 (d, 3J(H,H)=9.2 Hz, 1H; CHpy), 8.31–8.18 (m, 4H;
2
CHpy), 8.08 (m, 4H; CHpy), 7.58–7.32 (m, 5H; CHPh), 6.61 (d, J(H,H)=
Synthesis complex 9: A mixture of 2 (70.0 mg, 0.16 mmol),
[{RhCl(COD)}2] (39.43 mg, 0.08 mmol), K2CO3 (67.0 mg, 0.48 mmol)
and KBr (70 mg) in THF/DMF (7:2 mL) was stirred at 758C for 8 h
under N2. The resulting suspension was cooled to room tempera-
ture and the solvent was removed under vacuum. The pure com-
pound 9 was precipitated in a mixture of dichloromethane/hexane
to give a yellow solid. Yield 60 mg (56%); 1H NMR (300 MHz,
CDCl3): d=8.53 (d, 3J(H,H)=9.3 Hz, 1H; CHpyr), 8.32–8.14 (m, 4H;
CHpyr), 8.06 (m, 4H; CHpyr), 7.37 (m, 5H; CHPh), 6.69 (d, 2J(H,H)=
14.8 Hz, 1H; CH2), 6.59 (d, 2J(H,H)=1.9 Hz, 1H; CH), 6.46 (d,
2J(H,H)=2.0 Hz, 1H; CH), 6.33 (d, 2J(H,H)=14.8 Hz, 1H; CH2), 5.87
(d, 2J(H,H)=2.6 Hz, 2H; CHCOD), 5.23 (s, 2H; CH), 3.58 (m, 1H;
CHCOD), 3.48 (m, 1H; CHCOD), 2.47–2.12 (m, 4H; CH2,COD), 1.89 ppm
14.4 Hz, 2H; CH2, CHimidazole), 6.46 (s, 1H), 6.16 (d, 2J(H,H)=
14.8 Hz,1H; CH2) 5.79 (d, 2J(H,H)=15.0 Hz, 1H: CH2), 5.68 (d,
2J(H,H)=14.8 Hz, 1H; CH2), 4.82 (s, 2H; CHCOD), 3.21 (s, 1H; CHCOD),
3.09 (s, 1H; CHCOD), 2.15 (m, 4H; CH2,COD), 1.6 ppm (m, 4H; CH2,COD),
13C NMR (126 MHz, CDCl3): d=180.9 (IrCcarbene), 136.4 (Pyr), 131.8
(Pyr), 131.3 (Pyr), 130.9 (Pyr), 129.6 (Pyr), 198.0 (Pyr), 128.9 (Pyr),
128.7 (Pyr), 128.4 (Pyr), 128.2 (Pyr), 128.1 (Pyr), 127.7 (Pyr), 127.4
(Pyr), 126.3 (Pyr), 125.7 (Pyr), 125.7 (Pyr), 125.1 (CPh), 125.0 (CHPh),
124.7 (CHPh), 123.2 (CHPh), 120.6 (CHimidazole), 120.3 (CHimidazole), 84.8
(CHCOD), 84.8 (CHCOD), 54.6 (CH2), 53.0 (CH2), 52.7 (CH,COD), 52.4
(CH,COD), 33.4 (CH2,COD), 29.8 ppm (CH2,COD); elemental analysis calcd
(%) for C35H32N2BrIrH2O (770.7783): C 54.74, H 4.44, N 3.50; found:
C 54.53, H 4.44, N 3.63; ESI-MS (cone 20 V): m/z: 671.2[MÀBr]+.
1
(m, 4H; CH2,COD). 13C NMR (75 MHz, CDCl3): d=183.6 (d, J=50 Hz,
RhCcarbene),140.5 (Pyr), 135.9 (Pyr), 131.8(Pyr), 131.4(Pyr), 131.0(Pyr),
129.6 (Pyr), 129.00 (Pyr), 129.0 (Pyr), 128.9 (Pyr), 128.1(Pyr), 127.7
(Pyr), 127.4(Pyr), 126.4(Pyr), 125.8 (Pyr), 125.8 (Pyr), 125.1 (CPh),
125.1 (CPh), 124.7 (CHPh), 123.3 (CHPh), 121.3 (CHimidazole), 120.8
(CHimidazole), 98.4 (CHCOD), 98.6 (CHCOD), 98.5 (CHCOD), 69.9 (CHCOD),
69.7 (CHCOD), 69.6 (CHCOD), 69.4 (CHCOD), 54.7 (CH2), 52.8 (CH2), 32.9
(CH2,COD), 32.9 (CH2,COD), 29.2 (CH2,COD); elemental analysis calcd (%)
for C35H32N2BrRhCHCl3 (782.8291): C 55.23, H 4.24, N 3.57; found: C
55.27, H 4.15, N 3.29; ESI-MS (cone 20 V): m/z: 583.2 [MÀBr]+.
Synthesis of complex 7: A mixture of 3 (80.0 mg, 0.10 mmol),
[{IrCl(COD)}2] (64.5 mg, 0.10 mmol), K2CO3 (82.94 mg, 0.6 mmol) and
KBr (70 mg) in THF/DMF (10:3 mL) was stirred at 758C for 8 h
under N2. The resulting suspension was cooled to room tempera-
ture and the solvent was removed under vacuum. The crude prod-
uct was purified by column chromatography. The pure compound
7 was eluted with dichloromethane/ethyl acetate (8:2) and precipi-
tated in a mixture of dichloromethane/hexane to give a yellow
solid. Yield 90 mg (65%); 1H NMR (300 MHz, CDCl3): d=8.48 (dd,
2J(H,H)=9.3, 3.2 Hz, 2H; CHpyr), 8.23–8.13 (m, 8H; CHpyr), 8.09–7.98
(m, 8H; CHpyr), 7.43 (s, 4H; CHPh), 6.58 (m, 4H; CHimidazole, CH2), 6.44
(d, 2J(H,H)=8.0 Hz, 2H; CHimidazole), 6.14 (m, 2H; CH2), 5.85 (dd,
Preparation of 6-rGO: rGO (150 mg) and CH2Cl2 (10 mL) were
added to a round-bottom flask and the suspension was sonicated
for 30 min. Compound 6 (20 mg) was then added and the mixture
was stirred at room temperature for 12 h until the solution
became clear. The black solid was filtered and washed with CH2Cl2
(2”25 mL) to afford 6-rGO as a black solid. The filtrates were com-
bined and evaporated to dryness under reduced pressure to give
2
2J(H,H)=14.6, 11.0 Hz, 2H; CH2), 5.59 (dd, J(H,H)=15.1, 9.5 Hz, 2H;
CH2), 4.80 (s, 4H; CHCOD), 3.18 (s, 2H; CHCOD), 3.06 (s, 2H; CHCOD),
2.12 (m, 8H; CH2,COD), 1.72 (m, 8H; CH2,COD), 1.48 ppm (m, 8H;
CH2,COD); 13C NMR (75 MHz, CDCl3): d 180.8 (IrCcarbene), 136.3 (Pyr),
131.6 (Pyr), 131.2 (Pyr), 130.7 (Pyr), 129.4 (Pyr), 128.7 (Pyr), 128.4
(Pyr), 127.9 (Pyr), 127.57 (Pyr), 127.23 (Pyr), 126.2 (CPh), 125.6 (CHPh),
124.8(CHimidazole), 124.5, (Pyr), 123.0 (CHimidazole), 120.6(Pyr), 120.1
(Pyr), 84.8 (CHCOD), 84.6 (CHCOD), 54.0 (CH2), 53.0 (CH2,COD), 52.6
(CH2,COD), 52.2 (CH2), 33.3 (CH2,COD), 29.7 ppm (CH2,COD); elemental
analysis calcd (%) for C64H58N4Br2Ir23CHCl3 (1904.9247): C 42.87, H
3.28, N 2.90.; found: C 42.56, H 2.75, N 3.14; ESI-MS (cone 20 V): m/
z: 1347 [MÀBr]+.
Synthesis of 8: A mixture of 4 (80.0 mg, 0.09 mmol), [{IrCl(COD)}2]
(67.17 mg, 0.09 mmol), K2CO3 (82.94 mg, 0.6 mmol) and KBr
(70 mg) in THF/DMF (10:3 mL) was stirred at 758C for 8 h under N2.
The resulting suspension was cooled to room temperature and the
solvent was removed under vacuum. The crude product was puri-
1
unsupported compound 6, which was analysed by H NMR using
anisole as the internal standard. Integration of the characteristic
signal of anisole (-OMe) versus CH2-pyrene revealed the amount of
complex that was deposited on the rGO. The exact amount of
complex supported was determined by ICP-MS analysis.
Preparation of 7-rGO: rGO (240 mg) and CH2Cl2 (10 mL) were
added to a round-bottomed flask and the suspension was sonicat-
ed for 30 min. Compound 7 (27 mg) was then added and the mix-
ture was stirred at room temperature for 12 h until the solution
became clear. The black solid was filtered and washed with CH2Cl2
(2”25 mL) to afford 7-rGO as a black solid. The filtrates were com-
bined and evaporated to dryness under reduced pressure to give
1
unsupported compound 7, which was analysed by H NMR using
Chem. Eur. J. 2015, 21, 15263 – 15271
www.chemeurj.org
15270
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[4] a) S. Gonell, M. Poyatos, E. Peris, Angew. Chem. Int. Ed. 2013, 52, 7009–
7013; Angew. Chem. 2013, 125, 7147–7151; b) G. Guisado-Barrios, J.
Hiller, E. Peris, Chem. Eur. J. 2013, 19, 10405–10411; c) H. ValdØs, M.
Poyatos, G. Ujaque, E. Peris, Chem. Eur. J. 2015, 21, 1578–1588; d) H.
ValdØs, M. Poyatos, E. Peris, Inorg. Chem. 2015, 54, 3654–3659; e) S.
Ruiz-Botella, E. Peris, Organometallics 2014, 33, 5509–5516.
[5] a) S. Sabater, J. Mata, E. Peris, Organometallics 2015, 34, 1186–1190;
b) S. Sabater, J. A. Mata, E. Peris, ACS Catal. 2014, 4, 2038–2047.
[6] a) Q. Teng, H. Han Vinh, Inorg. Chem. 2014, 53, 10964–10973; b) S. Guo,
H. Sivaram, D. Yuan, H. Han Vinh, Organometallics 2013, 32, 3685–3696.
[7] a) M. Hamid, P. A. Slatford, J. M. J. Williams, Adv. Synth. Catal. 2007, 349,
1555–1575; b) M. G. Edwards, R. F. R. Jazzar, B. M. Paine, D. J. Shermer,
M. K. Whittlesey, J. M. J. Williams, D. D. Edney, Chem. Commun. 2004,
90–91; c) G. Guillena, D. J. Ramon, M. Yus, Angew. Chem. Int. Ed. 2007,
46, 2358–2364; Angew. Chem. 2007, 119, 2410–2416; d) P. J. Black, G.
Cami-Kobeci, M. G. Edwards, P. A. Slatford, M. K. Whittlesey, J. M. J. Wil-
liams, Org. Biomol. Chem. 2006, 4, 116–125; e) O. Saidi, J. M. J. Williams
in Iridium-Catalyzed Hydrogen Transfer Reactions, Vol. 34 (Ed.: P. G. An-
dersson), 2011, pp. 77–106; f) T. Suzuki, Chem. Rev. 2011, 111, 1825–
1845.
anisole as the internal standard. Integration of the characteristic
signal of anisole (-OMe) versus CH2-pyrene revealed the amount of
complex and Ir that was deposited on the rGO. The exact amount
of Ir supported was determined by ICP-MS analysis.
General procedure for the transfer hydrogenation: The catalyt
(0.5 mol% based on the metal) and KOH (0.05 mol) were placed in
a Schlenk tube with a Teflon cap. The tube was then degassed and
filled with nitrogen three times. 2-Propanol (2 mL) and the corre-
sponding ketone (0.5 mmol) were added, and the mixture was
stirred at 1008C for 1 or 2 h. Yields were determined by GC analy-
ses using anisole (0.5 mmol) as the internal standard.
General procedure for the b-alkylation of secondary alcohols
with primary alcohols: The catalyst (0.5 mol% based on the
metal) and KOH (1.0 mmol) were placed in a Schlenk tube with
a Teflon cap. The tube was then degassed and filled with nitrogen
three times. 1-phenylethanol (1 mmol), benzyl alcohol (1 mmol)
and dry toluene (0.5 mL) were added and the mixture was stirred
at 1008C for 8 h. Yields were determined by GC analyses using ani-
sole (1.0 mmol) as the internal standard.
[8] U. Hintermair, J. Campos, T. P. Brewster, L. M. Pratt, N. D. Schley, R. H.
Crabtree, ACS Catal. 2014, 4, 99–108.
Recycling experiments: In a round-bottomed flask, a mixture of 1-
phenylethanol (1 mmol), benzyl alcohol (1 mmol), KOH (1.0 mmol),
anisole (1.0 mmol) as internal standard and the 6-rGO or 7-rGO
(0.5 mol% based on the metal) was refluxed in toluene (1.5 mL) for
8 h. The monitoring of the reaction, yields, and conversions were
determined by GC analyses. After completion of each run (8 h), the
reaction mixture was allowed to reach room temperature and was
then filtered. The remaining solid was washed thoroughly with
CH2Cl2, dried under reduced pressure and reused in the following
run.
[9] a) P. Foley, R. Dicosimo, G. M. Whitesides, J. Am. Chem. Soc. 1980, 102,
6713–6725; b) N. D. Schley, J. D. Blakemore, N. K. Subbaiyan, C. D. Incar-
vito, F. D’Souza, R. H. Crabtree, G. W. Brudvig, J. Am. Chem. Soc. 2011,
133, 10473–10481.
[10] a) R. S. Srivastava, K. M. Nicholas, Organometallics 2005, 24, 1563–1568;
b) E. L. Hegg, S. H. Mortimore, C. L. Cheung, J. E. Huyett, D. R. Powell,
J. N. Burstyn, Inorg. Chem. 1999, 38, 2961–2968.
[11] a) V. Georgakilas, M. Otyepka, A. B. Bourlinos, V. Chandra, N. Kim, K. C.
Kemp, P. Hobza, R. Zboril, K. S. Kim, Chem. Rev. 2012, 112, 6156–6214;
b) L. Rodríguez-PØrez, M. Angeles Herranz, N. Martin, Chem. Commun.
2013, 49, 3721–3735.
[12] a) A. Le Goff, B. Reuillard, S. Cosnier, Langmuir 2013, 29, 8736–8742;
b) J. A. Mann, J. Rodriguez-Lopez, H. D. Abruna, W. R. Dichtel, J. Am.
Chem. Soc. 2011, 133, 17614–17617.
Acknowledgements
[13] a) M. Keller, V. Colliere, O. Reiser, A.-M. Caminade, J.-P. Majoral, A. Ouali,
Angew. Chem. Int. Ed. 2013, 52, 3626–3629; Angew. Chem. 2013, 125,
3714–3717; b) S. Wittmann, A. Schaetz, R. N. Grass, W. J. Stark, O. Reiser,
Angew. Chem. Int. Ed. 2010, 49, 1867–1870; Angew. Chem. 2010, 122,
1911–1914; c) P. Kang, S. H. Zhang, T. J. Meyer, M. Brookhart, Angew.
Chem. Int. Ed. 2014, 53, 8709–8713; Angew. Chem. 2014, 126, 8853–
8857; d) L. Xing, J.-H. Xie, Y.-S. Chen, L.-X. Wang, Q.-L. Zhou, Adv. Synth.
Catal. 2008, 350, 1013–1016.
[14] a) C. Ehli, D. M. Guldi, M. A. Herranz, N. Martin, S. Campidelli, M. Prato, J.
Mater. Chem. 2008, 18, 1498–1503; b) M. A. Herranz, C. Ehli, S. Campi-
delli, M. Gutierrez, G. L. Hug, K. Ohkubo, S. Fukuzumi, M. Prato, N.
Martin, D. M. Guldi, J. Am. Chem. Soc. 2008, 130, 66–73; c) J. S. Kavakka,
S. Heikkinen, I. Kilpelainen, M. Mattila, H. Lipsanen, J. Helaja, Chem.
Commun. 2007, 519–521.
We gratefully acknowledge financial support from MEC of
Spain (CTQ2014-51999-P) and the Universitat Jaume
I
(P11B2014-02). We are grateful to the Serveis Centrals d’Instru-
mentació Científica (SCIC) of the Universitat Jaume I for provid-
ing with spectroscopic facilities. We would also like to thank
the Generalitat Valenciana for a fellowship (S.R.B.).
Keywords: borrowing-hydrogen · graphene · iridium · N-
heterocyclic carbenes · pi interactions
[1] M. Raynal, P. Ballester, A. Vidal-Ferran, P. W. N. M. van Leeuwen, Chem.
Soc. Rev. 2014, 43, 1660–1733.
[15] a) H. H. Kung, R. J. Pellet, R. L. Burwell, J. Am. Chem. Soc. 1976, 98,
5603–5611; b) T. R. Lee, G. M. Whitesides, Acc. Chem. Res. 1992, 25,
266–272; c) F. Zaera, Progress in Surface Science 2001, 69, 1–98; d) A. D.
Pomogailo, Catalysis by Polymer-Immobilized Metal Complexes, Gordon
and Breach Science Publishers, Amsterdam, 1998.
[2] a) P. Dydio, J. N. H. Reek, Chem. Sci. 2014, 5, 2135–2145; b) S. Carboni,
C. Gennari, L. Pignataro, U. Piarulli, Dalton Trans. 2011, 40, 4355–4373.
[3] a) L. G. Ma, P. Hao, Q. H. Zhang, D. M. Du, H. X. Xu, Tetrahedron: Asym-
metry 2007, 18, 878–884; b) P. Brandt, P. Roth, P. G. Andersson, J. Org.
Chem. 2004, 69, 4885–4890; c) R. W. Quan, Z. Li, E. N. Jacobsen, J. Am.
Chem. Soc. 1996, 118, 8156–8157; d) G. B. Jones, B. J. Chapman, Synthe-
sis 1995, 475–497; e) H. C. Kolb, P. G. Andersson, K. B. Sharpless, J. Am.
Chem. Soc. 1994, 116, 1278–1291; f) L. A. Castonguay, A. K. Rappe, C. J.
Casewit, J. Am. Chem. Soc. 1991, 113, 7177–7183.
Received: July 27, 2015
Published online on September 10, 2015
Chem. Eur. J. 2015, 21, 15263 – 15271
www.chemeurj.org
15271
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim