Cooperative or Oxidative Hydrogen Addition to 2-Hydroxypyridonate Iridium Complexes: Dependence on Oxidation State

Article abstract of DOI:10.1002/ejic.201700534

Iridium(III)–pyridone complexes are commonly found to react in a cooperative and redox-neutral manner with dihydrogen and alcohols. In this work, the reactivity preferences of Ir^I–pyridone complexes were investigated under a variety of conditions. We have found that, in contrast to Ir^III–pyridones, Ir^I–pyridone complexes display a strong preference to react non-cooperatively. With a new chelating 2-hydroxy-8-diphenylphosphinoquinoline ligand that does not dissociate after hydrogen addition, oxidative addition is still preferred. In the preparation of mono- and bidentate neutral and anionic pyridone ligands, Vaska's complex was used as a point of reference. We expect these findings to have implications for catalyst development in the field of metal–ligand cooperation.

Full text of DOI:10.1002/ejic.201700534

DOI: 10.1002/ejic.201700534
Communication
Cooperative Phenomena
Cooperative or Oxidative Hydrogen Addition to
2-Hydroxypyridonate Iridium Complexes: Dependence on
Oxidation State
Sebastian J. K. Forrest,[a] Seetharaman Manojveer,[a] and Magnus T. Johnson*[a]
Abstract: Iridium(III)–pyridone complexes are commonly found phinoquinoline ligand that does not dissociate after hydrogen
to react in a cooperative and redox-neutral manner with di-
hydrogen and alcohols. In this work, the reactivity preferences
of IrI–pyridone complexes were investigated under a variety of
conditions. We have found that, in contrast to IrIII–pyridones, IrI–
pyridone complexes display a strong preference to react non-
cooperatively. With a new chelating 2-hydroxy-8-diphenylphos-
addition, oxidative addition is still preferred. In the preparation
of mono- and bidentate neutral and anionic pyridone ligands,
Vaska's complex was used as a point of reference. We expect
these findings to have implications for catalyst development in
the field of metal–ligand cooperation.
Introduction
nated by iridium. A recent example with a slightly more com-
plex ligand architecture, 2-hydroxy-6-methyl(diphenylphos-
phino)pyridine, was reported by Achard et al.[5] To the best of
our knowledge, all reported work relating to cooperative irid-
ium–pyridone catalysis involves iridium(III) metal centres.
Rauchfuss has reported the reaction of Cp*Ir(κ2-2-pyridonate)Cl
with dihydrogen and hydrogen donors.[6] It was observed that
dihydrogen was added cooperatively across the iridium centre
and the oxygen of the pyridonate ligand, under very mild con-
ditions (Scheme 1b). The field of iridium–pyridone catalysis has
been pioneered by Fujita, who has found application in a num-
ber of reactions, for example, hydrogen production from water
and methanol,[7] acceptorless alcohol dehydrogenation[8] and
N-heterocycle (de-)hydrogenation.[9]
The use of cooperative ligands that actively participate in con-
junction with a metal centre to catalyse reactions in what is
often called metal–ligand cooperation (MLC) has become in-
creasingly important since the breakthrough of bifunctional
Noyori-type catalysts.[1] The majority of work utilising MLC to
cooperatively catalyse hydrogen and alcohol activation involves
ruthenium and iridium and was recently reviewed.[1a,2] Since
the discovery that a 2-hydroxypyridine moiety is present in the
active site of monoiron hydrogenase,[3] 2-hydroxypyridine li-
gands have gained interest in MLC catalysis.[2] For cooperative
addition to occur, the 2-hydroxypyridine ligand must first be
deprotonated to form the activated anionic 2-pyridonate li-
gand. In this way, the 2-pyridonate ligand can act as an internal
base to cooperate with the metal centre in the activation
of hydrogen–element bonds. The simple tautomeric pair 2-
hydroxypyridine/2-pyridone serves as a monodentate ligand to
be coordinated either through the nitrogen or the oxygen
atom.[4] Subsequent addition of either dihydrogen or an alcohol
to the metal–pyridonate generates the neutral 2-hydroxypyr-
idine and the metal hydride or metal alkoxide, respectively
(Scheme 1a). Relative to the monumental body of work con-
cerning cooperative ligands in general, reports utilising the 2-
hydroxypyridine moiety are relatively rare. This area was, how-
ever, recently thoroughly reviewed.[1a,2] The majority of reports
employs simple unsubstituted 2-hydroxypyridine, 6,6′-di-
hydroxy-2,2′-bipyridine or 2,9-dihydroxyphenanthroline, coordi-
Scheme 1. (a) General representation of a cooperative addition of hydrogen
or alcohols to a base-activated metal–pyridonate complex. (b) Cooperative
addition of dihydrogen to an IrIII–pyridonate complex, as observed by Rauch-
fuss and co-workers.[6]
[a] Centre for Analysis and Synthesis, Department of Chemistry,
Lund University,
Crabtree et al. previously demonstrated that the hydroxyl
proton of a 2-hydroxypyridine ligand can interact with iridium
hydrides and halides in IrIII complexes.[10] To the best of our
knowledge, there are only two reports in which IrIII complexes
have been isolated after hydrogen addition, both in the work
P. O. Box 124, 22100 Lund, Sweden
E-mail: magnus.johnson@chem.lu.se
http://www.kilu.lu.se/cas/research/the-johnson-group/
Supporting information and ORCID(s) from the author(s) for this article are
available on the WWW under https://doi.org/10.1002/ejic.201700534.
Eur. J. Inorg. Chem. 2017, 3239–3243
3239
© 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
by Fujita and Yamaguchi.[9] In both of these, the protonated and L1H. Complex 1-neutral was characterised by NMR spec-
form of the ligand was obtained together with the iridium troscopy, mass spectrometry and X-ray diffraction; however,
hydride, implying a cooperative addition pathway. The addition elemental analysis was unsuccessful. X-ray-quality crystals of
of dihydrogen to IrI complexes bearing hydroxypyridine/ 1-neutral were obtained by cooling an acetonitrile solution of
pyridone moieties is, to the best of our knowledge, unreported. 1-neutral to –25 °C overnight (Figure 1). The structure is square
We were therefore interested to investigate whether a trivalent planar, with the pyridone orientated so that the C–O bond is in
iridium centre is required for the cooperative addition to occur, the N–Ir–C plane, antiperiplanar to the coordination plane of
or whether this is also possible with a monovalent iridium com- iridium. The Ir–CO bond is longer in 1-neutral [1.913(14) Å]
plex. In order to obtain a better understanding of the properties than in 1-OTf [1.792(5) Å], and the C–O bond length is corre-
of the pyridone moiety in metal-mediated hydrogen activation spondingly shorter in 1-neutral [1.10(2) Å] than in 1-OTf
and transfer reactions, studies on the addition of dihydrogen to [1.167(6) Å] (Table 1). IR spectroscopy results corroborate these
iridium–pyridone/pyridonate complexes were performed. Ana- observations: there are carbonyl stretches at 1989 and 1971 cm–1
logues of Vaska's complex ligated by 2-hydroxypyridine (L1H) for complexes 1-neutral and 1-OTf, respectively.
and 2-pyridonate (L1) were used as a point of reference. Fur-
ther, in order to address the problems related to 2-hydroxy-
pyridine dissociation after hydrogen addition, as commonly
experienced, the bidentate ligand 2-hydroxy-8-diphenylphos-
phinoquinoline (L2H) was also prepared and investigated anal-
ogously.
Results and Discussion
Synthesis of Iridium Complexes
An overview of the metal complex synthesis is shown in
Scheme 2. The iridium triflate (1-OTf) and tetrafluoroborate
(1-BF4) salts with ligand L1H were both prepared in 53 % yield.
Compounds 1-OTf and 1-BF4 display indistinguishable chemi-
cal shifts in their respective 1H- and 31P NMR spectra, which
indicates that the cation structures are identical. We were un-
Figure 1. Thermal ellipsoid representations (50 % probability level) of com-
plexes (a) 1-neutral·2MeCN, (b) 1-OTf·2C6H6, (c) 2, and (d) the oxide of L2H
(L2-oxide). Hydrogen atoms, solvates and counteranions were excluded for
clarity.
successful in obtaining satisfactory elemental analysis results of
the 1-BF4 salt.
It was possible to grow X-ray-quality crystals of complex 1-
OTf by addition of benzene to an acetonitrile solution of 1-OTf
and allowing the resultant solution to concentrate slowly by
evaporation of the solvents. The structure shows that the com-
plex is square planar around the iridium centre, with the
hydroxyl group of the pyridine in the N–Ir–C plane.
Complex 1-neutral could be prepared either through direct
reaction of Vaska's complex with the silver pyridonate salt in a
52 % yield, or by reaction of Vaska's complex with silver oxide
Table 1. Selected bond lengths [Å] and angles [°] of the structures in Figure 1.
Ir–N
C–O(CO)
C–O(Hp/qui)
P1–Ir–N
P1–Ir–CO
1-neutral 2.15(1)
1.10(2)
1.167(6)
1.23(1)
–
1.29(2)
92.13(6)
90.2(1)
81.7(2)
–
87.87(6)
90.1(2)
94.5(2)
–
1-OTf
2
2.099(3)
2.141(5)
–
1.335(7)
1.235(8)
1.230(5)
L2-oxide
Scheme 2. Preparation of complexes 1-BF4, 1-OTf, 1-neutral and 2.
Eur. J. Inorg. Chem. 2017, 3239–3243
www.eurjic.org
3240
© 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
When attempting to form the iridium tetrafluoroborate of
L2H following the procedure used for the formation of 1-BF4,
a mixture of products was observed by 31P NMR spectroscopy.
Silver triflate, assigned tentatively as 2-OTf, was observed as
one major product in the 31P NMR spectrum, and the spectrum
was selectively decoupled to aromatic proton resonances to al-
low the coupling of the phosphines to the hydride. Two dou-
blets of doublets were observed, the larger coupling of each
peak, 316 Hz, was assigned as a trans coupling between the
two phosphines. The smaller coupling of each peak, 11 Hz, is
1
due to coupling to a hydride, which is detected in the H NMR
Scheme 3. Preparation of ligand L2H.
spectrum as a virtual triplet. Coordination of the quinoline li-
gand resulted in displacement of one equivalent of triphenyl-
phosphine and the oxidation of the iridium by protonation of
the iridium centre by the hydroxyl proton of L2H. Repeated
attempts at recrystallisation from a variety of solvents did not
yield pure 2-OTf, and the 31P NMR spectra of the solids ob-
tained after each recrystallisation indicated no change in the
relative distribution of 2-OTf and triphenylphosphine in solu-
tion.
Addition of triethylamine to an acetonitrile solution of 2-OTf
led to the precipitation of X-ray-quality crystals of five-coordi-
nate complex 2. The structure shows a trigonal bipyramidal irid-
ium(I) centre bearing the 2-hydroxy-8-diphenylphosphinoquin-
oline ligand. The carbonyl group is located trans to the nitrogen
atom occupying the axial sites, and the phosphines are all equa-
torial. This structure appears to be fluxional in solution at room
temperature, as the 31P NMR spectrum in [D8]toluene shows no
signals. Upon cooling to 240 K, it is possible to resolve the 31P
NMR spectrum to yield a triplet at δ = 29.5 ppm and a doublet
at δ = 14.5 ppm with corresponding coupling constants of
154 Hz for both. It should be noted that 2 is a rare example of a
five-coordinate IrI complex with phosphine ligands. It is further
assumed that the fluxionality observed at room temperature is
due to the decoordination of a triphenylphosphine to form a
four-coordinate square-planar complex. In addition to X-ray dif-
fraction, the complex was characterised in the solid state by
elemental analysis and IR spectroscopy, the IR spectrum show-
ing a band at 1985 cm–1.
halogen exchange being suppressed by the initial OH deproto-
nation by the nBuLi. In an effort to improve the overall yield,
milder deprotonation of the hydroxyl group was also attempted
using sodium hydride prior to addition of nBuLi, without suc-
cess. Protection with tert-butyldimethylsilyl ether (TBDMS) was
also unsuccessful. The yield could possibly have been improved
through further protecting group optimisation; nevertheless,
we proceeded with the metallation reactions directly.
Hydrogen Addition Reactions
The addition of dihydrogen to complexes 1-BF4, 1-OTf, 1-neu-
tral and 2 was carried out and the products were characterised
by in situ NMR spectroscopy and X-ray diffraction (3-BF4 and
5-fac, Scheme 4). Upon reaction of 1-OTf and 1-BF4 with
dihydrogen in [D3]acetonitrile, the reaction mixtures displayed
identical 1H-, 31P{1H}- and 31P NMR spectra, showing full conver-
sion of the starting material. The 31P NMR spectra displayed a
1
single doublet of doublets. The H NMR spectra displayed two
hydride peaks, which were triplets of doublets with coupling
constants corresponding to each other and to the peak in the
31P NMR spectrum. The 1H- and 13C NMR spectra also displayed
signals for free L1H and the signals for the coordinated triphen-
ylphosphines. The formation of 3 occurs because of the lability
of the 2-hydroxypyridine, the ligand disassociating either prior
to or after oxidative addition of H2.
Upon cooling of the 1-BF4 reaction mixture, crystals of the
complex 3-BF4 suitable for X-ray diffraction were isolated. The
X-ray structure shows that decoordination of the pyridine from
the iridium centre has occurred and an equivalent of aceto-
Synthesis of 2-Hydroxy-8-diphenylphosphinoquinoline
(L2H)
The synthesis of 2-hydroxy-8-diphenylphosphinoquinoline nitrile has coordinated in its place (Figure 2). Oxidative addition
(L2H) was carried out in three steps from 2-bromoaniline of H2 is suggested by the arrangement of the hydrides, the
(Scheme 3). Reaction of 2-bromoaniline with E-cinnamoyl chlor- position of which are implied from the geometry of the metal
ide proceeded to give the amide intermediate in 66 % yield. centre and are located as such. Redissolving the crystals yielded
Cyclisation of the amide intermediate generated the 2-hydroxy- the identical NMR spectra, however without any of the signals
8-bromoquinoline in 85 % yield.[11]
In the last step, three equivalents of nBuLi were added to
induce lithium–halogen exchange and deprotonation of the
of L1H confirming that formation of 3 occurs upon reaction of
1-OTf or 1-BF4 with H2.
Reaction of 1-neutral with dihydrogen in [D3]acetonitrile
lactam. Addition of excess diphenylchlorophosphine followed also resulted in oxidative addition at the iridium centre with
1
by a hydrogen carbonate wash and recrystallisation gave L2H complete conversion of the starting material. The H-, 31P{1H}-
in 13 % yield. Ligand L2H was further found to be air-sensitive, and 31P NMR spectra indicate the formation of a cis-dihydride
and X-ray-quality crystals of the oxide of L2H (L2-oxide) were complex 4, and 1H NMR spectra indicate that the pyridone moi-
obtained, while elemental analysis was attempted without suc- ety remains coordinated and deprotonated. In combination
cess. The low yield of the final step is attributed to the lithium with NMR spectroscopic data, we assigned the structure as irid-
Eur. J. Inorg. Chem. 2017, 3239–3243
www.eurjic.org
3241
© 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Scheme 4. Reaction of iridium 2-pyridine/pyridone complexes with H2.
hydrogen leads to the formation of 5-mer and 5-fac. This proc-
ess could be assisted by the basic pyridonate; nevertheless, de-
spite low-temperature experiments, we were not able to detect
any intermediates supporting this hypothesis.
The reaction of 2 with H2 was carried out in both [D3]MeCN
and [D8]toluene. Complex 2 is only sparingly soluble in MeCN;
however, at –33 °C, it is still possible to observe signals in the
1H- and 31P NMR spectra that match those observed for com-
plex 2 in [D8]toluene. As the five-coordinate complex is dy-
namic at room temperature, addition of H2 was carried out at
–78 °C, and the sample was loaded directly into the NMR spec-
trometer, which was precooled to –33 °C. The 1H NMR spectrum
displays two hydride signals both doublets of doublets, which
are singlets in the 1H{31P} NMR spectrum. The 31P{1H} NMR spec-
trum shows three phosphorus environments, two doublets at
δ = 12.9 and 6.5 ppm with coupling constants of 13 Hz and a
broadened singlet at δ = –5.2 ppm, assigned to free triphenyl-
phosphine. Coupling between the phosphorus environments at
δ = 12.9 and 6.5 ppm and both hydride environments was
shown in the 1H-31P HMBC spectrum. The COSY experiment
shows cross peaks, at low intensity due to the cis-relationship,
between the two hydride environments. The lack of hydride–
Figure 2. Thermal ellipsoid representations (50 % probability level) of com-
plexes (a) 3-BF4 and (b) 5-fac. Hydrogen atoms and solvates are excluded
for clarity.
ium complex 4. In the conversion of 1-neutral to 4, no interme-
diates could be observed even at low temperature, either be-
cause they were nonexistent or elusive with a very short life-
time. When crystals of 4 were redissolved [D6]benzene, the re-
sulting NMR spectra matched the in situ reaction of 1-neutral
with H2, confirming the formation of 4.
Complex 4 slowly decomposes in a H2 atmosphere at room
temperature upon standing. The 1H-, 1H{31P}- and COSY NMR
spectra show four new resonances in the hydride region all
between δ = –9 and –11 ppm. The peaks at δ = –9.2 ppm, a
triplet of doublets, and at δ = –9.9 ppm, a triplet of triplets,
are shown to couple and have relative intensities of 2:1. The
peaks at δ = –9.4 and –10.4 ppm are a triplet and a multiplet,
respectively, with relative intensities of 1:2 and are shown to
couple. We assign these peaks by comparison to literature
data as the meridional (5-mer) and facial (5-fac) isomers of
Ir(H)3(PPh3)2(CO), respectively, in a 1:5 ratio (Scheme 4).[12] This
assignment was corroborated by 31P- and 31P-1H HMBC NMR
1
hydride coupling observed in the H NMR spectra is attributed
to the fact that the cis-coupling constant between the two
hydrides is very small. It is also possible to observe a cross peak
between the hydrides and a 13C NMR resonance at δ = 173 ppm
in the 1H-13C HMBC spectrum, interpreted as coupling between
the hydrides and the carbonyl ligand. Taken together, the in
situ data indicates that the cis-dihydride 6 is the product of
oxidative addition of H2 to 2. Several attempts were made to
isolate 6 from solution; however, removal of the hydrogen at-
mosphere resulted in decomposition of 6, and all attempts at
crystallisation under a hydrogen atmosphere were unsuccessful.
However, it is clear from the experiments that the bidentate
quinoline ligand does not dissociate from the metal centre as is
commonly observed with the simple unsubstituted 2-hydroxy-
1
spectra. The H- and HSQC NMR spectra also show signals for
free 2-hydroxypyridine. By cooling of the reaction mixture it
was possible to isolate X-ray-quality crystals of 5-fac, Figure 2.
When the procedure was repeated in C6D6, identical results
were obtained.
The lability of the 2-hydroxypyridine/pyridone moiety in the pyridine.
above examples is similar to the observed lability of the 2-pyr-
idone reported by Rauchfuss.[6] The formation of 3-BF4 and 3-
OTf can be attributed to direct oxidative addition and ligand
Conclusion
substitution. The formation of 4 shows that in the H2 addition
to IrI complex 1-neutral, direct oxidative addition occurs. After In summary, we have shown that iridium(I) complexes with 2-
IrIII complex 4 forms, the addition of a second equivalent of pyridonate-containing ligand frameworks add hydrogen oxid-
Eur. J. Inorg. Chem. 2017, 3239–3243
www.eurjic.org
3242
© 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Int. Ed. 2008, 47, 1814–1818; Angew. Chem. 2008, 120, 1838; c) R. H.
Crabtree, New J. Chem. 2011, 35, 18–23; d) B. Zhao, Z. Han, K. Ding,
Angew. Chem. Int. Ed. 2013, 52, 4744–4788; Angew. Chem. 2013, 125,
4844.
atively and non-cooperatively under a variety of conditions. This
is the case even when there is an option for cooperative addi-
tion, as with the preactivated basic amide functionality in 1-
neutral and 2. When the metal centre instead is in oxidation
state +III, the addition occurs cooperatively, as observed and
discussed with reference to the literature and as shown with
complex 4. With regard to the background that the majority of
reported work on pyridone-ligated catalysts occurs with iridium
metal centres, we believe that this insight is highly relevant for
further catalyst development in this field.
[2] a) C. M. Moore, E. W. Dahl, N. K. Szymczak, Curr. Opin. Chem. Biol. 2015,
25, 9–17; b) W.-H. Wang, J. T. Muckerman, E. Fujita, Y. Himeda, New J.
Chem. 2013, 37, 1860–1866.
[3] a) S. Shima, E. J. Lyon, M. Sordel-Klippert, M. Kauß, J. Kahnt, R. K. Thauer,
K. Steinbach, X. Xie, L. Verdier, C. Griesinger, Angew. Chem. Int. Ed. 2004,
43, 2547–2551; Angew. Chem. 2004, 116, 2601; b) S. Shima, O. Pilak, S.
Vogt, M. Schick, M. S. Stagni, W. Meyer-Klaucke, E. Warkentin, R. K.
Thauer, U. Ermler, Science 2008, 321, 572–575.
[4] a) J. M. Rawson, R. E. P. Winpenny, Coord. Chem. Rev. 1995, 139, 313–374;
b) S. Parsons, R. E. P. Winpenny, Acc. Chem. Res. 1997, 30, 89–95.
[5] F. Jiang, M. Achard, T. Roisnel, V. Dorcet, C. Bruneau, Eur. J. Inorg. Chem.
2015, 4312–4317.
[6] A. M. Royer, T. B. Rauchfuss, D. L. Gray, Organometallics 2010, 29, 6763–
6768.
CCDC 1519842 (for 1-OTf), 1519839 (for 1-neutral), 1519843 (for
2), 1519840 (for L2-oxide), 1519841 (for 3-BF4) and 1534634 (for 5-
fac) contain the supplementary crystallographic data for this paper.
These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre.
[7] K.-i. Fujita, R. Kawahara, T. Aikawa, R. Yamaguchi, Angew. Chem. Int. Ed.
2015, 54, 9057–9060; Angew. Chem. 2015, 127, 9185.
[8] a) K.-i. Fujita, N. Tanino, R. Yamaguchi, Org. Lett. 2007, 9, 109–111;
b) K.-i. Fujita, T. Yoshida, Y. Imori, R. Yamaguchi, Org. Lett. 2011, 13, 2278–
2281; c) R. Kawahara, K.-i. Fujita, R. Yamaguchi, Angew. Chem. Int. Ed.
2012, 51, 12790–12794; Angew. Chem. 2012, 124, 12962; d) R. Kawahara,
K. Fujita, R. Yamaguchi, J. Am. Chem. Soc. 2012, 134, 3643–3646; e) G.
Zeng, S. Sakaki, K.-i. Fujita, H. Sano, R. Yamaguchi, ACS Catal. 2014, 4,
1010–1020.
[9] a) R. Yamaguchi, C. Ikeda, Y. Takahashi, K.-i. Fujita, J. Am. Chem. Soc. 2009,
131, 8410–8412; b) K.-i. Fujita, Y. Tanaka, M. Kobayashi, R. Yamaguchi, J.
Am. Chem. Soc. 2014, 136, 4829–4832.
Acknowledgments
Financial support by the Swedish Research Council, the Olle
Engkvist Byggmästare Foundation, the Royal Swedish Academy
of Forestry and Agriculture, the Royal Physiographic Society in
Lund and the Science and Engineering Research Board, India,
the Magnus Bergvall Foundation (SB/OS/PDF-004/2015-16)
(S. M.), is gratefully acknowledged.
[10] E. Peris, J. C. Lee, J. R. Rambo, O. Eisenstein, R. H. Crabtree, J. Am. Chem.
Soc. 1995, 117, 3485–3491.
[11] L. D. Wickramasinghe, R. Zhou, R. Zong, P. Vo, K. J. Gagnon, R. P. Thum-
mel, J. Am. Chem. Soc. 2015, 137, 13260–13263.
Keywords: Iridium · Hydrides · N ligands · Metal–ligand
cooperation
[12] S. Hasnip, S. B. Duckett, S. B. Duckett, D. R. Taylor, M. J. Taylor, Chem.
Commun. 1998, 923–924.
[1] a) J. R. Khusnutdinova, D. Milstein, Angew. Chem. Int. Ed. 2015, 54, 12236–
12273; Angew. Chem. 2015, 127, 12406; b) H. Grützmacher, Angew. Chem.
Received: May 10, 2017
Eur. J. Inorg. Chem. 2017, 3239–3243
www.eurjic.org
3243
© 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim