Design of maleimide-functionalised electrodes for covalent attachment of proteins through free surface cysteine groups

Article abstract of DOI:10.1002/chem.201400246

Mixed two-component monolayers on glassy carbon are prepared by electrochemical oxidation of N-(2-aminoethyl)acetamide and mono-N-Boc- hexamethylenediamine in mixed solution. Subsequent N-deprotection, amide coupling and solid-phase synthetic steps lead to electrode-surface functionalisation with maleimide, with controlled partial coverage of this cysteine-binding group at appropriate dilution for covalent immobilisation of a model redox-active protein, cytochrome c, with high coverage (≈7.5pmolcm ^-2). In the mix: A controlled approach to the modification of carbon electrodes with maleimide groups is described for the efficient immobilisation of bovine heart cytochrome c, which is a redox-active model for proteins with free surface cysteine residues (see figure; PG=protecting group, GC=glassy carbon).

Full text of DOI:10.1002/chem.201400246

DOI: 10.1002/chem.201400246
Communication
&
Electrochemistry
Design of Maleimide-Functionalised Electrodes for Covalent
Attachment of Proteins through Free Surface Cysteine Groups
Emma J. Wright,[a] Maciej Sosna,[a, c] Sally Bloodworth,*[a] Jeremy D. Kilburn,*[b] and
Philip N. Bartlett*[a]
electrode surface around the binding group should be com-
Abstract: Mixed two-component monolayers on glassy
carbon are prepared by electrochemical oxidation of N-(2-
patible with the protein surface around the target functional
group of the protein to prevent the repulsion or denaturation
aminoethyl)acetamide and mono-N-Boc-hexamethylenedi-
of the enzyme.
amine in mixed solution. Subsequent N-deprotection,
A number of reliable procedures exist to create close-packed
amide coupling and solid-phase synthetic steps lead to
and mixed monolayers on electrode surfaces.[1–3] Monolayers
electrode-surface functionalisation with maleimide, with
can be readily assembled on gold from thiols or disulfides in
controlled partial coverage of this cysteine-binding group
solution,[4–6] however the layers thus formed are dynamic due
at appropriate dilution for covalent immobilisation of
to the mobility of thiols on gold and thiol exchange can occur
a model redox-active protein, cytochrome c, with high
coverage (ꢀ7.5 pmolcmÀ2).
in mixed monolayers, with the fraction of longer chains on the
surface increasing upon prolonged incubation in the deposi-
tion solution, making the surface composition different to that
of the solution.[6,7] Mixed thiols on the surface also segregate
Modification of electrode surfaces is fundamental to the devel-
opment of new, more efficient electrochemical devices for
sensing, catalysis and energy conversion. The desired redox
catalyst or recognition molecules must be stable on the sur-
face, preferably through covalent immobilisation, and their
redox state easily modulated by the potential applied to the
underlying bulk electrode. The particular case of modification
of electrodes with large biomolecules, especially redox en-
zymes, poses special challenges. To address the redox centre
of the enzyme directly, its orientation on the surface has to be
controlled in order to minimise the electron transfer distance
and thus avoid the necessity of using diffusing redox media-
tors. This requires surfaces that can selectively bind a specific
functional group (often artificially introduced by protein engi-
neering) of the enzyme. In addition, steric considerations must
be taken into account. The binding group on the surface
should be able to access the target functional group of the
protein and be laterally distributed to accommodate the bulky
macromolecule. Moreover, the polarity/charge of the modified
into homogeneous islands over time, and monolayer stability
is limited, particularly for short thiols and mixtures.[8] These
problems can be avoided with covalently modified carbon sub-
strates, but there has been significantly less research into
mixed monolayers on carbon. Gooding et al. have investigated
the formation of mixed layers from reaction of glassy carbon
(GC) surfaces with mixtures of differently substituted aryl di-
azonium salts, generated in situ. X-ray photoelectron spectros-
copy (XPS) indicates formation of the mixed layer, although
the ratio on the surface differs from that of the solution, with
the more easily reduced diazonium salt giving greater surface
coverage than expected from its concentration in solution.[9]
Mixed 4-carboxyphenyl/phenyl layers have been formed from
reduction of a mixture of the corresponding (pre-prepared)
benzoic acid/benzene diazonium salts on GC electrodes and
ferrocene methylamine subsequently coupled to the carboxylic
acid functionalised surface.[10] Comparison with analogous fer-
rocene-terminated thiol layers on gold reveals the functional-
ised GC electrodes to have lower electron transfer rates, sug-
gesting formation of multilayers, which is commonly observed
for diazonium derived organic layers.[11,12] Other mixed layers
have also been prepared for protein electrochemistry; mixed
polyethylene glycol (PEG) and oligo(phenylethynylene) (molec-
ular wire) layers have been prepared by reduction of a mixture
of PEG- and oligo(phenylethynylene)-based diazonium salts
onto GC electrodes. PEG was used as a group to stop the non-
specific adsorption of proteins onto the electrode and ferro-
cene methylamine or horseradish peroxidise was coupled to
the molecular wire, allowing direct electron transfer.[13] This
work did not focus on precise control of the partial coverage;
mixtures containing 5% (molar content) of the molecular wire
were used and the surface layers, interrogated by voltammetry,
shown to contain 3.7% covalently coupled ferrocene. Assess-
[a] E. J. Wright, Dr. M. Sosna, Dr. S. Bloodworth, Prof. P. N. Bartlett
Chemistry, Faculty of Natural and Environmental Sciences
University of Southampton, Southampton, SO17 1BJ (UK)
E-mail: S.Bloodworth@soton.ac.uk
P.N.Bartlett@soton.ac.uk
[b] Prof. J. D. Kilburn
School of Biological and Chemical Sciences
Queen Mary University of London, Mile End Road, London, E1 4NS (UK)
E-mail: j.kilburn@qmul.ac.uk
[c] Dr. M. Sosna
Present address:
Interdisciplinary Nanoscience Center (iNANO)
Aarhus University, Gustav Wieds Vej 14, DK-8000, Aarhus C (Denmark)
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201400246.
Chem. Eur. J. 2014, 20, 5550 – 5554
5550
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
ment of the extent of control
over surface coverage relied on
the assumption that the relative
surface coverages directly re-
flected the relative concentra-
tions of the two diazonium salts
in solution. Most recently, the
same group has investigated the
use of supramolecular interac-
tions between two diazonium
salts to form binary films with
a 1:1 composition.[14]
In our approach, we function-
alise GC by the oxidation of
mono-Boc-protected
diamines
which, following deprotection,
can be further modified using
solid-phase synthesis to achieve
the desired molecular architec-
tures.[15–20] The method of
carbon modification using amine
oxidation was introduced in
1990 by Barbier et al.,[21] but to
our knowledge, preparation of
mixed monolayers from amine
mixtures has not been attempt-
ed. In this Communication, we
report functionalisation of GC
electrodes with two-component
Scheme 1. Sequential electrochemical and solid-phase preparation of functionalised electrodes 6, 10, 11 and 17–
20. Subscripts a, b and c refer to n=1, 2 and 3, respectively. Prior to modification, 3 mm diameter (0.071 cm2)
glassy carbon (GC) rod electrodes were individually polished with silicon carbide paper (grade P1200, 3m) fol-
lowed by alumina lapping film (5 mm, 3m) and alumina slurries (1.0 and 0.3 mm) on polishing cloths. Reagents and
conditions: a) N-(2-aminoethyl)acetamide:mono-N-Boc-hexamethylenediamine (mono-N-Boc-HDA) in mixed solu-
tion (1:0, 9:1, 8:2, 1:1 or 0:1 ratio in MeCN) with a total amine concentration of 10 mm, nBu4N+ BF4À(150 mm),
2.1 V vs. Ag/AgCl, 180 s; b) 4.0m HCl in 1,4-dioxane, 1 h; c) 7, HBTU, iPr2NEt, DMF, 16 h; d) mCPBA, MeCN, 08C, 1 h
then RT, 1 h; e) 12a, b or c, HBTU, iPr2NEt, DMF, 16 h; f) cytochrome c (20 mm in phosphate buffer pH 7), 48C,
16 h. The ratio “Capped:Free” refers to the ratio of the concentrations of the two amines used in the initial elec-
trochemical immobilisation step (step a).
monolayers
derived
from
a mixed amine solution in which
one component is a mono-Boc-
protected diamine. Subsequent
Boc-removal results in a mixed
monolayer containing fractional coverage of a free amine, facil-
itating further stepwise coupling of spacer and terminal malei-
mide functionality. In this way, we obtain electrodes with con-
trolled fractional coverage of a maleimide group for cysteine
binding. An important feature of this strategy is that it gives
a modified electrode which can spontaneously react with the
engineered protein in buffer solution, allowing the most effi-
cient use of small quantities of purified enzyme. The influence
of maleimide coverage and the total length of the linker (con-
trolled by the introduction of a range of spacers) upon protein
binding is assessed using cytochrome c as a model redox pro-
tein. Cytochrome c from bovine heart was chosen due to an
accessible free cysteine at its surface (Cys 17) in proximity to
the haem redox cofactor.[22]
phenylselenyl group. Direct amide coupling of surfaces bearing
free amine functionality with 3-maleimidopropionic acid
(MPA)[23] would obviate the need for subsequent maleimide de-
protection (i.e., for selenyl cleavage), however coupling of MPA
with 5 was capricious in our hands and the use of 7, readily
prepared by phenylselenation of MPA (Supporting Informa-
tion), proved convenient since oxidative elimination of the se-
lenyl protecting group was facile under very mild conditions.[24]
With the free maleimide moiety thus exposed, electrodes 9
and 16a–c were finally treated with cytochrome c and its pres-
ence and quantity on the surface of the final redox protein-
functionalised electrodes 11 and 20a–c was investigated by
cyclic voltammetry (Supporting Information).
Figure 1a shows typical results for the voltammetry of the
cytochrome c modified electrodes. The baseline currents were
subtracted using the algorithm described in the Supporting In-
formation. This fits the background to the voltammogram
using anchor points lying away from the position of cycto-
chrome c peaks and using a B-spline function. The effective-
ness of this procedure was checked using an unmodified
glassy carbon electrode treated in the same way. Figure 1b
shows the background-subtracted voltammetry. In all cases,
the cytochrome c peaks were centred at ꢀ0.01 V versus SCE in
We first prepared a series of electrodes, each with a full mal-
eimide-terminated monolayer, but of varying length of tether
of the terminal maleimide group to the surface (Scheme 1). In
each case, the GC surface was functionalised by oxidation of
mono-N-Boc-1,6-hexanediamine (HDA), followed by N-depro-
tection to give 5. Direct HBTU-mediated coupling of 5 with 7,
or sequential coupling of a further N-Boc protected spacer 12
followed by N-deprotection before coupling with 7, gave elec-
trodes 9 and 16a–c, respectively, following cleavage of the
Chem. Eur. J. 2014, 20, 5550 – 5554
www.chemeurj.org
5551
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
layer of uniform thickness and binding between maleimide
and the Cys17 residue of cytochrome c is hindered since the
protein bulk cannot be sterically accommodated. Small increas-
es in coverage observed for electrodes 20b and 20c with the
longer linkers could result from greater disorder in the layer,
with a higher number of individual maleimide groups sterically
accessible. Figure 1c also shows analogous results obtained
with lower maleimide coverages (light bars), and we will return
to these data below.
For cytochrome c orientated towards the electrode by at-
tachment through Cys 17, the theoretical surface coverage, as-
suming the crystallographic dimensions and close packing of
the molecules on the surface, is about 11 pmolcmÀ2. This is
significantly larger than the values in Figure 1c for cytochrome
c coverage of 2–3 pmolcmÀ2 on electrodes 11 and 20a–c and
suggests that close packing of the maleimide linkers might
block their accessibility towards Cys 17 since accessibility is
compromised by a local concave protein surface topography.
To test this idea, we investigated the effect of diluting the cov-
erage of the maleimide linker on the electrode surface by
making mixed monolayers. The HDA linker and length of the
co-immobilised acetyl capped linker were selected based on
the cytochrome c crystal structure and location of Cys 17 (Sup-
porting Information).
Mixed two component monolayers on the GC surface were
prepared by electrochemical oxidation of a mixed solution of
N-(2-aminoethyl)acetamide and mono-N-Boc-HDA, by stepping
the potential to 2.1 V. From preliminary voltammetric investiga-
tion, it is known that at this potential both amines are oxidised
at a diffusion controlled rate, and therefore stoichiometric
amounts of the corresponding amine radicals are formed. If we
assume similar reactivities for the two radicals formed, we
expect the ratio of the surface coverages of the two amines to
be similar to the ratio of their solution concentrations. Follow-
ing N-deprotection, the resulting electrodes 2, 3, 4 and 5 are
therefore expected to bear surface monolayers containing, re-
spectively, 10, 20, 50 and 100% free amine functionality avail-
able for the subsequent solid-phase steps, reflecting the per-
centage component of mono-N-Boc-HDA in the first oxidation
step. Electrodes 2–5 were each coupled with spacer 12a
before N-deprotection of the spacer unit, coupling with 7, phe-
nylselenyl cleavage and binding of each member of the result-
ing electrode series 13a–16a with cytochrome c to give 17a–
20a (Scheme 1). Cyclic voltammetry of 17a–20a clearly indi-
cates that the surface concentration of maleimide in 13a–16a
leads to a pronounced effect on the amount of immobilised
cytochrome c (Figure 2, Supporting Information).[26] Dilution of
the binding group on the surface results in increased coverage
of the protein, reaching (7.5Æ1.4) pmolcmÀ2 for surfaces
modified from a mixture of amines containing 10% HDA. In an
attempt to maximise the cytochrome c coverage, a 1% HDA
coverage was also investigated, however the reproducibility of
the voltammetric signal of the corresponding final (cytochrome
c bound) electrodes was poor, and is not included in the
analysis.
Figure 1. a) Raw (black) cyclic voltammogram of electrode 17a overlaid with
baseline from a polished GC control electrode (red). b) Background-correct-
ed cyclic voltammogram of electrode 17a derived from baseline subtraction
from the raw data (Supporting Information). Voltammograms were recorded
at pH 7, in 20 mm phosphate buffer with 100 mm NaClO4 at 20 mVsÀ1. c)
Dark bars: Cytochrome c coverage on electrodes modified with a full mono-
layer of maleimide-terminated linker of variable length 11, 20a–c. Cyto-
chrome c coverage on a (maleimide-free) ‘capped’ electrode (6) is given for
comparison. Light bars: Cytochrome c coverage on electrodes modified with
a 10% mixture of maleimide-terminated linker of variable spacer length (10,
17a–c). Standard deviation (3 replicates of each electrode) is indicated.
agreement with published studies.[25] The surface coverage of
cytochrome c was estimated from the area under the curve.
Surface coverages of cytochrome c for full coverage with
maleimide linkers of different length (11, 20a, 20b and 20c)
are shown in Figure 1c (dark bars). The coverage on a control
electrode 6, obtained by treatment of a fully N-acetyl ‘capped’
surface (bearing no maleimide group) with cytochrome c is
also included. Similar values of coverage, 2–3 pmolcmÀ2, are
obtained, irrespective of the spacer used. Moreover, the
amount of cytochrome c on the surface in functionalised elec-
trodes 11 and 20a–c is comparable to that found on control
electrode 6. This suggests that non-specific binding is the pre-
dominant mode of protein retention on the electrode, perhaps
because the maleimide terminated linkers form a compact
Electrode 2, modified with a mixed monolayer containing
10% free amine, was thus chosen as a starting point to investi-
Chem. Eur. J. 2014, 20, 5550 – 5554
www.chemeurj.org
5552
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
electrode surface and independently vary the length of the
linker to this binding group. The efficacy of this approach has
been demonstrated using bovine heart cytochrome c as
a redox-active model protein. Our results show that for full sur-
face coverage of maleimide the final surface coverage of the
protein is close to that for non-specific adsorption and we pos-
tulate that this is because the high surface coverage of malei-
mide leads to significant steric hindrance towards the reaction
of the cysteine on the cytochrome c surface. Consistent with
this hypothesis, when the maleimide groups on the electrode
surface are diluted by co-immobilisation of mono-Boc-protect-
ed diamine with N-(2-aminoethyl)acetamide, we find signifi-
cantly higher coverages for cyctochrome c, as judged by the
electrochemistry of the immobilised protein. Further support
for this interpretation is provided by study of the effect of the
length of the electrode surface attachment linker for malei-
mide, upon the electroactive coverage. We find that for the
shortest linker the coverage is again lower, and steric restric-
tions upon coupling between maleimide and cysteine are con-
sistent with the crystal structure of the protein (Supporting
Information).
Figure 2. Cytochrome c coverage on electrodes of fixed tether length 17a–
20a, derived from electrodes 2–5, respectively. ‘HDA% molar content’ refers
to the composition of mixed amine solutions used to prepare 2 (10% HDA),
3 (20% HDA), 4 (50% HDA) and 5 (100% HDA) in the first surface-functional-
isation step. Standard deviation (3 replicates of each electrode) is indicated.
gate the effect of the total length of a maleimide-terminated
linker on binding of cytochrome c. Four different variants were
prepared; the shortest linker was obtained by coupling 7 di-
rectly with 2 to yield 8, and for longer linkers the coupling of 7
was preceded by the introduction of a spacer (one of 12a–c)
to eventually give maleimide-terminated electrodes 13a–c.
Upon reaction of 8 and 13a–c with the protein, surfaces 10
and 17a–c (also denoted as ‘no spacer’ and n=1–3, respec-
tively) were obtained, with cytochrome c bound via four differ-
ent tethers of distinct length. Surface coverage of cytochrome
c upon each electrode was determined by cyclic voltammetry
(Figure 1c, light bars, Supporting Information). The increase in
the length of the linker is initially followed by a similar trend in
the coverage of cytochrome c. The coverage doubles from
electrode 10 to 17a, that is, after introduction of the shortest
spacer, followed by a moderate increase with further elonga-
tion of the linker (17b). The use of the longest linker, however,
results in a drop in the cytochrome c coverage (17c) to a level
comparable with that of 17a. Whilst a longer, more flexible
linker can improve the access and therefore the binding of
a maleimide group to the cysteine residue of cytochrome c,
the trend appears to have a limitation. The longest linkers are
prone to the formation of more disordered layers with higher
likelihood of entanglement of aliphatic chains. This in turn can
cause a decreased efficiency of the subsequent binding steps,
in our case coupling of 7 to the terminal amine as well as
binding of cytochrome c to the terminal maleimide. Similar be-
haviour was previously reported for electrodes modified with
tethered metal complexes.[19]
Our results emphasise the importance of careful design of
the electrode surface architecture for efficient attachment of
cysteine-modified proteins, and demonstrate that this can be
achieved by the flexible synthetic approach described herein.
Experimental Section
Detailed experimental procedures for solid-phase preparation of
electrodes 6, 10, 11 and 17–20, cyclic voltammograms, and meth-
ods of peak extraction for electrodes 6, 10, 11 and 17–20, are pro-
vided in the Supporting Information.
Acknowledgements
We thank H. Hamzah for running background voltammograms
on the unmodified glassy carbon electrodes. This work was
funded by the European Research Council, project 3D-Nano-
BioDevice NMP4-SL-2009–22925.
Keywords: cyclic voltammetry · electrochemistry · proteins ·
solid-phase synthesis
[1] J. J. Gooding, F. Mearns, W. Yang, J. Liu, Electroanalysis 2003, 15, 81–96.
[2] N. K. Chaki, K. Vijayamohanan, Biosens. Bioelectron. 2002, 17, 1–12.
[3] A. J. Downard, Electroanalysis 2000, 12, 1085–1096.
[4] J. Heeg, U. Schubert, F. Kꢁchenmeister, Fresenius J. Anal. Chem. 1999,
365, 272–276.
[5] C. D. Bain, G. M. Whitesides, J. Am. Chem. Soc. 1988, 110, 6560–6561.
[6] C. D. Bain, G. M. Whitesides, J. Am. Chem. Soc. 1989, 111, 7164–7175.
[7] P. E. Laibinis, M. A. Fox, J. P. Folkers, G. M. Whitesides, Langmuir 1991, 7,
3167–3173.
[8] C. D. Bain, E. B. Troughton, Y. T. Tao, J. Evall, G. M. Whitesides, R. G.
Nuzzo, J. Am. Chem. Soc. 1989, 111, 321–335.
[9] G. Liu, M. Chockalingham, S. M. Khor, A. L. Gui, J. J. Gooding, Electro-
analysis 2010, 22, 918–926.
In conclusion, we have described a generic, modular ap-
proach to the modification of carbon electrodes with malei-
mide groups. The maleimide group is chosen because it under-
goes spontaneous reaction with free cysteine groups at room
temperature in aqueous buffer, making it an excellent choice
for the immobilisation of proteins genetically engineered to
contain free cysteine at suitable surface positions. By using
electrochemical oxidation of mixtures of mono-N-Boc-diamine
and N-(2-aminoethyl)acetamide, followed by N-deprotection
and subsequent solid-phase synthesis steps, we have been
able to control dilution of the maleimide group upon the GC
[10] G. Liu, J. Liu, T. Bçcking, P. K. Eggers, J. J. Gooding, Chem. Phys. 2005,
319, 136–146.
Chem. Eur. J. 2014, 20, 5550 – 5554
www.chemeurj.org
5553
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
[11] C. Saby, B. Ortiz, G. Y. Champagne, D. Bꢂlanger, Langmuir 1997, 13,
6805–6813.
[22] N. Mirkin, J. Jaconcic, V. Stojanoff, A. Moreno, Proteins Struct. Funct.
Bioinf. 2007, 70, 83–92.
[23] X. Zhang, G. Sun, M. Hovestꢃdt, V. Syritski, N. Esser, R. Volkmer, S. Jan-
ietz, J. Rappich, K. Hinrichs, Electrochem. Commun. 2010, 12, 1403–
1406.
[12] P. Doppelt, G. Hallais, J. Pinson, F. Podvorica, S. Verneyre, Chem. Mater.
2007, 19, 4570–4575.
[13] G. Liu, J. J. Gooding, Langmuir 2006, 22, 7421–7430.
[14] A. L. Gui, H. M. Yau, D. S. Thomas, M. Chockalingam, J. B. Harper, J. J.
Gooding, Langmuir 2013, 29, 4772–4781.
[15] J. M. Chrꢂtien, M. A. Ghanem, P. N. Bartlett, J. D. Kilburn, Chem. Eur. J.
2008, 14, 2548–2556.
[16] J. M. Chrꢂtien, M. A. Ghanem, P. N. Bartlett, J. D. Kilburn, Chem. Eur. J.
2009, 15, 11928–11936.
[17] M. A. Ghanem, J. M. Chrꢂtien, J. D. Kilburn, P. N. Bartlett, Bioelectrochem-
istry 2009, 76, 115–125.
[18] M. A. Ghanem, J. M. Chretien, A. Pinczewska, J. D. Kilburn, P. N. Bartlett,
J. Mater. Chem. 2008, 18, 4917–4927.
[19] A. Pinczewska, M. Sosna, S. Bloodworth, J. D. Kilburn, P. N. Bartlett, J.
Am. Chem. Soc. 2012, 134, 18022–18033.
[20] M. Sosna, J.-M. Chretien, J. D. Kilburn, P. N. Bartlett, Phys. Chem. Chem.
Phys. 2010, 12, 10018–10026.
[24] M. L. Durrant, E. J. Thomas, J. Chem. Soc. Perkin Trans. 1 1984, 901–904.
[25] A. Ion, F. Banica, J. Solid State Electrochem. 2001, 5, 431–436.
[26] Coupling of ferrocenyl hexanethiol to a GC surface functionalised with
a full maleimide-terminated monolayer was attempted in order to es-
tablish a model with which to then identify any impact of the mixed
monolayer upon the efficiency of thiol binding. However, voltammetric
investigation (electrode cycling in CH3CN with nBu4N+BF4Àfrom 0 to
1.1 V vs. Ag/AgCl for an initial 10 cycles at 100 mVsÀ1) indicated that
non-specific absorption of the redox-active species compromises accu-
rate measurement of coupling efficiency since a large and broad-shoul-
dered current at 0.74 V vs. Ag/AgCl, not seen in the reverse scan, ap-
pears which we attribute to loosely bound multiple ferrocene species
that desorb from the surface quickly.
[21] B. Barbier, J. Pinson, G. Desarmot, M. Sanchez, J. Electrochem. Soc. 1990,
137, 1757–1764.
Received: January 21, 2014
Published online on April 10, 2014
Chem. Eur. J. 2014, 20, 5550 – 5554
www.chemeurj.org
5554
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim