A fullerene helical peptide: synthesis, characterization and formation of self-assembled monolayers on gold surfaces

Article abstract of DOI:10.1039/c8nj04599b

The synthesis of a C₆₀-peptide using “clickable” fullerene and peptide derivatives is described. The peptide is composed of a repeating sequence of α-aminoisobutyric acid (Aib) and two l-alanine (Ala) residues, promoting the formation of a helical conformation, which has been confirmed by IR absorption, NMR and circular dichroism measurements. In addition, the presence of a lipoyl moiety, at the end of the peptide sequence, allows the formation of self-assembled monolayers of the C₆₀-peptide and the parent peptide on a gold surface. A C₆₀-alkyl derivative was also prepared to compare the self-assembly properties of fullerene derivatives containing peptides or alkyl chains. The fullerene assembly on a gold substrate characterized by quartz crystal microbalance and by cyclic voltammetry show that the monolayers containing alkyl chains are slightly less well packed than the peptide monolayers. Finally, polarization modulation reflection adsorption infra-red spectroscopy measurements indicate that the C₆₀-peptide tends to be more vertical than the parent peptide which could originate from complementary C₆₀-C₆₀ and helical-helical interactions.

Full text of DOI:10.1039/c8nj04599b

NJC
View Article Online
View Journal | View Issue
PAPER
A fullerene helical peptide: synthesis,
characterization and formation of self-assembled
monolayers on gold surfaces†
Cite this: New J. Chem., 2018,
42, 19423
Houssein Nasrallah,a Jad Rabah,a Van Bui-Thi-Tuyet,a Krystyna Baczko,a
Helene Fensterbank,a Flavien Bourdreux,a Anne-Marie Goncalves,a
´ `
b
c
a
Valerie Declerck,
Souhir Boujday,c Vincent Humblot,
Karen Wright,
*
´
a
Anne Vallee*a and Emmanuel Allard
*
´
The synthesis of a C60-peptide using ‘‘clickable’’ fullerene and peptide derivatives is described. The peptide is
composed of a repeating sequence of a-aminoisobutyric acid (Aib) and two L-alanine (Ala) residues,
promoting the formation of a helical conformation, which has been confirmed by IR absorption, NMR and
circular dichroism measurements. In addition, the presence of a lipoyl moiety, at the end of the peptide
sequence, allows the formation of self-assembled monolayers of the C60-peptide and the parent peptide on
a gold surface. A C60-alkyl derivative was also prepared to compare the self-assembly properties of fullerene
derivatives containing peptides or alkyl chains. The fullerene assembly on a gold substrate characterized by
quartz crystal microbalance and by cyclic voltammetry show that the monolayers containing alkyl chains are
slightly less well packed than the peptide monolayers. Finally, polarization modulation reflection adsorption
infra-red spectroscopy measurements indicate that the C60-peptide tends to be more vertical than the
parent peptide which could originate from complementary C60–C60 and helical–helical interactions.
Received 10th September 2018,
Accepted 19th October 2018
DOI: 10.1039/c8nj04599b
rsc.li/njc
on solid surfaces.2 Fullerene C60-based self-assembled monolayers
have been involved in the fabrication of solar cells as the electron
Introduction
Since their discovery, fullerene C60 and its derivatives have selective layer, noticeably improving the efficiency of the device,3
attracted the interest of many researchers due to their out- in field-effect transistors4 and in molecular rectifiers.5 The
standing physical and chemical properties, and have promising immobilization of donor–fullerene systems on surfaces for
applications in the fields of life and materials sciences. In the construction of photoelectrochemical devices has also been
particular, the surface modification with C60-based thin films considered.6 The occurrence of photo-induced energy and/or
or monolayers allows the transfer of their unique properties electron transfer in these donor–fullerene systems was used to
to solid surfaces, which may lead to practical applications generate photocurrent. More recently, it has been proposed
especially in organic electronics. The main methods to prepare that the charge separated state formed after illumination of
monolayers or thin films of fullerenes are based on the donor–fullerene SAMs could also function as a hole-blocking
Langmuir and Langmuir–Blodgett deposition techniques, as well layer and an electron transporting layer for photovoltaic cell
as the formation of self-assembled monolayers (SAMs) on solid applications.7
surfaces.1,2 The latter approach has attracted much attention,
The design of [60]fullerene-based self-assembled mono-
since self-assembled monolayers are spontaneously formed by layers on a variety of substrates, in particular gold conductive
chemisorption, giving rise to robust and well-defined structures surfaces, has been extensively studied.8 Among the approaches
that have been proposed to form SAMs containing fullerene C60
on gold surfaces, the direct strategy yields the most stable as
a
´
ILV, UMR-CNRS 8180, Universite de Versailles-Saint-Quentin-en-Yvelines,
well as the most homogeneous monolayers.2 This method
relies on the use of C60 derivatives fitted with anchoring units
such as thiols or disulfides, which are well-known to bind
strongly to gold surfaces.9 Usually, fullerene C60 and its sulfur
anchoring unit are connected by a spacer, including alkyl
chains and p-conjugated systems. Very often in the literature,
SAMs of molecular assemblies including fullerene derivatives
45 avenue des Etats-Unis, 78035 Versailles cedex, France.
E-mail: emmanuel.allard@uvsq.fr
b
c
´
CP3A Organic Synthesis Group, ICMMO, UMR-CNRS 8182, Universite Paris Sud,
´
Universite Paris-Saclay, 15 rue Georges Clemenceau 91405, Orsay cedex, France
´
´
´
Sorbonne Universite, UMR-CNRS 7197, Laboratoire de Reactivite de Surface (LRS),
4 Place Jussieu, F-75005 Paris, France
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c8nj04599b
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018 New J. Chem., 2018, 42, 19423--19432 | 19423

View Article Online
Paper
NJC
are depicted as idealized representations without defects. How- allows the grafting of our fullerene building block to the peptide in
ever, Tour and co-workers have shown that the immobilization one step, unlike previous strategies in which multiple steps were
of fullerene C60 on gold surfaces leads to binding anomalies necessary to complete the required fullerene–peptide conjugates.14
where the desired monolayers can coexist with multi-layers In addition, a reference compound (C60-alkyl), in which an alkyl
and/or head-to-tail assemblies.10 These anomalies could modify chain replaced the peptide, was also prepared in order to compare
the properties of the devices, thereby limiting the use of C60-SAMs the self-assembly properties of the corresponding fullerene
in a number of applications.
derivatives. The SAMs of the starting peptide, C60-alkyl and
C60-peptide were then immobilized on a gold surface and their
In this context, we propose to use peptides as surface
structuring agents, as it has been shown that peptides adopting properties were studied using surface characterization and
a helical conformation in solution and in the solid state may electrochemical techniques.
form well-organized and compact monolayers. Moreover, the
diameters of the fullerene C602 and of a peptide helix, whatever
its nature (310 or a-helix),11 are of the same order of magnitude
(close to 1.0 nm). In this case, organization within the mono-
Results and discussion
Synthesis
layer should be governed by two types of interactions, C60–C60
and peptide–peptide interactions, which could lead to a Fullerene building block 5 was prepared in a similar way to our
reduction in head-to-tail anomalies. In addition, a helical previously reported fullerene platform, developed for sequential
peptide appears to be a good electron mediator compared to orthogonal transformations (Scheme S1, ESI†).15 The alkyl spacer
an alkyl chain. Electron transfer through a peptide bridge is between the alkyne moiety and the C60 unit is shorter than in the
also strongly asymmetric, as transfer along the dipole moment previously described platform, in order to reduce the C60-surface
in a helical peptide is much faster than against the dipole distance in the final assembly. In the present C60 derivative, a
moment. These properties could be very important for organic maleimide moiety was introduced but remained protected. Lipoyl
electronic applications, especially for the design of electron azide 7 was prepared in three steps from 4-aminobutanol. Reaction
selective interfaces, in electrochemical photocurrent generation with hydrogen bromide followed by substitution with sodium azide
as well as in molecular rectifiers.
afforded an azido amine18 which was then coupled with (R)-lipoic
The combination of peptides and C60 has been largely explored, acid in the presence of HATU (Scheme S2, ESI†). Hexapeptide 15
mainly for biological applications and in the field of materials was synthesised step-by-step in solution (Scheme S3, ESI†), using
science.12,13 However, immobilization of C60-peptides on surfaces O-(7-azabenzo-1,2,3-triazol-1-yl)-1,1,3,3-tetramethyluronium hexa-
in which the peptide adopts a helical conformation has been fluorophosphate (HATU) as the coupling reagent and an Na-Boc
scarcely studied. To our knowledge, only one example has been protecting group strategy. The C-terminal alkyl azide chain was
reported by Kimura. In this study, porphyrin- and fullerene- introduced by reaction of the N-Boc-alanine succinimide ester
terminated 16-mer helical peptides were used to form homo- with 4-aminobutanol, followed by mesylation and substitution
geneous binary SAMs. The formation and characterization of with sodium azide, to give the desired protected azido-amide 9.
homogeneous SAMs based exclusively on fullerene terminated Peptide chain elongation was performed by sequential deprotec-
peptides were not reported in this paper.14
tion of the N-terminal Boc group with trifluoroacetic acid and
Herein, we describe the synthesis of a C60-peptide using coupling of the following amino acid residue in the presence of
‘‘clickable’’ fullerene and peptide derivatives. The fullerene HATU. The hexapeptide was finally substituted with (R)-lipoic
building block bears an alkyne moiety and a protected maleimide. acid at its N-terminal.
This derivative can be easily functionalized with azido and thiol
The reaction conditions for the copper-catalyzed alkyne-
derivatives after deprotection of the maleimide,15 which represents azide cycloaddition (CuAAC) reaction between the fullerene
a significant advantage over previously used derivatives that allow alkyne and the peptide azide or alkyl azide were investigated.
grafting of only one unit.14 Moreover, we used a simple hexa- Since the presence of sulfur atoms in substrates may lead to
peptide containing the repeating sequence a-aminoisobutyric partial deactivation of the CuAAC catalytic system,15a,19 we
acid (Aib) and two L-alanine (Ala) residues. The choice of the performed preliminary experiments using copper(I) salts and
peptide length, its composition and the position of the Aib various ligands in dichloromethane to determine the optimum
residues in the sequence were determined by the conforma- conditions. From these experiments, it appeared that the best
tional properties of Ca-tetrasubstituted amino acids, known to yields were obtained with bulky ligands such as (hexabenzyl)tren20
promote the formation of 310-helical structures in very short and tris-(benzyltriazolylmethyl)amine (TBTA).21 The preferred con-
peptides.16 This hexapeptide carries a lipoyl moiety (at the ditions for the CuAAC reaction between the alkyne C60 derivative 5
N-terminus) which acts as a bidentate ligand on the gold surface, and the azide-peptide 15 or lipoyl azide 7 involved the use of the
and an azide moiety (at the C-terminus) allowing a copper- preformed complex [Cu(I)ligand(hexabenzyl)tren]Br, prepared
catalyzed alkyne-azide 1,3-dipolar cycloaddition (CuAAC) reaction from CuBr and the hexabenzyltren ligand, in dichloromethane at
with the fullerene building block. The CuAAC reaction is a very room temperature under an inert atmosphere (Scheme 1). The
powerful synthetic tool in fullerene chemistry, which has allowed fullerene–alkyl and fullerene–peptide adducts were isolated in 40%
the efficient synthesis of a variety of fullerene derivatives, some of and 60% yields respectively, consistent with previous results
which cannot be obtained by classical methods.17 This approach obtained for this type of derivatization. This approach allowed
19424 | New J. Chem., 2018, 42, 19423--19432 This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018

View Article Online
NJC
Paper
Scheme 1 Preparation of C60-alkyl and C60-peptide. Reagents and conditions: (i) CuBr, hexabenzyltren, CH2Cl2, rt, Ar atmosphere, 2 days (C60-alkyl:
40%; C60-peptide: 60%).
the preparation of a significant amount of both the fullerene hydrogen bonds) and a weaker band at 3430 cmÀ1 (free N–H
building block, the fullerene–alkyl adduct and the fullerene– groups). The amide I region (Fig. S2, ESI†) of the derivatives
peptide, and exhibits a higher yield than a previously reported shows a strong absorption band with its maximum at 1655–
method.14 Complete characterization of these derivatives using 1660 cmÀ1, in accordance with the carbonyl stretching of a
NMR and mass spectrometry was carried out.22,23
peptide adopting a helical conformation. However, it is difficult
to determine the nature of the helix using this technique since
the absorption maximum of a 310-helix and a-helix is close to
Solution conformational analysis
The solution conformation of the reference hexapeptide 15 and the 1666–1663 cmÀ1 and 1663–1658 cmÀ1 region, respectively.25
of the C60-peptide was studied by using a combination of infra- While the carbonyl stretching band is located in the a-helix region,
red absorption, NMR and circular dichroism measurements.
the 310-helix conformation may also contribute to our short
IR absorption spectroscopy. The FT-IR spectra of the hexa- peptide structures.
peptide 15 and of the C60-peptide were recorded in CDCl3 at a
1H NMR spectroscopy. The NMR analysis of the starting
concentration of 1 mM. The absorption bands of the two hexapeptide 15 was carried out in CDCl3 at a concentration of
derivatives are located at the same position in the amide 50 mM (see below) using a combination of COSY and ROESY
A and amide I region, indicating as previously reported that experiments (Fig. S3 and S5–S7, ESI†). The use of these com-
the conformation of the peptide is not modified by the plementary 2D experiments allowed the complete assignment
presence of the C60 unit.24 In the N–H stretching region of all proton resonances. The cross peaks observed in the
(Fig. S1, ESI†), the FT-IR spectra show a strong absorption ROESY spectrum between NH–NH (i - i + 1) protons permitted
band at 3315 cmÀ1 (N–H groups in intramolecular CQOÁ Á ÁH–N the identification of all the amide protons of the peptide.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018 New J. Chem., 2018, 42, 19423--19432 | 19425

View Article Online
Paper
NJC
Fig. 1 Stack plot of the partial 1H NMR spectra in which DMSO-d6 was added in a 1 mM solution of the peptide 15 in CDCl3 (300 MHz, 298 K). The peaks
labelled with an asterisk are due to satellite bands.
These NH–NH (i - i + 1) and some aH–NH (i - i + 1) positions from the N-terminus (denoted as N(I)H and N(II)H)
interactions also indicate that the peptide adopts a prevailing underwent significant downfield shifts (Dd = 1.25 and 0.47 ppm,
folded structure. To complement the 2D experiments, DMSO-d6 respectively) while the other amide protons displayed minimal
titrations were performed on 1 mM solutions of the hexapeptide shifts (Dd = 0.05 to 0.19 ppm). This suggests a 310-helical con-
15 in CDCl3 (Fig. 1 and Fig. S11, ESI†). Similar titration experi- formation in solution where most NH groups (from N(III)H
ments have already been used to verify that peptides containing onward) are in a stable, intramolecular H-bonding network, while
Aib moieties adopt a 310-helical conformation in solution.26 Upon the two NH groups closest to the N-terminus are exposed to
the addition of the hydrogen bond acceptor DMSO-d6, the two intermolecular interactions. In addition, the concentration-
1
amide NH protons of the Aib and Ala residues at the 1st and 2nd dependence of the H NMR resonances of the amide NH groups
Fig. 2 Stack plot of the partial 1H NMR spectra of peptide 15 taken at varying concentrations from 1 mM to 100 mM (300 MHz, CDCl3, 298 K). The peaks
labelled with an asterisk are due to satellite bands.
19426 | New J. Chem., 2018, 42, 19423--19432 This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018

View Article Online
NJC
Paper
of 15 in CDCl3 was examined (Fig. 2 and Fig. S9, ESI†). Such of C60-peptide and C60-alkyl in dichloromethane. To determine
experiments have already been used to study the aggregation of the contribution of the peptide to the quality of the fullerene
helical peptides in non-polar aprotic media.27 As previously SAM, the SAMs were characterized in depth.
reported, the concentration-dependence study followed the trend
Cyclic voltammetry (CV). The molecular packing of the
observed during the DMSO dilution studies, with the N(I)H and fullerene SAMs was evaluated by CV (blocking) experiments
N(II)H protons being more involved in intermolecular interactions using bare and modified gold substrates as the working elec-
undergoing significant shifts as the concentration was raised from trode. Fig. 5 shows the CV responses of K3[Fe(CN)6] aqueous
1 to 100 mM (Dd = 1.51 and 1.17 ppm, respectively), while the other solution on different substrates. The peaks for the ferricyanide/
NH protons, involved in the H-bonding network, showed smaller ferrocyanide redox signal are clearly observed for the naked
shifts (Dd = 0.03 to 0.23 ppm, respectively).
gold while these peaks are nearly abolished for the gold surface
For the C60-peptide, NMR analyses were performed in CDCl3 at functionalized with C60-peptide molecules, highlighting the
a concentration of 30 mM (Fig. S3, ESI†). These experiments compactness of the C60-peptide SAM. In contrast, in the case
allowed the complete assignment of all proton resonances except of the C60-alkyl SAM the signal is lower than that of naked gold
those belonging to the amide protons, as some were too close to but is still present. The gold surface is accessible to the redox
allow clear identification using the ROESY experiment. In order to probe which implies that the C60-alkyl molecules formed a less
assign the amide protons and to confirm the helical conformation dense monolayer than C60-peptide molecules.30
of the C60-peptide, NMR experiments were further conducted in
Quartz crystal microbalance (QCM). To monitor the assembly
C5D5N for this compound, along with the reference hexapeptide 15 of fullerene derivatives on gold surfaces, QCM measurements were
(Fig. S14–S21, ESI†). For both compounds, the amide protons are performed in situ using dichloromethane as a solvent (Fig. 6). The
well-dispersed and well-defined (Fig. 3 and 4). The use of com- amount of C60-peptide and C60-alkyl in the SAMs was calculated
plementary 2D experiments (COSY, NOESY and ROESY) allowed from the frequency shifts by applying the Sauerbrey equation. After
the complete assignments of all proton resonances for 15 and rinsing the quartz diskette with dichloromethane, the measured
C60-peptide. In particular, the NOESY spectrum of 15 showed two frequency variations are 475 ng cmÀ2 and 301 ng cmÀ2, corres-
aH-NH (i - i + 2) cross peaks, suggesting specifically a 310-helical ponding to densities of 2.6 Â 10À10 and 2.15 Â 10À10 mol cmÀ2
,
structure. In addition, the cross peaks observed in the ROESY which are close to the reported values for fullerene compounds.7,31
spectrum of the C60-peptide between the NH–NH (i - i + 1) These results suggest a higher density of molecules adsorbed on
protons demonstrated that the peptide adopts a similar prevailing, the surface with the C60-peptide. The lower C60-alkyl SAM coverage
likely helical, conformation to the reference hexapeptide 15.
compared to the C60-peptide SAM is consistent with the less dense
Circular dichroism. A far-UV CD spectrum of the hexapep- monolayer observed by cyclic voltammetry which shows clearly
tide 15 was acquired in MeOH solution (Fig. S22, ESI†). A strong that the peptide moiety allows the formation of a more densely
negative Cotton effect at 203 nm was observed. It is widely packed layer.
recognised that negative Cotton effects below 200 nm (peptide
p–p* transition) indicate an unordered peptide conformation,
X-ray photoelectron spectroscopy (XPS). The C60-peptide and
C60-alkyl SAMs on gold were characterized using XPS to assess
while this band is shifted in peptides folded into helical the SAM composition and the bonds with the gold surface.
conformations to 203–208 nm.28 These results suggest an at The XPS spectrum of the hexapeptide reference molecule 15
least partially helical conformation for 15, consistent with adsorbed on gold was also recorded to determine how the
previous studies of short Aib-containing peptides.29
fullerene moiety influences the organization of the SAM. For
the three samples, after SAM formation, the Au4f peak
decreased in intensity while the C1s, O1s, N1s and the S2p
peaks appeared, indicating the presence of an additional layer
on the surface. High resolution N1s and S2p XPS signals are
Self-assembled monolayer formation and characterization
The C60-peptide and C60-alkyl were self-assembled on gold
surfaces by exposing a gold coated glass substrate to a solution
Fig. 3 1H NMR spectrum of 15 in C5D5N (300 MHz, 298 K).
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018 New J. Chem., 2018, 42, 19423--19432 | 19427

View Article Online
Paper
NJC
Fig. 4 1H NMR spectrum of the C60-peptide in C5D5N (300 MHz, 298 K). The peaks labelled with an asterisk are due to solvent residues.
Fig. 6 Resonance frequency changes of the Au quartz crystal during
C60-alkyl and C60-peptide SAM formation;
a 0.1 mM solution of
C60-alkyl and C60-peptide in dichloromethane was flowed for 55 min.
Fig. 5 CVs of K3[Fe(CN)6] for naked gold electrode (black), C60-alkyl
(blue) and C60-peptide (red) SAMs on a gold electrode.
the molecule N-amide/N-azide is equal to 3.5, above the expected
value (2.3). This can be explained by partial decomposition upon
shown in Fig. 7. The N1s peak of the peptide azide terminated X-ray radiation of the azide moiety, leading to a contribution at a
SAM presents four contributions at 400.0, 400.3, 401.3 and lower bonding energy probably included in the amide contribution
404.4 Æ 0.1 eV. The major contribution at 400.0 Æ 0.1 eV is at 400 eV.33
attributed to the N–H amide groups of the peptide (B78%).32
In the case of the C60-peptide or C60-alkyl SAMs, the N1s
The presence of the azide function is unambiguously revealed peaks are best fitted with three contributions. The main con-
by the three other components at 400.3, 401.3 and 404.4 Æ 0.1 eV.33 tribution at 399.9 Æ 0.1 eV is assigned to the N–H of the amide,
One can note here that the relative intensity ratio characteristic of to the OQC–N– of the maleimide or to the N–N–C of the
19428 | New J. Chem., 2018, 42, 19423--19432 This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018

View Article Online
NJC
Paper
Fig. 7 High resolution S2p and N1s XPS spectra of C60-alkyl, C60-peptide and hexapeptide 15 SAMs (from the top to the bottom).
Table
1
XPS sulfur contribution comparison for hexapeptide 15,
triazole. The two contributions at higher binding energies of
400.6 eV and 401.8 Æ 0.1 eV in an approximate ratio of 1 : 1 are
attributed to the triazole NQN–N–C and NQN–N–C, respec-
tively, showing that the molecules were intact when adsorbed
on the gold surface. To explore the binding mode of the
C60-peptide and C60-alkyl SAMs on gold
S2p components
Hexapeptide 15
C60-peptide
C60-alkyl
% S bound
% S unbound
48.5
37.4
14.1
47.6
36.9
15.5
73.3
26.7
molecules on the gold surfaces, the S2p signals were recorded. % S oxidized
The S2p signals of the two SAMs with a peptide moiety are very
similar and are best fitted with four S2p3/2,1/2 doublets. The first
two doublets at a lower binding energy e.g. 161.1 Æ 0.1 eV
Polarization modulation reflection adsorption infrared
(green) and 162.0 Æ 0.1 eV (S2p3/2) (blue) are characteristics of spectroscopy (PM-RAIRS). Information on the orientation and
multi-coordinated sulfur bonds to the gold surface and thiolate– order of molecules in the three SAMs (hexapeptide 15,
gold bonds, respectively.34 In the two SAMs, a third S2p signal at
C60-peptide and the C60-alkyl SAMs) were then probed by
163.6 Æ 0.1 eV (orange) is present, attributed to disulfide sulfur PM-RAIRS. First, the general spectroscopic profiles for the
atoms, suggesting that a fraction of the disulfide groups are not two molecules either as bulk, recorded by IR-ATR, or after
bonded to the surface despite the intensive rinsing of the assembly on gold, probed by PM-RAIRS, are shown in
samples.34,35 Also, a broad S2p peak showing a significant Fig. S23–S25 (ESI†). The ATR and PM-RAIRS profiles are com-
presence of unbound sulfur was systematically observed in XPS parable, which suggests successful adsorption on the gold
for the helical peptide adsorbed on gold.36 Finally, a small doublet surface. In the three cases, there is a quite good agreement
at a higher binding energy, 168.1 Æ 0.1 eV (red), is characteristic of between both spectra, and the differences observed can be
oxidized sulfur. The presence of unbound disulfide can originate explained by the specificity of both IR techniques. While ATR
from different species in the SAMs. In the case of fullerene provides information from the bulk, the RAIRS experiments,
derivative SAMs, the strong attraction between fullerene C60 units which follow metal surface selection rules (MSSRs),37 imply
and the gold surface can produce head-to-tail arrangements and that absorption bands with a transition moment perpendicular
disorganised multi-layers.10 However, the unbound disulfide to the surface are enhanced, whereas parallel modes experience
proportion in the S2p signal is of the same order of magnitude the opposite effect. The PM-RAIRS spectrum of the hexapeptide
in the hexapeptide 15 and C60-peptide SAMs as shown in Table 1. 15 adsorbed on gold is shown in Fig. 8. The band at 2093 cmÀ1
,
This finding indicates that the head to tail arrangements of the characteristic of the azide moiety, confirmed the formation of
C60-peptide are not likely to be the major contribution of unbound the hexapeptide 15 SAM.38,39 The spectrum is dominated by the
thiols. The presence of unbound thiols in the peptide and amide I and II bands at 1658 and 1542 cmÀ1 respectively
C60-peptide SAMs could originate from lateral interactions characteristic of a helical peptide.26a The upshift of the amide
between peptide chains or fullerene–fullerene contacts that I band on the SAM compared to the bulk (Fig. S24 and S25,
hamper the binding of the thiol to the gold surface.
ESI†) can be attributed to vibrational coupling between the
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018 New J. Chem., 2018, 42, 19423--19432 | 19429

View Article Online
Paper
NJC
are calculated to be 48.5 and 37.4 Æ 21 for the hexapeptide 15 and
C60-peptide SAMs, respectively. The tilt angle of the peptide SAM is
very close to the tilt angle reported in the literature for a short
chain peptide (491 for a peptide containing the [Ala–Aib]4
sequence).14,26a,42 Surprisingly, the tilt angle of the C60-peptide
SAM is closer to the tilt angle of a longer peptide (361 for a peptide
containing the [Ala–Aib]16 sequence). This result suggests that the
interactions between the fullerene units are added to the inter-
actions between the short peptides, leading to intermolecular
interactions that can be assimilated to the stronger intermolecular
interactions of longer peptides, which finally result in a greater
vertical orientation (Fig. 9).42 Interestingly, the presence of the C60
moiety increases the density of the monolayer compared to the
hexapeptide 15 since the C60-peptide SAMs are more densely
packed than the peptide SAMs.43
Fig. 8 PM-RAIRS spectra of the C60-alkyl, C60-peptide and hexapeptide
15 SAMs (from top to bottom).
Conclusion
In summary, we have prepared a fullerene-terminated helical
peptide and a fullerene–alkyl derivative through the use of
copper-catalyzed alkyne-azide cycloaddition reactions. The peptide
azide and alkyl azide were grafted on a fullerene platform that
bears on one side an alkyne unit, and a protected maleimide
amide groups of the nearby helices.40 This suggests the forma-
tion of a relatively dense monolayer of molecules on the sur-
face. The C60-peptide SAM PM-RAIRS spectra highlight the
amide I and II vibration bands at very similar wave-numbers
to that in the case of the hexapeptide 15 SAM, suggesting the
formation of a similar dense layer on the surface and confirm-
ing the dense layer observed by cyclic voltammetry.
Additional bands characteristic of the C60-peptide molecule
are also present. Three bands characteristic of the C60 malonate
at 1740, 1267 and 1181 cmÀ1 are clearly seen in the spectrum.41
The bands at 1740 and 1267 cmÀ1 correspond to the malonate
CQO and C–O stretching, respectively, while the band at
1181 cmÀ1 can be assigned to the o3 C60 cage vibration.7 Other
bands characteristic of the maleimide and triazole groups
of the C60-peptide molecule are visible in the spectra and
summarized in Table S1 (ESI†). The PM-RAIRS spectrum of
the C60-alkyl SAM highlights all the same bands characteristic
of the molecule as in the C60-peptide SAM, with some differences
in intensity. The amide I and II bands at 1644 and 1542 cmÀ1 are
not as intense as in the spectrum of the C60-peptide SAM, due to
the presence of only one amide group in the C60-alkyl molecule. A
careful analysis of the amide I and II band absorbance ratio in the
SAMs involving the peptide, using a distribution function model,
shows that the tilt angles of the helix axis from the surface normal
moiety on the other, allowing further functional modification.15
A
copper(I) source based on the preformed complex [Cu(I)ligand-
(hexabenzyl)tren]Br, prepared from CuBr and the hexabenzyltren
ligand, was used for this reaction, which led to 40% and 60%
yields for the C60-alkyl and C60-peptide, respectively. This approach
allowed the preparation of a significant amount of these deriva-
tives, in particular the C60-peptide, at a higher yield than the
previously reported strategy.14 The solution conformation analysis
by IR absorption, NMR and circular dichroism measurements, of
the starting hexapeptide and C60-peptide, confirmed their helical
conformation, likely a 310-helical structure from NMR experiments.
In addition, the starting hexapeptide, C60-alkyl and C60-peptide
were designed with a lipoyl moiety at the end of the alkyl chain or
the peptide N-terminus, which allowed their grafting on a gold
surface. The XPS and PM-RAIRS data showed that the monolayers
of these derivatives could be successfully prepared, and their self-
assembly properties were studied and compared using surface
characterization and electrochemical measurements. The beneficial
effect of the peptide on SAM formation was also confirmed. The
helical C60-peptide allowed the formation of well-packed fullerene
SAMs compared to the C60-alkyl, as determined from the coverage
rate obtained by QCM and the difference in the intensity of the CV
3À/4À
responses of the Fe(CN)6
redox probe. In addition to demon-
strating the contribution of the peptide towards the formation of a
denser fullerene layer, the comparison of angles obtained from the
PM-RAIRS measurements indicated that the C60-peptide tended to
be more vertical than the parent hexapeptide. This finding confirms
the starting assumption that the organization within the mono-
layers is governed both by fullerene interactions and peptide
interactions.
Fig. 9 Schematic representation of SAMs of the hexapeptide 15 and
We currently envision using the electron mediator properties of
the helical peptides along with their organising properties in
C
60-peptide adopting a 310-helical conformation along with the tilt angles
of the helices from the surface normal.
19430 | New J. Chem., 2018, 42, 19423--19432 This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018

View Article Online
NJC
Paper
Y.-K. Han, T. Joo and J. T. Park, Chem. – Eur. J., 2010, 16,
5586–5599.
7 M. Hosseini, D. M. Rivera-Nazario and L. A. Echegoyen,
ACS Appl. Mater. Interfaces, 2014, 6, 3712–3720.
fullerene-terminated peptide SAMs for organic electronic applica-
tions. In particular, we are planning to graft a donor–fullerene
dyad on a helical peptide in order to develop photoelectrochemical
devices. In such systems, the use of a short peptide, as used in this
study, able to assume a helical conformation should be beneficial
since the rate of electron transfer exponentially decreases with
increasing distance of the photo- and electro-active units from the
metal surface. The length of the short peptide should also be
sufficient not only to diminish the quenching of the photo-excited
chromophores by energy transfer from the gold surface, but also to
reduce charge recombination processes. In addition, the strongly
directional characteristic of electron transfer through the peptide
chain could also be exploited with fullerene-based SAMs for the
design of molecular rectifiers.
8 (a) H. Imahori, T. Azuma, S. Ozawa, H. Yamada, K. Ushida,
A. Ajavakom, H. Norieda and Y. Sakata, Chem. Commun.,
1999, 557–558; (b) H. Imahori, T. Azuma, A. Ajavakom,
H. Norieda, H. Yamada and Y. Sakata, J. Phys. Chem. B,
1999, 103, 7233–7237; (c) T. Gu, J. K. Whitesell and M. A.
Fox, J. Org. Chem., 2004, 69, 4075–4080; (d) A. S. Viana,
S. Leupold, C. Eberle, T. Shokati, F.-P. Montforts and
L. M. Abrantes, Surf. Sci., 2007, 601, 5062–5068; (e) Y. Shirai,
J. M. Guerrero, T. Sasaki, T. He, H. Ding, G. Vives, B.-C. Yu,
L. Cheng, A. K. Flatt, P. G. Taylor, Y. Gao and J. M. Tour, J. Org.
Chem., 2009, 74, 7885–7897; ( f ) M. del Carmen Gimenez-
¨ ¨
Lopez, M. T. Raisanen, T. W. Chamberlain, U. Weber,
Experimental section
M. Lebedeva, G. A. Rance, G. A. D. Briggs, D. Pettifor,
V. Burlakov, M. Buck and A. N. Khlobystov, Langmuir, 2011,
27, 10977–10985; (g) P. Piotrowski, J. Pawłowska, J. Pawłowski,
L. J. Opuchlik, R. Bilewicz and A. Kaim, RSC Adv., 2014, 6,
64310–64318; (h) M. V. Solano, E. A. Della Pia, M. Jevric,
C. Schubert, X. Wang, C. van der Pol, A. Kadziola,
K. Nørgaard, D. M. Guldi, M. B. Nielsen and J. O. Jeppesen,
Chem. – Eur. J., 2014, 20, 9918–9929.
The synthesis of compounds 1–15, C60-alkyl and C60 peptide
and the experimental details for SAM formation and character-
ization are described in the ESI.† The NMR spectra of the
hexapeptide 15 and C60-peptide are depicted in Fig. S3–S21
(ESI†). The IR spectra of hexapeptide 15 and C60-peptide in
CDCl3 solution, in ATR and on gold surfaces are depicted in
Fig. S1, S2 and S23–S25 (ESI†). The CD spectrum of hexapeptide
15 is also provided in Fig. S22 (ESI†).
9 J. C. Love, L. A. Estroff, J. K. Kriebel, R. G. Nuzzo and
G. M. Whitesides, Chem. Rev., 2005, 105, 1103–1169.
10 Y. Shirai, L. Cheng, B. Chen and J. M. Tour, J. Am. Chem.
Soc., 2006, 128, 13479–13489.
Conflicts of interest
11 (a) E. Gatto, A. Quatela, M. Caruso, R. Tagliaferro, M. De
Zotti, F. Formaggio, C. Toniolo, A. Di Carlo and M. Venanzi,
ChemPhysChem, 2014, 15, 64–68; (b) S. Yasutomi, T. Morita,
Y. Imanishi and S. Kimura, Science, 2004, 304, 1944–1947.
12 X. Yang, A. Ebrahimi, J. Li and Q. Cui, Int. J. Nanomed., 2014,
9, 77–92.
13 S. Jennepalli, S. G. Pyne and P. A. Keller, RSC Adv., 2014, 4,
46383–46398.
14 H. Uji, K. Tanaka and S. Kimura, J. Phys. Chem. C, 2016, 120,
3684–3689.
There are no conflicts of interest to declare.
Acknowledgements
This work was supported by the CHARMMMAT ANR-11-LABX-
0039-grant for the post-doctoral fellowships to HN and VB-T-T.
`
´
JR thanks the ‘‘Ministere de l’Enseignement Superieur de la
Recherche et de l’Innovation’’ for the PhD fellowship.
15 (a) C. Constant, S. Albert, N. Zivic, K. Baczko, H. Fensterbank and
E. Allard, Tetrahedron, 2014, 70, 3023–3029; (b) H. Fensterbank,
References
´
K. Baczko, C. Constant, N. Idttalbe, F. Bourdreux, A. Vallee, A.-M.
1 J.-F. Nierengarten, Top. Curr. Chem., 2003, 228, 87–110.
2 D. Bonifazi, O. Enger and F. Diederich, Chem. Soc. Rev.,
2007, 36, 390–414.
´
Goncalves, R. Meallet-Renault, G. Clavier, K. Wright and
E. Allard, J. Org. Chem., 2016, 81, 8222–8233.
3 S. K. Hau, Y.-J. Cheng, H.-L. Yip, Y. Zhang, H. Ma and A. K.-Y. 16 (a) I. L. Karle and P. Balaram, Biochemistry, 1990, 29,
Jen, ACS Appl. Mater. Interfaces, 2010, 2, 1892–1902.
6747–6755; (b) B. Pispisa, L. Stella, M. Venanzi, A. Palleschi,
C. Viappiani, A. Polese, F. Formaggio and C. Toniolo, Macro-
molecules, 2000, 33, 906–915; (c) G. Bocchinfuso, P. Conflitti,
S. Raniolo, M. Caruso, C. Mazzuca, E. Gatto, E. Placidi,
F. Formaggio, C. Toniolo, M. Venanzi and A. Palleschi,
J. Pept. Sci., 2014, 20, 494–507.
4 (a) T. Schmaltz, A. Khassanov, H.-G. Steinru¨ck, A. Margel,
A. Hirsch and M. Halik, Nanoscale, 2014, 6, 13022–13027;
(b) G. V. Dubacheva, M. Devynck, G. Raffy, L. Hirsch, A. Del
Guerzo and D. M. Bassani, Small, 2014, 10, 454–461.
5 L. Qiu, Y. Zhang, T. L. Krijger, X. Qiu, P. van’t Hof, J. C.
Hummelen and R. C. Chiechi, Chem. Sci., 2017, 8, 17 I. Nierengarten and J.-F. Nierengarten, Chem. Rec., 2015, 15,
2365–2372. 31–51.
6 (a) H. Imahori and S. Fukuzumi, Adv. Funct. Mater., 2004, 18 R. Srinivasan, L. P. Tan, H. Wu, P.-Y. Yang, K. A. Kalesh and
14, 525–536; (b) K.-S. Kim, M.-S. Kang, H. Ma and S. Q. Yao, Org. Biomol. Chem., 2009, 7, 1821–1828.
A. K. Y. Jen, Chem. Mater., 2004, 16, 5058–5062; (c) C. Y. Lee, 19 M. M. Kose, S. Onbulak, I. I. Yilmaz and A. Sanyal, Macro-
J. K. Jang, C. H. Kim, J. Jung, B. K. Park, J. Park, W. Choi, molecules, 2011, 44, 2707–2714.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018 New J. Chem., 2018, 42, 19423--19432 | 19431

View Article Online
Paper
NJC
20 L. Liang, J. Ruiz and D. Astruc, Adv. Synth. Catal., 2011, 353,
F. Formaggio, M. Crisma and M. Venanzi, J. Pept. Sci., 2011,
17, 124–131.
3434–3450.
21 T. R. Chan, R. Hilgraf, K. B. Sharpless and V. V. Fokin, Org. 31 W. B. Caldwell, K. Chen, C. A. Mirkin and S. J. Babinec,
Lett., 2004, 6, 2853–2855. Langmuir, 1993, 9, 1945–1947.
22 We have noticed that the fullerene derivatives containing 32 A. Vallee, V. Humblot, R. Al Housseiny, S. Boujday and
lipoic acids led to the formation of by-products detected by C.-M. Pradier, Colloids Surf., B, 2013, 109, 136–142.
TLC after allowing them to stand for several weeks in the 33 T. Heinrich, C. H.-H. Traulsen, E. Darlatt, S. Richter,
´
solid state or after several hours in oxygenated solution.
These by-products, which have been identified by mass
spectrometry, could result from the photosensitized oxida-
J. Poppenberg, N. L. Traulsen, I. Linder, A. Lippitz,
P. M. Dietrich, B. Dib, W. E. S. Unger and C. A. Schalley,
RSC Adv., 2014, 4, 17694–17702.
´
tion of the lipoic acid unit leading to the formation of 34 E. Beford, V. Humblot, C. Methivier, C.-M. Pradier, F. Gu,
thiosulfinates and thiolsulfonates.23 The fullerene unit acts
as a photosensitizer and triggers the formation of singlet
oxygen that promotes the oxidation of the lipoic moiety.
23 E. F. Stary, S. L. Jindal and R. W. Murray, J. Org. Chem., 1975,
40, 58–62.
F. Tielens and S. Boujday, Chem. – Eur. J., 2015, 21,
14555–14561.
35 D. G. Castner, K. Hinds and D. W. Grainger, Langmuir, 1996,
12, 5083–5086.
36 (a) M. Venanzi, G. Pace, A. Palleschi, L. Stella, P. Castrucci,
M. Scarselli, M. De Crescenzi, F. Formaggio, C. Toniolo and
G. Marletta, Surf. Sci., 2006, 600, 409–416; (b) E. Longo,
K. Wright, M. Caruso, E. Gatto, A. Palleschi, M. Scarselli,
M. De Crescenzi, M. Crisma, F. Formaggio, C. Toniolo and
M. Venanzi, Nanoscale, 2015, 7, 15495–15506.
37 R. G. Greenler, D. R. Snider, D. Witt and R. S. Sorbello,
Surf. Sci., 1982, 118, 415–428.
24 A. Polese, S. Mondini, A. Bianco, C. Toniolo, G. Scorrano,
D. M. Guldi and M. Maggini, J. Am. Chem. Soc., 1999, 121,
3446–3452.
25 M. De Zotti, K. Wright, E. D’Aboville, A. Toffoletti, C. Toniolo,
G. Longhi, G. Mazzeo, S. Abbate and F. Formaggio, J. Org.
Chem., 2017, 82, 10033–10042.
26 (a) K. Yanagisawa, T. Morita and S. Kimura, J. Am. Chem.
Soc., 2004, 126, 12780–12781; (b) L. Garbuio, S. Antonello, 38 P. Srivastava, M. Ghasemi, N. Ray, A. Sarkar, J. Kocabova,
I. Guryanov, Y. Li, M. Ruzzi, N. J. Turro and F. Maran, J. Am.
Chem. Soc., 2012, 134, 10628–10637; (c) N. Ousaka,
F. Mamiya, Y. Iwata, K. Nishimura and E. Yashima, Angew.
Chem., Int. Ed., 2017, 56, 791–795.
S. Lachmanova, M. Hromadova, S. Boujday, S. Cauteruccio,
P. Thakare, E. Licandro, C. Fosse and M. Salmain, Appl. Surf. Sci.,
2016, 385, 47–55.
39 P. Srivastava, A. Sarkar, S. Sawoo, A. Chakraborty, P. Dutta
and M. Salmain, J. Organomet. Chem., 2011, 696, 1102–1107.
40 E. Karjalainen and A. Barth, J. Phys. Chem. B, 2012, 116,
4448–4456.
27 S. J. Pike, V. Diemer, J. Raftery, S. J. Webb and J. Clayden,
Chem. – Eur. J., 2014, 20, 15981–15990.
28 (a) M. Goodman, A. S. Verdini, C. Toniolo, W. D. Phillips and
¨
¨ ¨
¨
F. A. Bovey, Proc. Natl. Acad. Sci., 1969, 64, 444–450; 41 P. Damlin, M. Hatonen, A. E. Dominguez, T. Aaritalo,
¨ ¨
H. Kivela and C. Kvarnstrom, RSC Adv., 2014, 4, 8391–8401.
(b) G. Holzwarth and P. Doty, J. Am. Chem. Soc., 1965, 87, 218–228.
29 E. Longo, A. Moretto, F. Formaggio and C. Toniolo, Chirality, 42 Y. Arikuma, H. Nakayama, T. Morita and S. Kimura, Angew.
2011, 23, 756–760. Chemie Int. Ed., 2010, 49, 1800–1804.
30 (a) M. Kai, K. Tadeka and S. Kimura, J. Pept. Sci., 2008, 14, 43 P. Gobbo, S. Antonello, I. Guryanov and F. Polo,
192–202; (b) E. Gatto, M. Caruso, A. Porchetta, C. Toniolo, ChemElectroChem, 2016, 3, 1–9.
19432 | New J. Chem., 2018, 42, 19423--19432 This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018