Triazine-Carbon Nanotubes: New Platforms for the Design of Flavin Receptors

Article abstract of DOI:10.1002/chem.201600630

The synthesis of functionalised carbon nanotubes as receptors for riboflavin (RBF) is reported. Carbon nanotubes, both single-walled and multi-walled, have been functionalised with 1,3,5-triazines and p-tolyl chains by aryl radical addition under microwav

Full text of DOI:10.1002/chem.201600630

DOI: 10.1002/chem.201600630
Full Paper
&
Molecular Recognition
Triazine-Carbon Nanotubes: New Platforms for the Design of
Flavin Receptors
[
a, b]
[a, b]
[a]
[a]
María Isabel Lucío,
Federica Pichler,
JosØ Ramón Ramírez, Antonio de la Hoz,
[a]
[b]
[c]
[b]
Ana Sµnchez-Migallón, Caroline Hadad, Mildred Quintana, Angela Giulani,
Maria Victoria Bracamonte,
María Antonia Herrero, Maurizio Prato, and Ester Vµzquez*
[b, d]
[e]
[b]
Jose L. G. Fierro, Claudio Tavagnacco,
[a]
[b]
[a]
Abstract: The synthesis of functionalised carbon nanotubes
as receptors for riboflavin (RBF) is reported. Carbon nano-
tubes, both single-walled and multi-walled, have been func-
tionalised with 1,3,5-triazines and p-tolyl chains by aryl radi-
cal addition under microwave irradiation and the derivatives
have been fully characterised by using a range of tech-
niques. The interactions between riboflavin and the hybrids
were analysed by using fluorescence and UV/Vis spectro-
scopic techniques. The results show that the attached func-
tional groups minimise the p-p stacking interactions be-
tween riboflavin and the nanotube walls. Comparison of p-
tolyl groups with the triazine groups shows that the latter
have stronger interactions with riboflavin because of the
presence of hydrogen bonds. Moreover, the triazine deriva-
tives follow the Stern–Volmer relationship and show a high
association constant with riboflavin. In this way, artificial re-
ceptors in catalytic processes could be designed through
specific control of the interaction between functionalised
carbon nanotubes and riboflavin.
Introduction
cesses and its redox behaviour depends on the covalent and
noncovalent interactions with the apoenzyme. Therefore, mod-
ulation of the environment of flavins can be used to set their
Flavoproteins are redox enzymes related to metabolic process-
[
1]
[2]
es, electron transfer and regulation of neurotransmitters. The
interdependence between redox events and molecular recog-
nition is a prevalent theme when studying these systems. The
flavin cofactor of proteins is the active site in the catalytic pro-
selectivity towards a specific analyte and to understand the
behaviour of enzymes. Historically, the most widely studied
flavin has been riboflavin (RBF) because of its chemical and
[
3]
biological versatility. Covalent attachment of flavins to en-
zymes is known to enhance selectively a reaction pathway and
[
4]
[
a] Dr. M. I. Lucío, F. Pichler, Dr. J. R. Ramírez, Prof. A. de la Hoz,
Dr. A. Sµnchez-Migallón, Dr. M. A. Herrero, Dr. E. Vµzquez
Departamento de Química Orgµnica, Inorgµnica y Bioquímica, Facultad de
Ciencias y Tecnologías Químicas
suppress unwanted side reactions. As a consequence, differ-
ent covalent modifications have been carried out on the cen-
tral nucleus of flavins to study their versatility in the design of
[
5]
[6]
sensors and alterations in the redox and fluorescence re-
sponses have been observed. Similarly, several receptors have
been developed to observe the molecular recognition of fla-
IRICA Universidad de Castilla-La Mancha Campus Universitario, 13071
Ciudad Real (Spain)
E-mail: Ester.Vazquez@uclm.es
[
7]
[
b] Dr. M. I. Lucío, F. Pichler, Dr. C. Hadad, A. Giulani, Dr. M. V. Bracamonte,
Prof. C. Tavagnacco, Prof. M. Prato
vins by noncovalent approaches. For example, the 2,4-dia-
mine-1,3,5-triazines are useful because they have an appropri-
ate D (H-donor)-A (acceptor) matrix to form hydrogen bonds
Dipartimento di Scienze Chimiche e Farmaceutiche Center of Excellence for
Nanostructured Materials (CENMAT) & Italian Interuniversity Consortium on
Materials Science and Technology (INSTM - Unit of Trieste) Università degli
Studi di Trieste Piazzale Europa 1, 34127 Trieste (Italy)
[
8]
with flavins. The 1,3,5-triazine nucleus is attractive in supra-
molecular chemistry because of the possibility of forming mul-
[
9]
[10]
[
c] Dr. M. Quintana
tiple noncovalent interactions by coordinate bonds, elec-
Instituto de Física Universidad Autónoma de San Luis Potosí Manuel Nava
[11]
[12]
trostatic interactions,
hydrogen bonds,
and p-p stack-
6
, Zona Universitaria 78290, San Luis Potosí, SLP (Mexico)
[
13,14]
ing.
tems can be easily modified.
In addition, from a synthetic point of view, these sys-
[
d] Dr. M. V. Bracamonte
[
15–18]
Instituto de Física Enrique Gaviola (CONICET) and FaMAF
Universidad Nacional de Córdoba Medina Allende s/n, X5000HUA Córdoba
It is known that three-dimensional (3D) patterns, such as sili-
cate matrices, effectively replicate both the isolation and preor-
(
Argentina)
[
19]
[
e] Prof. J. L. G. Fierro
ganisation found in the active sites of flavoenzymes. Accord-
ingly, the fabrication of 3D structures could make a difference
in the recognition processes of flavins and, apart from the bio-
logical mimics, the flavin-receptor complex could also be used
Instituto de Catµlisis y Petroleoquímica CSIC Cantoblanco
2
8049 Madrid (Spain)
Supporting information and the ORCID identification number for one of the
authors of this article can be found under http://dx.doi.org/10.1002/
chem.201600630.
[
20]
in the design of sensors and new catalysts.
Chem. Eur. J. 2016, 22, 8879 – 8888
8879
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
28–30]
Carbon nanotubes (CNTs) have been applied in different
lent modifications.
Two kinds of nanotube derivatives
[21]
[22]
fields such as electronics, biomedicine and sensor build-
were prepared (Scheme 1) with the aim of comparing the be-
haviour of covalent triazine-functionalised carbon nanotubes
with nanotubes functionalised to the same extent but without
the possibility of forming hydrogen bonds.
[23]
ing. The characteristic that makes these materials special is
the combination of dimension, structure, and topology, which
[24]
translates into a whole range of superior properties. More-
over, these materials have a large surface area that can be
Triazine 1 or p-toluidine (2) were reacted with pristine
HiPco SWCNTs under the conditions described in the experi-
[25]
ꢁ
easily modified in different ways. Covalent functionalisation
is a helpful tool to immobilise different useful molecules to
mental section to give SWCNTs-1 and SWCNTs-2. Thermogravi-
metric analysis shows the weight loss from samples as the
temperature increases. The TGA results for SWCNTs-1 and
SWCNTs-2 are shown in Figure 1a. The weight loss at 6008C
can be used to quantify the number of functional groups at-
tached to the carbon nanotube. SWCNTs-1 showed a weight
loss of 14.5% and this corresponds to 392 mmol of functional
groups per gram of carbon nanotube. The degree of function-
alisation is 1 functional group per 181 carbon atoms. SWCNTs-
2 showed a weight loss of 15.7%, which corresponds to
934 mmol of functional groups per gram of carbon nanotube.
This corresponds to 1 functional group per 75 carbon atoms.
The higher degree of functionalisation can be attributed to the
small size of the p-tolyl chain when compared with the sterical-
[26]
build supramolecular structures.
In addition, the electro-
chemical reactivity of some molecules—as well as some elec-
tron-transfer processes—are known to be improved by carbon
nanotubes, thus making them attractive for the design of bio-
[
27]
sensors.
We report here the synthesis and noncovalent interactions
of triazine-carbon nanotube derivatives with RBF. Both single-
walled carbon nanotubes (SWCNTs) and multi-walled carbon
nanotubes (MWCNTs) were functionalised by a radical arylation
with microwave activation. Different functional groups were
immobilised, namely 1,3,5-triazine chains, which are capable of
forming hydrogen bonds, and p-tolyl chains, which cannot par-
ticipate in this type of interaction. The characterisation of all of
the derivatives was carried out by using a range of techniques
such as thermogravimetric analysis (TGA), Raman spectroscopy,
UV/Vis/NIR spectroscopy, transmission electron microscopy
[
31,32]
ly more hindered triazine unit.
The electron-withdrawing
properties of the triazine ring and the lower solubility of tria-
[
33,34]
zine 1 may also reduce its reactivity.
(
TEM) and X-ray photoelectron spectroscopy (XPS). As a further
Raman spectroscopy provides information about the cova-
lent functionalisation of carbon nanotubes. The increase in the
step in understanding the molecular recognition processes,
the triazine-functionalised MWCNTs were applied as synthetic
hydrogen-bond receptors and their interactions with RBF were
analysed.
À1
3
D-band at 1314 cm can be related to the introduction of sp
2
[35]
defects into the sp network of the nanotube. The Raman
spectra of pristine and functionalised SWCNT derivatives are
shown in Figure 1b. SWCNTs-1 exhibits a more intense D-band
than p-SWCNTs and a more intense band is observed for
SWCNTs-2 as the degree of functionalisation increases. These
results are consistent with those of the TGA analysis.
Results and Discussion
Synthesis and Characterisation of Carbon Nanotube
Derivatives
Both SWCNT derivatives were analysed by UV/Vis spectros-
copy, which is an excellent method to monitor the electronic
perturbation of nanotubes. The van Hove singularities of pris-
There are a very few examples in which triazine moieties are
attached to nanotubes, and most of these concern noncova-
2
tine nanotubes disappeared due to rehybridisation of the sp
Scheme 1. Functionalisation of p-SWCNTs and p-MWCNTs.
Chem. Eur. J. 2016, 22, 8879 – 8888
www.chemeurj.org
8880
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
to 386 mmol of functional groups per gram of carbon nano-
tube, which is a value that is consistent with that obtained by
TGA analysis (392 mmol of functional groups per gram of
carbon nanotube).
Once the reaction had been successfully carried out with
SWCNTs and the products were fully characterised, p-MWCNTs
were functionalised with the same organic chains to yield de-
rivatives MWCNTs-1 and MWCNTs-2. The TGA traces for
MWCNT derivatives versus p-MWCNTs are shown in Figure S2.
MWCNTs-1 exhibited a weight loss of 13% and this corre-
sponds to 667 mmol of functional group per gram of carbon
nanotube. The degree of functionalisation is 1 functional
group per 110 carbon atoms. MWCNTs-2 showed a weight loss
of 11 % at 6008C, which corresponds to 1178 mmol of function-
al group per gram of carbon nanotube. This equates to 1 func-
tional group per 61 carbon atoms. The reaction with p-tolui-
dine 2 introduces more functional groups than the reaction
with triazine 1, as observed with SWCNTs. Functionalised
MWCNTs were also analysed by XPS. As expected, MWCNTs-2
did not contain nitrogen and MWCNTs-1 contained 5.0% nitro-
gen, which corresponds to 714 mmol of functional groups per
gram of carbon nanotube. This value is consistent with that
obtained by TGA analysis (667 mmol of functional groups per
gram of carbon nanotube) (Table S1).
The higher solubility of the final MWCNT derivatives in water
in comparison with the SWCNT derivatives, led us to use the
former nanotubes to analyse the molecular recognition of RBF.
It is known that the degree of functionalisation is an impor-
tant factor when the behaviour of CNTs is studied. With the
aim of comparing the influence of the types of functional
groups, three successive addition reactions were carried out
with triazine 1 on the same MWCNT sample. Similar degrees of
functionalisation were obtained in the hybrid with triazine and
the hybrid with p-tolyl groups. The TGA data for the latter de-
rivative, which is denoted as MWCNTs-3, are shown in Fig-
ure S2 along with those of the other MWCNT derivatives. The
degree of functionalisation of each MWCNT derivative is sum-
marised in Table 1.
Figure 1. a) Thermogravimetric analysis of p-SWCNTs, SWCNTs-1 and
SWCNTs-2. b) Normalised Raman spectra of p-SWCNTs, SWCNTs-1 and
SWCNTs-2 by using a laser source at 633 nm. (c) UV/Vis/NIR absorption spec-
tra of p-SWCNTs, SWCNTs-1 and SWCNTs-2 in solutions of sodium dodecyl
sulfate (2%) normalised to the local minimum at 930 nm.
It is known that functionalised carbon nanotubes that can
form hydrogen bonds in AD-DA (acceptor donor-donor accept-
or) systems produce good dispersions in aprotic polar systems
such as N,N-dimethylformamide (DMF), whereas they self-as-
semble to form superstructures in solvents in which they do
3
[31]
carbon atoms to sp in the functionalised carbon nanotubes.
A decrease in the intensity of the van Hove bands from p-
SWCNTs to SWCNTs-1 and SWCNTs-2 can be observed in Fig-
ure 1c because of the increase in the extent of functionalisa-
tion. This effect is more evident in SWCNTs-2 because of the
higher degree of functionalisation in comparison to SWCNTs-1.
The different derivatives were also analysed by X-ray photo-
electron spectroscopy (XPS). This semiquantitative technique
provides information about the elemental composition of the
[
37]
not form hydrogen bonds, such as dichloromethane. Conse-
quently, the different dispersibility of the nanotubes in differ-
ent solvents provides an insight into the kind of interactions
between them and the type of functional groups. Hence, the
behaviour of MWCNTs-1, MWCNTs-2 and MWCNTs-3 in different
solvents was analysed by transmission electron microscopy
(Figure 2). Both triazine-MWCNT derivatives formed good dis-
persions in water, MWCNTs-1 (Figure 2a) and MWCNTs-3 (Fig-
ure 2b), because of hydrogen-bonding interactions between
the organic chains and the solvent. However, the p-tolyl-func-
tionalised nanotubes MWCNTs-2 formed aggregates in water
[36]
sample as well as the types of bonds present. As expected,
the SWCNTs-2 derivative did not contain nitrogen and the
hybrid SWCNTs-1 contained 2.7% nitrogen, thus confirming
the functionalisation (Table S1). The N1s peak was deconvolut-
ed into two contributions at 398.3 and 400.1 eV (Figure S1).
These were assigned to the photoelectrons emitted by the N= (Figure 2c) as interactions with the solvent did not occur,
C and NÀC bonds present in the triazine rings, respectively.
whereas in nonpolar solvents such as 1,1,2,2-tetrachloroethane
the opposite behaviour was observed. The triazine-functional-
Furthermore, the percentage of nitrogen (2.7%) corresponds
Chem. Eur. J. 2016, 22, 8879 – 8888
www.chemeurj.org
8881
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
al groups (Figure 2d,e), whereas the p-tolyl-function-
alised nanotubes MWCNTs-2 formed a very good dis-
persion (Figure 2 f).
Table 1. Summary of the functionalised MWCNTs.
[b]
Product Structure
Weight loss FGC [mmol/
[
a]
[c]
[%]
g]
Molecular Recognition
The study of the molecular recognition of flavins by
a receptor is important in the development of sen-
sors as well as in the development of chemical
models to understand the role of the noncovalent in-
teractions in enzymatic processes. To obtain a qualita-
tive idea of the interaction between triazine-function-
alised carbon nanotubes MWCNTs-1 and RBF, an
excess of RBF (5 mg) was added to a 1,1,2,2-tetra-
chloroethane solution of MWCNTs-1 (20 mL,
MWCNTs-
1
1
3
110
667
À2
À1
2
.5·10 mgmL ) and, after washing the mixture, the
MWCNTs-
2
11
61
1178
recognition processes were analysed by TEM. High
dispersibility of nanotubes after the addition of RBF
can be observed in Figure 3. As expected, the triple
ADA-DAD hydrogen-bonding interactions between
the functional groups of the nanotubes and RBF
break the bundles.
In general, the methodologies used to observe the
molecular recognition are based on physical and
chemical changes in the receptor system when it in-
teracts with the acceptor system. Numerous studies
have been carried out to analyse the recognition of
flavin by different receptors and, for example, the
MWCNTs-
3
19.5
63
1056
[
38,39]
quenching of its fluorescence,
the shift of bands
[
40]
in the UV spectra, the shift of peaks in the IR spec-
[
a] TGA weight loss (%) at 6008C. [b] Functional group coverage: Number of carbon
[
41]
[8]
atoms per functional group=[(100À%weightloss)·Mmfunctionalgroup]/12%weightloss. [c] mmol of
tra and NMR spectra or the modification of the
[
42]
functional group per gram of carbon nanotube.
redox potential by cyclic voltammetry have been
observed. In particular, the linear optical properties of
[
43–45]
RBF have been widely studied.
ised nanotubes, MWCNTs-1 and MWCNTs-3, self-assembled be-
The UV/Vis spectra from a titration of RBF with increasing
cause of hydrogen-bonding interactions between the function-
concentrations of MWCNTs-3 in water are shown in Figure 4.
Figure 2. TEM Images of a) MWCNTs-1 in H
tetrachloroethane, and f) MWCNTs-2 in 1,1,2,2-tetrachloroethane. The concentration of nanotubes in all samples was 2.5·10 mgmL
2 2 2
O, b) MWCNTs-3 in H O, c) MWCNTs-2 in H O, d) MWCNTs-1 in 1,1,2,2-tetrachloroethane, e) MWCNTs-3 in 1,1,2,2-
À2
À1
.
Chem. Eur. J. 2016, 22, 8879 – 8888
www.chemeurj.org
8882
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
À2
À1
Figure 3. TEM Images of an MWCNTs-1 solution in 1,1,2,2-tetrachloroethane (2.5·10 mgmL ) a) before and b) after the addition of an excess of RBF. Expect-
ed interaction between MWCNTs-1 c) before and d) after the addition of RBF in 1,1,2,2-tetrachloroethane.
trum of RBF due to the superposition of the two absorbance
spectra (Figure 4). On the other hand, blueshifts in the bands
at 374 and 444 nm are observed on increasing the nanotube
concentration (Figure 4). The shifts are 16 nm for the first band
and 11 nm for the latter. The band at 374 nm is attributed to
the p-p* and n-p* transition of RBF and the band at 444 nm to
[
46]
the p-p* transition.
The peak at 444 nm hardly changes
when the hydrogen-bonding interactions between RBF and its
environment change, whereas the maximum at 374 nm does
[
47]
exhibit shifts in the wavelength. The small movements in
[
39,48]
our system are consistent with reported data
and confirm
the existence of supramolecular interactions in our complex. In
addition, the more marked shift of the peak at 374 nm corrob-
orates the modification of the hydrogen-bonding environment
of the RBF.
Figure 4. UV/Vis spectra showing a monotonic increase in intensity of RBF
À3
À1
with increasing concentration of MWCNTs-3. [RBF]=5.0·10 mgmL
,
À3
À3
À3
À3
À3
À3
[
MWCNTs-3]=0, 1.7·10 , 2.9·10 , 3.7·10 , 4.4·10 , 5.0·10 , 5.5·10 ,
À3
À3
À3
À2
À1
5
.8·10 , 6.1·10 , 6.4·10 and 1·10 mgmL .
Fluorescence spectroscopy is a more sensitive technique
than UV/Vis spectroscopy. Hence, the interaction between the
triazine-functionalised carbon nanotubes MWCNTs-3 and RBF
was also analysed by using this technique. The effect of in-
creasing the concentrations of MWCNTs-3 on the emission
properties of RBF is shown in Figure 5. As one would expect,
the addition of MWCNTs-3 to the solution of RBF resulted in
Although water is not an ideal solvent for this study, it was
chosen to obtain stable dispersions of RBF during the whole
complexation process. The increase in the concentration of
nanotubes results in an enhancement in the absorbance spec-
Chem. Eur. J. 2016, 22, 8879 – 8888
www.chemeurj.org
8883
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
54–56]
[57–62]
form p-p interactions
even with flavins
and the high
values of the association constants in our system could be due
to the presence of p-p stacking together with the formation of
hydrogen bonds.
A series of nanotube hybrids was studied to analyse the in-
fluence that the different kinds of interactions had on the
quenching of the fluorescence of RBF. These systems were tria-
zine-functionalised carbon nanotubes with different degrees of
functionalisation, i.e., MWCNTs-1 and MWCNTs-3, p-tolyl-func-
tionalised carbon nanotubes MWCNTs-2, which cannot form
hydrogen bonds, and p-MWCNTs (Figure 6). Pristine nanotubes
produced the highest quenching of RBF (black solid line) and
the p-tolyl-functionalised nanotubes the lowest (line with cir-
cles). Presumably, the presence of functional groups in the p-
tolyl-functionalised carbon nanotubes precludes the approach
of the planar rings of RBF to the walls. This situation is consis-
tent with other studies in which a lower quenching effect of
functionalised nanotubes was observed when compared with
Figure 5. Fluorescence spectra showing a monotonic decrease in intensity of
RBF on increasing the concentration of MWCNTs-3 (lexc =444 nm).
À3
À1
À4
À4
À4
À3
[
RBF]=1.0·10 mgmL , [MWCNTs-3]=5.7·10 , 7.5·10 , 8.9·10 , 1·10 ,
À3
À3
À1
1
.3·10 and 2·10 mgmL .
[
63]
pristine nanotubes. On the other hand, the quenching was
higher for the triazine-functionalised carbon nanotubes (solid
grey line and line with triangles) in comparison to the p-tolyl
system (line with circles). Notably, comparison of MWCNTs-3
and MWCNTs-2 shows a far higher quenching for the triazine
derivative in comparison to the p-tolyl derivative. The fact that
both derivatives have similar degrees of functionalisation indi-
cates that the highest quenching can be related to the ability
to form hydrogen bonds in the triazine hybrids. The titration
of riboflavin with increasing concentrations of MWCNTs-2 gave
results that are consistent with this assumption (Figure S7 and
S8). The association constant between these two components
partial quenching of its fluorescence emission because of the
interaction between the two species.
The maximum of the emission spectrum at 524 nm was
used to analyse the quenching data by using the Stern–Volmer
[45]
equation (1) (Figure S3).
0
F =F ¼ 1þK ½QŠ
ð1Þ
I
I
SV
0
I
where F and F are the fluorescence intensities in the absence
I
and presence of the quencher, respectively, KSV is the Stern–
Volmer quenching constant, and [Q] is the concentration of
À1
(riboflavin–MWCNTs-2) is 82.6Æ5.2 Lg and this equates to
3
À1
the quencher, MWCNTs-3, in the sample. The Stern–Volmer
(70.1Æ4.4)·10 Lmol
. This value is high because of the
p-tolyl
À1
quenching constant (K ) was 128.2Æ6.8 Lg and this equates
possibility of forming p-p stacking interactions with the func-
SV
3
À1
to (121.4Æ6.4)·10 Lmol
. These studies were also carried
tionalised nanotubes but it is lower than that found for the ri-
triazine
À1
out with MWCNTs-1, which has fewer triazine chains attached
to the walls of the nanotubes. The analysis showed the same
trend as observed for the derivative MWCNTs-3. However,
MWCNTs-1 exhibited a weaker interaction than MWCNTs-3 be-
cause the blueshift in the UV/Vis spectrum was smaller (5 nm
at 374 nm and 2 nm at 444 nm) and the quenching of RBF was
not very high (Figure S4 and S5). This weaker interaction was
confirmed by the association constant calculated from the
boflavin–MWCNTs-3 systems; that is, (128.2Æ6.8) Lg
or
3
À1
(121.4Æ6.4) 10 Lmol
. These differences reveal that the
triazine
hydrogen bonds make a clear contribution to the quenching
effect. Similar behaviour has previously been observed in other
examples in which the quenching of the fluorescence of flavins
is dependent on the ability of their receptors to form hydro-
[
64,65]
gen bonds and p-p interactions.
In this way, we demon-
strate here that the recognition of the RBF in our system is
modulated by the introduction of triazines with the ability to
form hydrogen bonds.
À1
Stern–Volmer plot (Figure S6). The value was 94.7Æ4.7 Lg
for the derivative MWCNTs-1.
In molecular systems the association constants between fla-
Once the formation of complexes between functionalised
MWCNTs and RBF had been demonstrated, we carried out
some preliminary studies aimed at characterising their electro-
chemical behaviour in solution. Electrochemical studies on fla-
vins in aqueous solutions are complex and studies concerning
interactions with other components are scarce. However, we
chose this solvent because of the aforementioned stability of
the RBF.
vins and 2,4-diamino-1,3,5-triazines bound by hydrogen bonds
À1 [8,49,50]
have values between 1 and 744 Lmol ,
depending on
the substituent(s) on the triazine rings. In addition, the pres-
ence of supplementary supramolecular interactions in the
system can affect the stabilisation of the hydrogen-bonding in-
teractions and these influences are used to modulate such
[
51]
bonding. As a consequence, the aromatic stacking interac-
[52]
À3
À1
À3
tions have been widely studied. Higher association constants
between flavins and 2,4-diamino-1,3,5-triazines with different
aromatic substituents are observed as a consequence of the in-
crease in the p-overlap between these compounds (values
The DPV profiles of 1.0·10 mgmL RBF and (1.0·10 –
À3
À1
2.0·10 mgmL ) RBF-MWCNT hybrids at a bare GCE in 0.1m
À1
phosphate buffer solution pH 7.4 with a scan rate of 0.100 Vs
are shown in Figure 7. RBF complexes formed with p-MWCNTs
À1 [53]
from 394 to 17600 Lmol ). Nanotubes are easily able to
and MWCNTs-3 were tested.
Chem. Eur. J. 2016, 22, 8879 – 8888
www.chemeurj.org
8884
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
À3
À1
À3
À3
À1
Figure 6. a) Fluorescence spectra of RBF (1.0·10 mgmL ) and RBF-nanotube (1.0·10 –2.0·10 mgmL ) solutions (lexc =444 nm). Schematic representation
of possible structural interactions of RBF with the different hybrids: b) MWCNTs-1 or MWNCTs-3; (c) MWCNTs-2 and (d) p-MWCNTs.
Several differences in the intensity and the position of the
maximum peak potentials were observed for the compounds.
RBF in solution exhibits only one oxidation peak at À0.468Æ
0
.006 V and this is in good agreement with the behaviour re-
[
67]
ported previously. After the interaction of the RBF with p-
MWCNTs a shift in the peak potential was not observed
(
À0.468Æ0.002 V). Although the quenching of the fluores-
cence of flavins by aromatic systems has been verified in differ-
ent studies in which interactions have been demonstrated, p-p
stacking has not been considered alone as a modulator of the
redox behaviour of flavins in solution. Nonetheless, the influ-
ence of these interactions on the redox properties of flavins
has been widely studied when they participate in complexes
involving other noncovalent interactions. However, the effect
was not linearly correlated with the p-overlap because it is not
Figure 7. Differential pulse voltammetry of a glassy carbon electrode in RBF
grey solid line), RBF–p-MWCNTs (black solid line), and RBF–MWCNTs-3 (dash
line).
(
[
52]
based on this kind of interaction alone. When RBF interacts
Chem. Eur. J. 2016, 22, 8879 – 8888
www.chemeurj.org
8885 ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
with the derivative MWCNTs-3, a shift of 19 mV in the peak po-
sition to more negative potential was observed (À0.487Æ
water, at RT, by the addition of known microlitre aliquots of a solu-
À3
À1
tion containing MWCNTs and RBF (2·10 mgmL
and
À3
À3
À1
À1
1
·10 mgmL , respectively) to 1 mL of a solution containing RBF
0
.009 V), which indicates that the RBF is more difficult to
(
1·10 mgmL ). UV/Vis titration experiments were carried out in
reduce. The difference of 19 mV in the redox potentials of RBF
is in agreement with reported data. Movements of 10–60 mV
to more negative potentials have been associated with the sta-
bilisation of the flavin through receptors that are able to take
water, at RT, by addition of known microlitre aliquots of a solution
À2
À1
À3
À1
containing MWCNTs and RBF (1·10 mgmL and 5·10 mgmL ,
À3
À1
respectively) to 1 mL of a solution containing RBF (5·10 mgmL ).
The thermogravimetric analyses were performed with a TGA Q50
(TA Instruments) at 108C min under a N2 atmosphere. Raman
[53]
À1
part in both bonding and p-p stacking.
In light of these preliminary electrochemistry results, the de-
rivative MWCNTs-3 is revealed as a potential hybrid for the
modulation of the redox behaviour of RBF. This proposal is
consistent with the UV/Vis and fluorescence results and corrob-
orates the influence of the attached organic chains in the sys-
tems. Nevertheless, groups that interfere to a greater extent
spectra were recorded with an Invia Renishaw microspectrometer
equipped with a He-Ne laser at 633 nm excitation wavelength and
a laser power of 0.17 mW. For the TEM analyses, several drops of
À2
À1
nanotube solutions (2.5·10 mgmL ) in different solvents were
placed on a copper grid (3.00 mm, 200 mesh, coated with carbon
film). After being air-dried, the sample was investigated with a TEM
Philips EM 208, accelerating voltage of 100 kV. The NMR spectra
were recorded with a Varian Unity 500 MHz spectrometer with TMS
as an internal standard. Photoelectron spectra (XPS) were obtained
with a VG Escalab 200R spectrometer equipped with a hemispheri-
cal electron analyser with a pass energy of 50 eV and a MgKa
(
i.e., strongly electron-withdrawing or electron-donating
groups) should be attached to the triazine rings to produce
more pronounced variations in the RBF behaviour. In addition,
the interaction through hydrogen bonds could become more
selective by introducing surfactants that subdue some p-p in-
(hn=1254.6 eV) X-ray source, powered at 120 W. Binding energies
[
66]
were calibrated relative to the C1s peak at 284.8 eV. High-resolu-
tion spectra envelopes were obtained by curve fitting synthetic
peak components with the software “XPS peak”. Symmetric Gaussi-
an–Lorentzian curves were used to approximate the line shapes of
the fitting components. Atomic ratios were computed from experi-
mental intensity ratios and normalised by atomic sensitivity factors.
Differential pulse voltammetry (DPV) was performed with an Auto-
lab 302b electrochemical workstation (Autolab, The Netherlands).
A conventional three-electrode system was employed with a bare
Glassy Carbon Electrode (GCE) as working electrode, a platinum
wire as auxiliary electrode, and the reference electrode was a satu-
rated Ag/AgCl (3m KCl). All measurements were performed from
À0.7 to 0.4 V using 4 mV step potential and 40 mV as modulation
teractions.
Conclusion
We have successfully synthesised and characterised 1,3,5-tria-
zine-CNT and p-tolyl-CNT derivatives (both SWCNTs and
MWCNTs) by means of an aryl radical addition. The self-assem-
bly of these hybrids has been analysed by transmission elec-
tron microscopy. It was observed that the 1,3,5-triazine deriva-
tives form good dispersions in water and self-assemble in non-
polar solvents because of the DA-AD hydrogen-bonding recog-
nition, whereas the p-tolyl derivatives show better dispersibility
in organic solvents and aggregate in polar solvents. On the
other hand, the ability of the different nanotubes to recognise
RBF has been studied by fluorescence spectroscopy and the
scope of the different noncovalent interactions has been ana-
lysed. The functionalisation of nanotubes by using a covalent
approach decreases their ability to form p-p stacking interac-
tions, thus allowing hydrogen-bonding interactions to play an
important role in the recognition processes between the com-
ponents. Preliminary electrochemical results show differences
in the response of RBF when it interacts with triazine-modified
nanotubes.
potential, at RT and under a N atmosphere.
2
Materials: Solvents were purchased from SDS and Fluka. Chemicals
were purchased from Sigma–Aldrich and were used as received.
ꢁ
HipCo SWCNTs were purchased from Carbon Nanotechnologies,
Inc., Lot R0513 and used without purification. MWCNTs 7000 series
were purchased from Nanocyl (lot 31825, www.nanocyl.com) and
were used without purification.
Preparation of 6-Amino-2,4-diamine-1,3,5-triazine (1): The prepa-
ration of this compound was carried out by using a modified re-
[68]
duction
protocol.
6-Nitro-2,4-diamine-1,3,5-triazine
(1 g,
4
.3 mmol) and hydrazine monohydrate (0.84 mL, 17.2 mmol) were
dissolved in EtOH (300 mL) and stirred after the addition of cat.
Pd/C (10%) at 808C overnight. The reaction solution was filtered
ꢁ
The possibility of modifying flavin and triazine substituents
in our system pave the way for the design of new flavin-based
molecular devices as chemical models for flavoenzymes in
which recognition and function could be directly correlated.
through Celite , which was washed with ethanol, and the solvent
was evaporated under vacuum to give compound 1. Yield: 825 mg
1
(
95%); yellow powder; m.p. 200–2028C; H NMR (500 MHz, [D6]
DMSO, 258C): d=7.96 (d, J(H,H)=8.8 Hz, 2H; CH), 6.55 (d, J(H,H)=
.8 Hz, 2H; CH), 6.48 (s, 4H; NH ), 5.59 ppm (s, 2H, NH ) (Fig-
8
2
2
13
ure S9); CNMR (500 MHz, [D ] DMSO, 258C): d=170.25 (s, C2),
6
1
67.22 (C6), 151.71 (C4’), 129.28 (C2’), 123.93 (C1’), 112.67 ppm
Experimental Section
À1
(
(
C3’) (Figure S10); IR (neat): n˜ =3700, 1740, 1734, 1717, 810 cm
Figure S11).
Techniques: Microwave irradiation was carried out with a CEM Dis-
cover reactor with an infrared pyrometer, pressure control system,
stirring and air-cooling option. UV/Vis/NIR spectra were recorded
with a Varian Cary 5000 spectrophotometer. Fluorescence spectra
were recorded with excitation at 444 nm, an emission filter from
Preparation of SWCNTs-1 and MWCNTs-1: Pristine CNTs (25 mg)
were sonicated in deionised water together with triazine
1 (210 mg, 1.1 mmol) for 10 min in a microwave glass vessel. Iso-
amyl nitrite (3) (0.56 mL, 4.16 mmol) was added and a reflux con-
denser was fitted. The mixture was heated at 808C by irradiation
at 100 W for 30 min and, after adding a new aliquot of isoamyl ni-
4
50 to 700 nm and a slit width of 3 nm. For absorption and fluo-
rescence experiments, quartz cuvettes with a 10 mm path-length
were used. Fluorescence titration experiments were carried out in
[32]
trite (3), at 30 W for 60 min. The mixture was cooled to RT and
Chem. Eur. J. 2016, 22, 8879 – 8888
www.chemeurj.org
8886 ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
the crude product was filtered off on a Millipore membrane (PTFE,
.2 mm). The product was removed from the filter and washed with
[17] A. Ruiz-Carretero, J. R. Ramírez, A. Sµnchez-Migallón, A. de La Hoz, Tetra-
hedron 2014, 70, 1733–1739.
0
[
[
[
18] . Díaz-Ortiz, J. Elguero, A. de La Hoz, A. JimØnez, A. Moreno, S.
Moreno, A. Sµnchez-Migallón, QSAR Comb. Sci. 2005, 24, 649–659.
19] M. D. Greaves, T. H. Galow, V. M. Rotello, Chem. Commun. 1999, 169–
dimethylformamide, methanol and dichloromethane (sonicated
and filtered) until the filtrate was clear. The solid was finally
washed with diethyl ether to afford SWCNTs-1 (23 mg) and
MWCNTs-1 (24 mg).
170.
20] L. Vial, P. Dumy, New J. Chem. 2009, 33, 939–946.
Preparation of SWCNTs-2 and MWCNTs-2: Pristine CNTs (25 mg)
were sonicated in deionised water together with p-toluidine (2)
[21] M. M. Shulaker, G. Hills, N. Patil, H. Wei, H.-Y. Chen, H.-S. P. Wong, S.
Mitra, Nature 2013, 501, 526–530.
(
120 mg, 1.1 mmol) for 10 min in a microwave glass vessel. Isoamyl
[22] A. Battigelli, C. MØnard-Moyon, T. Da Ros, M. Prato, A. Bianco, Adv. Drug
Delivery Rev. 2013, 65, 1899–1920.
nitrite (3) (0.45 mL, 3.34 mmol) was added and a condenser was
fitted. The mixture was heated at 808C by irradiation at 100 W for
3
3
product was filtered off on a Millipore membrane (PTFE, 0.2 mm).
The product was removed from the filter and washed with dime-
thylformamide, methanol and dichloromethane (sonicated and fil-
tered) until the filtrate was clear. The solid was finally washed with
diethyl ether to afford SWCNTs-2 (22 mg) and MWCNTs-2 (24 mg).
[23] P. YµÇez-SedeÇo, J. M. Pingarrón, J. Riu, F. X. Rius, TrAC Trends Anal.
Chem. 2010, 29, 939–953.
[24] P. M. Ajayan, Chem. Rev. 1999, 99, 1787–1800.
0 min and, after adding a new aliquot of isoamyl nitrite (3), at
[32]
0 W for 60 min. The mixture was cooled to RT and the crude
[25] P. Singh, S. Campidelli, S. Giordani, D. Bonifazi, A. Bianco, M. Prato,
Chem. Soc. Rev. 2009, 38, 2214–2230.
[
[
[
26] A. Micoli, M. Quintana, M. Prato, Supramol. Chem. 2013, 25, 567–573.
27] J. Wang, Electroanalysis 2005, 17, 7–14.
28] G. Ke, Carbohydr. Polym. 2010, 79, 775–782.
[29] L. A. S. D. A. Prado, A. De La Vega, J. Sumfleth, K. Schulte, J. Polym. Sci.
Part B 2009, 47, 1860–1868.
[30] S. S. Mahapatra, S. K. Yadav, H. J. Yoo, J. W. Cho, J.-S. Park, Composites
Part B 2013, 45, 165–171.
Acknowledgements
[31] R. K. Saini, I. W. Chiang, H. Peng, R. E. Smalley, W. E. Billups, R. H. Hauge,
J. L. Margrave, J. Am. Chem. Soc. 2003, 125, 3617–3621.
M.V.B. undertook this work with the support of the “ICTP TRIL
Programme, Trieste, Italy”. Financial support from the Spanish
MINECO (projects CTQ2011-22410 and CTQ2014-53600-R), the
FEDER Project PEII-2014-002A, the University of Castilla-la
Mancha, the University of Trieste and the Italian Ministry of
Education MIUR (cofin Prot. 2010N3T9M4 and Firb RBAP11ET-
KA) is gratefully acknowledged. We also thank Dr. Emilio M.
PØrez for help with Raman spectroscopy.
[32] F. G. Brunetti, M. A. Herrero, J. D. M. MuÇoz, A. Díaz-Ortiz, J. Alfonsi, M.
Meneghetti, M. Prato, E. Vµzquez, J. Am. Chem. Soc. 2008, 130, 8094–
8100.
[
33] M. E. Lipi n´ ska, S. L. H. Rebelo, M. F. R. Pereira, J. a. N. F. Gomes, C. Freire,
J. L. Figueiredo, Carbon 2012, 50, 3280–3294.
[34] J. L. Bahr, J. Yang, D. V. Kosynkin, M. J. Bronikowski, R. E. Smalley, J. M.
Tour, J. Am. Chem. Soc. 2001, 123, 6536–6542.
[
[
35] R. Graupner, J. Raman Spectrosc. 2007, 38, 673–683.
36] S. Giordani, J.-F. Colomer, F. Cattaruzza, J. Alfonsi, M. Meneghetti, M.
Prato, D. Bonifazi, Carbon 2009, 47, 578–588.
[
[
[
37] M. Quintana, M. Prato, Chem. Commun. 2009, 6005–6007.
38] M. D. Greaves, V. M. Rotello, J. Am. Chem. Soc. 1997, 119, 10569–10572.
39] S. Dutta Choudhury, J. Mohanty, A. C. Bhasikuttan, H. Pal, J. Phys. Chem.
B 2010, 114, 10717–10727.
Keywords: donor–acceptor systems · molecular recognition ·
nanotubes · noncovalent interactions · receptors
[
[
[
40] B. J. Jordan, G. Cooke, J. F. Garety, M. A. Pollier, N. Kryvokhyzha, A. Bayir,
G. Rabani, V. M. Rotello, Chem. Commun. 2007, 1248–1250.
41] K. Ariga, A. Kamino, H. Koyano, T. Kunitake, J. Mater. Chem. 1997, 7,
[
[
1] R. Hille, S. Miller, B. Palfey, Handbook of Flavoproteins, Berlin/Boston,
012.
2] R. F. Carvalhal, R. K. Mendes, L. T. Kubota, Int. J. Electrochem. Sci. 2007, 2,
73–985.
2
1
155–1161.
42] E. Breinlinger, A. Niemz, V. M. Rotello, J. Am. Chem. Soc. 1995, 117,
379–5380.
9
5
[
[
[
3] V. Massey, Biochem. Soc. Trans. 2000, 28, 283–296.
[
[
43] G. Weber, Biochem. J. 1950, 47, 114–121.
44] M. Quick, A. Weigel, N. P. Ernsting, J. Phys. Chem. B 2013, 117, 5441–
4] M. Mewies, W. S. McIntire, N. S. Scrutton, Protein Sci. 1998, 7, 7–20.
5] J. B. Carroll, G. Cooke, J. F. Garety, B. J. Jordan, S. Mabruk, V. M. Rotello,
Chem. Commun. 2005, 3838–3840.
6] N. Zainalabdeen, B. Fitzpatrick, M. M. Kareem, V. Nandwana, G. Cooke,
V. M. Rotello, Int. J. Mol. Sci. 2013, 14, 7468–7479.
7] R. Singh, Geetanjali, C. R. Babu, Chem. Biodiversity 2005, 2, 429–446.
8] R. Deans, G. Cooke, V. M. Rotello, J. Org. Chem. 1997, 62, 836–839.
9] T. J. Mooibroek, P. Gamez, Inorg. Chim. Acta 2007, 360, 381–404.
5447.
[
45] V. V. Mokashi, L. S. Walekar, P. V. Anbhule, S. H. Lee, S. R. Patil, G. B. Kole-
kar, J. Nanopart. Res. 2014, 16, 2291.
46] P. F. Heelis, Chem. Soc. Rev. 1982, 11, 15–39.
[
[
[
[
[
[
47] K. Yagi, N. Ohishi, K. Nishimoto, J. Do Choi, P.-S. Song, Biochemistry
1980, 19, 1553–1557.
[
[
48] A. Saha, S. Manna, A. K. Nandi, Langmuir 2007, 23, 13126–13135.
49] E. Jeoung, H. Augier de Cremiers, R. Deans, G. Cooke, S. L. Heath, P. E.
Vanderstraeten, V. M. Rotello, Tetrahedron Lett. 2001, 42, 7357–7359.
50] R. Deans, A. O. Cuello, T. H. Galow, M. Ober, V. M. Rotello, J. Chem. Soc.
Perkin Trans. 2 2000, 1309–1313.
[
[
10] B. R. Manzano, F. A. Jalón, M. L. Soriano, M. C. Carrión, M. P. Carranza, K.
Mereiter, A. M. Rodríguez, A. de La Hoz, A. Sµnchez-Migallón, Inorg.
Chem. 2008, 47, 8957–8971.
11] P. de Hoog, P. Gamez, I. Mutikainen, U. Turpeinen, J. Reedijk, Angew.
Chem. Int. Ed. 2004, 43, 5815–5817; Angew. Chem. 2004, 116, 5939–
[
[
[
51] G. Cooke, V. M. Rotello, Chem. Soc. Rev. 2002, 31, 275–286.
52] V. Nandwana, I. Samuel, G. Cooke, V. M. Rotello, Acc. Chem. Res. 2013,
46, 1000–1009.
5
941.
[
[
[
[
12] B. R. Manzano, F. a. Jalón, M. L. Soriano, A. M. Rodríguez, A. de La Hoz,
A. Sµnchez-Migallón, Cryst. Growth Des. 2008, 8, 1585–1594.
13] V. León, M. Quintana, M. A. Herrero, J. L. G. Fierro, A. de La Hoz, M.
Prato, E. Vµzquez, Chem. Commun. 2011, 47, 10936–10938.
[53] E. C. Breinlinger, V. M. Rotello, J. Am. Chem. Soc. 1997, 119, 1165–1166.
[54] A. Mateo-Alonso, C. Ehli, K. H. Chen, D. M. Guldi, M. Prato, J. Phys. Chem.
A 2007, 111, 12669–12673.
[55] S.-N. Ding, D. Shan, S. Cosnier, A. Le Goff, Chem. Eur. J. 2012, 18, 11564–
11568.
14] V. León, A. M. Rodriguez, P. Prieto, M. Prato, E. Vµzquez, ACS Nano 2014,
8
, 563–571.
15] . Díaz-Ortiz, J. Elguero, C. Foces-Foces, A. de La Hoz, A. Moreno, M.
del C. Mateo, A. Sµnchez-Migallón, G. Valiente, New J. Chem. 2004, 28,
[56] P. D. Petrov, G. L. Georgiev, A. H. E. Müller, Polymer 2012, 53, 5502–
5506.
9
52–958.
[
16] A. Ruiz-Carretero, O. Noguez, T. Herrera, J. R. Ramírez, A. Sµnchez-Migal-
[57] C. S. Lin, R. Q. Zhang, T. A. Niehaus, T. Frauenheim, J. Phys. Chem. C
2007, 111, 4069–4073.
lón, A. de La Hoz, J. Org. Chem. 2014, 79, 4909–4919.
Chem. Eur. J. 2016, 22, 8879 – 8888
www.chemeurj.org
8887
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
[
58] S.-Y. Ju, J. Doll, I. Sharma, F. Papadimitrakopoulos, Nat. Nanotechnol.
008, 3, 356–362.
59] S.-Y. Ju, D. C. Abanulo, C. A. Badalucco, J. A. Gascón, F. Papadimitrako-
poulos, J. Am. Chem. Soc. 2012, 134, 13196–13199.
[65] J. Carroll, M. Gray, K. Bardon, H. Nakade, V. Rotello, Lett. Org. Chem.
2
2004, 1, 227–230.
[66] A. Di Crescenzo, V. Ettorre, A. Fontana, Beilstein J. Nanotechnol. 2014, 5,
1675–1690.
[
[
60] S. Ju, F. Papadimitrakopoulos, J. Am. Chem. Soc. 2008, 130, 655–664.
61] R. Sharifi, D. C. Abanulo, F. Papadimitrakopoulos, Langmuir 2013, 29,
[67] H. Zhang, J. Zhao, H. Liu, H. Wang, R. Liu, J. Liu, Int. J. Electrochem. Sci.
2010, 5, 295–301.
7
209–7215.
62] J. Sim, H. Oh, E. Koo, S.-Y. Ju, Phys. Chem. Chem. Phys. 2013, 15, 19169–
9179.
[68] H. Weissauer, W. Braun, A. Tratter, Dyestuffs Containing Diamino Triazine
Groups, 1965.
[
1
[
[
63] D. K. Singh, P. K. Iyer, P. K. Giri, Carbon 2012, 50, 4495–4505.
64] C. Subramani, G. Yesilbag, B. J. Jordan, X. Li, A. Khorasani, G. Cooke, A.
Sanyal, V. M. Rotello, Chem. Commun. 2010, 46, 2067–2069.
Received: February 10, 2016
Published online on May 11, 2016
Chem. Eur. J. 2016, 22, 8879 – 8888
www.chemeurj.org
8888
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim