Indirect grafting of acetonitrile-derived films on metallic substrates

Article abstract of DOI:10.1021/cm100295n

Strongly bonded organic films with amino groups are obtained on gold, copper, and silicon surfaces by reduction of 2,6-dimethyl benzenediazonium in acetonitrile (ACN). The sterically hindered 2,6-dimethylphenyl radical is unable to attach to the surface, but it abstracts an hydrogen atom from ACN to give the cyanomethyl radical (.CH₂CN) that reacts with the surface. A spontaneous reaction is also possible on copper. The film is characterized by IR spectroscopy, scanning electron microscopy, ellipsometry, water contact angles, and cyclic voltammetry. A mechanism is elaborated that accounts for the formation, grafting of the cyanomethyl radical, and finally formation of amino multilayers.

Full text of DOI:10.1021/cm100295n

2962 Chem. Mater. 2010, 22, 2962–2969
DOI:10.1021/cm100295n
Indirect Grafting of Acetonitrile-Derived Films on Metallic Substrates
†
†
†
§
Avni Berisha,†,‡ Catherine Combellas, Frederic Kanoufi, Jean Pinson, Stephane Ustaze,
ꢀ ꢀ
ꢀ
and Fetah I. Podvorica*,†,‡
†Physico-Chimie des Electrolytes, des Colloides et Sciences Analytiques, ESPCI ParisTech, CNRS UMR
7195, 10 rue Vauquelin, 75231 Paris Cedex 05, France, ‡Chemistry Department of Natural Sciences Faculty,
University of Prishtina, rr. “Ne€na Tereze” nr. 5, 10 000 Prishtina, Kosovo, and Alchimer, 15 rue du Buisson
§
aux Fraises, 91300, Massy, France
Received January 29, 2010. Revised Manuscript Received March 17, 2010
Strongly bonded organic films with amino groups are obtained on gold, copper, and silicon surfaces
by reduction of 2,6-dimethyl benzenediazonium in acetonitrile (ACN). The sterically hindered
2,6-dimethylphenyl radical is unable to attach to the surface, but it abstracts an hydrogen atom from
ACN to give the cyanomethyl radical ( CH2CN) that reacts with the surface. A spontaneous reaction is
3
also possible on copper. The film is characterized by IR spectroscopy, scanning electron microscopy,
ellipsometry, watercontactangles, and cyclicvoltammetry. A mechanism iselaboratedthataccountsfor
the formation, grafting of the cyanomethyl radical, and finally formation of amino multilayers.
Scheme 1
Introduction
The reduction of aryl diazonium1-7 and diaryliodo-
nium8 salts on various materials provides an easy method
to ensure the covalent bonding of aryl layers to surfaces.
The key point is the formation of aryl radicals, either by
*Corresponding author. E-mail: fetah.podvorica@fshmn.uni-pr.edu.
(1) (a) Pinson, J.; Podvorica, F. Chem. Soc. Rev. 2005, 34, 429.
electrochemistry, or spontaneously1b (through a dissocia-
tive electron transfer and N2 evolution9), or in the pre-
sence of an added reducing agent.10 Once produced in
aprotic or aqueous media, part of those aryl radicals
reacts with carbon,2b metal,2h semiconductor,2d and poly-
mer10b surfaces according to Scheme 1.
ꢁ
(b) Barriere, F.; Downard, A. J. Solid State Electrochem. 2008, 12,
1231. (c) McCreery, R., L. Chem. Rev. 2008, 108, 2646. (d) Gooding,
J. J. Electroanalysis 2008, 20, 582.
ꢀ
(2) (a) Delamar, M.; Hitmi, R.; Pinson, J.; Saveant, J.-M. J. Am.
Chem. Soc. 1992, 114, 5883. (b) Allongue, P.; Delamar, M.; Desbat, B.;
ꢀ
Fagebaume, O.; Hitmi, R.; Pinson, J.; Saveant, J.-M. J. Am. Chem.
Soc. 1997, 119, 201. (c) Henry de Villeneuve, C.; Pinson, J.; Bernard,
M. C.; Allongue, P. J. Phys. Chem. B 1997, 101, 2415. (d) Allongue, P.;
Henry de Villeneuve, C.; Pinson, J.; Ozanam, F.; Chazalviel, J. N.;
Wallart, X. Electrochim. Acta 1998, 43, 2791. (e) Adenier, A.; Bernard,
M. C.; Chehimi, M. M.; Cabet-Deliry, E.; Desbat, B.; Fagebaume, O.;
Pinson, J.; Podvorica, F. J. Am. Chem. Soc. 2001, 123, 4541.
ꢀ
(4) (a) Saby, C.; Ortiz, B.; Champagne, G. Y.; Belanger, D. Langmuir
ꢀ
1997, 13, 3837. (b) D'Amours, M.; Belanger, D. J. Phys. Chem B 2003,
ꢀ
107, 4811. (c) Baranton, S.; Belanger, D. J. Phys. Chem B 2005, 109,
ꢀ
(f) Chausse, A.; Chehimi, M. M.; Karsi, N.; Pinson, J.; Podvorica, F.;
ꢀ
24401. (d) Chamoulaud, G.; Belanger, D. Langmuir 2004, 20, 4989.
Vautrin-Ul, C. Chem. Mater. 2002, 14, 392. (g) Adenier, A.; Cabet-
Deliry, E.; Lalot, T.; Pinson, J.; Podvorica, F. Chem. Mater. 2002, 14,
ꢀ
(e) Toupin, B.; Belanger, D. Langmuir 2008, 24, 1910. (f) Cougnon, C.;
ꢀ
Mauzeroll, J.; Belanger, D. Angew. Chem., Int. Ed. 2009, 48, 7395.
ꢀ
4576. (h) Bernard, M. C.; Chausse, A.; Cabet-Deliry, E.; Chehimi, M.
(5) (a) Downard, A. J. Langmuir 2000, 16, 9680. (b) Downard, A. J.;
Prince, M. J. Langmuir 2001, 17, 5581. (c) Brooksby, P. A.; Downard,
A. J. Langmuir 2004, 20, 5038. (d) Paulik, M. G.; Brooksby, P. A.;
Abell, A. D.; Downard, A. J. J. Phys. Chem. C 2007, 111, 7808.
(e) Downard, A. J.; Garrett, D. J.; Tan, E. S. Q. Langmuir 2006, 22,
10739. (f) Lehr, J.; Williamson, B. E.; Flavel, B. S.; Downard, A. J.
Langmuir 2009, 25, 13503.
M.; Pinson, J.; Podvorica, F.; Vautrin-Ul, C. Chem. Mater. 2003, 15,
3450. (i) Allongue, P.; Henry de Villeneuve, C.; Cherouvrier, G.; Cortes,
R. J. Electroanal. Chem. 2003, 550, 161. (j) Combellas, C.; Kanoufi,
F.; Pinson, J.; Podvorica, F. I. Langmuir 2005, 21, 280. (k) Adenier, A.;
ꢀ
Cabet-Deliry, E.; Chausse, A.; Griveau, S.; Mercier, F.; Pinson, J.;
Vautrin-Ul, C. Chem. Mater. 2005, 17, 491. (l) Combellas, C.;
Kanoufi, F.; Pinson, J.; Podvorica, F. I. Chem. Mater. 2005, 17,
3968. (m) Combellas, C.; Kanoufi, F.; Pinson, J.; Podvorica, F. I.
(6) (a) Nielsen, L. T.; Vase, K. H.; Dong, M.; Besenbacher, F.;
Pedersen, S. U.; Daasbjerg, K. J. Am. Chem. Soc. 2007, 129, 1888.
(b) Peng, Z.; Holm, A. H.; Nielsen, L. T.; Pedersen, S. U.; Daasbjerg, K.
Chem. Mater. 2008, 20, 6068. (c) Malmos, K.; Iruthayaraj, J.; Pedersen,
S. U.; Daasbjerg, K. J. Am. Chem. Soc. 2009, 131, 13926.
(7) (a) Kariuki, J. K.; McDermott, M. T. Langmuir 1999, 15, 6534.
(b) Kariuki, J. K.; McDermott, M. T. Langmuir 2001, 17, 5947.
(c) Shewchuk, D. M.; McDermott, M. T. Langmuir 2009, 25, 4556.
(8) (a) Vase, K. H.; Holm, A. H.; Norman, K.; Pedersen, S. U.;
Daasbjerg, K. Langmuir 2007, 23, 3786. (b) Matrab, T.; Combellas,
C.; Kanoufi, F. Electrochem. Commun. 2008, 10, 1230.
(9) Andrieux, C.; Pinson, J. J. Am. Chem. Soc. 2003, 125, 14801.
(10) (a) Wildgoose, G. G.; Pandurangappa, M.; Lawrence, N. S.; Jiang,
L.; Jones, T. G. J.; Compton, R. G. Talanta 2003, 60, 887.
(b) Bureau, C.; Pinson, J. Fr Pat 0510914, PCT WO2007048894
(to Alchimer).
ꢀ
Chem. Mater. 2006, 18, 2021. (n) Adenier, A.; Barre, N.; Cabet-Deliry,
ꢀ
E.; Chausse, A.; Griveau, S.; Mercier, F.; Pinson, J.; Vautrin-Ul, C. Surf.
Sci. 2006, 600, 4801. (o) Podvorica, F. I.; Kanoufi, F.; Pinson, J.;
Combellas, C. Electrochim. Acta 2009, 54, 2164.
(3) (a) Liu, Y. C.; McCreery, R. L. J. Am. Chem. Soc. 1995, 117, 11254.
(b) Liu, Y. C.; McCreery, R. L. Anal. Chem. 1997, 69, 2091. (c) DuVall,
S. H.; McCreery, R. L. J. Am. Chem. Soc. 2000, 122, 6759.
(d) Ranganathan, S.; Steidel, I.; Anariba, F.; McCreery, R. L. Nano
Lett. 2001, 1, 491. (e) Itoh, T.; McCreery, R. L. J. Am. Chem. Soc.
2002, 124, 1089. (f) Anariba, F.; DuVall, S. H.; McCreery, R. L. Anal.
Chem. 2003, 75, 3837. (g) Hurley, B. L.; McCreery, R. L. J. Electro-
chem. Soc. 2004, 151, B252. (h) Anariba, F.; Viswanathan, U.; Bocian,
D. F.; McCreery, R. L. Anal. Chem. 2006, 78, 3104. (i) Mahmoud,
A. M.; Bergren, A. J.; McCreery, R. L. Anal. Chem. 2009, 81, 6972.
r
2010 American Chemical Society
pubs.acs.org/cm
Published on Web 03/31/2010

Article
Chem. Mater., Vol. 22, No. 9, 2010 2963
Scheme 2
In an ACN solution, the aryl radicals generated from
film of controllable thickness. In this case, an indirect
grafting of an easily available solvent takes place through
a diazonium salt.
diazonium salts can undergo various reactions that are
presented in Scheme 2: (i) electrografting with the surface
(ksurf), (ii) dimerization in solution (kdim), (iii) reaction
with another aromatic group (either from an aromatic
compound in the solution or already grafted onto the
surface) to give dimers or oligomers in solution or at-
tached to the surface (kAr, Gomberg-Bachmann11 reac-
tion that involves a cyclohexadienyl intermediate), and
(iv) abstraction of an hydrogen atom from the solvent,
Experimental Section
Chemicals. Anhydrous ACN (99.8%; <0.05% water) was
from Aldrich as well as tetrabutylammonium tetrafluoroborate,
which was kept in an oven at 95 °C. Milli-Q water (>18 MΩ cm)
was used for rinsing the samples. 2,6-DMBD was prepared by
the standard procedure.14b
Samples. One (or 2) ꢀ 1 cm2 gold-coated silicon wafer plates
ACN, to give the cyanomethyl radical CH2CN (kH).12
3
˚
(Aldrich, 1000 A coating) were cleaned in a “piranha” solution
If we consider only the reactions with the surface
(i and iii), the reduction of diazonium salts (by electro-
chemistry, a reducing agent, or a reducing surface) pro-
vides aryl radicals that for one part react with the surface,
and for the other part attack the first attached aryl groups,
leading to the growth of a polyphenylene film.13 Pre-
viously, we showed that the steric hindrance is one of the
parameters that controls the polyphenylene film growth:
the two bulky tert-butyl groups of 3,5- di tert-butyl benze-
nediazonium prevent the growth of the film (kAr ≈ 0) and a
monolayer can be obtained;14a the two methyl groups of
2,6-dimethyl benzenediazonium (2,6-DMBD) prevent the
attachment of the radical to the surface (kSurf ≈ 0).14b
However, these two methyl groups should not prevent the
abstraction of an hydrogen atom from the solvent and the
formation of the cyanomethyl radical (iv or Scheme 3
[R2]). Besides, such a radical should be stabilized in the
medium bythe isotopicreaction withACN itself(Scheme 3
[R3]). This paper deals with the reaction of this radical with
the surface and the subsequent formation of an organic
(1/3 v/v H2O2/H2SO4 for 10 min), and rinsed one time with
Milli-Q water under sonication (10 min).
Massive 1 (or 2) ꢀ 1 cm2 copper plates were polished with
different grades of polishing paper and finally with a 0.04 μm
alumina slurry on a polishing cloth, using a Presi Mecatech
234 polishing machine. After polishing, the plates were rinsed
with Milli-Q water, left for 10 min in a 5% aqueous solution of
citric acid, rinsed again with Milli Q water, and finally sonicated
for 10 min in acetone.
P-doped silicon samples (Siltronic, resistivity 0.016 Ω cm)
were hydrogenated by dipping for 5 min in 3% HF, and then
rinsed in Milli-Q water for 5 min under sonication.
The electrodes for cyclic voltammetry were Cu or Au
(diameter, 1 mm) wires imbedded in epoxy resin or a small
silicon shard. Metallic electrodes were polished and rinsed as the
copper plate above.
Electrochemical experiments were performed with an EG&G
263A potentiostat/galvanostat and an Echem 4.30 version soft-
ware. Electrografting was performed by chronoamperometry in
ACN þ 0.1 M NBu4BF4 solutions with a 10 mM solution of
2,6-DMBD. The reference electrode was Ag/AgCl and the
counter electrode a platinum foil. In the case of Si, the modifica-
tion was achieved by chronopotentiometry at E = -0.9 V/Ag/
AgCl for 900 s under light irradiation. The Si sample was
scratched and connected in the back with silver conducting
polymer and a piece of copper tape.
(11) Gomberg, M.; Bachmann, W. E. J. Am. Chem. Soc. 1924, 45, 2339.
ꢀ
(12) M’Halla, F.; Pinson, J.; Saveant, J.-M. J. Am. Chem. Soc. 1980,
102, 4120.
(13) Doppelt, P.; Hallais, G.; Pinson, J.; Podvorica, F.; Verneyre, S.
Chem. Mater. 2007, 19, 4570.
Spontaneous grafting of the copper plates was carried out by
immersing the plates into ACNþ 0.1 M NBu4BF4 solutions in
the presence of 10 or 100 mM of 2,6-DMBD for 30 min.
(14) (a) Combellas, C.; Kanoufi, F.; Pinson, J.; Podvorica, F. I. J. Am.
Chem. Soc. 2008, 130, 8576. (b) Combellas, C.; Jiang, D.; Kanoufi, F.;
Pinson, J.; Podvorica, F. I. Langmuir 2009, 25, 286.

2964 Chem. Mater., Vol. 22, No. 9, 2010
Berisha et al.
Scheme 3
IR spectra of the modified samples were recorded using a
Jasco FT/IR-6100 Fourier Transform Infra Red Spectrometer
equipped with a MCT detector. For each spectrum, 1000 scans
electrografting. On Au, the thickness was also measured from the
SEM image of the limit of the layer of a tilted sample. Thicknesses
of films grafted onto Si were measured by carving a sharp furrow
through the film with a sharp tip and measuring the depth of this
furrow with an R Step IQ profilometer from KLA Tencor. The
length of a chain with multiple CH2CH(NH2) groups was calcu-
were accumulated with a spectral resolution of 4 cm-1
.
TOF-SIMS spectra were obtained with an ION-TOF IV with
Auþ primary ions at 25 keV. The analyzed zone was 150 μm2,
and the acquisition time 75 s. Blank samples were analyzed in the
same runas the modified samples. Thepeak intensity refersto the
area of the peak normalized to the total intensity of the spectrum.
Thicknesses of films grafted onto Au and Cu were measured
with a mono wavelength ellipsometer Sentech SE400. The follow-
ing values were taken for gold, ns = 0.182, ks= 3.489; copper, ns =
0.265, ks = 3.108, and the polymeric layer ns = 1.210, ks = 0.
These values were measured on clean surfaces before electrograft-
ing, and the thickness was determined from the same plate after
˚
lated with ACD ChemSketch 12 after optimization (11.23 A for ten
groups).
SEM images were obtained using a Zeiss Ultra II electron
microscope.
Contact angles were measured by a standard technique, in
which the substrate was cooled by a Pelletier element and a drop
of a few microliters of water was placed on it, using a micro-
syringe. A stereomicroscope allowed the droplet image enlarge-
ment and the contact angle measurement.

Article
Chem. Mater., Vol. 22, No. 9, 2010 2965
under these conditions are shown in the insert of
Figure 2c together with the corresponding bands under
dry conditions (Figure 2a). The IR bands have decreased
by ∼40%.
The electrochemical modification of a gold plate at
E = -0.9 V/Ag/AgCl for 480 s is also possible as shown
by IRRAS in Figure 3a. The IRRAS spectrum is similar
to that obtained on copper (Figure 2a); however, the band
at 1397 cm-1 is absent.
To ascertain the presence of amino groups on the sur-
face, a gold plate modified with ACN was treated with
trifluoroacetic anhydride (5% solution in THF) overnight
to transform primary amine groups into trifluoroaceta-
mide functions (-NH2þ(CF3CdO)2OT-NHC(O)CF3).
The IR spectrum of Figure 3b shows that the NH
deformation at 1617 cm-1 has disappeared and the band
at 3243 cm-1 has decreased; it now corresponds to the NH
stretching of amide functions. The presence of CF3(CO)
groups in the film is evidenced by the CdO band at
1785 cm-1 and the CF3 stretching bands at 1210 and
1170 cm-1 (∼3350, 3200, 1705, 1200, 1155 cm-1 for
trifluoroacetamide, CF3C(dO)NH2).
An ACN-derived film can also be grafted on copper
and iron electrodes by spontaneous reduction of
2,6-DMBD in an ACN solution (without supporting
electrolyte). Figure 4 shows the presence of amino bands
at 3300 and 1600 cm-1 and also of a strong CdO band at
1693 cm-1 as in the case of the electrochemical grafting at
E = -0.45 V/Ag/AgCl. For such a spontaneous grafting,
the intensity of the IR bands increases with the concen-
tration of 2,6-DMBD (from 10 to 100 mM) (compare
Figures 4 a and b).
Figure 1. Chronoamperometry of a copper plate in a 10 mM 2,6-DMBD
solution in ACN þ0.1 M NBu4BF4. Electrode biased at E = -0.9 V/Ag/
AgCl for 480 s.
Trifluoroacetylation. A modified gold plate was immersed
into 5 mL of THF þ 0.5 mL of (CF3CO)2O and a few drops of
triethylamine and left at room temperature overnight.15 The
plate was then rinsed with acetone and distilled water under
sonication.
Results
Electrochemical Measurements. The chronoampero-
metry of a copper plate (s = 1 cm2) in an ACN þ 0.1 M
NBu4BF4 þ 10 mM 2,6-DMBD solution is shown in
Figure 1. It is very different from the chronoamperogram
usually observed with diazonium salts where for t > 5 s,
the current is ∼0. On the contrary, here, for t > 5 s, the
absolute value of the current increases slightly with time,
which indicates that the film formed on the electrode
through a cathodic reaction is not blocking.2m At the end
of the electrolysis, a gray film is observed on the surface.
Similar results were observed with gold and hydrogenated
silicon and the chronoamperograms are presented in
Figures S1 and S2 (see the Supporting Information).
IRRAS Measurements. The IR spectrum of a 2 cm2
copper plate submitted to a constant potential of -0.9 V/
Ag/AgCl for 480 s is presented in Figure 2a. The main
features of this spectrum are the absence of a -CtN stretch-
ing band16 (2254 cm-1 for ACN itself) and the presence of
bands corresponding to grafted amine groups at 1595 cm-1
(NH deformation) and at 3243 cm-1 (-NH stretching). A
band at 1397 cm-1 (CH2 scissoring) is also observed.17
The IR spectrum of a 2 cm2 copper plate submitted to
a lower potential (E = -0.45 V/Ag/AgCl) for 480 s is
presented in Figure 2b. The latter amino bands at 3260
and 1595 cm-1 are still observed, although smaller, and a
new strong CdO band is present at 1693 cm-1. This
indicates that changing the potential for electrografting
results in a different structure for the grafted film.
ToF-SIMS. We have recorded the Tof-SIMS spectrum
of a copper plate spontaneously grafted in ACN þ
100 mM 2,6-DMBD for 30 min (without any supporting
electrolyte). The main features of this spectrum are
summarized in Table 1 and some significant peaks are
presented in Figure 5.
Thickness of the Films. They were measured by ellipso-
metry, profilometry, or SEM and are gathered in Table 2.
The thickness of the layers corresponds to a quite large
number of -CH2-CH(NH2)- groups (see below). Be-
˚
cause the size of this group is 1.1 A along the chain axis,
there should be ∼500 of such groups on both gold and
copper, indicating a very efficient process for the growth
of the layer.
SEM Results. The acetonitrile derived films grafted
electrochemically on gold or copper surfaces via 2,6-di-
methyl phenyl radicals were observed by SEM (Figures 6a,
b and c, d for gold and copper respectively). Figure 6b
shows the interface between the film and the unmodified
gold surface. Other images are given in the Supporting
Information (see Figures S3-S5). The films are homoge-
neous and dense. On gold (Figure 6a,b), the surface is
rather homogeneous with some bumps, and on copper a
clear structure is visible, the origin of which is still unclear
(Figures 6c,d and S5).
Grafting on copper was also achieved under the same
conditions as for Figure 2a except that ACN contained
1% H2O. The main IR bands of the plate modified
(15) Smith, M. B.; March, J. March’s Advanced Organic Chemistry, 5th
ed.; Wiley Interscience: New York, 2001; p 510.
(16) The nitrile band of acetonitrile is located at 2254 cm-1. This
position is very close to that of residual CO2, but careful examina-
tion of the spectra before CO2 correction indicates that, in any case,
the CN band should not exceed the background noise.
Contact Angle Measurements. The water contact angle
measurements before and after treatment are gathered in
(17) Socrates, G. Infrared and Raman Characteristic Group Frequencies,
3rd ed.; John Wiley & Sons: New York, 2008.

2966 Chem. Mater., Vol. 22, No. 9, 2010
Berisha et al.
Figure 2. IRRAS spectra of a copper plate after electrolysis in an ACN solution of 10 mM 2,6-DMBD. Plate biased at E = (a) -0.90 and (b) -0.45 V/Ag/
AgCl for 480 s under dry conditions; inset (c) in the presence of 1% H2O. IR assignments in cm-1: 3243, NH stretching; 1595, NH deformation; 1693, CdO
stretching.
Figure 3. IRRAS spectra of a gold plate after electrolysis in a 10 mM 2,6-DMBD ACN solution. Plate biased at E = -0.9 V/Ag/AgCl for 480 s (a) before
and (b) after reaction overnight with trifluoroacetic anhydride (CF3CdO)2O in a THF solution. IR assignments in cm-1: 3243, NH stretching; 1597, NH
deformation; 1170 and 1210, CF3 stretching.
Table 3 for gold and copper. They indicate that the
surface has been modified by the treatment into a more
hydrophobic one.
previously that, when using a 4 mM solution of
2,6-DMBD, the radical obtained does not attach to metal
surfaces.14b
Besides, when diazonium salts are grafted to a metal,
the structure of the film does not change with the grafting
potential;1 this is not the case with the ACN film pre-
sented above, as the IRRAS spectra are different when
the grafting is performed at -0.45 or -0.90 V (Figure 2).
In addition, in the case of spontaneous grafting, ACN is
the only compound present in the solution besides the
diazonium salt. Therefore, the observed grafted organic
film should originate from ACN.
Discussion
The 2,6-DMBD salt is electrochemically reduced14b
at Ep = -0.25 V/Ag/AgCl (gold electrode, 10 mM
solution in ACN þ 0.1 M NBu4BF4, scan rate v =
100 mV s-1; see Figure S6 in the Supporting In-
formation). In the above experiments, the electrolyses
were performed at -0.9 or -0.45 V/Ag/AgCl to reduce
the aryldiazonium salt into the aryl radical. We showed

Article
Chem. Mater., Vol. 22, No. 9, 2010 2967
Figure 4. IRRAS spectra of a copper plate dipped for 30 min in (a) 10 and (b) 100 mM 2,6-DMBD in an ACN solution. IR assignments in cm-1: 3250, NH
stretching; 1595, NH deformation; 1693, CdO stretching.
Table 1. Tof-SIMS Spectrum of a Copper Plate Modified in ACN þ
anion can be obtained in two ways: by reduction of the
radical or by deprotonation of ACN. The redox potential
2,6-DMBD (positive fragments)
E°( CH2CN/-CH2CN) = -0.69 V/SCE that is -0.66 V/
intensity on intensity on
bare Cu
3
Ag/AgCl has been measured by Gennaro;19 therefore, at
the electrolysis potential of -0.9 V, the cyanomethyl
radical should be reduced to its anion by the electrode
[R5]. At this potential, the 2,6-dimethyl phenyl radical
should also be reduced to its anion [R6]20 and this aryl
anion should also be able to deprotonate ACN [R7].21
In the case of spontaneous reactions, the problem is
more complicated, the only reducing agent in the experi-
ment is copper. The open circuit potential measured for
a copper plate in a 100 mM ACN solution of 2,6-DMBD
is -0.44 V/Ag/AgCl; therefore, reduction of the cyano-
methyl radical should be possible, at least in part. The
cyanomethyl anion should attack the first grafted cyano-
methyl group in the same [R8] reaction as in electrochem-
istry. The attack of the grafted cyanomethyl group by the
m/z
assignment
modified Cu
62.93
64.93
77.04
78.04
91.05
63Cu
65Cu
C6H5
C6H6
5034
2229
289
63
223
120
70
766
336
1423
390
751
1514
620
950
254
C6H4-CH3
105.07 CH3-C6H4-CH3
115.05 (CH3)2C6H4-CH
146.10 CH2-C6H4-CH-CH-NH2
267.17 [C6H4(CH3)-CH-CH(NH2)]2
14
12
In addition, the spectra do not indicate the presence of
nitrile groups, alternatively they clearly present the sig-
nature of amine groups, and this is confirmed by the
trifluoroacetylation reported above. Tof-SIMS spectra
also present nitrogen containing fragments that can be
assigned to amino groups. To account for these observa-
tions, we propose the mechanism reported in Scheme 3.
After its formation [R1], the 2,6-dimethylphenyl radi-
cal can dimerize in solution but the abstraction of an
hydrogen atom from the solvent [R2] is favored by the
large concentration of ACN (17 M) and by the fast rate of
hydrogen atom abstraction (the pseudo first order rate
constant of the reaction of the cyanophenyl radical
measured in ACN12 is kH = 4 ꢀ 107 s-1). Once formed,
the cyanomethyl radical is stabilized by the isotopic
reaction [R3]18 and reacts with the surface in reaction
[R4]. However, the IR spectra do not show any signi-
ficant -CtN bands but present strong amino bands. To
rationalize this fact, we propose that the surface-CH2CN
be attacked by the cyanomethyl anion as in reaction [R8].
When the reaction is performed electrochemically, this
cyanomethyl anion is known as the Thorpe reaction,22
a
reaction analogous to the aldol reaction. In both cases-
electrochemical or spontaneous-one obtains an iminium
ion that should be protonated by the very small amount of
residual water or by abstraction of a proton from ACN,
(19) Isse, A. A.; Gennaro, A. J. Phys. Chem. A 2004, 108, 4180.
(20) (a) The reduction potential of the phenyl radical to its anion
was measured by Andrieux,20b Ep = -0.64 V/SCE, and Vase,20c
Ep = -0.95 V/SCE. The reduction potential of the 2,6-dimethyl
phenyl radical should be slightly more negative because of the effect
of the methyl groups. (b) Andrieux, C. P.; Pinson, J. J. Am. Chem.
Soc. 2003, 125, 14801. (c) Vase, K. Ph.D. Thesis, University of Aarhus,
Aarhus, Denmark, 2007.
(21) (a) The pKa of acetonitrile is 31.321b and that of phenyllitium is
about 43;21c therefore, the 2,6-dimethylbezene anion should be able
to deprotonate acetonitrile to give the cyanomethyl anion.
(b) Rossi, L.; Feroci, M.; Inesi, A. Mini-Rev. Org. Chem. 2005, 2,
79. (c) Smith, M. B.; March, J. March’s Advanced Organic Chemistry,
5th ed.; Wiley Interscience: New York, 2001; p 331.
(18) The increase in the current observed during the chronopotentio-
metry is most likely related to the increasing concentration of the
cyanomethyl radical in the solution .
(22) Smith, M. B.; March, J. March’s Advanced Organic Chemistry,
5th ed.; Wiley Interscience: New York, 2001; p 1238.

2968 Chem. Mater., Vol. 22, No. 9, 2010
Berisha et al.
Figure 5. Some characteristic ToF-SIMS peaks.
bands at 848 and 770 cm-1 can be attributed to aromatic
groups, but the 848 cm-1 signal could also correspond
to an aliphatic amine (strong 856 cm-1 band for iso-
propylamine). This indicates that the number of aryl
groups attached to the chain is limited or small. Indeed,
once formed the 2,6-dimethylphenyl radical can react
either with acetonitrile (17 M) or with the thin organic
layer present on the surface; in view of the relative
concentrations, the reaction with ACN is favored.
The results described here can be compared with the
reaction observed when hydrogenated diamond is treated
with peroxides in ACN.25 The mechanism, which is
shown in Scheme 4, involves the formation of the peroxy
radical that abstracts an hydrogen atom from the dia-
mond surface; the diamond radical also abstracts an
hydrogen from ACN and recombination of a diamond
radical with the cyanomethyl radical provides a surface
that is functionalized with -CH2CN groups. The pre-
sence of the -CtN group is observed through its vibra-
tion at 2271 cm-1 that increases with time. But the IR
spectrum25 does not show any signal corresponding to an
amine. Even if the grafted radical is the same, the layer is
quite different from that observed here; this is due to the
absence of an anion that could attack the first grafted
cyano group.
Table 2. Thickness of the Layers Prepared by Reduction of 2,6-DMBD in
ACN at E = -0.90 V/Ag/AgCl for 480 s
Aua
Cub
Sic
56 ( 1 nm
49 ( 3 nm
9 ( 4 nm
a Measured by SEM. b Measured by ellipsometry. c Measured by
profilometry. [2,6-DMBD] = (a,b) 100 mM, (c) 10 mM.
regenerating the cyanomethyl anion [R9]. Imines are
known to be reducible,23 leading to amines via [R10] that
are characterized through their IR spectra and their trans-
formation into trifluoroacetamide. However, during the
spontaneous reaction, the potential is less negative than
during the electrochemical reaction and only part of the
imines is reduced. As imines are known to be very easily
hydrolyzed,24 theunreducedpartistransformedtoketones
[R11], most probably during the rinsing of the samples as
the experiments were performed under dry conditions.
This is in agreement with the IR spectra that show the
formation of both amines and ketones. The absence of IR
bands corresponding to the terminal nitrile group is a
consequence of the low proportion of nitrile groups com-
pared to -CH2CH(NH2)- groups (1 for 500).
In addition, the ToF-SIMS spectra indicate that the
aryl radical is able to react with the polymeric chain
obtained from acetonitrile, in agreement with some frag-
ments presented in Table 1. Such reaction, if sufficiently
effective, should also be observed by a less sensitive
method such as IR. Then bands similar to those of
2,6-dimethylaniline could be expected, such as CH3
stretching at 2967 and 2921 cm-1 and ring vibrations at
1622 and 1480 cm-1; on the spectrum of Figure 2, the
2967, 2921, and 1480 cm-1 bands are clearly absent. The
This reaction can also be compared with that described
by Deniau et al. who were able to form composite
polynitrophenylene-polyvinyl films by simultaneous re-
duction of the 4-nitrobenzenediazonium salt and a vinylic
compound.26 In this case, the nitrophenyl radicals initiate
the radical polymerization of vinylic monomers under
protic conditions. This reaction has been called SEEP
(surface electroinitiated emulsion polymerization). The
main difference with our procedure is that the 2,6-DMBD
(23) (a) Lund, H. Reduction of Azomethines in Organic Electrochemistry,
4th ed.; Lund, H., Hammerich, O., Eds.; Marcel Dekker: New York,
2001, p435. (b) Reed, R. C.; Wightman, R. M. In Encyclopedia of
Electrochemistry of the Elements; Bard, A. J., Lund, H., Eds.; Marcel
Dekker: New York, Vol. XV.1, pp 54. (c) Largeron, M.; Fleury, M.-B.
Org. Lett. 2009, 11, 883. (d) Largeron, M.; Chiaroni, A.; Fleury, M.-B.
Chem. Eur. J. 2008, 14, 996.
(25) Tsubota, T.; Shintaro, I.; Osamu, H.; Nagaoka, S.; Masanori, M.;
Yasumichi, M. Phys. Chem. Chem. Phys. 2002, 4, 3881.
(26) (a) Deniau, G.; Azoulay, L.; Bougerolles, L.; Palacin, S. Chem.
Mater. 2006, 18, 5421. (b) Tessier, L.; Deniau, G.; Charleux, B.;
Palacin, S. Chem. Mater. 2009, 21, 4261.
(24) Smith, M. B.; March, J. March’s Advanced Organic Chemistry,
5th ed.; Wiley Interscience: New York, 2001; p 1177.

Article
Chem. Mater., Vol. 22, No. 9, 2010 2969
Figure 6. SEM of (a, b) gold and (c, d) copper plates immersed into a 100 mM 2,6-DMBD þ 0.1 M NBu4BF4 ACN solution for (a, b) 900 s at E = -0.9 V/
Ag/AgCl, (c, d) 1800 s without electrochemical assistance. Scale bars: (a) 100 nm, (b) 10 μm (c) 200 nm, (d) 1 μm. Secondary electron analysis.
Table 3. Water Contact Angles
The reaction is triggered by electrochemistry or started
spontaneously with reducing metals such as copper. It
Au (deg)
Cu (deg)
should also be triggered by other reducing metals, such as
iron, nickel, zinc, etc. The growth of the layer results from
the further attack of the cyanomethyl anion on the first
grafted cyanomethyl group. This provides a surface con-
taining amino groups and, in some cases, keto groups. A
mechanism is proposed to account for the formation of
such grafted surfaces.
bare surface
modified surface
44 ( 2
53 ( 4
35 ( 2
52 ( 4
Scheme 4. Grafting of the Cyanomethyl Radical on Hydroge-
nated Diamond Surfaces25
As both amino and keto groups are very reactive, this
method provides surfaces that can be easily further
derivatized, for example, for controlled protein immobi-
lization.27 The method can be extended to other mole-
cules (RH) that can transfer an hydrogen atom to the
radical itself is not grafted on the electrode surface, but is
able to provide radicals from the solvent by hydrogen
atom abstraction. A film derived from acetonitrile is
formed, but in our case, aryl radicals are not incorporated
into the film, at least, to a large extent.
2,6-dimethylphenyl radical to give a new radical (R ) able
to react with a surface. Work is in progress to explore new
grafting reactions along these lines.
3
Acknowledgment. We are grateful to the Alchimer society
for the use of the SEM microscope and to Dr. Xavier Joyeux
for recording the images. The “Agence Nationale de la
Recherche” is gratefully acknowledged for its financial sup-
port via the ANR-06-BLAN-0368 project.
Conclusion
Under electrochemical or spontaneous grafting condi-
tions, the 2,6-dimethyl benzenediazonium salt does not
lead to the modification of the surface by 2,6-dimethyl-
phenyl groups. This is related to the steric hindrance in the
positions ortho to the nitrogen atom. The 2,6-dimethyl
phenyl radical is formed but instead of reacting with the
surface, it abstracts an hydrogen atom from ACN to
give the cyanomethyl radical, that reacts with the surface.
Supporting Information Available: Additional figures (PDF).
This material is available free of charge via the Internet at http://
pubs.acs.org.
(27) Kim, J.; Cho, J.; Seidler, P.; Kurland, N.; Yadavalli, V. Langmuir
2010, 26, 2599.