Novel tripod ligands for prickly self-assembled monolayers

Article abstract of DOI:10.1039/b515727g

The new tridentate thioether ligands PhSi(CH₂SMe)₃ (1) and Ph-p-C₆H₄Si(CH₂SMe)₃ (2) have been synthesised and used for the preparation of the chelates fac-[W(κ³-1)(CO)₃

Full text of DOI:10.1039/b515727g

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Novel tripod ligands for prickly self-assembled monolayers†
Tobias Weidner,a,c Andreas Kra¨mer,b,c Clemens Bruhn,b,c Michael Zharnikov,d Andrey Shaporenko,d
Ulrich Siemeling*b,c and Frank Tra¨ger*a,c
Received 7th November 2005, Accepted 10th January 2006
First published as an Advance Article on the web 10th May 2006
DOI: 10.1039/b515727g
The new tridentate thioether ligands PhSi(CH2SMe)3 (1) and Ph-p-C6H4Si(CH2SMe)3 (2) have been
synthesised and used for the preparation of the chelates fac-[W(j3-1)(CO)3] and fac-[W(j3-2)(CO)3],
which were characterised by single-crystal X-ray diffraction. 1 and 2 were used as tripodal adsorbate
molecules for the fabrication of self-assembled monolayers (SAMs) on gold. Film formation from
solution was investigated in situ by second harmonic generation (SHG) and ellipsometry, which
revealed a two-stepped process (fast adsorption, followed by slow film ordering). SAMs of 2 on gold
were further investigated by ex situ methods, viz. high-resolution X-ray photoelectron spectroscopy
(HRXPS), Fourier transform infrared reflection absorption spectroscopy (FTIRRAS), and scanning
tunneling microscopy (STM). The latter two methods indicated dense packing of the tripodal anchor
groups on the surface, with a substantially lower density of the biphenyl pricks. HRXPS showed three
different binding states of sulfur, including a standard thiolate-type and a coordination-type state.
comparatively low mole fractions of functionalised adsorbate
Introduction
molecules. Even then, however, the terminal units may be partly
Self-assembled monolayers (SAMs) on solid surfaces are of great
buried inside the film due to the flexibility of the oligomethylene
current interest in science and technology.1 SAMs formed by the
aggregation of organosulfur compounds on gold have received
most attention, ever since it was first reported that SAMs on
gold can be prepared conveniently by chemisorption of di-n-alkyl
disulfides or related cyclic derivatives from dilute solutions.2
Adsorbate molecules carrying terminal functional units can
lead to functional SAMs1b,c suitable for such diverse applications
as, for example, electrode modification,3 biosensors,4 molecu-
lar machines5 as well as molecular electronic components and
devices.6 A regular arrangement of functional units is a prerequi-
site for a well-defined and uniform response of functional SAMs
to external stimuli. Highly ordered and stable SAMs have been
obtained especially with n-alkanethiols of suitable chain lengths
(≥ca. 10 carbon atoms), which constitute the archetypal platform
in this context. However, functional units tend to disturb the
lateral interactions responsible for SAM ordering, which is due
to their physical dimensions, polarity, chemical reactivity etc.1b
A popular strategy to circumvent this problem has been the
preparation of mixed SAMs composed of n-alkanethiols of specific
chain length and longer n-alkanethiols carrying a functional unit.
This approach can afford a statistical (and therefore grossly
uniform) distribution of essentially isolated functional units at
tether.7 An alternative single-component strategy is the use of
binding units which cover a surface area whose diameter ideally
exceeds that of the functional unit.8 Adsorbate molecules with
multiple attachment points have been widely used in this context.9
Since a plane is defined by three points, the most straightforward
approach is based on adsorbate molecules with three identical
anchor groups,9a–f although four9g,10 or more11 have also been
realised. Owing to a “surface chelate” effect, oligodentate adsor-
bate molecules can bind particularly strongly to the substrate.
As a result, stable and well packed SAMs on gold have been
reported not only for thiols, but also for thioethers,12 although
these usually lead to films of inferior stability since their interaction
with a gold substrate is weaker and more reversible as compared
with the corresponding thiol functionality.13 Four9g,10 or even only
three9f thioether moieties have been found to be sufficient in this
context.
Depending on the relative size of the binding unit, SAMs
based on the single-component approach just described can
exhibit considerable void space between individual molecular
chains which protrude from the surface. Although the softness
and disorder which arises from the conformational flexibility of
oligomethylene units in loosely packed SAMs of this kind can be
beneficial for certain biomedical applications,9a a more rigid, well
ordered arrangement of terminal units is usually preferred for a
uniform response to external stimuli and is especially desirable in
certain cases such as, for example, high-density data storage based
on arrays of individual molecules.11a,14
aInstitute of Physics, University of Kassel, Heinrich-Plett-Str. 40, D-34132,
Kassel, Germany. E-mail: traeger@physik.uni-kassel.de
bInstitute of Chemistry, University of Kassel, Heinrich-Plett-Str. 40,
D-34132, Kassel, Germany. E-mail: siemeling@uni-kassel.de
cCenter for Interdisciplinary Nanostructure Science and Technology, Univer-
sity of Kassel, Heinrich-Plett-Str. 40, D-34132, Kassel, Germany
We have decided to use rigid-rod type units to minimise the
degree of orientational disorder in loosely packed SAMs on
gold which are based on molecules that contain a tridentate
anchor group. We have chosen the thioether functionality for
surface binding, since thioethers offer several practical advantages.
dApplied Physical Chemistry, University of Heidelberg, Im Neuenheimer
Feld 253, D-69120, Heidelberg, Germany
† Based on the presentation given at Dalton Discussion No. 9, 19–21st
April 2006, Hulme Hall, Manchester, UK.
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 2767–2777 | 2767
©

View Article Online
They are chemically more robust and easier to handle than thiols,
disulfides and thioacetates, and the synthesis of unsymmetrical
thioethers is rather straightforward. Tripodal thioethers have
already proved useful for SAM formation in this context (vide
supra). Our ultimate goal is to fabricate well ordered SAMs with
laterally separated, non-interacting functional units. We report
here our results obtained in an elementary study using C6H5 and
C6H5-p-C6H4 as rigid-rod type units connected to a central Si
atom at the vertex of the tripodal binding unit. We have chosen
Si,9c instead of the more usual C, as the branching atom, since the
ease and versatility of substitution reactions at silicon is highly
attractive from a preparative point of view. On top of that, silicon
may be used as an additional probe in XPS investigations of
SAMs, complementing the information gained from carbon and
sulfur.
In addition to XPS the SAMs have been characterised by optical
second harmonic generation (SHG), Fourier transform infrared
reflection absorption spectroscopy (FTIRRAS), ellipsometry and
scanning tunneling microscopy (STM). Particularly valuable
information has been obtained by second harmonic generation
owing to the high sensitivity and interface specificity of this
technique.15,16
in THF (150 ml). This solution was added dropwise to a stirred
solution of SiCl4 (17.38 g, 102.3 mmol) in THF (200 ml) at −90 ◦C
(liquid N2/n-propanol bath). The reaction mixture was allowed to
warm to room temperature. After 24 h volatile components were
removed in vacuo. The residue was extracted with n-hexane (3 ×
100 ml). The volume of the combined extracts was reduced to
ca. 80 ml. Crystallisation at −20 ◦C afforded the product as a
1
colourless, microcrystalline solid. Yield 7.58 g (78%). H NMR
(CDCl3): d 7.42 (m, 1 H), 7.50 (m, 2 H), 7.62 (“d”, apparent
J = 8.3 Hz, 2 H), 7.74 (“d”, apparent J = 8.3 Hz, 2 H), 7.90
1
(“d”, apparent J = 7.8 Hz, 2 H). 13C{ H} NMR (CDCl3): d 127.2,
127.3, 128.3, 129.0, 130.0, 133.7, 139.9, 145.6. Calc. for C12H9Cl3Si
(287.65): C, 50.11; H, 3.15. Found: C, 51.33; H, 3.27%.
{Tris[(methylthio)methyl]silyl}benzene (1).
A solution of
phenyltrichlorosilane (1.95 g, 9.2 mmol) in n-hexane (100 ml)
cooled to −90 ◦C was added dropwise over a period of 2 h
to a stirred suspension of (methylthio)methyllithium (2.66 g,
39.1 mmol) in n-hexane (100 ml) at −90 ◦C (liquid N2/n-propanol
bath). The reaction mixture was allowed to warm to room
temperature and was filtered after 14 h. Volatile components were
removed in vacuo. The product was obtained as a colourless liquid
after distillation at 10−3 mbar (bp ca. 120 ◦C). Yield 2.06 g (78%).
1H NMR (CDCl3): d 2.18 (s, 9 H, Me), 2.24 (s, 6 H, CH2), 7.40 (m,
Experimental
1
3 H, Ph), 7.68 (m, 2 H, Ph). 13C{ H} NMR (CDCl3): d 16.8 (Me);
20.4 (CH2); 127.9, 130.2, 133.0, 134.2 (Ph). Calc. for C12H20S3Si
(288.58): C, 49.95; H, 6.99. Found: C, 49.84; H, 6.93%.
Preparative work
General considerations. Synthetic work involving air-sensitive
compounds was performed under an atmosphere of dry ni-
trogen by using standard Schlenk techniques or a conven-
tional glovebox. Solvents were appropriately dried and puri-
fied. Phenyltrichlorosilane was procured from Lancaster, dis-
tilled and stored over magnesium turnings. 4-Bromobiphenyl
and magnesium turnings were purchased from Merck and used
as received. [W(CO)3(EtCN)3] was prepared according to the
standard protocol.17 (Methylthio)methyllithium was synthesised
from dimethyl sulfide by slight modification of a published two-
step procedure;18 we found that washing of the final product
with benzene (typically 3 × 20 ml for 100 mmol of LiCH2SCH3)
was necessary to remove tenacious n-Bu4Sn. NMR spectra were
recorded with a Varian Unity INOVA 500 spectrometer operating
at 500.13 MHz for 1H. IR spectra were obtained with a BIO-RAD
FTS-40A spectrometer. The purity of 1 and 2 was checked by
GC-MS analysis using a Varian Saturn GC/MS equipped with
an ion trap using EI (70 eV) with internal ionisation. The spectra
were recorded and processed with the Varian Saturn II software
Version 4.0 by the accumulation and averaging of several single
spectra for each GC peak using background subtraction. The
GC used a 25 m silica capillary column HP 5 MS (0.2 mm ID,
4-{Tris[(methylthio)methyl]silyl}biphenyl (2). This compound
(1.72 g, 73%) was obtained as a colorless liquid (bp ca. 190 ◦C at
10−3 mbar) from (methylthio)methyllithium (1.94 g, 28.5 mmol)
and 4-(trichlorosilyl)biphenyl (1.85 g, 6.43 mmol) by a procedure
1
analogous to that described for the preparation of 1. H NMR
(CDCl3): d 2.21 (s, 9 H, Me), 2.28 (s, 6 H, CH2), 7.36 (“t”, apparent
J = 7.3 Hz, 1 H, C12H9), 7.44 (m, 2 H, C12H9), 7.61 (m, 4 H,
1
C12H9), 7.76 (“d”, apparent J = 8.3 Hz, 2 H, C12H9). 13C{ H}
NMR (CDCl3): d 16.9 (Me); 20.6 (CH2); 126.7, 127.2, 127.6, 128.8,
131.6, 134.8, 140.8, 142.9 (C12H9). Calc. for C18H24S3Si (364.67):
C, 59.29; H, 6.63. Found: C, 59.38; H, 6.65%.
fac-[W(j3-1)(CO)3]. 1 (130 mg, 0.45 mmol) was added to a
stirred solution of [W(CO)3(EtCN)3] (182 mg, 0.42 mmol) in a
mixture of benzene (10 ml) and THF (40 ml). After 20 h the
volume of the solution was reduced to ca. 2 ml. The pale greenish
yellow crude product precipitated upon addition of diethyl ether
(2 ml) and was purified by dissolving it in THF (5 ml) and layering
1
the solution with diethyl ether (3 ml). Yield 200 mg (83%). H
NMR (THF-d8): d 2.56 (s, 6 H, CH2), 2.84 (s, 9 H, Me), 7.40 (m,
1
3 H, Ph), 7.52 (m, 2 H, Ph). 13C{ H} NMR (THF-d8): d 18.0 (Me);
◦
31.2 (CH2); 129.1, 131.5, 133.5, 134.7 (Ph); 215.7 (CO). IR (THF):
m(CO) 1912 (vs), 1785 (vs). Calc. for C15H20O3S3SiW (556.48): C,
32.38; H, 3.62. Found: C, 32.19; H, 3.65%.
0.3◦lm film) at a heating rate of 10 C min−1 (start temperature
◦
80 C, final temperature 280 C for 25 min). Elemental analyses
were performed by Mikroanalytisches Laboratorium H. Kolbe
(Mu¨lheim an der Ruhr).
fac-[W(j3-2)(CO)3]. This compound was prepared in 78%
yield in analogy to the synthesis described for fac-[W(j3-1)(CO)3].
1H NMR (THF-d8): d 2.60 (s, 6 H, CH2), 2.85 (s, 9 H, Me), 7.34
(“t”, apparent J = 7.3 Hz, 1 H, C12H9), 7.43 (m, 2 H, C12H9),
7.62 (m, 4 H, C12H9), 7.69 (“d”, apparent J = 8.3 Hz, 2 H,
4-(Trichlorosilyl)biphenyl. Previously reported syntheses of
this compound are cumbersome, since they involve high-
temperature reactions.19 We have therefore developed a conve-
nient wet-chemical route, which involves a Grignard reaction. A
solution of the Grignard reagent was prepared from Mg turnings
(0.85 g, 35.0 mmol) and 4-bromobiphenyl (7.90 g, 33.9 mmol)
1
C12H9). 13C{ H} NMR (THF-d8): d 18.3 (Me); 31.5 (CH2); 127.9,
128.1, 129.0, 130.0, 132.3, 135.6, 141.5, 144.7 (C12H9); 216.0 (CO).
2768 | Dalton Trans., 2006, 2767–2777
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
IR (THF): m(CO) 1915 (vs), 1789 (vs). Calc. for C21H24O3S3SiW
Sample preparation for physical studies
(632.55): C, 39.88; H, 3.82. Found: C, 39.80; H, 3.94%.
In our optical investigations by second harmonic generation and
ellipsometry, gold films with a thickness of 200 nm were used
as substrates for the self-assembled monolayers. The films were
prepared by the deposition of Au atoms on a Si(100) wafer.
A 15 nm titanium interlayer served for adhesion promotion.
Pieces with a size of 1 × 1 cm2 were cut from the gold-covered
silicon wafer, extensively rinsed with acetonitrile and dried in a
stream of high purity argon. They were then mounted in a home
built TeflonTM cuvette, whose quartz glass (SuprasilTM) window
was sealed with a chemrazTM O-ring. The cuvette was initially
filled with pure acetonitrile (uvasolTM, Merck). Preparation of
SAMs on the Au films was accomplished by replacing the pure
solvent by a solution of 1 or 2 in acetonitrile (uvasolTM grade,
Merck).
Samples for Fourier transform infrared reflection absorption
spectroscopy (FTIRRAS) were prepared by immersion of the
gold-covered silicon wafer pieces in an acetonitrile solution of
2 (30 mmol l−1) for 24 h and subsequent rinsing with pure
acetonitrile. The samples were dried in a stream of pure N2
and stored under Ar atmosphere until they were loaded into the
spectrometer.
In addition to the optical investigations, we have also carried out
experiments by high-resolution scanning tunneling microscopy
(STM) to study the lateral ordering of the SAMs. For these
measurements atomically flat gold substrates, rather than films
with their intrinsic roughness, are needed. We have prepared such
single crystal gold surfaces by a template-stripping method.22 For
SAM preparation, these Au surfaces were immersed in a solution
of 2 in acetonitrile (3 mmol l−1) for 2 h. They were subsequently
rinsed with pure acetonitrile and dried in an Ar flow.
X-Ray crystallography
A single crystal of fac-[W(j3-1)(CO)3] and fac-[W(j3-2)(CO)3]·
0.5THF, respectively, was mounted on a glass fibre and all geomet-
ric and intensity data were taken from this sample. Data collection
˚
using Mo-Karadiation (k = 0.71073 A) was performed on a STOE
IPDS II diffractometer equipped with a two-circle goniometer and
an area detector. The data sets were corrected for Lorentz and
polarisation effects. The structures were solved by direct methods
(SHELXS97) and refined using alternating cycles of least squares
refinements against F2 (SHELXL97).20 All non H atoms were
found in difference Fourier maps and were refined with anisotropic
displacement parameters, except those of the disordered sulfur
atoms in fac-[W(j3-2)(CO)3], which occur on two positions with
72 and 28% probability, respectively. The positions with the higher
occupancy were refined with anisotropic displacement parameters;
those with minor occupancy were refined with isotropic displace-
ment parameters. The disorder of the sulfur atoms automatically
affects the three carbonyl ligands. This effect was included in
the model through unusually high displacement parameters. As a
consequence, W–CO and C–O bond lengths were refined shorter
than expected. As a further consequence, the S–W–CO angles
in the disordered species are meaningless. For both structures
H atoms were placed in constrained positions according to the
riding model with the 1.2 fold isotropic displacement parameters.
H-Atom positions of the disordered molecule were included in
the model only regarding the positions with higher occupancy.
Graphical representations were made using ORTEP-3 win.21
Crystallographic data for fac-[W(j3-1)(CO)3]
Pale green platelet, dimensions 0.32 × 0.25 × 0.10 mm, T =
Optical second harmonic generation
˚
213(2) K, a = 13.1471(8), b = 23.761(2), c = 12.7676(8) A,
Film formation and ordering was monitored in situ and in real time
by optical second harmonic generation (SHG). The measurements
were performed with a pulsed Nd:YAG laser (Quanta-Ray GCR-
170, Spectra Physics) operating at a repetition rate of 10 Hz and
at a wavelength of 1064 nm. The pulse duration was 7 ns and the
laser beam was incident on the sample surface under an angle of
45◦. The second harmonic light generated was separated from the
fundamental light beam by several colour glasses (BG 39, OG 590,
Schott) and filters in combination with a monochromator, and was
finally detected by a photomultiplier operating in single photon
counting mode. Its signal was processed by a boxcar averager and
stored in a computer. Full details concerning the experimental
set-up have been published elsewhere.23–26
A small fraction of the light emitted by the laser was branched
off by a dielectric beam splitter and directed onto a y-cut quartz
plate in order to generate an SH reference signal. It was detected
and processed as explained above and served to normalise the
second harmonic light intensity of the sample in order to minimise
the influence of intensity fluctuations of the laser light.
b = 103.895(5)◦, monoclinic, space group P21/c (no. 14), V =
3
−3
˚
3871.7(5) A , Z = 8, F(000) = 2160, Dc = 1.909 g cm , l =
6.362 mm−1, Tmax/Tmin = 0.57/0.24. Within the h range of 1.6 and
25.5◦ 26213 reflections were collected, 7152 unique (Rint = 0.0435),
of which 5825 were observed [I > 2r(I)]. At final convergence, R1 =
0.0193, wR2 = 0.0382 [I > 2r(I)] and R1 = 0.0283, wR2 = 0.0395
(all data), for 422 parameters.
Crystallographic data for fac-[W(j3-2)(CO)3]·0.5THF
Pale green platelet, dimensions 0.09 × 0.18 × 0.30 mm, T = 173(2)
◦
˚
K, a = 9.8433(4), b = 41.729(2), c = 13.1267(5) A, b = 91.271(3) ,
3
˚
monoclinic, space group P21/n (no. 14), V = 5390.4(4) A , Z = 8,
F(000) = 2640, Dc = 1.648 g cm−3, l = 4.586 mm−1, Tmax/Tmin
=
0.697/0.336. Within the h range of 1.6 and 25.0◦ 45135 reflections
were collected, 9369 unique (Rint = 0.0803), of which 6676 were
observed [I > 2r(I)]. At final convergence, R1 = 0.0403, wR2 =
0.0858 [I > 2r(I)] and R1 = 0.0643, wR2 = 0.0926 (all data), for
585 parameters.
In addition to the SH intensity, we have determined the phase of
the nonlinear signal at several stages during film formation. Details
of the experimental arrangement have been described elsewhere.27
A gas cell was used for varying the delay time and a y-cut quartz
plate behind the sample served as a second SH source.
CCDC reference numbers 269773 and 269774 for fac-[W(j3-
1)(CO)3] and fac-[W(j3-2)(CO)3]·0.5THF, respectively.
For crystallographic data in CIF or other electronic format see
DOI: 10.1039/b515727g
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 2767–2777 | 2769
©

View Article Online
Ellipsometry
which is noticeably smaller than the full widths at half max-
imum (FWHM) of the photoemission peaks addressed in this
study.
The adsorption of 1 on gold was investigated in situ by ellipsometry
with a Woollam variable angle spectrometric ellipsometer. The
measurements were carried out in situ in the TeflonTM cell also used
for the SHG experiments (vide supra). The angle of the incident
light was set to 70◦. The evolution of the ellipsometric parameterW
was followed as a function of immersion time at a wavelength of
750 nm.
HRXPS spectra were fitted by symmetric Voigt functions and
either a Shirley type or linear background. To fit the S 2p3/2,1/2
doublets we used a pair of such peaks with the same FWHM, a
branching ratio of 2 (2p3/2/2p1/2), and spin–orbit splitting (verified
by fit) of ca.1.18 eV (2p3/2/2p1/2).32 The fits were performed self-
consistently; the same peak parameters were used for identical
spectral regions. The accuracy of the resulting BE/FWHM values
is 0.02–0.03 eV.
Fourier transform infrared spectroscopy
Fourier transform infrared spectra were recorded with a BIO-
RAD FTS-40a spectrometer in attenuated total reflection (ATR)
geometry. The spectral resolution was 2 cm−1. Fourier transform
infrared reflection absorption spectra of the SAMs on gold were
obtained using an evacuated Bruker IFS 66v/S spectrometer
equipped with a liquid nitrogen cooled mercury cadmium telluride
detector. p-Polarised light was incident on the sample under an
angle of 84◦. A total of 2000 scans were measured. The spectra
obtained were averaged over three neighbouring data points. Prior
to the SAM a spectrum of a clean, gold covered wafer piece was
recorded as a reference.
Results and discussion
Synthetic work and crystal structures
Compounds 1 and 2 were prepared in analogy to the synthesis of
33
MeSi(CH2SMe)3 by reaction of the respective aryltrichlorosi-
lane with 3 equiv. of lithiated dimethyl sulfide in n-hexane
(Scheme 1). Both compounds were isolated in ca. 75% yield as
colourless liquids. In previous investigations of SAMs fabricated
from thioethers it was noted that thiol impurities coordinate to
gold more strongly and therefore concentrate at the interface
preferentially.12,34 Consequently, in order to obtain meaningful
results in studies with thioethers, it is crucial to exclude such
impurities. Although the presence of thiol impurities seemed rather
unlikely considering the synthetic route used for the preparation
of 1 and 2, we checked the purity of both compounds by GC-
MS analysis. The thiol concentration was below detection limit
in each case. 1 contained n-Bu4Sn as contaminant (ca. 0.4%).
This is due to the fact that lithiated dimethyl sulfide, which was
prepared by transmetallation from n-BuLi and n-Me3SnCH2SMe,
contained trace amounts of n-Bu4Sn. The biphenyl derivative 2
also contained traces of n-Bu4Sn (ca. 0.1%); the major contam-
inant was biphenyl (ca. 1.0%), which was already present in the
4-(trichlorosilyl)biphenyl starting material.
Scanning tunneling microscopy
STM measurements of SAMs of 2 on gold were carried out under
ambient conditions at room temperature using a Veeco CP–R
Thermomicroscope equipped with a small area scanner. Tips were
prepared by mechanical cutting of a Pt : Ir (80 : 20) wire with
a diameter of 0.25 mm. The scanner was calibrated by recording
images of graphite surfaces with atomic resolution. The images
of the SAM were taken in constant current mode using a tip bias
of 1.5 V and a tunneling current of 100 pA. The scan speed was
10 nm s−1.
X-Ray photoelectron spectroscopy
The SAMs were characterised by synchrotron-based high-
resolution X-ray photoelectron spectroscopy (HRXPS). SAMs
formed from 4-biphenylthiol (BPT) on Au(111)/mica by following
a published procedure28 were used for reference purposes. All
experiments were performed at room temperature. The measure-
ments were carried out under UHV conditions at a base pressure
better than 1.5 × 10−9 mbar. The spectra acquisition time was
selected in such a way that no noticeable damage by the primary
X-rays occurred during the measurements.29
Scheme 1
The HRXPS experiments were performed at the bending
magnet beamline D1011 at the MAX II storage ring of the MAX-
lab synchrotron radiation facility in Lund, Sweden. The HRXPS
spectra were collected in normal emission geometry at photon
energies of 350 and 580 eV for the C 1s range and 350 eV for the
S 2p and Si 2p regions. In addition, Au 4f spectra were acquired
and the O 1s range was monitored. The binding energy (BE) scale
of every spectrum was individually calibrated using the Au 4f7/2
emission line of an alkanethiol-covered Au substrate at 83.95 eV.
The latter value is given by the latest ISO standard.30 This is
very close to a value of 83.93 eV, which has been obtained by
us for Au 4f7/2 using a separate calibration to the Fermi edge of a
clean Pt foil.31 The energy resolution was better than 100 meV,
In order to probe the binding capacity of 1 and 2 to-
wards zero-valent metal atoms, 1 and 2 were reacted with fac-
[W(CO)3(EtCN)3] in THF. The reactions proceeded cleanly and
swiftly at room temperature, affording fac-[W(j3-1)(CO)3] and fac-
[W(j3-2)(CO)3], respectively, in ca. 80% yield as pale greenish yel-
low crystals. Their structures were determined by X-ray diffraction
and are similar to that of fac-[W{j3-MeSi(CH2SMe)3}(CO)3].33 In
each case, the coordination of the tungsten atom is best described
as distorted pseudo-octahedral, with a tridentate face-capping
thioether ligand. Both asymmetric units contain two different
molecules, whose bond parameters are similar. In the case of fac-
[W(j3-2)(CO)3], one of the two molecules in the asymmetric unit
2770 | Dalton Trans., 2006, 2767–2777
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
exhibits disorder of the MeS and CO units, which restricts a
detailed discussion of bond parameters to the non-disordered
molecule shown in Fig. 1.
Film measurements
Second harmonic generation. The SHG signal of the clean
substrate immersed in pure acetonitrile was recorded for reference
purposes. Subsequently, the pure solvent was replaced by a
solution of the adsorbate molecules in acetonitrile, and the SHG
signal was instantly monitored in situ as a function of time.
Fig. 2 examplarily shows the time dependent SHG signal recorded
during immersion of the substrate in an acetonitrile solution of
1 at a concentration of 30 lmol l−1. The SHG intensity of the
clean surface has been set to unity. At t < 0 the signal of the clean
substrate in acetonitrile is shown. Distortions during exchange of
the pure solvent for the solution of 1 at t = 0 s leads to an abrupt
decrease of the signal for a short period of time. Subsequently, the
actual change of the SHG signal is detected. It decreases from 1.0
to a value of 0.55 0.04 within approximately (4000 1200) s.
For comparison, Fig. 2 also displays the SHG signal recorded
during film formation of 2 on gold under identical conditions. In
this case, the signal drops off to 0.49 0.04 of its original value
within approximately (15000 1000) s. The decrease of the SHG
intensity to a slightly lower value in the latter case suggests that a
higher final surface coverage is achieved with 2.
Fig. 1 Structure of the non-disordered molecule of fac-[W(j3-2)(CO)3] in
the crystal.
In both compounds the three carbonyl ligands are roughly
opposite to the three sulfur atoms, with S–W–Ctrans angles ranging
from ca. 170 to 178◦. The W–C bond lengths are almost identical,
˚
their mean value being 1.96 A. The W–S bond lengths are also
˚
very similar to each other, with an average value of 2.55 A. In
the case of fac-[W(j3-1)(CO)3] the S–W–S angles range from
84.17(3) to 91.20(3)◦ for the first molecule and from 81.72(3)◦
to 92.06(3) for the second, which indicates that coordinated 1
still exhibits conformational flexibility. This is further supported
by the range of W–S–C angles observed (between 107.9(1) and
112.3(2)◦ for W–S–CH3, between 111.5(1) and 113.9(1)◦ for W–
S–CH2). In contrast, the C–S–C angles are very similar to each
other. Their mean value of 100.7◦ deviates only slightly from the
values of 99.1 and 99.2◦ found for Me2S in the gas phase35 and
in the solid state,36 respectively. Obviously, metal coordination
has no significant distorting effect on the thioether moieties and
still allows considerable conformational flexibility. Essentially the
same conclusion may be drawn from the structural parameters
of the non-disordered molecule of fac-[W(j3-2)(CO)3] (S–W–S
angles between 83.06(6) and 87.50(6)◦; W–S–CH3 angles between
108.3(3) and 111.5(3)◦, W–S–CH2 angles between 111.7(3) and
113.6(2)◦; C–S–C angles between 99.9(4) and 101.8(4)◦).
Both tungsten complexes exhibit two very strong m(C–O) bands
at ca. 1915 cm−1 (A1 mode) and 1785 cm−1 (E mode) in the
solution IR spectrum, characteristic for hexacoordinate metal
tricarbonyl complexes with time-averaged C3v symmetry. It is
interesting to note that coordination of a W(CO)3 fragment leads
to a pronounced downfield shift (Dd ca. 0.6 ppm) of the CH3
resonance signal in the 1H NMR spectrum of 1 and 2, respectively.
Fig. 2 SHG signal recorded in real time during adsorption of 1 and 2,
respectively, on a gold substrate from acetonitrile solution (30 lmol l−1).
In general, the total second-order nonlinear susceptibility
(2)
(2)
vtot
,
27,37 is the sum of three contributions (eqn (1)). vsub(2), vin . and
(2)
vads denote the susceptibilities of the substrate, of the chemical
interaction between the substrate and the adsorbate, and of the
adsorbate, respectively; H is the surface coverage.
(2)
(2)
vtot = vsub + vin(2)(H) + vads(2)(H)
(1)
(2)
The susceptibility of the substrate vsub remains constant
upon adsorption. On the other hand, the second-order nonlinear
susceptibility of phenyl derivatives at the applied fundamental
(2)
wavelength of 1064 nm is negligible, i.e. vads = 0. This is
supported by results that we obtained from SHG measurements
using MeSi(CH2SMe)3, which is the methyl analogue of 1.33 CH3 is
known to have a negligible second-order nonlinear susceptibility
at k = 1064 nm.38 The CH3 tripod gave rise to exactly the
same decrease of the SHG signal as the C6H5 (1) or C6H5-
(2)
p-C6H4 tripod (2). We therefore conclude that vads vanishes.
Consequently, the temporal variation of the total second-order
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 2767–2777 | 2771
©

View Article Online
nonlinear susceptibility is exclusively due to the interaction term
(2)
vin(2). In fact, vin reflects the change of the nonlinear electronic
properties of the substrate/film interface induced by interaction
with the adsorbate molecules.
A drop-off of the SHG signal, which is observed in our
study (Fig. 2), has already been noted before for thiols27,39 and
thioethers34a and can be explained, broadly speaking, by a decrease
of the number of free electrons at the interface through localisation
in chemical bonds. However, the exact nature of the chemical
interaction of the tripodal thioethers 1 and 2 with a gold surface
cannot be probed by SHG at present. Our results from molecular
coordination chemistry demonstrate that 1 and 2 are suitable for
the coordination of zero-valent metal centres (vide supra). One can
therefore expect a coordination-type interaction of such adsorbate
molecules with a gold surface, which should be inherently weaker
than that shown by covalently bound thiolate units on gold.
In fact, desorption enthalpies determined for SAMs fabricated
from monodentate thiols and thioethers, respectively, on gold
reveal a difference of ca. 50 kJ mol−1 between both bond types.13
The coordination-type interaction exhibited by thioethers may be
classified as a weak chemisorption in view of the corresponding
desorption enthalpy values, which lie close to ca. 80 kJ mol−1 for
this class of compounds; the chemisorption of thiols, on the other
hand, can be classified as strong.
Fig. 3 Concentration dependence of the saturation time of the SHG
signal. The saturation time remains constant at high concentrations, where
adsorption is not rate limiting. The high saturation times result from
ordering processes after adsorption.
above 10000 lmol l−1). This is rather surprising since, according to
our experience with sulfur-based adsorbate molecules, saturation
of the SHG signal is reached extremely rapidly at such high concen-
trations and consequently the time resolution of the experiment
does not allow to follow the adsorption process. We note that
prior to our study SHG measurements performed with thiols and
thioethers have revealed that thioethers give rise to comparatively
long saturation times.34a In these investigations, however, the
thioether concentration (1000 lmol l−1) was more than one order
of magnitude lower than the maximum concentration used in the
present work.
The results obtained for the second concentration regime
indicate that the adsorption process is not rate limiting at higher
concentrations. This means that monolayer adsorption proceeds
on a timescale of only a few seconds and molecular ordering
and conformational alignment of the initially randomly adsorbed
molecules becomes the most important factor for film assembly
at high concentrations. In the case of conventional thiols and
related monodentate adsorbate molecules the ordering process
mainly concerns the lateral arrangement of the functional and
spacer groups. In contrast, the self-organisation described here for
1 and 2 also affects the binding of the anchor groups. This view
is further supported by the identical constant assembly time of 1
and 2 at high concentrations and by ellipsometric measurements
(vide infra).
A change of the nonlinear susceptibility by chemical bonds
(2)
implies a linear dependence of vin on the coverage H(t), where
the term “coverage” refers to the number of chemically bound
adsorbate molecules per unit area, as long as we make the plausible
assumption of independent binding sites. The second harmonic
signal ISHG is proportional to the square of the second-order
nonlinear susceptibility. The coverage dependence is described by
eqn (2), where C is a constant and Ifund denotes the intensity of the
fundamental light.27,37
ISHG = C|vsub − vin(H = 1)(2)H(t)2|I2
(2)
(2)
fund
The second-order nonlinear susceptibilities are complex third
rank tensors that are characterised by their amplitude and phase.
If phase changes occur because of nearby resonances, the coverage
dependence of the signal can be quite complicated27,40 and does
not readily allow the straightforward interpretation described
above. Here, however, this complication can be ruled out since
the constant relative phase between the substrate and interaction
contributions during chemisorption of 1 and 2 has been verified
experimentally at several stages of film formation.
We have also performed concentration dependent SHG mea-
surements for 1 and 2 in order to learn more about kinetic
aspects of film formation and ordering. In each case, different
concentrations give rise to different times needed until the SHG
signal reaches its final saturation value. This is shown in Fig. 3,
where two different concentration regimes can be identified for
both compounds. We note a low concentration regime with a
strong dependence of the saturation time in the range between
3 and 30 lmol l−1 for 1 and between 30 and 1000 lmol l−1 for 2.
The saturation times decrease with growing concentration in this
regime and vary between approximately 15000 and 4000 s, quite
typical for SAMs.39 Adsorption takes place nearly two times faster
in the case of 1 as compared to 2. At higher concentrations a second
regime can be made out, where the saturation time was found
to remain essentially constant (even at very high concentrations
Ellipsometry
Fig. 4 displays the ellipsometric parameter W measured during
adsorption of 1 on gold from an acetonitrile solution (3 lmol
l−1). W increases until saturation is reached after about (1500
500) s. The final value of the ellipsometric signal is reached
approximately seven times faster than that of the SHG signal at
the same concentration (Fig. 3).
We pause to note that the ellipsometric data was collected at
a particularly low concentration. Higher values as used in most
of our experiments gave rise to an immediate, step-like change
of the ellipsometric signal, completely lacking time resolution. In
order to understand the different temporal behaviour observed by
SHG and ellipsometry, we briefly have to consider the origin of the
2772 | Dalton Trans., 2006, 2767–2777
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Fig. 4 Ellipsometric parameter W recorded in situ during adsorption of
1 on a gold substrate from acetonitrile solution (3 lmol l−1). After (1500
500) s the signal reaches saturation, which corresponds to monolayer
coverage.
signal in ellipsometry; W is related to the ellipsometric parameter
q by41
Fig. 5 Model of the film formation process operative for 1 (not shown)
and 2. (a) Molecules adsorb in a random orientation and conformation
until monolayer coverage is reached. Since not all thioether anchor groups
interact with the gold substrate, the SHG signal does not saturate even
though the adsorption process is essentially complete. (b) Molecular
ordering and conformational adjustment leads to an array of adsorbed
molecules at the interface. Adsorption only takes place at sites that have
become accessible during the ordering process and is not rate limiting.
The SHG signal saturates after this step has ended, because a maximum
number of thioether moieties now interacts with the substrate.
˜
˜
q = Rp/Rs = tan W exp(iD)
(3)
˜
˜
Rp and Rs are the effective complex reflection coefficients of
the interface for p- and s-polarised light, respectively; tan W is the
˜
˜
absolute value of the ratio of Rp and Rs, while D denotes their
phase relation. The value of W shows a complicated dependence
on surface coverage as well as layer thickness.41 W changes during
SAM formation from solution as a function of growing layer
thickness and saturates at full monolayer coverage. The time
needed for signal saturation is a direct measure for the adsorption
time of one monolayer. In other words, in our experiment,
ellipsometry basically provides a measure of the amount of 1 that
is in contact with the surface, irrespective of the nature of the
substrate interaction. In contrast, the change of the SHG signal
reflects the number of chemical bonds of thioether anchor groups
with the substrate.
Fourier transform infrared spectroscopy
While compounds 1 and 2 give reproducible and matching results
by applying the in situ methods described above, 1 does not show a
consistent picture in our ex situ investigations that will be described
in the following. The reason for this unsatisfactory behaviour of
1 is not quite clear and requires further studies. We will therefore
concentrate on the ex situ measurements performed with SAMs
of 2.
Fig. 6 shows the Fourier transform infrared reflection absorp-
tion spectrum of a monolayer of 2 along with an FTIR spectrum
of the bulk compound. Table 1 compares the vibrational bands
Different saturation times of the SH and ellipsometric sat-
uration therefore imply that (weak) adsorption of a complete
monolayer of 1 takes place much more rapidly than full binding
of all anchor groups. The slow decrease of the SHG signal is
compatible with the notion that after adsorption the molecules
rotate and undergo conformational changes until the maximum
possible number of anchor groups have finally come into contact
with the surface forming chemical bonds.
We can conclude at this stage, film formation basically proceeds
in two steps. First, the molecules adsorb in a random orientation
until monolayer coverage is reached (Fig. 5(a)). This process is
concentration dependent and takes only a few seconds at high
concentrations. The second step (Fig. 5(b)) consists of molecular
reorientation and conformational adjustment to allow a maximum
number of thioether moieties access to the surface. Adsorption
only takes place at those sites that have become accessible during
the ordering process and is not rate limiting. This process is
independent of the concentration, takes approximately 4000 s and
ends up with a maximum number of sulfur-containing moieties
attached to the surface.
Fig. 6 Top: FTIR reflection absorption spectrum of a monolayer of 2 on
a gold substrate. Bottom: ATR-FTIR spectrum of neat 2.
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 2767–2777 | 2773
©

View Article Online
Table 1 Comparison of IR data of bulk 2 with those of 2 adsorbed on
the monolayer spectrum except for the dominant C–H wagging
mode around 1260 cm−1. The dominance of aliphatic modes
indicates that even close packing of the sterically demanding
tripodal anchor groups leads to a comparatively low density
of the aromatic groups. This view is supported by the fact
that the aromatic C–H stretching has been clearly observed for
monolayers fabricated from PhSH43 and Ph-p-C6H4SH,44 which
is due to their high density of aromatic units. Alternatively, the
low intensity of the aromatic C–H stretching modes could also
be due to the orientation of the molecules (with the aromatic
rings lying essentially flat on the surface). According to FTIRRAS
theory, only those vibrations can be excited on surfaces that have
transition dipole moments (TDM) parallel to the surface normal.45
The aromatic C–C and Si–C stretching modes can be clearly made
out in the SAM spectrum of 2 at around 1100 cm−1. Both modes,
like the aromatic C–H stretching modes discussed above, have
TDMs parallel to the Si–Cipso axis. We can therefore conclude
that the diminution of aromatic C–H stretching modes is not due
to molecular orientation on the surface, but indeed reflects the
comparatively low surface density of biphenyl units.
golda
Bulk
SAM
Mode
Direction of TDM
3022
2970
2910
2878
1596
1543
1485
1435
1422
1384
1313
1263
1137 sh
1115
1073
3022
2952
2910
nd
1601
1548
1495
nd
nd
nd
nd
1252
1137
1104
nd
m C–Harom
ma, Me, ma CH2
ms Me
ꢀ Si–Cipso axis
ms
m C–Carom
m C–Carom
m C–Carom
m Si–phenyl
ꢀ Si–Cipso axis
ꢀ Si–Cipso axis
ꢀ Si–Cipso axis
ꢀ Si–Cipso axis
ꢁ
da Me
d CH2
ds Me
x CH2
m C–Carom
m Si–C/C–Carom
m C–Carom
ꢀ Si–Cipso axis
ꢀ Si–Cipso axis
ꢀ Si–Cipso axis
a Only modes relevant for the interpretation of the spectral data are listed.
Band assignments are based on refs 42. nd: not detected.
In summary, the IR spectra suggest a dense packing of the
tripodal head groups and a lower density if biphenyl groups in the
monolayer.
of monolayer and bulk spectra which are relevant to the interpre-
tation with respect to molecular packing. The assignment of the
group frequencies was done by following standard procedures.42
The bulk spectra basically consist of the three parts, viz. (i) the
aromatic and aliphatic C–H stretching vibrations observed above
and below 3000 cm−1; (ii) the C–C stretching bands of aromatic
rings found around 1500 cm−1, where semicircle and two-quadrant
stretching modes are observed and around 1000 cm−1, where
more semicircle stretching modes are located; (iii) the bending
vibrations of the aliphatic moieties, which give rise to bands
around 1300 cm−1.
In the spectrum of the monolayer film characteristic differences
as compared to the bulk are observed. The aromatic C–H
stretching vibration at 3022 cm−1 in the bulk spectrum can be
found just above the noise level in the SAM spectrum. In contrast,
the aliphatic C–H stretching vibrations can be observed quite
clearly; only the aliphatic C–H bending modes are absent in
Scanning tunneling microscopy
Fig. 7 displays a constant-current STM topography of a SAM
of 2 on a template-stripped gold substrate. The image shows the
lateral arrangement of 2 on a Au(111) terrace. Its interpretation
is based on the plausible assumption that the terminal phenyl
group is imaged here. This follows from earlier work on SAMs
of biphenyl-based thiols,46 which were imaged under comparable
tunneling conditions and may be expected to exhibit a similar
conductivity as the molecules studied here.
As inferred from height profiles along the lines (a) and (b) shown
in Fig. 7, the molecules exhibit a regular intermolecular spacing
˚
of (7.9
0.8) A along cross section (a), while along direction
Fig. 7 STM image of a SAM of 2 on an Au(111) single crystal surface. Tunneling current 100 pA, bias voltage 1.5 V, scan speed 10 nm s−1. High
frequency noise has been removed. The intermolecular distances were obtained from profiles in the directions of lines (a) and (b) and were found to be
◦
˚
7.9 and ca. 13.2 A. (a) and (b) include an angle of 85 .
2774 | Dalton Trans., 2006, 2767–2777
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
˚
the shoulder can be deduced from a weak asymmetry of the
observed C 1s peak. The BE position of the main emission in
2 on Au (284.6 eV) is significantly higher than that in BPT on Au
(284.17 eV), which suggests that the aromatic layers are somewhat
different. The FWHM of the main C 1s emission in 2 is larger
than that observed for BPT on gold, which also implies a larger
structural inhomogeneity of the former film.
(b) a broader distribution of lateral distances of (13.2 2.5) A
is observed. The lines (a) and (b) include an angle of 85◦. These
results indicate a structural incommensurability with the subjacent
Au(111) lattice.
The mean surface area occupied per molecule is (102
30)
2
˚
A , as compared to the corresponding van der Waals dimensions
of a Ph-p-C6H4 unit of only 21.1 A (cross section in the plane
2
˚
perpendicular to the aromatic ring plane).46 In accord with our IR
spectroscopic results (vide supra), the density of Ph-p-C6H4 units
on the surface is much lower than in comparable SAMs of
monodentate thiol-based biphenyl derivatives, which typically
show packing densities corresponding to mean surface areas
The S 2p spectra of BPT and 2 on gold shown in Fig. 8 exhibit
several S 2p3/2,1/2 doublets related to individual sulfur moieties.
There are doublets at ca. 161.0–161.2, ca. 162.0 and ca. 163.3–
163.4 eV (S 2p3/2). A doublet at ca. 162.0 eV is commonly assigned
to a thiolate species.31,50 It is a highly dominant feature in the
S 2p spectrum of BPT on gold, which suggests a thiolate-type
attachment to this substrate for the vast majority of the BPT
molecules. In the case of 2 on gold, the intensity of the thiolate-
related doublet is comparable to those of the features at ca. 161.0–
161.2 and ca. 163.3–163.4 eV, which implies a variety of different
chemical states for the anchor groups of 2 in the SAM. This is
compatible with previous observations which show that thioethers
can adsorb non-destructively on gold,34b,c,51 although this need
not be the case, because C–S bond cleavage can also occur. In
the case of (n-H25C12)2S the dichotomy of C–S bond cleavage vs.
intact adsorption has been shown to depend on subtleties of film
preparation.52 If all adsorbate molecules had the same adsorption
character, the relative weights of the three doublets observed in
the S 2p spectrum of 2 on gold would follow a 1 : 1 : 1 ratio. This
is clearly not the case. Consequently, either a distribution of the
adsorption geometries for the individual molecules or formation
of domains with different molecular arrangements and anchor
characters can be concluded.
2
46,47
˚
between ca. 21.6 and 27.5 A occupied per molecule.
A rough
estimate of the surface area occupied by a single tripodal anchor
group can be obtained from the van der Waals dimensions of the
(methylthio)methyl unit. Applying pr2 on an estimated length of
ca. 5.8 A for this unit affords a surface area of ca. 106 A , in accord
with the STM data.
In summary, 2 forms ordered SAMs on gold. The molecules are
densely packed in one direction forming rows along line (a) shown
in Fig. 7. The distances between these rows are larger and also
exhibit larger variations.
2
˚
˚
High-resolution X-ray photoelectron spectroscopy
Normalised C 1s and S 2p HRXP spectra of 4-biphenylthiol
(BPT) and 2, respectively, on gold are presented in Fig. 8, along
with the corresponding fits. Si 2p HRXP spectra have also been
measured and show the presence of the Si branching atom of the
tripodal anchor group. The effective thicknesses of BPT and 2 on
gold derived from the IC 1s/IAu 4f intensity ratios using previously
The chemical identity of the species associated with the doublets
at ca. 161.0–161.2 and 163.3–163.4 eV deserves special consider-
ation. A binding energy of ca. 163.3–163.4 eV is characteristic
of unbound S, but a weak coordination-type binding is more
likely.34b The feature at ca. 161 eV can be ascribed either to atomic
sulfur53 or to a differently (other than by a typical thiolate bond)
adsorbed SAM constituent.54 The former assignment stems from
reported attenuation lengths48 are 11.7 and 10.04 A, respectively.
˚
The main emission observed in the C 1s spectra of both SAMs can
be assigned to the aromatic backbone present in both BPT and 2;
a weak shoulder on the high binding energy side (at 285.11 eV
for BPT/Au), merged with the main peak, can be tentatively
ascribed to the carbon atoms bonded to S.49 The presence of
Fig. 8 Left: S 2p HRXPS spectra of BPT and 2 on Au acquired at a photon energy of 350 eV, along with the corresponding fits (solid lines). The spectra
are decomposed into the individual contributions (solid lines) related to different sulfur species (see text for details). A background is shown. Right: C 1s
HRXPS spectra of BPT and 2 on Au acquired at a photon energy of 350 eV.
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 2767–2777 | 2775
©

View Article Online
experiments with n-alkanethiolate SAMs on Au annealed at 415–
490 K. It was observed that the thiolate-related doublet shifted
to ca. 161.0 eV upon cleavage of the C–S bond in the SAM
constituents.53 The latter assignment is based on kinetic studies
of the SAM formation. The doublet at ca. 161.0 eV was observed
to appear at the early stage of molecular assembly, i.e. at short
immersion times, and disappear later on.54 Under the conditions of
our experiments, it is quite unlikely that the molecules decompose
on the Au surface to chemisorbed sulfur. Therefore, we favour
the other possibility that the SAM of 2 contains some thiolate-
type units in a non-standard binding mode. The difference between
the two putative thiolate-type species, which give rise to the
doublets at ca. 161.0–161.2 and 162.0 eV, is probably due to a
hybridisation change or to adsorption site differences, adsorption
at a defect site (e.g. at the step edge), or a different molecular
orientation (see Fig. 5(a)).
We note that doing HRXPS we frequently observed the doublet
at 161.0 eV for different thiol-derived SAMs on both Au(111) and
Ag(111).55 This doublet is also present in the S 2p spectra of BPT
on gold. All this indicates that such differently adsorbed sulfur
seems to be a general phenomenon. Further studies combining
HRXPS and STM will be necessary to find out the exact nature
of this species.
the monodentate thiol Me-(p-C6H4)2SH,47 and essentially the
same holds true for SAMs based on n-alkane thiols carrying a
terminal Ph-p-C6H4 unit.46 The HRXPS S 2p spectra obtained
from SAMs of 2 on gold reflect three different binding states of the
sulfur, viz. (i) unbound or, much more likely, coordinatively bound
sulfur;34b (ii) standard thiolate-type attachment of the sulfur to the
substrate; (iii) a third binding type which most likely is due to
thiolate-type sulfur in a hybridisation different from that of the
standard type present in (ii). These three binding types have all
been observed before for thioethers on gold.52,54 Since no clear
1 : 1 : 1 stoichiometry of these states has been observed for 2, a
uniform binding behaviour of the adsorbate molecules does not
occur. At most, different domains are present in the SAM, each
one containing individual molecules which show an essentially
identical binding behaviour.
Further ex situ experiments with ligand 1 will be carried out in
order to elucidate the reasons for the rather poor SAM quality
once the films are exposed to ambient conditions.
Our results obtained with 2 augur well for the validity of the
general concept to fabricate prickly SAMs. It is evident that the size
of the anchor group will have to be increased in order to achieve
non-interacting pricks. We are currently extending the structural
concept used here to molecules carrying terminal functional
groups such as, for example, redox-active units and metal ion
binding sites. We envisage that such compounds can be utilised
as building blocks for larger architectures and multiresponsive
functional surfaces. We also conclude from our present study
that the tripodal anchor unit will have to be modified in order
to facilitate dense packing of the adsorbate molecules through
lateral interactions. This may potentially be achieved by replacing
the methyl groups of the thioether moieties by longer n-alkyl chains
or by introducing aromatic substituents capable of -stacking. We
will report on our findings in due course.
Summary and conclusions
Wehave preparedPhSi(CH2SMe)3 (1) and Ph-p-C6H4Si(CH2SMe)3
(2) in high yield from the respective aryltrichlorosilane. 1 and 2
are tripodal rigid-rod type molecules, which contain three identical
thioether units attached to a central Si branching atom. They act
as tridentate ligands towards the W(CO)3 fragment. Coordination
of the zero-valent metal centre still allows a considerable degree
of conformational flexibility for the tridentate binding unit.
We have used 1 and 2 for the preparation of self-assembled
monolayers on gold surfaces. Adsorption at the interface from
solution takes place on a timescale of typically several minutes.
Film growth is accompanied by formation of ordered structures
on a much longer timescale. We have been able to distinguish
between adsorption and ordering of the molecules by using in
situ techniques, in particular optical second harmonic generation.
The experimental data are in accord with the following scenario
for film formation. In a first step, the adsorbate molecules
adsorb in a random orientation, not necessarily all of their three
thioether units pointing towards the surface. This concentration
dependent process is rate limiting at low concentrations. At high
concentrations it is completed within a few seconds. The second
step involves conformational adjustment until a maximum number
of thioether moieties are bound to the surface. This step takes
approximately 4000 s for both compounds, irrespective of their
aryl substituent, and is rate limiting at high concentrations.
The stability of SAMs formed from 1 on gold was insufficient
to obtain meaningful results from ex situ FTIRRAS, STM and
HRXPS measurements. In the case of the biphenyl derivative 2,
however, these methods demonstrate that this compound indeed
forms stable SAMs on gold. Furthermore, FTIRRAS and STM
reveal dense packing of the tripodal anchor groups on the gold
surface, while the density of the biphenyl groups by necessity is
substantially lower. For comparison, the density of biphenyl units
is four to five times larger in the case of SAMs fabricated from
Acknowledgements
We are grateful to Prof. A. Pucci and D. Enders (University
of Heidelberg) for providing the experimental setup for the
FTIRRAS measurements and to Dr M. Letzel (University of
Bielefeld) for expertly performing the GC-MS analyses. We also
thank T. Spehr (University of Kassel) for helping us with the
ellipsometric measurements. The substrates were provided by Dr
T. Saragi and D. Guthermuth (University of Kassel), which we
gratefully acknowledge. We thank Prof. M. Grunze (University of
Heidelberg) for the support of this work, A. Ku¨ller (University of
Heidelberg) for providing us with a sample of BPT, and MAX-lab
staff for the assistance during the synchrotron-based experiments.
U. S. thanks Prof. K. Weising (Chair of Botany, University of
Kassel) for explaining to him the differences between thorns,
spines and pricks. This work has been supported by the German
BMBF (05KS4VHA/4) and European Community (Access to
Research Infrastructure action of the Improving Human Potential
Programme) and by the Fonds der Chemischen Industrie.
References
1 For reviews, see: (a) S. Onclin, B. J. Ravoo and D. N. Reinhoudt, Angew.
Chem., 2005, 117, 6438; S. Onclin, B. J. Ravoo and D. N. Reinhoudt,
Angew. Chem., Int. Ed., 2005, 44, 6282; (b) J. C. Love, L. A. Estroff,
J. K. Kriebel, R. G. Nuzzo and G. M. Whitesides, Chem. Rev., 2005,
2776 | Dalton Trans., 2006, 2767–2777
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
105, 1103; (c) F. Schreiber, J. Phys.: Condens. Matter, 2004, 16, R881;
(d) F. Schreiber, Prog. Surf. Sci., 2000, 65, 151; (e) V. Chechik, R. M.
Crooks and C. J. M. Stirling, Adv. Mater., 2000, 12, 1161; (f) A. Ulman,
Chem. Rev., 1996, 96, 1533.
2 R. G. Nuzzo and D. L. Allara, J. Am. Chem. Soc., 1983, 105, 4481.
3 For a recent review, see: J. J. Gooding, F. Mearns, W. Yang and J. Liu,
Electroanalysis, 2003, 15, 81.
28 A. Shaporenko, K. Heister, A. Ulman, M. Grunze and M. Zharnikov,
J. Phys. Chem. B, 2005, 109, 4096.
29 (a) M. Zharnikov and M. J. Grunze, Vac. Sci. Technol. B, 2002, 20, 1793;
(b) M. Wirde, U. Gelius, T. Dunbar and D. L. Allara, Nucl. Instrum.
Methods Phys. Res. B, 1997, 131, 245; (c) B. Ja¨ger, H. Schu¨rmann, H. U.
Mu¨ller, H.-J. Himmel, M. Neumann, M. Grunze and C. Wo¨ll, Z. Phys.
Chem., 1997, 202, 263; (d) K. Heister, M. Zharnikov, M. Grunze, L. S. O.
Johansson and A. Ulman, Langmuir, 2001, 17, 8.
4 For a recent review, see: N. K. Chaki and K. Vijayamohanan, Biosens.
Bioelectron., 2002, 17, 1.
30 Surface chemical analysis—X-ray photoelectron spectrometers—
Calibration of the energy scales, ISO 15472:2001.
5 See, for example: (a) S. Saha, L. E. Johansson, A. H. Flood, H.-R.
Tseng, J. I. Zink and J. F. Stoddard, Small, 2005, 1, 87; (b) K. Kim,
W. S. Jeon, J.-K. Kang, J. W. Lee, S. Y. Jon, T. Kim and K. Kim, Angew.
Chem., 2003, 115, 2395; K. Kim, W. S. Jeon, J.-K. Kang, J. W. Lee, S. Y.
Jon, T. Kim and K. Kim, Angew. Chem., Int. Ed., 2003, 42, 2293.
6 For a review, see:J. M. Tour, Acc. Chem. Res., 2000, 33, 791.
7 For a characteristic example, see: C. E. D. Chidsey, C. R. Bertozzi,
T. M. Putvinski and A. M. Mujsce, J. Am. Chem. Soc., 1990, 112, 4301.
8 Seminal paper: J. K. Whitesell and H. K. Chang, Science, 1993, 261,
73.
9 Selected recent references for SAMs on gold: (a) J.-S. Park, A. N. Vo, D.
Barriet, Y.-S. Shon and T. R. Lee, Langmuir, 2005, 21, 2902; (b) L. Wei,
K. Padmaja, W. J. Youngblood, A. B. Lysenko, J. S. Lindsey and D. F.
Bocian, J. Org. Chem., 2004, 69, 1461; (c) H. Jian and J. M. Tour, J. Org.
Chem., 2003, 68, 5091; (d) L. Zhu, H. Tang, Y. Harima, K. Yamashita,
Y. Aso and T. Otsubo, J. Mater. Chem., 2002, 12, 2250; (e) D. Hirayama,
K. Takimiya, Y. Aso, T. Otsubo, T. Hasobe, H. Yamada, H. Imahori, S.
Fukuzumi and Y. Sakata, J. Am. Chem. Soc., 2002, 124, 532; (f) K. W.
Kittredge, M. A. Minton, M. A. Fox and J. K. Whitesell, Helv. Chim.
Acta, 2002, 85, 788; (g) H. Scho¨nherr, G. J. Vansco, B.-H. Huisman,
F. C. M. van Veggel and D. N. Reinhoudt, Langmuir, 1999, 15, 5541.
10 (a) B.-H. Huisman, D. M. Rudkevich, F. C. J. M. van Veggel and
D. N. Reinhoudt, J. Am. Chem. Soc., 1996, 118, 3523; (b) B.-H.
Huisman, E. U. Thoden van Velzen, J. F. J. Engbersen and D. N.
Reinhoudt, Tetrahedron Lett., 1995, 36, 3273; (c) E. U. Thoden van
Velzen, F. C. J. M. Van Veggel, J. F. J. Engbersen and D. N. Reinhoudt,
J. Am. Chem. Soc., 1994, 116, 3597.
31 K. Heister, M. Zharnikov, M. Grunze and L. S. O. Johansson, J. Phys.
Chem. B, 2001, 105, 4058.
32 J. F. Moulder, W. E. Stickle, P. E. Sobol and K. D. Bomben, Handbook of
X-ray Photoelectron Spectroscopy, ed. J. Chastian, Perkin-Elmer Corp.,
Eden Prairie, MN, 1992.
33 H. W. Yim, L. M. Tran, E. D. Dobbin, D. Rabinovich, L. M. Liable-
Sands, C. D. Incarvito, K.-C. Lam and A. L. Rheingold, Inorg. Chem.,
1999, 38, 2211.
34 (a) C. Jung, O. Dannenberger, Y. Xu, M. Buck and M. Grunze,
Langmuir, 1998, 14, 1103; (b) C.-J. Zhong, R. C. Brush, J. Anderegg
and M. D. Porter, Langmuir, 1999, 15, 518; (c) J. L. Trevor, K. R. Lykke,
M. J. Pellin and L. Hanley, Langmuir, 1998, 14, 1664.
35 T. Iijima, S. Tsuchiya and M. Kimura, Bull. Chem. Soc. Jpn., 1977, 50,
2564.
36 N. W. Mitzel and U. Losehand, Z. Naturforsch., Teil B, 2004, 59, 635.
37 (a) R. M. Corn and D. A. Higgins, Chem. Rev., 1994, 94, 107; (b) K. B.
Eisenthal, Chem. Rev., 1996, 96, 1343.
38 Nonlinear Optical Properties of Organic Molecules and Crystals, ed.
D. S. Chemla and J. Zyss, Academic Press, New York, 1987, vol. 1.
39 O. Dannenberger, M. Buck and M. Grunze, J. Phys. Chem. B, 1999,
103, 2202.
40 M. Buck, F. Eisert, M. Grunze and F. Tra¨ger, Ber. Bunsenges. Phys.
Chem., 1997, 3, 399.
41 R. M. A. Azzam, Ellipsometry and polarized light, North-Holland,
Amsterdam, 1987.
42 (a) D. Lin-Vien, H. B. Colthup, W. G. Fateley and J. G. Grasselli, The
Handbook of Infrared and Raman Characteristic Frequencies of Organic
Molecules, Academic Press, San Diego, CA, 1991; (b) G. Varsanyi,
Assignments for Vibrational Spectra of Seven Hundred Benzene Deriva-
tives, Adam Hilger, London, 1974; (c) N. P. G. Roeges, A Guide to
the Complete Interpretation of Infrared Spectra of Organic Structures,
Wiley, Chichester, 1994.
11 See, for example: (a) L. Zobbi, M. Mannini, M. Pacchioni, G.
Chastanet, D. Bonacchi, C. Zanardi, R. Biagi, U. Del Pennino, D.
Gatteschi, A. Cornia and R. Sessoli, Chem. Commun., 2005, 12, 1640;
(b) M. W. J. Beulen, J. Bu¨gler, B. Lammerink, F. A. J. Geurts, E. M. E. F.
Biemond, K. G. C. van Leerdam, F. C. J. M. van Veggel, J. F. J.
Engbersen and D. N. Reinhoudt, Langmuir, 1998, 14, 6424.
12 Seminal paper: E. B. Troughton, C. D. Bain, G. M. Whitesides, R. G.
Nuzzo, D. L. Allara and M. D. Porter, Langmuir, 1988, 4, 365.
13 D. J. Lavrich, S. M. Wetterer, S. L. Bernasek and G. Scoles, J. Phys.
Chem. B, 1998, 102, 3456.
43 L.-J. Wan, M. Terashima, H. Noda and M. Osawa, J. Phys. Chem. B,
2000, 104, 3563.
44 H.-T. Rong, S. Frey, Y.-J. Yang, M. Zharnikov, M. Buck, M. Wu¨hn, C.
Wo¨ll and G. Helmchen, Langmuir, 2001, 17, 1582.
14 C. Li, J. Ly, B. Lei, W. Fan, D. Zhang, J. Han, M. Meyyappan, M.
Thompson and C. Zhou, J. Phys. Chem. B, 2004, 108, 9646.
15 R. W. Boyd, Nonlinear Optics, Academic Press, London, 1992.
16 Y. R. Shen, Annu. Rev. Phys. Chem., 1989, 40, 327.
17 G. J. Kubas, L. S. van der Sluys, R. A. Doyle and R. J. Angelici, Inorg.
Synth., 1990, 28, 29.
18 D. J. Peterson, J. Organomet. Chem., 1971, 26, 215.
19 See, for example: B. I. Vainshtein, L. P. Bogovtseva, V. L. Rogachevskii,
T. L. Krasnova and E. A. Chernyshev, Russ. J. Gen. Chem., 1977, 47,
2345.
45 (a) J. T. Young, F. J. Boerio, Z. Zhang and T. L. Beck, Langmuir, 1996,
12, 1219; (b) M. K. Debe, J. Appl. Phys., 1984, 55, 3354.
46 P. Cyganik, M. Buck, W. Azzam and C. Wo¨ll, J. Phys. Chem. B, 2004,
108, 4989.
47 T. Y. P. Leung, P. Schwartz, G. Scoles, F. Schreiber and A. Ulman, Surf.
Sci., 2000, 458, 34.
48 C. L. A. Lamont and J. Wilkes, Langmuir, 1999, 15, 2037.
49 A. Shaporenko, M. Grunze and M. Zharnikov, J. Electron. Spectrosc.
Rel. Phenom., in press.
50 P. E. Laibinis, G. M. Whitesides, D. L. Allara, Y.-T. Tao, A. N. Parikh
and R. G. Nuzzo, J. Am. Chem. Soc., 1991, 113,, 7152.
51 M. W. J. Beulen, B.-H. Huisman, P. A. van der Heijden, F. C. J. M. van
Veggel, M. G. Simons, E. M. E. F. Biemond, P. J. de Lange and D. N.
Reinhoudt, Langmuir, 1996, 12, 6170.
20 SHELXS97 and SHELXL97, Programs for crystal structure solution
and refinement, G. M. Sheldrick, University of Go¨ttingen, Germany,
1997.
21 L. J. Farrugia, J. Appl. Crystallogr., 1999, 32, 837.
22 P. Wagner, M. Hegner, H.-J. Gu¨nterodt and S. Semenza, Langmuir,
1995, 11, 3867.
23 U. Siemeling, D. Rother, C. Bruhn, H. Fink, T. Weidner, F. Tra¨ger, A.
Rothenberger, D. Fenske, A. Priebe, J. Maurer and R. Winter, J. Am.
Chem. Soc., 2005, 127, 1102.
24 M. Buck, M. Grunze, F. Eisert, J. Fischer and F. Tra¨ger, J. Vac. Sci.
Technol. A, 1992, 10, 926.
25 O. Dannenberger, M. Buck and M. Grunze, J. Phys. Chem. B, 1999,
103, 2202.
52 H. Takiguchi, K. Sato, T. Ishida, K. Abe, K. Yase and K. Tamada,
Langmuir, 2000, 16, 1703.
53 Y.-W. Yang and L.-J. Fan, Langmuir, 2002, 18, 1157.
54 (a) M. Himmelhaus, I. Gauss, M. Buck, F. Eisert, Ch. Wo¨ll and M.
Grunze, J. Electron. Spectrosc. Rel. Phenom., 1998, 92, 139; (b) T.
Ishida, M. Hara, I. Kojima, S. Tsuneda, N. Nishida, H. Sasabe and W.
Knoll, Langmuir, 1998, 14, 2092; (c) T. Ishida, N. Choi, W. Mizutani,
H. Tokumoto, I. Kojima, H. Azehara, H. Hokari, U. Akiba and M.
Fujihira, Langmuir, 1999, 15, 6799.
26 M. Buck, F. Eisert, J. Fischer, M. Grunze and F. Tra¨ger, Appl. Phys. A,
1991, 53, 552.
27 M. Buck, F. Eisert, M. Grunze and F. Tra¨ger, Appl. Phys. A, 1995, 60,
1.
55 See for example: (a) A. Shaporenko, M. Brunnbauer, A. Terfort, L. S. O.
Johansson, M. Grunze and M. Zharnikov, Langmuir, 2005, 21,, 4370;
(b) K. Heister, H.-T. Rong, M. Buck, M. Zharnikov, M. Grunze and
L. S. O. Johansson, J. Phys. Chem. B, 2001, 105, 6888.
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 2767–2777 | 2777
©