ARTICLE pubs.acs.org/Langmuir
Functionalized Fullerenes in Self-Assembled Monolayers Maria del Carmen Gimenez-Lopez,† Minna T. R€ais€anen,‡ Thomas W. Chamberlain,† Uli Weber,§ Maria Lebedeva,† Graham A. Rance,† G. Andrew D. Briggs,§ David Pettifor,§ Victor Burlakov,§ Manfred Buck,*,‡ and Andrei N. Khlobystov*,† †
School of Chemistry, University of Nottingham, NG7 2RD, United Kingdom EaStCHEM School of Chemistry, University of St. Andrews, North Haugh, St. Andrews, KY16 9ST, United Kingdom § Department of Materials, University of Oxford, Oxford, OX1 3PH, United Kingdom ‡
bS Supporting Information ABSTRACT: Anisotropy of intermolecular and molecule substrate interactions holds the key to controlling the arrangement of fullerenes into 2D self-assembled monolayers (SAMs). The chemical reactivity of fullerenes allows functionalization of the carbon cages with sulfur-containing groups, thiols and thioethers, which facilitates the reliable adsorption of these molecules on gold substrates. A series of structurally related molecules, eight of which are new fullerene compounds, allows systematic investigation of the structural and functional parameters defining the geometry of fullerene SAMs. Scanning tunnelling microscopy (STM) measurements reveal that the chemical nature of the anchoring group appears to be crucial for the long-range order in fullerenes: the assembly of thiol-functionalized fullerenes is governed by strong molecule surface interactions, which prohibit formation of ordered molecular arrays, while thioether-functionalized fullerenes, which have a weaker interaction with the surface than the thiols, form a variety of ordered 2D molecular arrays owing to noncovalent intermolecular interactions. A linear row of fullerene molecules is a recurring structural feature of the ordered SAMs, but the relative alignment and the spacing between the fullerene rows is strongly dependent on the size and shape of the spacer group linking the fullerene cage and the anchoring group. Careful control of the chemical functionality on the carbon cages enables positioning of fullerenes into at least four different packing arrangements, none of which have been observed before. Our new strategy for the controlled arrangement of fullerenes on surfaces at the molecular level will advance the development of practical applications for these nanomaterials.
’ INTRODUCTION The iconic shape of fullerenes—hollow polyhedra built of sp2hybridized carbon atoms—is strongly associated with the field of molecular nanotechnology. Apart from their aesthetic appeal, these molecules and their derivatives possess a range of electronic, optical, magnetic, and chemical properties that can be exploited for a wide spectrum of technological applications ranging from photovoltaics to quantum computing. In most cases, the properties of fullerenes are studied for random ensembles in solution. However, the full technological potential of these materials can only be harnessed if the arrangement of fullerenes can be controlled.1 Pristine, unfunctionalized fullerenes readily self-assemble into hexagonal close packed (hcp) arrays (Figure 1a) when deposited on surfaces.2 Hcp arrays are formed almost invariably as the fullerene fullerene and fullerene substrate interactions are typically governed by low directional van der Waals forces. In our approach, we utilize chemical modification of the exterior of the fullerene cage to address the lack of structural diversity in 2D fullerene arrays. A chemical group attached to the fullerene r 2011 American Chemical Society
surface can in principle be used to introduce a specific, highly directional interaction with the substrate (Figure 1b) or to induce anisotropy in the intermolecular interactions within the 2D arrays, thus providing a mechanism for directing the assembly of fullerenes into 2D geometries other than hcp. There is a wide spectrum of different functional groups that may modify the energy and directionality of interactions between fullerene and a chosen surface. Thiols, for example, are known to form strong, directional interactions with many metal surfaces, leading to the spontaneous formation of highly ordered SAMs.3 Therefore, a thiol or any other sulfur-containing group grafted to the C60 cage is expected to play a powerful role in directing the surface assembly of such fullerenes (Figure 1b). However, the majority of studies detailing thiol-functionalized fullerenes deposited on surfaces observe the formation of disordered structures.4 10 The two examples of ordered fullerene SAMs Received: February 19, 2011 Revised: July 8, 2011 Published: July 11, 2011 10977
dx.doi.org/10.1021/la200654n | Langmuir 2011, 27, 10977–10985
Langmuir
ARTICLE
Scheme 1. The Versatile Nature of 1,3-Dipolar Cycloaddition Reactions Enables the Introduction of a S-Containing Anchoring Group to Either the Nitrogen Atom (a) or the Carbon Atom (b) of the Pyrrolidine Ring, Depending on the Structure of the Aldehyde and Amino Acid Precursors Figure 1. Schematic representations of (a) unfunctionalised fullerene molecules adopting a hexagonal close packed array on a Au(111) surface and (b) a functionalized fullerene molecule attached to a Au(111) surface through a sulfur based anchoring group (X = H or alkyl chain).
Scheme 2. Synthesis of N-(4-Methylthiophenyl)glycine (3SR-A)
Figure 2. Thioether and thiol functionalized fullerene molecules investigated in this study.
reported (for alkylthiol- and anthrylphenylacetylenethiol-functionalized C60)11,12 show SAMs with very similar long-range geometry to the hcp arrays formed by unmodified C60. In this study, we synthesize a series of structurally related thioland thioether-derivatized fullerenes (Figure 2) and systematically investigate their surface assembly as a function of the chemical group. We vary the length, shape, and flexibility of the functional group attached to the fullerene cage to determine the optimum structural parameters for construction of ordered fullerene SAMs on Au(111). Our approach offers a very effective and versatile methodology for fullerene self-assembly, as we report formation of 2D square and parallelogram arrays of fullerenes, which are the first examples of highly organized fullerene SAMs with long-range order other than hexagonal.
’ RESULTS AND DISCUSSION Design, Preparation, and Characterization of the S-Functionalized Fullerenes. Fullerenes exhibit rich chemical reactiv-
ity in a variety of organic reactions, including 1,3-dipolar cycloaddition, one of the most reliable and high-yielding reactions involving fullerenes, which introduces a five-membered pyrrolidine ring attached directly to the carbon cage (Scheme 1). Recently, we have demonstrated that the grafting of a pyrrolidine group onto the fullerene cage does not perturb the magnetic13 or optical properties14 of endohedral fullerenes. As the 1,3-dipolar cycloaddition reaction proceeds via the condensation of R-amino
acids with aldehydes, this process can be used to attach a sulfurbased anchoring group to either the carbon atom or the nitrogen atom of the pyrrolidine ring, depending on whether the anchoring group initially forms part of the amino acid or aldehyde (Scheme 1). This gives control over the symmetry of the final functionalized fullerene. The structural diversity of potential aldehyde and amino acid precursors offers a wide choice of spacers of various lengths and different conformational rigidity. Both the symmetry and the flexibility of the functionalized fullerenes are expected to have a significant effect on the structure of the resultant 2D molecular arrays. To illustrate this methodology, we have developed a synthetic approach for the N-(4-methylthiophenyl)glycine precursor, 3SR-A (Scheme 2), that can be reacted with C60 in the presence of formaldehyde to yield a fullerene, 3-SR, possessing a conformationally rigid spacer unit with a mirror plane of symmetry. Alternatively, using N-(4-mercaptophenyl)glycine in the cycloaddition reaction, which can be prepared in a similar way, leads to the formation of a structurally related fullerene 1-SH with a thiol anchoring group instead of a thioether. The four hydrogen atoms of the pyrrolidine ring in 3-SR and 1-SH appear as a single peak in the 1H NMR spectra [δ = 5.17 ppm for 3-SR (Figure 3a) and δ = 5.07 ppm for 1-SH], thus confirming the symmetry of these molecules. However, if the same amino acid, N-(4-mercaptophenyl)glycine, is reacted with acetaldehyde instead of formaldehyde, the symmetry of the functionalized fullerene molecule (2-SH) is lowered by the introduction of a methyl group on one side of the pyrrolidine ring. This is demonstrated by the splitting of the single 1H NMR 10978
dx.doi.org/10.1021/la200654n |Langmuir 2011, 27, 10977–10985
Langmuir
ARTICLE
Figure 3. 1H NMR spectra of the pyrrolidine ring in 3-SR (a) and 2-SH (b).
Scheme 3. Synthesis of the Thioether and Thiol Aldehyde Precursors
signal into three peaks [two doublets at δ = 5.05 and 4.86 ppm and a quartet peak at δ = 5.55 ppm (Figure 3b)]. Functionalized fullerenes with an angular geometry can be synthesized if a thiol or thioether anchoring group is introduced as part of the aldehyde precursor in the 1,3-dipolar cycloaddition reaction (Scheme 1). The conformational flexibility of the spacer is expected to play an important role in the subsequent formation of molecular arrays and can be conveniently controlled by the length of the alkyl chain between the fullerene cage and the anchoring group. To illustrate this strategy, we prepared thiol and thioether fullerene derivatives 4-SR and 3-SH by reacting C60 and N-methylglycine with 2-(2-methylsulfanylethoxy)benzaldehyde and 2-(2-mercaptoethoxy)benzaldehyde, respectively. The aldehyde precursor 4SR-A was obtained by the nucleophilic substitution of 2-chloroethyl methyl sulfide with 2-hydroxybenzaldehyde (Scheme 3a). However, two steps were required to obtain the precursor 3SH-A, through the monosubstitution of 1,2-dibromoethane with 2-hydroxybenzaldehyde followed by the nucleophilic substitution of 3SH-A1 with hexamethyldisilathiane (Scheme 3b). It should be noted that 3SH-A can also be formed via an alternative route from 3SH-A1 by reaction with thiourea in the presence of sodium hydroxide; however, this results in a significantly lower yield (∼10%). Para-substituted isomers of the above aldehydes, 4SH-A and 5SR-A (Scheme 3c, d), were synthesized in a similar way from 4-hydroxybenzaldehyde. In an analogous fashion to the orthosubstituted analogues, they react smoothly with C60 in the presence of glycine to form the corresponding thiol and thioether fullerene derivatives (5-SR, 4-SH) with a more angular shape.
Figure 4. (a) The proposed reaction scheme for the formation of the thiazolidine ring. (b) DEPT 135 13C NMR of 1-SR (only the aliphatic carbon atom region is shown).
The short-chain aliphatic aldehyde 3-(methylthio)propionaldehyde reacts efficiently with C60 in the presence of glycine to form the conformationally more rigid, angular-shaped spacer 2-SR. However, an attempt to introduce a short aliphatic spacer bearing a thiol group led to an unexpected result. The reaction of L-cysteine and paraformaldehyde with C60 yielded the thioetherfunctionalized fullerene 1-SR instead of the expected thiol. A cyclic thiazolidine group appears to form in two successive steps prior to the cycloaddition (Figure 4a). The heterocyclic amino acid formed in situ reacts readily with another equivalent of formaldehyde and forms a pyrrolidine ring attached to the C60 cage, resulting in the formation of fullerene 1-SR. 10979
dx.doi.org/10.1021/la200654n |Langmuir 2011, 27, 10977–10985
Langmuir The presence of two fused five-membered rings in fullerene 1-SR is evident from the DEPT 135 13C NMR spectrum, which exhibits one tertiary and three secondary carbon atoms, confirming the cyclic structure of the spacer. As the length of the spacer is significantly reduced in this molecule, 1-SR adsorbed on the surface is expected to adopt fewer possible conformations than the alkyl spacer containing fullerenes 4-SR or 5-SR (during preparation of our manuscript we became aware of a recent study reporting cyclic thioether formation on fullerene: Chen, C.; Li, X.; Yang, S. New J. Chem. 2010, 34, 331 336.). The series of nine novel, structurally related fullerenes (Figure 2) bearing different spacer and anchoring groups were isolated in isomerically pure forms by column chromatography and fully characterized by NMR spectroscopy and mass spectrometry prior to deposition on Au(111) substrates.
ARTICLE
Molecular Monolayers of Functionalized Fullerenes. Characterization by STM. To evaluate the effect that different
anchoring and spacer groups have on the structure of the resultant fullerene monolayers, the synthesized fullerenes were deposited on a Au(111)/mica substrate by immersion in dilute toluene or 1,2,4-trichlorobenzene (TCB) solutions of the respective fullerene at room temperature. For all molecules studied, molecular monolayers formed spontaneously, irrespective of the type of anchoring group, i.e., thioether or thiol. The SAM structures observed are very different from the hcp array adopted by unfunctionalized C60 and exhibit a pronounced dependence on the details of the molecular structure with both the spacer and anchoring group having a strong influence. All 2D fullerene arrays formed in this study can be qualitatively divided into ordered (Figure 6) and disordered SAMs (Figure 5).
Figure 5. STM images of SAMs formed on Au(111)/mica by thiol- and thioether-functionalized fullerenes: (a) 1-SH (I = 12 pA, Utip = 800 mV), (b) 2-SH (I = 13 pA, Utip = 860 mV), (c) 3-SH (I = 6 pA, Utip = 800 mV), (d) 4-SR (I = 4 pA, Utip = 1 V), and (e) 4-SH (I = 6 pA, Utip = 800 mV). Scale bars are 20 nm. The insets of parts a and b show close-ups of the ordered islands with the unit cells indicated by white squares. Scale bars are 5 nm. 10980
dx.doi.org/10.1021/la200654n |Langmuir 2011, 27, 10977–10985
Langmuir
ARTICLE
Figure 6. Large-scale STM images of the SAMs formed on Au(111)/mica by thioether-functionalized fullerenes: (a) 1-SR (I = 42 pA, Utip = 300 mV), (b) 5-SR (I = 33 pA, Utip = 520 mV), (c) 3-SR (I = 7 pA, Utip = 1 V), and (d) 2-SR (I = 39 pA, Utip = 860 mV). Insets show high-magnification images (a c) and Fourier transforms reflecting the crystalline structure of the SAMs (a, d) with unit cells indicated by white boxes (a d). (e) Diagrammatic model of 2-SR with the fullerene moieties represented by large dark circles, the SAM unit cell depicted by a black rectangle, the square arrangement of four fullerene molecules highlighted with a dashed square, and the nonorthogonal lattice vectors shown in white with the arrows highlighting the directions associated with the diffraction spots in the Fourier transform image inset in part d. Scale bars of the large-scale and inset images are 20 and 5 nm, respectively.
It can be clearly seen that the tendency to order is crucially dependent on the chemical nature of the anchoring group, as all fullerene thiols form disordered structures and all fullerene thioethers (with the exception of 4-SR) form more ordered arrays. To explain this observation it is tempting to use a simple kinetic argument that the stronger bonding of thiols with gold, as compared to thioethers,15 hinders the surface mobility of thiolfunctionalized fullerenes, thus not allowing this type of molecule to form regular 2D arrays. However, this argument conflicts with the fact that some thiols (for example, 2-SH) form small ordered domains, and some thioethers (such as, 4-SR) form no ordered structures. Clearly, the nature of the anchoring group and the
surface mobility of functionalized fullerenes are not the only factors determining the structure of SAMs. The fullerene SAMs, under our experimental conditions, should be considered as thermodynamic products of the molecular assembly, whose structures are dictated by a fine balance of molecule substrate and molecule molecule interactions. If intermolecular forces (dipole dipole, van der Waals, chargetransfer interactions) are dominant, they will inevitably steer the SAM formation toward an ordered structure, in an analogous fashion to the assembly of molecules into a 3D ordered crystal lattice (e.g., crystallization). But the nature of the molecule substrate interactions is much more complex and thus is less 10981
dx.doi.org/10.1021/la200654n |Langmuir 2011, 27, 10977–10985
Langmuir
ARTICLE
Table 1. Structural Parameters of the SAMs Formed by Thioether Functionalized Fullerenes on Au(111)/Micaa fullerene
a (Å)
b (Å)
R (deg)
N
A (Å2)
comments on structure √ √ c(2 3 2 3)
C60
5
5
60
1
86
1-SR
11
10.4
90
1
114.4
2-SR
43.4
10
90
4
109
intrarow
10.1
3.5 aAu
interrow
12
4 aAu
primitive unit cell, different orientations √ c(15 2 3), aligned with Au
different orientations
3-SR rectangular
18.8
16.3
90
2
153
centred unit cell
oblique 5-SR
33.5 19
16.5 15.5
60 90
3 2
160 147
domains not aligned with [110] different orientations centred unit cell domains not aligned with [110]
a
Values have an error of typically 5% (a, b, and R are the crystal-lattice parameters; N is the number of molecules per unit cell; A is the area per molecule).
predictable. These interactions are strongly dependent on the energy landscape of the adsorption sites, as reflected by the anchoring group substrate corrugation potential, and the bonding geometry, which exerts a directing force on the molecule.16 Either a good structural match of the size/shape of the molecule with the geometry of the underlying substrate or other ways to accommodate mismatch3 are a prerequisite for formation of ordered SAMs under the influence of molecule substrate interactions. Any mismatch between the molecule and substrate lattices leads to stress and, consequently, limits the extent to which ordered domains can form. These thermodynamic considerations provide a satisfactory explanation of our experimental observations: the surface behavior of fullerenes with thioether anchoring groups is controlled by the intermolecular interactions ensuring formation of more ordered SAMs, while thiolfunctionalized fullerenes are mostly influenced by the more structurally demanding molecule substrate interactions, leading to the formation of less ordered structures. For example, the observed difference between molecules 3-SR (Figure 6c) and 2-SH/1-SH (Figure 5a, b) fits into the thermodynamic framework as 3-SR forms ordered structures over an extended range, while any order in the thiol analogues of this molecule are limited to small domains. This phenomenon is even more pronounced for thioether 5-SR (Figure 6b) and its thiol analogue 4-SH (Figure 5e). All ordered SAMs appear to consist of straight rows of thioether fullerene molecules (Figure 9), but the relative alignment and spacing of neighboring rows varies widely (Table 1) depending on the structure of the spacer moiety between the fullerene cage and the anchoring group. For example, fullerene 3-SR forms two types of ordered structures: one with adjacent rows of fullerenes being shifted by half the intermolecular distance along the lattice vector b and the inter-row distances alternating between two values, thus giving rise to a double row structure, and the other one exhibiting a grouping of four rows of molecules separated by dislocations (parallelogram structure) (Figure 6c). Similarly, groups of rows are formed by 2-SR (Figure 6d), where the number of rows grouped together varies, typically between two and three. For the double row structure, there is an interesting difference between 2-SR and 3-SR. While paired rows of 3-SR are staggered within each pair, for 2-SR this staggering occurs only between the pairs of rows. In contrast, in 5-SR SAMs (Figure 6b, inset) all rows are equidistant (the impression of pairing arises from an asymmetry in the STM
Figure 7. Schematic representation of the different modes of an adsorbed functionalized fullerene enabling either (a) C60 3 3 3 Au or (b) S Au interactions or, alternatively in the case of (c), both interactions simultaneously.
contrast), which is in contrast to the pairwise arrangement in the 3-SR structure. Such variety of SAM geometries for structurally related molecules cannot be explained by the interplay between intermolecular and interfacial forces derived from general concepts of simple adsorbate layers,17 and therefore, the shape and size of the spacer moiety must be taken into account. The spacer not only affects the intermolecular interactions, e.g., by adding an anisotropic component to the isotropic interactions of the fullerenes, but also may affect the molecule substrate interactions. The role of the spacer is clearly demonstrated by the comparison of the positional isomers 5-SR (Figure 6b), forming well-ordered extended domains, and 4-SR (Figure 5d), forming disordered assemblies. The substantial differences in SAMs of 1-SR, 2-SR, 3-SR, and 5-SR can also be attributed to the nature of the spacer. The surface packing density of fullerene molecules is a useful parameter for the comparison of the different packing arrangements, as it scales with the size of the spacer. For 1-SR (Figure 6a) and 2-SR (Figure 6d) the area per molecule increases by 30% and 20%, respectively, as compared with surface packing of unfunctionalised C60. This strongly suggests a canted orientation (i.e., “lying-down” mode, Figure 7c) of the molecules rather than an upright orientation (Figure 7b), as a canted orientation is the only arrangement in which a scaling of the molecular area with the size of the spacer is expected. This is also consistent with 10982
dx.doi.org/10.1021/la200654n |Langmuir 2011, 27, 10977–10985
Langmuir the 70 90% increase in surface area required for 5-SR (Figure 6b) and 3-SR (Figure 6c). Such a drastic increase is fully expected for the 5-SR structure, but less so for 3-SR, indicating that the conformational rigidity of the spacer plays as important a role as its physical size. Even though the geometry of fullerene SAMs can be clearly visualized by STM, detailed structural analysis can be hampered by the coexistence of different packing arrangements for the same type of functionalized fullerene (for example, 3-SR, Figure 6c). Another factor is that the SAM lattices, in general, do not match the lattice parameters along the main symmetry axes of the substrate, resulting in rotated domains for all but one of the SAMs (2-SR) in this study. This problem is highlighted for 5-SR (Figure 6b), where the arrows, which are parallel to the lattice vector a, indicate three different domain orientations. Angles of approximately 70°, 40°, and 15° measured for three different domains with the [110] direction show that the ordered domains are neither aligned parallel to the symmetry axes of the substrate nor exhibit a unique rotation with respect to the substrate lattice. A final factor complicating the accurate structural evaluation of SAMs is the distortion of the molecular lattices, as shown by the Fourier transform of 1-SR (Figure 6a, inset). The square (within the experimental error) unit cell results in diffraction spots that are smeared as a consequence of this lattice distortion, i.e., the positions of the individual molecules deviate slightly from the ideal lattice points, confirming that the SAM structures are defined by a complex interplay of different interactions. Thioether-functionalized fullerene 2-SR is the only molecule that forms sufficiently ordered arrays to enable a detailed model to be described. Without comprehensive consideration of the specific orientation of the spacer, which is discussed in detail in the next section, we are only able to display the fullerene moieties as circles (Figure 6e; considering that the exact adsorption sites of the molecules are unknown, the substrate lattice can be used only as a reference for the√dimensions). The 2D array of 2-SR is described by a c(15 2 3) unit cell (solid rectangle), which comprises the squared arrangement of the molecules within the double rows (dashed square) and the staggered alignment between them. The regular arrangement of shifted double rows gives rise to nonorthogonal lattice vectors, which is illustrated by the white lines and corresponding arrows indicating the directions associated with the diffraction spots in the Fourier transform image (Figure 6d, inset). The primed values in the Fourier transform denote symmetry-equivalent rotated domains. The alternation in the inter-row distances is consistent with the molecules being tilted and the functional group pointing toward the gap between the staggered rows, which would lead to a mirror symmetric arrangement of the molecule in the double row structure. While this description gives a good appreciation of the structural features of this SAM, only detailed theoretical calculations can explain the origin of any given molecular packing. Theoretical Modeling. Any correlation between the observed packing geometries of the ordered fullerene SAMs with the atomic structures of functional groups attached to the fullerene is significantly complicated by the existence of the multiple potential orientations each adsorbed molecule can adopt with respect to its nearest neighbors. As an illustrative example, we consider two possible packing arrangements for the thioether fullerene 2-SR on a Au(111) substrate (Figure 8). It is expected that a “lying-down” adsorption mode (Figure 8b) will be thermodynamically preferred over a “standing-up” mode
ARTICLE
Figure 8. “Standing-up” (a) and “lying-down” (b) modes of 2-SR (top and side views) and their schematic representations (fullerene cages are represented as circles and the functional groups as rectangles).
(Figure 8a). The stronger C60 3 3 3 Au interaction compared to the intermolecular C60 interactions in combination with the weaker interactions of thioethers compared to thiols should result in a higher energy per unit area. However, for the sake of completeness, both modes have been considered in our Monte Carlo molecular dynamics calculations to simulate the most efficient packing for the 2-SR molecule on a two-dimensional plane. The theoretical packing arrangements predicted for the lying-down mode (Figure 9a, b) match the rectangular pattern observed by STM well. The ethylmethylsulfane group ( CH2CH2 S CH3) attached to the side of the pyrrolidine ring and the ethyl group ( CH2CH3) attached to the N-atom of the pyrrolidine ring slot between the fullerene cages effectively increases the inter-row center-to-center separation between them beyond the 1 nm distance, which is typically observed in a hcp array of unmodified C60. As a result, van der Waals interactions between neighboring fullerene cages, which would be possible for the standing-up mode (Figure 9c), are inhibited in the rectangular 2D array of the lying-down mode. The energy gain of the C60 3 3 3 Au interaction (109 251 kJ/mol)18 present in the lying-down mode clearly outweighs the loss of any C60 3 3 3 C60 interactions (∼16 kJ/mol per pair),19 leading to the experimentally observed rectangular arrays of 2-SR. Detailed calculations for other molecules are more difficult to perform due to the higher complexity of their structures and the challenges posed by the STM characterization described in the previous section. The case of the thioether 2-SR, however, illustrates well the range of parameters that should be taken into consideration for a full structural characterization of fullerene SAMs. XPS Characterization. As expected from the similarity in their chemical composition, all SAMs give XPS spectra that essentially display the same features, which can be illustrated by considering the spectra of 3-SR (Figure 10). The S3/2 peak appears at around 163 ( 0.1 eV, which is characteristic for thioethers,15 indicating that the S C bonds remain intact and the thioether groups retain their chemical nature upon adsorption on the substrate. In order to fit the sulfur 3/2 and 1/2 doublet, the peak separation was fixed at 1.2 eV, values for fwhm were constrained to be identical, and the intensity ratio was fixed at 2:1. Fits for the other elements were carried out using a single peak (Figure 10b,c). As expected, the carbon and nitrogen peaks appear around 284.9 ( 0.1 eV and 399.5 ( 0.3 eV, respectively. While there is considerable uncertainty in the 10983
dx.doi.org/10.1021/la200654n |Langmuir 2011, 27, 10977–10985
Langmuir
ARTICLE
Figure 9. (a, b) Predicted 2D packing patterns for a lying-down mode of 2-SR superimposed on an experimental STM image of a SAM of 2-SR on Au(111)/mica demonstrating good correlation, and (c) the predicted 2D pattern of a standing-up mode of 2-SR is included for comparison (exhibiting much smaller separations between fullerene cages that are governed by purely isotropic van der Waals interactions). Fullerenes are shown as red circles in each case.
’ CONCLUSIONS Organization of fullerene molecules into ordered mesoscopic structures still poses a significant challenge. The anisotropy of intermolecular and molecule substrate interactions holds the key for the directed arrangement of fullerenes and, as demonstrated in our study, can be effectively introduced and controlled through chemical functionalization of fullerene cages with sulfur-containing groups: thiols and thioethers. We have synthesized a series of structurally related molecules, which allowed systematic investigation of the parameters defining the geometry of fullerene SAMs. The chemical nature of anchoring groups appears to be crucial for the long-range order in fullerenes: the assembly of thiol-functionalized fullerenes is governed by strong molecule surface interactions, which prohibit formation of ordered molecular arrays, while in the case of thioether-functionalized fullerenes, a variety of ordered 2D molecular arrays can be formed owing to noncovalent intermolecular interactions. The precise structure of SAMs is highly dependent on the shape and size of the spacer group linking the anchoring group with the fullerene cage. We have demonstrated that through the careful control of the chemical functionality on the carbon cages, it is possible to arrange fullerenes in at least four different packing arrangements, none of which have been observed before. Our study offers a new experimental strategy for the controlled arrangement of fullerenes on surfaces. ’ ASSOCIATED CONTENT
bS
Figure 10. XPS spectra of a SAM of 3-SR on Au(111)/mica in the (a) S2p, (b) C1s, and (c) N1s regions. Black solid lines are experimental data, and gray lines are simulated fits.
intensity of the S2p and N1s signals due to a poor signal/noise ratio, the differences in packing density observed with STM are mirrored by the intensity of the C1s signal for each SAM; e.g., 3-SR exhibits a lower signal than 1-SR.
Supporting Information. Additional experimental information of SAM formation of functionalized fullerenes on Au(111)/mica, STM, XPS, and model considerations; details on the synthesis, characterization, and the spectral data of all new compounds including N-(4-methylthiophenyl)glycine (3SR-A), the aldehyde precursors (3SH-A, 4SR-A, and 4SH-A), and the thioether and thiol functionalized fullerenes [the (4-mercaptophenylamino)acetic acid precursor was prepared according to the literature procedure20]. This information is available free of charge via Internet at http://pubs.acs.org/.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected] (M.B); andrei.khlobystov@ nottingham.ac.uk (A.N.K.). Fax: (+44) 01159513555. 10984
dx.doi.org/10.1021/la200654n |Langmuir 2011, 27, 10977–10985
Langmuir
ARTICLE
’ ACKNOWLEDGMENT This work was supported by the Engineering and Physical Sciences Research Council (EPSRC grant EP/D048761/01), the European Science Foundation (ESF), and the Royal Society. ’ REFERENCES (1) Benjamin, S. C.; Ardavan, A.; Briggs, G. A. D.; Britz, D. A.; Gunlycke, D.; Jefferson, J.; Jones, M. A. G.; Leigh, D. F.; Lovett, B. W.; Khlobystov, A. N.; Lyon, S. A.; Morton, J. J. L.; Porfyrakis, K.; Sambrook, M. R.; Tyryshkin, A. M. J. Phys.: Condens. Matter 2006, 18, S867–S883. (2) Bonifazi, D.; Enger, O.; Diederich, F. Chem. Soc. Rev. 2007, 36, 390–414. (3) Cyganik, P.; Buck, M.; Wilton-Ely, J. D. E. T.; W€oll, C. J. Phys. Chem. B 2005, 109, 10902–10908. (4) Viana, A. S.; Leupold, S.; Eberle, C.; Shokati, T.; Montforts, F.-P.; Abrantes, L. M. Surf. Sci. 2007, 601, 5062–5068. (5) Patnaik, A.; Okudaira, K. K.; Kera, S.; Setoyama, H.; Mase, K.; Ueno, N. J. Chem. Phys. 2005, 122, 154703–1/9. (6) Lee, S.; Shon, Y.-S.; Lee, T. R.; Perry, S. Thin Solid Films 2000, 358, 152–158. (7) Gu, T.; Whitesell, J. K.; Fox, M. A. J. Org. Chem. 2004, 69, 4075–4080. (8) Shon, Y.-S.; Kelly, K. F.; Halas, N. J.; Lee, T. R. Langmuir 1999, 15, 5329–5332. (9) Kelly, K. F.; Shon, Y.-S.; Lee, T. R.; Halas, N. J. Phys. Chem. B 1999, 103, 8639–8642. (10) Shirai, Y.; Cheng, L.; Chen, B.; Tour, J. M. J. Am. Chem. Soc. 2006, 128, 13479–13489. (11) Shi, X.; Caldwell, W. B.; Chen, K.; Mirkin, C. A. J. Am. Chem. Soc. 1994, 116, 11598–11599. (12) Kang, S. H.; Ma, H.; Kang, M. S.; Kim, K. S.; Jen, A. K. Y.; Zareie, M. H.; Sarikaya, M. Angew. Chem., Int. Ed. 2004, 43, 1512–1516. (13) Jinying Zhang, J.; Morton, J. J. L.; Sambrook, M. R.; Porfyrakis, K.; Ardavan, A.; Briggs, G. A. D. Chem. Phys. Lett. 2006, 432, 523–527. (14) Gimenez-Lopez, M. C.; Gardener, J. A.; Shaw, A. Q.; IwasiewiczWabnig, A.; Porfyrakis, K.; Balmer, C.; Dantelle, G.; Hadjipanayi, M.; Crossley, A.; Champness, N. R.; Castell, M. R.; Briggs, G. A. D.; Khlobystov, A. N. Phys. Chem. Chem. Phys. 2010, 12, 123–131. (15) Weidner, T.; Ballav, N.; Siemeling, U.; Troegel, D.; Walter, T.; Tacke, R.; Castner, D. G.; Zharnikov, M. J. Phys. Chem. C 2009, 113, 19609–19617. (16) Cyganik, P.; Buck, M.; Azzam, W.; W€oll, C. J. Phys. Chem. B 2004, 108, 4989–4996. (17) Braun, O. M.; Kivshan, Y. S. The Frenkel Kontorova Model. Concepts, Methods and Applications: Springer Verlag: Berlin, 2004. (18) Altman, E. I.; Colton, R. J. Surf. Sci. 1993, 295, 13–33. (19) Lu, J. P.; Li, X.-P.; Martin, R. M. Phys. Rev. Lett. 1992, 68, 1551–1554. (20) Gough, J. D.; Gargano, J. M.; Donofrio, A. E.; Lees, W. J. Biochemistry 2003, 42, 11787–11797.
10985
dx.doi.org/10.1021/la200654n |Langmuir 2011, 27, 10977–10985
