Diffraction from physisorbed layers

Current Opinion in Colloid & Interface Science 17 (2012) 23–32

Contents lists available at SciVerse ScienceDirect

Current Opinion in Colloid & Interface Science journal homepage: www.elsevier.com/locate/cocis

Diffraction from physisorbed layers Thomas Arnold a,⁎, Stuart M. Clarke b a b

Diamond Light Source, Harwell Science & Innovation Campus, Chilton, Didcot, OX11 0DE, United Kingdom BP Institute and Department of Chemistry, University of Cambridge, Madingley Road, Cambridge, CB3 0ZE, United Kingdom

a r t i c l e

i n f o

Article history: Received 14 October 2011 Received in revised form 16 November 2011 Accepted 21 November 2011 Available online 26 November 2011 Keywords: Physisorption Monolayer X-ray and neutron diffraction Scanning tunnelling microscopy LEED GIXD Atomic force microscopy Hydrocarbon Alkane

a b s t r a c t Diffraction techniques used to study the structures of atoms and molecules physically adsorbed onto a variety of solid surfaces are reviewed. This is part of an important topic that includes the thermodynamics, dynamics and simulations/calculations of physisorption. We identify that there has been an interesting recent expansion in the variety of surfaces and molecular adsorbates that have been studied, extending previous monocomponent studies to studies of binary mixtures with particular focus on novel intermolecular interactions. Technically, improvements in access to large centralised facilities with appropriate diffraction instrumentation have the potential to allow previously unfeasible measurements to be made and the beginnings of work in this area is summarised. There have also been significant advances in related techniques, such as scanning tunnelling microscopy and atomic force microscopy, that provide important and complementary structural information. © 2011 Elsevier Ltd. All rights reserved.

1. Introduction The physical adsorption of species at solid interfaces is of fundamental importance to a very wide range of important phenomena including oil recovery, detergency and floatation. The presence of the adsorbed material, although representing a very small amount of the total material present, can dominate the behaviour. For example, by changing the nature of the surface from water to oil wet. Physisorption, where the enthalpy of adsorption is rather small and similar to the enthalpy of vaporisation of the adsorbate, can occur from both the vapour and condensed phases. The study of both of these interfaces represent particular experimental challenges, however, it is probably fair to say that the study of adsorption from liquids represents a particular challenge, given that the adsorbed layer is more inaccessible, encompassed by condensed phases on both sides. However, in recent years a combination of theory, simulation and experimental techniques have been developed to probe the structure, dynamics, thermodynamics and electronic properties of physisorbed layers under a variety of conditions. This combination of approaches now gives significantly more insight into these adsorbed layers and their behaviour, both as pure materials and as mixtures.

⁎ Corresponding author. Tel.: + 44 1235 778543. E-mail address: [email protected] (T. Arnold). 1359-0294/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.cocis.2011.11.003

In this short review we must limit our focus to developments over the last 5–10 years particularly in the use of diffraction techniques to to study the structures of physically adsorbed pure materials and mixtures from the vapour and liquids onto solid substrates. It is important though, not to forget that structural studies are usually done to facilitate a better understanding of physical phenomena. As such this review is just a subset of the studies of physical adsorption and as such the reader may also wish to consult a number of reviews on related material including those by Bruch et al. [••1], Clarke [2] and Inaba [3]. There have also been several reviews of Scanning Tunnelling Microscopy (STM) of organic layers [4,5] and there continues to be significant effort in this area. We have not included chemisorbed systems or other important topics such as self-assembled monolayers (SAM's) on metal surfaces [6,7]. Initially we will review the principle advances in the techniques relevant to this area. We will then consider the different experimental systems that have been addressed in the recent literature. We will consider each adsorbate on different substrates as this provides a convenient grouping to present and compare the topics. 2. Experimental approaches There are now a fairly large number of experimental methods for the study of physisorbed layers. It is convenient to consider them in two broad classes. The first are surface specific techniques that provide detailed, atomic resolution, information about the adsorbed species and its position relative to a well characterised (often) single

24

T. Arnold, S.M. Clarke / Current Opinion in Colloid & Interface Science 17 (2012) 23–32

crystal substrate. The second is to use high-specific-surface-area powdered substrates, which mean that the very small contribution from a monolayer is large enough to be detected by experimental techniques that are not surface-specific. In the first category, although beam-damage has historically been a problem, Low Energy Electron Diffraction (LEED) has seen some recent developments which improve the measurement of beam-sensitive layers. For example fibre-optic LEED [8] and delay-line-detector LEED [9]. These techniques both allow high sensitivity diffraction patterns with a low-flux electron beam. This is particularly important for physisorbed layers since absorption of high-energy electrons can result in desorption of the monolayer. Helium atom scattering (HAS) is a rather unusual and potentially powerful experimental approach that can both provide structural and dynamic information on a monolayer simultaneously. There are some recent studies of dynamics of physisorbed layers (e.g. benzene on graphite using both helium and neutron spin-echo spectroscopy [10]), but to our knowledge helium atom diffraction has not recently seen much use in the area of physisorbed layers. Synchrotron grazing incidence X-ray diffraction (GIXD) is increasingly attractive, partly with the availability of new third generation facilities but also for the rich in-plane structural information that can be obtained. This technique was developed in ultra-high vacuum (UHV) but has increasingly been used to study non-UHV systems. Using this, the related methods of X-ray reflectivity and grazing incidence small angle X-ray scattering (GISAXS), it is possible to obtain detailed structures of thin films adsorbed on a range of surfaces from solid– solid to liquid interfaces over a range of length scales, including polymers [11] and lipids [12]. Technical advances in this area should mean that detailed structural and dynamic studies of these types of complex system will become increasingly viable in the future. A particularly interesting example in which GIXD has been used is Xenon physisorbed on Ag(111) [••13]. In this study scans of the Xe(10) Bragg rod were able to confirm the structure of a Xenon monolayer with higher resolution than LEED. Further, the out-of plane structure during film growth was also investigated: Scans along the Xe (0 1 l) rod were able to show the stacking sequence of the film, and the (0 0 3/2) anti-Bragg position were able to show layer-by-layer film growth. The specular reflectivity scans were able to precisely measure the film thickness during growth. This study clearly demonstrates the level of detailed structural information that is potentially available using this method, and the technique has since been applied to some other systems. However, similar studies have been rare, perhaps due to the difficulties associated with this approach. A small UHV chamber was required with in-situ sputtering, LEED and Auger capability together with sample cooling to 30 K. In general, near-UHV conditions must be considered with some care, as they can lead to desorption of physisorbed species. A common way to address this issue was used in this study; the samples were prepared using a dynamic approach in which the surface is continually dosed to compensate for losses due to continual pumping. This results in a “quasi-equilibrium”, in which the local conditions at the sample are nominally equivalent to the presence of a 3D gas. Alternatively, experiments can be performed at temperatures for which the vapour pressure is very small so that desorption is negligible. However, both of these approaches mean that there is some question over thermodynamic equilibrium. Some other new techniques for studies of surfaces in near-UHV conditions have been developed recently. These include Very High Resolution Ellipsometry (VHRE) [••14] which can quickly and accurately measure the thickness of an adsorbed film and provides a useful addition to the complementary techniques of X-ray and neutron reflectometry. Additionally Environmental Scanning Electron Microscopy (ESEM) has also been used with physisorbed systems [15]. Although the electron gun and most of the beam path are in UHV, the sample can be maintained at a modest vapour pressure preventing desorption.

Measurements with powdered substrates can be made under equilibrium conditions in the presence liquid or vapour, and therefore offers some advantages. When collecting powder diffraction from adsorbed monolayers it is preferable if one uses powders that have just one crystal-facet available for adsorption, historically Graphite and Boron Nitride (which are ‘lamella’ in nature with an exposed basal plane), and 3D crystals such as Magnesium Oxide which don't show significant faceting. Careful subtraction of patterns obtained from samples with and without adsorbate is used to obtain the scattering pattern of monolayers with peaks which have a distinctive “saw-tooth” line shape that is characteristic of a two-dimensional lattice, which can be used to confirm that an adsorbate is present as a monolayer and not as 3D crystallites. Fig. 1 shows examples of recent synchrotron and neutron measurements that illustrate the high quality of data that can be obtained. These diffraction patterns can be solved using analogous approaches to 3D. However, in most cases the diffraction patterns do not have many peaks and so structural determinations are limited in their complexity and can only be fitted by highly constrained models. Confidence in structural assignments can be improved with additional information in the form of combined X-ray and neutron measurements or by comparing the structures of homologous series of adsorbates where there is a systematic variation of alky chain length (for example). This approach does not directly measure registry with the substrate, but a comparison of the unit cell parameters can be used to infer whether a monolayer structure is commensurate. It is important to note that these techniques are best used in combination with other supporting techniques. As well as thermodynamic measurements (calorimetry and adsorption isotherms in particular) other techniques such as inelastic and quasi-elastic neutron or helium

Fig. 1. Two examples of high quality diffraction data due to improved X-ray and neutron instrumentation. Top) Butane on MgO measured at OSIRIS at the ISIS neutron source. Reprinted from [••64] Copyright 2006, American Physical Society. Bottom) A halogen bonded binary mixture on graphite measured on I11 at Diamond Light Source. From [•84] — reproduced by permission of The Royal Society of Chemistry.

T. Arnold, S.M. Clarke / Current Opinion in Colloid & Interface Science 17 (2012) 23–32

atom scattering which probe the dynamics of adsorbed films have been widely used. These experimental studies are greatly enhanced in combination with theoretical methods. For example, molecular dynamics calculations have historically proved very useful providing proposed mechanisms for monolayer melting. Conversely, traditional DFT approaches have had difficulty in accounting for the van der Waals interactions that dominate physisorption. However, recently there are reports where the use of an additional correction for dispersion forces has meant that the use of DFT for physisorbed molecules has become increasingly viable [16]. In recent years the use of scanning probe techniques has become increasingly prevalent. Atomic force microscopy (AFM) and scanning tunnelling microscopy (STM) are direct, real-space, probes of the surface structure complementary to the ‘reciprocal space’ diffraction techniques outlined above. They are both surface specific techniques that can give atomic level structural information and can be used in ambient conditions and/or in the presence of bulk liquid. AFM measures the forces involved as a tip approaches a surface and the perturbation due to the presence of any adsorbed layer. STM measures the tiny electrical current that passes between a surface and a tip when a voltage difference is applied and the two are not quite in contact so that the electrons tunnel between them. STM images are a subtle combination of tip, surface and monolayer wavefunctions and some care is required in their interpretation. In addition, many samples require preparation from solution (typically phenyloctane), which can complicate matters. It can be challenging to study adsorbed layers over a wide range of temperatures and STM requires a conducting surface. Meanwhile, the quality of AFM data has improved remarkably, to the extent that it is now possible to obtain atomic resolution of molecules adsorbed on surfaces using both STM (e.g. [17]) and AFM [•18,19,20] (see Fig. 2) and these provide important tools for the study of physisorbed systems.

25

Symmetry considerations in determining two-dimensional closepacked structures has been long established [21,22]. A recent review [•23] of structures obtained by STM, seems to indicate that the plane groups adopted by physisorbed layers of a very wide range of molecular adsorbates is indeed governed by the molecular symmetry and the requirement for close packing in two-dimensions. Importantly exceptions to these rules have now been identified, with ‘honeycomb’ networks observed, in which ‘gaps’ within the monolayer exist that reveal the ‘bare’ graphite. These only occur when there is a strong intermolecular driving force such as an extensive hydrogen bonding network (e.g. Triacid benzene [24]), or due to strong steric effects such as inter-digitation [25]. However, even in these cases, solvent molecules or other species will try and fill the ‘gaps’ where possible. 3. New substrates Well characterised substrates have generally been important whichever technique is used. Interestingly there has recently been a significant expansion in the types of substrate that have been studied with diffraction (e.g. mesoporous materials [26,27]) as well as more extensive exploration of existing materials. Carbon nanotubes and related substrates in particular have received considerable attention [28,29]. There are other recent reviews of this subject since these materials are of interest for the variety of intermolecular interactions and geometries available and the potential applications that they imply. Interestingly nanotubes have actually been used directly to study physisorption, in the development of ‘nano-mechanical resonators’ based on single- or double-walled nanotubes [30,31,•32]. In a way analogous to quartz-crystal-microbalances, these devices use measurable changes in the nanotube resonant frequency to give extreme mass sensitivity to adsorption. In addition, measurements of the electrical properties of the nanotube can be made simultaneously. Gas adsorption and phase changes within an adsorbed film change these properties and so can be measured with high precision. For example, coverage dependent commensurate and incommensurate solids can be identified from Ar or Kr adsorption onto a nanotube resonator [•32]. One class of non-traditional substrates of interest are quasicrystalline surfaces. These are an interesting case in that they can be well characterised but do not show long range ordering. As such the close packing of adsorbed layers (generally six-fold symmetry) can be frustrated by the unusual symmetry of the underlying substrate (e.g. five-fold symmetry in the case of AlNiCo [•33]). In effect, the level of ordering of monolayers adsorbed on these substrates is governed by the relative length scales of the substrate ordering compared to the size adsorbate molecules. This could have some very interesting consequences for monolayers of alkanes or other hydrocarbon adsorbates. 4. Studies of adsorption of molecular species Over 30–40 years there have been a large number of studies of small molecules and noble gases physisorbed on graphite while the literature on other substrates is much more modest (for details see [••1] and other reviews referenced therein). Here we concentrate on two particular systems of interest that have seen some recent developments. 4.1. Hydrogen and methane on MgO

Fig. 2. Examples of atomic resolution STM and AFM. Top) STM of bromoalkanes. Reprinted with permission from [90]. Copyright 2009 American Chemical Society. Bottom) AFM of naphthalocyanine molecule adsorbed on NaCl(2ML)/Cu(111). Image reproduced by permission of IBM Research — Zurich. Unauthorised use not permitted. Reprinted from [19] Copyright 2011 American Institute of Physics.

Two recent studies of hydrogen and methane on MgO demonstrate the power of using complementary techniques, including diffraction, thermodynamics, inelastic neutron scattering measurements and theoretical results. Both systems show extensive layered isotherms [•34] which indicate complete wetting of the surface. Diffraction data is reported to show that, at least in part, this is due to the coincidental matching of the inter-molecular spacing with the lattice parameters of

26

T. Arnold, S.M. Clarke / Current Opinion in Colloid & Interface Science 17 (2012) 23–32

the MgO (100) surface. This means that there is no surface induced strain between a growing film as it tends towards the bulk. Further, inelastic neutron scattering data in combination with theoretical calculations has been used to obtain an extremely detailed understanding of the surface potential and its interaction with the adsorbed molecules [•34,35–37]. In the case of methane the molecules are adsorbed over the Mg2 + ions in a “dipod” configuration, (i.e. with two rather than three hydrogen atoms nearer the surface). Recent DFT results [35,36] confirmed this arrangement and additionally suggested that adjacent molecules are oriented with their C2v axes rotated by 90° with respect to each other. For hydrogen [•34] the adsorption site is also above the Mg 2 + ions, and the interaction with the surface effectively distorts or flattens the molecular rotor so that the measured ortho-para transition shifts markedly from that seen for a free molecule or one adsorbed on other substrates such as graphite. 4.2. Acetylene Recently the structures of Acetylene (Ethyne) on the surface of NaCl and KCl have been determined by LEED, HAS, PIRS (polarisation infrared spectroscopy, a technique that is sensitive to the orientation of molecules on a surface) [•38,39]. These structures make for an interesting comparison to older data and calculations for acetylene on graphite [40,41] and MgO [42,43]. Subtle variations between structures on different substrates of acetylene are important indicators of the role played by the surface potentials and cell parameters. Acetylene is reported to form commensurate layers with the molecules found to lie with their axis parallel to the surface and adopt a commensurate “herringbone” or “square-tee” structure, similar to (001) face of the orthorhombic bulk phase. The distance between (001) planes of the low temperature phase of bulk acetylene is similar to cell parameters on each surface (see Fig. 3). The intermolecular interactions are dominated by the strong quadrupole moment of the acetylene molecule which favours the tee-shaped arrangement of the molecules and a nearest neighbour intermolecular separation similar to that of the bulk. The next most significant interaction is that with the substrate. On the ionic materials, the molecules adsorb above the cations and the overall (commensurate) structure is thus determined by the lattice parameters of the substrate. Compression of the surface cell parameters results in a perturbation to the preferred tee-shaped intermolecular interaction. On graphite the substrate potential is weaker and so the preference for a commensurate structure is weaker. This weaker interaction may in part explain the possible coverage dependent phase behaviour on this substrate.

understanding of many types of physisorbed alkyl hydrocarbons and is discussed here in some detail. The n-alkanes have been studied most of all organic molecules and studies on graphite are the most widespread. 4.3.1. n-alkane structures on graphite Diffraction measurements show a distinct odd even effect in the monolayer structures of alkanes adsorbed on graphite. Experimental measurements (for ethane and butane [44]) or potential energy calculations (for propane [45]) report that the very short chain alkanes adopt herringbone structures. Pentane, however, was reported to have an anomalously unstable monolayer with a centred rectangular rather than the herringbone structure [••14,46,47]. Thereafter all of the odd alkanes adopt this symmetry structure, while the even alkanes continue to form herringbone structures up until around tetradecane [17,46,48]. Interestingly dodecane is seen to undergo a monolayer phase transition prior to melting that may be related to a transition between a herringbone and a parallel structure. For all longer alkanes STM results show centred (parallel) structures. This balance between structures is also seen for other short functionalised alkyl-species in which the interactions at the ends of the molecules are relatively large compared to the interactions between the alkyl chains. There continues to be some debate over the orientation of the plane of the zig–zag of the carbon backbone with respect to the graphite surface. Diffraction measurements have been used to suggest that the plane of the carbons is close to parallel to the graphite surface, or in some cases alternates between parallel and perpendicular for adjacent molecules within the lamellae [49]. However, diffraction is not very sensitive to this particular aspect of the structure. Recently very high resolution STM images [50] showed tetradecane orientated with this plane perpendicular to the substrate. In this case the STM image is interpreted as a parallel structure in which the principal molecular axis of the molecule is tilted by 60° relative to the lamellar axis. Such a structure is inconsistent with the herringbone structure determined by diffraction [48]. In the same study, hexadecane is seen to adopt a centred structure that is more reminiscent of the odd-alkane structures mentioned earlier and other longer alkanes STM images seen elsewhere. In this case the molecular axis is 90° to the lamellar direction while the zig–zag plane is parallel to the surface. These results clearly show that there is a very delicate balance between the demands of molecular packing and the interaction with the substrate that changes as the molecules get longer. Somewhere between dodecane and hexadecane there seems to be a transition between favouring the herringbone and favouring the parallel structure.

4.3. Normal-alkanes The alkyl chain chemical group is predominantly governed by van der Waals forces and is an important reference system for our

Fig. 3. Schematic representation (approximately to scale and derived from results in [•38–40,42,43]) of the similar “herringbone” or “square-tee” structures of acetylene on MgO (2 × 2), KCl (√2 × √2)R45, NaCl (3√2 × √2)R45 (note for NaCl the full unit cell proposed has 5 molecules per unit cell) and Phase I (square-tee structure) of acetylene on graphite. For reference, the substrate unit cell parameters are 4.212 Å for MgO, 5.64 Å for NaCl and 6.29 Å for KCl, while for graphite the C–C distance is 2.46 Å.

4.3.2. n-alkane structures on other substrates Comparison of the structures adopted by n-alkane species on different surfaces is interesting as it gives insight into the balance between molecule-substrate and intermolecular interactions. Here we limit ourselves to the physisorption regime (a wider review of the interaction of low molecular weight alkanes on metal surfaces is available [51]). Adsorption energies of alkanes on metals have been reported to vary with chain length (approximately 6–7 kJ mol − 1 per CH2 group [••1,51]). For a given alkane (e.g. hexane) heats of adsorption are fairly similar (~ 68.5 kJ mol − 1 on graphite [••1], 36.2 kJ mol − 1 on MgO [52], either 50.4 kJ mol − 1 [51] or 65.7 kJ mol − 1 [53] on Cu(111), 61.7 kJ mol − 1 on Pt(111) [51] and 62.7 kJ mol − 1 on Au(111) [51]). In each case, the following systems can therefore be considered as physisorption. The hexagonal Pt(111) crystal face (spacing Pt–Pt of 2.77 Å compared to C–C in graphite of 2.46 Å) adsorbs alkanes (butane–octane) with the molecules parallel to each other (not herringbone like on graphite) [54]. In general the structures are commensurate but defined by close-packing of the alkanes rather than cell parameters of the surface. Butane has several phases with the molecules perhaps

T. Arnold, S.M. Clarke / Current Opinion in Colloid & Interface Science 17 (2012) 23–32

tilting away from the surface at high coverage (from LEED and TPD [55]). The structures of alkanes (pentane–octane) on Ag(111) (also hexagonal symmetry with a spacing of Ag–Ag of 2.89 Å) were also determined by LEED measurements [56] and report structures are very similar to those seen on Pt(111). However, recently the structures of butane, hexane and heptane were revisited using X-ray diffraction [57] and some alternative structures have been proposed. The butane structure, for example, was proposed to adopt a herringbone structure similar to that proposed on graphite [44] but with different cell parameters. Interestingly Helium atom diffraction of tetracosane (C24H50) on a hydrogen terminated Pt(111) surface has indicated a herringbone structure [58]. This molecule is much longer than the longest alkane for which the herringbone structure is seen on graphite The structures for octane and nonane adsorbed on Cu(111) determined by LEED and HAS [59], show that octane adopts a herringbone structure, while nonane a centred parallel structure. If this is the general case, then it may indicate that the structure is importantly influenced by the surface cell parameters, since Copper is significantly smaller (Cu–Cu spacing of 2.55 Å in the (111) surface), than Pt and Ag, and much closer to that of graphite on which similar structures are seen. On Au(111), STM and theoretical studies [60,61] show the longalkane monolayers have a preference for the parallel type structure, but a herringbone structure was seen for decane [•62]. As on other metals, the structures are generally close-packed and consequently incommensurate and this is particularly evident in Moiré patterns in the STM images. On MoS2 and MoSe2 [63], surfaces are similar to graphite but with significantly larger equivalent spacings of 3.16 Å and 3.29 Å respectively. In both cases STM images of dotriacontane (C32) adsorbed from phenyloctane show incommensurate layers with both herringbone and parallel lamellae observed with the molecular axis about 30° to the lamellar direction. The details of the molecular packing vary slightly between each substrate but are approximately 10% less dense than on graphite. In summary we conclude that the structures adopted by alkanes on these hexagonal surfaces depend on the relative strength of surface-adsorbate and adsorbate–adsorbate interactions. Close packing within the adsorbed monolayer can lead to incommensurate structures. For surfaces with small atomic spacings herringbone structures for the shorter alkanes are observed. As the molecule length (or perhaps the surface atomic spacing) increases, the parallel structures begin to dominate. It is not clear why a herringbone structure is seen for dotriacontane on MoS2 and MoSe2 but given that both structure types were observed, it is possible one of these is metastable. MgO(100) provides an interesting contrast to the surfaces so far discussed, since it is ionic and has a square surface symmetry. Neutron diffraction data (shown in Fig. 1) indicate that despite these differences, butane still adopts a commensurate herringbone structure [••64] that is very similar to the structures of even alkanes on graphite. Structures of pentane and hexane have also recently measured and show a similar odd–even effect as observed on graphite. The particular reasons for this intriguing result are not clear but future studies of alkanes on other ionic surfaces may be helpful in this area. 4.3.3. Multilayer film growth The way in which a monolayer of physisorbed molecules grows from 2D to 3D is of fundamental interest and closely related to the physics of wetting. A detailed discussion of wetting phenomena is beyond the scope of this review (see other reviews on this topic [65]) but in short, these effects depend on the balance between the relative strength of interactions of an adsorbed molecule with the surface compared to intermolecular interactions in a bulk liquid or crystal. A range of coverage and temperature dependent transitions is

27

possible but for many systems a good understanding of these is yet to be obtained. Here we report some interesting recent studies of growth of hydrocarbon films on different substrates. Pentane on graphite [••14] shows up to seven adsorbed layers at low temperature. While at room temperature, depth sensitive AFM of a fluid hexadecane film on graphite [66] shows at least four oscillations separated by about 4 Å, confirming that these layers grow with their molecular axis parallel to the substrate. Unlike methane and hydrogen that show extensive layering (as mentioned earlier), the other alkanes on MgO only show 2–3 layers [52]. The commensurate herringbone structure of butane [••64] is considerably different from any face of its 3D crystal, which may suggest that there is some frustration between the two structures as the film grows. Of course this also depends on whether the film is solid or fluid, and in each case isotherms (at temperatures where the multilayers are fluid) do not show dewetting behaviour. This is not the case for all hydrocarbons, ethylene for example. At low temperatures this molecule does not wet either MgO or graphite above 1 monolayer, but as the temperature is increased towards the bulk critical point, a wetting transition is seen with up to 3 layers observed [•67]. On Ag(111) [57] alkanes have been reported to grow up to three layers with their long axis parallel to the substrate, before the formation of preferentially oriented bulk crystallites. This transition at about 2–3 layers maybe due to the gradual shielding of the surface potential so that beyond this the intermolecular interactions in a bulk-like environment begin to dominate. Similar behaviour has also been seen on SiO2. In this case, after the first two complete layers of molecules form with their axes parallel to the surface, preferentially oriented particles are reported to form in coexistence with incomplete so-called “standing-up layers”. These layers are reproducibly formed, have been observed using a number of different techniques [68–72] and consist of molecules with their principal axes perpendicular to the surface. This phenomenon is clearly intricately related to the wetting (thermodynamics and kinetics) of the film and the layers may be metastable since when such samples are thermally annealed, the coexisting nanoparticles grow while the perpendicular layers reduce in size [73]. In fact they are only really stable over a narrow temperature range near the alkane melting point. Above the melting point bulk liquid droplets are seen, while significantly below bulk crystallites form. 4.3.4. Melting The melting of physisorbed systems is of interest both from a theoretical point of view and as a guide to the relative stability of the systems. Experimentally monitoring the evolution of diffraction peaks as a function of temperature [49,52,74] can provide information about the melting process. It is important to note that the exclusive use of diffraction to determine the order of the melting transition is not advisable. These measurements measure the loss of long-range coherent ordering and do not wholly account for the onset of molecular motion during melting. Consequently a combination of additional techniques are often used, including calorimetry [75–77], quasielastic neutron scattering [52,78], NMR [79], STM [•62], AFM [71] and Interference Enhanced Reflection Light Microscopy [80]. Each of the techniques have their advantages and limitations, but together can provide valuable evidence of the mechanisms for melting transitions in two-dimensions. These mechanisms have been of considerable interest for many years, but the details of these and the related diffusion properties of physisorbed molecules is beyond the scope of this review. It is important to point out that these systems are not truly two dimensional, and as such the melting process can include the third dimension [81,82]. Thus, whether melting is a discontinuous or continuous transition is intimately tied to the vacancy creation process. In this quasi 2D situation three vacancy creation mechanisms have been proposed: edge melting, promotion of molecules to the second layer and/or the reduction of the projected area of the

28

T. Arnold, S.M. Clarke / Current Opinion in Colloid & Interface Science 17 (2012) 23–32

molecule within the monolayer (either by tilting away from the surface or by some conformational change). Each of these processes creates a vacancy within the monolayer solid and thereby allows for the onset of translational diffusion. Many years of research has established a rule of thumb that a monolayer coverage solid film in thermodynamic equilibrium with its surroundings will melt at a temperature of approximately 70–80% of its bulk phase melting point. This rule has proved reasonably robust for systems that are dominated by van der Waals interactions. Recently it has been shown that for some alkyl-amide monolayers adsorbed onto graphite in the sub-monolayer regime actually melt at temperatures close to or even above the bulk melting point [83]. This unusual behaviour may arise because of the extensive hydrogen bonding possible within the monolayer solid which considerably stabilises these layers. Halogen bonding has also resulted in this kind of stabilisation [•84]. At higher coverage, on graphite, hydrocarbon monolayers have been found to melt at temperatures between 5 and 20% higher than the bulk melting point. The degree of stabilisation here seems indicative of the relative stability of the monolayer and is influenced by both the intermolecular and molecule-substrate interactions. Thus monolayers that have a weak interaction with graphite and each other such as fluoroalkanes melt only 3–5% above the bulk [85], while the alkyl-amides, due to the extensive hydrogen-bonding network mentioned above, show exceptional stabilisation, melting up to 20% above the bulk [77]. In some cases hydrogen bonding may even extend out of the layer and influence the melting of the second and even third layers as reported for the layer by layer melting/freezing of alcohols on graphite [86]. Unlike graphite, MgO shows no similar stabilisation [52], with alkane films melting at or close to the bulk melting point. On gold there is some evidence from an STM study [•62] of stabilisation above the bulk melting point, but the extent of this is much more chain-length dependent than on graphite. It was proposed that this was due to the mismatch in the Au–Au period and that of the carbon backbone that is increasingly important for longer molecules. In fact this may directly lead to a two-step melting mechanism with first melting uniaxially along their molecular axis. 4.4. Functionalised hydrocarbons In recent years there has been a considerable advance in obtaining structural information from functionalised alkyl and unsaturated molecular systems adsorbed principally on graphite. 4.4.1. Halogenated-alkanes Although superficially similar to alkanes, the perfluoroalkanes (CnFn + 2) have a number of different properties which make them an interesting comparison. Because the fluorine atoms are somewhat bigger than hydrogen atoms, the packing in a molecule leads to a helical molecular structure, which has the potential to lead to chiral adsorbed layers. They also have weak van der Waals interactions that result in a relatively weak physical adsorption. As with alkanes, perfluoroalkanes (C6F8–C16F18) form solid monolayers when adsorbed on graphite from their liquid. They adsorb with their principal axis parallel to the graphite surface and show both herringbone and parallel structures [87], although the odd–even dependence is reversed compared to the alkanes. In many cases, transitions to a relatively stable rotator phase at higher temperatures are observed prior to melting, and these structures all have the parallel molecular arrangement. The adsorption of the series of molecules 1-fluoro-, 1-chloro-, 1bromo- and 1-iodo-alkanes have all been studied on graphite. Early STM studies of halo-alkanes were done as part of surveys of various adsorbates on graphite [5,88,89]. In general these images show the halogen as a light or a dark region relative to the contrast of the

alkyl chain. The longer molecules tend to adopt lamellae-like structures with the molecular axis perpendicular to the lamellar axis, very similar to the parallel centred structures seen for alkanes. The halogen groups are usually paired so that the molecules lie head-tohead. More detailed and high resolution studies of bromoalkanes have been performed recently [17,90,91]. These have revealed an odd–even effect for the shorter molecules with 1-bromoheptane adopting a herringbone structure while 1-bromohexane a parallel structure. Again this is notably the reverse of the situation seen for alkanes. The molecules in the parallel structure are not necessarily perpendicular to the direction of the lamellae and the precise angle can be dependent on whether the film is prepared in vacuum or from a solvent [90,91]. 4.4.2. Alcohols Like the alkanes, the monolayers of alkyl alcohols on graphite at high coverages melt above their bulk melting points [86,92] and as mentioned earlier exhibit layer-by-layer melting or freezing [86]. In general the monolayers form lamellae structures with the molecular axis at an angle to the lamellar direction and the alcohol groups paired with each other through hydrogen bonding. Both herringbone and parallel structures are seen, depending on the chain length. In some cases both structures can coexist for the same molecule [93] (although this may be a kinetic effect) and may depend on whether adsorption is from pure liquid or solution. X-ray data of the shorter homologues (n-butanol to n-nonanol) indicate that they all form herringbone structures [94,95] and this type of structure is supported by STM for alcohols at least up to n-undecanol [92]. Longer alcohols show significantly more parallel structures and there is a delicate balance between these two structure types, which may be governed by the relative strength of hydrogen bonding compared to the intermolecular van der Waals bonding [93]. 4.4.3. Carboxylic acids As above the normal-carboxylic acids (saturated fatty acids) on graphite are stabilised relative to their bulk at high coverages with elevated melting points [96]. The crystal structures are again lamellar in form but are dominated by dimerisation due to strong intermolecular hydrogen bonding. The structures [97–99] are also lamellar but with interdigitated dimers. This leads to an interesting odd–even structure dependence that is evident in the melting points of the monolayers [96]. Because of the dimer formation, the evenchain length acids are found to adopt two enantiomeric structures which can be distinguished in STM images. However for the same reasons the odd-chain length acid structures adopt a single racemic structure [100,•101]. Studies of the unsaturated fatty acids (carboxylic acids in which the carbon backbone contains at least one unsaturated bond) adsorbed on graphite are much fewer and to our knowledge have only been studied by STM [•101]. The double bond can, of course, exist in the cis- or trans- form and this has an impact on the structures that are adopted. The molecular form of the trans-molecules is not too dissimilar from the saturated case, but the cis-double bond introduces a 120 degree kink in the carbon backbone that is incorporated in the general lamellar structure. However, this does introduce some additional packing constraints that influence the detailed structure. For example, an asymmetric position of the double bond within the alkyl-chain can mean that the interdigitated dimer structures mentioned above are not possible. Instead all the acid groups locate at the same side of the lamellae, with the molecular axis tilted relative to the lamellar direction in order to reduce steric interactions between acid groups. Further, the surface structures also show some interesting surface induced 2D chirality. The structures observed show a complicated dependence on the number of carbon atoms in their alkyl chains, particularly between the acid group and the double bond, with enantiomeric or racemic structures observed. In fact

T. Arnold, S.M. Clarke / Current Opinion in Colloid & Interface Science 17 (2012) 23–32

29

these authors [•101] predict that such odd–even steric effects can, in some cases, lead to a failure to form any stable structure.

the surface so that the ester carbonyl group is positioned away from the graphite surface [115].

4.4.4. Amides As already mentioned, the alkyl amides form extremely stable monolayers on graphite, which melt up to 15–20% higher than the bulk, even at submonolayer coverage, depending on the presence or absence of unsaturated bonds in the alkyl chain [77]. The monolayer stability can be attributed to extensive hydrogen bonding networks forming both intra-dimer and inter-dimer links. This conclusion seems to be confirmed by the structures determined by X-ray diffraction [83,102–104] and STM [5,105,106] and is also true for N-alkyl amides (in which an alkyl chain is bonded to the nitrogen in the amide group) [107]. In fact the existence of this extensive hydrogen-bonding has led to some interesting molecular assemblies being proposed for complex amides [108]. There is evidence of both herringbone and parallel lamellar structures for saturated alkyl amides. For example both structures can be imaged for stearamide (octadecanamide) using STM although the herringbone structure appears to be metastable as it disappears on annealing [106]. For the shorter amides we again see an odd–even effect that disappears for longer amides. The even-amides and long odd-amides (>heptanamide) show a parallel structure while pentanamide and heptanamide both have the herringbone structure [103,104]. There is clearly a delicate balance between these two structures and heptanamide is reported to undergo a phase transition from herringbone to parallel prior to melting. [103]. Meanwhile, trans-monounsaturated amides adopt structures that are very similar to the equivalent saturated amide and the cis-monounsaturated amides, which have a large kink in their carbon backbone also pack well on the surface [93,106].

4.4.8. Ethers and thioethers Ethers and thioethers seem to behave in a way that is similar to the alkanes, forming parallel lamellar structures that have the molecular axis perpendicular (or close to perpendicular) to the lamellar direction [116–119].

4.4.5. Amines Alkyl-amine monolayers melt above the bulk for hexanamine and longer chain lengths [109]. The degree of stabilisation increases with chain length and suggests that the amine group is a destabilising factor even though the observed structures suggest hydrogen bonding in these layers [93,110,111]. The structures show lamellar arrangement with head-to-head arrangement. However the hydrogen bond lengths are much longer than those seen for amides, acids or alcohols and this may explain the relatively less stable monolayer. Reaction with CO2 in the air must be rigorously excluded with the amines. 4.4.6. Aldehydes Aldehydes form stable layers that melt above the bulk melting point for octanal and longer, and are relatively less stable than alkanes. Monolayer structures reported for members of this series show lamellae of parallel molecules with their molecular axis perpendicular to the lamellar direction. Despite extensive diffraction and STM investigation [112,113] the precise symmetry of the structures cannot be determined since the high symmetry structure means that there are very few diffraction peaks or distinguishing features in the STM images. 4.4.7. Esters There has been relatively little work on ester molecules adsorbed on graphite. STM studies [114,115] have reported parallel style lamellar structures. The molecules seem to be distorted into a linear shape when physisorbed and it was suggested that this distortion maximises the interaction with the substrate and improves the commensurateness with the graphite lattice. The molecular axis was seen to form a range of angles with the lamellar direction. Based on these results and supporting calculations it was proposed that these molecules adsorb with the zig–zag plane of molecules perpendicular to

5. Preferential adsorption and mixing in two-dimensions When adsorbed from solution it is usually the case that a surface will preferentially adsorb the longer molecules, an effect that can be attributed to maximising the translational entropy of the system, among other reasons. However for molecules of a similar size this effect is small and so co-adsorption is possible [120]. The consequent phase behaviour has been studied in detail on graphite using a range of techniques and some of these studies are now outlined. 5.1. Alkanes and perfluoroalkanes The coefficient of isomorphism is a useful concept originally applied to three-dimensional structures that is also found to predict mixing behaviour in two-dimensions for n-alkanes adsorbed on graphite [121,122] reasonably well. This model compares the unit cells of the pure components to predict whether structures of the same symmetry will mix. Where the size and shape of the two structures are significantly different (e.g. for odd–even combinations), eutectic-like behaviour is observed [120,123,124]. Mixtures of perfluoroalkane show similar behaviour, but with more ideal mixing than seen in the bulk or for their hydrocarbon analogues. In fact the monolayers do not completely phase separate even for relatively large differences in chain length [••125]. This was explained in an analogous fashion to the bulk behaviour using a subtle molecular replacement model that allows for molecule shape, size and deformability. 5.2. Carboxylic acids, alcohols and amides The rules defined for the alkanes also seem to apply well to binary mixtures of saturated carboxylic acids, primary alcohols and saturated alkyl-amides. Mixing and eutectic behaviour are reasonably well predicted according to the coefficient of isomorphism, (at least for the acids which are adsorbed as dimers) [77,92,96,102,126–128]. However, when mixing does occur there can be some more complicated behaviour than seen for n-alkanes. For example, in alcohol mixtures which differ by a single carbon atom in chain length, where the shorter component is odd, ideal mixing is observed, but when the shorter component is even, co-crystals are formed. For unsaturated compounds we see some surprising behaviour. As might be expected from their similar molecular structures, the saturated and trans-unsaturated amides show good mixing. The cisunsaturated alkyl amides, however, show much greater miscibility than might otherwise be expected from the large kink in their carbon backbone. Following on from diffraction results, STM [106] has recently been used to investigate the local origin of this behaviour. This study shows that the seemingly thermodynamically stable phase is a co-crystal involving lamellar arrays of hetero-dimers (i.e. dimers formed between the two different components rather than with two of the same component). 5.3. Mixed components We have already stated that the relative strengths of adsorption seem to be evident in the relative stabilisation of the monolayer above the bulk melting points. It is interesting, then, to examine

30

T. Arnold, S.M. Clarke / Current Opinion in Colloid & Interface Science 17 (2012) 23–32

binary mixtures in which two different components directly compete for the surface. Thus, competition between perfluoroalkanes and alkanes almost exclusive favours the alkane [129]. Conversely, alcohol/alkane [130] and amide/alkane mixtures [77] show strong preferential adsorption against the alkane. When co-adsorption is possible, due to heavily biased bulk solution composition, both mixing and phase separation are seen, although phase separation is more prevalent. A strong interaction between adsorbates can result in more complicated phase behaviour. For example, acid–amine mixtures show several solid complexes and eutectic, peritectic, dystetic, and incongruent melting behaviours [111]. Halogen bonding (interactions between Iodine containing species and Lewis bases such as N, O and S) provides another example recently reported [•84], showing the structure of a particularly stable co-crystal between di-iodotetrafluorobenzene and 4,4′bipyridine. This interaction is very directional and can be controlled by appropriate use of electron donating and withdrawing substituents and hence represents an interesting direction for the future.

6. Summary and outlook In this short review we have highlighted that the number and availability of diffraction techniques to study adsorbed layers are still growing, although some remain at a modest level due to the technical difficulties involved. Particular advances have come from the increased availability of synchrotron and neutrons scattering facilities as well as technical improvements to other techniques. This has lead to the number of substrates and adsorbates studied that are continuing to grow and it is particularly interesting to see new substrates and the increased use of multi-component adsorbate combinations, which represent an important increase to the complexity of systems being studied. This work continues to provide insight into the subtle influences of intermolecular and substrate-molecule interactions on physisorbed molecular structures. Diffraction studies are an important tool and benefit from use in combination with other complementary techniques. This area continues to attract attention and we look forward to the outcome of the developments we have outlined. Acknowledgements We would like to thank a number of colleagues for their helpful and stimulating discussions during the preparation of this review, particularly J.Z. Larese, H. Taub, H. Hedgeland, S.M. Driver, S. J. Jenkins, and J. E. Parker. References •,•• [1] ••Bruch LW, Diehl RD, Venables JA. Progress in the measurement and modeling of physisorbed layers. Rev Mod Phys 2007;79:1381–454.Comprehensive review of the subject of physisorbed layers, taking a broader view of the field. The paper considers dynamics and theoretical studies in depth and covers most applicable techniques used in the study of such systems. [2] Clarke SM. Neutron diffraction and incoherent neutron scattering from adsorbed layers. Curr Opin Colloid Interface Sci 2001;6:118–25. [3] Inaba A. Structure and phase behavior of two-dimensional solids formed at interfaces. Pure Appl Chem 2006;78:1025–37. [4] Wang D, Wan LJ, Bai CL. Formation and structural transition of molecular selfassembly on solid surface investigated by scanning tunneling microscopy. Mater Sci Eng R Rep 2010;70:169–87. [5] Giancarlo L, Cyr D, Muyskens K, Flynn GW. Scanning tunneling microscopy of molecular adsorbates at the liquid–solid interface: functional group variations in image contrast. Langmuir 1998;14:1465–71. [6] Kuhnle A. Self-assembly of organic molecules at metal surfaces. Curr Opin Colloid Interface Sci 2009;14:157–68. [7] Schreiber F. Structure and growth of self-assembling monolayers. Prog Surf Sci 2000;65:151–257. • ••

Of special interest. Of outstanding interest.

[8] Driver SM, King DA. Methanethiolate structural phases on Cu(111) observed using a novel fibre-optic low-energy electron diffraction instrument. Surf Sci 2007;601:510–7. [9] Human D, Hu XF, Hirschmugl CJ, Ociepa J, Hall G, Jagutzki O, Ullmann-Pfleger K. Low energy electron diffraction using an electronic delay-line detector; 2006. [10] Hedgeland H, Fouquet P, Jardine AP, Alexandrowicz G, Allison W, Ellis J. Measurement of single-molecule frictional dissipation in a prototypical nanoscale system. Nat Phys 2009;5:561–4. [11] Prasad S, Jiang Z, Sinha SK, Dhinojwala A. Partial crystallinity in alkyl side chain polymers dictates surface freezing. Phys Rev Lett 2008;101. [12] Ziblat R, Leiserowitz L, Addadi L. Crystalline lipid domains: characterization by X-ray diffraction and their relation to biology. Angew Chem Int Ed 2011;50: 3620–9. [13] ••Dai PC, Wu Z, Angot T, Wang SK, Taub H, Ehrlich SN. Synchrotron X-raydiffraction study of the structure and growth of Xe films adsorbed on the Ag(111) surface. Phys Rev B 1999;59:15464–79.This paper represents an elegant example of the complex structural information that can be provided by GIXD. [14] ••Kruchten F, Knorr K, Volkmann UG, Taub H, Hansen FY, Matthies B, Herwig KW. Ellipsometric and neutron diffraction study of pentane physisorbed on graphite. Langmuir 2005;21:7507–12.This study nicely combines neutron diffraction with very-high-resolution ellipsometry to provide very detailed structural information for pentane on graphite. [15] Dragnevski KI, Donald AM, Clarke SM, Maltby A. Novel applications of ESEM and EDX for the study of molecularly thin amide monolayers on polymer films. Colloids Surf A Physicochem Eng Aspects 2009;337:47–51. [16] Murdachaew G, de Gironcoli S, Scoles G. Toward an accurate and efficient theory of physisorption. I. Development of an augmented density-functional theory model. J Phys Chem A 2008;112:9993–10005. [17] Florio GM, Werblowskyf TL, Ilan B, Muller T, Berne BJ, Flynn GW. Chain-length effects on the self-assembly of short 1-bromoalkane and n-alkane monolayers on graphite. J Phys Chem C 2008;112:18067–75. [18] •Gross L, Mohn F, Moll N, Liljeroth P, Meyer G. The chemical structure of a molecule resolved by atomic force microscopy. Science 2009;325:1110–4.This is one of the finest examples of atomic-resolution AFM and really demonstrates the potential power of this technique for future studies. [19] Mohn F, Gross L, Meyer G. Measuring the short-range force field above a single molecule with atomic resolution. Appl Phys Lett 2011;99:053106. [20] Gross L, Mohn F, Moll N, Meyer G, Ebel R, Abdel-Mageed WM, Jaspars M. Organic structure determination using atomic-resolution scanning probe microscopy. Nat Chem 2010;2:821–5. [21] Kitaigorodskii AI. Mixed crystals. Berlin: Springer Verlag; 1984. [22] Clarke SM, Thomas RK. The structure of a bromomethane monolayer adsorbed on graphite. Mol Phys 1991;72:413–23. [23] •Plass KE, Grzesiak AL, Matzger AJ. Molecular packing and symmetry of twodimensional crystals. Acc Chem Res 2007;40:287–93.Comprehensive survey of close-packing in published STM images. This paper has lead to a useful database of these structures. [24] Elemans J, De Feyter S. Structure and function revealed with submolecular resolution at the liquid–solid interface. Soft Matter 2009;5:721–35. [25] Arrigoni C, Schull G, Bleger D, Douillard L, Fiorini-Debuisschert C, Mathevet F, Kreher D, Attias AJ, Charra F. Structure and epitaxial registry on graphite of a series of nanoporous self-assembled molecular monolayers. J Phys Chem Lett 2010;1:190–4. [26] Roussel T, Pellenq RJM, Bienfait M, Vix-Guterl C, Gadiou R, Beguin F, Johnson M. Thermodynamic and neutron scattering study of hydrogen adsorption in two mesoporous ordered carbons. Langmuir 2006;22:4614–9. [27] Morishige K, Tateishi M. Accurate relations between pore size and the pressure of capillary condensation and the evaporation of nitrogen in cylindrical pores. Langmuir 2006;22:4165–9. [28] Gatica SM, Calbi MM, Diehl RD, Cole MW. Review: novel physics of gases near carbon nanotubes and buckyballs. J Low Temp Phys 2008;152:89–107. [29] Rols S, Johnson MR, Zeppenfeld P, Bienfait M, Vilches OE, Schneble J. Argon adsorption in open-ended single-wall carbon nanotubes. Phys Rev B 2005;71:5411. [30] Jensen K, Kim K, Zettl A. An atomic-resolution nanomechanical mass sensor. Nat Nanotechnol 2008;3:533–7. [31] Chiu H-Y, Hung P, Postma HWC, Bockrath M. Atomic-scale mass sensing using carbon nanotube resonators. Nano Lett 2008;8:4342–6. [32] •Wang ZH, Wei J, Morse P, Dash JG, Vilches OE, Cobden DH. Phase transitions of adsorbed atoms on the surface of a carbon nanotube. Science 2010;327:552–5.A particularly nice demonstration of the use of the nanotube resonator for studies of the properties of physically adsorbed species. [33] •Diehl RD, Setyawan W, Curtarolo S. Gas adsorption on quasicrystalline surfaces. J Phys Condens Matter 2008;20.Quasicrystals represent a substrate of significant potential for future studies because of their relevance to phenomena such as “superlubricity”. This paper intriguingly summarises current work on gas adsorption on these substrates and recommends future studies of hydrocarbons. [34] •Larese JZ, Arnold T, Frazier L, Hinde RJ, Ramirez-Cuesta AJ. Direct observation of H2 binding to a metal oxide surface. Phys Rev Lett 2008;101:165302.This study of Hydrogen on MgO shows a very unusual shift in the position of the ortho-para transition of hydrogen. Combination with thermodynamic and diffraction results allowed a detailed understanding of the system that is of particular interest for other studies of hydrogen adsorption on other substrates. [35] Drummond ML, Sumpter BG, Shelton WA, Larese JZ. Electronic structure investigation of surface-adsorbate and adsorbate–adsorbate interactions in multilayers of CH4 on MgO(100). J Phys Chem C 2007;111:966–76.

T. Arnold, S.M. Clarke / Current Opinion in Colloid & Interface Science 17 (2012) 23–32 [36] Drummond ML, Sumpter BG, Shelton WA, Larese JZ. Density functional investigation of the adsorption of a methane monolayer on an MgO(100) surface. Phys Rev B 2006;73. [37] Larese JZ, Marero DMY, Sivia DS, Carlile CJ. Tracking the evolution of interatomic potentials with high resolution inelastic neutron spectroscopy. Phys Rev Lett 2001;8720. [38] •Vogt J, Weiss H. LEED and PIRS structure analysis of physisorbed molecules on insulators: monolayer C2D2/KCl(100). Phys Rev B 2008;77:125415.This study takes advantage of recent LEED advances to provide very nice data of weakly adsorbed molecules on an insulating surface. [39] Cabello-Cartagena AG, Vogt J, Weiss H. Structure and infrared absorption of the first layer C2H2 on the NaCl(100) single-crystal surface. J Chem Phys 2010;132: 4706. [40] Peters C, Morrison JA, Klein ML. The adsorption of acetylene on a graphite surface. Surf Sci 1986;165:355–74. [41] Elmiger M, Larese JZ, Zhang QM. Bull Am Phys Soc 1990;35:593. [42] Coulomb JP, Trabelsi M, Larher Y, Mirebeau I. Neutron diffraction investigation of the structure of the C2D2 film physisorbed upon MgO(100). Mol Phys 1994;81:1259. [43] Ferry D, Hoang PNM, Suzanne J, Bibérian JP, Van Hove MA. Structure of physisorbed molecules on an oxide surface from potential calculations and dynamical low-energy electron diffraction analysis: acetylene on MgO(100). Phys Rev Lett 1997;78:4237. [44] Herwig KW, Newton JC, Taub H. Structure and growth of butane films adsorbed on graphite. Phys Rev B 1994;50:15287–97. [45] Bruch LW. Monolayer solids of short alkanes and perfluoroalkanes. J Phys Chem C 2009;113:17399–406. [46] Arnold T, Dong CC, Thomas RK, Castro MA, Perdigon A, Clarke SM, Inaba A. The crystalline structures of the odd alkanes pentane, heptane, nonane, undecane, tridecane and pentadecane monolayers adsorbed on graphite at submonolayer coverages and from the liquid. Phys Chem Chem Phys 2002;4:3430–5. [47] Castro MA, Clarke SM, Inaba A, Arnold T, Thomas RK. Anomalous behaviour of pentane adsorbed at the graphite/liquid interface. Phys Chem Chem Phys 1999;1: 5203–7. [48] Arnold T, Thomas RK, Castro MA, Clarke SM, Messe L, Inaba A. The crystalline structures of the even alkanes hexane, octane, decane, dodecane and tetradecane monolayers adsorbed on graphite at submonolayer coverages and from the liquid. Phys Chem Chem Phys 2002;4:345–51. [49] Diama A, Matthies B, Herwig KW, Hansen FY, Criswell L, Mo H, Bai M, Taub H. Structure and phase transitions of monolayers of intermediate-length n-alkanes on graphite studied by neutron diffraction and molecular dynamics simulation. J Chem Phys 2009;131:084707. [50] Zhao M, Jiang P, Deng K, Yu AF, Hao YZ, Xie SS, Sun JL. Insight into STM image contrast of n-tetradecane and n-hexadecane molecules on highly oriented pyrolytic graphite. Appl Surf Sci 2011;257:3243–7. [51] Weaver JF, Carlsson AF, Madix RJ. The adsorption and reaction of low molecular weight alkanes on metallic single crystal surfaces. Surf Sci Rep 2003;50:107–99. [52] Arnold T, Barbour A, Chanaa S, Cook RE, Fernandez-Canato D, Landry P, Seydel T, Yaron P, Larese JZ. Melting of thin films of alkanes on magnesium oxide. Eur Phys J Spec Top 2009;169:143–50. [53] Lei RZ, Gellman AJ, Koel BE. Desorption energies of linear and cyclic alkanes on surfaces: anomalous scaling with length. Surf Sci 2004;554:125–40. [54] Firment LE, Somorjai GA. Surface structures of normal paraffins and cyclohexane monolayers and thin crystals grown on the (111) crystal face of platinum. A lowenergy electron diffraction study. J Chem Phys 1977;66:2901–13. [55] Weaver JF, Ikai M, Carlsson A, Madix RJ. Molecular adsorption and growth of n-butane adlayers on Pt(111). Surf Sci 2001;470:226–42. [56] Firment LE, Somorjai GA. Low-energy electron diffraction study of the surface of thin crystals and monolayers of normal paraffins and cyclohexane on the Ag(111) crystal surface. J Chem Phys 1978;69:3940–52. [57] Wu Z, Ehrlich SN, Matthies B, Herwig KW, Dai PC, Volkmann UG, Hansen FY, Taub H. Growth of n-alkane films on a single-crystal substrate. Chem Phys Lett 2001;348:168–74. [58] Fuhrmann D, Graham AP, Criswell L, Mo H, Matthies B, Herwig KW, Taub H. Effects of chain branching on the monolayer structure of alkanes at interfaces: a neutron and helium atom scattering study. Surf Sci 2001;482:77–82. [59] Fuhrmann D, Gerlach R, Rubahn HG, Woll C. Structure and phase transitions of ultrathin films of alkanes adsorbed on Cu(111). Surf Sci 1999;424:145–54. [60] Zhang HM, Xie ZX, Mao BW, Xu X. Self-assembly of normal alkanes on the Au(111) surfaces. Chem Eur J 2004;10:1415–22. [61] Baxter RJ, Teobaldi G, Zerbetto F. Modeling the adsorption of alkanes on an Au(111) surface. Langmuir 2003;19:7335–40. [62] •Cousty J, Marchenko A. Substrate-induced freezing of alkane monolayers adsorbed on Au(111) dependant on the alkane/gold misfit. Surf Sci 2002;520: 128–36.This paper demonstrates that melting transitions can be observed using STM. It shows that on gold there may be additional transitions prior to melting that are not seen on other substrates and may arise due to the corrugation of the surface potential. [63] Cincotti S, Rabe JP. Self-assembled alkane monolayers on Mose2 and Mos2. Appl Phys Lett 1993;62:3531–3. [64] ••Arnold T, Chanaa S, Larese JZ, Cook RE, Clarke SM. Structure of an n-butane monolayer adsorbed on magnesium oxide (100). Phys Rev B 2006;74: 085421.This paper nicely shows that herringbone structures of alkanes can exist on surfaces with square symmetry. It is rare to see such good powder diffraction data on non-graphitic substrates. [65] Gatica SM, Cole MW. To wet or not to wet: that is the question. J Low Temp Phys 2009;157:111–36.

31

[66] Van LP, Kyrylyuk V, Polesel-Maris J, Thoyer F, Lubin C, Cousty J. Experimental three-dimensional description of the liquid hexadecane/graphite interface. Langmuir 2009;25:639–42. [67] •Barbour AM, Telling MTF, Larese JZ. Investigation of the behavior of ethylene molecular films using high resolution adsorption isotherms and neutron scattering. Langmuir 2010;26:8113–21.This paper presents a nice illustration of the subtle behaviour of these physisorbed systems including wetting transitions that are very similar despite very different substrates (MgO compared to graphite). [68] Bai M, Trogisch S, Magonov S, Taub H. Explanation and correction of false step heights in amplitude modulation atomic force microscopy measurements on alkane films. Ultramicroscopy 2008;108:946–52. [69] Cisternas EA, Corrales TP, del Campo V, Soza PA, Volkmann UG, Bai MJ, Taub H, Hansen FY. Crystalline-to-plastic phase transitions in molecularly thin n-dotriacontane films adsorbed on solid surfaces. J Chem Phys 2009;131:4705. [70] del Campo V, Cisternas E, Taub H, Vergara I, Corrales T, Soza P, Volkmann UG, Bai MJ, Wang SK, Hansen FY, et al. Structure and growth of vapor-deposited n-dotriacontane films studied by X-ray reflectivity. Langmuir 2009;25:12962–7. [71] Lu LB, Zander KJ, Cai YG. Stability of the parallel layer during alkane spreading and the domain structures of the standing-up layer. Langmuir 2010;26:5624–31. [72] Yamamoto T, Nozaki K, Yamaguchi A, Urakami N. Molecular simulation of crystallization in n-alkane ultrathin films: effects of film thickness and substrate attraction. J Chem Phys 2007;127:4704. [73] Wang SK, Mamontov E, Bai M, Hansen FY, Taub H, Copley JRD, Sakai VG, Gasparovic G, Jenkins T, Tyagi M, et al. Localized diffusive motion on two different time scales in solid alkane nanoparticles. Epl 2010;91:66007. [74] Enevoldsen AD, Hansen FY, Diama A, Criswell L, Taub H. Comparative study of normal and branched alkane monolayer films adsorbed on a solid surface. I. Structure. J Chem Phys 2007;126. [75] Clarke SM, Inaba A, Arnold T, Thomas RK. Calorimetric investigation of the monolayers formed at solid–liquid interface. J Therm Anal Calorim 1999;57: 643–51. [76] Espeau P, White JW. Thermodynamic properties of n-alkanes in porous graphite. J Chem Soc Faraday Trans 1997;93:3197–200. [77] Arnold T, Clarke SM. Thermodynamic investigation of the adsorption of amides on graphite from their liquids and binary mixtures. Langmuir 2008;24:3325–35. [78] Enevoldsen AD, Hansen FY, Diama A, Taub H, Dimeo RM, Neumann DA, Copley JRD. Comparative study of normal and branched alkane monolayer films adsorbed on a solid surface. II. Dynamics. J Chem Phys 2007;126. [79] Alba MD, Castro MA, Clarke SM, Perdigon AC. NMR study of n-dodecane adsorbed on graphite. Solid State Nucl Magn Reson 2003;23:174–81. [80] Mutihac RC, Riegler H. Phase transition broadening due to interfacial premelting: a new quantitative access to intermolecular interactions within submonolayer films at solid/vapor interfaces. Langmuir 2010;26:6394–9. [81] Larese JZ, Passell L, Heidemann AD, Richter D, Wicksted JP. Melting in 2-dimensions — the ethylene-on-graphite system. Phys Rev Lett 1988;61:432–5. [82] Hansen FY, Taub H. Melting mechanism in monolayers of flexible rod-shaped molecules. Phys Rev Lett 1992;69:652–5. [83] Bhinde T, Arnold T, Clarke SM. The structure of dodecanamide monolayers adsorbed on graphite. Prog Colloid Polym Sci 2010;137:5–8. [84] •Clarke SM, Friscic T, Jones W, Mandal A, Sun CG, Parker JE. Observation of a twodimensional halogen-bonded cocrystal at sub-monolayer coverage using synchrotron X-ray diffraction. Chem Commun 2011;47:2526–8.This paper indicates that halogen bonding provides a significant new variety of intermolecular interaction in physisorbed monolayers. This has potential to be as interesting and important as hydrogen bonding which presently dominates studies of noncovalent interactions in monolayers. [85] Parker JE, Clarke SM. Solid monolayers of fluorocarbons adsorbed on graphite from liquids. Colloids Surf A Physicochem Eng Aspects 2007;298:145–7. [86] Messe L, Perdigon A, Clarke SM, Castro MA, Inaba A. Layer-by-layer surface freezing of linear alcohols at the graphite/liquid interface. J Colloid Interface Sci 2003;266:19–27. [87] Parker JE, Clarke SM, Perdigon AC, Inaba A. The crystalline structures of fluoroalkane monolayers adsorbed on graphite at submonolayer coverages. J Phys Chem C 2009;113:21396–405. [88] Cyr DM, Venkataraman B, Flynn GW. STM investigations of organic molecules physisorbed at the liquid–solid interface. Chem Mater 1996;8:1600–15. [89] Claypool CL, Faglioni F, Goddard III WA, Gray HB, Lewis NS, Marcus RA. Source of image contrast in STM images of functionalized alkanes on graphite: a systematic functional group approach. J Phys Chem B 1997;101. [90] Florio GM, Ilan B, Muller T, Baker TA, Rothman A, Werblowsky TL, Berne BJ, Flynn GW. Solvent effects on the self-assembly of 1-bromoeicosane on graphite. Part I. Scanning tunneling microscopy. J Phys Chem C 2009;113:3631–40. [91] Ilan B, Florio GA, Werblowsky TL, Muller T, Hybertsen MS, Berne BJ, Flynn GW. Solvent effects on the self-assembly of 1-bromoeicosane on graphite. Part II. Theory. J Phys Chem C 2009;113:3641–9. [92] Wang GJ, Lei SB, De Feyter S, Feldman R, Parker JE, Clarke SM. Behavior of binary alcohol mixtures adsorbed on graphite using calorimetry and scanning tunneling microscopy. Langmuir 2008;24:2501–8. [93] Miao XR, Chen CM, Zhou J, Deng WL. Influence of hydrogen bonds and double bonds on the alkane and alkene derivatives self-assembled monolayers on HOPG surface: STM observation and computer simulation. Appl Surf Sci 2010;256:4647–55. [94] Morishige K, Kato T. Chain-length dependence of melting of n-alcohol monolayers adsorbed on graphite: n-hexanol, n-heptanol, n-octanol, and n-nonanol. J Chem Phys 1999;111:7095–102.

32

T. Arnold, S.M. Clarke / Current Opinion in Colloid & Interface Science 17 (2012) 23–32

[95] Morishige K, Sakamoto Y. Melting of n-butanol and n-pentanol monolayers adsorbed on graphite — effect of molecular length on melting. J Chem Phys 1995;103:2354–60. [96] Bickerstaffe A, Messe L, Clarke SM, Parker J, Perdigon A, Cheah NP, Inaba A. Mixing behaviour of carboxylic acids adsorbed on graphite. Phys Chem Chem Phys 2004;6:3545–50. [97] Bickerstaffe AK, Cheah NP, Clarke SM, Parker JE, Perdigon A, Messe L, Inaba A. The crystalline structures of carboxylic acid monolayers adsorbed on graphite. J Phys Chem B 2006;110:5570–5. [98] Senenko AI, Stryzheus DV, Marchenko AA, Naumovets AG. A comparative STMinvestigation of monolayers of n-alkanes and arachidic acid on the liquid tetradecane–graphite interface. Materialwiss Werkst 2011;42:55–8. [99] Mali KS, Lava K, Binnemans K, Feyter S. Hydrogen bonding versus van der Waals interactions: competitive influence of noncovalent interactions on 2D selfassembly at the liquid–solid interface. Chem Eur J 2010;16:14447–58. [100] Hibino M, Sumi A, Tsuchiya H, Hatta I. Microscopic origin of the odd–even effect in monolayer of fatty acids formed on a graphite surface by scanning tunneling microscopy. J Phys Chem B 1998;102:4544–7. [101] •Tao F, Goswami J, Bernasek SL. Self-assembly and odd–even effects of cisunsaturated carboxylic acids on highly oriented pyrolytic graphite. J Phys Chem B 2006;110:4199–206.Very interesting STM analysis of chirality seen in the monolayers of cis-unsaturated fatty acids, that shows behaviour is intricately linked to the position of the double bond within the carbon backbone. This kind of information is not possible using traditional powder diffraction methods. [102] Bhinde T, Brewer AY, Clarke SM, Phillips TK, Arnold T, Parker JE. Adsorption of unsaturated amides on a graphite surface: trans-unsaturated amides. J Phys Chem C 2011;115:6682–9. [103] Bhinde T, Phillips TK, Clarke SM, Arnold T, Parker J. Phase behavior of heptanamide adsorbed on a graphite substrate. Langmuir 2011;27:15–8. [104] Bhinde T, Clarke SM, Phillips TK, Arnold T, Parker JE. Crystalline structures of alkylamide monolayers adsorbed on the surface of graphite. Langmuir 2010;26:8201–6. [105] Takeuchi H, Kawauchi S, Ikai A. Differentiation of chemically functional groups in stearoyl amide and anilide with scanning tunneling microscopy. Jpn J Appl Phys 1 1996;35:3754–8. [106] Mali KS, Van Averbeke B, Bhinde T, Brewer AY, Arnold T, Lazzaroni R, Clarke SM, De Feyter S. To Mix or Not To Mix: 2D Crystallization and Mixing Behavior of Saturated and Unsaturated Aliphatic Primary Amides. ACS Nano 2011;5:9122–37. [107] Zou B, Dreger K, Muck-Lichtenfeld C, Grimme S, Schafer HJ, Fuchs H, Chi LF. Simple and complex lattices of N-alkyl fatty acid amides on a highly oriented pyrolytic graphite surface. Langmuir 2005;21:1364–70. [108] Ahn S, Matzger AJ. Six different assemblies from one building block: twodimensional crystallization of an amide amphiphile. J Am Chem Soc 2010;132: 11364–71. [109] Cheah NP, Messe L, Clarke SM. The formation of solid monolayers of linear amines adsorbed on graphite from the liquid. J Phys Chem B 2004;108:4466–9. [110] Cyr DM, Venkataraman B, Flynn GW, Black A, Whitesides GM. Functional group identification in scanning tunneling microscopy of molecular adsorbates. J Phys Chem 1996;100:13747–59. [111] Sun C, Bojdys MJ, Clarke SM, Harper LD, Jefferson A, Castro MA, Medina S. Bulk and adsorbed monolayer phase behavior of binary mixtures of undecanoic acid and undecylamine: catanionic monolayers. Langmuir 2011;27:3626–37. [112] Phillips TK, Bhinde T, Clarke SM, Lee SY, Mali KS, De Feyter S. Adsorption of aldehydes on a graphite substrate: combined thermodynamic study of C6–C13

[113]

[114]

[115]

[116] [117]

[118]

[119] [120] [121]

[122] [123]

[124]

[125]

[126]

[127]

[128]

[129]

[130]

homologues with a structural and dynamical study of dodecanal. J Phys Chem C 2010;114:6027–34. Phillips TK, Clarke SM, Bhinde T, Castro MA, Millan C, Medina S. Monolayer structures of alkyl aldehydes: odd-membered homologues. Thin Solid Films 2011;519:3123–7. Tao F, Cai Y, Bernasek SL. Scanning tunneling microscopy studies of the selfassembly of carboxylic esters on graphite: linear distortion and multiple adsorption structures. Langmuir 2005;21:1269–76. Miao XR, Xu L, Liao CY, Li ZM, Zhou JA, Deng WL. Two-dimensional self-assembly of esters with different configurations at the liquid–solid interface. Appl Surf Sci 2011;257:4559–65. Duim WC, Clarke SM. Adsorption and mixing behavior of ethers and alkanes at the solid/liquid interface. J Phys Chem B 2006;110:23853–9. Papadantonakis KM, Brunschwig BS, Lewis NS. Scanning tunneling microscopy studies of monolayer templates: alkylthioethers and alkylethers. Langmuir 2008;24:10543–8. Papadantonakis KA, Brunschwig BS, Lewis NS. Use of alkane monolayer templates to modify the structure of alkyl ether monolayers on highly ordered pyrolytic graphite. Langmuir 2008;24:857–61. Padowitz DF, Messmore BW. STM observations of exchange dynamics at the solid/liquid interface using a molecular tracer. J Phys Chem B 2000;104:9943–6. Castro MA, Clarke SM, Inaba A, Arnold T, Thomas RK. Competitive adsorption of simple linear alkane mixtures onto graphite. J Phys Chem B 1998;102:10528–34. Clarke SM, Messe L, Adams J, Inaba A, Arnold T, Thomas RK. A quantitative parameter for predicting mixing behaviour in adsorbed layers: the 2D isomorphism coefficient. Chem Phys Lett 2003;373:480–5. Espeau P, White JW, Papoular RJ. The structure of n-alkane binary mixtures adsorbed on graphite. Appl Surf Sci 2005;252:1350–9. Castro MA, Clarke SM, Inaba A, Arnold T, Thomas RK. The investigation of mixed monolayers adsorbed from solution: octane and nonane mixtures on graphite. Phys Chem Chem Phys 1999;1:5017–23. Castro M, Clarke SM, Inaba A, Thomas RK, Arnold T. Adsorption behaviour of the binary mixtures of octane and nonane at sub-monolayer coverage on graphite. Phys Chem Chem Phys 2001;3:3774–7. ••Parker J, Clarke SM. Mixing in adsorbed monolayers: perfluorinated alkanes. Langmuir 2008;24:4833–44.This paper presents a detailed insight into mixing studies of binary adsorbed layers beyond the usual regular solution theory and including details of the replacement model of mixing with deformation of the molecules. Bickerstaffe AK, Clarke SM. The interpretation of mixing behaviour in carboxylic acid monolayers adsorbed on graphite using a regular solution description. Colloids Surf A Physicochem Eng Asp 2007;298:80–2. Messe L, Clarke SM, Arnold T, Dong C, Thomas RK, Inaba A. Mixing behavior at the solid/liquid interface: binary monolayers of linear alcohols adsorbed on graphite. Langmuir 2002;18:4010–3. Messe L, Clarke SM, Dong CC, Thomas RK, Inaba A, Alba MD, Castro MA. Mixing behavior at the solid/liquid interface: binary alcohol monolayers on graphite. Langmuir 2002;18:9429–33. Parker JE, Clarke SM, Perdigon AC. Preferential adsorption of solid monolayers of hydrocarbons over fluorocarbons at the solid/liquid interface. Surf Sci 2007;601: 4149–53. Messe L, Perdigon A, Clarke SM, Inaba A, Arnold T. Alkane/alcohol mixed monolayers at the solid/liquid interface. Langmuir 2005;21:5085–93.