Adsorption of Aldehydes on a Graphite Substrate - American Chemical

J. Phys. Chem. C 2010, 114, 6027–6034

6027

Adsorption of Aldehydes on a Graphite Substrate: Combined Thermodynamic Study of C6-C13 Homologues with a Structural and Dynamical Study of Dodecanal Tamsin K. Phillips, Tej Bhinde, Stuart M. Clarke,* and Seung Y. Lee BP Institute and Department of Chemistry, UniVersity of Cambridge, Madingley Rise, Madingley Road, Cambridge CB3 0HE, United Kingdom

Kunal S. Mali and Steven De Feyter DiVision of Molecular and Nanomaterials and INPAC-Institute for Nanoscale Physics and Chemistry, Katholieke UniVersiteit LeuVen, Celestijnenlaan 200 F, B-3001 HeVerlee, Belgium ReceiVed: NoVember 21, 2009; ReVised Manuscript ReceiVed: February 4, 2010

In this work we present a study on the adsorption of linear alkyl aldehydes physisorbed from their bulk liquid onto a graphite substrate combining calorimetry for all homologues from C6 to C13, with more detailed diffraction, incoherent neutron scattering, and scanning tunneling microscopy techniques for one (C12) representative member. We identify solid monolayer formation for some of these species for alkyl chain lengths of 6 to 13 carbons at high surface coverages. The C12 monolayer structure is determined to be most likely Pgg and this structure is discussed in terms of the importance of dipolar interactions. 1. Introduction The physisorption of organic liquids to a solid substrate forming a solid monolayer at the interface has been investigated for a variety of chemical species of increasing chemical complexity, including alkanes,1–8 alcohols,9–15 ethers,16 carboxylic acids,17–19 and amides.20 The solid monolayer formed in this process can have a significant effect on the surface properties and is therefore important in many areas including lubrication, wetting, and detergency. Alkanes are the simplest type of hydrocarbons to be investigated from the essentially spherical methane to the long linear chains of the higher homologues. Alkanes are found to form solid monolayers at both submonolayer and multilayer coverages, although presolidification (formation of a solid layer coexisting with the bulk liquid) is not observed for alkyl chain lengths of 7 or fewer carbon atoms. Adsorbed alkanols show the formation of hydrogen bonds in the adsorbed monolayer9,12 and are found to be more stable than the equivalent alkane of the same alkyl chain length. In contrast, ethers, isomers of the alcohols, do not form such stable solid monolayers,16 a behavior attributed mainly to the lack of hydrogen bonding and the lone pair repulsion of adjacent ether oxygen atoms in the solid monolayers. In this work, we focus on adsorption of another important oxygen-containing group, the carbonyl, CdO. This is a key component of a number of important surface-active species such as the carboxylic acids19 (-CO2H) and amides20 (-CONH2), both of which form very stable solid monolayers, attributed to very significant hydrogen bonding and in-plane dimerization. As part of this investigation, we present here a study of the adsorption of straight chain aldehydes (-COH) to a graphite surface. Studies of related species, ketones (-CO-) and esters (-CO2R), will appear separately. We show that these carbonylcontaining aldehydes do indeed form stable solid monolayers * To whom correspondence should be addressed. E-mail: stuart@ bpi.cam.ac.uk.

at both submonolayer and multilayer coverages adsorbed on graphite and present the monolayer structure. Key characteristics of the carbonyl group that may affect the adsorption properties include the dipole moments. For carbon monoxide the dipole moment is rather small (0.11 D21) compared to other carbonyl species such as formaldehyde, 2.332 D, ethanal, 2.750 D, and propanal 2.72 D.22 Dipoles are known to favor particular molecular orientations, where the dipoles can align, “head to tail”. Interestingly, the dipole moment of carbon monoxide is rather small compared to the quadrupole moment, which would favor other molecular orientations where the CdO bonds are perpendicular to each other, in a herringbone arrangement. The monolayer structure of CO adsorbed on graphite has been studied and highlights these issues.23 The carbonyl bond is also a double bond and unsaturation may also influence the interaction of the adsorbed species with the surface, for example, the π orbitals of the carbonyl and the π system of the graphite. Interestingly, studies of alkenes and phenyl alkenes have indicated that although unsaturated species can form solid monolayers, longer homologues are required to observe presolidification compared to the alkanes.24 2. Materials and Methods 2.1. Materials. C6-C13 aldehydes were purchased from Sigma-Aldrich, UK, and used without further purification (purities: C6 99.8%, C7 92.79%, C8 99.4%, C10 99%, C11 97.6%, C12 98.9%, and C13 99.2%). Perdeuterated dodecanal was prepared in the Department of Chemistry, University of Cambridge, by oxidation of the alcohol using pyridine dichromate. The purity was assessed by NMR, elemental analysis, and IR spectroscopy (C12 97%). The substrate “Papyex”, a lightly recompressed exfoliated graphite, was purchased from Le Carbone Lorraine, characterized by nitrogen adsorption isotherm measurements, and found to have a specific surface area of 29.9 m2 g-1. In this work Cn refers to an aldehyde with n carbon atoms in the alkyl chain. For DSC measurements the aldehydes were dosed on the substrate by weighing approximately 10 mg of the aldehydes and approximately 15-20

10.1021/jp911069t  2010 American Chemical Society Published on Web 03/04/2010

6028

J. Phys. Chem. C, Vol. 114, No. 13, 2010

mg of the graphite substrate into a aluminum sample pans which were hermetically sealed. The samples were annealed prior to use at a temperature above the expected monolayer melting point (approximately 20-30% above the bulk melting point of the adsorbate) close to the bulk boiling point to ensure equilibrium has been attained, and then they were cooled slowly. 2.2. Methods. Here we have exploited a combination of scattering by neutrons and X-rays, scanning tunneling microscopy (STM), and calorimetry to identify and characterize these monolayers. Differential scanning calorimetry (DSC) has been demonstrated to be able to provide important information on the formation of solid monolayers adsorbed from liquids to a solid substrate.25–27 The approach has been described in detail elsewhere.14 In outline, small heat changes when an adsorbed layer undergoes a phase change (e.g., melting) can be observed. This approach therefore is an important indicator of the formation of adsorbed solid monolayers but can also provide quantitative data on the monolayer phase transition, such as the monolayer melting point and melting enthalpy.26,28 Surface phase behavior can also be extracted from composition-dependent measurements. If, at high coverages, there are significant differences in the melting points of the monolayer and the bulk adsorbate, calorimetry can show these transitions separately and the monolayer behavior is reasonably clearly identified. The DSC device used here was a Pyris 1, power-compensated DSC from Perkin-Elmer calibrated using standard procedures for temperature with indium and heptane and for enthalpy by indium. The device measures the difference in energy required to heat the sample and an inert reference at the same rate. Phase changes in the sample give rise to more energy being supplied to the sample side and a peak is observed in the thermogram. The dynamic nature of the technique means that for conventional first-order bulk-phase transitions the peak in the thermogram does not represent the best estimate of the transition temperature, rather the “onset”, where the peak first leaves the baseline, is a better estimate. For the rather weak peaks of the monolayers of interest here, the onset is somewhat difficult to extract reliably and the peak temperature can be quoted instead. However, for the small monolayer peaks of interest here the difference in temperature of the peak and onset is typically 2 °C. The samples were run at 20 K/min, unless otherwise stated, representing a compromise between sensitivity and temperature precision. X-ray diffraction measurements were made at the Paul Scherrer Institut in Switzerland, using the instrument X04SApowder, a power diffraction beamline with a banana multidetector. The incident wavelength was selected to be approximately 0.1 nm, and at the Diamond Light Source in the UK, using the I11 beamline with a crystal analyzer detector and a wavelength of 0.12 nm. The samples were small 2 mm diameter discs of graphite sealed in Lindeman glass capillaries. The temperature was controlled using a standard nitrogen cryostream. The neutron diffraction was carried out at the Institut Laue-Langevin in Grenoble, France, using instrument D20, a powder diffractometer with banana multidetector and incident wavelength of 0.241 nm.29 The samples were approximately 12 g of graphite in an aluminum sample can sealed with indium. The temperature was controlled using a standard orange cryostat. Diffraction is ideal for the study of adsorbed monolayers at submonolayer coverages where the sawtooth line shape characteristic of 2D monolayers clearly indicates the formation of solid layers. The use of diffraction for the study of monolayers coexisting with their liquids is complicated by the presence of the very intense broad fluid scattering of the bulk fluid swamping the much weaker signal from the adsorbed layer.30–32 In

Phillips et al. favorable cases the scattering from the monolayer can be obtained by careful subtraction of the liquid scattering. Diffraction patterns obtained using X-rays and neutrons differ because these two types of radiation are scattered from different components of the structure: X-rays are scattered by the electrons and neutrons by the nuclei. This complementarity is often of benefit in monolayer studies, given that the number of reflections accessible in a 2D diffraction pattern is usually small. Because of the strong incoherent scattering of neutrons by hydrogen, neutron diffraction requires perdeuterated species. Normal protonated species can be used for synchrotron X-ray studies. Incoherent elastic neutron scattering can also be used to confirm the presence of a solid adsorbed monolayer in the presence of bulk liquid adsorbate.33 In brief, the intensity of incoherent elastic scattering from a sample is a measure of how much solid sample is present. From a sample initially at very low temperatures where the entire sample is frozen one expects to observe high elastic scattering. On heating the sample through the bulk melting point when most of the adsorbate becomes mobile, the elastic intensity is expected to fall. However, if a solid monolayer is present, then the elastic intensity will not fall to zero, reflecting the small contribution from the solid monolayer. This residual intensity will only vanish above the monolayer melting point. Hence, the temperature dependence of the elastic incoherent scattering from a sample should indicate the presence of a solid monolayer coexisting with its liquid. In this work the high energy resolution instrument IN10 at the Institute Laue-Langevin, France, has been used. The incident and scattered neutrons at a wavelength of 6.271 Å are selected using backscattering from Si(111) crystals giving the high energy resolution of 1 µeV. The samples are dosed on to approximately 7 g of graphite (Papyex) cut into rings and annealed in glassware under reduced pressure before sealing into aluminum cans with indium wire. STM samples were prepared by heating the aldehyde just above the melting point and then applying a drop of the neat aldehyde onto a freshly cleaved highly oriented pyrolytic graphite surface (HOPG, grade ZYB, Advanced Ceramics Inc., Cleveland, OH). The STM experiments were carried out using a PicoSPM (Agilent) operating in the constant-current mode with the tip immersed in the liquid at room temperature. STM tips were prepared by mechanical cutting from Pt/Ir wire (80%/ 20%, diameter 0.2 mm). For analysis purposes, the imaging of a molecular layer was immediately followed by recording at a lower bias voltage the graphite lattice, under otherwise identical experimental conditions. Drift effects were corrected by using Scanning Probe Image Processor (SPIP) software (Image Metrology A/S). Initial models for the calculated diffraction patterns used a simple geometry for the dodecanal molecule devised from known average bond lengths and angles for aldehydes.34 The fine adjustments of the model calculations were carried out using a molecular geometry optimized using DFT in the Gaussian 03 program35 using the B3LYP36 functionals and the 6-31G(d,p) basis set.37,38 3. Results: Pure Aldehyde Samples 3.1. Calorimetry. Aldehydes of alkyl chain lengths from 6 to 13 were added to graphite samples and the DSC thermograms collected, as outlined above. Figure 1 illustrates a representative DSC thermogram from C12 aldehyde adsorbed on graphite. This sample has a dosing corresponding to approximately 40 equivalent monolayers of the aldehyde. The two larger peaks

Adsorption of Aldehydes on a Graphite Substrate

Figure 1. DSC thermogram for approximately 40 equivalent monolayers of dodecanal adsorbed on graphite. Inset shows an enlargement of the monolayer melting transition peak.

Figure 2. Calorimetry results showing bulk and monolayer melting transition temperatures for straight chain aldehydes C6 to C13. The monolayer melting transitions are approximately 10-15% higher than the bulk transition temperatures.

(onset at approximately 270 and 285 K) are observed in the thermogram of the pure aldehyde alone and hence are attributed to bulk transitions (a solid-solid phase transition, typical in long-chain alkyl species, and the bulk melting point). The weaker peak at higher temperature (onset at approximately 317 K) is not present in the thermogram of the graphite nor the pure adsorbate; hence, we attribute this feature to the melting of a solid monolayer as observed for alkanes,39–41 alcohols,14,42 carboxylic acids,19 amines,43 amides,20 and other adsorbed materials confirmed by diffraction and other techniques. No additional peaks are observed that could be attributed to a solid-solid transition in the monolayer or a multilayer melting transition as seen in other systems.14,44,45 For members of the series from C8 to C13 the thermograms were similar in nature to that in Figure 1, indicating formation of a solid monolayer which melts at a temperature significantly above the bulk melting point. This behavior is typical of a number of other alkyl species discussed in the Introduction where other techniques, such as diffraction and scanning tunneling microscopy, have confirmed solid monolayer formation. Hence, we tentatively assign this feature to the formation of a solid monolayer and its melting. Figure 2 shows a graphical representation of the bulk and monolayer melting points for a number of these materials as a function of the alkyl chain length. Figure 2 indicates that the longer homologues with more than 8 carbons in the alkyl chain all appear to form solid monolayers. Interestingly, the shorter homologues did not show any evidence of a solid monolayer

J. Phys. Chem. C, Vol. 114, No. 13, 2010 6029

Figure 3. (a, lower pattern) X-ray and (b, upper pattern) neutron diffraction patterns from 0.5 monolayers of dodecanal adsorbed on graphite. The solid lines represent the results of a structural fit to the data using a simple primitive unit cell model discussed in the text below.

melting transition. This presolidification behavior of the longer aldehydes (relative to the shorter homologues) is similar to that observed with alkanes, alcohols, and other organic molecules,7 with the melting transitions for the monolayers tending to a value between 10 and 15% higher than the melting temperature of the bulk.46 It is noted that although we do not observe a transition above the bulk melting point arising from the monolayer melting for the shorter homologues, this does not mean that solid monolayers do not form for these species. It could be that the monolayer melting point is the same as or close to the bulk melting point and cannot be resolved in the DSC thermograms presented. This behavior has been observed with pentane.7 In addition, some members (C11 and longer chains) show additional bulk thermodynamic transitions in the DSC data, which are attributed to solid-solid phase transitions, typical of many alkyl species. 3.2. X-ray and Neutron Diffraction: Low Coverage. X-ray and neutron diffraction patterns at low, submonolayer coverage (0.5 equivalent monolayers) and low temperature for dodecanal aldehyde are given in Figure 3a,b. These patterns have been obtained after subtraction of the scattering from the graphite without adsorbate and hence the peaks essentially correspond to the scattering from the solid monolayer alone. The X-ray pattern was collected at 260 K and the neutron pattern at 10 K. The variation in the “background” at low angles corresponds to the imperfect subtraction of the small-angle scattering (Porod scattering) from the graphite crystallites. Results are presented in q (4π sin θ/λ, where θ is half the scattering angle and λ the radiation wavelength) to allow direct comparison with neutron diffraction data, collected at a different wavelength. Figure 3 clearly illustrates diffraction peaks with the characteristic sawtooth line shape typical of 2D monolayers for both these samples, clearly confirming the formation of solid monolayers of these materials at this coverage and temperature. This is in good agreement with the DSC data presented above. Figure 3a,b shows a 2D peak in the samples at roughly 0.13 nm-1. This peak corresponds to a d spacing of 0.479 nm and is related to the side-to-side separation of the alkyl chains in the lattice. A peak is evident at low q in the neutron pattern, Figure 3b, indicating that this molecule lies flat on the graphite. There are no similar large low-angle reflections for dodecanal in the X-ray pattern, although there is a small-angle peak at approximately q ) 0.71 Å-1, which is approximately double that of the lowest order reflection in the neutron pattern. However, this and related weak features are very close to the noise and should be treated

6030


Figure 4. Temperature variation of the diffracted intensity at 3 equivalent monolayers coverage of dodecanal on graphite (the lowest temperature is the top pattern).

with caution. In addition, the peak at low q in the neutron pattern are also difficult to distinguish against the intense background Porod scattering and the difficulty in resolving low q angles close to the straight-through beam and should also be considered with some care. In comparing the data from these two radiation sources, angular calibration was performed on the separate instruments by use of standard samples. In this work, the position of the graphite (002) reflection also provides an internal standard (at 300 K (002) should appear at q ) 1.8744 Å-1, d ) 3.35 Å). With these adjustments, the positions of the two large monolayer peaks at approximately q ) 1.32 Å -1 agree to within 0.02 Å-1. To facilitate the fitting process, the X-ray pattern has been shifted in q by 0.02 Å-1 to enable these peaks to coincide. At high q, this represents a very small change in lattice parameter (ca. 0.07 Å); however, at the lowest q values this shift can represent a significant variation (ca. 1.0 Å). The variation in the monolayer peak with temperature was also studied. The peak at approximately q ) 1.3 Å-1 moves to a lower angle with increasing temperature, as the layer apparently expands in the b direction, the short axis of the unit cell. The rate of expansion is found to be 7 × 10-4 Å/K. The variation in the a parameter could not be reliably extracted. The monolayer peak begins to disappear as the layer melts at a temperature of approximately 270 K at this coverage. This is higher than expected for many physisorbed monolayers, which melt at approximately 75% of the bulk melting point. It is to be noted that the neutron diffraction monolayer peak at 0.13 Å-1 is broader than calculated, the low-angle reflection in the neutron pattern, and the equivalent X-ray reflection. The neutron peaks are usually a little broader than the X-ray peaks due to the instrument resolution, which has been included in the calculations. There is some evidence that the threedimensional nature of the monolayer, and/or bilayer, formation can produce similar broadened peak shapes. 3.3. X-ray Diffraction: High Coverage. The temperaturedependent diffraction patterns from approximately 3 equivalent monolayers of dodecanal adsorbed on graphite are presented in Figure 4. The diffraction pattern with an excess of dodecanal

Phillips et al.

Figure 5. Incoherent elastic scattering from 3, 6, and 10 monolayers of dodecanal adsorbed on graphite as a function of temperature compared to the pure aldehyde. Insert shows the region of solid monolayer and bulk fluid coexistence.

has also been collected to identify the bulk diffraction peaks. Hence, the data in Figure 4 shows the evolution of the bulk and monolayer peaks with increasing temperature. At the lowest temperatures we can identify relatively intense peaks at q ) 1.36 and 1.42 Å-1, which are from the bulk material. We note that not all bulk peaks are observed at this coverage, possibly due to some preferred orientation of the bulk crystallites or the crystals being too small to fully satisfy the Laue conditions. The peak at q ) 1.33 Å-1 is from the coexisting adsorbed monolayer. As the temperature is raised, the bulk peaks disappear (at 278-280 K) to form a broad peak typical of liquid scattering but the weak monolayer peak at 1.33 Å-1 is still evident, characterized by the sharp attack of the reflection at the low q side. This peak does not disappear until a temperature of 315-325 K when the monolayer melts. These findings are in excellent agreement with the DSC results above (bulk melting point of 280 K and monolayer melting point at 317 K) and clearly support the formation of a solid monolayer coexisting with the bulk liquid aldehyde. There is some evidence that the monolayer is slightly compressed on increasing the coverage from submonolayer. For example, the low-angle peak in the neutron pattern moves by approximately 0.01 Å-1 to higher q on increasing the coverage and the higher angle peak in the X-ray data moves from 1.30 to 1.33 Å-1. However, these are small changes in already weak features, in the presence of significant bulk scattering, and hence these conclusions should be considered tentative. 3.4. Incoherent Elastic Scattering. Figure 5 presents the variation in incoherent elastic scattering from 3, 6, and 10 monolayers of dodecanal on graphite with increasing temperature. The scattering from the bare graphite has been subtracted in preparing this figure. It also includes the scattering from a sample of pure aldehyde, “bulk”. The data in Figure 5 show the behavior expected for the coexistence of a solid monolayer and the liquid adsorbate. At low temperatures where all adsorbate is solid, the elastic intensity is high. Above the bulk melting point the scattering intensity falls sharply as the bulk material melts to form a liquid. However, the scattering does not fall to zero because of the presence of the solid monolayer, which is in good agreement with the DSC and X-ray data


Figure 6. STM image of dodecanal adsorbed from neat liquid on a graphite surface at room temperature. Tunneling conditions: Iset ) 0.161 nA, Vbias ) 0.85 V. The main symmetry axes of graphite are indicated in the lower left corner of the STM image.

presented above. The intensity only falls to zero when the solid monolayer melts. A key indication is the temperature region where the scattering from the pure bulk sharply falls below the scattering of the aldehyde with the graphite, highlighted in the inset. The transition temperatures are somewhat higher than those observed in the DSC and X-ray data. This is attributed to the nature of the dynamics in this system. The incoherent experiment measures the point at which the motion of the protonated species is faster than the time scale of the experiment, and not the melting point specifically. In the solid regions, bulk and monolayer, the elastic intensity also falls slightly with increasing temperature. This is attributed to the increased thermal motion of the atoms in the crystal lattice, a Debye-Waller factor. 3.5. STM. Figure 6 shows an STM image of dodecanal adsorbed on HOPG at room temperature. Dodecanal molecules were adsorbed on HOPG from the neat liquid. The molecules are ordered in lamellae, which run parallel to each other and are separated by narrow troughs. The width of the lamellae is 1.70 ( 0.1 nm, which is in good agreement with the length of the molecules (∼1.60 nm), indicating that the molecules are adsorbed with the carbon backbone parallel to the surface, which is in excellent agreement with the diffraction results presented above. The lamellae run at an angle of 26.4° ( 0.8° with respect to one of the underlying symmetry axes of graphite. The alkyl chains are oriented at 90° ( 1.2° with respect to the lamellar axis and also along one of the main graphite symmetry axes. This arrangement of molecules is similar to the one that has been observed for alkanes and carboxylic acids adsorbed on graphite. The intermolecular distance between the molecules along the lamella axis is 0.45 ( 0.02 nm. The aldehyde functionality in the molecule is not directly visualized by STM. Because of the lack of contrast between the aldehyde functional group and the rest of the molecule, the position of the aldehyde function could not be ascertained unambiguously. The influence of the substrate lattice on the STM measurements is reflected in the contrast of the adsorbed aldehyde monolayer. The STM image displayed in Figure 6 clearly shows a contrast modulation along the lamella axis and perpendicular


Figure 7. Correlation averaged STM image of dodecanal adsorbed from neat liquid on a graphite surface at room temperature. Tunneling conditions: Iset ) 0.161 nA, Vbias ) 0.85 V. The yellow boxes indicate separate molecules with the long axis of the box parallel to the molecular backbone. It is evident that the molecular backbone is not always collinear in adjacent lamellae.

to the molecular axis. This modulation is attributed to the near but not perfect commensurate packing of the aldehyde molecules with respect to the underlying graphite lattice along the lamella axis. This minor mismatch creates an interference effect between the carbon backbones of the molecules and the underlying substrate lattice, giving rise to a Moire´ pattern along the lamella axis.47 Typically, the period of the contrast modulation in the present case is ∼5-6 molecules (2.93 ( 0.1 nm). This implies that every fifth or sixth molecule is adsorbed on a site that is equivalent to that of the first molecule in the Moire´ period (∆M). As already mentioned, the alkyl chains of the molecules run parallel to one of the main graphite axes. On the basis of this information, the Moire´ pattern can be analyzed and it allows us to determine the intermolecular distances with reasonable accuracy. However, in the present case the exact period of contrast variation (5 or 6 molecules per period) could not be established unambiguously due to lack of good contrast in the STM image. Nevertheless, if we consider the intermolecular distance obtained from the direct analysis of the STM image (0.45 ( 0.02 nm), a value of ∼4.62 Å obtained from the Moire´ period (∆M corresponding to 6 molecules) appears to be reasonable. The observation of Moire´ pattern along the lamellar axis is an indication that the bulky head groups (the aldehyde functions) of the molecules require relatively more space when they lie on the graphite surface,48,49 as compared to those of the corresponding alkanes wherein the Moire´ pattern appears along the molecular axis.47 A close inspection of the STM image reveals that although it is not straightforward to observe the aldehyde function of the molecule due to lack of image contrast, one can still discern between individual molecular backbones. Figure 7 displays a correlation averaged STM image of dodecanal on graphite at higher resolution. The fundamental molecular unit in the assembly is indicated by the yellow boxes. It is evident from the figure that there appears to be no regular pattern of molecular arrangement in terms of the alignment of the alkyl chain

6032


Figure 8. Crystal structure determined for dodecanal adsorbed on graphite at submonolayer coverage, showing the unit cell.

backbones of the molecules in the adjacent lamellae. The molecular backbones are collinear in the case of most of the adjacent lamellae whereas in a few other cases they appear to be shifted slightly. However, this conjecture should be considered with some care as it is not straightforward to define the fundamental molecular unit in these STM images. 4. Discussion The X-ray and neutron diffraction patterns at low coverage can be simultaneously fitted to a single structural model that closely resembles the structure deduced from the STM data, obtained at high coverages. The calculated patterns are presented in Figure 3a,b and the structure schematically illustrated in Figure 8. In this fitting procedure, the structural solution has been as highly constrained as possible, using the fewest fitted variables where possible in combination with the bond lengths and bond angles taken from the bulk. The low-angle diffraction peaks allow the a parameter of the unit cell to be determined as approximately 17.55 Å, in good agreement with the interlamellae spacing deduced from the STM image. The value is somewhat larger than the STM estimate, reflecting the compression in the unit cell on increasing the coverage and the uncertainty in the absolute values of the d spacings from the monolayer diffraction patterns. The remaining reflections can be indexed on a rectangular unit cell a ) 17.55 Å, b ) 4.83 Å, and V ) 90°, with an area per molecule of 84.77 Å2. The area per molecule for the equivalent, C12 alkane, dodecane is 73.36 Å2, reflecting the increase in area of 11.4 Å2 on replacing a hydrogen atom by the larger oxygen atom. This difference is somewhat larger than expected based on the projected areas of spherical hydrogen and oxygen atoms (with van de Waals radii50,51 of 1.1 and 1.52 Å, respectively) of 3.45 Å2. A unit cell of these dimensions can contain only a single molecule, a primitive unit cell. With the molecular bond lengths and angles structure constrained to their bulk values, there are only two variables to be fitted in this model: (i) The orientation of the molecule about its long axis; (ii) the rotation of the molecule about the normal to the surface plane. The scattering patterns are relatively sensitive to the inclination of the molecular axis to the a axis of the unit cell and we conclude that the molecules are parallel to the a axis to less than (5°. The patterns are relatively insensitive to the rotation of the molecules about their long axis ( 40° and are considered to lie with the plane of the C-C-C skeleton parallel to the graphite. This conjecture is also substantiated by the STM image of dodecanal, which clearly shows double rows of bright spots that run parallel to each other. These features can be attributed to the protruding hydrogen atoms of the alkyl chains since the electron density is primarily dominated by the protruding hydrogen atoms.52 It has been shown recently that the intensity modulation observed in the case of the substituted and unsubstituted alkanes is a result of electronic effects and not due to

Phillips et al. tilting of the alkane backbone with respect to the graphite substrate.53 Moreover, the calculated distance between adjacent alkyl chains when the molecules are lying parallel to the graphite surface is ∼4.4 Å, which is in reasonable agreement with the distance obtained in the present study (∼4.6 Å). The corresponding distance between the perpendicularly oriented chains is ∼3.8 Å, which rules out the possibility of perpendicular orientation.53 Thus, on the basis of these arguments, it is not unreasonable to assume that the aldehyde molecules lie with the C-C-C alkyl skeleton parallel to the graphite surface. The structures in Figure 8 have the CdO carbonyl dipoles aligned on the surface. Elementary calculations54 indicate that this arrangement is enthalpically favorable with an interdipole attraction of approximately -5.3 kJ/mol. Dipoles separated by the unit cell in the a direction have a much larger separation and will have 50 times weaker interaction. In addition, the relative orientation of the dipoles leads to an unfavorable enthalpy, approximately +22 J/mol. Heats of adsorption of molecular species like these are usually similar to the heat of vaporization, which is 68.3 kJ/mol for dodecanal.55 Hence, we conclude that this dipolar interaction is expected to be a significant contribution to the monolayer stability and hence exert an influence on the monolayer structure. In this structural analysis we have considered the simplest structural solution that is in accord with the diffraction data, and in reasonable agreement with STM observations. Significantly, both X-ray and neutron patterns have been successfully fitted even though they have quite different characteristics. However, the resulting structure does not appear to be physically reasonable, as illustrated in Figure 8, mainly due to the lines of molecules that are formed in this arrangement. Usually molecules favor low-energy close-packed and six-coordinate arrangements in a 2D layer.56,57 The structure in Figure 8 is only four-coordinate. In addition, there seems to be an unfavorable overlap of the molecules with each other. Hence, we have also considered more complicated monolayer structures that are also found to fit the experimental data reasonably well, in particular, where there is more than one molecule in a larger unit cell. There are several possible molecular symmetries possible with two molecules per unit cell: (i) centered cell, (ii) a single glide line in the a direction, Pg(a), and (iii) single glide line in the b direction Pg(b). The Pgg plane group with four molecules per unit cell is also considered. With aldehyde molecules rotated such that the plane of the carbonyl group is perpendicular to the graphite surface, the plane centered plane group becomes Cm and plane groups (ii) Pg(a) and (iii) Pg(b) become Pgm and Pmg, respectively. All of the monolayer structures in Figure 9 with the CdO group lying in the plane of the graphite have very similar favorable dipolar interactions arising from the interaction in the b direction of the unit cells, -5.3 kJ/mol. There are other interactions in the a direction but these dipoles are well separated and the interactions are much weaker (by a factor of 50). Those structures where all CdO groups are parallel and “upright” pointing up from the surface have strongly unfavorable interactions. There is the possibility that these upright arrangements consist of alternating “up” and “down” dipoles that would be more favorable. However, the diffraction data do not allow us to evaluate these options. Besides this, as discussed earlier, the STM data also rule out the possibility of the upright orientation. The only structures with significantly additional favorable interactions are the Pg(b) and Pgg. It is the interaction between dipoles along the glide axis separated by 1.5 unit cells that significantly enhances (-1.33 kJ/mol) these arrangements over



Figure 9. Other possible monolayer structures which are compatible with the experimental diffraction data and STM image. Unit cells are indicated on the schematic structures. The unit cells employed for these calculations were a ) 35.0, b ) 4.83, and V ) 90 for the unit cells containing 2 molecules and a ) 70.0, b ) 4.83, and V ) 90 for the unit cell containing 4 molecules.

the others considered. In the Pg(b) arrangement the CdO dipoles form a zigzag chain with an overall dipole moment in the b direction. Hence, the Pg(b) layer will have an overall dipole moment. The Pgg arrangement is slightly lower in energy than the Pg(b) arrangement, as alternate zigzag chains point in opposite directions and the layer will not have an overall dipole moment. On the basis of these very elementary calculations, the Pg(b) or Pgg plane group structures are considered to be the preferred arrangements. Fine adjustment of the model for the Pgg structure shows that the closest agreement with the experimental pattern occurs when the carbonyl groups are as far apart as possible without overlapping the ends of the alkyl chain. The model shows no difference in the diffraction pattern with the adjacent carbonyls in the same direction or opposite direction, but the dipole energy favors having neighboring dipoles in the same orientation; see Figure 10. Nevertheless, the STM data reveal that these molecular arrangements do not prevail exclusively on the surface of graphite and there is a significant contribution from structures in which the molecules are aligned in collinear fashion in adjacent lamellae.

Figure 10. Molecular arrangement with neighboring dipoles in the same orientation.

5. Conclusions A combination of DSC, X-ray and neutron diffraction, neutron scattering, and STM have been used to demonstrate the formation of solid monolayers of alkyl aldehydes on graphite at submonolayer coverage and at multilayer coverage. The layers at high coverage are stable at temperatures above the bulk melting point, similar to other related alkanes. The shortest aldehyde to show such behavior is C8. Analysis of the low coverage diffraction and high coverage STM data indicate the molecules lie flat on the graphite surface parallel to each other

6034


TABLE 1: Approximate Dipole-Dipole Energies for Different Possible Structures of Dodecanal Monolayers structure

dipolar energy per molecule (J/mol)

primitive, P1 centered, C Pg(b) Pg(a) Pgg

-5255 -5262 -6540 -5163 -6540 (see text for discussion)

“upright” structures P1, Cm, Pmg, and Pgm

see text for discussion

and the unit cell a axis. Even with this structural information no conclusions can be drawn on the details of the monolayer structure. However, elementary calculations indicate that the Pg(b) [Figure 9d] or Pgg [Figure 9g and 10] plane group structures are most likely as they favor effective dipole-dipole interactions. Acknowledgment. We thank EPSRC (TP) and the Cambridge Commonwealth Trust (TB) for funding, the BP Institute and the University of Cambridge for financial support and facilities, and the PSI and ILL staff and scientists Antonio Cervellino, Fabia Gozzo, Thomas Hansen, Jacques Torregrossa, and Tilo Seydel for beam time and technical support. This work is based on experiments performed at the Swiss Light Source, Paul Scherrer Institute, Villigen, Switzerland. The research leading to these results has received funding from the European Community’s Seventh Framework Programme (FP7/2007-2013) under grant agreement no. 226716. K.S.M. thanks the F.W.O.Flanders (Belgium) for a visiting postdoctoral fellowship. References and Notes (1) Gilbert, E. P.; White, J. W.; Senden, T. J. Chem. Phys. Lett. 1994, 227, 443. (2) Arnold, T.; Cook, R. E.; Chanaa, S.; Clarke, S. M.; Farinelli, M.; Yaron, P.; Larese, J. Z. Physica B 2006, 385-386, 205. (3) Yang, T.; Berber, S.; Liu, J.-F.; Miller, G. P.; Tomanek, D. J. Chem. Phys. 2008, 128, 124709. (4) Larese, J. Z.; Passell, L.; Heidemann, A. D.; Richter, D.; Wicksted, J. P. Phys. ReV. Lett. 1988, 61, 432. (5) Arnold, T.; Chanaa, S.; Clarke, S. M.; Cook, R. E.; Larese, J. Z. Phys. ReV. B 2006, 74, 085421. (6) Arnold, T.; Thomas, R. K.; Castro, M. A.; Clarke, S. M.; Messe, L.; Inaba, A. Phys. Chem. Chem. Phys. 2002, 4, 345. (7) Arnold, T.; Dong, C. C.; Thomas, R. K.; Castro, M. A.; Perdigon, A.; Clarke, S. M.; Inaba, A. Phys. Chem. Chem. Phys. 2002, 4, 3430. (8) Morishige, K.; Takami, Y.; Yokota, Y. Phys. ReV. B 1993, 48, 8277. (9) Buchholz, S.; Rabe, J. P. Angew. Chem., Int. Ed. Engl. 1992, 31, 189. (10) Cyr, D. M.; Venkataraman, B.; Flynn, G. W. Chem. Mater. 1996, 8, 1600. (11) Morishige, K. J. Chem. Phys. 1994, 100, 3252. (12) Gunning, A. P.; Kirby, A. R.; Mallard, X.; Morris, V. J. J. Chem. Soc., Faraday Trans. 1994, 90, 2551. (13) Clarke, S. M.; Messe, L.; Whitehead, C.; Inaba, A.; Thomas, R.; Arnold, T. Appl. Phys., A 2002, 74, 1072. (14) Messe, L.; Perdigon, A.; Clarke, S. M.; Castro, M. A.; Inaba, A. J. Colloid Interface Sci. 2003, 266, 19. (15) Morishige, K.; Sakamoto, Y. J. Chem. Phys. 1995, 103, 2354. (16) Duim, W. C.; Clarke, S. M. J. Phys. Chem. B 2006, 110, 23853. (17) Tao, F.; Goswami, J.; Bernasek, S. L. J. Phys. Chem. B 2006, 110, 19562. (18) Bickerstaffe, A.; Messe, L.; Clarke, S. M.; Parker, J.; Perdigon, A.; Cheah, N. P.; Inaba, A. Phys. Chem. Chem. Phys. 2004, 6, 3545. (19) Bickerstaffe, A. K.; Cheah, N. P.; Clarke, S. M.; Parker, J. E.; Perdigon, A.; Messe, L.; Inaba, A. J. Phys. Chem. B 2006, 110, 5570. (20) Arnold, T.; Clarke, S. M. Langmuir 2008, 24, 3325. (21) Muenter, J. S. J. Mol. Spectrosc. 1975, 55, 490.

Phillips et al. (22) CRC Handbook of Chemistry and Physics, 81st ed.; Lide, D. R., Ed.; CRC Press: Boca Raton, FL, 2000. (23) You, H.; Fain, S. C. J. Surf. Sci. 1985, 151, 361. (24) Castro, M. A.; Clarke, S. M.; Inaba, A.; Perdigon, A.; Prestidge, A.; Thomas, R. K. Stud. Surf. Sci. Catal. 2001, 132, 873. (25) Espeau, P.; White, J. W.; Papoular, R. J. Appl. Surf. Sci. 2005, 252, 1350. (26) Clarke, S. M.; Inaba, A.; Arnold, T.; Thomas, R. K. J. Therm. Anal. Calorim. 1999, 57, 643. (27) Castro, M. A.; Clarke, S. M.; Inaba, A.; Arnold, T.; Thomas, R. K. Phys. Chem. Chem. Phys. 1999, 1, 5017. (28) Espeau, P.; White, J. W. Carbon 2005, 43, 1885. (29) Hansen, T. C.; Henry, P. F.; Fischer, H. E.; Torregrossa, J.; Convert, P. The D20 instrument at the ILL: a Versatile high-intensity two-axis neutron diffractometer, 4th European Conference on Neutron Scattering, Lund, Sweden, 2007. (30) Clarke, S. M. Curr. Opin. Colloid Interface Sci. 2001, 6, 118. (31) Castro, M. A.; Clarke, S. M.; Inaba, A.; Thomas, R. K. J. Phys. Chem. B 1997, 101, 8878. (32) Inaba, A. Pure Appl. Chem. 2006, 78, 1025. (33) Castro, M. A.; Clarke, S. M.; Inaba, A.; Thomas, R. K. Physica B 1998, 241-243, 1086. (34) Ana, M. G.; Ricardo, A. M. J. Chem. Phys. 1999, 110, 6606. (35) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Rob, M. A.; Cheeseman, J. R.; Jr.; , J. A. M.; T. Vreven, K. N. K.; J. C. Burant, J. M. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; *Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; J. V. Ortiz, Q. C.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A., Gaussian 03; Gaussian, Inc.: Wallingford, CT, 2003. (36) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (37) Petersson, G. A.; Andrew, B.; Thomas, G. T.; Mohammad, A. A.L.; William, A. S.; John, M. J. Chem. Phys. 1988, 89, 2193. (38) Petersson, G. A.; Mohammad, A. A.-L. J. Chem. Phys. 1991, 94, 6081. (39) Espeau, P.; White, J. W. J. Chem. Soc., Faraday Trans. 1997, 93, 3197. (40) Rabe, J. P. In 6th International Conf. on Scanning Tunneling Microscopy (Stm 91); Elsevier Science Bv: Interlaken, Switzerland, 1991; p 41. (41) Taub, H. Neutron Scattering Studies of the Structure, Phase Transitions and Dynamics of Commensurate Herringbone Monolayers Physisorbed on Graphite. In The Time Domain in Surface and Structural Dynamics; Long, G. J., Grandjean, F., Eds., 1988; Vol. 228. (42) Morishige, K.; Kato, T. J. Chem. Phys. 1999, 111, 7095. (43) Cheah, N. P.; Messe, L.; Clarke, S. M. J. Phys. Chem. B 2004, 108, 4466. (44) Parker, J. E.; Clarke, S. M.; Perdigon, A. C. Surf. Sci. 2007, 601, 4149. (45) Morishige, K.; Kobayashi, N. J. Chem. Phys. 1994, 101, 5209. (46) Castro, M. A.; Clarke, S. M.; Inaba, A.; Arnold, T.; Thomas, R. K. J. Phys. Chem. B 1998, 102, 10528. (47) Rabe, J. P.; Buchholz, S. Science 1991, 253, 424. (48) Batra, I. P.; Ciraci, S. J. Vac. Sci. Technol. A-Vac. Surf. Films 1988, 6, 313. (49) Hibino, M.; Sumi, A.; Hatta, I. Jpn. J. Appl. Phys. Part 1 1995, 34, 3354. (50) Rowland, R. S.; Taylor, R. J. Phys. Chem. 1996, 100, 7384. (51) Bondi, A. J. Phys. Chem. 1964, 68, 441. (52) Liang, W.; Whangbo, M. H.; Wawkuschewski, A.; Cantow, H. J.; Magonov, S. N. AdV. Mater. 1993, 5, 817. (53) Ilan, B.; Florio, G. M.; Hybertsen, M. S.; Berne, B. J.; Flynn, G. W. Nano Lett. 2008, 8, 3160. (54) Israelachvili, J. N. Intermolecular and Surface Forces; Academic Press, 1991. (55) Verevkin, S. P.; Krasnykh, E. L.; Vasiltsova, T. V.; Koutek, B.; Doubsky, J.; Heintz, A. Fluid Phase Equilib. 2003, 206, 331. (56) Kitaigorodski, A. I. Organic chemical crystallography; Consultants Bureau: New York, 1961. (57) Clarke, S. M.; Thomas, R. K. Mol. Phys. 1991, 72, 413.

JP911069T

Adsorption of Aldehydes on a Graphite Substrate - American Chemical

Recommend Documents