J. CHEM. SOC. FARADAY TRANS., 1994, 90(7), 969-972 Formation and Structure of Langmuir-Blodgett Films of C,, and Arachidic Acid Ciaran Ewins and Brian Stewart Department of Chemistry and Chemical Engineering, University of Paisley, High Street, Paisley, UK PA I 2BE ~~ ~~ ~~ Langmuir films have been prepared using pure c60 and arachidic acid mixed with c60 in the ratio of 1 : 1 and 4.2 :1. Langmuir-Blodgett (LB) films were then deposited on a variety of substrates with Y-type deposition being observed. Films containing bilayers of arachidic acid alternating with both single and double monolayers of C,, were prepared. The structure of the arachidic acid component in the mixed films was investigated by grazing- incidence FTIR spectroscopy and was shown to be similar to that in an LB film of pure arachidic acid.The absence of structural disruption in the presence of C,, provides strong evidence for an ordered alternant arrangement of fullerenes and arachidic acid in the films. LB films incorporating c60 have been reported by several groups. c60 is one of an increasing number of new com- pounds that form LB films but are not amphiphiles with hydrophobic and hydrophilic parts to the molecule. It has been found that C,, forms stable monolayerl and multi-layer films' at an air/water interface but that these films do not transfer well to solid substrates. Transfer promoters such as arachidic acid' (CH3C18H36C02H) and stearyl alcohol3 (C,,H,,OH) have been used to aid deposition of good quality films. Transfer promoters have in the past been suc- cessfully applied to the deposition of other materials such as phthal~cyanines.~ Different values for the area per molecule (A,) have been reported for LB films of c60 ranging from 32 A2 to 96 A2.The theoretical A, calculated' for an ideal close-packed monolayer is 92 A'. The lower observed values have been attributed to the formation of floating multilayer films. The chosen solvent and the concentration of the spreading solu- tion seem to have an effect on this value.2 The transfer of these films to glass has given variable results, with no transfer' and 60% Z-type deposition2 being reported. The theoretical ratio of A, values for arachidic acid : c60 is 4.2 : 1. The use of a solution containing arachidic acid and c60 in this mole ratio is expected to produce Langmuir films in which both components occupy the same total area.The transfer ratio for such a mixture is improved to almost The interesting point about these films is that the value for A, corresponds to that expected for the arachidic acid alone. It has been suggested that this is due to the c60 being 'squeezed out' of the arachidic acid monolayer during compression to sit on top of the alkyl chains. Langmuir films of this type have been deposited in a Y-type manner, suggest- ing an alternant bilayer structure of arachidic acid and c60 [Fig. l(a)]. Some evidence for such a film structure has come from film thickness measurements. In the present work we have investigated the structure of the alternate bilayer films using grazing incidence (GI) FTIR spectroscopy.In addition we have prepared an LB film con- taining monolayers of c60 sandwiched between bilayers of arachidic acid [Fig. l(b)]. This was achieved using an alter- nate layer Langmuir trough, with separate films of arachidic acid and 4.2 :1 arachidic acid : c6,. The proposed structure is supported by thickness measurements, UV-VIS absorb-ance and GI FTIR spectroscopy. GI FTIR Spectroscopy GI FTIR has been used extensively to study LB films and can provide information about the orientation of molecules rela- tive to a metal surface.6 The success of this technique relies on the fact that the component of the IR radiation that is polarised perpendicular (p-polarised) to the metal substrate gives an enhanced absorption, while the component polarised parallel (s-polarisation) to the surface gives no measurable absorption, due to a 180" phase change upon reflection causing destructive interference.The polarisation is con-trolled by a polariser and the angle of incidence is usually 80" (also giving greater pathlength and thus sensitivity). For a fatty acid molecular chain perpendicular to a surface, the transition dipole of the methyl symmetric stretch v, (CH,) is nearly perpendicular to the surface and thus absorbs p- polarised light whereas the symmetric and asymmetric stretches of the methylenes v, (CH,) and v, (CH,) are parallel to the surface and so will not absorb p-polarised light.This allows the intensity of the CH, absorption to be measured in two different polarisations (transmission for parallel and GI for perpendicular) and hence the orientation of the alkyl chains can be calculated., Arachidic acid has been studied by this technique and found to be tilted at an angle, 0 = 25" to the surface normal.8 GI FTIR can also be used alone in a more qualitative manner to assess the order and orientation of an LB film by comparison with films of known structure. Experimental The LB trough used was a NIMA alternate layer troughg located in a class lo00 clean room. The trough is made of PTFE and the surface pressure was measured by a Wilhelmy balance. The toluene used was scintillation grade, the arachi- dic acid was purchased from Aldrich (99%)and was used as 000000 000000 I I I alternant bilayer film CG0monolayer film Fig.1 (a) Proposed structure of 4.2 :1 arachidic acid-C,, LB film. (b)Proposed structure of arachidic acid<,, monolayer film. received. The c60 was prepared in this laboratory and puri- fied by column chromatography over silica/graphite" fol-lowed by recrystallisation from cyclohexane. Hydrophobic quartz slides were prepared by treating Spectrosil B (Thermal Syndicates) with hexamethyldisilazene. Aluminium-coated glass slides were prepared in an RF sputter coater. The water was purified on a deionisation-reverse osmosis-UV sterilisa- tion system (Elgastat/UHP) and had a resistivity of 18 MR cm-'.The subphase water pH was 5.8 and the temperature was 22°C.Typical concentrations of the solutions used were: c60, 0.35 mg cm-3 (4.86 x mol dm-3) and arachidic acid, 1.1 mg cm-3 (3.52 x mol dm-3) in toluene. To prepare the LB films, the solutions were mixed and spread on the water surface uia micro-syringe. The solvent was allowed to evaporate for 20 min and the floating films were then com- pressed at a rate of between 40 and 70 A2 molecule-'min-'. Langmuir films were prepared from pure c60 and from the mixed solutions of molar ratios, 1 :1 c60 : arachidic acid and 4.2 : 1 arachidic acid : c60. The floating films were transferred at a surface pressure, n = 20 mN m-', for the pure c60 and n: = 30 mN m-' for the mixed films. To prepare the monolayer c60 LB films a two-compartment alternate layer trough was used.Solutions of arachidic acid and of 4.2 :1 arachidic acid :c60 were spread in compartments A and B, respectively. Films were deposited by alternately passing a cleaned slide down through A and then up through B until the required number of c60 mono-layers were deposited. The films produced were studied by UV-VIS spectroscopy and GI FTIR. The spectrometers used were a Perkin-Elmer Lamda 9 spectrophotometer for transmission UV-VIS spec-troscopy and a BIO-RAD FTS 40 at 2 cm-' resolution for GI FTIR. Film thickness measurements were carried out with a Watson-Barnett interferometer fitted to a Nikon Optiphot Microscope. This method assesses film thickness by measur- ing the shift in interference fringes across a step in film height.To aid this the LB films were coated with a thin film of vacuum-evaporated silver. Results and Discussion Cs0 Film C,, appears to form a compressed monolayer at the air/ water interface under the conditions stated. The isotherm [Fig. 2(a)] gives A, = 90 A2. This is obtained by extrapo- lating the linear part of the plot back to zero pressure and compares well with the calculated value of 92 A*. The float- ing film is brittle and does not transfer well, giving low trans- fer ratios and an uneven film on glass. 1 :1 Molar Ratio C,,-Arachidic Acid Film The 1 :1 C60-araChidiC acid film gave the isotherm shown in Fig. 2(b). Extrapolating back to zero pressure gives A, x 60 A', which is close to the calculated value.Since the isotherm is not that of two immiscible liquids'' this suggests that the C60 and arachidic acid are mixed in a monolayer film with the arachidic acid uniformly dispersed throughout the film. The presence of the transfer-promoter arachidic acid allowed the deposition of good quality films onto aluminium-coated slides. The transfer ratios were approximately 100% but the deposition was of Y-type, i.e. deposition on the upstroke and downstroke. GI FTIR spectroscopy of 20 layers on an aluminium-coated glass slide [Fig. 3(a)] confirms the presence of the c60 J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 8ol70 7 60-E z E 50-1s? 40-2?;30-z;20-10 -n, 10 30 50 70 90 110 130 A, fA2 Fig.2 (a) Isotherm of C,, on ultrapure water. (b) Isotherm of 1 : 1 arachidic acid-C,, .(c) Isotherm of 4.2 : 1 arachidic acid<,, . in the film with all fullerene bands occurring at the same wavenumbers as in a pure sample.12 Fig. 4(a) shows the C-H stretch region. The intensities of the v, and v,(CH,) bands (transition moments perpendicular to the chain axis) are greater than would be expected in a pure arachidic acid film of the same number of layers (where 0 = 25"), suggesting greater tilt and/or disorder. The v,(CH,) bands are weak and the v,(CH,) band is very weak. This indicates disorder of the 0.08 Q, 0.06 I t I I I I I I 1 1800 1600 1400 1200 1000 800 600 400 wavenumber/cm-' Fig. 3 GI FTIR absorption: (a) 1 :1 arachidic acid-C,,; (b) arachi-dic acid-€,, monolayer LB film; (c) pure arachidic acid; (d) 4.2 : 1 arachidic acid-c,, all on aluminium 0.07O.O81 0.03--3100 3d50 3000 2$50 2900 2850 2800 27150 2700 wavenumber/cm-' Fig.4 GI FTIR absorption in the C-H stretch region for: (a) 1 : 1 arachidic acid<,,; (b) arachidic acid<,, monolayer LB film; (c) arachidic acid LB film; (d) 4.2: 1 arachidic acid-C,, all on alu-minium J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 chain ends which is expected to result from methyl-fullerene packing. The C=O stretch is only observed at 1707 cm-', indica-tive of 'facing' dimers in the carboxylic acid groups. Facing dimers are the norm for a pure arachidic acid LB film.There is a very weak band at 3020 cm-' due to aromatic C-H stretching. This is most likely to be toluene solvent trapped in the film. 4.2 :1 Molar Ratio Film The 4.2 :1 arachidic acid-C,, mixed film gives the isotherm shown in Fig. 2(c). It is identical to that previously reported' and confirms that the total area occupied is the same as that for arachidic acid alone. The film is very stable, the total area only falling by 1.5% over 2 h at 30 mN m-'. Aluminium, silicon and quartz substrates gave Y-type deposition. The presence of the c60 was confirmed by transmission UV-VIS spectroscopy and GI FTIR. The GI FTIR spectrum [Fig. 3(d)] shows the four c60 bands in the expected positions. The c60 absorbance at 1428 cm-' is somewhat obscured by the v,(CO,) at 1429 cm-'.Only one type of carbonyl is visible. This is the 'facing dimer' stretch at 1706 cm-'. Fig. 4(d) shows the C-H region for the mixed film and Fig. 4(c) the same region for a film of pure arachidic acid prepared under identical conditions. The overall similarity in the shape and relative intensities of the two spectra in the C-H region shows that the arrangement of the arachidic acid is little different from that in a pure arachidic acid film. This is only possible if the film is indeed composed of alternating double layers of c60 and arachidic acid with carboxylic acid groups facing each other. The two main differences are that the v,(CH,)ip (in-plane) stretch is weaker relative to the out-of-plane stretch [v,(CH,)op] com-pared with the mixed film and that the intensities of all C-H stretching bands are weaker overall in the mixed film.The differences in the v(CH,) in-plane and out-of-plane intensities at 2963 and 2954 cm-' are most likely due to the presence of the c60 on top of the methyl groups. Observed wavenumbers and assignments of vibrational bands are given in Table 1. C,, Monolayer Films LB films containing monomolecular layers of c60 between bilayers of arachidic acid [Fig. l(a)] were prepared with good transfer ratios using the alternate trough method. The first dip was down through a monolayer of arachidic acid and the second dip was then up through the arachidic acid-C,, film. Films were prepared on hydrophobic quartz with 5, 10 and 20 monolayers of C,, by dipping a total of 10, 20 and 40 times, respectively.A plot of absorbance (at 343 nm) us. Table 1 GI FTIR main absorbances of 4.2 :1 arachidic acid :(20-layer LB film on aluminium) _____~ wavenumber/cm - assignment notes 2963 2955 2917 2872 2850 1707 1461-1450 1429 1354-1175 1183 577 528 ~~ ~ Identical band positions (to within +2 cm-') are observed for a 1 : 1 film. 971 0.3 ~ 0.2 0 (0-e2 -", 0.1 0 0 U I I , I I 0 10 20 30 40 number of dips Fig. 5 Absorbance at 341 nm us. the number of dips for an arachi- dic acid-C,, monolayer LB film number of dips (Fig. 5) is approximately linear (gradient = 0.0048 absorbance units per dip) and is evidence of good film quality in the monomolecular layer c60 film.A similar plot for a 4.2:l arachidic acid :C60 LB film (c60 bilayer film) by Williams et aL2 has approximately twice the gradient (0.012),corresponding to the same absorbance per c60 layer. The absorbance of a multilayer system can be used to determine the layer coverage (0)if the absorption cross- section and the layer number density are known: @ = A/ (2.303aN),where A is the decadic absorbance [10g~~(~~/Z)], a is the cross-section in cm2 molecule-' and N is the mono- layer number density in molecules cm-2. The absorption cross-section may be obtained from the molar absorptivity in solution provided that there is no change in the nature of the transition in going to the solid state.For c60, the bands at 343, 268 and 220 nm show no detectable differences in wave- length or bandwidth between solution in toluene or cyclo- hexane and the solid-state film. It therefore seems justified to obtain a value of the solid-state absorption cross-section from a solution molar absorption coefficient (E) measurement. A small solvent effect on E is observed between toluene and cyclohexane. However, toluene solvation is probably a better model for the fullerene close-packing environment. The molar absorption coeficient at 343 nm is found to be (56 316 f3%) dm3 mol-' cm-' for c60 [>99% by high-performance liquid chromatography (HPLC)] in toluene [average of four determinations in the concentration range (1.69-4.15) x 10-mol dm-,].This value is considerably larger, and therefore more reliable, than that previously reported. The absorp- tion cross-section is calculated from the relation : lOOOE c=-2.3O3NA Using our solution molar absorption coefficient, = 9.352 x cm2 molecule-'. Using this, together with the molecular density in the crystalline (1 11) planes;14 Nccp= 1.21 x lo-'' molecule cm-', the calculated monolayer coverages for the three films are 5.02, 10.33 and 23.53 L. This compares well with the ideal values: 5, 10 and 20, perhaps indicating that some multilayering is occurring during the time required for higher numbers of dips. The UV-VIS absorption spectrum of the monolayer c60 film (Fig. 6) is very similar to that of an alternant bilayer film J.CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 Oa5 t 0.4i r! f 0.3I’M 0tI I I 1 I I ,‘@I 200 300 400 500 600 700 800 /lf/nm Fig. 6 UV-VIS spectrum of arachidic acid-C,, (b)bilayer LB films containing 20 layers of C,, (a) m onolayer and with no significant shifts in peak positions being observed. The absorption background at 200 nm is approximately twice that observed in the bilayer films since double the number of arachidic acid layers are present. The thickness of a film produced from a total of 41 dips containing 20 monolayers of c60 is expected to be 126.6 nm, taking the thicknesses of the C60 and arachidic acid mono- layers to be 1.0 and 2.6 nm, respectively. The actual film thickness was measured to be 124 nm by interferometry, offering further good evidence for the proposed structure.GI FTIR was carried out on a film containing 10 mono-layers of C,o and 20 layers of arachidic acid deposited on aluminium [Fig. 3(b)]. The c60 bands are present in their usual positions with the expected intensities. In the C-H stretch region [Fig. qb)]the spectrum is very close to that of a pure arachidic acid LB film [Fig. 4(c)] and is indicative of a well ordered film. Conclusions We have shown that LB films of different c6, :arachidic acid ratios can be deposited. GI FTIR spectroscopy of films deposited on aluminium shows that the arachidic acid chains are oriented in a manner similar to that in an LB film of pure arachidic acid.This result, together with previous film thick- ness measurements, confirms that in the 4.2:1 mixed film the deposited film consists of alternating double layers of arachi-dic acid and c60. The arachidic acid/C6, monolayer films also contain well ordered bilayers of arachidic acid but only single monolayers of c60 which is confirmed by our film thickness and absorbance measurements. We have demonstrated an ability to control the formation of well ordered layers of fullerenes of varying thickness which are insulated from each other by arachidic acid. This offers a valuable tool for probing the effects of dimensionality on the electrical, optical and spectroscopic properties of fullerene films. We would like to thank Denny Wernham for preparing and purifying the 0.References 1 Y. S. Obeng and A. J. Bard, J. Am. Chem. SOC.,1991,113,6279. 2 G. Williams, C. Pearson, M. R. Bryce and M. C. Petty, Thin Solid Films, 1992,209, 150. 3 J. Milliken, D. D. Dominquez, H. H. Nelson and W. R. Barger, Chem. Muter., 1992,4,252. 4 Yansong Fu and A. B. P. Lever, J. Phys. Chem., 1991,956979. 5 J. M. Hawkins, A. Meyer, T. A. Lewis, S. Loren and F. Hollander, Science, 1991,252, 312. 6 An Introduction to Ultrathin Organic Films, ed. A. Ullman, Aca- demic Press, New York, 1991. 7 P. A. Chollet, Thin Solid Films, 1980,68, 13. 8 D. L. Allara and R. G. Nuzzo, Langmuir, 1985,1,45. 9 F. Grunfeld, Rev. Sci. Instrum., 1993,64, 548. 10 W. A. Scrivens, P. V. Bedworth and J. M. Tour, J. Am. Chem. SOC., 1992, 114, 7917. 11 Langmuir Blodgett Films, ed. G. Roberts, Plenum, New York, 1990. 12 P. Bhyrappa, A. Penicaud, M. Kawamoto and C. Reed, J. Chem. SOC.,Chem. Commun., 1992,936. 13 D. R. Haynes, A. Tokmakoff and S. M. George, Chem. Phys. Lett., 1993,214, 50. 14 W. Kratschmer, L. D. Lamb, K. Fostiropoulos and D. R. Huffman, Nature (London), 1990,347,354. Paper 3/07018B; Received 25th November, 1993