Mendeleev Communications Electronic Version, Issue 4, 1998 (pp. 129–168) Saturated vapour pressure and enthalpy of sublimation of C84 Olga V. Boltalina,*a Vitaliy Yu. Markov,a Andrey Ya. Borschevskii,a Vladimir Ya. Davydov,a Lev N. Sidorov,a Valery N. Bezmelnitsin,b Alexander V. Eletskiib and Roger Taylorc a Department of Chemistry, M. V. Lomonosov Moscow State University, 119899 Moscow, Russian Federation.Fax:+7 095 939 1240; e-mail: ovb@thermo.chem.msu.su b Institute of Applied Chemical Physics, Russian Research Centre ‘Kurchatov Institute’, 123182 Moscow, Russian Federation c The Chemistry Laboratory, CPES School, Sussex University, Brighton, BN1 9QJ, UK The vapour pressure and sublimation enthalpy of HPLC-purified C84 have been determined by Knudsen cell mass spectrometry in the temperature range 658–980 K; the temperature–pressure equation is ln(p/Pa) = (–24337±539)/(T/K) + (25.15±0.64), and the mean sublimation enthalpy at 853 K is 202±4 kJ mol–1.There has been considerable interest in the thermodynamic properties of pure C60 and C70. The enthalpies of formation, heat capacity, enthalpies of sublimation and the saturated vapour pressures have been reported.1–7 However, separation and purification of higher fullerenes Cn (n > 70) still remains a complicated and time- and labourconsuming procedure, and this hinders investigations of their physical and chemical properties. Only a few papers have been published on the thermodynamic studies of higher fullerenes and their mixtures.8,9 Here we report our determination of the sublimation enthalpy of pure C84 and partial-mole sublimation enthalpies of C76, C78 and C84 determined from a mixture of higher fullerenes.Sample preparation. Two samples were investigated. Sample 1 was prepared as follows. The crude fullerenes were Soxhletextracted with chloroform from the fullerene-containing soot. The extract was then dissolved in toluene and purified directly by HPLC (without pre-separation of C60 and C70) using a 10 mm×25 cm Cosmosil Buckyprep column, toluene eluent, operated at 4.5 ml min–1 flow rate.The retention time of C84 was 24.95 min (C60 elutes at 7.8 min under the same conditions). The product from the initial separation was recycled (same conditions) to remove traces of other fullerenes present due to tailing during the initial run.The re-purified material showed (HPLC) complete absence of other fullerenes (except that traces of C82 may be present since it is well-known to co-elute with C84 using this column and conditions, which are now the universal standard for separation of this fullerene). Sample 2 was obtained after HPLC removal of C60 and C70 from the fullerene-containing soot extract.Knudsen cell mass spectrometry study. A magnetic sector MI-1201 mass spectrometer (Russia) equipped with a high temperature ion source (electron impact ionization, EI) was used in the sublimation studies; details of the instrumentation are described elsewhere.10 A weighed amount (3–4 mg) of the sample was placed into the first chamber of the twin effusion cell. Both Ni and Pt effusion cells were incorporated in different series, the hole diameters being 0.5 and 0.25 mm, respectively.The evaporation/effusion surface ratios were estimated to be not less than 400 and 100, respectively. Prior to our experiments with the twin cells, they were checked for the equivalence of flows from the two chambers. [60]Fullerene was placed into both chambers, and C+ 60 ion currents were measured and found to be equal. Either [60]fullerene or caesium iodide, used as standards in the experiments with samples 1 and 2, respectively, were placed into the second chamber of the twin cell.The cell was resistively heated and the temperature was measured with a Pt/Pt–Rh thermocouple. The EI mass spectra of samples 1 and 2 are presented in Figures 1 and 2, respectively.The composition of sample 2 was estimated from the HPLC data, assuming equivalence of the extinction coefficients of C76, C78 and C84 (Table 1). The amount of residue after completion of the vapour pressure measurements comprised 10–15% of the starting sample (5–8 mg). The residue dissolved partially in toluene and the solution obtained was analysed by HPLC, which showed a relative increase in C60 and C70 contents compared to the initial composition.This contradicts expectation based on their higher volatilities and arises in part due to the known thermal degradation of higher fullerenes to lower ones, and possibly also from differential solubilities in the toluene- Figure 1 Electron impact mass spectrum of sample 1 at T = 980 K. Mass range 700–1030 amu. 250 200 150 100 50 700 750 800 850 900 950 1000 m/z Intensity (arbitrary units) C+ 60 C+ 82 C+ 84 Figure 2 Electron impact mass spectrum of sample 2 at T = 822 K. Mass range 700–1030 amu. 700 750 800 850 900 950 1000 m/z Intensity (arbitrary units) C+ 60 C+ 70 C+ 76 C+ 78 C+ 82 C+ 84 1.0 0.8 0.6 0.4 0.2 Table 1 Composition of sample 2 before and after evaporation. Fullerene Mole fraction Before evaporation After evaporation C60 0.05 0.10 C70 0.12 0.24 C76 0.19 0.14 C78 0.23 0.16 C84 0.41 0.36Mendeleev Communications Electronic Version, Issue 4, 1998 (pp. 129-168) insoluble carbonaceous residue. This residue, widely observed in fullerene thermolysis, is believed to be composed of degraded or polymerised material. The temperature dependence of C+ 84 ion current for sample 1 was measured in the temperature interval 658–980 K.The partial pressure values of C84 were obtained from comparison of the intensities of C+ 84, and of C+ 60 of the standard, using the known vapour pressure of C60.6 The correction for ionisation cross section was made using the additivity rule.11 The recommended temperature variation of the saturated vapour pressure of C84 in the temperature range 658–980 K, and the corresponding value of the sublimation enthalpy, are presented in Table 2.We consider the pressure equation already mentioned and the sublimation enthalpy to describe the thermodynamic characteristics of pure C84 and disregard the influence of 10% of C82 (always present unless a two-stage HPLC secondary Buckylutcher column is used).For sample 2 the temperature dependence was measured in the temperature range 700–862 K. The partial pressures were determined from the measured ion currents of C+ 84 and Cs2I+ of the standard (CsI) using the known vapour pressure of Cs2I2.12 The correction for ionisation cross sections was made using the additivity rule.11 The temperature dependence of the partial pressure of C84 and the corresponding partial molar enthalpy of sublimation are also presented in Table 2.The error values were taken from the least square correlation as t0.95s, where t0.95 is the Student coefficient and s is the standard deviation. The plots of ln[p(C84)/Pa] versus 1/T for samples 1 and 2 are presented in Figure 3. The values of the enthalpies of sublimation of C84 obtained from the pure sample and from the higher fullerene mixture were compared with the sublimation enthalpy of Moalem et al.8 and that of Piacente et al.9 (Table 2). The former result was determined in the study of a mixture of higher fullerenes [the composition estimated from surface analysis by laser-induced desorption (SALI) mass spectrometry was as follows: C60, 0.05%; C70, 13%; C78, 5.7%; C84, 44% and C96, 5.8%].This mixture was assumed to behave as an ideal solution and the value obtained was regarded as the sublimation enthalpy of pure C84. Our result for the sublimation enthalpy of separated C84 differs considerably from the values obtained for the mixture (see Table 2). This suggests that the higher fullerene mixture cannot be considered as an ideal solution.Using the data on the composition at the end of the experiment, the activity coefficient of C84 in sample 2 at 741 K was determined as g(C84) = 0.05±0.02 (Table 1). These results roughly satisfy a regular solution equation (1): where is the partial molar enthalpy of dissolution. We also determined the saturated vapour pressures and partial molar sublimation enthalpies of C76 and C78 for the higher fullerene mixture (see Table 2).Results for C60 and C70 were rather unexpected. Values of their activities were ca. 10–3, but the sublimation enthalpies were very close to the values obtained for pure substances.6,7 This can be ascribed to diffusion in the solid phase, i.e. the diffusion process controls the evaporation rate of the volatile component.It results in the decrease in concentration of volatile components C60 and C70 in the surface layer and leads, subsequently, to underestimated values of the activities of these components and underestimation of the partial molar sublimation enthalpies of C60 and C70. This effect would appear not to influence the results on the vapour pressure and sublimation enthalpy of the major components of the mixture, i.e.C84, C78 and C76, which are less volatile than C60 and C70. We are grateful to the Russian Foundation for Basic Research (grant nos. 97-03-003391a and 97-03-33682a) and the Russian Research Program ‘Fullerenes and Atomic Clusters’ for the partial financial support of this work. References 1 C.K.Mathews, M. Sai Baba, T. S. LakshmiNarasimhan, R.Balasubramanian, N. Sivaraman and P. R. Vasudeva Rao, J. Phys. Chem., 1992, 96, 3566. 2 M. Sai Baba, T. S. Lakshmi Narasimhan, R. Balasubramanian, N. Sivaraman and C. K. Mathews, J. Phys. Chem., 1994, 98, 1333. 3 J. Abrefah, D. R. Olander, M. Balooch and W. J. Siekhaus, Appl. Phys. Lett., 1992, 60, 1313. 4 A. Popovic, G. Drasic and J. Marsel, Rapid Commun. Mass Spectrom., 1994, 985. 5 E. V. Skokan, O. V. Boltalina, P. A. Dorozhko, L. M. Khomich, M. V. Korobov and L. N. Sidorov, Mol. Mat., 1994, 4, 221. 6 V. Piacente, G. Gigli, P. Scardala, A. Gustini and D. Ferro, J. Phys. Chem., 1995, 99, 14052. 7 V. Piacente, G. Gigli, P. Scardala, A. Gustini and P. Bardi, J. Phys. Chem., 1996, 100, 9815. 8 M. Moalem, M. Balooch, A. V. Hamza and R. S. Ruoff, J.Phys. Chem., 1995, 99, 16736. 9 V. Piacente, C. Palchetti, G. Gigli and P. Scardala, J. Phys. Chem., 1997, 101, 4303. 10 N. S. Chilingarov, M. V. Korobov, L. N. Sidorov, V. N. Mit’kin, V. A. Shipachev and S. V. Zemskov, J. Chem. Thermodyn., 1984, 16, 965. 11 J. W. Otwos and D. P. Stevenson, J. Am. Chem. Soc., 1956, 78, 536. 12 Termodinamicheskie Svoistva Individual’nykh Veshchestv (Thermodynamic Properties of Individual Substances), ed.V. P. Glushko, Nauka, Moscow, 1978–1982 (in Russian). aThe parameter of the temperature equation was estimated without errors. Table 2 Vapour pressures and sublimation enthalpies of C60, C70 and higher fullerenes. Work Fullerene System Temperature interval/K lg(p/Pa) = –B/T + A T/K DsubH0T /kJ mol–1 B/K A This C76 mixture 700–862 11586±913 10.32±1.21 756 222±17 This C78 mixture 700–862 11588±842 10.25±1.12 756 222±16 8 C84 mixture 800–950 12902±1306 15a — 247±25 9 C84 pure 920–1190 10950±300 10.92±0.30 950 210±6 This C84 pure 658–980 10570±234 10.92±0.28 853 202±4 This C84 mixture 700–862 12101±949 11.15±1.26 756 231±18 2 0 –2 –4 –6 –8 –10 –12 –14 1.0 1.1 1.2 1.3 1.4 1.5 1.6 ln[p(C84)/Pa] T–1/103 K–1 Figure 3 The plots of ln(p/Pa) versus 1/T for C84 over sample 1 ( ) and over sample 2 ( ). DsolH0 T(C84)m=RTln[g(C84)] (1) DsolH0 T(C84)m Received: Moscow, 7th May 1998 Cambridge, 8th June 1998; Com. 8/03647K