P.E. Vickers et al. Carbon 38(2000)675-689 Table 8 Advancing and receding contact angles 25% Medium 4.2±12.3 52.0+113 73.0±8.0 35.8±84 67.3+5.1 41.3±6.2 Glycerol 4±1.1 8.9±8.6 38.5±4.0 37.1±11.3 144±5.2 C.H-BI 25.0±13.4 48.1±3.0 29.1+142 the high-energy sites on the surface. Roughness of the angles recorded using liquids of known dispersive and surface generally has the effect of increasing the recorded polar surface free energy, a routine described by the advancing contact angle and decreasing the receding angle, equation below a result of geometric effects of a rough surface on an inclined plane Fig 6(a) shows the dCaa trace obtained for the 100% 1+cs=2y)2+2y2 treated fibre in water. The advancing curve may be identified by the trace marked by the zero depth Therefore, if the relevant properties are known for a range immersion (zdoi) marker, and the effect of hysteresis is of probe liquids, i.e. y2 and y 2, this allows the estimation observed by the change in force recorded by the balance of y i and y The surface free energy(generically known when the motor is rey versed and the half-cvcle of emersion as the surface tension, sl, or here, )2 )has been determined begins. Most interesting to note are the sharp deviations in for a range of predominantly dispersive liquids, such as force reading present on the receding plateau of the trace, I-bromonaphthalene, and polar substances, e.g. water which are a form of fine structure. The fluctuation in the glycerol, as listed in Table 2. Calculation of yi and yd is orce reading can either be a result of chemical hetero- therefore possible by solving the resulting simultaneous geneity, porosity or surface roughness. It is likely that the equations, a routine automatically performed by the soft receding curve(upper) is susceptible to huge effects ware present on the data collection system. Table 9, shows because of hydrophilicity of the surface, and the advancing that the dispersive contribution to the surface free energy is trace (lower) shows a fluctuating force reading more as a relatively unchanged by the oxidation, whereas, the polar result of surface roughness. Large excursions of this type contribution is greatly enhanced during the initial stages of are observed consistently on the receding plateau of all oxidation, but this does not significantly increase with three types of fibre analysed, whereas only small fuctua- further treatment tions are observed on the advancing plateau. Spikes observed on the force curves may also occur because of asperities on the surface either dissolving or interacting 4. Discussion with the contact medium. It is possible that the features observed on the receding plateau are an effect of pitting of Chemical changes in the surface composition of fibres as the fibre observed during oxidation [5], or a patch of the a result of various degrees of treatment affect the surface surface that is rich in cliffs that result from the spinning thermodynamics at the microscopic (adsorption of mole- process in the manufacture of graphite fibres from PAN, cules at infinite dilution) and macroscopic (wettability) where the edge planes of the graphitic structure are exposed [42]. Fig. 6(b) shows a DCAA spectrum of a Table 6 shows that ys values invariably decrease upon 100% treated fibre in 1-bromonaphthalene. In contrast to treatment, an effect which is illustrated in Fig. 7. This the force curve obtained with water, there is very little behaviour contrasts with the frequent observation in the hysteresis observed, because the low surface tension liquid literature that surface oxidation of carbon fibres yield an wets the surface moderately easily. In this case however, ncrease in y. values. However, ys values as low as there is also no fine structure observed on either the 50-60 mJ m-(at ca. 50.C)were reported 31, 45] advancing or receding plateaux. This is probably because untreated C fibres. Such a low value is indicative of strong l-bromonaphthalene does not interact strongly with hydro- surface organic contamination. It is therefore possible that philic sites, and small enough to enter the micro- pores on the fibre surface. This assumption is supported by duces high-energy functional groups to the surface. In our he spectra of 100% fibre in glycerol, shown in Fig. 6(c), case, we can claim that the untreated carbon fibres are where moderate contact angle hysteresis is observed, but relatively clean since we obtained a fairly high y, valt no fine structure on either plateau (104.5 mJ m at 50C. The decrease in the ys value A measure of the surface free energy may be obtained upon treatment is due to the grafting of functional groups [43, 44 by calculating the geometric mean of contact of the C-OH, C=O, COoH types as well as nitrogenP.E. Vickers et al. / Carbon 38 (2000) 675 –689 685 Table 8 Advancing and receding contact angles 0% 25% 100% Medium u u uu u u adv rec adv rec adv rec Water 94.2612.3 52.0611.3 73.068.0 35.868.4 67.365.1 41.366.2 Glycerol 58.461.1 8.968.6 38.564.0 16.662.2 37.1611.3 14.465.2 C H Br 42.4 10 7 63.4 25.0613.4 48.163.0 40.9612.2 43.564.4 29.1614.2 the high-energy sites on the surface. Roughness of the angles recorded using liquids of known dispersive and surface generally has the effect of increasing the recorded polar surface free energy, a routine described by the advancing contact angle and decreasing the receding angle, equation below. a result of geometric effects of a rough surface on an d p 1/2 d p 1/2 2(g g ) 2(g g ) inclined plane. ]]12 12 1 1 cos u 5 1 ] ]]] (10) Fig. 6(a) shows the DCAA trace obtained for the 100% 2 2 g treated fibre in water. The advancing curve may be Therefore, if the relevant properties are known for a range identified by the trace marked by the zero depth of d p of probe liquids, i.e. g 2 2 and g , this allows the estimation immersion (zdoi) marker, and the effect of hysteresis is p d of g 1 1 and g . The surface free energy (generically known observed by the change in force recorded by the balance as the surface tension, gsl 2 , or here, g ) has been determined when the motor is reversed and the half-cycle of emersion for a range of predominantly dispersive liquids, such as begins. Most interesting to note are the sharp deviations in 1-bromonaphthalene, and polar substances, e.g. water, force reading present on the receding plateau of the trace, p d glycerol, as listed in Table 2. Calculation of g and g is which are a form of fine structure. The fluctuation in the 1 1 therefore possible by solving the resulting simultaneous force reading can either be a result of chemical hetero- equations, a routine automatically performed by the soft- geneity, porosity or surface roughness. It is likely that the ware present on the data collection system. Table 9, shows receding curve (upper) is susceptible to huge effects that the dispersive contribution to the surface free energy is because of hydrophilicity of the surface, and the advancing relatively unchanged by the oxidation, whereas, the polar trace (lower) shows a fluctuating force reading more as a contribution is greatly enhanced during the initial stages of result of surface roughness. Large excursions of this type oxidation, but this does not significantly increase with are observed consistently on the receding plateau of all further treatment. three types of fibre analysed, whereas only small fluctua￾tions are observed on the advancing plateau. Spikes observed on the force curves may also occur because of asperities on the surface either dissolving or interacting 4. Discussion with the contact medium. It is possible that the features observed on the receding plateau are an effect of pitting of Chemical changes in the surface composition of fibres as the fibre observed during oxidation [5], or a patch of the a result of various degrees of treatment affect the surface surface that is rich in cliffs that result from the spinning thermodynamics at the microscopic (adsorption of mole￾process in the manufacture of graphite fibres from PAN, cules at infinite dilution) and macroscopic (wettability) where the edge planes of the graphitic structure are scales. d exposed [42]. Fig. 6(b) shows a DCAA spectrum of a Table 6 shows that g s values invariably decrease upon 100% treated fibre in 1-bromonaphthalene. In contrast to treatment, an effect which is illustrated in Fig. 7. This the force curve obtained with water, there is very little behaviour contrasts with the frequent observation in the hysteresis observed, because the low surface tension liquid literature that surface oxidation of carbon fibres yield an d d wets the surface moderately easily. In this case however, increase in g values. However, g values as low as s s 22 there is also no fine structure observed on either the 50–60 mJ m (at ca. 508C) were reported [31,45] for advancing or receding plateaux. This is probably because untreated C fibres. Such a low value is indicative of strong 1-bromonaphthalene does not interact strongly with hydro- surface organic contamination. It is therefore possible that philic sites, and is not small enough to enter the micro- surface treatment removes organic contaminants and intro￾pores on the fibre surface. This assumption is supported by duces high-energy functional groups to the surface. In our the spectra of 100% fibre in glycerol, shown in Fig. 6(c), case, we can claim that the untreated carbon fibres are d where moderate contact angle hysteresis is observed, but relatively clean since we obtained a fairly high g value s 22 d no fine structure on either plateau. (104.5 mJ m ) at 508C. The decrease in the g value s A measure of the surface free energy may be obtained upon treatment is due to the grafting of functional groups [43,44] by calculating the geometric mean of contact of the C–OH, C=O, COOH types as well as nitrogen