FILLER-ELASTOMER INTERACTIONS. PART I. CARBON-BLACK-SURFACE ENERGIES AND INTERACTIONS WITH ELASTOMER ANALOGS* Meno-Jiao WANG AND SincrRieD WOLFF Degussa AG, INORGANIC Cieaical ProDucts Division, APPLIED REStAncH AND TECHNICAL SERVICE FoR FILLERS AND RUBBER CHEMICALS, KOLMER Staassi 122, W-6047 Wrst AND Jean-Baprisre DoNNET CavTRE ne RECHERCHES SUR LA PHYSIC0-CHME nes SURFACES SoLIDES, 24, AVENUE DU PRESIDENT KENNEDY, F-68200 MULNDUSE, FRANCE INTRODUCTION Soon after carbon black was discovered to be an active filler in rubber at the beginning of this century, it became one of the most important components in the manufacture of rubber products, with a consumption second only to rubber itself. This situation will probably continue into the next century, and there is no indication of carbon black being replaced completely with other materials. Carbon blacks have held their position mainly thanks to their unique ability to enhance the physical properties of elastomers. This well-documented phenomenon’, termed “reinforcement,” has a profound effect on today's tire and rubber industry, During the last 50 years, an extensive amount of information on elastomer reinforcement with carbon blacks has been published. Although much attention has been focussed on the morphology of the blacks (particle size and structure) and its effect on rubber properties, it, has been recognized that the main feature of carbon blacks, distinguishing them from other fillers, is their high surface activity, which has an essential effect on the nature of the interaction able to take place between filler surface and polymer matrix. Surface activity, a poorly defined term, but widely used in the filler field, can, in a chemical sense, be related to different chemical groups on the surface, such as carboxyl, quinone, phenol and lactone, In a physical sense, variations in surface energy determine the adsorptive capacity and energy of adsorption. Unlike white fillers, whose chemical groups (hydroxy! and/or metal oxide groups) play an important role in their reinforcing ability for elastomers, the surface chemistry of carbon blacks has a significant effect only on the vulcanization of compounds. No direct correlation has been demonstrated between chemical groups on the black surface and rubber-related properties. However, it is generally believed that the surface energy of blacks is of much greater importance than their chemical nature with regard to the mechanical properties of the filled elastomers, particularly when general-purpose hy- drocarbon rubbers are concerned. Thus, the investigation of carbon-black-surface energy is, of great significance. In this study, the surface energies of a series of carbon blacks widely used in the rubber industry have been measured by inverse gas-solid chromatography at infinite dilution (zero coverage). Furthermore, the interactions between carbon blacks and elastomers were esti- mated from the energies of adsorptions of their low-molecular-weight analogs. The results will be compared with those of silicas, another important. group of fillers, so that the different, reinforcing behaviors of these fillers in polymers can be better understood. + Presented at a meeting of the Rubber Division, American Chemical Society, Washington, D.C., October 9-12, 1990.6. 714FILLER-ELASTOMER INTERACTIONS 715 EXPERIMENTAL MATERIALS ‘The carbon blacks used in this study are all rubber-grade products, ranging from N110 (SAF) to N990 (MT). The analytical properties of these blacks are described in Table I for the fundamental study. All samples are dry-pelletized without any pelletizing agent, unless otherwise indicated. A series of commercial products pelletized with a pelletizing agent were also tested. A channel black, $301, and a graphitized N375 (manufactured by Ringsdorff- Werke, Bonn, Germany) along with the initial black were included in order to evaluate the influence of the surface chemistry and crystallographic structure on surface free energy. ‘The preliminary experiments showed a great influence of the surface state and the washing method on the results. The dry-pelletized samples were washed twice with distilled water and then dried in a vacuum oven. The dried blacks were subsequently extracted with toluene at least 6 times. All extractions were carried out at room temperature (cold extraction) by mixing carbon black in solvent. After about 1 h extraction, the carbon blacks were centrifuged at 12 000 rpm for 30 min and the solvent was then decanted. After the last separation, the carbon blacks were placed in a vacuum oven at 115°C and dried to a constant weight, typically for 16 h. The probes used for the characterization of black surfaces and the elastomer analogs are listed in Table II, All probes were chromatographic-grade products and were used without further purification. CHROMATOGRAPHIC MEASUREMENT The equipment, the preparation of the columns, and the procedure of the measurements have been previously described’, For carbon blacks, the columns were conditioned at 180°C for about 20 h in a helium stream before the measurement. Since the experimental temperature has a great effect on surface energy, as will be dis- cussed later, the temperature of the columns was limited to 210°C. For most carbon blacks, Tante 1 ANALYTICAL PROPERTIES OF CARBON BLACKS Iodine Compressed ‘Tinting Carbon Surface area Surface area, adsorption, DBP adsorption, _ strength, black Na m/e CTAB, mg mig cL?/100 g % pH Nu10 140.0 127.0 146.0 93 us 89 N110 HS. 138.0 147.0 94 123 a7 N10 LS 140.0 = 147.0 7 18 87 N220 118.0 118.0 116.0 94 47 96 N24 120.0 120.0 122.0 98 116 93 N826 86.5 85.5 66 ut 92 778 82.2 81 101 92 914 86.9 39 110 8.9 80.3 89.2 94 95 94 87.5 88.0 92 105 89 BT 82.6 38 91 94 40.4 44.0 74 64 9.0 42.2 39.2 70 68 72 43.8 37.5 86 63 73 34.8 30.9 5B. 63 75 40.6 344 79 62 56716 RUBBER CHEMISTRY AND TECHNOLOGY Vo. 64 ‘Tape IT ‘THE PRones FoR CARBON-BLACK-SURFACE CHARACTERIZATION AND ELASTOMER ANALO( Probe Structure Rubber n-Alkanes ANAN Re n-trans-3-Alkenes AA AR ARS NR BR n-1-Alkenes BN AN AOS CNR BR 3-Methyl-t-alkenes LOC CT aver 2,2-Dimethyl-alkanes Nee pee UR Alkylbenzenes Nitriles the chromatographic measurements were performed at 5 temperatures ranging from 150 to 210°C, In the case of the thermal black N990, the lowest temperature used was 105°C. CALCULATION OF THE ADSORPTION PARAMETERS AND SURFACE PRE ENERGIES OF CARBON BLACKS* ‘The change in free energies of adsorption of probes on filler surfaces, AG’, was obtained directly from the net retention volume: ag? = -RT-n 2, a Sg where Vy is net retention volume corrected for gas compressibility and temperature; S is specific surface area of the filler; g is mass of the filler packed in the column; R is the gas constant; 7 is temperature in K; and C is a constant which is 299 when De Boer’s definition of the standard state is adopted‘, Surface energy of the filler, y,, consists of two components, the dispersive yi, and the specific (or polar) y;” component: went @) The y! was obtained from the free energy of adsorption per mole of methylene group, AGox,, Which was derived from the slope of the linear plot of AG” of a series of n-alkanes vs, their carbon-atom number*; 4 Abn Gabaron .) where acy, is the area covered by a ~CH,~ unit; yeu, is the surface free energy of a surface composed entirely of -CH.- units; and N is Avogadro's number. The value of 77” was estimated from the specific interaction, 7”, which the filler is able to exchange with the polar probe. T can be calculated from the difference between the adsorption energies of polar probes and n-alkanes (real or hypothetical) having the same molecular surface area, which is termed AAG. Thus one obtains: a ‘Na, where a, is the surface area of the polar probe. 1° , (4)FILLER-ELASTOMER INTERACTIONS TT RESULTS AND DISCUSSION DISPERSIVE COMPONENTS OF CARBON-BLACK-SURPACE ENERGIES, Ys The effect of solvent extraction and heat treatment on y;.—Figure 1 shows the effects of the sample treatments of carbon black N330 on y3 which were obtained from free energies of adsorption of a series of normal-alkanes at 180°C. The difference in y; between extracted and nonextracted blacks is remarkable. Since no pelletizing agent. was used in the posttreat- ment of the black and since the chemical groups on the filler surface cannot be changed during the extraction at room temperature, such an effect can only be attributed to some extractables which are believed to be some oily residues, so-called tars, left on the blacks when carbon-forming and dehydrogenation reactions are quenched in the furnace just as these reactions are nearing completion®. However, it has not been clarified, so far, how these substances influence the surface free energy. One of the possible reasons may be the com- plication of the surface, especially on the crystallographic plane of quasi-graphite crystallites, by these extractable matters. The hot extraction was carried out in a Soxhlet extractor for 16 h with toluene as the solvent. The slightly higher 7; obtained with hot extraction, as compared to the results of the cold-extracted sample, is probably associated with the loss of some chemical groups on the black surface during extraction, which may increase the numbers of high-energy sites, as will be discussed later. In a previous study, Donnet and his coworkers’ were able to show the effect of toluene extraction at boiling temperature. They found that the chemical group concentration on the black surface decreased after extraction and did not reach an equilibrium until two weeks later. Heat treatments were performed in the carrier gas at given temperatures for 20 h, and the surface free energies were then measured at 180°C. The data yielded by this treatment suggest the following 1. Below 210°C, the heat treatment has no influence on 7. 600 500 400 300 ¢- Unextracted ©- Cold extracted ©- Hot extracted TC 100 100 200 300 400 500 Fig, 1. Effect of toluene extraction and heat treatment on 4f of N330 carbon black 200)718 RUBBER CHEMISTRY AND TECHNOLOGY Vou. 64 2, Beyond 210°C, the dispersive component of the black surface energy increases with rising temperature, whether the samples were extracted or not. It is obvious that the change in vf at high temperatures would reflect a change of the black surface, particularly of the surface chemistry, which has been the subject of many studies published. Garten and Weiss* demonstrated that the number of acid groups of carbon blacks started to decrease beyond 200°C, and others'**' showed that the oxygen groups decompose during thermal treatment (from 200 to 800°C), accompanied by the evolution of CO, and CO. Therefore it is logical to assume that after decomposition of the oxygen groups, some high-energy sites would be left and/or that some original energetic spots which are screened by the oxygen groups could be revealed. ‘Thus, in order to secure the black surface for ensuring reliable data comparison, the temperature of conditioning and measurement should be limited to below 210°C. The dependence of ¥; on surface area and structure.—The surface energies and some adsorption parameters at 150°C are given in Tables III and IV for all carbon blacks inves- tigated. The heats of adsorption are listed in Table V. In Figures 2 and 3, 7 is plotted vs. 24 M4 DBP values and nitrogen surface area, respectively. While y? seems to be independent of carbon-black structure, it increases with surface area, more or less, in a linear fashion. Such a result is inconsistent with other investigations of surface activity of carbon blacks and is related to rubber reinforcement. With microcalorimetry, constant integral heat of immersion per unit surface area was reported for different carbon blacks upon contact with liquid hydrocarbons ranging from benzene and n-decane to liquid rubber’"*, The results of anus I ‘f oF Finuers anp Titan Speciric ee IWTERACTION WITH POLAR PRODES Free ENeRotps oF Ansonprion oF Some PRones ar 150°C on Fiuuzs aT 150°C he Weegee Heptane, Heptene, Benzene, Acetonitrile, Filler mJ/m®_mJ/m?_mJ/m? Filler ki/mol_—_kJ/mol__—_kJ/mol__—_—kdJ/mmol N10 2704 120.0 2126 N110 347 33.9 32.9 207 NIOHS 206.7 1077 189.9 NI1OHS 35.0 33.3 29.8 159 NIOLS 284.2 1010-1822 NIMOLS 36.0 345 29.9 16.9 N220 295.2 108.9 179.1 N220 318 20.8 28.5 16.1 N24 226.9 90.4 180.5 N234 315 29.9 26.0 166 N26 186.5 90.2. 164.8 NB26 29.1 27.8 26.7 158 330 196.9 85.9 193.7 N330 28.4 271 237 180 N332 2073 91.3 1826 Naga 278 268 24.0 154 339 232.4 100.8 178.0 N939 316 303 278 16.0 N47 1929 87.9 160.7_NB47 28.4 271 243 141 N3T5 196.0 90.7 158.2 NB75 29.8 28.2 25.6 144 N539 1682 77.1 1764 _NB89 25.9 247 218 16.9 N50 1344 75.0 165.0. NB5O 243 23.5 au 16.1 660 1247 TLL 158.0 N660 23.9 23.1 205 158 N633 125.6 70.5 155.9 N83 24.1 23.2 15.6 N762 1264 77.7 165.6 N762 247 23.6 16.7 N765 1289 77.9 189.8 N765 245 23.6 173 N990 718 56.6 1220 NO90 198 18.9 133 Silica PI 22.9 = 64.0 252.0 SilicaP1 = 9.2 6 206 343° 719 285.0 SilicaP2 «9.6 124 317 307 45.4 155.9 Siliea Al 10.0 2 16.2 158 Silica A2* 44.3 55.1 173.0 SilicaA2_——'106 17 * Products of Degussa AG: Pi, Ul- trasil VN 2; P2, Ultrasil VN 3; Al, Aerosil 130; A2, Aerosil 200.FILLER-ELASTOMER INTERACTIONS 719 ‘TasLeV HEATS OF ADSORPTION OF SOME PROBES ON FILLERS: Heptane, Heptene, Benzene, Acetonitrile, Filler kJ/mol kJ/mol kJ/mol kJ/mol Nilo 78.8 116 812 60.8 N110 HS. 79.4 mA 74.4 64.1 N11OLS 76.2 74.3 72.0 87.5 N220 74.4 67.3 73.2 50.5 N234 72.7 67.6 612 46.2 826 70.4 67.4 66.3 50.8 N330 677 672 619 63.1 332 m4 68.5 64.9 65.5 Na39 73.3 70.5 19 60.9 N347 66.6 63.8 615 48.1 N35 64.4 63.0 60.4 36.3 N539 598.1 54.9 55.0 52.6 N550 512 49.3 49.8 40.7 N66 497 48.2 45.1 37.4 683 55.0 54.2 516 50.1 N762 87.4 54.7 60.3 46.6 N765 58.7 57 58.5 53.9 990 67.1 56.6 47.3 59.9 Silica P1 47.6 618 60.7 - Silica P2 56.5 69.5 66.3 — Silica Al 42.9 53.0 43.8 62.9 Silica A2 514 59.5 513 59.3 bound rubber per unit surface area even showed a decrease of surface activity with increasing surface area of the carbon blacks". It is necessary to point out that the values for y; mea- sured chromatographically are results averaged over the whole surface. In the case of bound rubber, the result may sometimes be misleading as a measure of surface activity because, ‘on the one hand, only part of the apparently bound rubber is truly adsorbed on the black surface and, on the other hand, multiattachment of the rubber molecule on the black surface makes the estimation of surface activity difficult. It can be understood that the rubber mol- ecule may attach to the filler surface over and over again at new sites, This, however, would reduce the effectiveness of the surface, since only one attachment would make the whole molecule inextractable in solvent. This multiattachment of single chains may occur on dif- ferent black particles (aggregates) close to each other. It depends, of course, on the distance between filler particles. This distance decreases as the filler particle size decreases at. the same loading, which in turn reduces the efficiency of the formation of bound rubber. Inastress-strain study of filled rubber, Kilian and co-workers" found that the parameter Ju, which is related to slippage energy of polymer chains along the filler surface during stretching, is directly proportional to the y{ of carbon blacks. The dependence of y$ on the microstructure of carbon blacks.—It is known that in carbon black, the carbon atoms are largely considered to be parts of quasi-graphitic crystallites consisting of several turbostratic layers at an interplanar distance, d, of 0.35 to 0.38 nm. ‘The crystallite dimension is characterized by the average stacking height of the parallel planes in the c direction, I, and the average diameter of the parallel layers in ab plane, L. ‘The difference in ys between carbon blacks may be associated with differences in their mi- crostructure rather than their surface area or particle size. The apparent surface area de- pendence of 7! may reflect the particle-size dependence of the microstructure of the blacks. Significant variations in the erystallographic parameters of carbon blacks have been720 RUBBER CHEMISTRY AND TECHNOLOGY Vou. 64 400 ya;md/m2 300 e . . 8. 200 2 o ee ° e 774 e 100 DBP, mI/100g 0 \ L 1 40 60 80 100 120 Fic. 2.—yf vs. 24M4 DBPA for dry-pelletized carbon blacks at 150°C. reported'*”, generally showing a decrease of the crystallite dimensions with increasing surface area. In Figure 4, 7/ is presented as a function of L., indicating that surface energy decreases with increasing crystallite dimensions. This is understandable, since more crystal edges and 500 400 300 200 100 1 4 1 Ss ” mig 50 100 150 200 s. nitrogen surface area for dry-pelletized carbon blacks at 150°C,FILLER-ELASTOMER INTERACTIONS. 721 400 ¥g,mJ/m? 150°C 300F \_, . a 200 . ™y .° . 100 Le,nm oO 1 1 1 14 16 18 20 Fi. 4. yf as a function of Le for dey-pelletized carbon blacks at 160°C a greater unsaturated charge, which can be considered as high-energy sites, would be expected in smaller crystallites, corresponding to the smaller particle-size products. This ean be con- firmed, on the one hand, by the surface free energy of graphitized carbon black (it will be discussed later), and, on the other hand, from the comparison with surface energies of other carbonaceous materials When the y; in Figure 3 is extrapolated to zero surface area, a yz value of 74 mJ/m? is obtained. This value is comparable to that of N990 (71.8 mJ/m*). If we consider the tem: perature dependence, the yf of “'zero surface area” carbon black at room temperature would fall into the range of graphite and carbon fibers. The selected data from the literature are Tanus VI +f! oF Carbon Proovers x ‘Temperature, Product mJ /m? °C Method Reference Graphite powder 80-100 water adsorption 18 Graphite powder 115-132, water adsorption 19 Graphite 151 two-phase liquid 20, Graphite 123-139 60 icc 21 Graphite 168 20 contact angle ai Carbon fiber 109-120 20 icc 22 Carbon fiber m1 IGe 23 N990 7 Icc N980° 7 IGc Iec Carbon black” 74 © Extrapolated value. “The value extrapolated to zero surface area.722 RUBBER CHEMISTRY AND TECHNOLOGY Vou. 64 600 400 200 S,m¥g ° \ L 1 L L L i oO 40 80 120 160 Fra. 6. ‘vs, nitrogen surface area for wet-pelletized carbon blacks at 180°C. @, furnace carbon blacks O, channel black given in Table VI. Thus if the 7; of graphite or the “zero surface area” black is taken as a reference, the difference in y; between a given black and the reference may be attributed to imperfections of the graphite lattice structure of the products. Therefore, it follows that the defects of the crystal lattice or the development of a quasi-graphitic structure play a very important role in the high surface free energies of the carbon blacks. ‘The above discussion shows that the surface-area dependence or particle-size dependence of carbon black surface energies is more representative of the effect of microstructure. Nev- ertheless, for the sake of convenience, we will continue using the surface area to correlate the data. The 4 of commercial carbon blacks.—The 7 of 26 commercial carbon blacks, measured at 180°C, are presented in Figure 5 and also given in Table VII along with adsorption pa- rameters. These carbon blacks were pelletized with a pelletizing agent (wet pelletizing process) and extracted in a Soxhlet extractor with toluene (hot extraction) prior to chromatographic measurements. Although these carbon blacks follow the same trend in the surface area dependence as observed for the dry-pelletized and cold-extracted products, i.e., ys increases as the surface area increases, the values of the surface energies are much higher for the wet-pelletized carbon blacks, especially for the smaller particle-size products. The posttreat- ment processes changing from dry to wet, for instance, raise the y; from about 285 mJ/m? to 400 mJ/m? for N110-series carbon blacks at 180°C. Additional work is in progress to clarify the reason for such a big difference, such as the effect of posttreatment on the surface state, the extraction procedure, and temperature. The yi of channel carbon black.—Only one channel carbon black, $301, was investigated, and the results are presented in Figure 5 and Table VIL. There is no significant difference in yi! compared to commercial furnace blacks, even though the surface chemistry of channel blacks is quite different from that of furnace carbon blacks on whose surface the concentration of oxygen groups is much lower.FILLER-ELASTOMER INTERACTIONS 723 ‘Tame Vit SunFACE ENERGIES AND ADSORPTION PARAMETEIS OF SOME PROBES ON COMMERCIAL Canwow BLacKs at 180°C Carbon Surface area a Pana Te 5. 5, black No, m'/g. mJ /m? mJ /m? mJ/m® benzene MeCn Ns 1370 ain 105 169 127 1.44 Black 1 149.0 429.0 105 173 127 144 NI2t 1210 361.6 102 207 1.28 156 110 378.0 107 186 1.28 1.49 1110 395.8 U3 i90 1.20 149 126.0 403.1 105 197 127 181 N339 87.7 286.6 99 181 130 155 N326-1 78.2 260.2 100 175 132 1.56 N326-2 735 271.2 97 178 130 158 Na32 887 100 183, ual 157 N3a7 86.9 97 17 1:30 1.55 NavB 92.9 98 12 131 154 Na30-1 80.5 98 183 130 137 N330-2 782 99 186, 131 187 Black 2 789 99 187 133 1.62 Black 3 874 103 187 32 158 Black 4 79.9 98 194 130 1.60 Black 5 872 106 180 132 1.52 N55O 39.0 88 Ml 1.32 1.60 N660 349 ot 137 129 1.60 No83 38.1 76 167 131 1.69 N72 356 66 150 129 167 69 146 130 1.62 N765 36.0 The 5 of graphitized carbon black.——The effect of graphitization of carbon black on surface energy is summarized in Table VIII. A drastic decrease of 7s from 257 to 189 mJ/m? at 180°C upon graphitization is obviously due to the growth of crystallites which may be another evidence demonstrating the microstructure dependence of the surface free energy Tante VII ERPEcr oF GkaPHIniZATiON oF N375 oN SURFACE ye, md/m? BG boncene, KJ/MOL AG huey, KI/mol Ghar, k3/mol AG eae, KI/mol Tawney TS Fifeow , 105/10 Spomzone Srmvcw Syme Spence Original NAS 287.1 24.1 13.9 18.2 16.7 93. 183 1.30 1.58 1a 1.26 ENERGY AND ADSORPTION PRopeRriEs ar 180°C Graphitized N375, 188.9 19.2 45 126 13.5 60 100 1.26 1.36 1.20 1.23724 RUBBER CHEMISTRY AND TECHNOLOGY Vou. 64 200 7 1°, mJ/me Benzene 150°C 150 100 ° 7 ro a ame 50 e : S,myg ) 50 100 150 200 Fio. 6.1” of benzene vs. nitrogen surface area for dry-pelletized carbon blacks at 150°C. of blacks. The fact that the 7! of the graphitized black is still higher than that of graphite may be caused by the incomplete graphitization and/or limited growth of the crystallites" ‘THE SPECIFIC COMPONENTS OF CARBON-BLACK-SURFACE FREE ENERGIES ‘The specific components of surface free energies of carbon blacks are evaluated by the parameter ” of polar probes, benzene and acetonitrile (MeCN). Since benzene is capable of specific interaction with the black surface via the electronic localization of a big x-bond system and since, in the case of acetonitrile, hydrogen bonding may be involved, the I” of benzene and MeGN should reflect the specific components of surface free energies of the carbon blacks A comparison of carbon blacks is shown in Figures 6 and 7, where the J”? of benzene and acetonitrile, respectively, are plotted against the surface areas of carbon blacks. The results suggest that the smaller-particle products exhibit greater specific interaction with polar probes, even though the data are somewhat scattered. This pattern is similar for wet-pel: letized carbon blacks. However, similar to the dispersive components of surface energies, the surface-area dependence of the specific interaction is more representative of surface state rather than surface area itself. This is illustrated by the fact that the specific interaction of the channel black with the polar probes is slightly but definitely higher than expected from its surface-area dependence in the case of furnace blacks (see Table VII). ‘This is un- doubtedly related to the high concentration of oxygen groups, which are polar sites by nature Itis also found that the graphitization of carbon black produces a considerable decrease in the specific component of surface energy (see J” in Table VITD), which is probably due to the mechanism of growth of crystallites (decrease of crystal-structure defect) as well as a diminished number of oxygen groups, which are completely decomposed below 800°C. INTERACTIONS OF CARBON BLACKS WITH LOW-MOLECULAR-WEIGHT ANALOGS OF ELASTOMERS The free energies of adsorption of low-molecular-weight analogs of rubbers could provide information concerning the ability of the filler surface to interact with polymers. For thisFILLER-ELASTOMER INTERACTIONS 725 400) 1, mum? 200 S, mg % 50 100 150 200 Fic. 7.1" of acetonitrile vs. nitrogen surface area for dry-pelletized earbon blacks at 150°C purpose, olefins were used as model compounds of unsaturated rubbers such as NR and BR, and alkanes as those of saturated rubbers such as EPR and IIR. A series of alkylated benzenes and homologous nitriles were used to evaluate the contribution of an aromatic ring to in- 50 -aG:, kJ/mol N110 150°C ” 40} J a |. www 30 oa eee ann a Wf 20+ e Vi / ° ANCA 10} S.A.,nm2 c i 0.15 0.25 0.35 0.45 055 Fio, 8.—Energies of adsorption of a variety of probes on N110 2s. their molecular surface areas at 160°C.726 RUBBER CHEMISTRY AND TECHNOLOGY Vou. 64 50 -aG kJ/mol N330 150°C 40;- 30 a Lo 20 a LE on 10 Rice a i S.A,nm? 0.15 0.25 035 045 0.55 Fig, 8. Energies of adsorption of a variety of probes on N380 ws. their molecular surface areas at 150°C. teractions of SBR with carbon blacks and the contribution of the ~CN group to interactions of NBR with blacks, respectively. Figures 8, 9, 10, and 11 show the free energies of adsorption of a variety of rubber analogs on the carbon blacks N110, N330, N550, and N990 as a function of the surface area covered by the probe molecule. a ~4G; kJ/mol N550 150°C 40 Sof 4 aan sok 2X —_ ° OF oe ee 6@ 0 20 aNC™ & a ao 10 S.A, nm? G38 0.25 0.35 0.45 055 Fio. 10.—Enengies of adsorption of a variety of probes on NB50 vs. theit molecular surface areas at 150°C.FILLER-ELASTOMER INTERACTIONS 727 50 -4G; kJ/mol N990 150°C 40h en~ aa~ wae 30 20 10 rs 0.25 0.35 045 0.55 Fig. 11.—Bnergies of adsorption of a variety of probes on N9G0 1s, their molecular surface areas at 150°C. It can be seen that the plots of all homologous probes are straight lines, which can be treated as family plots. The difference in the ordinate between the given family plot and the plot of n-alkanes is related to the contribution of the functional groups. The slopes of the family plots of the homologous probes could be representative of the contribution of AG? of the ~CH,- group per unit surface area if the interaction between filler and functional group is kept constant. Probably caused by the high polarity of the nitriles as well as their ability to form hy- drogen bonds with the black surface, the -CN group shows a much greater specific interaction than the benzene ring, which is only capable of conjugated r-bond specific interaction with the filler surface. The low slope of the nitrile family plot suggests a strong attenuation of the specific interaction due to the increasing number of alkyl chains. This may be interpreted as resulting from the increase of the distance of the -CN group from the filler surface caused by increasing methylene content or from a different configuration of the alkyl chains which do not fit on the filler surface, as in the case of n-alkanes. This effect is less pronounced for the aromatic homologs. For the aliphatic chemicals, the family plots are much higher than those obtained from silicas®, This, of course, results from the high / of carbon blacks. Moreover, their family plots almost crowd together, indicating a lesser importance of the double bond. This is also different from what is observed on silicas, where the r-bond plays an important role in olefin adsorption. This may be related to the low polarity of carbon black, making the double bonds less polarized. The comparison of AG between aliphatic hydrocarbon probes with the same carbon atoms suggests that the olefins have lower energies of adsorption than normal alkanes, This can be seen clearly from Figure 12, in which all probes have six carbon atoms. The lower AG? of olefins may be interpreted as their smaller molecular surface area and/or their con- figuration on the black surface. When the values of AG” are normalized only by the molecular surface area, specific interaction is found between carbon blacks and the double bond of normal alkenes (Figure 13). It is noted that the end x-bond has a stronger specific interaction than the double bond inside the molecular chains (n-1-hexene vs. n-trans-3-hexene). The728 RUBBER CHEMISTRY AND TECHNOLOGY Vou. 64 -aG} kJ/mol 180°C 22 N330 180 20 18 16 aN RI ow ar FO Fro, 12.—Energies of adsorption of some probes on NB3O at 160°C. probes having 5, 7, and 8 carbon atoms follow the same pattern. This leads to the conclusion that the effect of the methylene group on increases in the electron density of the x-bond is not as pronounced as the effect of the molecular configuration on the surface. In the case of Lalkanes, the end double bond would fit better on the adsorbent. Probably for the same reason, cis-isomers have a higher AG° than érans-isomers. ‘The steric effect of the branched probes, i.e., weakening of the interaction between filler and probe by the side group through the mechanism of increasing distance between the mass centers of adsorbate and filler surface, is observed for both alkanes and olefins. The specific interaction of double bonds could be partly counteracted by the steric effect for cis- I?mdJ/m? N330 180°C 10: oN PI oe ow FO rie Fi. 13.—Specific adsorption energies of some probes on N330 at 180°C.PILLER-ELASTOMER INTERACTIONS 729 50 -aG', kJ/mol N330 150°C 40}- 4 et | ao} LO ae _ 20+ oe J on. 10; S.A.,nm? 0 : 1 P i 0.15 0.25 0.35 0.45 055 Fic, 14.—Comparison of free energies of adsorption between probes with different x-bonds ‘and saturated chemicals on N30. 2-hexene and completely offset for 3-methyl-I-hexene, whose side methyl group is isolated from the x-bond. In the same way, the model compounds of EPR and IIR show the weakest interaction with carbon blacks. In order to confirm the specific interaction of the x-bond with the black surface, a series of cyclo-chemicals, which can be considered to have about, the same configuration on a solid surface, were used as probes. The results in Figure 14 indicate that the introduction of double bonds in cyclo-hexane results in an increase of the energy of adsorption. The greater the number of double bonds introduced, the higher the specific interaction observed, even though the molecular surface areas become smaller. On the other hand, the conjugated 7- system seems to be more effective in this respect. All carbon blacks, whatever the production process and posttreatment, show the same pattern, but, as mentioned previously, the level of the interaction energies for all probes decreases with increasing particle size, corresponding to their dispersive and specific com- ponents of surface energy. From the above-mentioned results of filler/rubber analog interactions, the elastomers can be classified in the same order as that obtained with silicas, i¢., NBR and SBR exhibit strong interactions with carbon blacks, followed by unsaturated rubbers, such as NR and BR. The saturated rubbers, particularly branched polymers, ¢.g., IIR, exhibit low interaction with fillers. This observation is in good agreement with the results of bound rubber, indicating a higher activity for SBR, followed by BR. IIR gives the lowest bound-rubber content'*, In an electron microscope observation of stretched vulcanizates of NR, BR, and SBR, filled with carbon blacks, Todani and Sagaye™ were able to show an agglomeration of the carbon-black particles under elongation, forming a very heterogeneous structure in the polymer matrix for NR and BR, but this phenomenon was much less developed in SBR. We believe that one of the reasons for this phenomenon is the stronger interaction between SBR and carbon black,730 RUBBER CHEMISTRY AND TECHNOLOGY Vou. 64 COMPARISON OF THE SURFACE ENERGIES OF CARBON BLACKS AND SILICAS Surface energies.—It is now possible to discuss the difference between carbon blacks and silicas from the surface energy point of view, particularly in relation to the reinforcement of general-purpose hydrocarbon rubbers, which are the most important elastomers used in the rubber industry. In Tables III, IV, and V, the surface energies and some adsorption results of silicas, both precipitated and fumed, are also included. It can be seen that, on the one hand, 7$ is extremely high for carbon blacks. At 150°C the yf of N110 is about 10 times higher than those of the silicas P1 and Al, which have a comparable surface area. This means that carbon blacks may exhibit stronger interaction with non- or low-polar materials. On the other hand, very high specific interaction is found for silicas, particularly when dealing with highly polar probes. This is an indication of a very high 7 of the silica surfaces and strong interaction with polar materials. S,factor.—While fillers have polar surfaces, the hydrocarbon rubbers are generally non- (or very low-) polar materials. The higher adsorption energy of a nonpolar probe, such as heptane, resulting from a higher / of the filler surface, would indicate a greater ability of the filler to interact with hydrocarbon rubbers. However, the higher adsorption energy of ahighly polar probe, such as acetonitrile, resulting from the high 7”, would be representative of filler particle-particle interaction in hydrocarbon rubbers. In Figure 15, the energies of adsorption of acetonitrile, AGi,cv, are plotted against AGéz, the energies of adsorption of heptane. It is evident that much information can be extracted from the comparison of the data between these two types of fillers, which are located at different regions on the plot. It can be understood that when AG2, is constant, the increase of AG‘ecv Would promote incompatibility of the filler with hydrocarbon rubbers, thus enhancing aggregate agglomeration. At the same level of AGiwcw, the higher AG? would indicate stronger filler/polymer interaction, hence a lesser degree of association of the ag- gregates would be expected. Based on this consideration, we have defined a specific interaction 50 ~4Givecw, ku/mol 150°C c 2 5% 40 g 2 Silicas ‘o 30 2g 5 2 20 3 ee = “carton blacks g 10 . 2 ~sG¢,,ku/mol ro 1 LL 1. . 0 10 20 30 40 Hydrocarbon rubber/filler interaction ——————> Fi, 15.—Energies of adsorption of acetonitrile vs. those of heptane for dry-pelletized carbon blacks and silicas.FILLER-ELASTOMER INTERACTIONS 731 Fares oo kanes aGaik -2G5 om 7 \ S.A. Fig, 16. Definition of the $, factor, factor, S,, as the adsorption energy of a given probe, AG®, divided by that of an alkane (real or hypothetical), the surface area of which is identical with that of the given probe, AG2 (Figure 16): S255 © 4 Se-senzene 450°C 3 2 Q ° oe ee athe wea 1 S.myg 0! 1 1 4 1 1 : 1 0 40 80 120 160 200 Fi, 17.—Sof benzene on carbon backs and silica a 150°C732 RUBBER CHEMISTRY AND TECHNOLOGY Vou. 64 Sar 4a2 2 ‘CARBON BLACK ° a eae o— o3 —_______ 1 S,myg fo} 1 1 4 1 1 1 L 0 40 80 120 180 200 Fig, 18 —S,of acetonitrile on carbon blacks and silicas at 160°C, In this definition, the reference state for calculation of AG’ is the AG’ of a hypothetical alkane with zero surface area, which is extrapolated from AG? of a series of n-alkanes. For a highly polar probe, the higher the 5; of a filler, the more developed is the aggregate ag- glomeration in hydrocarbon rubbers. 50 -aGi,kJ/mol N330 150°C 40 LS 30 | 20 eect = pne fae McHCly cite,“ “S Mcotone 10 i __ S.A, nm? Sas 0.25 0.35 045 055 Fro. 19. energies of adsorption of a variety of probes with different acidity on N3B0 at 150°C,FILLER-ELASTOMER INTERACTIONS 733, 50] .G; ku/mol 450°C Pt 40 30 scan “em Pe thoe 20 10 S.A.,nm? 35 0.25 0.35 0.45 055 Fic, 20.—Energies of adsorption of a variety of probes with different acidity on the silica PI at 160°C. ‘The S, of some polar probes on carbon blacks and silicas are compared in Table IX. The data of benzene and acetonitrile are illustrated in Figures 17 and 18, In the case of benzene adsorption, silicas, especially precipitated silicas, show higher relative polarity than carbon Tanue IX 5, VALUES oF Som POLAR PROBES ON FiLtires aT 150°C Filler Benzene MeCN THF Ether __ Acetone EtAc CHCl, CC Nilo 1.36 1.63 1.36 1.09 Lal 1.26 1.25 aa N110 HS. 1.31 LBB 1.25 1.05 1.28 115 1.25 14 NILOLS 1.30 1.64 124 1.04 1.28 115 1.25 14 220 1.34 1.58 131 1.05 1.35 1.20 1.25 115 N234 1.30 1.59 126 1.06 1.30 LIT 1.27 1.16 326 1.33 1.60 1.25 1.05 1.30 Liz 1.26 116 330 1.30 1.68 1.26 1,05 1.30 115 1.26 116 NaB2 131 1.63 127 1.07 132 118 127 116 N39 1.33 1.58 1.25 1.05 1.28 115 1.25 Lu NB47 1.31 157 127 1.05 1.29 1.16 1.26 115 B75 1.32 1.56 1.25 1.05 1.27 1.15 me iy 539 1.30 1.69 141 140 1.46 1.28 1.25 LAS BBO 1.32 1.69 1.09 144 1.29 1.26 1B N660 131 1.69 1.07 134 1.20 1.28 1.15 Ness. 131 168 1.08 Lat 1.26 1.27 116 N762 1.34 1.67 1.09 1.39 1.22 1.29 116 N765 1.88 173 1.09 144 1.27 129 116 N990 1.32 1.70 1.09 a2 124 1.37 121 Silica PL 1.66 3.58 201 291 2.50 1.45 1.24 Silica P2 1.60 3.87 2.08 2.84 2.49 149 1.28 Silica Al 4.40 2.38 1.66 219 191 1.42 1.23 Silica A2 1.40 2.26 1.58 2.08 1.84 1.42 1.24784 RUBBER CHEMISTRY AND TECHNOLOGY Vou. 64 5° G,ku/mol -aG;ku/mo! ” ° A 150°C 40 30 rue etac 20 Mcslone Wing, 1 ogee coal OCs WW ccs Pec, ger es S.A., nm? 0 015 0.25 0.35 045 055 Fig, 21.—Energies of adsorption of a variety of probes with different acidity on the sillea Al at 160°C. blacks. The values of S; seem to be constant for all carbon blacks, only very slightly higher for the large-particle blacks. The same can be seen from nitrile adsorption, but the difference in S, between carbon 3 Sf-cuicts 150°C ak [ og OH ee ° “srs 24, af S,mig ° \ L . 1 1 L 1 ital Oo 40 80 120 160 200 Fio, 22.—S,of CHCl, on carbon blacks and silicas at 160°C.FILLER-ELASTOMER INTERACTIONS. 735 4 oi 150°C 3k Spy Opp at oa One rv CARBON BLACK ab | S,mzg 0 \ 1 i : L \ it oO 40 80 120 160 200 Fic. 23.—S, of THF on carbon blacks and silicas at 150°C. blacks and silicas is much greater, due to the strong polar interaction as well as hydrogen bonding between silica surface and the probe, as mentioned earlier. The very high S; of silicas suggests a strong tendency of the silicas to form a developed filler network (or sec ondary network) or cluster, even if it is not continuous, in the polymer matrix. Another conclusion drawn from these data is that from the energetic point of view, the agglomeration tendencies are the same for all carbon blacks. Thus the agglomeration, i.e., the dispersion of the carbon black in a given polymer, would only be determined by its particle size and structure, both of which determine the distance between particles. It is very interesting to note that S, appears to be independent of temperature, at least over the range of the temperatures investigated. ‘Acidity.—Acid-base properties of filler surfaces were also assessed from the adsorption energies of various chemicals. Among the probes used, THF and ether are basic, and acetone and ethyl acetate are amphoteric. Chloroform and carbon tetrachloride are taken as the acid probes. The results are presented in Figures 19, 20, and 21, taking the carbon black N330 and the silicas P1 and Al as examples, It can be seen that contrary to the carbon blacks where no great difference between the probes was found, the adsorption energies of basic and amphoteric probes are much higher in the case of silicas than those of acid chemicals, indicating the high surface acidity of the silicas. Furthermore, this behavior is more pro- nounced for the precipitated silica. ‘The same conclusion can be drawn from the S, factors of these probes. In Figures 22 and 28, the S, of CHCl, and THF are plotted ws. the surface area of the fillers. Here again there is no significant difference between silicas and carbon blacks for the acid probes. When dealing with the basic probe THF, however, the pattern is quite different. The very high S, of THF on the silicas suggests a strong acid-base interaction between them. It is noted that the S; values of each probe on all furnace carbon blacks are essentially identical.736 RUBBER CHEMISTRY AND TECHNOLOGY Vou. 64 SUMMARY ‘The surface energies, both the dispersive component, 77, and the specific component, ‘fof dry- and wet-pelletized carbon blacks, ranging from N110 to N990, were evaluated by inverse gas-solid chromatography at infinite dilution. The results indicate that the dis- persive components of the surface energy of carbon blacks increase with increasing surface area. This dependence may essentially reflect an effect of microstructure on the surface energies, which can be confirmed by the relationship between the crystallographic parameters of crystallites and the graphitization of the carbon blacks. It was found that smaller crys- tallites characterized by a lower value of L- lead to higher surface energy, whereas graphit- ization of the carbon black points toward lower surface energy, perhaps resulting from the growth of the quasi-graphite structure, Surface area dependence of the specific component of the surface energy characterized by the specific energy of adsorption of a polar probe follows the same pattern as was observed for the dispersive component, i.e, 7 increases with surface area. This is believed to be related to the crystallographic structure and the surface chemistry. Studies on adsorption energies of the low-molecular-weight analogs of elastomers gen- erally show that the interactions between carbon blacks and rubbers depend not only on filler surface energies but also on the structure of the elastomers. Due to their polar functional groups, NBR and SBR show a stronger interaction with blacks than unsaturated rubbers. Among the rubbers simulated, IIR would have the lowest interaction with the filler. ‘A comparison of the surface energies of carbon blacks and silicas points toward a very high yf for blacks which may show strong interaction with nonpolar- or low-polar polymers, while the very high S; value of the silicas, especially precipitated silicas, a measure of the relative polarity of their surface, is considered to be representative of strong particle-particle interaction, leading to the formation of a filler network. ACKNOWLEDGMENT ‘This investigation was carried out at the Deutsches Institut fiir Kautschuktechnologie, Hannover, Germany. We gratefully acknowledge the assistance of Dr. R. Badura and Dr. R. H. Schuster. REFERENCES “J.-B, Donnet and A. Voet, “Carbon Black, Physics, Chemistry, and Elastomer Reinforcement," Marcel Dekker, New York, 1976. 2G, Kraus, Ed,, “Reinforcement of Elastomers,” John S. Wiley and Sons, New York, 1965. *M. -J. Wang, 8. Wolff, and J.-B. Donnet, Ruuper Cea. Tzcaxor. 64, 559 (1991). 45. H De Boer, “The Dynamical Character of Adsorption,” Oxford University Press, London, 1953, 5G, M. Dorris and D. G. Gray, J. Colloid Interface Sci. 71, 99 (1979), ©. Janzen, RuaneR Cia, TECHNOL. 88, 669 (1982). 7 4.-B, Donnet, unpublished “V. A Garten and D. E, Weiss, Aust. J. Chem. 10, 309 (1957). ®R.B, Anderson and P. H. Emmett, J. Phys. Chem, 66, 753 (1952). %©M.T. Coltharp and Hackermann, J. Phys. Chem. 72, 1171 (1968). " D, Rivin, Runner Cen, TechwoL. 44, 307 (1971), 'W, HL Wade and MLL, Deviney de., RUBBER Cis, TECHNOL. 44, 218 (1971) © WH. Wade, M. L. Deviney Jr., W. A. Brown, M.H. Hnoosh, and D. R. Wallace, RusweR Cem. TucuNoL. 48, 117 (1972) “8M. Dannenberg, Ruse Cres. Tecunot, 69, 512 (1986). GH. Kilian, unpublished. © A. E. Austin, Proc. $rd Cong. Carbon, p. 389, 1958, © M, Gerapacher and C. M. Lansinger, Paper No. 7 presented at a meeting of the Rubber Division, American Chemical Society, Dallas, Texas, April 19-22, 1988. ™ p. Staszeauk, Mater. Phys. 14, 279 (1980), ° A.C, Zettlemoyer, in “Hydrophobic Surfaces," F. M. Fowkes, Rd., Academic Press, 1969, p. 9 © ©, Cazeneuve, Thesis of Docteur Ingénieur, Université de Haute Alsace, France, 1980. * J.B, Donnet, Carbon 25, 757 (1987), #8. Dong, Ph.D. Thesis, Université de Haute Alsace, France, 1989, *G, Guilpain, Ph.D. Thesis, Université de Haute Alsace, France, 1988. MY. Todani and 8. Sagye, Nippon Gomeu Kyokaishi 46, 1081 (1973),