摘要:

Heats of Mixing and the Solid-state Transition in [(C,H,)3PCH3],-~[(C,H,),AsCH3I~(TCNQ)i, (0 X I), Anion Radical Salts BY Y~ICHI IIDA* Department of Chemistry, Faculty of Science, Hokkaido University, Sapporo 060, Japan Received 4th March, 1977 The phase transitions of solid solutions were investigated with the anion radical salts of [(C6H5)3PCH3]1-i [(C~H&ASCH~]; (TCNQ):, (0 < x < 1). The behaviour of the phase transi- tions with respect to the composition parameter (x) was studied by using a regular binary solid solution model. It was found that the difference in the heats of mixing between the low-temperature (ccy) and high-temperature (&) phases of the solid solutions was very important in theexplanation of the line shape of the observed phase diagram. Much attention has been paid to the solid anion radical salts of 7,7,8,8-tetra- cyanoquinodimethane (TCNQ) because of their prominent electronic properties. 1-9 In particular, the anion radical salts containing mixed cations represented by [(C6H5)3PCH3]1-: [(C,H,),AsCH,]; (TCNQ);, (0 < x < l), are known to under- go phase transitions at 1 atm pressure in the solid state, and show discontinuities in 440 I I I 700 600 500 rl 400 % E CI 300 Q 200 100 0 0.0 0.2 0.4 0.6 0.8 1.0 composition parameter, x FIG.1 .-The experimental relations (a) of transition temperature (Tc) to composition parameter (x), and (6) of heat of the phase transition (AH) to x in[ (C6H5)3PCH3]+-: [(C6H5)3hCH3]+z (TCNQ)T, (0 Q x =+G l), anion radical salts at 1 atm pressure. The solid h e of (a) shows the theoretical relation between Tc and x , as predicted by eqn (4), see ref.(6) and text. 190Y. IIDA 191 the temperature dependence of the electrical conductivity and magnetic susceptibility at the transition temperat~re.~” The phase transition of pure methyltriphenyl- phosphonium salt, where x = 0.00, takes place at 315.7 K. Heat-capacity measure- ments of this phase transition have been made by Kosaki et d3 The transition has thus been found to be of the first order. The enthalpy and the total entropy change associated with the phase transition were experimentally determined to be 485.18 cal mol-1 and 1.7206 cal deg-l mol-l, re~pectively.~ For the mixed crystals, we earlier found that the transition temperature (Tc) increased, while the magnitude of the heat of transition (AH) decreased, progressively with an increase in the composition parameter (x), and that pure methyltriphenylarsonium salt, where x = 1.00, has no such phase transition up to a decomposition temperature of ~ 4 8 0 K at 1 atm pressure.2-6 The experimental results of this thermodynamic behaviour plotted against the composition parameter are illustrated in fig.1. In this paper, we have attempted to understand the mechanism of the phase transitions of [(C,H,),PCH,],-~ [(C6H5)&CH3]: (TCNQ)?, (0 < x < l), anion radical salts and to explain their experimental phase diagram in fig. 1 by the use of a regular binary solid solution model. The difference in the heat of mixing between the high- and low-temperature phases of those solid solutions was derived from the experimental relation between AH and x and also by analysing the phase diagram in terms of such a model.THEORETICAL In a previous paper,6 we proposed a thermodynamical theory of an ideal solid solution model for [(C,H,),PCH,],-~ [(C,H,),AsCH,]f; (TCNQ)?, (0 < x < l), anion radical salts. This model assumes that the phase transition of a solid solution does not change the manner of ideal mixing of the two components, because the methyltriphenylphosphonium and methyltriphenylarsonium cations are so bulky that we cannot expect cation exchange in the phase transition. In this respect, we have to note that the phase transition of our system is not the usual order-disorder type with respect to the mixing of the. two components.Moreover, we could assign the low- and high-temperature phases of the solid solutions at 1 atm pressure as ccy and phases, respectively.6 Their Gibbs free energies per mol were expressed by G’Y(T,p, X) = (1-X) G:(T,p)+xG$ (T,p)+ RT ((1 -x) In (1 --x)+x In x], (i = a, p), (1) where (1 - x ) and x are the mole fractions of the component [(C6H5),PCH3]+ (TCNQ)? and [(C,H5)3AsCH3]+ (TCNQ);, respectively. G;((T, p ) and Gf(T, p ) are the Gibbs free energies per mol for the low-temperature (a) and high-temperature (p) phases of pure phosphonium salt, respectively, while G;(T,p) is that for the (y) phase of pure arsonium salt at 1 atm pressure. In order to explain the experimental relation between T , and x of fig. 1, this kind of ideal solid solution model is not sufficient,’ because the phase transition is related to the cooperative interaction between the two components of methyltriphenyl- phosphonium and methyltriphenylarsonium salts.Therefore, we introduce into eqn (1) a heat of mixing effect between the two components and consider a regular solid solution model, as expressed by GiY(T, p , x) = (1 -x) Gi(T, p ) + xG,Y(T, p ) +Hz(l - X ) * X + RT ((I -x) In (1 - x ) + x In x}, (i = a, p), (2) where Hz is the heat of mixing per mol for each phase, and is assumed to be independent of temperature.192 PHASE TRANSITIONS I N TCNQ SALTS First, we examine how the effect of heat of mixing may influence the heat of transition for the ay 4 By phase transition of the solid solutions. If we do not consider any heat of mixing effect, as expressed by eqn (I), the heat of transition per mol, AH, against x of the solid solution is simply given by AH = (1 -x).(Hf - H;"), where H: and Hf are the enthalpies per mol for the a and /? phases of pure phss- phonium salt, respectively.6 Therefore, if the (Hf -H;) value remains constant, the heat of transition AH of the solid solutions will have a maximum value of 485.18 cal mol-1 for the salt with x = 0.00,3 decrease linearly with increase in x, converging to zero at x = 1 .OO.However, fig. 1 shows that, although the AH value is 485.18 cal mol-1 for x = 0.00 and converges to zero at x = 1.00, the experimental relation between AH and x is different from the predicted linear relation, having a broad hump over the region of 0.00 < x -= 1.00. This deviation in the heat of transition may arise from the difference in the heats of mixing between the ay and /?y phases, that is, (HB,'-H;)-(l - x ) - x .If this term is fitted to the experimental hump of the deviated heat, the ( H g - H z ) value is estimated to be 180+ 60 cal mol-I. Next, we consider the phase diagram of the solid solutions at 1 atm pressure. For the ay + B y phase transition, the phase equilibrium condition at the transition point is GaY(T,p, x) = GBY(T,p, x ) . If the temperature and the composition para- meter vary as T 4 T+ dT and x -+ x+ dx under a constant pressure of p = 1 atm, then GaY+dGar = GB"+dG@Y. In this case, from eqn (2), dGaY and dGBY can be expressed by eqn (3) dG" = - ( l - ~ ) S i dT-Gi dx-xSi dT+G,Y dx+H2(1-2x)dx+ R((1 -x) In (1-x)+x In x) dT+RT In - (1 ") dx, where i = a or j?, and where Si or S2y is the entropy per mol for each phase.There- fore, the condition dGaY = dG@Y leads to a theoretical relation between the transition temperature (Tc) and the composition parameter (x), as expressed by eqn (4) dTc - (GY-G~)+(HE-HZ)(~-~X) - - !dx (SE -s;)(l -x) (4) For pure phosphonium salt, where x = 0.00, the experimental values of Tc and Sf-ST are 315.7 K and 1.7206 cal deg-l mol-l, re~pectively,~ but since G4 - Gf = 0, the slope value of eqn (4) is reduced to (dTc/dx) = (HE-Hky)/(Sf-S;) at T, = 315.7 K. From fig. 1, the experimental slope value of (dTJdx) at x = 0.00 is found to be 82.5 & 2.5 K per composition unit, so that the value of f12 - HZ is estimated to be 142 5 cal mol-1 by putting Sf - S; = 1.7206 cal deg-l mol-1 value into the reduced equation.The (HF-HZY) value thus estimated is found to agree, within experimental error, with the 180 +_ 60 cal mol-1 value which was previously derived from the relation between the heat of transition and composition parameters of the solid solutions. We consider the phase diagram of the solid solutions. As the value of x increases, the denominator value of (Sf- Sq).(l -x) in eqn (4) will be positive and decrease linearly, while the (G;--G!) value of the numerator will be positive and increase progressively, because the a phase of pure phosphonium salt is the unstable phase in the temperature range above 315.7 K. On the other hand, the (Hg--HkY)*(l-2x) term of the numerator is greatest (i.e., Hg-HzY) at x = 0.00, decreases linearly with the increase in x, and is zero at x = 0.50.In the range 0.50 < x < 1.00, this (HE - Hz).( 1 - 2x) term becomes negative, decreases linearly with increase in x and has the most negative value of -(HE--Hky) at x = 1.00. These situations areY. IIDA 193 demonstrated schematically in fig. 2, where the (a) line shows the relation of (G; - Gf) to x, while the (b) line, that of (H:--H:Y)*(1-2x) to x. Then, as is shown by the (c) line of fig. 2, the net value of the numerator, (G: - Gf) + (H:-H27)*(1-2~), in eqn (4) will be greatest at x = 0.00, gradually decrease with the increase in x and have a finite positive value at x = 1.00. As has been mentioned, the denominator value of eqn (4) will be positive and decrease linearly with increase in x, converging to zero at x = 1 .OO.Therefore, the value of the slope (dTc.dx) in eqn (4) will be 82.5 & 2.5 K per composition unit at x = 0.00 and gradually increase with increase in x, diverging to infinity for the salt with x = 1.00. This theoretical relation between Tc and x is schematically demonstrated in the solid (a) line of fig. 1. composition parameter, x FIG. 2.-A schematic energy diagram for the components of the numerator of eqn (4) against the composition parameter ( x ) in the solid solutions of [(C6H5),PCH3I1-~ [(C6H5)3A~CH3]: (TCNQ)i, (0 < x < l), anion radical salts (see text). (a), G;- Gf ; (b), (Hg-H&”) (1 -2x) ; (c), (a)+@). On the other hand, the experimental relation between Tc and x of the solid solutions at 1 atm pressure, as given in fig.1, shows that the Tc value is lowest (i.e., 315.7 K) at x = 0.00 and that the slope value of (dTc/dx) has a finite positive value at this point. Moreover, this slope value increases gradually with increase in x, and the T, value reaches very high temperatures (presumably up to infinity) for the salt with x = 1.00. Therefore, the above-mentioned theoretical phase diagram for the solid solutions can explain well the experimental relation between Tc and x. of [(C6H5),PcH3],-z [(C6H5),ASCH3]z (TCNQ);, (0 < x < l), anion radical salts DISCUSSION In this section, we consider the meaning of the estimated (HE-HZY) value on the basis of eqn (2). As for the phosphonium and arsonium components of the low-temperature (ay) phase of the solid solutions, the crystal and molecular structures as well as the chemical properties of the a phase of the phosphonium salt are known to be very similar to those of the y phase of the arsonium salt.8 This implies a very low value of H z in eqn (2) for the ay phase.However, as for the phosphonium and arsonium components of the high-temperature (By) phase of the solid solutions, only the structural change of the a phase into the phase takes place in the phosphonium component, while there is no change in the arsonium part.6 The crystal and molecular structures of the /? phase differ from the a phase, in that remarkable internal rotation of the phenyl substituents takes place in the methyltriphenylphosphonium cation, while no significant difference was observed in the TCNQ parts.9 This fact means that there exists appreciable structural difference between the /? phase of the 1-7194 PHASE TRANSITIONS I N TCNQ SALTS phosphonium salt and the y phase of the arsonium salt. This situation will lead to, in eqn (2), a large value of HZ for the B y phase of the solid solutions. The difference in the heat of mixing between the ay and j3y phases then makes an important contribu- tion, not only to the heat of transition, but also to the line shape of the phase diagram of those solid solutions. L. R. Melby, R. J. Harder, W R. Hertler, W. Mahler, R. E. Benson and W. E. Mochel, J. Amer. Chem. SOC., 1962, 84, 3374. W. J. Siemons, P. E. Bierstedt and R. G. Kepler, J. Chem. Phys., 1963,39,3523 ; R. G. Kepler, J. Chem. Phys., 1963,39,3528. A. Kosaki, Y . Iida, M. Sorai, H. Suga and S. Seki, Bull. Chern. SOC. Japan, 1970, 43, 2280. Y. Iida, J. Chem. Phys., 1973, 59, 1607, and the references therein. Y . Iida, Bull. Chem. SOC. Japan, 1970, 43, 3685. Y . Tida, J. Phys. Chem., 1976,80,2944. ’ Y . Iida, Mol. Cryst. and Liq. Cryst., 1977, 39,195. A. T. McPhail, G. M. Semeniuk and D. B. Chesnut, J. Chem. SOC. A, 1971, 2174. M. Konno and Y . Saito, Actu Cryst., 1973, B29,2815. (PAPER 7/387)

ISSN:0300-9599    

DOI:10.1039/F19787400190

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Hydrogenation of Acetylene over Supported Metal Catalysts Part 1 .-Adsorption of [14C]Acetylene and [14C]Ethylene on Silica Supported Rhodium, Iridium and Palladium and Alumina Supported Palladium BY ASAD S. AL-AMMAR AND GEOFFREY WEBB* Department of Chemistry, The University, Glasgow, G12 SQQ Received 28th April, 1977 The adsorption of [' 4C]ethylene and [ 14C]acetylene on supported palladium, rhodium and iridium catalysts occurs irreversibly at 298 K in two distinct stages ; a non-linear primary region, in which the species are predominantly dissociatively adsorbed, and a linear secondary region. Hydrogenation catalysis is associated with the hydrocarbon species adsorbed on the secondary region. From [ L4C]carbon monoxide adsorptions it is concluded that the hydrocarbon primary region is associated with the metal, whilst the secondary region probably involves the formation of overlayers on the primary adsorbed species.The co-adsorption of ethylene and acetylene shows that, under acetylene hydrogenation conditions, both are adsorbed at separate sites and undergo hydrogenation inde- pendently of each other. The relevance of these observations to the selective hydrogenation of acetylene is discussed. The metal-catalysed hydrogenation of acetylene has been the subject of a large number of studies which have been reviewed by Bond and Wells and, more recently, by Wells and by Webb.3 In general the selectivity for ethylene formation has been considered to be a consequence of a thermodynamic factor, which controls the relative surface coverages of acetylene and ethylene, the same surface sites being assumed to be involved in the adsorption of both hydrocarbons, and a mechanistic factor, which determines the extent to which the ethylene is further hydrogenated to ethane without undergoing intermediate de~orption.~'~ In a recent study7 of the hydrogenation of acetylene in the presence of large excesses of ethylene over palladium + alumina catalysts it has been suggested that the ethane is formed by the hydrogenation of ethylene, rather than directly from acetylene, and that this hydrogenation may occur at sites which are independent of those involved in the adsorption and hydrogenation of acetylene.Similarly, in the selective hydro- genation of propadiene over a series of silica-supported metals * and of buta- 1,3-diene over alumina-supported palladium, rhodium and platinum, it has been suggested that the sites responsible for alkene formation may be distinct from those involved in alkane formation.It has also been suggested that the availability of hydrogen at these distinct sites may be different, possibly due to the occlusion of hydrogen within the metal. In a previous study of the adsorption of [14C]acetylene and [14C]ethylene on supported rhodium catalysts,10 it was observed that the adsorption isotherms showed two distinct regions : a steep primary region which, from related [14C]carbon monoxide adsorption, was considered to represent monolayer coverage of the metal by hydrocarbon, and a linear region thought to be associated with the support.It was also observed that acetylene could only displace 3 10% of the preadsorbed ethylene from the surface. 195196 ADSORPTION OF [14C]ACETYLENE AND [14C]ETHYLENE The present studies were undertaken in an attempt to determine (a) the relevance of the primary and secondary adsorption regions to the hydrogenation of ethylene and acetylene, and (b) to examine the co-adsorption of ethylene and acetylene under hydrogenation conditions. From such studies it was hoped to gain a further insight into the factors which determine the selective behaviour of metal catalysts in hydro- genation reactions. EXPERIMENTAL CATALYSTS The catalysts containing 5 % w]w metal supported on Aerosil silica (Degussa) or y- alumina (Degussa) were prepared by adding an aqueous solution of the metal chloride, containing the required weight of metal, to an aqueous suspension of the support.Excess water was evaporated and the catalyst finally dried in an air oven at 393-413 K. The catalysts were stored as the supported salt until required. Before use the supported salt was reduced in a stream of hydrogen ( m 10 cn13 min-') at 473 K for 12 h. Activation was completed by heating in an atmosphere of hydrogen at 623 K for a further 6 h. Finally the catalyst was evacuated at 623 K for 6 h and cooled to ambient temperature in vacuo. This procedure lead to reproducible activity from sample to sample. MATERIALS Acetylene (B.O.C.) contained both acetone and air. These were removed by a series of bulb to bulb distillations; the purified acetylene contained no impurities detectable by gas chromatography. Ethylene (Matheson) contained no detectable impurities and was merely degassed before use.[14C]labelled hydrocarbons (Radiochemical Centre, Amersham) were diluted to the required specific activity with the purified nonactive hydrocarbon before use. [14C]~arbon monoxide was prepared by the reduction of [14C]carbon dioxide (Radio- chemical Centre, Amersham) with metallic zinc.'l Cylinder hydrogen was purified by diffusion through a heated palladium thimble. APPARATUS The principal feature of the apparatus was that it permitted the direct observation of the adsorbed species during the adsorption process and in subsequent reactions. The reaction vessel is shown in fig. 1. The catalyst was spread uniformly on the glass boat (B), which FIG.1.-Reaction vessel used for the direct monitoring of surface adsorbed species. (GM, Geiger- Muller counter ; F, furnace region ; B, catalyst boat). could be moved into the furnace region (F) for activation and then withdrawn into position under the intercalibrated Geiger-Muller counters (Mullard MX 168/01) during the reaction. By suitable placement of the boat, the gas phase and the combined gas phase and surface count rates could be determined simultaneously.A . S . AL-AMMAR AND G . WEBB 197 The reaction vessel (volume 533 cm3) was connected to a mercury-free conventional high vacuum system, Pressures were measured using a calibrated differential transducer (Akers Electronics). The vessel was also connected to a gas sampling system (volume 5.5 cm3) which was then coupled to a combined gas chroinztograph-proportional c0unter.l This permitted the sampling and analysis of the reaction products throughout the course of the reaction.Analysis was performed using a 1 m column packed with 30-60 mesh activated silica ge!. The column was operated at 333 K with helium as carrier gas at a flow rate of 60 cm3 niin-l. On elution from the column the eluant was mixed with the required amount of methane before entering the gas proportional counter. EXPERIMENTAL PROCEDURE Adsorption of [14C]labelled species was investigated by admitting small aliquots of the labelled adsorbate to the reaction vessel and determining the surface and gas phase count rates after each addition. In the hydrogenation reactions a premixed sample of hydrogen and acetylene, in the ratio 3 : 1, was admitted to the reaction vessel to the required pressure.The progress of the reaction was monitored by the pressure fall observed with the transducer. At the desired pressure fall the reaction vessel was connected to the sample volume and a sample of the reaction mixture analysed. This procedure was repeated several times during the course of each reaction. RESULTS All adsorption and hydrogenation measurements were carried out at 294-298 K using hydrocarbon pressures in the range 0 to 12.5 Torr. On freshly reduced catalysts the adsorption isotherms for [14C]acetylene, ['"C]- ethylene and [14C]carbon monoxide were of the form shown in fig. 2. The shapes II 53 12 W 0 1 2 gas-phase count rate/10-3 min-' FIG.2.-Adsorption isotherms for [14C]carbon monoxide (@), [14C]acetylene (0) and [14C]ethylene (B) on freshly reduced (a) 0.20 g Rh/silica ; (b) 0.51 g Ir/silica ; (c) 0.10 g Pd/silica and (d) 0.10 g Pd/alumina. { [14C]specific activity was 0.10 mCi (mmol)-l in each case).2 98 ADSORPTION OF [14C]ACETYLENE AND [l4C]ETHYLENE of these isotherms are similar to those observed previously,1o the hydrocarbon isotherms consisting of a steep non-linear primary region followed by a linear second- ary region. With both acetylene and ethylene the gas in contact with the surface during the build up of the isotherm was analysed. This showed that with both acetylene and ethylene over rhodium and iridium, and with ethylene over palladium the gas in equilibrium with the surface primary adsorbed species consisted entirely of ethane.With acetylene over both palladium catalysts ethane was the sole gaseous species present up to 80% of the total primary region. During the subsequent build up of the isotherms the gas in contact with the surface wasethane, formed in the primary region, together with the adsorbate hydrocarbon. No ethylene was ever observed as a self-hydrogenation product from the adsorption of acetylene on any of the catalysts. gas-phase count rate/10-3 min-I FIG. 3.-Adsorption isotherms for [14C]carbon monoxide (E) ; [14C]acetylene (0) and [14C]ethylene (e) on steady state catalysts (a) 0.20 g Rh/silica ; (6) 0.51 g Ir/silica and (c) 0.10 g Pd/silica. ([14C]carbon specific activity was 0.10 mCi (mrnol)-l in each case). Catalysts which had been activated as stated above, but which had been allowed to cool to ambient temperature in hydrogen before evacuation, showed similar acetylene and ethylene isotherms to those shown in fig.2, except that the extent of the primary region was substantially reduced as shown in table 1. The gradients of the secondary isotherms were independent of the cooling-evacuation procedure. In the hydrogenation of acetylene over each catalyst it was observed that, from reaction to reaction, the catalyst became progressively deactivated until eventually a constant activity was attained. Thus, for example, with Pd/silica (0.10 g), Pd/alumina (0.10), Rh/silica (0.25) and Ir/silica (0.51 g) the activities, expressed as the first order rate constant (klmin-I), decreased from initial values of 1.4 x lo-', 1.45 x 2.70 x and 8.35 x to constant limiting values of 1.59 x 1.50 x 0.62 x and 2.70 x respectively.For catalysts which had been deactivated to this constant limiting activity, which will be termed the catalyst steady state, adsorption isotherms of [14C]acetylene, [14C]ethylene and [14C]carbon monoxide were determined. Investigations with [14C]acetylene showed that, although only a part of pre-adsorbed [14C]acetylene participated in a subsequent hydrogenation reaction, a further fraction of the pre-adsorbed acetylene could be removed by allowingA . S . AL-AMMAR AND G . WEBB 199 TABLE 1 .-EXTENT OF PRIMARY ADSORPTION OF [' 4C]ACETYLENE AND [' 4C]ETHYLENE ON CATALYSTS EVACUATED AT 623 K AND 298 K FOLLOWING A ~ A T I O N adsorbate temperature/K (0.1 g) (0.1 g) (0.2 g) (0.51 9) surface count rate at turning point/min-1 evacuation Pd/A1203 Pd /SiO z Rh/SiOz Ir/SiO2 C2H2 623 2610 293 3 5688 801 2 C2H2 298 1975 2442 3867 6391 C2H4 623 500 848 5232 6706 CzH4 298 190 325 3497 533 1 the catalyst to stand under either hydrogen or the hydrogen-rich reaction product mixture following the hydrogenation of 12.5 Torr acetylene with 37.5 Torr hydrogen.Consequently, the adsorption isotherms for the deactivated steady state catalysts, shown in fig. 3, were determined after the catalyst had been allowed to stand under the hydrogen-rich reaction product mixture for 6 h, this being the time for the com- plete removal of all of the " removable " preadsorbed acetylene. Comparison of theiisotherms shown in fig. 2 and 3 shows that catalysts in their steady states showed TABLE 2.-vARIATION OF f14C]ACETYLENE AND [ 14C]ETHYLENE SURFACE COUNT RATES WITH VARIOUS TREATMENTS AT 298 K gas pressure /Torr 1.98 0.51 1.12 5.07 6.46 1.37 3.77 0.63 0.67 0.59 0.61 0.46 0.71 2.00 0.3 1 0.59 surface count ratelmin-1 initial evacuation CzHz C& 2225 2179 2195 - 3518 3457 3463 - 7117 7090 7082 - - 820 782 - 1248 1212 1248 1248 4513 4369 4513 4513 2322 2267 2320 - 5655 5508 5655 - 482 482 481 - 2887 2823 2799 2887 4943 4921 4841 - 5112 5009 5112 5112 10350 10173 10350 10350 150 150 - - 180 178 82 180 1546 1528 - 1545 1612 1600 1612 - 2135 2112 1667 2135 3313 3267 2831 3313 9315 9298 9156 9315 1835 1794 1835 1835 3350 3297 3350 3350 Hz - I - 710 - 2176 4370 5513 - - - 3760 9023 - - - 1187 - - 1315 50 H2+C2H2 1060 1818 3557 - - - - - 115 2212 3 843 - - 25 - - - 875 1363 3838 - [l4C1 specific activities : 0.025 mCi(mmoI)-' (Rh), or 0.10 mCi (mmol)-l ( F W ) ; 0.05 mCi (mmol)-l (Pd) and 0.10 mCi (mmol)-l (Ir).200 ADSORPTION OF ['4C]ACETYLENE AND [l4C]ETHY LENE either a much reduced (- 50% with acetylene over palladium) or absent primary region in acetylene or ethylene adsorption. The steady state catalysts also showed a much reduced capacity for carbon monoxide adsorption compared with those of the freshly reduced catalysts. Evacuation for 1 h of steady state surfaces which had been equilibrated with varying pressures of [14C]acetylene or [I4C]ethylene had little effect on the surface count rates, as shown in table 2 (column 5 ) .Subsequent admission of 12.5 Torr non-radioactive acetylene to the [14C]acetylene precovcred surfaces, or of 12.5 Torr ethylene to the [14C]ethylene precovered surfaces, also had little effect on the surface count rate (table 2, column 6).Admission of 12.5 Torr acetylene to [l 4C]ethylene precovered surfaces, following evacuation of the gas phase [14C]ethylene, resulted in the displacement of small amounts of radioactivity, as shown in table 2 (column 6). With all catalysts this displacement was complete within 1 h and the amount of radioactivity displaced was independent of the surface concentration of the [14C]ethylene. I I I I 0 20 LO 60 80 103 time/& FIG. 4.-Effect of hydrogen on the adsorbed [14C]acetylene adsorption isotherms for Rh/silica (e), Ir/silica (H) and Pd/silica (0) in the steady state.The effects of hydrogen (37.5 Torr) on surfaces which had been equilibrated with various pressures of [14C]acetylene, the pressures being such that each steady state surface was covered to beyond the primary region, are shown in fig. 4. With each catalyst there is a rapid (10-20 min) initial removal of the secondary adsorbed species, followed by a further slow (- 2 h) decrease in the surface count rate. The decrease in count rate at the end of the rapid stage is independent of the initial surface con- centration of the secondary adsorbed acetylenic species, as shown in table 2 (column 8). Admission of 50 Torr of a hydrogen : acetylene (3 : 1) reaction mixture to a [ 4C]acetylene precovered steady state surface, following evacuation of the residual gas phase [l 4C]acetylene, yeilded similar results to those observed when hydrogen alone was used (table 2, column 9).In order to examine the adsorption of ethylene in the presence of acetylene, thereby obtaining information regarding the possible thermodynamic effect in acetylene hydrogenation, [14C]ethylene was admitted to a steady state catalyst in the presence of 12.5 Torr of gas-phase acetylene. As shown in fig. 5, the extent of adsorption of the ethylene on the steady state palladium and iridium catalysts wasA . S . AL-AMMAR AND G . WEBB 201 almost the same as that on a similar catalyst in the absence of acetylene. With rhodium, the presence of the acetylene reduced the adsorptive capacity of the surface for ethylene by z 30 %, although this value is much larger than the amount of [‘“CI- ethylene displaced by the admission of 12.5 Torr acetylene to an ethylene precovered steady state rhodium surface.The effect of acetylene hydrogenation upon the surface concentration of [‘“CI- ethylene was investigated by admission of 50 Torr of a hydrogen : acetylene (3 : 1) reaction mixture to a steady state catalyst which had been previously exposed to various pressures of [‘“Clethylene. When, before admission of the reaction mixture, the residual C1”C]ethylene in the gas phase had been evacuated from the reaction vessel, gas-phase count rate/10-3 min-I FIG. 5.-Adsorption of [14C]ethylene on (a) 0.51 g Irlsilica, (b) 0.20 g Rh/silica and (c) 0.10 g Pdlsilica in the steady state in the absence (0) and presence of 12.5 Torr acetylene (0).the surface count rate decreased quite rapidly on addition of the reaction mixture to a limiting value as shown in table 2 (column 9). However, when the residual [l”C]ethylene was allowed to remain in the reaction vessel, or 5 Torr of [14C]ethylene was admitted to the vessel after the reaction mixture, the surface count rate decreased only fractionally until the pressure fall in the system corresponded to the virtual removal of all of the acetylene. At this point, as shown in fig. 6 , the surface count rate decreased quite rapidly. During the acetylene hydrogenation the selectivity was 0.94 (Pd), 0.75 (Rh) and 0.16 (Ir), values similar to those reported previous1y.l Throughout the acetylene hydrogenation small yields of [I4C]ethane, which increased linearly with pressure fall, were observed.With a gas pressure of 5.0Torr [l4“C]- ethylene, the [I4C]ethane yield constituted 12.7 % (Pd), 9.5 % (Rh) and 2.8 % (Ir) of the total ethane yield. The reactivity of the acetylenic species adsorbed on the primary region of a freshly reduced catalyst was examined by covering the surface with [14C]acetylene to a point corresponding to the completion of the primary region. A mixture of hydrogen and202 ADSORPTION OF ['4C]ACETYLENE AND ['4C]ETHYLENE timelmin (0 and 0 ) 0 40 80 120 160 200 240 280 320 I I 8 1 01 I I I I I I I I I 0 20 40 60 80 100 120 140 160 timelmin FIG. 6.-Variation of the [14C]ethylene surface count rate during the hydrogenation of 12.5 Torr acetylene and 37.5 Torr hydrogen in the presence of 5 Torr [l4C]ethylene for silica-supported rhodium (O), iridium (a) and palladium (U).acetylene (3 : 1 ratio) was admitted to the catalyst to a pressure of 50 Torr and the surface and gas phase monitored for radioactivity. No change in surface count rate was detectable over a period of 6 h (time of reaction, 2-4 h), with any of the catalysts, although when the reaction product mixture was left in contact with the surface for 72 h, some of the surface species had been hydrogenated to [I4C]ethane, as shown in table 3. TABLE 3 .-EXTENT OF PERMANENT RETENTION OF [14C]ACETYLENE IN THE PRIMARY ADSORPTION REGION AT 298 K time of treatment in hydrogen = 72 h surface count ratelmin-1 % catalyst initial final retention PdIf3J203 2595 63 1 24 Pd/Si02 2905 491 17 Rh/SiOz 5696 1821 32 Ir/Si02 8007 3904 49 DISCUSSION From the results presented above a number of features emerge regarding the adsorption of acetylene and ethylene, both separately and in competition, on sup- ported palladium, rhodium and iridium catalysts.In particular, these results show that, in agreement with our previous observations using supported rhodium catalysts,l O over all the catalysts studied the hydrocarbon adsorption isotherms show two distinct regions : a steep primary region followed by a linear secondary region. The relevance of these two regions to hydrogenation will be discussed below. The results also show that when acetylene and ethylene are allowed to compete for the same surface,A . S . AL-AMMAR AND G . WEBB 203 each is adsorbed independently of the presence of the other.This observation sug- gests that, contrary to earlier suggestion^,^-^ thermodynamic factors may not play an important role in determining the selectivity observed in acetylene hydrogenation. From the adsorption isotherms shown in fig. 3 and 5 it is apparent that on steady state catalysts the adsorption of ethylene is limited to the secondary region. Similarly, for steady state rhodium and iridium catalysts the adsorption of acetylene is also limited to the secondary region, although with palladium there appears to be a part of the primary region which can be regenerated with hydrogen at ambient temperature. These observations, together with the low reactivity of the primary adsorbed acetylenic species and the independent hydrogenation of ethylene and acetylene, during the hydrogenation of the latter, indicate that the species of importance from the stand- point of hydrogenation is located on the secondary region.The question arises as to the nature of the species which constitute the primary and secondary regions and as to their location on the surface. The absence of an ethylene primary region on the steady state catalysts, suggests that in this region the acetylene and ethylene are adsorbed at the same surface sites, although as shown in table 1 the actual amounts of ethylene and acetylene adsorbed on these sites on freshly reduced catalysts differ widely from metal to metal. From the amounts of ethane formed during the primary adsorption of acetylene, the average composition of the adsorbed acetylenic species in this region can be calculated as C2Hle4(Pd) ; CzHl .,(Rh) and C2Hl.&r), assuming that, as observed previ- ~ u s l y , ~ ~ * l4 the amount of catalyst hydrogen available for hydrogenation is negligible. The composition of the primary adsorbed species, calculated as above, is also inde- pendent of the extent of coverage of the primary region. These results indicate that the species adsorbed on the acetylenic primary region are predominantly dissociatively adsorbed. The amounts of carbon monoxide adsorbed on the freshly reduced catalysts correspond to CO to total metal atom ratios of 4.27 (Pd), 2.44 (Rh) and 1.88 (ir, these values being similar to those expected from the particle sizes as determined by electron microscopy, which showed that the average particle sizes were 85 A (Pd), 25 A (Rh) and 18 A (Ir), assuming that the carbon monoxide is adsorbed in the linear mode.15 Further, in all cases the relative amounts of carbon monoxide and acetylene adsorbed on the primary region are in the approximate ratio of 2 : 1.These observations are consistent with the postulate that the primary adsorption occurs directly on the exposed metal and that, as suggested previously,1° the turning point in the adsorption isotherms corresponds to monolayer coverage of the metal with hydrocarbon. The small amounts of carbon monoxide adsorbed on the steady state rhodium and iridium catalysts, 5.9% of that adsorbed on the freshly reduced surfaces, can be interpreted in terms of the adsorption of carbon monoxide on single metal sites left vacant following the dissociative adsorption of acetylene. With the steady state palladium catalysts the extent of carbon monoxide adsorption is 59% of that of the freshly reduced catalysts, in agreement with the observation that, with this metal, the primary region is reduced to - 50% by the deactivation process.According to the above suggestions, the secondary adsorbed species arise from the adsorption of amounts of ethylene and acetylene in excess of that required for monolayer coverage. Indeed from the adsorption isotherms it can be deduced that, for example, at a gas pressure of 5 Torr of acetylene, the ratio of the adsorbed acetylene to the total number of metal atoms is in excess of unity. It was suggested previously lo that the secondary adsorption arose from " spill-over " of the hydrocarbon from the metal on to the support.However, the present results show that the extent of secondary adsorption is similar over all the catalysts and, with palladium-silica and204 ADSORPTION OF [14C]ACETYLENE AND [l4c]ETHYLENE palladium-alumina, appears to be independent of the support. Furthermore, on catalysts which have been treated with hexamethyldisilazane, which effectively covers the support surface with trimethylsilyl groups, the extent of secondary adsorption is the same as on untreated catalysts.16 Since, as noted by Levy and Boudart,17 the phenomenon of " spill-over " requires suitable sites on the support, it seems unlikely that the present observations can be satisfactorily interpreted in terms of " spill-over ".It has recently been suggested from LEED studies of acetylene adsorption on Pt(l11) surfaces that, at high pressures, the acetylene forms an overlayer on the initially adsorbed acetylene on the meta1.l' Although, in the present studies, the precise chemical nature cf the secondary adsorbed acetylene or ethylene could not be deduced, the results are not inconsistent with the suggestion that the secondary adsorbed species arise from the formation of overlayers on the primary adsorbed material. The results also lend support to the recent suggestions of Thomson and Webb regarding the importance of the formation of dissociatively adsorbed carbonaceous species in catalytic hydrogenation of unsaturated hydrocarbons by metals.Their model for hydrogenation depends upon hydrogen transfer be tween adsorbed C&, which is permanently retained on the surface, and the hydrocarbon undergoing hydrogenation. There is in this paper what seems to be convincing evidence for formation of the primary adsorbed species followed by further hydrocarbon adsorp- tion on that layer and the subsequent hydrogenation of this secondary species. The observation that, even with the secondary adsorbed ethylene or acetylene, the adsorbed hydrocarbon would not undergo molecular exchange with gaseous hydrocarbon, nor could it be removed by evacuation, shows that the adsorption was effectively irreversible. Whilst this conclusion is in agreement with those drawn far acetylene adsorption from studies of acetylene-deuterium exchange,20 it is in sharp contrast with the behaviour of ethylene adsorption over alumina-supported rhodium and palladium catalysts at low temperature,21 as deduced from ethylene- deuterium exchange studies, where it was concluded that ethylene adsorption was readily reversible.These observations, together with the lack of displacement of any appreciable amounts of ['4C]ethylene by acetylene, may suggest, at first sight, that the species formed during the adsorption process are of little relevance to hydro- genation. However, this is not the case, since the results obtained are independent of whether or not the catalyst was evacuated at high temperature following activation, or, more importantly, whether or not hydrogen was present during the adsorption of the [14C]ethylene.The results obtained from the admission of [14C]ethylene to a catalyst in contact with an acetylene + hydrogen reaction mixture were identical with those obtained when the ethylene was admilted first. Rather surprisingly, the only effect of preadsorbed hydrogen, assumed to be present after the catalyst was allowed to cool to ambient temperature in hydrogen before evacuation, was to reduce the extent of the primary region. Whilst the reasons for this are not readily apparent, it would appear that the presence of preadsorbed hydrogen in some way acts as a poison for the formation of the dissociatively ad- sorbed primary species. From the behaviour of added [I Tlethyiene during acetylene hydrogenation we conclude that the adsorptions of acetylene and ethylene occur at independent sites on the secondary region.Since, with palladium and iridium, the extent of ethylene adsorption is the same in the presence or absence of acetylene, it may be concluded that the differences in the rates of ethylene hydrogenation in the presence and absence of acetylene, observed previously 5 3 and in the present work,16 are not due to any variation in the surface coverage of the hydrocarbon, but are probably due to differ- ing hydrogen availabilities in the two systems. Similar effects have been observedA . S . AL-AMMAR AND G. WEBB 205 in the competitive hydrogenation of cycloalkenes over metal catalysts.22 This conclusion will be discussed more fully in a subsequent paper, when the effects of added ethylene and of the poisoning effects of carbon monoxide on acetylene hydro- genation will be discussed.In conclusion, the studies reported in this paper have shown that on the catalysts studied, the hydrogenation of acetylene and ethylene is associated with the secondary adsorbed species, which are not adsorbed directly on the metal, but which probably exist as an overlayer on dissociatively adsorbed hydrocarbon. Further, under acetylene hydrogenation conditions, the adsorption isotherms show that the con- centration of adsorbed species is directly proportional to the gas phase pressure up to pressures > 22 Torr. The competitive adsorption of ethylene and acetylene also shows that in the secondary region on the surface, the adsorption of ethylene is not affected by the presence of acetylene and, therefore, that thermodynamic effects of the type suggested previously are unlikely to play any part in determining the selectivity in acetylene hydrogenation.The authors are grateful to the Government of Iraq for financial support to one of us (A. S. Al-M.). We are also grateful to Prof. S. J. Thomson for many helpful and stimulating discussions. G. C. Bond and P. B. Wells, Adv. Catalysis, 1963, 15, 91. P. B. Wells, Surface and Defect Properties of Solids (Specialist Periodical Reports, Chemical Society, London, 1972), vol. 1, p. 24. G. Webb, in Comprehensive Chemical Kinetics, ed. C. H. Bamford and C. F. H. Tipper (Elsevier, Amsterdam), vol. 20, chap. 1, in press. G. C. Bond, D. A. Dowden and N. Mackenzie, Trans. Faraday Soc., 1958,54,1537. G. C. Bond and P. B. Wells, J. Catalysis, 1965, 4, 211 ; 1966, 5, 65 ; 1966, 5, 419. G. C. Bond, G. Webb, P. B. Wells and J. M. Winterbottom, J. Catalysis, 1962, 1, 74. C. Kemball, W. T. McGown, D. A. Whan and M. S . Scurrell, J.C.S. Faraday I, 1977,73,632. * K. C. Khulbe and R. S. Mann, Proc. 6th Irzt. Congr, Catalysis (London, 1976), preprint no. A35. A. J. Bates, Z. K. Leszeynski, J. J. Phillipson and P. B. Wells, J. Chem. Soc. (A), 1970, 2453. R. B. Bernstein and T. I. TayIor, Science, 1947, 21,498. lo J. U. Reid, S. J. Thomson and G. Webb, J . Catalysis, 1973, 30, 372. l2 G. F. Taylor, S. J. Thomson and G. Webb, J. Catalysis, 1968, 12, 191. l3 J. A. Altham and G. Webb, J. Catalysis, 1970, 18, 133. l4 Z . Paal and S. J. Thomson, J. Catalysis, 1973, 30, 96. l 5 G. Blyholder and C. Marvin, J. Amer. Chem. Soc., 1969,91, 315. l 6 A. S. Al-Ammar and G. Webb, unpublished results. R. B. Levy and M. Boudart, J. Catalysis, 1974, 32, 304. E. L. Kesmodel, P. C . Stair, R. C . Baetzold and G . A. Somorjai, Phys. Rev. Letters, 1976,36, 1316. l 9 S. J. Thomson and G. Webb, J.C.S. Chem. Comm., 1976, 526. 2o G. C. Bond and P. B. Wells, J. Catalysis, 1966, 6, 397. 2 1 G. C. Bond, J. J. Phillipson, P. B. Wells and J. M. Winterbottom, Trans. Faraday SOC., 1966, 22 J. K. A. Clarke and J. J. Rooney, Adu. Catalysis, 1976, 25, 125. 62,443. (PAPER 7/714)

ISSN:0300-9599    

DOI:10.1039/F19787400195

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Acidic Properties of Mixed Tin+ Antimony Qxide Catalysts BY ELIZABETH A. IRVING AND DUNCAN TAYLOR* Chemistry Department, University of Edinburgh, Edinburgh EH9 355 Receiwd 27th Muy, 1977 Mixed oxides of tin and antimony have been used as catalysts in a static system and outgassed both at room temperature and at 698 K in a study of the approximately zero order stages of the isomerization of 3,3-dimethylbut-l-ene (373 K), cyclopropane (41 1 K), but-1-ene (293 K) and cis- but-2-ene (293 K) and of the dehydration of isopropanoI(343 and 408 K). With catalysts outgassed at room temperature, weakly acidic sites are present, and all the reactions probabiy occur by a carbonium ion type of mechanism with Bronsted acid sites as a source of protons. Rates increase to a maximum as the antimony content increases from zero to NN 50 atomic %, and then decline with further increase in the antimony content.Outgassing of the catalysts at 698 K increased the rate of isornerization of 3,3-dimethylbut-l-ene, but for cyclopropane and isopropanol decreased rates were observed due to poisoning by the propene product. For but-1-ene and cis-but-2-eneY the higher temperature outgassing procedure changed the rate against catalyst composition pattern considerabIy in that only catalysts with less than 50 % Sb were active, and a mechanism involving an allyl inter- mediate is proposed. Catalyst activity could be poisoned by treatment with bases or with sodium acetate. It is concluded on the basis of a proposed correlation between rates and acidity, that the catalyst composition corresponding to maximum acidity is different from that for maximum selective oxidation activity.Mixed oxides of tin and antimony, besides possessing selective olefin oxidation properties,l have been shown to exhibit acidic activity in the propene-D,Q exchange reaction provided the oxides were not pre-exposed to the olefin. The carbonium ion mechanism of the exchange changed after pre-treatment of the oxides with propene to one involving an allyl intermediate. Bronsted acidity has been reported to account for butene isomerization reactions also. To clarify the nature and distribu- tion of acidic centres as a function of catalyst composition, a range of mixed oxides outgassed at both ,low and high temperatures has been used in a study of the isomerization of 3,3-dimethylbut-l-eneY cyclopropane and butenes, and of the decomposition of isopropanol.EXPERIMENTAL All reactions were carried out in an electrically heated 270 cm3 cylindrical Pyrex reaction vessel attached to a conventional gas-handling vacuum system. A gas sampling valve allowed ~ 0 . 5 % of the vessel’s contents to be removed periodically for analysis by G.L.C. In experiments with 3,3-dimethylbut-l-ene, a 2 m column of 35 % propylene carbonate on Chromosorb P was used at 293 K ; the same column was used at 303 K in the cyclopropane work. For isopropanol reactions, a 2 m column of 25 % carbowax 1500 on Chromosorb W AW/DMCS was used at 338 K, and for reactions of butenes a 3 m column of 28 % bis-2-methoxyethyladipate on Chromosorb P at 298 K.The range of tin + antimony oxide catalysts was selected from those described previously. All catalyst compositions are given as atomic percent of antimony in the total metal content. For each run a fresh sample of catalyst of weight between 0.2 and 1.0 g was outgassed by one of two procedures. In experiments described below as series I, catalysts were evacuated at 133 x N m-2 at room temperature for 5 h ; in series I1 experiments evacuation was for 16 h at 698 K. After the outgassing process, the reaction vessel was brought to the reaction temperature 206E . A . IRVING AND D . TAYLOR 207 in vacuo before admission of reactant (purity >99 %) to initial pressures of 1.6 kN m-2 for 3,3-dimethylbut-l-ene and cyclopropane, 1.46 kN m-2 for but-lene and cis-but-2-eneY and 0.67 kN m-2 for isopropanol.Except for isopropanol, a very rapid reaction occurred in the &st 1-2 min, during which 5-10 % of the reactant was consumed. The extent of this rapid reaction depended on both the reactant and the cataIyst composition. The rate of the next 10-20 % of the reaction % Sb Frc, ].-Rates of isomerization of 3,3-dimethylbut-l-ene at 373 K on mixed Sn+Sb oxides. (A) series I catalysts ; (B) series I1 catalysts. was close to zero order, and was taken as a measure of catalyst activity to construct the activity patterns of fig. 1-3. Later in the reactions, slower rates, again approximately zero order, occurred but activity patterns using the slower rates did not differ significantly from those in the figures. For isopropanol reactions, the zero order rate of the first 15-20 % of the decomposition was used.the reactant, and were shown diffusion effects. Rates were reproducible within f 4 to f 8 % depending on by use of different weights of catalyst not to be limited by FIG. 2.- % Sb catalysts. -Rates of propene formation from isopropanol at 408 K on mixed Sn + Sb oxides, series II208 ACIDIC PROPERTIES OF Sn+Sb OXIDES In studies of the effects of poisons on rates, catalysts were either treated with 10 % aqueous sodium acetate followed by washing, drying, and outgassing at 698 K, or, after outgassing at room temperature or 698 K, were exposed in the reaction vessel at the reaction temperature to pyridine (0.13 or 0.4 kN m-2 for 30 min) or 2,6-dimethylpyridine (0.13 kN m-2 for 60 min) followed by 30 min evacuation.i0Q- N B E ri I .- 0 20 4 0 2 0 40 10 5 0 % Sb FIG. 3.-Rates of isomerization of but-1-ene (A) and cis-but-Zene (€3) at 293 K on mixed Sn+Sb oxides, series I1 catalysts. RESULTS AND DISCUSSION Since isomerization of 3,3-dimethylbut-l-ene is known to proceed by a carbonium ion mechanism requiring uptake of a proton by the olefin, the results at 373 K shown in fig. 1 indicate the presence on the mixed oxides of Bronsted acid sites. The fact that a 9-fold decrease in rate occurred after the 26.8 % Sb catalyst had been poisoned by the sodium acetate treatment not only provides support for the occurrence of Bronsted acid sites, but also indicates that the distribution of rates across the catalyst composition range runs parallel to the distribution of acid sites.By Benesi’s method the catalysts were clearly shown to possess acidic properties, but quantitative measure- ments of acidity to confirm the rate-acidity correlation were not possible due to the grey-blue colour of the original mixed oxides. However, indirect support for the correlation was provided by the results of the isopropanol dehydration reaction given below, a reaction for which activity against acidity correlations for catalysts have been established.6*7 Comparison of curves A and B in fig. 1 shows that outgassing of the catalysts at 698 K instead of room temperature caused an increase in the number and/or activity of the acid sites, an effect which may be associated with the removal of adsorbed water molecules and the occurrence of surface phase changes.The only isomers formed were 2,3-dimethylbut-2-ene and 2,3-dimethylbut- 1 -ene in proportions ( 5 : 1) slightly greater than the calculated thermodynamic ratio ; hence the acid sites are to be regarded as weak rather than ~ t r o n g . ~ Decomposition of isopropanol at 343 K in series I experiments also indicated the presence of Bronsted acidity since rates of formation of propene gave an activity pattern similar to that of fig. l(A), the only difference being a sharper maximum at 49.5 % Sb. The mole percent propene in the decomposition products ranged from 85 to 100, and of di-isopropyl ether from 2 to 12, depending on the catalyst composi-E. A . IRVING AND D. TAYLOR 209 tion. Never mor2 than 3 % of the alcohol was dehydrogenated to acetone.In series 11 experiments at 408 K, rates of propene formation gave the pattern in fig. 2. In tb,: catalyst composition range 19.6-75 % Sb, propene was the main product (87-100 %), with small amounts of di-isopropyl ether. Tin oxide and the 6.1 % Sb catalyst, however, gave acetone in 90 % yield, but as the antimony content of the catalysts increased above 6.1 %, the acetone selectivity fell sharply. The dehydration mechanism most probably involves uptake of a proton by the alcoholic OH group, but in spite of the conclusion above that in series I1 the catalysts possess higher Bronsted acidity than in series I, propene formation rates in those series I1 experi- ments, where propene was the major product, were some five times less than in corresponding runs in series I.The slower rates in series I1 are believed to be due to preferential poisoning by propene, an effect which has been reported previously with the same catalysts for the propene-D20 exchange reacti0n.l In view of the reported occurrence of Lewis acid sites on SnOz after evacuation at 508 I<,9 the series I1 mixed oxides may also possess Lewis acidity to an extent modified by the antimony content. Accordingly, dehydration by Gentry and Rudham's mechanism lo involving OH- abstraction from the alcohol by a Lewis site may also be occurring, and could explain the small rate of propene formation found with SnO, and the 6.1 % Sb catalyst, since the latter catalysts, in view of the results in fig. 1, appear to have no Bronsted acidity. However, for Sn02 itself, dehydrogenation sites are clearly more numerous and/or active than dehydration Lewis sites.Treatment with pyridine at 408 K of those series I1 catalysts which gave both acetone and dehydration products in significant amounts showed that dehydration sites were poisoned to a much greater extent that those causing dehydrogenation. This illustrates the difference between the two types of site, besides again indicating the existence of acidic sites. Rates of isomerization of cyclopropane to propene at 41 1 K in series I experiments gave an activity pattern similar to that of fig. 1(A) but with a pronounced maximum at 49.5 % Sb. These results are consistent with a carbonium ion type of mechanism, again involving the Bronsted acidity of the catalysts. Parallel to the results for isopropanol, isomerization rates at 411 K for series I1 were lower (between 4- and 100-fold depending on the % Sb) than rates for series I, caused most probably by preferential poisoning of the series II catalysts by the propene product.The isomerization of but-1-ene and of cis-but-2-ene at 293 K over series I catalysts both gave activity trends not significantly different from that in fig. l(A), but with sharp maxima at 49.5 %Sb. But-1-ene reacted about seven times faster than the cis-isomer ; in each reaction the product ratio (cis : trans from but-1-ene, trans: I-ene from cis-but-2-ene) was always approximately four. Treatment of the catalysts with 2,6-dimethylpyridine partially poisoned both reactions, but without change in the product ratios.These results are adequately accounted for by carbonium ion mechanisms involving the Bronsted acidity of the catalysts. Reaction of the butenes at 293 K over series I1 catalysts, gave activity patterns as shown in fig. 3 quite different from those in fig. 1(B) and 2, and, therefore, indicate a change of mechanism from that in series I. Only those catalysts containing less than 50 % Sb were active, the maximum rate occurred at 19.6 % Sb for but-1-ene isomerization and at 8.7 % Sb for cis-but-2-ene ; the product ratios were respectively M 1 and ~ 4 . The but-1-ene reaction was strongly poisoned by 2,6-dirnethylpyridine and by but-1-en@ itself, and also by butadiene which was formed in ~ 0 . 2 % amounts during the isomerization. The much slower reaction of the cis-isomer was poisoned only to a small extent by the base, and no butadiene was formed.In neither reaction was the product ratio affected by poisoning. In view of the inactivity in series I1210 ACIDIC PROPERTIES OF Sn+Sb OXIDES of the 49.5 and 75 % Sb catalysts compared with their pronounced acidic activity shown in fig. 1(B), it is unlikely that the butene isomerizations occur mainly via carbonium ion intermediates involving the Bronsted acidity of the catalysts. Mechanisms involving allyl intermediates are more probable ; it is suggested that the poisoning effects are caused by strong adsorption of the poisons at those metallic ion sites required for adsorption of the allyl species. It is significant that in the selective oxidation of propene over the same series of catalysts,l a reaction known to involve an allyl species, maximum oxidation activity was observed for catalysts with the same range of antimony content as cause maximum rates of isomerization of the butenes.The main indications from this work are that (1) mixed oxides of tin and antimony exhibit Bronsted acidity, particularly in the composition range 25-75 % Sb, (2) the acidity present after degassing at room temperature is increased by degassing at 698 K, and (3) maximum acidity occurs in a composition range different from that where maximum propene selective oxidation activity is observed. Acknowledgements are made to the B.P. Co. Ltd. for the supply of catalysts, to Dr. C . C . McCain for stimulating discussions, and to the S.R.C. for the award of a maintenance grant (to E. A. I.). G. W. Godin, C. C. McCain and E. A. Porter, Proc. 4th In?. Congr. Catalysis, Moscow, 1968 (Akademiai Kiado, Budapest, 1971), vol. 1, p. 271 ; J. R. Christie, D. Taylor and C . C. McCain, J.C.S. Faraduy I, 1976, 72, 334. B. Hughes, C. Kemball and K. J. Tyler, J.C.S. Faraduy I, 1975, 71, 1285. F. Sala and F. Trifiro, J. Catalysis, 1974, 34, 68. H. Pines and W. 0. Haag, J. Amer. Chem. SOC., 1960, 82, 2471, 2488 ; C. Kemball, H. F. Leach, B. Skundric and K. C. Taylor, J. Catalysis, 1972, 27, 416. H. A. Benesi, J, Phys. Chem., 1957, 61,970; J. Amer. Chem. Soc., 1966,78,5490. K. Tanabe, SoZid Acids and Buses (Academic Press, New York, 1970), chap. 5. ' M. Ai and S. Suzuki, J. Catalysis, 1973, 30, 362. * J. E. Kikpatrick, E. J. Prosen, K. S. Pitzer and J. D. Rossini, J. Res. Nat. Bur. Stand., 1946, 36, 559. E. W. Thornton and P. G. Harrison, J.C.S. Faraduy I, 1975,71,461. l o S. J. Gentry, R. Rudham and K. P. Wagstaff, J.C.S. Furaday I, 1975, 71, 657. (PAPER 7/917)

ISSN:0300-9599    

DOI:10.1039/F19787400206

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Kinetics of the Decomposition and Hydrogen Reduction of Nitric Oxide on Niobium, Nickel and Platinum Filaments at High Temperatures and Low Pressures BY JOHN A. MORGAN AND ANDREW J. B. ROBERTSON* Department of Chemistry, King’s College, Strand, London WC2R 2LS Received 2nd June, 1977 NO decomposition kinetics were studied with ultra-high vacuum techniques on niobium, nickel and platinum wire filaments at high temperatures with pressures between -1 and 100 p N m-2. Niobium readily decomposed NO with first-order kinetics, nitrogen and oxygen being deposited on the surface, resulting in a progressive loss of filament activity. Nitrogen was desorbed from nickel, which otherwise behaved similarly to niobium. Excess H2 affected the kinetics of NO reaction on these metals mainly by reaction with and subsequent removal of surface contaminants rather than by a direct reaction with NO.Platinum did not observably decompose NO at low pressures except in the presence of excess Hz. Above -1300 K the rate appeared to be determined by Hz atomisation, but at lower temperatures a reaction between H2 and NO may have been occurring. The rate-limiting step for this reaction may be the decomposition of H2 and addition of H atoms to adsorbed NO. The pre-exponential term for this reaction indicated that the transition state was mobile on the surface. The decomposition of NO to N2 and O2 and its reduction by H2 to form N2 and M20 are both thermodynamically favourable processes, even at 1000 K and for pressures in the region of 1 , ~ N r n - ~ . However, these reactions are slow in the absence of a catalyst.Owing to the occurrence of NO as an important pollutant in automobile exhausts, in recent years a great deal of interest has been directed at research into the catalytic decomposition and reduction of NO by a variety of catalysts. Reviews by Shelef and Kummer and Dwyer have described such work and its applicability to the control of automobile emissions. The scale of interest in NO decompositions may be judged by the fact that Winter has studied the decomposition on forty metal oxides. However, apart from a few papers such as the adsorption studies by Yates and made^,^ work in this field has been performed in conventional high-vacuum conditions with NO pressures usually above 10 N m-? Research of a more fundamental nature on the interaction of NO with clean surfaces, involving the use of lower pressures and ultra-high vacuum (u.h.v.) techniques, is lacking.The present paper describes the investigation of the kinetics of NO decomposition and reduction by H2 on niobium, nickel and platinum-wire filaments at high tempera- tures. Ultra-high vacuum techniques were used and NO pressures were typically between 1 and 100 ,uN m-2. EXPERIMENTAL APPARATUS A Pyrex glass flow system was constructed, incorporating greaseless stopcocks and mercury cut-offs. The reactant gases passed from separate gas-handling systems into the reaction vessel through calibrated, fine, capillary leaks. The reaction vessel was a 500 cm3 Pyrex bulb in which the filament was suspended in the form of a coiled loop, lashed to 21 12 12 NITRIC OXIDE ON NIOBIUM, NICKEL A N D PLATINUM tungsten leads by fine copper wire.Residual u.1i.v. pressures were measured with a Bayard- Alpert type gauge (I.O.G. 12), while partial pressures of reacting gases were monitored with a small 180" mass spectrometer (V.G. Micromass 2) incorporating an LaB6 coated filament. The m a s spectrometer was calibrated with a h4cLeod gauge, which could be removed from the system when u.h.v. experiments were performed. H20 pressures were derived from relative sensitivity data supplied by the manufacturers. The reaction vessel was pumped by a large mercury diffusion pump, 50nm diameter mouth, which was preceded by two large refrigerated traps. The reaction vessel, inlet leaks and pressure measuring devices were baked by an oven at -6600 I(.The traps and tubing up to the mouth of the pump were baked with heating tapes. The system was normally baked overnight and then cooled in sections, starting at the pump. Residual pressures below 10 nN m-2 could be attained. MATERIALS NO (Air Products) had purity >99 %. Apart froin traces of N2, no impurities were detected by mass spectrometry. The NO was purified by a freeze-and-thaw method. Cylinder hydrogen was purified by diffusion through palladium. In studies with nickel and platinum, the gases before entry into the reaction vessel were passed through a trap cooled with solid C02 to remove mercury vapour, which might have had a contaminating effect on these catalysts. The 130 pm diameter nickel and 70 pm diameter platinum wires (Johnson-Matthey) were of spectroscopic purity.The 150 pm diameter niobium wire contained various bulk impurities listed previ~usly.~ Nickel and platinum were cleaned by degassing at high temperatures, followed by cycles of oxidation in % 100 puN n r 2 of O2 at GZ 1000 K and reduction at the same temperature in % 500 N m-2 of pure HL. Oxidation was omitted from the cleaning procedure for niobium. All filaments were thoroughly degassed at high temperatures prior to use in experiments. The filaments were z 30 cm long. REACTION RATES When reactant gases reached steady pressures in the reaction vessel, reaction rates at a given filament temperature could be calculated, with a knowledge of the appropriate pumping speeds, from the decreases in reactant gas pressures or increases in product gas pressures which had occurred through reaction.6 This assumed that rates were negligible with the filament at room temperature compared with rates at the elevated filament temperatures studied.Reaction orders in NO or H2 were determined by the differential method and by the isolation technique. First-order rates were expressed in terms of the probability P of reaction of a molecule at a single collision with the catalytic surface.6 The catalytic areas were assumed to be equal to the geometric areas of the filaments. P was expressed as Y = B exp (-EIRT). FILAMENT TEMPERATURES Temperatures above red heat were determined by optical pyrometry, using average values of the spectral emissivities, at I. = 0.65 pm, obtained by Jain et al.7* for nickel and platinurn. The spectral emissivity of niobium (0.36) had been determined experimentally, as previously described5 Temperatures below red heat were interpolated from plots of the variation of filament resistance or current with filament temperatures, obtained by using values at incandescent temperatures, 373 K and room temperature.Corrections to filament teinpera- tures for heat losses, including conduction of heat to the leads, were calmlated to be of negligible importance. RESULTS NIOBIUM DECOMPOSITION Clean new filaments were very reactive to NO. No gaseous products of decoin- position were observed at temperatures below ~ 2 0 0 0 K. At ~ 2 0 0 0 I< N2 and NJ. A. MORGAN AND A . J. B. ROBERTSON 213 atoms were desorbed.Although NO appeared to be readily chemisorbed, probably ciissociatively, the product surface species were too strongly bound to niobium to be desorbed, except at the highest temperatures. Reaction rates for NO disappearance from the gas phase were calculated from changes in NO molecular-ion peak heights and were first order in NO for pressures up to xl mNm-2. At higher pressures and for low temperatures, rates of NO reaction decreased with time until the filament eventually became inactive. Decays of activity were more rapid for higher pressures and lower temperatures. Filament activity could be restored, to some extent, by heating the filament at a higher temperature in VQCUO or in NO, or by reducing the NO pressure. Even after prolonged heating at temperatures above 2200 K in uacuo, filaments became less active with repeated use.Very probably the products of reaction not removed by desorption diffused into the filament interior and the bulk and surface concentration of these Y 4 6 8 10 12 10" KIT FIG. 1.-Series of Arrhenius plots for NO decomposition on a niobium filament. Pressures before reaction occurred were, : run 1, 53 pN m-2 (0) ; run 2, 40 pN m-' (0) ; run 5, 2.6 pN m-2 (a) ; run 6, 4 p N m-2 (a) ; run 8, 4 pN m-2 ( x ). Runs 3, 4 and 7 are omitted for clarity. contaminants increased with use. As values of P for NO decomposition on clean niobium were close to unity, P values on used filaments probably reflected the uncontaminated surface areas of these filaments. Fig. 1 shows sets of Arrhenius plots for a filament, illustrating the loss of activity with use in consecutive runs.Results on clean filaments could be expressed as P = (1.8f0.6) exp (- 14f4 kJ mol-l/RT). The results show that all the filaments were extremely reactive towards NO. The rapid increase of P with temperature, on used filaments for temperatures above = 1900 K, corresponded with the occurrence of nitrogen desorption. This suggested that the nitrogen desorption increased the uncontaminated niobium surface area available for reaction. An Arrhenius plot for N, desorption, using rates calculated from m/e 14 peaks heights to avoid complications from possible CO production, gave a desorption energy of 420 & 40 kJ mol-I.214 NITRIC OXIDE ON NIOBIUM, NICKEL AND PLATINUM 0.0 -0.5 Qi t4 0 - - 1.0- REDUCTION The atomisation reaction of H2 was important at temperatures where H2 affected the reaction rate of NO, and rates of reaction of H2 could not be satisfactorily determined by monitoring the m/e 2 peak.Because of the ready adsorption on the reaction vessel walls of H20 produced during reaction, rates calculated from H 2 0 production were unreliable unless the reaction was allowed to continue for consider- able lengths of time until the H20 pressure stabilised. Therefore, NO reaction rates were measured using changes in the mJe 30 peak height in the presence of excess M2. On new filaments excess of H2 had virtually no discernible effect on the probability of NO reaction. If the probability of reaction in the absence of H2 is the probability of chemisorption of NO on niobium, then this would be the maximum probability of NO reaction, and the existence of a bimolecular surface reaction between chemi- sorbed NO and H2 or H atoms would not increase the value of P for NO.The lack of increase in P in the presence of excess H2 also indicated that a possible reaction on the walls of the vessel between adsorbed NO and H atoms desorbed from the filament was unimportant compared with reaction rates on the filament. - - 0 I I I I I 1 4 6 8 10 12 Ik FIG. 2.-Arrhenius plots illustrating the effect of excess H2 on NO reaction probabilities on a used niobium filament : NO pressures 4 p N rnd2 before reaction. H2 pressures were : no H2 (0) ; 2 mN m-2 (0) ; 14 mN m-2 (0). On used filaments the presence of H2 caused a marked enhancement in activity towards NO for temperatures between z I100 and 1400 K compared with rates in the absence of H2 (fig.2). Above 1400 K, P for NO in the presence of H2 became almost constant with increasing temperature at a value of about 0.6, and the activity was comparable to that observed on a clean unused filament. For relatively new filaments the reaction products when H2 was present were mixtures of NH3, H20 and some N2, and some N atoms at higher temperatures. NH3 was the predominant product, but its importance decreased with the contamination of a filament. The more used, and therefore contaminated, filaments produced H20 as the major product. Possibly, surface nitrogen was more easily removed from the surface of these filaments by diffusion to the interior.J .A . MORGAN AND A . J . B . ROBERTSON 21 5 NICKEL DECOMPOSITION On admission of NO to the reaction vessel containing a cleaned nickel filament at room temperature, no NO was at first observed by mass spectrometry. However, a large m/e28 peak due to N, was observed, which decreased in height to a steady value after about an hour, while the m/e peak due to NO had then appeared and reached a steady value. The mass spectrum of the gas in the reaction vessel then corresponded approximately to that of NO. It seems likely that NO readily decom- posed on clean nickeI at room temperature, producing N2 as a gaseous product and oxidising the nickel surface (see also Onchi and Farn~worth),~ The nickel probably became less active fur NO decomposition as the surface coverage of the inhibiting oxygen increased and so the nickel eventually became inactive after an hour or so.The nickel film thrown on the walls of the reaction vessel daring high temperature degassing was also able to decompose NO at room temperature. - 1.0 - 4 5 M - -1.5 - -2.0 - c I I I 10 12 14 16 104 KIT FIG. 3.-Arrhenius plots for NO decomposition on nickel : pressures before reaction 80 (0) ; 26 (0) ; 11 (0) : 4 PN m-2 to). Because of the above effects, it was not surprising that NO decomposition rates were difficult to reproduce unless low NO pressures in the region of 1 to 10 pN m-2 were used. Decays of activity, similar to those observed on niobium, were observed for nickel filaments. To overcome these problems, a special procedure for experi- ments with nickel filaments was developed.NO was admitted to the reaction vessel at a pressure higher than that required and the room temperature oxidation of the nickel surfaces allowed to continue until a steady NO pressure resulted. The NO pressure was then reduced to the pressure required for study and the filament was heated to 1100 K for = 10 or 15 min to allow surface oxygen on the filament to diffuse into the interior. After this treatment, the filament was ready for kinetic studies, whereas the still oxidised thin nickel film on the walls was inactive. When steady rates were observed at very low NO pressures the reaction was first order in NO and the only gaseous product detected was N2, with nearly one molecule of N2 produced per two molecules of NO decomposed (at 960 K).Arrhenius plots for NO decomposition are shown in fig. 3, indicating the reproducibility at lower NO21 6 NITRIC OXIDE ON NIOBIUM, NICKEL AND PLATINUM pressures and the loss of activity at higher pressures. Rates in a run were measured at increasing filament temperatures. The most reliable results for NO decomposition on a clean nickel surface, obtained at the lowest NO pressures, corresponded to p = 101.3fO.1 exp (- 34 k 2 kJ mol-l/RT) for temperatures between x 570 and 890 K. Above 890 K the curves became flattened to give lower E and B values. REDUCTION When H2 was present during NO decomposition H20 was produced. Using the isolation technique with H2 in excess, the rate of NO reaction in molecules cm-2 s-l was found to be proportional to NO pressure at 1040 and 960 K.With NO in at least a hundred-fold excess, the reaction of NO became first-order in HZ, rates being determined by following the H20 production. However, under these conditions the nickel surface was almost certainly heavily oxidised and the rate of NO reaction was similar to that observed in the absence of H2. For H2 pressures of the same order of magnitude as the reacting NO pressure the order in H2 was zero. 8 10 12 14 10' K/T FIG. 4.-Arrhenius plots illustrating the effect of H2 on NO reaction probabilities on an oxidised nickel filament : NO pressures 26 p N m-' before reaction. Hz pressures were : 3 mN m-' for increasing (0) and decreasing temperatures (e) ; no Hz present (a). Except at the most elevated temperatures, excess H2 did not appear to enhance the rate of NO reaction on clean nickel filaments when low NO pressures were used.For higher NO pressures or on much used and oxidised filaments, P values were not steady with time, tending to increase slowly as the reaction continued. Rates were calculated from NO pressures and results depended greatly on the previous history of the filament, being less reproducible for the more oxidised filaments. In fig. 4 a typical set of results for an oxidised filament is compared with results on the same filament in the absence of H2. For temperatures between M 750 and 900 K, P for NO reaction in the presence of H2 increased rapidly with increasing temperature, while above 900 K, P became almost constant at ~ 0 . 5 . This represented an increase in filament activity over that observed in the absence of H2 on clean filaments for temperatures above 900 K.A similar increase was also observed for clean filamentsJ . A . MORGAN A N D A . J . B. ROBERTSON 217 in the presence of excess H2 and indicated that P values were not only increased in the presence of H2 by the removal of surface oxygen by H20 production, but possibly also by a surface reaction between NO and H2 or H atoms. PLATINUM DECOMPOSITION There was no detectable reaction of NO on platinum for filament temperatures below x 1650 K. At higher temperatures, a reaction involving disappearance of NO was observed which was approximately half-order in NO. No products of this reaction could be detected. Arrhenius plots were linear and rates could be expressed as v = 1047*2p&0 exp (-SS0540 kJ mol-l/RT) molecule m-2 s-l, where u is the reaction rate and pNO the NO pressure in N m-2. This reaction was unlikely to be occurring on the filament as the apparent activation energy was greater than the N-0 bond dissociation energy (626 kJ mol-f).10 Homogeneous decomposition of NO in the gas phase, after collision with the heated filament, should be negligible with such a high bond dissociation energy. Calculations also indicated that the probability of reaction by collisions of platinum atoms evaporated from the filament and NO molecules in the gas phase was negligible.Therefore, it seems possible that a trapping reaction of NO molecules by evaporated platinum was occurring on the reaction vessel walls. REDUCTION The reaction between NO and H2 was studied at temperatures below 1650 K.At 1230K a reaction occurred that was first-order in NO and zero-order in H2, unless the NO pressure was vastly in excess of the H2 pressure when half-order dependence on H2 was observed. With NO in excess the rates were not reproducible, and so kinetics were only studied for conditions with H2 in excess and of zero-order dependence in H2. The only reaction products detected were N2 and H20. Reaction probabilities were normally calculated from changes in the NO pressures. Arrhenius 104 K/T FIG. 5.-Typical Arrhenius plot of NO reaction on platinum in the presence of H2 : NO pressure 26 pN m-,, H, pressure 1.3 mN m-,.21 8 NITRIC OXIDE ON NIOBIUM, NICKEL AND PLATINUM plots were linear for temperatures below M 1250 K.A typical plot is shown in fig. 5. E below 1250 K was 45+2 kJ mol-l and B varied between ~ 0 . 5 and 1.5. These results were in reasonable agreement with rates calculated from rates of H 2 0 production, which gave E = 47 kJ mol-1 and B = 2.7. Above 1250 K, NO reaction probabilities calculated from changes in NO pressures corresponded to P = 107*0k2*2 exp (-213 & 12 kJ mol-l/RT). This change in slope of Arrhenius plots corresponded to the temperatures at which H2 atomisation was noticeable, and results above 1250 K may relate to a reaction between NO and H atoms on the filament or on the reaction vessel walls. Cooling the reaction vessel walls to * 193 K did not produce any noticeable changes in the rate of NO reaction. This might indicate that the reaction was occurring on the filament rather than the walls. However, the reaction H + NO + M -+ HNO + M has an activation energy very close to zero in the gaseous phase.A surface reaction between H and NO on the vessel walls would also be likely to have no pronounced temperature dependence; such a reaction may contribute to the observed reaction rate. The observed E for NO reaction above 1250 K is in good agreement with the E for H2 atomisation on platinum (214 kJ rnol-l).l2 In view of these considerations the results for temperatures below 1250 K were probably representative of a reaction between NO and H2 molecules rather than atoms. DISCUSSION Yates and Madey have reported that the chemisorption of NO on tungsten is non-dissociative at room temperature. Decomposition of NO was observed at higher temperatures by the detection of gaseous N2.It is likely that a similar process occurs on niobium and nickel, although desorption of N2 was only observed by us at the highest temperatures on niobium. The observed probabilities of reaction of NO on niobium and, possibly, nickel may, therefore, correspond to the probability of chemisorption of NO on the metal surface in its state during reaction. The absence of N2 desorption from niobium indicates the greater strength of the surface metal-nitrogen bond on niobium than on nickel. When a decomposition reaction occurs on niobium or nickel, the surface becomes increasingly contaminated with nitrogen or oxygen, or just oxygen, and thus the probability of reaction decreases with time and may eventually become too small to measure as the filament becomes completely covered by surface species.On used filaments, the rates of decomposition are limited by the relative rates of deposition of contaminants, and removal of these by desorption, diffusion from the filament surface, or by reaction with a reducing agent, such as Ha, which may produce desorbable products. In contrast to niobium and nickel, platinum did not measurably decompose NO at low pressures. P was therefore less than x Other workers have observed measurable rates of decomposition. However, such studies have been performed at NO pressures between w300 N and 66.5 kN m-2, and reaction orders have varied between zero, for saturated coverage, and second order for a bimolecular decomposition of NO.At the very low pressures, and hence very low surface coverages used in the present work, the probability of a bimolecular surface reaction of NO would be very low. Therefore, it appears that the unimolecular decomposition of NO by a dissociative adsorption is an unfavourable process on platinum. Hydrogen may enhance the probability of NO reaction on contaminated niobium by two processes : by reaction with surface contaminants, or by direct reaction with adsorbed NO. The absence of an increase in P for NO reaction on clean new filaments in the presence of excess €3, would tend to eliminate the latter reaction. However, if P for reaction on clean niobium corresponds to probabilities of NO chemisorption,J. A . MORGAN AND A .J. B . ROBERTSON 219 there would not be an increase in P with excess H, present. Owing to inaccuracies in the measurement of rates of product formation, because of their ready adsorption on the reaction vessel walls, it was not possible to determine whether the increase in P on contaminated niobium filaments in the presence of H2 was due, in part, to a reaction of H2 with NO. If the enhancement of activity of contaminated niobium is entirely due to the removal of surface contaminants, P should increase with time of reaction as the surface becomes less and less contaminated. In fact P was relatively constant with time, which suggests that a nearly steady state occurs and the observed P is determined by the relative rates of removal of surface contaminants by reaction with M2, and their replenishment by deposition and diffusion from the interior to the surface.The same considerations apply in the discussion of the results obtained on nickel. P on clean nickel was increased in the presence of excess H2, indicating that an H,+NO reaction was occurring to a small extent. Because of these complicated considerations, kinetic parameters for the reaction of NO on niobium and nickel in the presence of excess H2 may have no ready interpretation and serve mainly to illustrate trends in P with changing filament temperatures. As platinum did not dissociate NO molecules in the absence of Hz, the observed NO reactions in the presence of H, are probably bimolecular surface reactions. At temperatures above = 1250 K the reaction rate seems to be determined by the rate of atomisation of H2 at the filament.For lower temperatures the reaction is probably between NO and H2 molecules and has an activation energy nearly equal to that observed for the reduction of O2 by H2 on platinum by ourselves (49-1- 1 kJ mo1-l)13 and by Gentry et al. (49+3 kJ mol-l).l4 The rate-determining step may be the same for both reactions, possibly the rupture of the H--H bond, with addition of pi atoms to NO or Q2, and the reduced species then decomposing rapidly to smaller fragments. Such a mechanism would explain the rate dependence on H2 atomisation at temperatures above 1250 K. An interesting aspect of the NO reduction by H, at platinum temperatures below 1250 K is that the term B has a value close to unity, which is in accordance with the transition state theory prediction for a mobile transition state.l This result contradicts the view of some workers, notably Laidler,16 who has stated that bimolecular surface reactions of all types may be satisfactorily explained in terms of an immobile layer. We thank the S.R.C. for their support. M. Shelef and J. T. Kummer, Amer. Znst. Chem. Eng., Chem. Eng. Progr. Symp. Series, 1971, 67, no. 115, 74. F. G. Dwyer, Catalyst Rev., 1972, 6, 261. E. R. S. Winter, J. Catalysis, 1971, 22, 158. ' J. T. Yates and T. E. Madey, J. Chern. Phys., 1966, 45, 1623. J. A. Morgan and A. J. B. Robertson, J.C.S. Furuday I, 1974, 70, 936. D. J. Fabian and A. J. B. Robertson, Proc. Roy. SOC. A, 1956, 237, 1. S. C. Jain and T. C. Goel, Brit. J. Appl. Phys., 1968, 1,573. S. C. Jab, T. C. Goel and V. Narayan, Brit. J. Appl. Phys., 1969,2,109. M. Onchi and H. E. Fansworth, Surface Sci., 1969,13,425. lo A. B. Callear and I. W. M. Smith, Disc. Faraday Soc., 1964,37,96. M. A. A. Clyne and B. A. Thrush, Disc. Faraduy Soc., 1962,33, 139. l2 D. Brennan, Adv. CataZysis, 1964, 15, 1. l 3 J. A. Morgan and A. J. B. Robertson, unpublished results. l4 S. J. Gentry, J. G. Firth and A. Jones, J.C.S. Faraday I, 1974, 70, 600. l 6 K. J. Laidler, Chemical Kinetics (McGraw-Hill, London, 1965), p. 292. A. J. B. Robertson, J. Colloid Sci., 1956, 11, 308. (PAPER 7/946)

ISSN:0300-9599    

DOI:10.1039/F19787400211

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Exchange Processes in Solutions of Nitroxide Surfact ants BY KATHARINE K. Fox Unilever Research, Port Sunlight Laboratory, Port Sunlight, Wirral, Merseyside L62 4XN Received 5th December, 1975 The temperature-dependent broadening of the electron paramagnetic resonance spectra observed in aqueous solutions of the paramagnetic surfactants 2,2,6,6-tetramethylpiperidine-oxidedecyldi- methylammonium bromide, 2,2,6,6-tetramethyIpiperidine-oxidedodecyldimethyla~onium bromide and 2,2,6,6-tetramethylpiperidine-oxidetetradecyldimethylammo~~ bromide is not caused by monomer-micelle exchange in those solutions. It is probably due to the different effectiveness of monomers and micelles in causing Heisenberg spin exchange, complicated by the temperature variation of the c.m.c. and by possible dimerization of the nitroxide head group.In a previous paper the electron paramagnetic resonance (e.p.r.) spectrum of the paramagnetic surfactant I (n = 11) was shown to have CH3 CH3 I / + I , X = O Br- 11, X = H n = 9, 11, 13 n NX / L/ I\ ' I '\ CH3 I '\ CH, CH3 temperature-dependent line broadening above the critical micelle concentration (c.m.c.) which could not be attributed to Heisenberg spin exchange. This increased line broadening was attributed to surfactant exchange between monomer and micellar environments. Further work described here 2 * indicates that this initial attribution is incorrect, and that other processes are responsible for the observed line width changes. The cationic nitroxides I (n = 9, 11 and 13) were prepared using slight modifica- tions of the procedure described previously,l* followed by recrystallisation from water.The c.m.c. values of the surfactants were determined by conductivity. The number of unpaired electrons per molecule was determined by comparing the e.p.r. spectra of the solid state nitroxide surfactants with the spectra of freshly grown CuS04. 5H20 crystal^,^ with spectral intensities determined by a nomogram method.5 The results are shown in table 1. The e.p.r. spectra were recorded with a Varian E-4 spectrometer fitted with a variable temperature control unit. Temperatures were measured to 0.1 K relative, kO.1 I< absolute, using a copper-constantan thermopile placed in the centre of the e.p.r. cavity before and after each run. Microwave power was kept at 5 mW to 220K . K .FOX 22 1 avoid saturation effects, and themodulation amplitude was kept at < 1/10 of the linewidth to avoid modulation broadening. The solutions were deoxygenated before the spectra were recorded. In the experiments described here, micellization was induced in the paramagnetic surfactant systems I (n = 9, 11, 13) not by increasing the concentration of I, but by adding enough of the appropriate diamagnetic surfactant I1 (n = 9, 11 , 13) to take the resultant system above its c.m.c. This method was adopted in order that the non-micellar system containing only I could serve as an upper limit for the Heisenberg spin exchange contribution in the I +I1 system, instead of as a lower limit as when micellisation was induced by increasing the concentration of I.The frequency with which monomeric I enters a micelle should be similar in both cases. TABLE 1 .-PROPERTIES OF NITROXIDE SURFACTANTS spins per c.m.c. X lO3lmol dm-3 compound molecule 298 K 313 K 333 K I ( n = 9) 1.00+0.02 41 51 42 f 1 43 f 1 I (n = 11) 0.9.5_+0.02 8.7f0.2 9.3f0.2 9.1f0.2 I(IZ = 13) l.OOf0.02 2.120.1 2.1f0.1 2.6k0.1 For the n = 9 compounds, two solutions each 1.8 x mol dm-3 in I(n = 9) were prepared. One solution was also 5.8 x mol dm-3 in II(n = 9), which increased the total surfactant concentration to above the combined c.m.c. for the surfactant mixture. Spectra of both solutions obtained at 349.8 K consisted of three broad lines. In the sample containing II(n = 9) the m, = 0 line was more intense relative to the mi = + 1 lines than the mi = 0 line in the other sample, a result which is consistent with the presence of an underlying micellar peak in the first sample.This effect, although diminished in magnitude, was still present at 327.9 K. Thus the sample containing II(n = 9) had micelles at both temperatures studied, although the micellar signal was narrower at the higher temperature, due to increased Heisenberg spin exchange between the surfactant molecules in the micelle. TABLE 2.-LINEWIDTH OF n = 11 SAMPLES linewidth/mT sample 318K 328K 337K 346K 357 K 366 K 7x mol dm-3 I (n = 11) 0.170 0.176 0.188 0.197 0.211 0.223 7x mol dm-3 I (n = 11)+ 8 x mol dm-3 0.168 0.174 0.182 0.193 0.210 0.226 11 (n = 11) At 327.9 K the MI = 0 line of the system which contained micelles was narrower (0.298 0.005 mT) than that of the non-micellar system (0.365 f 0.005 mT).The same pattern held at 349.8 K (0.398k0.005 mT against 0.465f0.005 mT). Thus the major effect of creating micelles is not the introduction of monomer-micelle exchange, but the removal of paramagnetic monomers from the solution with consequent reduction in Heisenberg spin exchange. It is possible to study I(n = 11) at lower nitroxide concentrations, where Heisenberg spin exchange between monomers should have a smaller effect on the observed linewidths. A solution containing 7 x mol d ~ n - ~ I(n = 11) has been compared with a solution containing 7 x mol mol dm-3 I(n = 11) and 8 x222 EXCHANGE I N NITROXIDE SURFACTANTS dm-3 II(n = 11). The e.p.r. spectra of the latter sample show one broad micellar peak at temperatures ranging between 317 and 366 K, as well as the usual three-lined monomer spectrum.The width of the MI = 0 line in each system is shown in table 2. The linewidths in the two systems are similar, with the width in the non-micellar system exceeding the width in the micellar system by - 3 times the experimental error in the region around 337 K. The major effect of the addition of micelles to this system is also the removal of paramagnetic monomers from solution, and the concomitant reduction in Heisenberg spin exchange. A similar experiment was performed for the n = 13 compounds. A solution containing 2 x mol dm-3 II(n = 13) at 343 K contained micelles whose e.p.r. spectrum consisted of one broad, exchange-narrowing line. Superimposed upon this were three monomer hyperfine structure lines, each of which exhibited proton hyperfine structure.At 298 K the micellar spectrum of this solution consisted of three broadened lines, since intra-micellar exchange was no longer great enough to cause exchange-narrowing. The three monomer lines were also present, with decreased proton hyperfine structure resolution. Monomer- micelle exchange in the solution at 298 K would cause less broadening of the monomer lines than exchange of the same magnitude in the 343 K solution, due to the different micellar signals.6 The observed effect of increased proton hyperfine structure resolution is due to the increased tumbling frequency of the monomers at the higher temperature, and the consequent increased averaging of the g- and a-tensor aniso- tropies.Monomer-micelle exchange is not fast enough to overcome this effect, and thus is <5 x lo4 s-I in this particular system. The examples given above show that this e.p.r. method cannot detect monomer- micelle exchange in the I(n = 9, 11, 13) systems when micellisation is induced by adding the diamagnetic surfactant II(n = 9, 11, 13). The effect on the linewidths previously attributed to monomer-micelle exchange is probably due to the different effectiveness of monomers and micelles in causing Heisenberg spin exchange, compli- cated by the temperature variation of the c.m.c. Dimerization of the nitroxide headgroup 2* mol dm-3 I(n = 13) and 8 x may also contribute to the observed linewidths. I would like to thank Prof. M. C . R. Symons and Mr. J. Clifford for helpful discussions during the course of this work. K. K. Fox, Trans. Faraday SOC., 1971, 67, 2802. A preliminary account of this work is published in, Chemical and Biological Applications of Relaxation Spectrometry, ed. E. Wyn-Jones (Proc. NATO Advanced Study Institute, University of Salford, 29 Aug.-12 Sept. 1974), p. 215. K. K. Fox, Ph.D. Thesis (University of Leicester, 1974). P. B. Ayscough, Electron Spin Resonance in Chemistry (Methuen, London, 1967), chap. 1. V. A. Tolkacher and A. I. Mikhailov, Probor. Tekhn. Eksper, 1963, 9, 95. J. R. Zimmerman and W. E. Brittin, J. Phys. Chem., 1957, 61, 1328. (PAPER 5/2367)

ISSN:0300-9599    

DOI:10.1039/F19787400220

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Hydrogen Sorption by Palladium-Gold Wires B Y DANIEL D. ELEY * AND EDWARD J. PEARSON Chemistry Department, University Park, Nottingham, NG7 2RD Received 30th September, 1974 Hydrogen sorption was studied on PdAu wires carefully cleaned in ultra high vacuum, at a hydrogen pressure of 6.25 pPa and wire temperatures of 100, 150,200 and 304 K. For Pd at 100 K the initial sticking coefficient was 0.16, falling to 0.006 at a fractional monolayer coverage 0 = 0.8. Sorption against time curves at 100 K for hydrogen continued smoothly to coverages 8 > 3.0. Deuterium for the same collision number ( P D ~ = 2/!&) gave a similar rate up to fl - 1.0 when the uptake became slower. Temperature programmed desorption gave one peak and desorption activation energies, which extrapolated to a 0 = 0 value of 100 kJ mol-1 for Pd, 90PdlOAu and 70Pd30Au and 45 kJ mol-1 for 55.2Pd44.8Au.These activation energies decreased sharply with coverage, to a value for the three Pd rich alloys of -20 kJ mol-I, comparable with the heat of solution. Hydrogen sorptions for a given exposure increased slightly from Pd to 90PdlOAu, decreasing sharply towards zero at 40Pd60Au, where the holes in the d band have disappeared. The results are discussed critically in relation to earlier data, and a tentative model advanced involving (mainly) /3 chemisorbed H (on Pd atoms, immobile at loOK), c( chemisorbed H2 (weakly held on octahedral holes in 100 planes) and dissolved H atoms under the surface. Although it is concluded that holes in the d band are necessary for strongly chemisorbed /3-H, a role for PdnAu, surface ensembles cannot be rulzd out.Starting with the ortho-para hydrogen conversion, we have now published work on eight reactions on our original set of 42 SWG polycrystalline PdAu wires,2 with three papers in preparation on further reactions, now carried out under ultrahigh vacuum condition^.^ In this paper we study the temperature programmed desorption of hydrogen from these wires, to obtain information on the bond strength and general behaviour relevant to our work on catalytic mechanisms. Since Pd and Au form a homogeneous set of solid solutions for all compositions, they are currently of great interest in academic studies, as witness recent reviews.4* In addition, Auger analysis has established that PdAu alloys have identical bulk and surface compositions' in vacuo,? whereas PdAg shows Ag enrichment in the surface with respect to the bulk.6 At the outset it should be realized that the main problem with Pd alloys is to distinguish adsorbed hydrogen on the metal surface from absorbed hydrogen in the bulk.In ordinary high vacuum conditions it was found necessary to add finely divided transfer catalysts, such as Pd black or UH3 7* * to secure a rapid surface reaction leading to equilibrium concentrations of dissolved H in reasonably short times, in Pd wires or foils at temperatures <120°C. More recently it has been found that equilibrium dissolved H in massive Pd may be rapidly achieved at temperatures as low as 177 K, by using ultrahigh vacuum equipment which excludes mercury vapour and tap grea~e.~*l These are the conditions adopted in the present research.Wagner 12- l3 discussed the solution of hydrogen in Pd terms of one or other of two successive steps. being rate determining. However, the surface kinetics involved still merit investi- gation. It appears that it is not the first strongly chemisorbed hydrogen but a H2(g) + 2H(ads), H(ads) $ H(disso1ved) 'r Preliminary work by Dr. B. Moore has confirmed this result on the present wires. 1-8 223224 H ADSORPTION BY Pd-Au WIRES successively weakly cheinisorbed hydrogen lo (either atoms lo or molecule^)^^ which is the precursor to dissolved M atoms. This is the background to the present investigation. A previous flash-desorption study of hydrogen from PdAu wires was published by Tardy and Teichner.'' EXPERIMENTAL A conventional ultrahigh vacuum equipment l6 was constructed from 3 cm-diameter glass tubing.This comprised a cylindrical vessel, 3 cm-diameter, holding axially a 15 cm length of 42 SWG PdAu wire, with an Alpert ion gauge and omegatron mass spectrometer closely adjacent. Hydrogen gas could enter this system at one end via a V.G. variable leak metal valve, leaving at the other end via a pinhole leak connected to the pumping system. The equilibrium or steady-state pressure pes = 6.25 pPa for hydrogen was fixed for each experiment by the setting of this metal valve. The Alpert gauge, a Mullard 106-17, containing a lanthanum boride coated filament, was run at a low enough temperature to avoid dissociation of the hydrogen into atoms.It was used to measure the desorption spectra. During bakeout the omegatron was used to monitor the background impurities, which on cooling usually yielded a vacuum below 53 nPa (4x 10-lo Torr). The 42 SWG alloy wires were from the original spools provided for Couper's work by Johnson Matthey. British Oxygen Grade X hydrogen, stated purity 99.999 %, was used without further purification. The filaments were heated electrically with an electronic control circuit that kept the average filament temperature to _+ 1°C in the steady state, and also allowed us to impose a linear heating regime In our experiments we used p = 2 K s-l, a relatively slow rate compared with N 100 K s-l used in much flash desorption work. Resistance against temperature calibration plots for the filaments were determined using a series of thermostat baths and furnaces, and agreed well with published data.15 The PdAu filament was initially cleaned by the following treatment: (1) outgassed (50 nPa) at 1000 K for 6 h, (2) heated in 6 mPa oxygen at 1000 K for 20 h, (3) pump, then 6 mPa hydrogen at 1000 K for 6 h, (4) pump, then 6 pPa oxygen for 15 min, (5) pump, then 6 pPa hydrogen for 15 niin, (6) pump to 50 nPa at 1000 K for 10 min.Between runs the wire was given 5 min reduction as for (5), followed by outgassing for 2 min as for (6), and every 2-3 days the wire was subjected to (4) followed by (9, repetition of this treatment always yielding the same final state. H atom bombardment followed by outgassing of the wire between runs gave virtually the same result as the oxidation-reduction technique. Sticking probability determinations by Ehrlich's method l6 involved equilibrating the wire at 1050 K and hydrogen pressure pes, then suddenly cooling the wire to the desired temperature and recording the p against t curve, adsorption coverages being calculated from the integrated p against t curve.Temperature-programmed desorption (TPD) data were analysed by one or other of three methods in terms of order of reaction a, activation energy E, and frequency factor v($, in the equation for the isothermal rate of desorption dnldt, where there are n adsorbed molecules per m2 dnldt = n'vg) exp (- E/RT). The methods were : (a), from Tp, the temperature at which the desorption rate is a maximum, e.g. for a = 2 from a plot of log,, noT; against l/Tp, no being the initial adsorbed concentra- tion.and (from Redhead)l7 (b), from a plot of log,, (radn/dt) against 1/T, which gives a line of slope E/2.303R, if E is independent of coverage and the correct value of a is used (see Degras) ;l (c), methods (a) and (b) fail to yield straight lines if E depends on coverage n, but the E against n relation may be derived from plots of log (- dn/dt)given against 1 /T.D . D . ELEY AND E. J . PEARSON 225 RESULTS SITE DENSITY To predict the uptake of hydrogen molecules in a chemisorbed monolayer, it is necessary to assume a density of adsorption sites on Pd and the PdAu alloys. We shall assume as in earlier work l * that the Pd wire has a roughness factor of 1.0 and exposes equal area of (1 10) and (100) places, giving 1.16 x lo1 Pd atoms or single sites per m2.Therefore assuming dissociative atomic adsorption of the hydrogen we predict a monolayer uptake of 5.8 x 10l8 molecules m-2, which we also apply to the PdAu alloys and denote by M where appropriate in the figures. There are clearly several uncertainties in this assumption. For example, Pd being f.c.c. it might have been more appropriate to assume equal areas of the lowest energy planes (1 11) and (loo), which would give a monolayer uptake of 7.2 x lo1 molecules m-'. The first assumption, however, appears to fit in better with the present results, besides being consistent with our earlier work. STICKING COEFFICIENT The cleaned Pd wire at 100 K was exposed to a pressure of 6.25 pPa hydrogen, and the result is shown in fig.1. 10 2 0 30 40 5 0 M 6 0 coverage/lOl' molecule m-2 monolayer coverage (see text). FIG. 1.-The sticking probability of hydrogen at 6.25 pPa on the Pd wire at 100 K. M denotes TEMPERATURE-PROGRAMMED DESORPTION The cleaned and outgassed wire was heated to 1000 K and exposed to 6.25 pPa hydrogen, conditions resulting in negligible sorption. It was then cooled to the required initial temperature, namely 100, 150,200 or 304 K, a suitable exposure time allowed for adsorption and absorption to occur, after which the linear temperature programmer was switched on and the desorption " spectrum " recorded. Exposures are recorded in Langmuir units, 1L = Fig. 2 then shows three TPD curves from pure Pd following exposures of 1.41,4.23 and 9.87 L at 100 K and 6.25 pPa hydrogen.By integration surface coverages were calculated for the given exposures, and coverage against time plots drawn. An estimated error for Torr s = 133 pPa s.226 22 0 20c 18C & 160 ‘3 140 8 ei a .-( E 120 \ E! 100 8 J E 8 0 8 8 r- r( H ADSORPTION BY Pd’-Au WIRES #A I00 2 0 0 3 0 0 t emperature/K FIG. 2.-Desorption spectra of hydrogen from a Pd wire initially exposed at 100 K and 6.25 pPa, for the number of Langmuirs (L) indicated. Each experiment starts at pes = 6.25 pPa.D. D. ELEY A N D E . J . PEARSON 227 these coverages is 5-7 %. Three such curves are shown for Pd, 90PdlOAu and 70Pd30Au at 100 K and 6.25 pPa in fig. 3. Fig. 4(a) shows the effect of exposure and alloy composition on surface coverages of the wire held at 100 K, and fig.4(6) N I E 100 9 0 8 0 70 6 0 5 0 40 100 9 0 8 0 7 0 6 0 5 0 4 0 (4 at %Pd (b) at % Pd FIG. 4.-(u) Effect of alloy composition on hydrogen uptake for various exposures (L) at 100K and 6.25 pPa. U,56 L ; V, 28 L ; 0,14 L ; x ,5.6 L. (b) Effect of alloy composition on hydrogen uptakes for 28 L exposure, at four temperatures and 6.25 pPa. The dotted line is the result due to Tardy and Teichner (see text). 0, 150; 0, 100; V, 200; x , 304 K. *0° t t F1 I E 0 5 10 15 2 0 2 5 3 0 35 4 0 adsorption timelmin FIG. 5.-Comparison of H2 uptake (0) at pHz = 6.25 pPa with Dz uptake ( A ) at P D ~ = 2/2, = 8.75 pPa, by a 70Pd30Au wire at 150 K.228 H ADSORPTION BY Pd-Au WIRES shows the effect of the temperature and composition of the wires on coverages follow- ing an exposure of 28 L (the apparently anomalous position of the 150 K curve being noted).In fig. 5 the coverage against time relationship for deuterium is compared with that for hydrogen on 70Pd30Au, the deuterium pressure pD2 being adjusted to 8.75 pPa, i.e. $pH2, so as to give the same impingement rate on the wire. It is apparent that the rate of sorption of deuterium is identical to that of hydrogen up to - 1 monolayer coverage, when it becomes noticeably slower. 15.5 16.0 16.5 1 0 5 1 ~ FIG. 6.-First order (A) and second order (0) Degras plots for desorption of hydrogen from a Pd wire. The initial sorption temperature was 200 K. 3 0 0 3 5 0 4 00 450 5 0 0 1 0 5 1 ~ FIG. 7.-Desorption activation energy plots for a Pd wire corresponding to the following coverages (left to right) x ,4.8 x lo1' ; 0, 15.0 x 1017 ; U, 24.5 x 10'' ; 0,45.0 x lox7 ; all in molecule m-2.D .D. ELEY AND E . J . PEARSON 229 Activation energy for desorption plots were made according to all 3 methods, methods (a) and (b) (the latter for both first and second order reactions) giving curved plots indicative of a dependence of activation energy on coverage, and fig. 6 shows an example of method (6) applied to pure Pd. It was therefore decided to analyse our data by method (c), which involves deriving (dnldt) for the same coverage from a number of runs at different temperatures. Here we had to be satisfied with straight lines based on 3 points, as in the examples in fig. 7, the slopes of which gave E12.303 R. Fig. 8 shows that the E values so derived for Pd, 90PdlOAu, and 70Pd30Au lie virtually on the same E against coverage curve, which extrapolates back to zero coverage to give an initial activation energy (" initial heat of adsorption ") of 100 kJ mol-1 with a visually estimated error of 10 kJ mol-'.In the fig. 8 inset we have plotted the initial (zero coverage) E values as a function of alloy composition. 2 0 40 8 0 100 120 140 ~overage/lO'~ molecule m-2 FIG. 8.-Desorption activation energy as a function of coverage for: CI, Pd ; 0, 90PdlOAu ; A, 70Pd30Au and x , 55.2Pd44.8Au wires. The inset plot gives the initial activation energies at zero coverage as a function of composition (estimated error bars have been inserted). DISCUSSION SOLUTION OF HYDROGEN We first enquire whether our conditions of pH2 = 6.25 pPa and TK are likely to give rise to the ap phase transition? in the palladium hydride, assuming an equilibrium concentration of dissolved hydrogen is reached.gives a linear plot of loglops, against 1/T(K) which we assume may be extrapolated below 194.5 K to our lowest temperature of 100 K. As a result we derive values of pslx as 300 K, 1333 Pa; 200 K, 0.43 Pa; 150 K, 133 pPa and 100 K, 0.013 nPa. From this it may be seen that at equilibrium only the a phase may be formed in our experiment at 300, 200 and 150 K although if a steady pressure of 6.25 pPa were maintained it should give the P-hydride phase at equilibrium at 100 K. However, the time duration of our experiments is such that only the a phase can be formed at 100 K. are used both to label the two palladium hydride phases, and to distinguish the weakly and strongly adsorbed hydrogen states in desorption spectra.Lewis t cc and230 H ADSORPTION BY Pd-Au WIRES give a formula relating n = [H]/[Pd] for solution equilibrium in the a hydrogen phase, as a function of T and p over the range + 75 to - 78°C. Assuming this formula is applicable to our temperatures, we calculated for the working pressure p = 6.25 pPa at 100 K, the value rz = 2 x and at 200 K, n = 3.7 x If the wire for simplicity exposed only the (100) face, and if r, wire radius = 0.1016 mm and Ypd the radius of the Pd atom = 0.137 nm then Wicke and co-workers7* rw 2 . 8 3 ~ ~ ~ = 2.6 x 105. N- total number of Pd atoms in wire number of Pd atom in wire surface - So assuming H = Pd for the surface, then at 100 K total dissolved H/surface H - 520 and at 200K the ratio is 0.96.Given therefore a sufficient initial exposure of the wire to hydrogen for equilibrium hydride formation, and sufficiently rapid de-solution kinetics, we should expect the flash desorption peak due to dissolved H to be dominant over that for adsorbed H following an initial exposure of the wire at 100 K, and the two peaks to be comparable following initial exposure at 200 K. There may, of course, be several adsorption peaks if there are several adsorbed species with different binding strengths. When only one peak is observed overall this may be due to overlap of the solution and adsorption peaks, or perhaps because diffusion is too slow to yield a solution peak within the time of the temperature scan.STICKING COEFFICIENT ON Pd Aldag and Schmidt 2 o found for hydrogen on a polycrystalline Pd wire an initial sticking coefficient of S = 0.1 3 falling to 0.001 at higher coverages, which later they associated with the weakly bound a state detected in TPD, which they regarded as a precursor to solution. Fig. 1 shows our very similar results, S = 0.16 falling to S = 0.006 at a coverage of 4.5 x 10'' molecules m-2, i.e. a fractional coverage 0 of 0.78. These values support the view that our initial Pd surface is clean. Tardy and Teichner on the other hand reported much lower initial sticking coefficients for Pd, -10-4.15 COVERAGE AGAINST TIME RELATION It may be seen from fig. 3 that the rate of uptake at 100 K and 6.25 pPa is closely similar for Pd, 90PdIOAu and 70Pd30Au wires, and that a monolayer coverage is reached at -8 min or 22.5 L exposure, with no indication of any knee at this point.The rate after the monolayer shows some scatter but it appears highest on 90PdlOAu, a result brought out in several experiments in fig. 4. A similar run on a 70Pd30Au wire at 150 M in fig. 5 compares deuterium (8.75 pPa) and hydrogen (6.25 pPa) rates of uptake, showing that the deuterium rate becomes the slower after -1 monolayer uptake. This suggests that the initial process is formation of a strongly bound chemisorbed monolayer, the p state of Aldag and Schmidt,20 followed by onset of the solution process, where we expect the rate of solution of deuterium to be determined by its permeability constant, which is less than that for The permeability constant itself is a product of solubility (where deuterium < hydrogen) 8 * 9* 23 and diffusion constant (where apparently conflicting ratios have been reported, as &ID, = 1.31 at 302.5"C 24 and &ID,, = 0.62 at 25"C, extrapolating to 1.0 at 300°C).The concept of an initial strongly chemisorbed p state, a precursor to dissolved hydrogen was suggested by Lynch and Flanagan lo* l1 and fits in with the deuterium effect as noted above. However, the continuity of the hydrogen sorption curve, and the fact that already at a coverage of 0.75 monolayer the activation energy for desorption has decreased to a value approximating to the 22D. D. ELEY A N D E . J . PEARSON 23 1 heat of solution, suggests that adsorption and solution cannot be separated, even during uptake of the first monolayer.Auer and Grabke I4 observed that the solution rate is first order in hydrogen and suggested that the solution precursor was a hydrogen molecule, adsorbed with its axis perpendicular to the surface with one atom embedded in an octahedral interstitial site in the surface, following a process H,(ads) + [H(ads) . . . H(sub-surface)] + 2H(dissolved). They found kinetic evidence for a blocking action of H atoms (or protons) adsorbed in these sub-surface interstices. On this view their H,(ads) may be equated with the a-state of Aldag and Schmidt,20 although these authors report that at 100 K the a-state was not saturated even at 10-4Torr, containing much more hydrogen than the P-state, leading them to identify it with dissolved H atoms remaining near the surface. Our fig.3 for 100 K and fig. 5 for 150 K show that uptake is still continuing after 3 monolayer equivalents of H, e.g. 1 monolayer H and 2 ‘‘ mono- layers ” of dissolved H, and one is led to wonder whether at such surface concentra- tions something analogous to a surface a -+ p phase transition needs to be considered in the top 2 or 3 layers of metal atoms. TPD PEAKS Most of our TPD peaks, e.g. those in fig. 2, refer to an initial coverage of <1 monolayer of hydrogen, so we may assume that the single maximum observed in the range 220-3OOK refers to adsorbed hydrogen. Conrad et aZ.26 working with H,-Pd (110) also found single desorption peaks but understandably at the higher temperature of 360 K, because of a higher heating rate of 20 IS s-l.Aldag and Schmidt 2o who adsorbed hydrogen on polycrystalline Pd at 200 K and applied an even greater heating rate - 100-500 K s-I (L. Schmidt, personal communication) found their main /3 peak (including 3 sub-peaks PI, p2, p3) centred around 450 K. The peaks observed by all three groups clearly refer to the same adsorbed species, which we all attribute to atomic hydrogen, Pd-H. Aldag and Schmidt,20 by adsorbing hydrogen at 100 K, were able to detect a weakly bound a state, regarded as a precursor to solution, around 250 K. This peak did not appear in Conrad’s or our desorption spectra, presumably because of our much lower heating rates. On the other hand, by using very long exposures of 600L at room temperature Conrad et al.found a second desorption peak setting in at 500-700 K which they associated with dissolved hydrogen. Presumably the high activation energy associated with a peak at this temperature, means the dissolved H is being “ desorbed ” as atoms. ACTIVATION ENERGIES OF DESORPTION Since chemisorption on transition metals usually has a negligible activation energy, it is usual to equate activation energies of single-step desorptions with the heats (negative enthalpies) of adsorption of the species concerned, here labelled P. On the other hand, where bulk diffusion is the rate limiting reaction, the temperature maximum may be related to the activation energy for diffusion by an equation analogous to Redhead’s equation.27 The activation energy against coverage graphs for Pd, 90PdlOAu and 70Pd30Au all extrapolate to an initial heat of 100 kJ mol-1 at 8 = 0, which may be compared with Aldag and Schmidt’s 2o P peak value for 202 kJ mol-l, Conrad’s 26 87 kJ mol-I for Pd (1 11) and 102 kJ mol-I for Pd (1 lo), Beeck’s 2 8 105 kJ mol-l for Pd films, and Vert’s 29 electrochemical value of 1 1 3 kJ mol-I.Our values fall sharply with increasing coverage, to values of 15 kJ mol-I232 H ADSORPTION BY Pd-Au WIRES for Pd, 17.5 kJ mol-1 for 90PdlOAu and 19 kJ mol-l for 70Pd30Au. These values are lower than the a state desorption activation energy of 54-58 kJ mol-l of Aldag and Schmidt,20 and the isosteric heat of Lynch and Flanagan’s lo precursor state to solution which was 36.4-44.8 kJ mol-I. They are in fact much closer to values of the isosteric heats of evaporation of dissolved hydrogen, given as 19.3 kJ mol-l for Pd, 29.0 kJ mol-1 for 84.7Pdl5.3Au, and 37.8 kJ mol-l for 73.5Pd26.5A~.~’ Other values for this latter quantity for pure Pd are 23.3 and 27.4 kJ m ~ l - l , ~ ~ all “ mols ” referring to HZ.The fall of adsorption energy with coverage observed here may denote the effect of a repulsive interaction between neighbouring H atoms in an immobile film, i.e. immobile at the adsorption temperature of 100 K.31 However, it may also arise from (a) chemisorption on different planes, (b) increased adsorption into the a state, and (c) increased solution contributions at the higher coverages. In contrast, Conrad et a1.26 report for H2-Pd (1 10) that the value of 102 kJ mol-’ remains constant from 8 = 0 to -8 = 0.5, thereafter falling, and similar behaviour appears to have been found b ~ V e r t .~ ~ This latter behaviour was shown by Roberts 31 to be expected for dissociative chemisorption on a uniform set of sites with nearest neighbour repulsion energy, when the film of atoms is mobile, as seems likely when prepared at room temperature as in these two cases. In the inset of fig. 8 we see that the zero coverage E shows a strong decrease after 70Pd30Au to 45 kJ mol-1 at 55.2Pd44.8Au, which if continued would reach 0 kJ mol-l at 45Pd55Au, a composi- tion approaching that where the 0.6 holes in the 4d band of Pd are filled by 5 s electrons from the Au atoms. We clearly cannot detect hydrogen uptake beyond this point (fig. 4) in agreement with Tardy and Teichner.15 CHEMISORPTION A N D ALLOY COMPOSITION In fig.4(a) and (b) we see that the main result is a profound decrease in the coverage of the Pd rich alloys which occurs around 55Pd45Au’ to a virtually zero value at 40Pd60Au, where the holes in the Pd d-band are completely filled by s-electrons from the Au. The dotted curve in fig. 4(b) is Teichner and Tardy’s result l5 for very different conditions, namely 6900 L exposure to hydrogen at an unspecified tempera- ture, which appears to have been 278 K. Their results show a similar absorption for Pd but a much more marked maximum at intermediate Pd-rich compositions, and also the same profound decrease in adsorption of 55Pd45Au. These results, together with the activation energies of desorption, support the view that holes in the d band play a dominant role in the bonding energy of the strongly chemisorbed ( p ) hydrogen, although the relation is clearly not a simple linear one.A flat peak over say Pd to 70Pd30Au has been shown by Dowden to result from adsorption on Pd,Au, Kobosev ensembles.32 On the other hand the maximum adsorption found in this region compels us to draw an analogy with the parallel finding for the a-phase in dissolved hydrogen.33 Although under some conditions this maximum is not found for Pd-rich PdAu alloys,34 nevertheless, it seems always present so long as p < pas, the critical pressure for the ccp hydride phase transition, which occurs in both Pd and PdAg and presumably PdAu, alloys.35 It appears to be due to the effect of added Ag(Au) in lowering the energy required to expand the octahedral interstitial sites which hold the dissolved protons in bulk Pd.35 It also seems possible that the a- adsorption of H2 molecules in surface octahedral interstitial sites will be similarly favoured.Clearly, the maximum uptake at 90PdlOAu is to be associated with subsurface dissolved protons, with a possible contribution at low exposure times from aH,, the precursor to dissolved hydrogen. This explanation seems more likely than that of a ‘‘ second-order geometric effect ” on the adsorption of the stronglyD. D. ELEY AND E . J . PEARSON 233 chemisorbed PH offered ear lie^,^ both for the adsorption of hydrogen, and the catalysis of the poH2 and H2 + D2 reactions. We shall explore the relevance of D2 weakly adsorbed on octahedral sites in a later paper ; it might be the species respon- sible for the D2 exchange reaction observed with strongly chemisorbed PH by A.and L. F a r k a ~ , ~ ~ in which case the (111) Pd surface, which has no octahedra1 interstices, should be relatively less active in this reaction. Referring to fig. 4(b) the decrease in coverages over the series 100, 200, 300 K (150 K being anomalous), is in line with expectation for " pseudo-equilibrium " values of QH at the very low ambient pressure used. Our coverages for pure Pd may be compared with the saturation coverages of Aldag and Schmidt,20 the latter at 0.66 pPa and 10 L exposure, in brackets, uiz. 100 K, 70 x 1017 molecules m-2 ; 200 K, 34 x 1017 (60 x 1017) ; and 304K, l o x 1017 (300K, 25x 1017).In view of the differences in assumptions, approximations and experimental conditions between the two laboratories, we may regard the agreement as satisfactory. The anomalously high coverage in our experi- ments at 150 K if confirmed by further work would suggest the operation of a small activation energy for adsorption of PH in reducing the coverage at 100 K below that for 150 K. A SURFACE MODEL In fig.9 we show a picture of the(100)plane of f.c.c. Pd. The strongly chemisorbed H atom is regarded as forming an electron pair bond to a single Pd atom, which is consistent with the agreement between the observed heat value for QH and Stevenson's calculated value 37 for the covalent bond H2 +2Pd + 2PdH AHads = -94.1 kJ mol-1 (calc) - 100 kJ mol-1 (expt). H,vW I FIG.9.-The (100) plane of Pd, with the various postulated chemisorbed PI, P2 and oc hydrogen species indicated by white spheres. The partial electronic charges are noted, H2 VW is van der Waals adsorbed hydrogen held over 2Pd sites. This is the hydrogen with a negative surface potential (Q;) according to Dus.~* When this film is virtually complete, we can have weak chemisorption of atoms and molecules, pz and a, both with positive surface potential^.^^ The existence of a234 H ADSORPTION BY Pd-Au WIRES strongly bound negative species followed by a weakly bound positive species was originally inferred from electrical conduction and thermoelectric 40 studies of hydrogen on palladium films. Whether it is sensible to postulate that the fli and a forms can also be chemisorbed on the interstices of the (1 11) plane is a matter for further study.So far as the bulk is concerned, the (1 11) interstices form the transition state for a dissolved proton diffusing from one octahedral hole to its neighbour in the lattice.41 Therefore it is likely that fl; and a hydrogen will be less stable on this (1 11) plane. Denoting 0, as a surface octahedral interstitial site, and nb by a similar site immediately under the surface, we summarise the available data for the processes in this discussion below. H2 OsOb = H * C]sH nb H2+Os + H 2 . 0 , AH = -44.8 kJ mo1-1 lo AH = -33.9 kJ mol-1 Eact = 28.4 kJ mol-1 l4 AGO - -21.3 kJmol-l 42 AHo - -23 kJmol-l l1 Eac,(diffusion) = 24.0 kJ m ~ l - ~ . ~ ~ H nsob * UsH ob H2$2C3b + 2H. Ob H a b a b +- nbH ob In fig.10 we have roughly sketched some potential-energy curves for these various species on the surface and sub-surface of a (100) Pd lattice plane, in the absence of lateral surface interactions. It seems clear from these curves that if HB is to be in equilibrium with dissolved H, it must be thermally activated into a surface mobility (perpendicular to the plane of the paper) so that it may freely diffuse into the metal surface FIG. 10.-A rough potential energy scheme for surface and sub-surface hydrogen on Pd (100). The section is taken through the surface octahedral holes (full lines), and through the neighbouring Pd atom (dotted lines). octahedral sites, and hence into the bulk metal. This mobility may just be setting in around 25"C, but the process is unlikely to be competitive with the much lower activation energy solution process involving a adsorbed H2 until much higher temperatures, when a p* rate of solution should set in.The effect of alloying Pd with Au is to lower slightly the bulk H potential energy, and after -70Pd30Au to raise that for chemisorbed HB. The important matter of surface mobility betweenD. D . ELEY AND E . J . PEARSON 235 weakly and strongly chemisorbed H atoms on adsorption and solution of hydrogen in Pd has been discussed recently by Bucur, Mecea and 111drea.~~ It seems quite likely that for Au rich alloys (beyond 40Pd60Au) that HB becomes more weakly held than Ha, and all the chemisorbed hydrogen H2 or H is weakly held on the octahedral hole sites. Thus although Au will catalyse pH2 and H, + D2 reactions indicating the presence of chemisorbed hydrogen, there is no measurable H2 uptake by ordinary volumetric methods.45 This model is, of course, extremely tentative, and does not involve any steps or other defect states which may be needed in a more complete consideration of the problem.A recent review 46 has concluded that “ the dynamical conditions of formation and decomposition of hydride phases ” are important in catalysis and the present work gives strong support to this statement. The authors wish to thank the S.R.C. for a studentship awarded to E. J. P., and I.C.I. for a grant to purchase the omegatron mass spectrometer. ’ A. Couper and D. D. Eley, Disc. Faraduy SOC., 1950, 8, 172. D. D. Eley, J. Res. Inst. Catalysis, Hokkaido University, Sapporo, 1968, 16, 101.D. D. Eley, Chem. and Znd., 1976, 12. R. L. Moss and L. Whalley, Ado. Catalysis, 1972, 22, 115. E. 6. Allison and G. C. Bond, Catalysis Rev., 1972, 7, 233. B. J. Wood and H. Wise, Surface Sci., 1975, 52, 151. E. Wicke and G. Bohmholdt, Z. Phys. Chem. N.F., 1964,42,115. E. Wicke and G. H. Nernst, Ber. Bunsenges Phys. Chem., 1964,68,224. J. D. Clewley, T. Curran, T. B. Flanagan and W. A. Oates, J.C.S. Faraday Z, 1973, 69, 449. lo J. F. Lynch and T. B. Flanagan, J. Phys. Chem., 1973,77,2628. ’’ J. F. Lynch and T. B. Flanagan, J.C.S. Faraday Z, 1974, 70, 814. l2 C. Wagner, 2. Phys. Chem. A, 1932,159,459. l 3 C. Wagner, Z. Elektrochem., 1938,44, 507. l4 W. Auer and H. J. Grabke, Ber. Bunsengw Phys. Chem., 1974, 78, 58. l5 B.Tardy and S. J. Teichner, J. Chim. Phys., 1970,67, 1962,1968. l6 G. Ehrlich, Adv. Catalysis, 1963, 14, 265. l 7 P. A. Redhead, Vacuum, 1962,12,203. l 8 D. A. Degas, 27iesis (Faculty of Science, Paris, 1966). l9 F. A. Lewis, The Palladium Hydrogen System (Academic Press, New York, 1967), p. 25. 2o A. W. Aldag and L. D. Schmidt, J. Catalysis, 1971, 22,260. ’ R. M. Barer, Dzj5usion in and Through Solids (Cambridge University Press, Cambridge, 1941), p. 184. 22 L. A. Rubin, Engelhard Ind. Inc. Tech. Bull., 1961,7, No. 122, 55. 23 M. J. B. Evans and D. H. Everett, Hydrogen in Metals (School of Metallurgy, Birmingham 24 W. Jost and A. Widmann, 2. Phys. Chem. B, 1935,29, 17; 1940,45,285. 2 5 G. Bohmholdt and E. Wicke, Z. Phys. Chem. N.F., 1967,56,133. 26 H. Conrad, G. Ertl and E. F. Latta, Surface Sci., 1974,41,435. 27 G. Farrell and G. Carter, Vacuum, 1967, 17, 15. 28 0. Beeck, Disc. Faraday SOC., 1950, 8, 118. 29 Z. L. Vert, I. A. Mosevitch and I. P. Tverdovskii, Doklady Akad. Nauk, S.S.S.R., 1961, 30 K. Allard, A. Maeland, J. W. Simons and T. B. Flanagan, J. Phys. Chem., 1968,72,136. 31 J. K. Roberts, Some Problems in Adsorption (Cambridge University Press, 1939). 32 D. A. Dowden, Catalysis, Proc. Fifth Znt. Congress on Catalysis (Miami Beach), 1972, 1, 621 3 3 A. Sieverts, E. Jurisch and A. Metz, 2. anorg. Chem., 1915,92, 329. 34 A. Maeland and T. B. Flanagan, J. Phys. Chem., 1965, 69, 3575. 35 A. Brodowsky and E. Poeschl, Z. Phys. Chem. N.F., 1965,44,143. 36 A. Farkas and L. Farkas, J. Amer. Chern. SOC., 1942,64, 1594. 37 D. P. Stevenson, J. Chem. Phys., 1955, 23,203. 38 R. Dug, Surface Sci., 1973, 42, 324. ” R. Suhrmann, G. Wedler and G. Schumiki, in Structure and Properties of Thin Films, ed. C. A. Neugebauer, J. B. Newkirk and D. A. Vermilyea (Wiley, New York, 1959), p. 268. University, Jan. 5 and 6, 1976), preprints. 140, 149. (North Holland Press, Amsterdam, 1973).236 H ADSORPTION BY Pd-Au WIRES 40 D. D. Eley and J. Petro, Natuw, 1966, 209, 501. 41 H. Buchold, E. Wicke and G. Sicking, Hydrogen in Metals (School of Metallurgy, Birmingham University, Jan. 5 and 6, 1976), preprints. 42 V. Breger and E. Gileadi, Electrochim. Acta, 1971, 16, 177. 43 H. Zuchner, 2. Naturforsch, 1970, 25a, 1490. 44 R. V. BUCW, V. Mecea and E. Indrea, Hydrogen in Metals (School of Metallurgy, Birmingham University, Jan. 5 and 6, 1976), preprints. 45 M. J. Chappell and D. D. Eley, in preparation. 46 W. Palczewska, Adv. Catalysis, 1975, 24, 245. (PAPER 4/2011)

ISSN:0300-9599    

DOI:10.1039/F19787400223

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Thermal Behaviour of y-MnO, and Some Reduced Forms in Oxygen BY JOHN A. LEE, COLIN E. NEWNHAM" AND FRANK L. TYE Ever Ready Co. (Holdings) Ltd., Central Laboratories, St. Ann's Road, London N15 3TJ AND FRANK S. STONE School of Chemistry, University of Bath, Claverton Down, Bath BA2 7AY Received 4th November, 1976 The products obtained by partial chemical reduction of an electrodeposited y-manganese dioxide have been studied by thermal analysis in an oxygen environment between room temperature and 725 K. Weight loss in the temperature range 430-520 K was independent of sample composition when up to 35 % of MnIv was replaced by MnIrI. Substitution of >50 % MnIv led to a dependence of weight loss on degree of reduction. With the aid of X-ray diffractometry and magnetic suscepti- bility measurements, these results could be explained by the formation of a single phase solid solution below -50 % reduction and a two-phase system when greater amounts of MnIII were incorporated into the solid.Weight loss in the neighbourhood of 630K could be resolved into two entities. The low temperature component was associated with the phase transition from y-Mn02 to j3-MnOz accom- panied by some loss of oxygen. The high temperature component was attributed to the loss of water resulting from the decomposition of Mn(OH),+ Previous thermal analysis studies on electrodeposited y-MnO, have identified three evolved entities designated I, II and 1II.l. Type I appeared at 390 K and was identified as the removal of molecular water reversibly adsorbed on the oxide surface.Type I1 was associated with the irreversible loss of water between 430 and 520 K. Experiments in static water vapour environments partially resolved type I1 into two components. The first component, IIa, appeared to arise from water held in the bulk of the oxide, and amounted to -20 mg g-l. The second type, IIb, was directly related in amount to the overall oxidation state of the oxide and was attributed to the removal of water by condensation of hydroxyls. Type I11 appeared at 590- 630K, but in near-vacuum, nitrogen or water vapour environments it was only poorly resolved and was not chemically assigned. The present paper reports t.g./d.t.g. studies of y-MnO, in an environment of oxygen. The first part focuses attention on the type I1 region and is especially concerned with the variations which arise with chemically-reduced y-MnO,. The t.g./d.t.g.technique has been exploited to throw light on the controversial question of whether y-MnO, and a-MnOOH form a complete range of solid solutions. This part is supplemented by X-ray diffraction and magnetic studies. The second part of the paper has been aimed at a better understanding of the type I11 region. The relevance of the oxygen environment is that it delays the onset of bulk oxide decom- position which otherwise obscures the type I11 feature. Earlier work had briefly 237238 THERMAL BEHAVIOUK OF ?t-MnG2 indicated that type I11 weight loss involved both water and oxygen. Hi' sc), the resulting oxide would either be a mixture of phases or be a highly non-stoichiometric dioxide, and this merited further investigation.EXPERIMENTAL MATERIALS The manganese dioxide used in this work was mainly the commercially electrodeposited y-phase Mn02 previously described.'. Chemical reduction of this dioxide was eEected by hydrazine sulphate or hydrazine h ~ d r a t e . ~ Hydrazine sulphate (0.24-1.47 g) was added to a suspension of y-Mn02 (9.5 g) in distilled water (100 cm3) and the whole stirred for 2 h at 24+0.2"C. The solid was separated by filtration and vacuum dried at ambient tempera- ture. Hydrazine sulphate could be replaced with hydrazine hydrate solution (6 % w/w) and an identical procedure carried out using aliquots in the range 0-19 cm3. Thermal behaviour and X-ray diffraction patterns of reduced samples of the same composition prepred by the two routes were identical.The level of reduction of y-Mn02, denoted by x in the following equivalent formulae ; MniV- Mn;" 0 2 - OH, ; (I - x) Mn02. xMnOOH ; MnOOH,, ranged from x = 0.1 1 to x = 0.95, where x = 1 .O signifies groutite, a-h,lnBOH. The commercially electrodeposited y-Mn02 used as starting material had a composition corresponding to x = 0.11, Chemically prepared y-Mn02 was used for certain experiments ; this was obtained from Mn304 by the method of Giovanoli, Maurer and Feitkne~ht.~ METHODS The t.g./d.t.g. apparatus has been described previously.' Prior to t.g./d.t.g. runs samples were evacuated at room temperature to pressures of 5-7x lo-' Torr and then oxygen was admitted. In most experiments 10Torr of oxygen was used to provide the static gas environment, although in certain circumstances pressures up to 450 Torr were employed. Characterization by X-ray diffractometry and chemical analysis was carried out on the starting materials, and also on the solids at intermediate stages of t.g./d.t.g.by arresting the linear heating programme and evacuating the oxygen. Analyses were made as soon as possible after cooling to room temperature. X-ray diffraction data of the powdered solids were recorded as diffractometer traces using a Philips PW 1130 unit with Ni-filtered Cu Kdl radiation and a goniometer scan rate normally 1" mi+. The procedure for chemical analysis of MnTV was based on methods employed by Vetter and Jaeger and Chapman and Freeman.6 The sample, typically 60mg, was refluxed in iron stabilised sodium oxalate solution (10 cm3, 0.025 mol dm-3) containing 2 cm3, 20 % v/v sulphuric acid for 30 min.This solution was diluted to 25 cm3, heated to 350 K and titrated with KMn04 (0.02 niol dm-3) ( Vl). To the same solution was added sodium pyrophosphate (-4g) which dissolved. The pM was adjusted to 6.5-7.0 by addition of sodium hydroxide pellets and the solution was titrated potentiometrically with 0.02 rnol ~ l m - ~ KMn04 (V2). Finally the volume of KMn04 required to standardise 10 cm3 of sodium oxalate solution was noted. x was calculated from the relationship x = 4[1-5( Vo- V1)/(24V2- SV,)]. Thermohygrometric analysis (THA) was performed with a Stanton Redcroft Thermo- balance TG750 and a differential d.t.g. unit using a linear heating rate of 15 K n1in-l and an oxygen flow (at atmospheric pressure) of 25 cm3 min-l, the gas being pre-dried with magnesium perchlorate and a molecular sieve.7 A commercial electrolytic hygrometer (Salford Electrical Instruments) coupled to the thermobalance was employed to andyse the evolved water.Magnetic susceptibility measurements were made using a Gouy balance constructed and described by Hagan.8 Temperature could be varied from 80 to 300 K and magnetic field from 3000 to 5500 G.J . A . LEE, C. E . NEWNHAM, F . L . TYE A N D F . S . STONE 239 RESULTS c HEMI c A L LY - RE D u CED y-MnO, T . G . / D . T . G . I N THE TYPE 11 REGION Oxygen environment t.g./d.t.g. curves for samples in the compositional range 0.11 6 x < 0.35 are given in fig. 1 .In parallel with the degree of chemical reduction of the manganese dioxides an associated increase in water loss in the region of 410 K was observed. This feature was also present in the previous water environment studies2 It reflects the presence of additional liydrogcn-bonded water adsorption following the increase in surface hydroxylation which accompanies chemical reduction. FIG. 1.-D.t.g. curves in 10Torr of oxygen for partially reduced oxides. Extent of reduction increases in the sequence (a) +. (e). x in (1 - x ) MnOa. xMnOOH is equal to (a) 0.11 (- - -1, (b) 0.14 (- -), (c) 0.18 (-), (d) 0.26 (--) and (e) 0.35 (- -). Chemical reduction appeared to induce a modification in the t.g./d.t.g. curve of the " as-received '' MnOz (x = 0.1 1) in the type I1 region.An initial reduction from x = 0.1 1 to x = 0.14 depressed the weight loss between 430 and 530 IS (fig. l), reflecting a decrease in the amount of type TI water. It is to be noted that the untreated unreduced y-MnO, differed in sample history from the other samples studied. The important observation however was made on the samples which had been reduced beyond x = 0.14 and especially those from x = 0.18 to x = 0.35. These showed no dependence of the type I1 feature upon the degree of reduction. The peak, however, shifted slightly to higher temperatures and is accordingly designated as type IIb., Brouillet et aL9 noted a similar invariance and proposed a reaction route involving dehydroxylation of the oxyhydroxide component of (1 - x ) MnO,. xMnOOM simultaneously with oxidation in accordance with the overall reaction : The net loss in weight resulting from reaction (1) is small, and therefore in oxygen in contrast to near-vacuum or water vapour environments, little dependence of type PIb upon x would be expected.Differential thermal analysis (d.t.a.) in oxygen, reported in an earlier paper,l indicated a net exothermic reaction at these temperatures (420-590K). Since the simple condensation of hydroxyls and removal as water is endothermic, this is further evidence for the presence of a simultaneous (exothermic) 4 MnOOH -!- Q2 + 4 MnO, + 2H20. (1)240 THERMAL BEHAVIOUR OF y-MnO, oxidation step. Finally, X-ray diffraction analyses on samples removed from the balance in the region of 500 K detected only y-MnO,. All the data are therefore consistent with reaction (1) for 0.14 < x < 0.35. Fig.2 shows the d.t.g. curves for the reduced oxides in the compositional range 0.54 < x < 0.91. Surprisingly, behaviour in the type I1 region is now dependent on chemical composition. Both the amount and the rate of weight loss increased with increase in the tervalent manganese content within the oxide. Fig. 3 shows the weight loss 0.1 6 - 0.14- 0.12- r( *k o.lo- M 0.08- 8 3 0.06- CI .- B 2 0.04- c 0 44 0.02 - sample temperaturelK FIG. 2.-D.t.g. curves in 10 Torr of oxygen for x equal to (a) 0.54 (- - -), (b) 0.70 (- .), (c) 0.76 (- -) and (d) 0.91 (-). I 100 - r( 8 0 - \ 8 6 0 - bo 0" v) - 0.2 0.4 0.6 0 . 8 1.0 X FIG. 3.-Weight loss up to 550 K as a function of degree of reduction. a, 10 Torr oxygen; 0, vacuum.J .A . LEE, C. E. NEWNHAM, F* L. TYE A N D F. S . STONE 241 associated with the type I and II features as a function of x, and the new behaviour at x > 0.5 is evident. First we should consider whether the dependence of type I1 on x arises from a buoyancy effect (volume change) or the poor accessibility of oxygen to samples as the extent of reduction is increased. The buoyancy effect was accordingly estimated for the particular experimental conditions; it was found to be too small to contribute. Accessibility of oxygen was examined by recourse to a specific set of experiments on the x = 0.91 sample where the oxygen pressure during t.g./d.t.g. was varied from 4 to 456 Torr (fig. 4). The results show that the type I1 feature is not an artefact caused by oxygen starvation.Since buoyancy and lack of oxygen accessibility are ruled out, we conclude that the increase in peak height for type I1 at x > 0.5 (fig. 2) indicates that the decomposition route in this range of x involves an intermediate which is less susceptible to oxidation than is the case for less-reduced oxide (x < 0.5) where the Brouiilet mechanism [reaction (1)J holds. 0.24- 1 I I 1 400 5 0 0 600 700 sample temperature/K Fig. 4.-D.t.g. curves for x = 0.91 at various oxygen pressures increasing in the sequence (a) -+ (e). (a) 4 (-), (b) 53 (- -), (c) 258 (- -), (d) 358 (- -) and (e) 456 Ton (- - -). 1 2 3 4 5 6 7 8 9 heating ratePC min-l FIG. 5.-Weight loss for x = 0.91 in 10 Torr of oxygen as a function of heating rate.242 THERMAL BEHAVIOUR OF y-MnO, A further important result in fig.2 is the evidence of an absorption process above the type 11 feature, clearly shown by traversal of the base line for the most reduced sample. The absorption process may well be present for the other samples but is masked by the type I1 evolution process. The total weight loss from a t.g./d.t.g. decomposition process should be independent of the linear heating rate.l O However, as fig. 5 shows, the weight loss corresponding with the decomposition of the x = 0.91 sample (MnOlSs4) in an oxygen environment up to the onset of the absorption region is a function of the heating rate. The occurrence of a dependence on heating rate is confirmatory evidence that two or more competing reactions are present.X-RAY DIFFRACTION The " as-received " y-MnO, and the chemically-reduced samples (0.1 1 < x < 0.95) were also studied by X-ray diffractometry. The diffraction data for samples with x between 0.1 1 and 0.46 showed five reflec- tions in the range 70" > 20 > lo", and these were assigned to the 110,021, 121,221 and 002 planes of the diaspore unit cell. In view of the relatively small number of observable reflections and also low intensity, it was more significant to examine the variations in individual d-spacings with x than to attempt to evaluate lattice para- meters. As shown in fig. 6, where the respective d-spacings for four of the above planes are plotted against x, the crystal lattice initially dilated uniformly as the mean I I 1 1.59 i- X 0.20.40.60.8 1.0 X 0.20.40.60.8 1.0 0.20.4 0.60.8 1.0 X X FIG.6.-InterpIanar separation ( d ) as a function of composition for partially reduced y-Mn02. 0, Experimental points; 0, natural groutite.J . A . LEE, C. E . NEWNHAM, F . L . TYE AND F . S . STONE 243 valency of the manganese decreased, No new reflections appeared as x increased from 0.11 to 0.46, so the results are consistent with the formation of solid solution of MnOOH in y-MnO,. For x > 0.50, however, there was a marked change in the X-ray pattern. This can be seen in fig. 7, which shows a comparison of the diffractometer traces in the range 33" < 26 < 57" for six chemically reduced samples with x between 0.11 and 0.95. For the compositional range x = 0.46 to x = 0.70 the diffraction peaks decreased in intensity : the five reflections mentioned above could only be identified with difficulty, and the structure appeared to be reorganizing.Beyond x = 0.70, new reflections appeared and increased in intensity as x increased (those in the range 57 45 33 (b) " A FIG. 7.-X-ray diffraction traces (CuKor) of partially reduced y-Mn02. x in (1-x) Mn02. xMnOOH (a) 0.11, (b) 0.26, (c) 0.39, (4, 0.70, (e) 0.76 and (f) 0.91. 33" < 28 < 57" are shown in fig. 7). All the new reflections could be ascribed to the a-MnBOH structure, though the d-spacings were smaller than those of mineral groutite. There was no significant evidence for change in the unit cell size with x. Fig. 6 shows data for four d-spacings of four samples with x between 0.69 and 0.95 ; the values are almost constant.The correlation of these d-spacings with those of the 221, 021, 121 and 002 spacings of the diaspore structure, as implied in fig. 6, is arbitrary. To sum up, the results presented in fig. 6 and 7 are in accord with single-phase solid solution over only part of the range from y-MnO, to a-MnOOH. Attempts to isolate an intermediate by terminating t.g/.d.t.g. runs for x = 0.91 which had been heated in oxygen at selected temperatures up to 520 K and character- izing by X-ray diffraction were unsuccessful. Diffractometer traces showed only oxide with x = 0.91 or the oxidation product y-MnO,. However, in a corresponding experiment using synthetic groutite (a-MnOOH, x = 1 .O) as starting material, analysis244 THERMAL BEHAVIOUR OF y-Mn02 of material from a run terminated at 530 K showed the presence of corundum-phase Mn,O,.Structural details will be given elsewhere ; the result is of special interest since Mn20, ordinarily crystallises with the C-structure. MAGNETIC MEASUREMENTS Further information on the chemically-reduced samples was obtained from magnetic susceptibility measurements. Susceptibilities of seven samples were measured and the magnetic moment p was evaluated from the slopes of Curie-Weiss plots (x-l against T). These plots were linear over the range from 140 to 290 K. The variation of p with extent of reduction is shown in fig. 8. 3.5 - 0.2 0.4 0.6 0.8 1.0 X FIG. 8.-Magnetic moment as a function of sample composition for partially reduced y-MnOz. (-) Experimental, (- - -) calculated.p increased with reduction, as is to be expected for the change in valency from MnxV (pspin only = 3.87 B.M.) to Mn"' (,uspin only = 4.88 B.M.). In the range 0.1 < x < 0.4 the magnetic moments formed a consistent set, and no appreciable orbital contributions to the magnetic moment are indicated. This is entirely com- patible with the existence of a solid solution. At x - 0.5, however, where the X-ray data indicate a solid solution limit, the magnetic moment is anomalously high; an orbital contribution would be consistent with structural reorganisation and distortion of octahedral symmetry. Gabano and Labat l2 have also observed a similar effect on p at x = 0.5 for a sample of y-Mn02 reduced electrochemically. y-MnO, HEATED ABOVE 520 K T.G./D.T.G. AND THA IN THE TYPE 111 REGION that an oxygen environment at 10 Torr pressure enables type I11 weight loss to be resolved as a separate peak.The effect of increasing oxygen pressure has been studied, and fig. 9 shows results for x = 0.18. At high pressures of oxygen type 111 resolved into two components. Mass spectrometric analysis of the evolved species indicated that oxygen was liberated above 530 K and became the major product beyond 670 K.l Fig. 10 presents the simultaneously recorded differential thermogravimetric and thermohygrometric analyses for x = 0.11 under a dynamic flow of oxygen at atmospheric pressure. Results appertaining to the type I1 region fit the expected pattern. The evolved water vapour analysis fingerprinted the dehydroxylation It has already been notedJ .A . LEE, C. E. NEWNHAM, F . L . TYE AND F . S . STONE 245 rl 350 400 450 5 0 0 550 6 0 0 650 700 sample temperature/# FIG. 9.-D.t.g. curves for x = 0.18 at different oxygen pressures. (-) 4 Ton, (- - -) 190 Torr. 400 5 0 0 6 0 0 700 8 0 0 9 0 0 sample temperature/K FIG. 10.-Thermal hygrometric THA (- -) and d.t.g. (-) curve of y-MnOz in a dynamic oxygen environment. The ordinate for the d.t.g. curve is as indicated, that of the THA is arbitrary. process as registered by the largest rate peak. The two components comprising type I11 were present in the d.t.g. trace between 570 and 670 K. The first component of I11 yielded a specific water entity since a peak in the THA curve matches in position the d.t.g. hump at 590-600 K (fig. 10). The second component of type 111 is presumed to involve the loss of oxygen as the mass spectrometric analyses of the evolved gases detected principally oxygen at these temperatures and no THA signal is seen.This oxygen desorption process appeared to be divorced from the main dioxide lattice decomposition occurring later (i.e. 4 M n 0 , j 2Mn203 + 0,) and identified by the large d.t.g. peak at -870 K. It is possible that carbon present as an impurity in the commercially electrodeposited material might contribute to type 111. This, however, can be dismissed as the reason for the type I11 behaviour since a carbon-free synthetic y-MnO, produced by oxidation of Mn304 exhibited a type I11 component. Further- more an artificially increased carbon content within the dioxide failed to modify the t.g./d.t .g.behaviour.246 THERMAL IIEHAVIOUR OF y-MnO, The position maximum and value of rate of weight loss for type 111 was indepen- dent of the degree of reduction. However, the peak narrowed on the low temperature side and was more symmetrical for the reduced samples. Since it has been established that the thermal dehydroxylation in oxygen of a-MnBOH produces y-MnO,,l I this implies that the sample composition just prior to type I11 in t.g./d.t.g. runs corresponds to y-MnQ2 irrespective of the original degree of reduction. The oxygen contribution to type I11 does not then depend upon the initial oxidation state. The peak narrowing appeared to reflect a diminution of type TI1 water. X-RAY DIFFRACTION AND ANALYTICAL DATA X-ray diffraction measurements on samples withdrawn from d.t.g.experiments at the rate minima on either side of type I11 indicated a manganese dioxide phase transformation of y -+ P-MnO, during the type 111 process (fig. 11). For example the diffractometer line at 28 = 29" which is characteristic of the @ modification is present only after the loss of type 111. Significantly application of a linear heating programme to P-Mn02 resulted in the absence of a detectable weight loss in the region of type 111. 28 CU Kg FIG. 11.-X-ray diffractometer traces before (top) and after (bottom) type I11 weight loss. Table 1 contains analytical data for the electrodeposited and chemically prepared y-MnO, samples at the t.g./d.t.g. minimum before and after type 111 weight loss. The previous postulate of some oxygen loss is supported by the diminution in oxidation states after type 111.Column 4 of table 1 records the calculated weight loss due to oxygen removal based on the composition changes indicated by analysis and column 5 the experimental values calculated by dropping verticals from the minima on the rate curves to the temperature axis and integrating. Inaccuracies are inevitable in the extrapolation procedure and small errors in assessment of oxidation states are amplified in predictions of weight loss due to oxygen desorption. Never- theless it is considered that the values in columns 4 and 5 for the synthetic dioxideJ . A . LEE, C . E . NEWNHAM, F . L. TYE AND F . S . STONE 241 probably indicate only oxygen loss whereas with the electrodeposited material the duplex nature of type I11 is manifest and the considerably larger experimental weight loss reflects the additional removal of water as detected by THA.TABLE 1 .-ANALYTICAL DATA PERTAINING TO THE TYPE I11 REGION MnlO ratio MnlO ratio a t d.t.g. at.d.t.g. mnimum mrumum prior to after type I11 type I11 wt. loss wt. loss electro- deposited y-Mn02 MnO1.96 MnO1.93 chemically prepared y-Mn02 MnO1.98 Mn01.91 calculated weight loss Img g- 1 6 12 experimental weight loss Img g-l 20 9 DISCUSSION The distinctly different thermal characteristics in oxygen of materials in the two compositional ranges 0.11 < x < 0.35 and 0.54 4 x 4 0.91 described in this paper, together with the X-ray diffraction evidence (fig. 6 and 7), indicate the absence of a complete range of solid solution.It is concluded however that solid solution exists for solids with x between x = O(Mn0,) and x = 0.50 in agreement with electrode potential investigations performed by Bell and Huber and diffraction data reported by Giovanoli and Leuenberger.14 This is contrary to the work of others such as Bode and Schmier l 5 who claim the existence of a solid solution over the entire range from y-Mn02 to a-MnOOH. Critical examination of their data, however, suggests that their X-ray diffraction results are not consistent with a single homo- geneous phase. The solid solution limit of x - 0.50 established in this work may conform to a stabilized hydrogen bonded structure.16 The structural disordering beyond the homogeneous solution limit possibly reflects the effect of Jahn-Teller distortion at high MnlI1 contents.The structural implications of Jahn-Teller distortion in MnlI1 compounds are well authenticated. In the range 0.54 < x < 0.91 the solid may be either monophasic but different from that for x < 0.50 or may exist as two phases, with groutite forming at the expense of the stabilised, solid solution limit material. The invariance in the position of the X-ray diffraction lines and the increase in intensity in going from x = 0.70 to 0.91 supports the premise of two phases. The groutite phase should be more accurately described as near-groutite, as its d-spacings are slightly less than naturally occurring groutite. It is proposed that the modified groutite component decomposes under the t.g./ d.t.g.conditions to produce a corundum form of Mn,O, as an intermediate between MnOOH and y-MnO,. This proposal is made in view of the isolation of a corundum- structure sesquioxide by controlled decomposition of synthetic groutite to be reported elsewhere. Thus the steps are : 2MnOOH -+ Mn,O, (corundum) + H20 (2) 2Mn,O, +02 -+ 4Mn0,. (3) This corundum-structure intermediate probably suffers no immediate oxidation and allows reaction (2) to proceed to a reasonable extent before (3) begins. Distinctive evolution and absorption steps therefore result as can be seen in fig. 2. Giovandi and Leuenberger l4 thermally decomposed groutite to the monoclinic oxide Mn,08,248 THERMAL BEHAVIOUR OF y-MnO, at 470-550 K in a stream of dry oxygen. Additionally they reported that this oxide oxidized at temperatures in excess of 570 K to produce p-Mn0,.We were unable to detect Mn,08 as an intermediate in this work and in the latter stages of the absorption step only the y-phase manganese dioxide was apparent. Thus the forma- tion of Mn,08 is unlikely to be responsible for the observed weight loss and weight gain phenomena present in t.g./d.t.g. runs. The effect of increasing the pressure of the oxygen environment (fig. 4) can be qualitatively understood in terms of reaction (2) and (3). As the oxygen pressure increases, reaction (3) proceeds more quickly and the observed rate of weight loss will decrease. The phase transformation y -+ P-Mn02 as detected in the Type I11 region has been observed by other workers,17 but there are no reports in the literature of the transition being accompanied by a loss in weight (type 111).P-MnO, is also usually more stoichiometric than its y counterpart. The accommodation of such a large oxygen deficiency may indicate a crystallographic shear (CS)l structure. P-MnO, has the rutile structure. Rutile (Ti02) is well known to undergo crystallographic shear,lg* 2o and of the other first row transition metal oxides possessing the rutile structure CS plane formation has also been reported for VO, 21 and Cr02.22 It appears to be a requisite for formation of a CS structure that the metal concerned can form a corundum-phase trivalent oxide. With the recent confirmation of a corundum- phase Mn,O, l1 this requirement is met also by Mn, and although CS systems based on MnO, are not reported so far it is possible that MnOlSg1 (table 1) is an oxide containing CS planes.The sheared structure may form directly as an intermediate in the y -+ p transformation or result from a thermally induced shear defect in the newly formed pyrolusite lattice. As previously noted the loss of water in the type 111 region was peculiar to the electrodeposited y-Mn02. Fleischmann, Thirsk and Tordesillas 23 have studied the electrodeposition of y-MnO, and argue that the slow stage in crystal growth involves the dehydration of Mn(OH),. It is possible that at this lattice formation step the dehydration is not always complete and the water entity in type 111 reflects the reaction 0 2 Mn(OH)* + MnO, +2H,O. (4) The observed decrease in type 111 water with increasing levels of initial chemical reduction may mirror the reductions Mn(OH), + H+ + e -+ Mn00H + 2H,O ( 5 ) Mn(OH),+2H++2e --+ Mn(OH), +2H20.(6) The authors thank the Directors of Ever Ready Co. (Holdings) Ltd for permission to publish this paper and Mrs. J. L. Hitchcock for obtaining the thermohygrometric data. J. A. Lee, C. E. Newnham and F. L. Tye, J. Colloid Interface Sci., 1973, 42, 372. J. A. Lee, C. E. Newnham, F. S. Stone and F. L. Tye, J. Colloid Interface Sci., 1973, 45, 289. W. Feitknecht, H. R. Oswald and U. Feitknecht-Steinmann, Helv. Chim. Acta, 1960, 43, 239. R. Giovanoli, R. Mauer and W. Feitknecht, Helv. Chim. Acta, 1967, 50, 1072. K. J. Vetter and N. Jaeger, Electrochim. Acta, 1966, 11, 401. D. S. Freeman and W. G. Chapman, Analyst, 1971, 96, 865. A. Hagan, Ph.D. Thesis (University of Bristol, 1974). Ph. Brouillet, A. Grund, F. Jolas and R. Mellet, in Batteries 2, ed. D. H. Collins (Pergamon Press, 1965), p. 189. ' J. L. Hitchcock and P. F. Pelter, 4th h t . Conf. on Thermal Analysis (Budapest, 1974). lo G. Carter, Vacuum, 1962, 12,245.J . A . LEE, C . E. NEWNHAM, F. L. TYEANDF. S. STONE 249 l 1 J. A. Lee, C. E. Newnham, F. S. Stone and F. L. Tye, to be published. l2 J. P. Gabano and J. Labat, Compt. rend., 1967, 264, 164. l3 G. S. Bell and R. Huber, J. Electrochem. SOC., 1964, 1, 11 1. l4 R. Giovanoli and U. Leuenberger, Helv. Chim. Acta, 1969, 52, 2333. l5 M. Bode and A. Schmier, in Proceedings of the 3rd International Symposium on Batteries, ed. D. H. Collins (Pergamon Press, 1962), p. 329. S. Atlung, Manganese Dioxide Symposium (The Electrochemical SOC., Cleveland, U.S.A., 1975), paper no. 2. J. M. A. Laudy and P. M. de Wolff, J. Appl. Sci. Res., 1963, 10, 157. l8 J. S. Anderson, in Surface and Defect Properties of Solids (The Chemical Society, 1972), vol. I, p. 1. S. Andersson, B. Collen, U. Kuylenstierna and A. Magneli, Acta Chem. Scand., 1957,11,1641. 'O B. G. Hyde, Pruc. 7th Int. Symp. on Reactivity of Solids (Bristol, 1972) (Chapman and Hall, London, 1972), p. 23. " A. D. Wadsley, Rev. Pure Appl. Chem., 1955, 5, 165. '' M. A. Alario Franco, J. M. Thomas and R. D. Shannon, J. Solid State Chem., 1974, 9, 261. 23 M. Fleischmann, M. R. Thirsk and M. Tordesillas, Trans. Faraday SOC., 1962, 58, 1865. (PAPER 6/2044)

ISSN:0300-9599    

DOI:10.1039/F19787400237

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Influence of Sodium on the Physico-chemical and Catalytic Properties of Magnesium Oxide BY JACEK KIJE~SKI AND STANISLAW MALINOWSKI* Institute of Organic Chemistry and Technology, Technical University (Politechnika), Koszykowa 75, 00 662 Warszawa, Poland Received 1 1 th February, 1977 Physico-chemical and catalytic properties of two series of catalysts comprising magnesia and sodium were determined. The first (I) series of catalysts was prepared by doping magnesia with varying amounts of NaOH. The second (11) series of catalysts was obtained by evaporating metallic sodium onto MgO preparations precalcined at different temperatures. The concentration and strengths of basic and acid sites, as well as the amounts of one-electron donor and one-electron acceptor sites, were measured, and the specific surface areas of the catalysts determined.Catalytic activity in isomerization of pent-1-ene, tram-pent-2-ene and the conversion of cumene was studied. It was concluded that the (11) series of catalysts displayed remarkably strong basic and one-electron donor properties. Also, it was proved that basic sites coexist on the surfaces of catalysts with the one-electron donor sites. Catalytic activity in alkene isomerization and cumene dehydrogenation was unambiguously associated with the presence of well-defined surface active sites. In previous papers it was shown that addition of alkali metal to oxide catalysts brings about a considerable change in their physical and catalytic properties, the variation being non-linear. 1-6 The aim of the present study was to investigate the sodium effect on the physico- chemical and catalytic properties of magnesium oxide.Magnesia is known to be one of the strongest solid bases,7* and oxygen anions 0’- of low coordination number are considered to be responsible for its basic properties. Taking into consideration the low first ionization energy of sodium compared with that of magnesium, it was anticipated that the reaction of sodium with the oxygen ions of MgO surfaces would result in an increase in the effective negative charge of oxygen anions. The immediate result of such a reaction should be a significant rise in both the basic and one-electron donor power of the catalytic system. In the present work, acid-base, radical and catalytic properties of two sets of MgO + Na catalysts have been studied.The catalysts were prepared (I) by impregna- tion of MgO with NaOH in aqueous solution, and (11) by evaporation of metallic sodium onto a MgO surface. EXPERIMENTAL PREPARATION OF CATALYSTS Magnesium hydroxide was obtained by hydrolysing Mg(NO& 6H20 with aqueous ammonia. The resulting precipitate was dried at 60 and 120°C and subsequently heat- treated at 550°C for 16 h to give pure magnesia. Both drying and calcination were performed in a flow of argon. Catalysts of the MgO+NaOH series were prepared by impregnating magnesia with aqueous NaOH solutions of varying concentrations. The preparations were treated in the manner described above. The amount of sodium introduced was measured 250J . K I J E ~ S K I AND s. MALINOWSKI 25 1 using an atomic absorption spectrometer.The catalysts under consideration were found to contain 0 ; 0.005 ; 0.01 ; 0.07 ; 0.35 and 0.82 mmol Na per g MgO respectively. Metallic sodium was evaporated onto the magnesia surface at 480°C under Torr pressure (1 Tom = 133.32 N m-2). In one run 200 mg of sodium was deposited on 3 g of MgO. After all the sodium was used up the temperature was raised to 550°C and maintained at this level for 2.5 h, the pressure being Torr. SPECIFIC SURFACE AREA MEASUREMENTS The method of thermal desorption of argon was employed to measure the specific surface areas of catalysts under examination. A gas mixture containing 5 % argon and 95 % hydrogen was used for the determinations. BASE A N D ACID PROPERTIES The number of basic sites at various strengths were determined by titrating the catalysts with benzoic acid solutions in benzene in the presence of appropriate Hammett indicators.The following were applied as basicity indicators : 2,4,6-trinitroaniline (PKa = 12.2) ; dinitroaniline (PKa = 15.0) ; 4-chloro-Znitroaniline (PKa = 17.6) ; 4-nitroaniline (PKa = 18.4) ; diphenylamine (PKa = 22.3) ; 4-chloraniline (PKa = 26.5) ; aniline (pKB = 27) ; triphenylmethane (PKa = 33); diphenylmethane (PKa = 35). Acidity was measured by titrating the surface with a benzene solution of n-butylamine. Dicinnamalacetone (PKa = - 3.0), benzalacetophenone (pKa = - 5.6) and anthraquinone (PKa = - 8.2) were used as zcidity indicators. Since nearly all the indicators used assumed the colour of the corres- ponding anions in the presence of extremely strong surface basic sites, it was impossible to determine the acidity of the catalysts with evaporated sodium.The only finding in this connection was that anthraquinone adsorbed on the MgO+ Namet catalysts does not undergo colour change, this being indicative of the absence of acid sites at a strength of Ho < -8.2. SURFACE FREE RADICAL PROPERTIES The one-electron donating or accepting properties of the catalysts were determined by adsorbing organic molecules, i.e. one-electron acceptors or donors respectively, onto the catalyst surface and recording the e.s.r. signals of the resulting ion radicals. Perylene (ionization energy (IE) 7.22 eV) and pyrene (IE 7.55 eV) are known to form cation radicals on oxide surfaces by donating one electron to the surface accepting s i t e ~ .~ - l l Tetra- cyanoethylene [electron affinity (EA) 2.8 eV], s-trinitrobenzene (EA 1.7 eV), m-dinitrobenzene (EA 1.4 eV) and nitrobenzene (EA 0.7 eV)? are capable of accepting one electron from the surface donor sites to give the corresponding anion radica1s.l’. l3 Tetracyanoethylene is thought to react readily with all one-electron donor centres (including those of low one- electron donating power) in contrast with TNB, DNB and NB which are only able to react with correspondingly stronger sites. Adsorption of electron donors or acceptors from benzene solutions was carried out under argon. The samples were kept under argon while recording e.s.r. spectra. REACTION PROCEDURE Conversion of cuniene was carried out in a conventional flow reactor with a fixed catalyst bed.The space velocity was 1 g cumene per 1 g catalyst per hour. Isomerization of pent-1-ene and trans-pent-2-ene was performed in a batch-type set-up with stirring at 20°C. Reaction products were analysed using a gas chromatograph Chrom 4 with a 50 m capillary Squalane column. Products of isomerizations were analysed at room temperature, whereas the analyses of cumene conversion products required a temperature of 90°C. Mass spectrometry was employed for additional identification of products. The mass spectra were recorded using Varian MAT 11 1 spectrometer. t s-TNB is sym-trinitrobenzene, rn-DNB is metu-dinitrobenzene and NB is nitrobenzene,252 Na INFLUENCE ON MgO PROPERTIES POISONING OF THE ACTIVE SITES The catalyst was suspended in a benzene solution of Hammett indicator or one-electron acceptor, the amount of which was stoichiometric with respect to the number of corres- ponding active sites on the catalyst surface.The suspension was stirred for 4 h at room temperature. Benzene was then distilled off in a vacuum of Torr and dry deoxidized argon was admitted to the flask with the catalyst. E.S.R. MEASUREMENTS The e.s.r. spectra were registered using a Jeol JES-MC3X spectrometer. Preparations with preadsorbed electron donors or acceptors were all studied at room temperature, while those with preadsorbed cumene or pent-1-ene were being examined at - 100°C. RESULTS The acid-base properties of catalysts prepared by impregnating MgO with aqueous solutions of NaOH are close to those of the starting material, i.e. pure MgO.The reaction of magnesium oxide with sodium ions occurring on the MgO surface does not affect the maximum acid-base strength of the system. As a result only the number of centres is modified to an extent dependent on the quantity of NaOH added (table 1). Maximum concentration of both basic and acid surface sites was noted for the catalyst containing 0.35 mmol Na+ per g MgO ; and the lowest concentration of both types of sites was found for the catalyst doped with 0.82 mmol Na+ per g MgO The basic properties of both pure MgO and MgO doped with NaOH weakened with increase in the temperature of calcination. Thus, at 1000°C, the strongest basic sites disappear (27 < H- < 33) and the concentration of basic sites at lower strengths decreases substantially.TABLE CO CONCENTRATIONS OF BASIC AND ACID SITES FOUND FOR CATALYSTS DOPED WITH DEFERENT AMOUNTS OF SODIUM HYDROXIDE concentrations sodium content calcination mmol m-2 of acid centres! in catalyst temperature specific surface mmol m-2 concentrations of basic centres at various strengths/ /(mmol/g MgO) 1°C area/m2 8-1 12.2 < H- 18.4 d H- 27 < H- Ho < -3.0 < 15 <22.3 <33 total 0 0.005 0.01 0.07 0.35 0.082 550 750 lo00 550 750 lo00 550 750 1000 550 750 lo00 550 750 lo00 550 750 1000 64 58 48 72 65 51 81 73 36 58 35 31 29 26 22 28 25 22 0.015 0.006 0.006 0.014 0.008 0.005 0.014 0.006 0.008 0.01 1 0.01 5 0.01 1 0.031 0.019 0.017 0.008 0.010 0.003 0.012 0.007 0.006 0.010 0.014 0.004 0.008 0.008 0.009 0.017 0.020 0.013 0.026 0.018 0.01 6 0.020 0.002 0.002 0.023 0.01 1 0 0.021 0.014 0 0.021 0.01 1 0 0.010 0.013 0 0.01 5 0.014 0 0.007 0.001 0 0.050 0.024 0.012 0.045 0.036 0.009 0.043 0,025 0.017 0.038 0.048 0.024 0.072 0.051 0.033 0.035 0.013 0.005 0.007 0.01 5 0.012 0.005 0.012 0.010 0.007 0.012 0.015 0.006 0.029 0.027 0.014 0.033 0.029 0.010 0.027 0.019J .K I J E ~ S K I AND s. MALINOWSKI 253 An insignificant rise in the concentration of acid sites can be observed when the calcination temperature is raised from 550 to 750°C. Catalysts calcined at 750 and 1 000°C appeared to have practically identical acid properties. The one-electron donating properties of magnesium oxide undergo an insignificant change upon doping with sodium hydroxide. The greatest concentration of one-electron donor centres capable of reducing TCNET was found in the catalyst containing 0.35 mmol Na+ per g MgO (fig.1 ).0 0,005 0.01 037 0.35 0.82 30 25 20 T: k ri 15 \ .f3 8 10 5 mmol Na per g MgO FIG. 1 .-Concentrations of one-electron donor centres on magnesia doped with varying quantities of NaOH and the yields of cumene transformation reactions over theselcatalysts. Since both the electron donor and the basic properties are most pronounced for the same catalyst in the series, i.e. the one containing 0.35 mmol Na+ per g MgO, it was necessary to find out if donating one electron or a pair of electrons can be attributed to one and the same active site on the surface. The fact that the observed concentra- tions of one-electron donor centres are much lower than those of basic sites seems to favour the idea that there must be different types of sites responsible for the two properties.Additionally, no e.s.r. signals of anion radicals are observed upon adsorption of Hammett indicators, which leads to the conclusion that the apparent colour change is caused by the process of donating an electron pair rather than one electron to the Hammett indicator molecule. In order to suppress all basic sites, 2,4,6-trinitroaniline (pK, = 12.2) was adsorbed on the catalyst surface, its amount corresponding stoichiometrically to the number of detected basic sites. TCNE was subsequently adsorbed on such a catalyst. E.s.r. spectroscopy revealed (TCNE)- anion radicals in a quantity equal to that observed for the catalyst containing no Hammett indicator.This result gives support to the supposition that there are two different types of sites : one responsible for one-electron donating properties and the other for the basic properties of the catalyst. t TCNE is tetracyanoethylene.254 The surface reaction between MgQ and metallic sodium results in the catalytic system having salient basic properties. Catalysts with evaporated sodium were found to possess basic sites of extreme strength (superbase properties If- 2 35). Na INFLUENCE ON MgO PROPERTIES 8 N z E X I N totol :.8, /---- MgO pretreatment temp./"(= FIG. 2.-Concentrations of basic sites at various strengths as found for the catalysts (MgO) doped with metallic sodium. ! MgO pretreatment temp /"C FIG. 3 .-Concentrations of one-electron donor centres on catalysts doped with metallic sodium.J .K I J E ~ S K I AND s. MALINOWSKI 255 To date, centres exhibiting such extraordinary strength have not been reported in the literature. The greatest concentration of superbase centres as well as those of strength 27 < H- < 33 were detected for catalysts containing magnesium oxide pretreated at relatively lower temperatures, i.e. 550 and 650°C. Both superbasic and basic centres at 27 4 H- < 33 disappear for catalysts prepared from MgO precalcined at higher temperatures; only the concentration of basic centres at a strength of 18.4 < H- < 22.3 is practically independent of the temperature of MgO pretreatment. The one-electron donor sites concentration against temperature plots for catalysts doped with metallic sodium are given in fig.3. Exceedingly strong sites of a one- electron donor type were found for catalysts belonging to this series ; they can reduce a nitrobenzene molecule. A maximum concentration of one-electron donor sites was noted for the catalytic system Mg0700+Namet. The general conclusion is that catalysts obtained by evaporating sodium onto magnesia pretreated at higher tempera- tures (700-1000°C) have considerably stronger free radical properties than those prepared from MgO activated at lower temperatures, 550 and 650°C. No one-electron accepting centres were found for catalysts prepared by doping magnesia with NaOH and metallic sodium. Paramagnetic cation radicals were not formed even upon adsorption of such strong electron donors as perylene and pyrene.CATALYTIC ACTIVITY I SOMERIZATION Pure magnesia and preparations of MgO doped with sodium hydroxide were completely inactive in isomerization of pent- 1-ene and trans-pent-2-ene under adopted conditions. Under the same conditions magnesia doped with metallic sodium was found to exhibit remarkable isomerizing activity (table 2). It follows from the results summarized in table 2 that increasing the precalcination temperature of MgO substantially influences the isomerizing activity of the catalyst. Both the initial &/trans ratio of pent-2-ene formed through isomerization and the ratio of isomers after 2 and 10 h of reaction undergo a significant change. TABLE 2 . V A L U E S OF PENT-2-ENE CiS/tUanS RATIO AND PENT-1-ENE CONVERSION OBTAINED OVER THE MgO+Na,,t SERIES CATALYSTS cisltrans ratio of pent- 1 -ene cisitruns pent-2-ene conversion initial ratio of pent-1-ene over catalyst over catalyst cisltrans pent-2-ene conversion poisoned with poisoned with specific surface ratio of after 2 after 2 TPM after 2 TPM after 2 catalyst area/m2 g-1 pent-2-ene and 10 h and 10 h/ % and 10 h and 10 h/ % In the case of catalysts with marked superbase properties (Mg0550 +Namet, Mg0650 +Name# the composition of the reaction mixture in steady state conditions approaches equilibrium composition at 20°C and the initial cisltrans ratio of pent-2- ene isomers becomes close to 2.With catalysts comprising MgO precalcined at higher temperatures, i.e. those with dominating radical properties, the selectivity t Mg055o+Namet and MgO65o+Namet etc.are catalysts obtained by evaporation of metallic sodium onto MgO surface calcined at respectively 550, 650°C, etc. 1-9256 Na INFLUENCE ON MgO PROPERTIES towards cis pent-2-ene increases, the initial cisltrans ratio of pent-2-ene isomers reaches - 3 and the composition of the reaction mixture under steady state conditions differs considerably from the equilibrium composition. Since neither pure magnesium oxide nor the MgO specimens doped with NaBH showed any isomerizing activity, it was reasonable to link this activity with active sites existing only on the metallic sodium doped surface of magnesia. Thus it became evident that the sites responsible for the isomerizing activity of the catalysts being studied are the superbase centres (H- 2 35) or the one-electron donor sites able to reduce nitrobenzene molecules.Triphenylmethane (pK, = 33) was adsorbed on the Mg05s0 + Namet catalyst surface (possessing a high concentration of superbase sites) and on the MgO700 + Namet catalyst characterized by the largest amount of one-electron donor centres. The quantity of TPM* corresponded to the amount of strongest basic sites. The results of pent-1-ene isomerization over these catalysts are presented in table 2. In the presence of Mg0550+Nam,t poisoned with TPM the value of the cis- pent-2-eneltrans-pent-2-ene ratio changed from 0.29 (the value being attained with unpoisoned catalyst) to 3.2 after a lapse of 2 h reaction time. The poisoned Mg0700 + Namet catalyst gave a smaller change in cisltrans ratio.Also, the conversion of pent-1-ene decreased to a lesser extent. The observed decrease in pent-1-ene conversion over poisoned Mg05 + Namet catalyst is presum- ably the result of both poisoning the superbase sites and suppressing the remaining active sites (one-electron donor centres) by physically adsorbed triphenylmethane. As the quantity of TPM introduced onto the surface of MgO,OO + Namet catalyst is markedly smaller and the amount of superbase sites on this catalyst is also much smaller, it is felt that the probability of accidentally suppressing the one-electron donor sites with the introduced poison was much lower. Hence we consider the diminishing amount of pent-1-ene conversion to be due to the elimination of superbase centres by the poison.Moreover, it should be emphasized that values of the cis/trmzs ratio attained over poisoned catalyst with TPM are close to those obtained over MgOlooo +Namet catalyst, in which no ionic sites at the basic strength of H- 3 35 were detected. The isomerization of trans-pent-2-ene to cis-pent-2-ene and pent-1-ene was carried out over MgOeSo + Namet and MgOlooo + Namet catalysts, the ones most distinctly varying in physicochemical properties. The catalyst MgOlooo + Namet exhibited no activity towards the isomerization of trans-pent-2-ene, whereas the reaction was observed to proceed over the h4g0650+Namet catalyst to yield, after 2 h, a mixture of cis-(11.5 %) and tuans-(87 7:) pent-2-ene and pent-1-ene (1.5 %). Blocking the superbase centres of this catalyst with a stoichioinetric quantity of TPM rendered the catalyst entirely inactive towards isomerization.The e.s.r. spectroscopic method revealed a signal for the organic radical resulting from pent-1-ene adsorbed on the surface of Mg0700 + Namet catalyst [fig. 4(a)]. The g value of this signal was estimated to be 2.0012. CUMENE TRANSFORMATIONS The yields of cumene transformation reactions over catalysts doped with sodium hydroxide are shown in fig. 1. The maximum conversion of cumene is reached with the catalyst containing 0.35 mmol Na+ per g MgO, the main product being a-methyl- styrene with all the catalysts studied. The only exception is the catalyst comprising 0.01 mmol NaOH over which the formation of a-methylstyrene is accompanied by equivalent yields of toluene.The variation in a-methylstyrene yields with the amount * TPM is triphenylmethaneJ . K I J E ~ S K I AND s . MALINOWSKI 257 (4 (b) FIG. 4.-E.s.r. siguals of organic radicals obtained after (a) pent-1-ene adsorption on the Mg,o,+ Namet surface, (b) cumene adsorption on the Mg0,50 t- Namet surface. I O E - 1-1 : : 553 650 7W 750 MgO pretreatment temp. /"C FIG. 5.-Yields of cumene transformations products obtained over MgO + Namet catalysts.258 Na INFLUENCE ON MgO PROPERTIES of sodium accurately corresponds to the variation in concentration of one-electron donor sites capable of reducing TCNE. The variations in cumene conversion to ethylbenzene and styrene are similar to those of a-methylstyrene yields. The results of the cumene conversion over catalysts with evaporated sodium are demonstrated in fig.5. Catalysts doped with metallic sodium were found to be more active than those doped with sodium hydroxide and the rezction appeared to be more selective, the main product being a-methylstyrene resulting from the dehydrogenation of cumene. Among the reaction products were ethylbenzene, n-propylbenzene, toluene, benzene and styrene. The largest yields of the majority of cumene conversion products are obtained over the hfgB750 + Namet catalyst. This catalyst possesses a relatively large number of one-electron donor sites and an insignificant concentration of saper- base sites (fig. 1 and 3). The formation of n-propylbenzene was favoured over Mg0650+Namet catalyst which was the most abundant in superbase centres (fig.I). Table 3 summarizes the results of the conversion of cumene over MgQ750 4- Namet (the most active) as a function of temperature. TABLE 3 .-CUMENE DEHYDROGENATION YIELDS AT VARIOUS TEMPERATURES reaction temperature/'C 20 250 350 450 500 550 yields* of a-methylstyrene/m2 - 0.03 0.09 0.22 0.27 0.91 other products - I - - benzene, ethyl benzene, toluene, styrene, n-propyl benzene 0.14 0.23 * mole per 100 moles of cumene The first product to appear in conversion of cumene at 250°C was a-methylstyrene. Other products were found only after the temperature was raised to 500°C. The reaction with cumene was carried out in the presence of the MgQ750 + Name* catalyst whose surface was devoid of one-electron donor centres by suppressing them with stoichiometric quantity of TCNE.As a result, a decrease in the yield of a-methyl- styrene was observed (table 4). The products did not contain styrene. The yield of toluene was the least affected. TABLE 4.-TCNE ADSORPTION INFLUENCE ON ACTIVITY OF Mg07;3+ Namet CATALYST IN CUMENE TRANSFORMATIONS cumene reactions products/(mole/ 100 moles of cumene) m2 ethyl- a-methyl- n-propyl- catalyst benzene toluene benzene styrene styrene benzene M g 0 7 5 o + N a m e t 0.03 0.04 0.12 0.01 0.91 0.03 MgO, 5 0 + Namet after TCNE adsorption I 0.01 0.01 - 0.10 - Cumene vapours were adsorbed at 250°C on the Mg0,50 +Namct catalyst surface. It should be remembered that the only reaction product at this temperature was a-methylstyrene (table 3). The e.s.r. spectra were measured for the catalyst with preadsorbed cumene.A strong signal originating from the organic radical was recorded [fig. 4(b), g = 2.0064, AH,,, = 1.3 GI. The narrow signal width isJ . KIJEASKI AND s. MALINOWSKI 259 presumably due to strong exchange interactions among adsorbed radical species. Formation of paramagnetic surface species may be considered sufficient evidence of the radical reaction being initiated at 250°C. DISCUSSION Sodium addition to magnesia brings about alterations in the physicocheinical properties of the oxide. Thus, if sodium is introduced in the ionic form (NaOH) the only apparent change is in the amount of active sites. If, on the other hand sodium is vaporized onto magnesia, the resulting system differs from the parent one in the quantity as well as quality of active sites (superbase or strong one-electron donating properties are generated).We suppose that in both cases the modifying action of sodium consists in its interacting with surface oxygen atom groupings (lattice oxygen anions or oxygen atoms from adsorbed water). We observed that suppressing the basic sites had no effect on the concentration of one-electron donor sites. This observation led us to conclude that these centres exist on the surface of MgO + NaOH catalyst entirely independently. It is however beyond any doubt that in the series of catalysts doped with NaOH there exists a parallelism between basic and one-electron donor activity of the surface, which was confirmed by Cordischi and Indovina who investigated the one-electron donor properties of a set of ionic oxides.14 For catalysts sputtered with sodium vapours we observed no parallelism between basic (maximum for MgO,,, + Namet catalyst) and one-electron donor activity (maximum for Mg0700 + Namet catalyst).A similar differentiation of properties can also be observed for pure MgO. The strongest basic properties were noted for MgO specimens calcined at temperatures from 550-600°C (heating over 600°C results in rapid decrease of catalyst basicity), which was also observed by Tanabe and Hattoris The maximum of one-electron donor properties of MgO falls in the calcination temperature range from 700-800°C.1 Comparison of the results of isomerization reactions with the variations in basic and one-electron donating properties in the series of catalysts being tested, as well as the results of experiments involving poisoning of active sites, warrant the statement that two different reaction mechanisms are operative in isomerization.These are : the ion-type mechanism (involving the superbase sites) and the radical-type mechanism (where one-electron donor sites are responsible for catalytic activity). Superbase sites catalyse isomerization of pent-1-ene to give a mixture of pent-2-eiie isomers and isomerization of cis-pent-2-ene towards trans-pent-2-ene as shown in the diagram : cis-pent-2-ene trans-pent-2-ene. pent- 1 -ene Jr It is most probable that the initial stage of these reactions involves proton abstraction from the alkene molecule by the superbase site (pK, = 37 for pentenes). Therefore a mechanism similar to that proposed by Pines and Schaap l5 can hold in this case a3 0 B + R--CH2--CH=CH2 -+ BH[R-cH-CH=CH2] (2) BH[R--C_H--CH=-CI3,] t) BH[R-CH=CH---C_H,] (3) BH[R-CH=CH-GH,] + B + R-CH=CH-CHS] (4) (3 8 fi3 8 8 e where B is a superbase centre.260 The initial cisltrans ratio of pent-2-ene isomers obtained in our experiments was found to be greater than unity.This result might be accounted for by the higher stability of the intermediate cis-ally1 carbanion as compared with that of trans-ally1 species.16 A further stage would be the removal of a proton from the resulting cis- or trans-pent-2-ene. As a result a new allyl-type carbanion would be formed : Na INFLUENCE ON MgO PROPERTIES el3 B + R-CH=CH-CH, -+ BH[R-CH==CH=-==CHJe ( 5 ) which on addition of one proton would transform into trans- or cis-pent-2-ene or pent-1-ene.The main product of the consecutive reaction is the most thermo- dynamically stable trans-pent-2-ene. The values of the initial cisltrans ratio and the cisltrans ratio after 2 and 10 h measured for Mg0550 +Namet and MgQ650 +Name, catalysts (table 4) support the reaction pathway described above. Superbase sites are considered to activate trans-pent-2-ene in the same manner ; as a result of proton abstraction from the trans-alkene molecule, an ally1 carbanion is formed which on subsequent addition of a proton may be transformed either to cis- and trans-pent-2-ene or to pent- 1 -ene. Over catalysts possessing one-electron donor sites isomerization of pent-1-ene gives a mixture of isomers with the cis-form as the major product of the reaction.In this case the reaction pathway may be envisaged as follows : (one-electron donor centre). + R-CH2-CH=CH2 -+ (one-electron donor centre)-H + R-kH-CH=CH2 (6) R-cH-CH=CH2 0 ++ R-CH=CH--dH2 (7) R-CH=CH-CH2 + (one-electron donor centre)-H -+ R-CH=CH-CH3 + (one-electron donor centre).. (8) According to Walling and Thaler l7 cis-pent-2-ene is the most favoured radical isomerization product from the statistical point of view. As suggested by these authors cis-ally1 radical (the intermediate for cis-pent-2-ene) is formed through hydrogen atom abstraction from one of the two equivalent gauche conformers of pent-1-ene. On the other hand, trans-ally1 radical is formed via hydrogen atom abstraction from one possible trans conformer of pent-1-ene. Hence there is a greater likelihood of cis pent-2-ene formation.A similar mechanism might play a substantial role in the reactions studied in this work. The cisltrans ratio of the pent-2-enes calculated at the end of the reaction does not correspond to the theoretical equilibrium ratio; this is probably the consequence of the absence of an internal double bond isomerization process, i.e. cis-pent-2-ene to trans-pent-2-ene. In favour of this view is the complete lack of trans-pent-2-ene transformation over catalysts possessing exclusively one-electron donor sites (MgO, ooo + Nilmet and Mg8 6 + Name* poisoned with TPM). The correlations found between the yields of cumene conversion products, i.e. a-methylstyrene, ethylbenzene and styrene and the variation in one-electron donating properties of catalysts doped with NaOH and metallic sodium are indicative of a free radical nature of these transformations.The idea of a radical type of reaction pathway is further supported by the fact that, after poisoning the catalyst with TCNE, dehydrogenation is no longer observed. Moreover, paramagnetic organic radicals are formed from cumene adsorbed on Mg0750 + Namet catalysts. The radical mechanism of cumene dehydrogenation may be analogous to that put forward by Krause * for ethylbenzene dehydrogenation on one-electron donor centres :J . K I J E ~ S K I AND s . MALINOWSKI 26 1 (one-electron donor centre). + C6H5C2H5 -+ (one-electron donor centre)-H + C6H5C2H4 + (one-electron donor centre)-H 4- C6H,&H4 (9 (10) 0 C&&CH=CH2 4 H2 + (one-electron donor centre). and n-propylbenzene seems to be formed in the ionic pathway. The greatest amount of this compound was obtained over catalysts with the strongest superbase properties (MgOS5* +Namet and Mg0650 +Name*), possibly formed by realkylation of benzene (or toluene) with propylene (or ethylene). According to the mechanism proposed by Pines and Schaap l5 one may write : %3 e B + C6H5CH3 -+ Bh-I[C6H5_CH2] G3 e J. M. Parera and N. S . Figoli, J. Catalysis, 1969, 14, 303. L. D. Scharme, J. Phys. Chem., 1974,20,2070. T. T. Chuang and I. G. DaIla Lona, J.C.S. Faraduy I, 1972,68,777. H. Bremer, K. H. Steinberg and K. D. Wendlant, 2. anorg. Chem., 1969, 366, 130. H. Pines and W. 0. Haag, J. Amer. Chem. Soc., 1960,82,2471. S . Santhangopalan and C. N. Pillai, Indian J. Chem., 1973, 11, 957. ' J. Take, N. Kikuchi and Y . Yoneda, J. Catalysis, 1971, 21, 164. H. Hattori, N. Yoshii and K. Tanabe, Proc. Fifrrh Jnt. Congress on Catalysis (Amsterdam, 1972), vol. 10, p. 1. B. D. Flockhart, I. A. N. Scott and R. C. Pink, Trans. Faraduy SOC., 1966, 62, 730. BH[C,H5CH2CH2CHZ] -+B + CbH5CW2CH2CHS (1 3) lo J. H. de Boer and J. Weissnian, J. Anzer. Chem. SOC., 1958,80, 4549. l 1 J. Kijehski, S. Malinowski and B. Zielihski, Kinetics Catalysis React. Letters, 1976, 4(2), 251. l2 A. I. Tench and R. L. Nelson, Trans. Faraday SOC., 1967,63,2254. l3 M. Che, C. Naccache and B. Imelik, J. Catalysis, 1972, 24, 328. l4 D. Cordischi and V. Indovina, J.C.S. Farada-v I, 1976, 72,2341. l 5 H. Pines and L. A. Schaap, Adu. CataZysis, 1960, XII, 117. l 6 S. Bank, A. Schriesheim and C. A. Rowe, J. Amer. Chem. SOC., 1965, 87, 3244; S. Bank, l 7 C. Walling and W. Thaler, J. Amer. Chem. Soc., 1961, 83, 3877. J . Amer. Chem. SOC., 1965,87, 3245. A. Krause, Sci. Pharm., 1970, 38, 266. (PAPER 71238)

ISSN:0300-9599    

DOI:10.1039/F19787400250

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Mg(63PJ Photosensitization of 3-Methylbut-1-enet Part 1 .-Reactions of Vibrationally Excited Triplet 3-Methylbut-1-ene and the 2-Methylbuta-l,3-diyl Biradical BY DEREK C. MONTAGUE* Department of Physical Chemistry, The University, Leeds, LS2 9JT Received 14th February, 1977 3-Methylbut-1-ene has been photosensitized by Hg(63P1) atoms in the gas phase at 245 1"C, producing both vibrationally excited 3-rnethylbuta-ly2-diyl and 2-methylbuta-1,3-diyl biradicals in the triplet state. The pressure dependences of the quantum yields of the various reaction products have been investigated and a computer modelling treatment applied to the proposed reaction mechanism, thereby allowing rate constants for the major decomposition and isomerization reactions of these two intermediates to be derived.In addition RRKM calculations have been performed, enabling rough estimates of critical energies for three of these reactions to be obtained. In a previous paper the results of a comprehensive quantitative study of the Hg(63P,) photosensitization of 3,3-dimethylbut-l-ene have been presented.l Rate constant values and critical energy estimates for various reactions of both the excited triplet olefin and the 2,2-dimethylbuta-l,2-diyl biradical were reported. The com- putational methods of data analysis developed for this previous investigation have now been applied to results from the Hg(Ci3P1) photosensitization of 3-methylbut-1 -me (3MB), thereby allowing similar rate constant and critical energy data to be obtained. Vibrationally excited triplet olefins, formed by energy transfer from Hg(63P1),2 react predominantly by the fission of weak C-H and/or C-C bonds, and by hydrogen atom migration reactions leading to isomerization products.Collisional deactiva- tion of the excited olefin also occurs, and this introduces a pressure dependence of the product yields. In the 3MB system a 174-hydrogen atom transfer will produce the 2-methylbuta- 1 ,3-diyl biradical (MBD13). This biradical has been postulated as an intermediate in the structural isomerization of 1,2-dimethylcyclopropane,4 the photochemical decompositions of both 2,3-dimethylcylobutanone and 3,4-dimethyl-A1-pyrazoline,6 and in the reaction of triplet methylene with but-2-ene.' A complete quantitative understanding of the reactive channels open to this intermediate is clearly essential to the elucidation of the often complex chemical behaviour of these various systems.Achievement of this goal, at least in part,-formed the primary objective of this study. In addition rate data for the reactions of vibrationally excited triplet 3MB were sought. The principal advantage of the method used here is that the reaction intermediates have a reasonably narrow spread in excitation energies, thereby enabling the experimental rate constants to be compared in a meaningful way with those calculated by RRKM theory. In contrast, triplet MBD13 formed from 2,3-dimethyl- cyclobutanone and 3,4-dimethyl-A'-pyrazoline has a much wider spread in vibrational energy due to energy partitioning during CO and N2 loss respectively. Presented in part at the Third International Symposium on Gas Kinetics, Brussels, 1973.262D. C. MONTAGUE 263 EXPERIMENTAL 3-Methylbut-1-ene (3MB), stated purity 99.9 %, was obtained from Matheson Gas Products. It contained no impurities that interfered with the gas chromatographic analysis of reaction products and was therefore used without further purification, other than thorough degassing in uacuo. The grease-free vacuum system, photolysis cell, lamp and analytical apparatus have all been described previously. The percentage conversion of the olefin was always <0.5 %. Reaction mixtures were usually analysed on a 100 m wall-coated polypropylene glycol (PPG) capillary column, fitted to a Perkin-Elmer 452 gas Chromatograph that was equipped with a flame ionisation detector.By operating this column at O'C, the yields of all major products from methane to 2,3-dimethylpentane could be quantitatively measured. Some analyses were also performed on a 45 ft x inch silicone oil and Borapak combination column, using a Perkin-Elmer F11 gas chromatograph with flame ionisation detector. Peak areas on the chart recorder trace, estimated by planimetry, were converted to absolute product yields using previously determined calibration factors. Products were identified by a comparison of their retention times with those of authentic samples of compounds thought to be present. The detector sensitivity was measured for ethane, but- 2-ene, 3MB, 3,3-dimethylbut-l-ene (33Dh4B) and 2-methylpent-2-ene (2MP2), and found to be proportional to the carbon number of the compound, within experimental error.Detector sensitivities for other hydrocarbon products were assumed to be similarly related. Actinornetry was carried out as described previously using the Hg(63P1) photosensitized isomerization of cis-but-2-ene. The mercury photosensitization quantum yield for this process has recently been confirmed as 0.5 at pressures in excess of 30 Torr* and low overall conversion level^.^ Using a value of 0.7 nm2 for the absolute quenching cross-section of 3MB, estimated by analogy with the corrected lo cross-section for pent-l-ene,l' a calculation similar to that carried out for 33DMB shows that, in the pressure range of these experiments, essentially all the excited Hg(63P,) atoms are quenched by the olefin. RESULTS In the pressure range 11 3-764 Torr, the Hg(63P,) photosensitization of 3MB yields as major products methane, ethane, trans- and cis-pent-2-ene (tP2 and cP2), 2-methylbut-2-ene (2MB2), 2-methylpent-2-ene (2MP2), 2,3-dimethylbutane (23DMB) and 1,2-dimethylcyclopropane (DMCP).Smaller amounts of 3,3-dimethylbut- 1 -ene (33DMB), 2-methylpentane (2MP), 2,3-dimethylpentane (23DMP), 4-methylpent- 2-ene, 2-methylbut-1-ene (2MBl) and isoprene were found, as were trace quantities (in order of decreasing yields) of buta-l,3-diene, but-2-ene, but-1-ene, isobutene, propane and propylene. Also observed were traces of 2,3-dimethylbut-1 -ene and 3-methylpent-2-ene at pressures below M 100 Torr, and an additional product that was formed in yields lower than those of 33DMB.Despite considerable effort this compound, which eluted from the PPG column between 2MB2 and isoprene, defied identification, many potential candidates, including ethylcyclopropane, buta-l,2-dieiie and 3-methylbuta-l,2-diene being eliminated. Its characterization using conventional techniques was unfortunately precluded by its small yields. Methylcyclobutane was not identified among the products. It cannot be positively ruled out, however, as if it were formed in trace 'quantities, it would not necessarily have been observed, as it would have been incompletely resolved from tP2 in the product chromatogram. Its maximum quantum yield is estimated as Similarly 1 ,I-dimethylcyclopropane could not be characterised as a reaction product as its retention time was almost identical with that of the parent 3MB.Attempts to find an alternative chromatographic column suitable for its analysis were hindered by complications arising from the presence of the other products in the mixture. Data * 1 Ton = 133.3 Pa264 3-METHYLBUT- 1 -ENE PHOTOSENSITIZATION from other mercury + olefin photosensitizations indicate that its yield would probably be minor, however.12 Higher molecular weight compounds were also almost certainly formed, but no attempt was made either to identify them or to assess their yields quantitatively. The mercury photosensitization quantum yields (Q) of the major products are presented in table I. Most of the product yields decrease, within experimental error, as the pressure increases. Notable exceptions are those of DMCP, 2MB2 and methane, all of which appear to pass through a maximum.Other similar examples of this behaviour have been noted previous1y.l Both trans- and cis-DMCP were formed, in a ratio that was pressure dependent. This finding is discussed in greater detail in the following paper.13 reactant pressure /Tom 11.8 24.7 37.5 48.9 86.9 101.0 127.8 140.8 177.0 187.3 221.2 228.8 251.6 274.6 350.2 374.1 407.6 451.2 463.4 497.9 548.8 577.5 599.0 691.7 764.4 TABLE 1 .-Hg(63P1) PHOTOSENSITIZATION PRODUCT (I QUANTUM YIELDS ( x lo3) CH4 C2H6 ZP2 cP2 2R.181 2hfB2 33DbLB 3MP2 23DM5 2MP 23DMP C&12 'DMCP nm nm nm nm 0.30 nm nm nm nm nm ntn nm n:n F.31 ntn nm nix nm 6.03 63.26 35.09 25.25 0.46 1.92 1.01 5.25 6.11 nm nm nm nm 0.37 nm nm nm nm 8.96 40.70 24.03 16.74 0.36 2.35 1.27 8.20 7.08 9.15 37.68 21.21 14.95 0.39 2.56 1.07 6.37 5.21 9.57 29.05 18.08 12.89 0.31 2.54 0.83 4.55 5.52 nm nm 16.24 11.77 0.19 2.31 nm nm nm 10.29 22.31 13.53 9.73 0.22 2.47 0.88 5.12 3.97 9.86 19.8U 12.23 8.93 0.23 2.59 1.06 6.05 4.56 10.01 17.50 10.77 8.06 0.33 2.33 0.88 5.67 4.19 nm nm nm nm 0.14 2.35 0.94 6.58 nm 10.01 14.40 9.56 6.92 0.31 2.39 0.97 5.48 4.13 10.26 12.28 7.89 5.85 0.18 2.25 0.76 4.82 2.96 9.01 9.90 7.34 5.15 0.25 2.23 0.67 4.01 2.77 9.64 8.31 5.25 3.79 0.26 2.08 0.58 3.65 2.11 9.20 8.53 5.95 4.14 0.20 2.25 0.60 4.05 2.26 9.20 7.41 4.86 3.50 0.34 2.14 0.53 3.64 1.81 9.12 6.42 4.28 3.00 0.24 2.05 0.44 3.22 1.7'3 8.74 6.07 4.05 2.86 0.25 2.09 0.53 3.70 1.85 8.35 5.36 3.42 2.42 0.25 2.08 0.39 3.11 1.41 8.97 5.59 3.48 2.46 0.26 2.08 0.33 2.28 1.32 8.13 4.36 2.66 1.96 0.27 1.93 0.35 2.39 1.32 8.51 4.40 2.54 1.90 0.20 2.13 0.34 2.49 1.19 7.02 3.15 1.55 1.31 0.18 1.88 0.25 1.96 0.85 a See text for explanation of symbols ; b 4-methylpent-2-ene ; nrn = nm nm nm nm 0.41 0.51 nm nm 1.14 0.44 0.73 0.64 0.64 nm nm nm 0.61 0.62 0.61 0.47 0.75 0.64 nm nm 0.83 0.45 0.64 0.74 0.69 0.29 0.62 0.61 0.66 0.61 0.59 0.33 0.64 0.54 0.57 0.71 0.57 0.53 0.56 0.48 0.60 0.49 0.59 0.39 0.58 0.31 not measured nm nm 0.19 0.28 0.38 0.25 0.08 nm 0.13 nm 0.i5 0.26 0.15 0.18 0.09 0.10 0.15 0.22 0.23 0.13 0.09 0.06 0.08 nm nm 13.00 14.05 13.49 14.23 14.40 14.30 13.89 13.89 13.00 13.60 12.60 12.71 12.02 12.00 10.71 9.99 10.00 9.50 9.13 8.57 8.16 7.87 7.79 7.05 6.37 A few runs were carried out in which the intensity of the light from the photolysis lamp was reduced by a factor of ten.In these experiments (CD)DMCP remained essentially unchanged at any given pressure. The quantum yield of methane increased, however, while that of ethane decreased considerably. In addition the ratio of Q23DMB to @2MP decreased, as did the values of (Dtp2 and QCp2, although Qtpz/cDcPz was constant. Particular care was therefore taken to ensure that under " normal " experimental conditions, photolyses were performed with as constant an intensity as possible. In a separate series of experiments, discussed elsewhere,13 varying amounts of oxygen were added to several different pressures of 3MB before photolysis. In these runs the yields of all major products, except DMCP and 2MB2, fell to trace levels, while those of C3 and C4 products increased slightly.DISCUSSION Vibrationally excited triplet olefins, formed by energy transfer from Hg(63P,), normally react by one of two processes. Examination of the product quantum yields observed at different photolysis light intensities reveals that both are present in thisD. C. MONTAGUE 265 system. Thus fragmentation to radicals is evidenced by the intensity dependent yields of those products with radical precursors, and, secondly, isomerization by the intensity independent yield of DMCP. In addition, the low overall total product quantum yield shows that many excited olefin molecules are collisionally deactivated before reaction can occur. FORMATION AND REMOVAL OF RADICALS The 3MB molecule contains two equivalent weak C-C bonds /3 to the n-bond.Fission of one of these bonds either in the vibrationally excited triplet olefin 3(3MB)*, or in the excited ground singlet state olefin (3MB)*, formed by inter-system crossing, leads to both methyl and 1-methylallyl radicals. The following mechanism surn- marises the possible reactions : Hg(63P,)+3MB +- Hg(G1So)f3(3MB)" 3(3hIB)* -4 (3MB)" 3(3MB)* 3 CH3 + CH3CH"CHE'-CH2 (3MB)* 3 CH3 + CH,CH"CH-CH, (3) (4) M 3(3MB)* 3 3(3MB) M (3MB)* +- 3MB 2CH3 4 C2H6 ( 7 ) tP2 ( 8 4 (86) 7 CH3 + CH,CH=CH"'CH2 4 cP2 'X 3MB. ( 8 4 1,l-Dimethylallyl radicals are also produced, by rupture of the weak C--H bond fi to the double bond. The hydrogen atoms thus liberated probably lead to the formation of molecular hydrogen, the yield of which was not measured, though a considerable fraction of them appear to be scavenged by the parent olefin, giving the 3-methylbut-2-yl and 3-methylbut-1-yl radicals.At the pressures of these experi- ments both of these radicals are essentially completely stabilised before decomposition can occur. The formation of ethane, tP2, cP2, 33DMB, 2MP2, 23DMB and 2MP can all be explained by combination of the various radicals 33DMB 2MP2 CH3 + (CH3)2CECHECH2 < CH3 + (CH3)2CH(?HCH3 3 23DMB CH3 + (CH3),CHCH2CH2 3 2MP as shown in reactions (7)-(11). The ratio, K,, of tP2 to cP2 was constant, within experimental error, over the pressure range 38-764 Torr. It has been argued that the value of K , , 1.3910.02, reflects the thermodynamic equilibrium ratio of trans- and cis-1-methylallyl radicals at 297+ 1 K, the temperature of these experiments.This value has therefore been combined with data obtained at other temperatures to estimate the differences in entropy and heat of formation of the two isomeric forms266 3-METHYLBUT-I-ENE PHOTOSENSITIZATION of this radical.* Obviously the yield of 3MB from reaction (8c) cannot be measured in these experiments. It can, however, be calculated using data from other experi- ments. The ratio of the quantum yields of 2MP2 and 33DMB formed in reaction (9) is a measure of the relative reactivity, a,,,, of the two reactive centres in the 1,l- dimethylallyl radical. The derivation of aIIi = 6.1 rfi0.6 from the data of these and other experiments has been discussed elsewhere. Alternative pathways leading to 23DMB and 2MP can be postulated.CH3 addition to 3MB gives both the 2-methylpent-3-yl radical and the 2,3-dimethylbut-l-yl radical. Various disproportionation and abstraction reactions of these radicals produce 23DMP and 2MP. Two observations suggest that these pathways are less important than reactions (10) and (1 l), however. First (CH3)2CHkHCH2CH3 (CH3)2CHCH(CH3)eH2 CH3 + 3MB < (CH3)2CHeHCH2CH3 CH3 +{(CH3)2CHCH(CH3)kH2 (1 3) the yields of 23DMP, produced by reaction (13), are low, suggesting that reactions (12~) and (12b) are relatively slow in this sytem ; and secondly, the quantum yield ratio oF23DMB to 2MP is always greater than unity, but decreases as the overall pressure increases, paralleling a concomitant decrease in the rate of other radical combination reactions, e.g.reactions (7)-@). These experimental findings are in line with the proposal that at low pressures the favoured production of 3-methyl-but-2-yl radicals by addition of EI atoms to 3MB, ensures that reaction (10) is important, whereas at higher pressures, the decreased rate of formation of M atoms and the lower overall steady state radical concentrations, lead to a greater fraction of CH3 radicals reacting by addition to 3MB, mainly via reaction (12a), resulting in a relatively increased yield of 2MP. Observations from the reduced light intensity experiments support this qualitative explanation. Under these conditions the rate of radical production becomes so low that radical-molecule reactions compete even more effectively with radical-radical interactions, thereby reducing Q2 DMB/@2MP still further.There are two potential pathways that can contribute to the formation of methane from CH3 radicals, viz., hydrogen abstraction from 3MB, and various disproportiona- tion reactions. The latter routes will be most important at low pressures, where radicals are removed predominantly by radical-radical processes. As the pressure of 3MB increases their importance will decrease relative to H abstraction, which will become dominant. The abstraction rate cannot be varying linearly with 3MB pressure, however, as QCzH6 is not constant. The product P = Q&L6 [3MB] reflects the form of the variation. Values of P rise at low pressures but reach a plateau at pressures around 300 Torr, thereafter increasing very much more slowly.In addition, the ratio of P values at 764.4 and 37.5 Torr is approximately four times larger than the corresponding ratio of methane quantum yields, indicating the substantial contribution of disproportionation processes at low pressures. No useful analytical expression can be derived to describe iDCH4 against pressure, as a result of the large number of reactions that influence the interdependent stationary state concentrations of the radicals involved in disproportionation. Radical disproportionation reactions account for the formation of the small yields of buta-l,3-diene, isoprene, 2MB2, 2MP2, 4-methyl-2-pentene, but-2-ene and but-1-ene. A further route to 2MB2 is proposed in the following section.D . C. MONTAGUE 267 ISOMERIZATION REACTIONS Four hydrogen atom translocation reactions are possible in 3(3MB)*.They are, a 1,4-shift (termed l5 a 5pp process) giving the 2-methylbuta-l,3-diyl biradical (MBD13), two 1,3-shifts, namely a 4tp and a 4ps process producing triplet 2MB2 and the 2-methylbuta-l,4diyl biradical (MBD 14) respectively, and a 3ts process giving the 3-methylbuta-l,3-diyl (3MBDl3) biradical. These isomerizations, summarized by reactions (14)-(17), would presumably yield vibrationally excited products in the triplet state : - r r 3(3MB)* + son 3[kH2CH(CH3)kH2CH3]* (14) 4tP '(3MB)* + 4PS 3(3MB)* $ 3(3MB)* + 3ts 3 ( 2 ~ ~ 2 ) * (1 5 ) 3[kH2CH(CH3)CH2cH2]* (1 6) 3[kH2CH2e(CH3)2]*. (17) The appreciable yields of DMCP throughout the experimental pressure range indicate that reaction (14) is important. Addition of oxygen to the reaction mixture only suppresses the yields of 2MB2 slightly, suggesting that reaction (1 5) probably provides the major route to this product.Excited MBD14 would be expected either to fragment to ethylene and propylene or collapse to methylcyclobutane. The failure to observe this latter compound among the products and the minor yields of propylene demonstrate that reaction (1 6) occurs to a negligible extent. Similarly, in the Hg(63P1) photosensitization of 33DMB, the rate of the analogous 5pp process greatly exceeded that of the 4ps process.1 As mentioned earlier analytical difficulties prevented an assessment of the almost certainly minor role played by reaction (17). Following its genesis, 3(MBDl 3)* undergoes either decomposition,l collisional deactivation, or intersystem crossing, as shown in reactions (1 8)-(20) : 3(MBD13)* + CH3 +CH,CH"CH"'CH2 (1 8) M 3(MBD13)* -+ 3(MBD13) + DMCP 3(MBD13)* -+ (MBD13)* -+ DMCP.Both reactions (19) and (20) are included in the mechanism because the ratio of trans- to cis-DMCP was observed to be pressure dependent. Such behaviour cannot otherwise be explained, geometrical isomerization of (DMCP)" being at least an order of magnitude too slow. Cyclisation of 3(MBD13)* takes place via inter-system crossing to the singlet state. The factors controlling processes of this type have been discussed.17 The nature of the " singlet biradicals " thus produced is still a matter of some debate,l* though in this case one may confidently assert that the presence of a high level of vibrational excitation in the molecule ensures that the rate of ring closure would be much faster than that of any decomposition reaction, irrespective of the exact magnitude of any possible cyclisation energy barrier.Thus the extensive decoinposition that occurs via reaction (1 8), vide infra, implies that the biradicals formed in reaction (14) must be in the triplet state, as presumed. Fragmentation by a hydrogen atom loss process analogous to reaction (18) cannot occur to any great extent, as the product 1,2-dimethylallyl radicals would go on to give 2,3-dimethylbut- 1-ene and 3-methylpent-2-ene by combination with CH3, compounds that were only observed in minor trace amounts at pressures below268 3-METHYLBUT- 1-ENE PHOTOSENSITIZATION z 100 Torr.Structural isomerisation of (DMCP)* can also be ruled out at the pressures of these experiments. The small yields of 2MB1 that are found must therefore arise by isomerization of (MBD13)" and/or MBD13. The contributions from 3(MBDB)* and 3(MBD13) are negligible as 1,Zhydrogen shifts in triplet biradicals are particularly unfavourable due to the high energy barrier.12* l9 The much lower barriers operating for singlet biradicals result from energy release accompanying n-bond formation as the isomerization proceeds along the reaction coordinate. If 2MB1 is indeed produced in this way, then approximately equivalent yields of 2MB2 would also be expected to be formed in a parallel rearrangement.? A schematic simplified representation of the major reactions of 3(3MB)* and 3(MBD13)* is shown in fig.1. Minor routes and bimolecular radical reactions have been omitted in the interests of clarity. 3(h) - h \ T M T \ FIG. 1.-Major reactions of 3(3MB)* and 3(MBD13)*. QUANTITATIVE FORMULATION OF THE MECHANISM If the usual assumptions of the stationary state hypothesis are made for the concentrations of 3(3MB)*, (3MB)*, 3(MBD13)*, (MBD13)* and MBD13 during photolysis, application of a kinetic analysis to the reaction mechanism enables the sum of the quantum yields of DMCP, 2MB1 and the 2MB2 formed by isomerization of the singlet biradical, to be expressed as (21) where [MI is the total pressure of 3MB and a = k20/k19, y = k5/k14:, 6 = k18/k19, E = k--14/k19 and p = k'/k14, where k' is the sum of the rate constants for reactions t In the gas phase thermal isomerization of both trans- and cis-DMCP, the ratio of the rate of formation of 2MB2 to that of 2MB1 at the mid-point of the temperature range of the study (727 K) was found to be 1.07.The Arrhenius parameters for both pathways are identical within experi- mental error.20 +@nmi + @Yi% = (a + EMI)/{(I +P +WI)(a + 8 + E + [MI)-&)D . C . MONTAGUE 269 (2j, (3) and (15) plus any other possible reactions of 3(3MB)*. If it is assumed (i) that k14 = 2k-14, i.e. that the rate constants per transferable hydrogen atom for reactions (14) and (- 14) are equivalent, (ii) that the collisional deactivation rate constants for 3(3MB)* and 3(MBD13)*, k5 and k19, are equal, and (iii) that the yields of 2MB2 and 2MB1 from singlet biradical isomerization are equivalent, eqn (21) can be simplified to eqn (22) %MCP+2@2MBi = (a+IMI)/((l +D+[MI/~&)(~+~+E+[MI)-&~. (22) An expression for the ratio of the mercury photosensitization quantum yield of methyl radicals to the sum of the yields of DMCP and the singlet biradical isomeriza- tion products 2MB1 and 2MB2 can also be derived.Again by assuming @)2MB1 = @y$&, this expression reduces to eqn (23) @Me/(@D,MCP+2@2MSl) = ((a+d+&+ [M1)(c+q/(l +e[M1)) +6)/(a+[M1) (23) where = k3/kI4, q = k2/k14 and 8 = ks/k4. QMe is given by @Me = @CH4 + 2a)CzHs + 1 *~(@,Pz + @cP2) + @2)2MP2 + @3 3DMB + Q23DMB + @2Mp + 2@)23DMp + a(4-methylpent-2-ene). (24) Eqn (24) assumes that no other products are formed by methyl radicals. Under these experimental conditions this assumption is reasonable, though minor yields of higher molecular weight compounds produced by the addition of the methyl adduct radicals 2-methylpent-3-yl and 2,3-dimethylbut- 1 -yl to 3MB may be produced. The observed high quantum yield of C2W6 and the low quantum yields of 2MP and 23DMP, imply that the yields would be very small, however.Eqn (22) and (23) have an identical form to those successfully used to analyse the data obtained in the Hg(63P,) photosensitization of 33DMB.l As the complex forms of both equations preclude the use of any simple method to obtain values for the unknown rate constant ratios, it was necessary to adopt a computer modelling approach which employed an iterative procedure that refined initial estimates of the unknown parameters. The " best " values thus obtained were defined as those minimising the sum of the squares of the deviations between the calculated and experimental data points.In view of the large number of unknowns in eqn (22) the error hypersurface defining the agreement between the computed and experimental data was explored using a series of fixed values for a and/or 6, thus enabling corres- ponding estimates for p and E to be obtained. Eqn (23) was similarly treated using the same fixed values for a and/or 6, together with the appropriate E value and 0 = 0.0826 Torr-l, calculated by RRKM theory (vide infra). It was found that the two pairs of a and 6 values that minimised the two error functions were similar, but not identical, presumably as a result of the slight scatter in the experimental data.Compromise values were therefore derived that nevertheless gave good agreement between the computed and experimental values for both sets of data. Table 2 shows the optimum values thus obtained. It is worthwhile re-emphasising that this method of data analysis can sometimes give misleading or meaningless results, as there may be more than one set of parameter values that are commensurate with the data. In this case, however, no evidence could be found for an alternative set of meaningful values, and it would therefore appear that within reasonable error limits, that proposed is unique. Of the unknown parameters the values of E and 5 were very insensitive to changes in ct and 6, only requiring alteration by up to 10 and 15 % respectively, in order to still achieve acceptable fits to the data.An attempt was made to place realistic error limits on the quoted optimum values of cc and 6 despite the difficulties inherent in2'70 3-METHYLBUT- 1 -ENE YHOTOSENSITIZA'TION such an undertaking. The computed results show that they are probably accurate to within factors of 1.5 and 1.2 respectively. Previous experience in treating the results from the H S ( ~ ~ P ~ ) photosensitization of 33DMB suggested that the values derived for 8 and q would be closely inter-related and that equally good fits would be obtained for many different pairs of values for these parameters. Consequently no attempt was made here to derive 8 from the data, its value being calculated using RRKM theory. The optimum fit with a = 33.6 Torr and 6 = 302 Torr gave y equal to zero.Increasing a first to 34 and then to 35 Torr whilst holding 6 constant, TABLE 2.-oPTIhaUM RATE CONSTANT RATIOS AND HALF-QUENCHING PRESSURES a a/Torr G/Torr &/Torr O/Torr-1 4 rl 33.6 302 4.145 0.0826 2.345 6 1 C See text for explanation of symbols ; b calculated value, see text ; C see text. resulted in q values of 0.04 and 0.2. Increasing 6 by 1 % gives y equal to zero and 0.14 for a < 34 and 35 Torr respectively. Clearly these data do not allow y to be defined with any precision, though an order-of-magnitude upper limit of unity could be estimated. Its true value is probably considerably less than this. If y is indeed zero then methyl radicals are not formed by (3MB)" decomposition, reaction (4), in this system. On the other hand the fragmentation of 3(3MB)* must be important, as satisfactory fits to the data could not be achieved with set equal to zero.Thus of the three postulated methyl forming reactions, (3), (18) and (4), only the latter can be classified as not being essential to the mechanism. Fig. 2 and 3 show plots of (QDMCp + 2@2MBI) and @MVle/(CDDMCp + 2@2MB1) respectively against pressure, together with computer generated " optimum fit " curves to the data points calculated from the parameter values listed in table 2. As noted previously it is possible that the exclusion of potential high molecular weight product quantum yields from eqn (24) might result in values of a)Me/(@DMCp+2@e)2MBI) that are low. The effect of increasing OMe by an amount DHMW, arbitrarily set equal to 2@23DMp, was therefore investigated.Using the same 8 value as before, the optimum 6 value 0 I I I I I 1 I 1 200 LOO 600 pressure/Torr FIG. 2.-Variation of (%MCP+ 22QM1-31) with pressure.D. C. MONTAGUE 27 I 0 200 400 600 pressure/Torr FIG. 3.-Variation of the ratio @)M~/(@)DMcP+ 2Qi2m 1) with pressure. was found to decrease by z 3 Torr, while that of a increased by xl Torr and [ by ~ 0 . 0 5 . These small variations are well within the error limits for these parameters. The ratio of rate constants for 3(3MB)'k fragmentation versus rearrangement to 3(MBD13)* is given by 5, while 5/y gives the relative rates of fragmentation and intersystem crossing. The uncertainty in rules out an exact determination of this ratio, but a value below 2 seems unlikely.The most favourable reaction channel available to 3(MBD13)* is decomposition, which is faster than ring closure by the factor 6/a = 9.0. The reactive pattern outlined by these results resembles that found for the mercury photosensitization of 33DMB, and, as in that system, the rate constant for CH, loss from the excited triplet biradical exceeds that from the excited triplet olefin, here measured by the factor 8/2~5 = 15.3, despite a statistical factor of two favouring the latter reaction. Estimating a collision diameter for 3MB of 560pm, by analogy with molecules of similar molecular weight and carbon skeleton, enables the collision rate constant TABLE 3 .-EXPERIMENTAL RATE CONSTANTS AND COMPUTED CRITICAL ENERGIES < 1.13 - - I (2) 226 (3) 2.64 (15.5+0.5) {:::;:$ 206 (4) (1.64+0.55)c (16.6+0.3)d (2S4_+4)e 471 r93 226 1.13 i:' 206 226 206 (128k 9) 226 (10518) 186 (1 8) 41 .O (15.1k0.5) { (20) 4.56 - - - a Estimated ; b error limits result from uncertainty in log (A/+) ; C computed value ; dreported value ;31 e calculated from reported activation energy ;31 fpreferred value.272 3-METHYLBUT- -ENE PHOTOSENSITIZATION to be calculated as 1.36 x lo7 s-I Torr-l (2.52 x 1014 cm3 mol-1 s-l) at 24°C.If it is assumed that the collision diameters of the reactive intermediates are also 560 pm, and that the strong collision assumption holds, the rate constants shown in table 3 can then be computed from the half-quenching pressures and rate constant ratios given in table 2. The value of the rate constant for intersystem crossing of 3(MBD13)* is, within the estimated error limits, the same as that found previously for intersystem crossing of the excited triplet 2,2-dimethylbuta-lY3-diyl biradical. Uncertainty in k , makes a comparison with the analogous rate constafit for 3(33DMB)'y more difficult, but it would appear probable that they are again of a similar magnitude.Quantitative data on the intersystem crossing rate constants for simple triplet acyclic olefins and propa-l,3-diyl biradicals is lacking. The values obtained here are, however, in accord with those estimated by considering the factors controlling the rates of these processes.17 T HER M 0 C 13 E MI S T R Y The heat of formation of the relaxed triplet 31MB biradical was calculated as 216 kJ in01 from the reported standard heats of hydrogenation 21 and formation 2 2 of 3MB (- 127 and -29 kJ mol-l respectively), by the same method as that used to compute the corresponding value for 3(33DMB).1 As reaction (14) is formally thermoneutral, AHf"[3(MBD13)] was also taken to be 216 kJ mol-I.It can be argued, however, that AH,"[3(3MB)] and AHf"[3(MBD13)] could be up to 20 and 40 kS 11101-1 greater respectively, by applying analogous reasoning to that previously used when discussing the heats of formation of 3(33DMB) and the triplet 2,2- dimethylbuta- 1,3-diyl biradica1.l These higher biradical AH," values imply a degree of interactive destabilization between the radical centres. If this interaction is indeed present, then it is probable that its magnitude is different in the two biradicals.In view of these uncertainties the RRKM calculations described below have been carried out using both the estimated minimum and maximum AH; values. The average total internal excitation energy, E'*, of initially formed 3(3MB)* and 3(MBD23)* can be readily computed to be at most 226 kJ mol-1 in excess of the residual average thermal energy, Eth, of 3MB, from the calculated thermochemistry and the known excitation energy of Hg(63P1) atoms (471 kJ mol-I). Using higher AHf" values for 3MB and/or MBD13 reduces accordingly. RRKM ESTIMATION OF CRITICAL ENERGIES RRKM theory enables the energy dependence of the specific rate constant, kE, of a unimolecular reaction, to be related to the critical energy, E,, if the A factor is known. For the triplet biradical reactions studied here the A-factors are not avail- able, and it was therefore necessary to estimate them in order to obtain rough values of the critical energies for these processes.This procedure is not as difficult as might be expected, as experimental pre-exponential factors have been measured for several similar reactions. Except for (3MB)* decomposition, the procedure adopted was to investigate the dependence of kE on Eo for fixed values of E*. In this way the value of Eo that resulted in equivalent calculated and experimental rate constants could be determined. The RRKM computations were performed assuming that all overall molecular rotations were inactive and that the reactants were monoenergetic. The semi-classical Whitten-Rabinovitch approximation was used to evaluate the sums and densities of the molecular quantum states.23 Details of the activated complex models used are given in the Appendix.Calculations were performed for four processes. TheD. C. MQNTAGUE 273 rate constant for (3MB)* fragmentation, k4, was first computed so that a value for 8 might be substituted in eqn (23), thereby reducing the number of variable parameters when computer fitting this equation to the experimental data. The other three reactions treated were the decompositions of 3(3MB)*, reaction (3), and 3(MBD 13)*, reaction (18), and the isomerization of 3(3MB)*, reaction (14). The results of the calculations are shown in table 3. One feature of the results is the difference in critical energies for the decomposition reactions (3) and (1 8).This disparity presumably accounts for the experimental observation that kI8 is some fifteen times larger than k3. A similar conclusion concerning the fragmentations of 3(33DMB)* and the excited triplet 2,2-dimethylbuta- 1,3-diyl biradical was rationalised by postulating that whereas the favoured orthogonal configuration of the two free electron orbitals in the triplet 1,2-biradical inhibits extensive resonance stabilisation in the incipient allylic system, the preferred (0,O) configuration of the propa-l,3-diyl biradical allows allylic stabilisation of the developing radical centre to be immediately realised. Implicit in this conclusion and explanation is the assumption that it is valid to compare calculated critical energies based upon equivalent AH; values for 3(3MB) and 3(MBD13).As discussed above it is quite possible that the method used to calculate these quantities may result in their being underestimated, especially in the case of the triplet olefin. Thus if AHt[3(3MB)] is greater than AHf"[3(MBD13)], the average total internal excitation energy, E*, of 3(3MB) would be less than that of 3(MBD13)* by an equivalent amount, and this could result in a correspondingly lower value for k3. Such an effect could partially explain the difference in k3 and k I 8 , and moreover would lead to a lower computed critical energy for reaction (3). The rate constant discrepancy is, however, too large to be entirely due to an effect of this type, even allowing for error in the relative A-factors for the two reactions. The critical energies for these two processes cannot therefore be the same. The computed critical energy for 3(3MB)* isomerization, reaction (14), lies in the range 82-93 kJ mol-l, the exact calculated value depending on the chosen A-factor and excitation energy.Direct comparison with the critical energies calculated for similar lY4-hydrogen shift processes involved in the isomerization of various alkyl and alkenyl radicals 24-26 is difficult, due to differences in the chosen activated complex models and reaction thermochemistries, though it would seem that they are of the same magnitude. 1,3 and 1,2-hydrogen migrations proceeding by 4sp, 4pv (v = vinyl), 3ss and 3sp processes, are reported to have critical energies of z 130 and 140 kJ mol-1 respect- ively.lgT 26-28 Reaction (16) is a 4ps process, and woufd be expected to be less thermochemically favoured.It is therefore not surprising that products such as methylcyclobutane and propylene, derived from the 2-methylbuta- 1,4-diyl biradical formed by this isomerization, were not detected. The experimental data suggests, however, that despite the low reaction path degeneracy, part of the 2MB2 yield arises via reaction (I 5), a 4tp hydrogen shift, presumably as a result of the exothermic nature of the process. A simple calculation using RRK theory shows that the critical energy for this reaction can be no more than M 115 kJ mol-l. Reaction (17), a potential route to the 3-methylbuta-l,3-diyl biradical, the precursor of 1,l-dimethyl- cyclopropane, is a 3ts hydrogen migration and as such would be expected to make virtually no contribution to the reaction mechanism if its energy of activation were similar to those for 3ss and 3sp processes, even if some allowance is made for its possibly greater exothermicity.A striking feature of all the critical energies calculated here is their remarkable similarity to those computed for the analogous reactions occurring in the Hg(63P1)274 3-METHYLBUT-1-ENE PHOTOSENSITIZATION photosensitization of 33DMB, the A-factors for each reaction pair having been estimated in a mutually consistent manner. Results are also available for the Hg(63P,) +2,3-dirnethylbut-l-ene system and it is intended to give a more detailed discussion of the similarities of these various olefin systems in the report of that study.Information concerning the biradical internal energy dependence of the ratio of trans- to cis-DMCP and the effect of oxygen on the reactions of 3(3MB):k and 3(MBD13)* is given in the following paper.13 APPENDIX RRKM C A L CUL AT1 0 NS (3 MB) * D E c o M P o s IT I o N The frequency assignment for 3MB is that deduced by Taylor and Simons 29 except for the inclusion of realistic energy barriers for the internal rotations about the three C-C single bonds. Activated complex frequencies were assigned by considering the bonding changes as the C(3)-C(4) bond is stretched by a factor of two. This bond extension results in Q,'/Ql = 1.69. The reaction coordinate for both this and the other two decompositions described below was taken to be the stretching frequency of the extended bond.Trenwith has revised his published 30 Arrhenius expression for the unimolecular decomposition of 3MB, now recommending 31 loglo (k/s-l) = 16.6kO.3-(35750 K/2.303 2"). Three complex models, I, I1 and 111, were proposed, corresponding to loglo (A/&) = 16.3, 16.6 and 16.9 (calculated at 712 K). The statistical parameter L* for this reaction is 2. Frequencies (wavenumber/cm-l) common to both reactant and complex were as folIows (degeneracies shown in parenthesis) : 3093 3012 2992 2968(3) 2946 2903(2) 2885 1468(2) 1451(2) 6420 1373(2) 1314 1295 1278 1165 955 913 748 575 418 375(2) additional reactant frequencies : 1650 1152 1000 904 850 790 418 250 211(2) 13 additional complex frequencies : complex I : 1230(2) 550 250 155 150 117 100 70 50 complex I1 : 1230(2) 550 250 155 150 100 70 58 50 complexIII: 1230(2) 550 250 155 150 100 70 50 29.In a few calculations the complex frequencies were modified, but in such a way that the corresponding loglo ( A / s - l ) values were unaltered. The results obtained were very similar to those computed using the listed assignments. 3(3MB)* D E c OM P o SI TI ON Vibrational frequencies for 3(3MB) were obtained by removing two C-H stretching, and four CH3 and CH, deformation modes from the frequency assignment for isopentane given by Snyder and Scha~htschneider.~' In addition the torsional frequencies associated with the C(2)-C(3) and C(3)-C(4) bonds were adjusted. With I,* = 2 and by assuming Q,' /Ql = 1.69, rate constants were calculated using three postulated activated complex models, w, v and VI, having vibrational frequency assignments corresponding to values for 1oglo(A/s-l) of 15.0, 15.5 and 16.0 respectively (calculated at 450 K), estimated to be intermediate between those for fragmentation of 3MB and the 3-methylbut-2-yl radical. Common reactant and complex frequencies :D.C. MONTAGUE 275 2962(3) 2952 2938 2926 2873(3) 2852 1384 1377 1366 1351 1337 1298 101 1 952 . 910 796 412 368 additional reactant frequencies : additional complex frequencies : 1268 1037 969 917 764 459 260 204 coniplex IV: 1220(2) 300 250 200(2) 155 complexV: 1220(2) 300 250 200(2) 155 complexVI: 1220(2) 250 195 155 1 SO(2) 1475 1455(3) 1176 1149 198 140 15 100 79 75 100 75 25 100 75 21. 3( MBD 13)* D E c o M P o s I TI o N Vibrational frequencies for 3(MBD13) and the three chosen activated complex models, VII, VIII and IX, corresponding to log,, (A/s-l) = 14.6, 15.1 and 15.6 respectively, were assumed to be identical to those assigned to 3(3MB) and complexes IV, V and VI, with the exception of one reactant frequency at 204 cm-' which was lowered to 150 cm-,.For this reaction E* = 1 and Q,'/Ql = 1.69. (3 MB)* IS o M E R I z AT I o N The vibrational frequency assignment of 3(3MB) is given above. Two five-membered ring activated complex models, X and XI were postulated, corresponding to log,o(A/s-') = 11.2 and 11.5, calculated at 450 K, their frequencies being estimated by analogy with those for the proposed ' complex models involved in the isomerization of 3(33DMB)*. The reaction coordinate was taken to be one of the ring deformation modes involving motion of the H atom.I,* = 6 and Q,'/Ql was assumed to be unity. Complex X frequencies : 2960(3) 2920(2) 2880(2) 2860(2) 1470(2) 1450(3) 1255(2) 1215 1150(3) llOO(2) 1040(2) 1020 lOOO(4) 930 895(2) 870 550 400 360 290 198 The two ring deformation modes at 870 and 290 cm-1 were reduced to 550 and 200 cm-I to obtain the frequency assignment for complex XI. The author is grateful both to Prof. H. M. Frey at the University of Reading, where much of the experimental work was carried out, and to Prof. P. Gray, for the provision of laboratory facilities. Acknowledgement is also due to P. E. Montague for assistance with computing. D. C. Montague, J.C.S. Faraday I, 1975,71,398. R. J. CvetanoviC, H. E. Gunning and E. W. R. Steacie, J. Chem. Phys., 1959, 31, 573. R. J. CvetanoviC, Prop. Reaction Kinetics, 1964, 2, 39. H. E. O'Neal and S. W. Benson, J. Phys. Chem., 1968,72,1866. J. Metcaife and E. K. C. Lee, J. Amer. Chem. SOC., 1972, 94, 7. R. Moore, A. Mishra and R. J. Crawford, Canad. J. Chem., 1968, 46, 3305; E. B. Klunder and R. W. Carr, Chem. Comm., 1971,742. ' F. J. Duncan and R. J. CvetanoviC, J. Amer. Chem. SOC., 1962, 84, 3593; F. S. Rowland, P. S.-T. Lee, D. C. Montague and R. L. Russell, Disc. Faradny SOC., 1972, 53, 111. * D. C. Montague, Int. J. Chem. Kinetics, 1973, 5, 513. M. Tennonia and G. R. De Mare, Chem. Phys. Letters, 1974, 25, 402. S. D. Gleditsch and J. V. Michael, J. Phys. Chem., 1975, 79,409. l 1 B. de B. Darwent, M. K. Phibbs and F. B. Hurtubise, J. Chem. Phys., 1954, 22,859. l2 P. Kebarle and M. Avrahami, J. Chem. Phys., 1963,38,700. l 3 D. C. Montague, J.C.S. Faraday I, 1978, 74,277. l4 R. J. CvetanoviC and L. C. Doyle, J. Chem. Phys., 1962,37, 543. l 5 E. A. Hardwidge, C. W. Larson and B. S. Rabinovitch, J. Amer. Chem. Soc., 1970, 92, 3278. l6 D. C. Montague and F. S. Rowland, J.C.S. Chem. Comm., 1972, 193.276 3-METHYLBUT-1-ENE PHOTOSENSITIZATION l7 L. Salem and C. Rowland, Angew. Chem. Int. Edn, 1972,11,92. I'D. C. Tardy, Int. J. Chem. Kinetics, 1974,4,291. 2o M. C. Flowers and H. M. Frey, Proc. Roy. SOC. A, 1961, 260,424. 21 G. B. Kistiakowsky, M. A. Dolliver, T. L. Gresham and W. E. Vaughan, J. Amer. Chem. SOC., 22 American Petroleum Institute Research Project 44 Tables (Thermodynamic Research Center, 23 G. Z . Whitten and B. S . Rabinovitch, J. Chem. Phys., 1963, 38,2466. 24 K. W. Watkins and D. R. Lawson, J. Phys. Chem., 1971,75,1632. 2 5 K. W. Watkins and L. A. O'Deen, J. Phys. Chem., 1971,75,2665. 26 W. P. L. Carter and D. C . Tardy, J. Phys. Chem., 1974,78,2201. 27 T. Ibuki, A. Tsuji and Y. Takezaki, J. Phys. Chem., 1976, 80, 8. 28 T. Ibuki, T. Murata and Y . Takezaki, J. Phys. Chem., 1974,78,2543. 29 G. W. Taylor and J. W. Simons, Int. J. Chem. Kinetics, 1971, 3, 453. 30 A. B. Trenwith, Trans. Furaday SOC., 1970, 66,2805. 31 A. B. Trenwith, Third International Symposium on Gus Kinetics (Brussels, 1973). 32 R. G. Snyder and J. H. Schachtschneider, Spectrochim. Acta, 1965, 21, 169. W. von E. Doering and K. Sachdev, J. Amer. Chem. SOC., 1974,96,1168. 1937, 59, 831. Texas A and M University, 1972), vol. V. (PAPER 7/247)

ISSN:0300-9599    

DOI:10.1039/F19787400262

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Hg(63P,) Photosensitization of 3-Methylbut- 1 -ene Part 2.-Intersystem Crossing and Cyclisation of the 2-Methylbuta- 1,3-diyl Biradical BY DEREK C. MONTAGUE Department of Physical Chemistry, The University, Leeds LS2 9JT Receiued 14th February, 1977 Hg((j3P1) photosensitization of 3-methylbut-1-ene (3MB) yields, among other products, trans- and cis-1,2-dimethylcyclopropane, in a strongly pressure dependent ratio, ranging from 1.37 at 2.5 TOIT to 3.44 at 789.2 Tom. This observation is interpreted in terms of a mechanism involving inter- system crossing and ring closure of both vibrationally excited and thermalised triplet 2-methylbuta- 1,3-diyl (MBD13) biradicals, produced by a 5pp hydrogen migration in excited triplet 3MB. Quanti- tative analysis of the data allows a value for the intersystem crossing rate constant for 3(MBD13)* to be obtained, together with the ratio of dimethylcyclopropanes formed from thermalised MBD13. Addition of oxygen to the reaction system considerably simplifies the observed product spectrum and enables lower limit estimates of the rate constants for intersystem crossing and cyclisation of thermalised triplet and singlet MBD13 respectively, to be derived.The results are compared to those obtained in other systems, and their relevance to the implications of thermochemical and quan- tum chemical analyses of the energetics of propa-l,3-diyl biradicals discussed. In the preceding paper the yields of the products formed by Hg(63P1) photo- sensitization of 3-methylbut-1 -ene (3MB) have been reported. The experimental data were interpreted in terms of reactions of 3-methylbuta-l,2-diyl and 2-methylbuta- 1,3-diyl (MBD 13) biradicals, both initially produced with excess vibrational energy in the triplet state.Application of a stationary-state computer modelling treatment to the postulated mechanism enabled certain isomerization and decomposition rate constants to be derived, and these were in turn used in conjunction with RRKM theory to estimate critical energies for three of the reactions involved. The possible importance of biradicals typified by MBD13 has been recognised for some time. MBD13 itself has been invoked as an intermediate in the structural isomerization of 1,2-dimethylcyclopropane 2 $ (DMCP), and as a precursor to this compound in the photochemical decompositions of both 2,3-dimethylcyclobutanone and 3,4-dimethyl-A1-pyrazoline,5 9 and in the reaction of triplet methylene with b~t-2-ene.~ In this paper attention is focused on the rates of the intersystem crossing and cyclisation processes involved in DMCP formation, both from vibrationally excited and thermalised triplet MBD13.The influence of excitation energy on the stereochemistry of this latter reaction has been recognised, but as briefly noted previously,8 the results now presented demonstrate that the magnitude of this effect is much greater than is generally appreciated. Additional insight into the reaction mechanism has been gained by carrying out photolyses in the presence of molecular oxygen. EXPERIMENTAL Experimental procedure, and the photolysis and analytical apparatus were the same as those described in Part 1.' Oxygen, used as a scavenger in some photolyses, was obtained from the British Oxygen Company, and was used without further purification.277278 3-MET H Y LB U T-1 -ENE PHOTOS ENS IT I Z A TI 0 N RESULTS Both trans- and cis-DMCP are major products in the H S ( ~ ~ P ~ ) photosensitization of 3MB, at all pressures within the range 2.5-790Torr.* The mercury photo- sensitization quantum yields of all the products have been listed in Part 1, over a less extensive pressure range (11.8-764 Torr). Values for the ratio of tDMCP to cDMCP, R, at different experimental pressures are shown in table I. R increases with increasing pressure and appears to approach a limiting value as the pressure tends to infinity.It is however, independent of photolysis light intensity. TABLE VA VARIATION OF THE RATIO tDMCP/cDMCP WITH PRESSURE pressure/Torr pressure/Torr pressure/Torr pressure/Torr pressurelTorr pressure/Torr R R R R R R 2.5 1.37 40.0 1.74 134.9 2.21 254.5 2.64 463.4 3.13 599.0 3.37 5.0 1.39 48.9 1.72 140.8 2.19 274.6 2.77 497.9 3.17 669.8 3.33 10.0 80.8 177.0 305.2 513.5 1.49 1.96 2.38 2.83 3.20 691.7 3.40 aR = tDMCP/( 11.8 86.9 187.3 350.2 536.5 764.4 1.58 2.00 2.43 2.90 3.28 3.48 :DMCP 20.0 1.56 101.1 2.06 2.54 3.03 3.27 3.44 221.2 374.1 542.0 789.2 24.7 127.8 228.8 407.6 548.8 1.65 2.19 2.55 3.03 3.15 37.5 131.5 251.6 451.2 577.5 1.59 2.21 2.66 3.21 3.32 In a second series of experiments various amounts of oxygen were added to several different pressures of 3MB before photolysis, with the result that the yields of all products formed from monoradical precursors in the unscavenged system, e.g.pent-2-ene, were reduced to trace levels, whereas those of DMCP and 2-methylbut- 2-ene (2MB2) were only partially decreased. In addition the spectrum of C2-C4 products, of which buta-1 ,3-diene7 but-2-ene7 but-I-ene and propylene could be positively identified, changed slightly. The small yields of these compounds prevented TABLE 2.-vARIATION OF THE RATIO tDMCP/cDMCP WITH 3MB AND OXYGEN PRESSURES a 3MB/Torr 0 2/Torr R' 204.9 200.5 299.5 397.0 398.2 402.0 398.1 398.2 402.2 500.0 501.4 500.0 498.8 26.8 54.9 75.5 41.2 98.6 113.9 121.7 137.6 151.9 30.5 60.9 86.4 170.4 2.20 2.04 2.23 2.61 2.32 2.25 2.15 2.21 2.14 2.98 2.65 2.57 2.23 a See text for explanation of symbols.* 1 Torr = 133.3 Pa.D . C . MONTAGUE 279 their quantitative assessment with any confidence, however. Minor amounts of 2-methylbut-1-ene (2MB1) and isoprene were also observed. Other higher molecular weight products were undoubtedly formed but no attempt was made either to detect or to identify them. The only data amenable to a meaningful quantitative interpreta- tion are those of the variation of the DMCP geometrical isomer ratio, with changes in the oxygen concentration and overall pressure. Experimental values of this ratio, symbolised as R’ to distinguish it from the ratio of the same products, R, in the absence of oxygen, are given in table 2. They show that for an approximately constant 3MB pressure, increasing the pressure of O2 decreases R’.DISCUSSION Vi brationally excited triplet acyclic mono-olefins produced by Hg(63P1) photo- sensitization can fragment, unless collisionally deactivated, producing mono-radicals, or isomerize by hydrogen atom migration processes proceeding by cyclic transition ~ t a t e s . ~ For 3MB, 5pp isomerization gives vibrationally excited triplet MBD 13, which can subsequently either decompose, undergo spin inversion leading to DMCP by ring closure, or, less importantly, isomerise by a 1 ,Zhydrogen translocation to 2MB 1 and, presumably, 2MB2 in approximately equivalent yields. OXYGEN FREE EXPERIMENTS The extensive variation of R with pressure can be rationalised by postulating first that the rate of intersystem crossing of excited triplet MBD13 to the ground singlet state is sufficiently slow, so that collisional deactivation often occurs prior to PRODUCTS A I k,5! 0, I k, 6 PRODUCTS A I I k , , 10: I 1‘41 * L 1 1 L k M 7 Pf w) L \ 4 I I [\a/] * k, M PRODUCTS FIG.1 .-Reaction scheme for 3(MBD13)*.280 3-MET HY L B U T-1 -E NE P H 0 T 0 SEN S I TI Z A T I0 N spin inversion, and secondly that biradical ring closure becomes more stereospecific as the biradical excitation decreases. Both deactivation and fragmentation of the excited singlet biradical are ruled out at the pressures of these experiments by its rapid rate of cyclisation. * Geometrical isomerization of newly formed excited DMCP serves only to modify the observed values of R, especially at low pressures; it alone cannot provide an explanation for the data, as the rates are too low (vide infra). The fate of excited triplet MBD13 can be discussed in terms of reactions (1)-(lo), schematically depicted in fig.1. '(MBD13)" and l(MBD13) are enclosed in square brackets to indicate that they may not possess an intrinsic stability. Routes to 2MB1 and 2MB2 from 3(MBD13)* and 3(MBD13) have been omitted as they are negligibly s1ow.l If the usual assumptions of the stationary state hypothesis are applied to the concentrations of '(MBD13)*, 3(MBD13), '(MBD13), (tDMCP)* and (cDMCP)*, then a kinetic analysis of the reaction mechanism enables the expression for R shown in eqn (1 1) to be derived. p(1 +a + cm[M] j + pP R = @(I+ a + p[ MI) + P where and a = k4t/k40 = kst/ksC, 6 = kl/kz, p = k9/k4, $ = k10/k5, p = k7/k--6 and o = ks/k6.It would seem likely that eqn (I 1) could only be fitted to the experimental data to yield meaningful information if values for at least four of the seven rate con- stant ratios are known. The inverse half-quenching pressures for geometrical iso- merization of the excited trans- and cis-DMCP formed in reaction (4), p and co, can be estimated by RRKM theory. The calculation was carried out in the same way as those performed previously, 3 using frequencies for the reactants and activated complexes based on those listed by Simons and Rabinovitch,ll and chosen to be commensurate with the Arrhenius A-factors for these rearrangernents.l Similarly critical energies were derived, in the usual way,13 from the measured activation energies. The computed rate constants were converted to the inverse half-quenching pressures p = 0.0110 Torr-l and co = 0.0824 Torr-I, using a calculated collision rate constant of 1.357 x s-l Torr-l (2.517 x 1014 cm3 mol-1 s-l).The magnitudes of p and $ reflect the relative rates of isomerization against cyclisation for excited and thermalised singlet MBD 13 respectively. A comparison of the small yields of 2MB1 with those for DMCP over the experimental pressure range, indicates that p and $ must both be of the order of 0.05. If they are equal then their magnitude is irrelevant, as the term (1 +p)/(l +$) in eqn (12) becomes unity. Their values can be estimated by fitting the variation with pressure of Q, the ratio of the sum of the quantum yields for the isomerization products 2MB1 and 2MB2 to that of DMCP, to eqn (13), which can also be derived from a kinetic analysis of the proposed mechanism.The data used were those tabulated in Part 1 and, as before, it was assumed that reactions (9) and (10) produce eqclivalent amounts of 2MB1 and 2MB2. The same computer modelling approach to that previously employed was used once again.l* lo It was found that good fits could beD . C . MONTAGUE 28 1 found for any value of 6 in the range 30-170 Torr. Thus if 6 = 33.6 Torr as found in Part 1, p = 0.0440 and $ = 0.0514 ; for 6 = 150 Torr, p = 0.0427 and II/ = 0.0554. Satisfactory fits could not be obtained however when either p or $ was set equal to zero. Biradical isomerization must therefore occur, albeit to a minor extent, both from l(MBD13) and l(MBD13)*. As the experimental data are scattered, detailed conclusions should not be drawn from the exact relative magnitude of p and $.Suffice it to say that the near equivalence of these parameters indicates that the energy barriers for isomerization and ring closure of MBD13 must be similar, the difference in the rates of the two reactions being largely brought about by a difference in pre- exponential factors. I I I I J 200 400 600 pressure/Torr FIG. 2.-Variation of Q D ~ P / C D ~ D M C P with pressure. Using the values obtained for p, coy p and $, eqn (11) can be fitted to the experi- mentally observed variation of R with pressure. As p and $ are not only both small but almost equal, their effect is minimal. It was found that the best fits could be obtained by also treating p as a variable parameter.The optimum value derived, 0.115 Torr-l, is within 4% of that calculated by RRKM theory. The optimum values for the other variables are listed in table 3. Fig. 2 shows the best-fit curve computed from these parameters. The effects of neglecting biradical isomerization and interconversion of tDMCP* and cDMCP* were investigated in turn, by equating first p and $, and then, in addition, w and p to zero. Almost equally good fits to282 3-METHY LB UT- 1 -ENE P HOT 0s ENS I TI Z A TI 0 N the data were found, generating the parameter values also listed in table 3 . Clearly a reaction mechanism of the level of sophistication proposed is not strictly necessary. The isomerization reactions (6), (- 6), (9) and (10) have nevertheless been included for the sake of completeness, thereby enabling their minor role to be illustrated. TABLE 3.-oPTIMUM RATE CONSTANT RATlQS AND HALF QUENCHING PRESSURES ’ a B G/Torr olTorr-1 p /Torr-1 P Y 1.368 4.435 150.0 0.0824 0.115 0.0427 0.0554 C 1.368 4.435 151.9 0.0824 0.115 O d 0 “ 1.386 4.451 156.0 0 ” 0 ” O d 0 “ See text for explanation of symbols ; b calculated value, see Text ; C fixed value, see text ; d selected value.The value obtained for 6 is a factor of w 4.5 larger than that derived in Part 1 from an analysis of the overall product yields (33.6 Torr, giving k l = 4.56 x lo8 s-l). Attempts to iit the data here treated using the lower value were entirely unsuccessful, as, it will be recalled, were efforts to explain the quantum yield data discussed in Part 1 using a value of 6 in excess of w 50 Torr.This apparent discrepancy can be resolved, however, if the strong collision assumption, invoked in both kinetic treat- ments of the reaction mechanism, does not necessarily apply, and If the energy barrier for triplet MBD13 intersystem crossing is considerably lower than that for its decomposition by reaction (14) The results from the experiments in which oxygen was added to the pliotslysis mixtures demonstrate that this latter postulate must be correct. They show that the minimum intersystem crossing rate constant for thermalised 3(MBD13) is 3.05 x lo6 s-I, approximately two orders of magnitude lower than that for 3(MBD13)‘k, whereas the rate constant for thermalised 3(MBD1 3) fragmentation, calculated from the computed Arrhenius parameters,1 is at most s-l, some fourteen orders of magnitude lower than its value for 3(MBD13)*.These rate constants illustrate that the rate of intersystem crossing is much less sensitive to the level of internal excitation than is that for decomposition. Thus, while a single deactivating collision of 3(MBD13)* with a bath gas molecule might well reduce the excitation level SO as effectively to prevent fragmentation, the rate of intersystem crossing will only be decreased slightly. The proposed reaction scheme implies that the ratio in which the products tDMCP and cDMCP are formed is dependent on the biradical excitation energy. This means that partially deactivated 3(MBD1 3)* will give these products in a ratio intermediate between a and p.Thus the true mechanism has been simplified by assuming that all triplet biradicals that have undergone a deactivating collision give tDMCP and cDMCP in the same ratio p, a condition that in fact only applies to thermalised biradicals formed by multi-step deactivation. It is a consequence of this assumption that the intersystem crossing rate constant for excited (3MBD 13)’:’, derived using eqn (1 l), is too large, its true value, (4.56 x lo8 s-l), being that measured relative to the decomposition rate constant in Part 1. Several conformations of roughly equivalent stability can be envisaged for triplet MBD13. Twisted configurations rather than the preferred (0,O) orientation of the unsubstituted propa-1,3-diyl biradical l4 are probably favoured as a result of the steric interaction introduced by the CH3 substituents.Even if a structure close to the (0,O) form is adopted, two rotational isomers are possible. The population distribution for these various configurations depends on their relative stabilities and 3(MBD13)* 3 CH3 +CK,CH CH *= CH2. (14)D . C . MONTAGUE 283 the internal energy of the biradical. Steric effects will also influence the course of the cyclisation reaction and the consequent relative proportions of trans- and cis- DMCP formed from each of the conformers present. At high levels of excitation these factors will be less important and it is therefore not surprising that less stereo- specifcity is observed under low pressure conditions where ring closure from non- deactivated biradicals predominates.a represents the relative yields of trans- to cis-DMCP" at the zero pressure limit. Its value is clearly energy dependent and may not be equal to the observed ratio of these products due to their subsequent geometrical isomerization. At the other pressure extreme, p is independent of the initial excitation energy, its value being determined by the relative overall cyclisation rates of those biradical configurations in thermal equilibrium. Thus if 3(MBD1 3)" could be produced in this system with differing initial internal energies, plots of tDMCP/cDMCP against pressure should form a family of curves, all of which approach the same /? value at infinite pressure. Moreover as a and 6 increase and decrease respectively with decreasing internal energy, the curves should not cross.Data from other experimental systems should also fit this hypothesis as long as the bath gases used are equally proficient at deactivating 3(MBD13)*. Results from the reaction of CH2(3B1) with both trans- and cis-but-2-ene do indeed exhibit the predicted b e h a v i o ~ r , ~ ~ but those from the Hg(63P1) photosensitizations of 2,3-dimethyl- cyci~butanone,~ 3,4-dimethyl-Al-pyrazoline and 3-methylpent-4-enal appear at first sight not to conform, in that the plots of tDMCP/cDMCP against pressure intersect that shown in fig. 2. This apparent discrepancy can be traced however to the fact that argon, a relatively inefficient deactivating bath gas, was used in these latter systems. If the experimental total pressures are modified using an argon collisional deactivation eficiency (&) of < 0.23 for the ketone and aldehyde experi- ments, and < 0.14 for the pyrazoline experiments, then the variations of the observed tIIMCP/cDMCP ratios with eflective pressure agree with the pattern suggested by the results from the Hg(63P1)/3MB and methylene systems.The values postulated for IJp are not unreasonable, as pp has been measured as 0.16 for the collisional de- activation of vibrationally excited ground state 2,3-dirnethyl~yclobutanone.~~ An alternative less likely explanation for the different behaviour observed in the ketone and aldehyde (or pyrazoline) systems is that the acyl-alkyl (or diazenyl-alkyl) bjradical initially formed from the triplet precursor molecules could participate directly in the formation of DMCP by a concerted CO (or N2) displacement reaction.Even if such a reaction were possible it would probably not compete effectively with direct loss of CO from the biradical. Estimates of the rate constant for this frag- mentation, computed by RRK theory using thermochemistry calculated from the reported enthalpy of cyclobutanone,18 standard group additivity procedures and Arrhenius parameters assumed equivalent to those for propionyl decomposition,20 show that the rate is very rapid at the excitation levels produced by Hg(63P1) sensit- ization. However at the: lower excitation level resulting from triplet benzene sensit- ization (3Blu, 353 kJmol-l), the calculated rate constant for CO loss from the biradical derived from trans-2,3-dimethylcyclobutanone is only 2 x 1 O6 s-l, indicating that decomposition could be effectively quenched at even moderate pressures.The RRK computations discussed here apply to the ground state acyl-alkyl biradical, in which the acyl group is bent. It is possible, however, that photosensitized de- carbonylation may proceed via an excited state biradical with a linear acyl group, if the theoretical postulates of Turro, Farneth and Devaquet are correct.21 If this is so then decarbonylation would be fast at all levels of excitation due to predissociation. It is unfortunate that product quantum yields were not measured for triplet benzene sensitization of the 2,3-dimethyl~yclobutanones,~~ as they might well have provided284 3-M E T 13 Y L B UT- 1 -EN E P HO T 0 SEN SI T I Z A T I 0 N information on the rate of acyl-alkyl biradical fragmentation, thereby enabling the relative importance as intermediates of the linear and bent forms of the biradical to be determined, OXYGEN SCAVENGED EXPERIMENTS The addition of oxygen to the photosensitization system virtually suppresses the formation of products such as pent-2-ene completely.This observation demonstrates not only that monoradicals are effectively scavenged in these experiments, but that intramolecular methyl migration reactions, analogous to those thought to be present in some other olefin direct photolysis are absent. The slightly reduced DMCP yields in the presence of oxygen presumably result from some O2 quenching of Hg (63P1) and partial removal of the triplet olefin and MBD13 biradical by scaveng- ing reactions, leading to unknown products.In the discussion that follows it is assumed that DMCP is neither formed nor removed by either of these latter processes, or by reactions of excited molecular oxygen produced by 0, + Wg(6’Pl) interaction. The results can be interpreted in terms of reactions (1)-(lo), together with the addi- tional reactions, (15)-(18), also shown in fig. 1. Scavenging of l(MED13)* need not be considered under the conditions of these experiments as it is removed extremely rapidly by the first order reactions (4) and (9).8 If a kinetic analysis is applied to the mechanism then eqn (19) can be derived by assuming that stationary state conditions are achieved. where K = k17/k3, A = kls/k,, 0 = k16/k2, and [MI is the effective pressure for collisional deactivation of (tDMCP)* and (cDMCP)*.Eqn (19) expresses the relationship between the experimentally observed ratio of trans- to cis-DMCP, R’, and the pressures of 3MB1 and O2 in terms of ten rate constant ratios. Of these all but K , 3, and Q have been determined previously (cf. table 3), and the latter can be estimated from data in the literature. If values of the pseudo-pressure Y are calculated, the variation of the magnitude of the 1.h.s. of eqn (19) with oxygen pressure can in principle be computer fitted, thereby generating K and A values. As strong collisions are again assumed to be operative for 3(MBD13)*, the appropriate value of 6 for substitution in eqn (20) is 150.0 Torr. 0 is estimated to be < 0.23 by analogy with other molecules and A lower limit of c = 0 is possible if collisions between 0, and 3(MBD13)* result solely in product formation.However, M was calculated assuming that the collisional efficiency of 0, relative to 3MBl for deactivation of (tDMCP)* and (cDMCP)*, was 0.23. Values of Y calculated from the data shown in table 2 were found to range from 105.5 to 245.3 Torr. Computer fitting eqn (19) to these values by the same iterative random-exploration least-squares method, using CF = 0.23, gave optimum K and A values of 6.64 x and 1.75 x low2 Torr-l respectively. Fig. 3 shows the data plotted according to eqn (19) together with the computer generated best-fit curve. If kl, and k18 have the same approxi- mate value as that suggested 25 as typical for the analogous reaction of alkyl mono- radicals, 1.0 x 10l2 cm3 mol-l s-1, then k3 and k, can be calculated from these optimum K and 3, values as 3.08 x lo6 and 8.13 x lo8 s-l respectively.It was found, however, that almost equally good fits could be obtained with either A or IC set equalD . C. MONTAGUE 285 to zero. Under these conditions maximum values of each concomitant parameter were obtained, viz. for A = 0 then K,,, = 1.77 x Torr-l, and for K = 0, then A,,, = 1.86 x Torr-l. Fixing either parameter at any intermediate value both determined that of the other and resulted in an alternative, but less acceptable fit to the data. The poorest fit, at the limits of acceptability, was found when K = 6.13 x Torr-l. The data therefore suggest that either K $- A or A @ K , and not K N A.Torr-I and A = 6.70 x 1 Y o I I 1 I I oxygen pressure/Torr FIG. 3.-Variation of ([3MB1] +0.23[Q2])/ Y with oxygen pressure. 0 50 100 150 If A is indeed close to zero, then cyclisation of thermalised l(MBD13) must be much faster than its removal by oxygen scavenging, and the corresponding K value, K , , ~ defines the minimum rate constant for ,(MBD13) intersystem crossing k,. Substituting k17 = 1.0 x 1OI2 cm3 mol-1 s-l as before gives k,(min) = 3.05 x lo6 s-l, very close to the optimum value. It is nevertheless only a factor of 2.9 lower than that corresponding to the least acceptable fit. Amax on the other hand allows the minimum ring closure rate constant of l(MBD13), k5, to be computed, and xmin the maximum value of k3. It seems unlikely that k,(max) could exceed kl, however, and therefore, assuming k3(max) = k , = 4.56 x lo8 s-I leads via Kmin = 1.18 x Torr-l and a revised A,,, = 1 .8 0 ~ Torr-l, to k,(min) = 3.00 x lo6 s-l. The errors in k,(min) and k,(min) are largely determined by those associated with kl, and k18 respectively. Thus, whereas k5(min) is largely unaffected by a decrease in k17, a reduction in k18 will lower k,(min) by a similar factor, and vice versa for k (min) . A similar set of K and A values, generated by assuming that o = 0, allows the minimum values 3.38 x lo6 and 3.21 x lo6 s-l to be computed for k, and k5 respec- tively, a change of only 10 % in these rate constants. Neglecting the effects of singlet biradical isomerization by setting t,b and p equal to zero produces even smaller changes.In summary, the results almost certainly demonstrate that 3.05 x lo6 -c k3 c 4.56 x lo8 s-l and k5 2 3.00 x lo6 s-l. The short lifetime of thermalised MBD13 biradicals ensures that their reaction with substrates other than those endowed with a par-286 3-METH Y L B U T- 1 -EN E P €3 0 TOS E N SI TI Z AT I 0 N ticularly high reactivity towards free radical species (with, say k 2 lo9 cm3 mol-I s-l), would not occur at 293 K for normal substrate pressures. Addition to the n-bond of a simple olefin to form a cyclopentane is therefore not feasible. Arrhenius parameters for cyclisation of singlet propa-l,3-diyl biradicals have been estimated by O’Neal and Benson from a thermochemical analysis of pyrolysis data for various substituted cycl~propanes.~ They deduce an activation energy of 38.9 kJ inol-l, and log (Als-l) = 13.6, calculated at 700 K.Several theoretical studies dispute these estimates however, suggesting instead that the ring closure activation energy lies in the range 0-10 kJ mol-? A lucid account of the implications of these conflicting proposals has been presented by Stephenson, Gibson and B r a ~ r n a n . ~ ~ The data obtained here can be used to support either postulate. Thus while the experimental k5(min) value is close to that calculated from O’Neal and Beplson’s suggested parameters, uiz., 5.8 x lo6 s-l, the higher, equally acceptable k5 values are compatible with lower cyclisation activation energies. The latter situation is perhaps to be favoured however, in that as intersystem crossing rate constants invariably show an energy dependence, it would be surprising if kl were not > k3, especially as the energy difference of ’(MBDl3)* and 3(MBD13) is some 186-226 kJ mol-l.This assumption implies that the optimum value for k3 is k3(min) (3.05 x lo6 s-l) rather than k3(max) (= kl), intermediate values having been shown to give less acceptable fits to eqn (19); the argument concludes that k5 approaches k,(max). This con- clusion does not allow the ring closure activation energy to be defined precisely however, as any value in the range 0-35 kJ mol-I will satisfy its proposals assuming the postulated A factor to be correct. The values of kl and k3 allow the “ Arrhenius parameters ” for 3(PJIBD13) inter- system crossing to be roughly estimated using RRM theory.Input data required for this calculation are the biradical excitation energy, initially taken to be 226 kJ mol-1 and the number of active vibrational modes, s. A value of 19 was chosen for s as this enabled good agreement to be obtained between the RRK and RRKM calculated rate constant for 3(MBD13)* deconiposition.l Use of the minimum value for k3 leads to upper limits for the activation energy, EISC, and A-factor. Thus by sub- stituting k3 = 3.05 x lo6 s-l, the intersystem crossing rate constant, kIsc, can be expressed as kIsc = 1.64 x lo9 exp (- 1870/T) s-l. Decreasing k17 by an order of magnitude reduces k3 by a similar factor and gives kIsc = 3.10 x lo9 exp (-2740/T) s-l . Uncertainty in the heat of formation of MBD 13 results in a minimum biradical excitation energy of approximately (1 86 + EIsc) kJ mol-l.A lower value would imply that the heat of formation of I(MBD13) would probably have to be larger than the maximum value dictated by the activation energy for DMCP structural isomer- ization. Using this minimum excitation energy leads to kIsc = 2.01 x lo9 exp (- 1930/T) s-I if k l , = 1 x 10l2 cm3 mol-1 s-l. These calculations show that the uncertainty in the upper limit of the intersystem crossing activation energy is largely brought about by error limits associated with the value of k l , rather than with the excitation energy. Theoretical studies have suggested that the ground state of MBD13 is a triplet.26* 27 It would therefore be anticipated that triplet-singlet intersystem crossing would involve a small positive activation energy.The maximum values postulated here are consistent with this expectation and go some way towards assessing the magnitude of the difference in lowest triplet and singlet state enthalpies. D. C. Montague, J.C.S. Faraday I, 1978, 74,262. M. C. Flowers and H. M. Frey, Proc. Roy. Soc. A, 1961,260,424. H. E. O’Neal and S. W. Benson, J. Phys. Chem., 1968, 72,1866. J. Metcalfe and E. K. C. Lee, J. Arner. Chern. Soc., 1972, 94, 7.D . C . MONTAGUE 287 R. Moore, A. Mishra and R. J. Crawford, Cunud. J. Chem., 1968, 46, 3305 ; E. B. Klunder and R. W. Carr, Chem. Comm., 1971, 742. E. B. Klunder and R. W. Carr, J. Amer. Chem. Soc., 1973, 95,7386. F. J. Duncan and R. J. CvetanoviC, J. Amer. Chem. Soc., 1962, 84, 3593 ; F. S. Rowland, P. S.-T. Lee, D. C. Montague and R. L. Russell, Disc. Faraduy SOC., 1972, 53, 111. D. C. Montague, J.C.S. Chem. Comm., 1972,615. R. J. CvetanoviC, Progr. Reaction Kinetics, 1964, 2, 39. lo D. C. Montague, J.C.S. Faruduy I, 1975, 71, 398. l 1 J. W. Simons and B. S . Rabinovitch, J. Phys. Chem., 1964, 68, 1322. l 2 M. C. Flowers and H. M. Frey, Proc. Roy. Soc. A, 1960, 257, 122. l 3 P. J. Robinson and K. A. Holbrook, Unimoleculur Reactions (Wiley, London, 1972). l4 R. G. Bergman, Free Radicals I, ed. J. K. Kochi (Wiley, London, 1973), p. 191. l 5 D. C. Montague and F. S. Rowland, to be published. l6 D. C. Montague, unpublished observations. l 7 J. Metcalfe, H. A. J. Carless and E. K. C. Lee, J. Amer. Chem. SOC., 1972, 94, 7235. G. Wolf, Helv. Chim. Actu, 1972, 55, 1446. l 9 S. W. Benson, Thermochemical Kinetics (Wiley, London, 1968). 2o J. A. Kerr and A. C. Lloyd, Trans. Faruduy Soc., 1967, 63,2480. 2 1 N. J. Turro, W. E. Farneth and A. Devaquet, J. Amer. Chem. Soc., 1976,98,7425. 22 H. A. J. Carless, J. Metcalfe and E. K. C. Lee, J. Amer. Chem. Soc., 1972, 94, 7221. 2 3 T. R. Fields and P. J. Kropp, J. Amer. Chem. Soc., 1974, 96, 7559. 24 S. C. Chan, B. S. Rabinovitch, J. T. Bryant, L. D. Spicer, T. Fujimoto, Y. N. Lin and S . P. 2 5 R. R. Baldwin and R. W. Walker, Fourteenth Symposium (International) on Combustion, 1973, 26 cf. the discussion given in ref. (14). 27 L. M. Stephenson, T. A. Gibson and J. I. Brauman, J. Amer. Chem. Soc., 1973,952849. Pavlou, J. Phys. Chem., 1970,74, 3160. 241. (PAPER 7/248) 1-10

ISSN:0300-9599    

DOI:10.1039/F19787400277

出版商:RSC    

年代:1978

数据来源: RSC