摘要:

J . Chem. SOC., Faraday Trans. I, 1985, 81, 3073-3080 Temperature-programmed Desorption Studies of Alcohol Decomposition on Zinc Oxide Propan-2-01 BY MICHAEL BOWKER,* RAYMOND W. PETTS AND KENNETH C. WAUGH Imperial Chemical Industries PLC, New Science Group, P.O. Box 11, The Heath, Runcorn, Cheshire WA7 4QE Received 18th February, 1985 Temperature-programmed desorption (t.p.d.) has been used to investigate the nature of the interaction of propan-2-01 with the zinc oxide surface. Adsorption at pressures of 10 Torr resulted in an averaged saturation coverage of 2 x 1014 molecule ern+ at 310 K. Temperature programming resulted in the desorption of propan-2-01 itself just above the adsorption temperature and a more strongly adsorbed intermediate species (an alkoxyl) decomposed at cu.480 K to yield products of acetone, propene and hydrogen. It is proposed that these products derive from a sequential reaction scheme ; first a-hydrogen elimination occurs, which yields acetone into the gas phase and an extremely reactive surface hydride. The latter then attacks remaining alkoxyl groups at a /I-H position, the resulting transition state then decomposes to yield the alkene and hydrogen into the gas phase and shows a small temperature lag in the t . p.d. experiments. In a previous paper in this series the results of a study of the decomposition of ethanol on reduced ZnO surfaces were reported.' Ethanol was adsorbed by the surface to produce a strongly bound intermediate which decomposed at 5 10 K to yield mainly ethylene, hydrogen and a small amount of acetaldehyde or ethylene oxide (which are difficult to distinguish in the mass spectrometer).Thus it appeared that the dehydration of ethanol was the dominant reaction route under the transient conditions used in temperature-programmed desorption (t .p .d.). recently reported results of steady-state microreactor experiments for the decomposition of propan-2-01 on zinc oxide which showed only a dehydrogenation of the molecule to acetone and hydrogen. In view of our earlier results with ethanol this selectivity appeared surprising, and so the study described in the present paper was embarked upon in order to determine whether any products of dehydration (propene) could be detected. The technique we have used is the same as in the earlier paper,' namely temperature-programmed desorption.Akiba et EXPERIMENTAL The apparatus used has been described in detail in a previous r e p ~ r t . ~ Briefly, it consists of an adsorption chamber which can be filled with dosing gas to atmospheric pressure. The sample holder, containing 0.2 g of catalyst, was mounted on an FC38 conflat flange in this chamber. This holder consists of a quartz tube (8 mm o.d., 3.8 mm i.d.) within which the catalyst is packed and around which is wound a Pt-Rh wire for sample heating. The heating is provided from a programmable power supply and the heating rate can be varied; a typical heating curve has been reported in an earlier paper4 and is linear above 400 K. 30733074 DECOMPOSITION OF PROPAN-2-OL ON ZnO I I I I I I 350 400 450 500 550 600 TIK Fig.1. Temperature-programmed desorption spectrum of products after the adsorption of propan-2-01 on ZnO at 310 K. The small contributions to any individual product peak from the other products has been subtracted for clarity and the desorption curves are offset vertically for the same reason. (a) Propene (m/e 41) x 3.3, (b) propan-2-01 (m/e 45) x 3.3, (c) hydrogen (m/e 2) x 1 and ( d ) acetone (m/e 58) x 10. After dosing the sample with gas and pumping the ambient away, the temperature was ramped and the products desorbed were measured in a Vacuum Generators QX200 quadrupole mass spectrometer which could be programmed to select four individual mass peaks for one run. Thus several dosing experiments were performed in order to gather complete cracking patterns for the various products desorbing from the surface.The propan-2-01 used was analytical-reagent grade (99.5% pure) and was subjected to pumping cycles to remove any impurities more volatile than propan-2-01 itself; the purity of alcohol was confirmed by in-situ mass-spectral analysis. RESULTS The ZnO sample was initially treated in the same way as described in the earlier papers,1y3*4 i.e. it was heated to 700 K to remove desorbable impurities (0.25% carbonate being the major quoted impurity) and was then reduced at 550 K in order to prescribe a constant initial surface state of the zinc oxide. After this pretreatment the propan-2-01 was dosed onto the catalyst at 10 Torr* for 900 s in a closed system with the sample ca. 310 K. The remaining gas was then pumped away to a base pressure of ca.low6 Torr and temperature programming was begun. The resulting desorption product distribution is shown in fig. 1 . The detectable products were propan-2-01, acetone, propene and hydrogen. Fig. 1 shows experimental curves obtained for four individual masses which are considered as representative of the four * 1 Torr = 101 325/760 Pa.M. BOWKER, R. W. PETTS AND K . C. WAUGH 3075 products seen. Many other mass peaks were also monitored to verify the correctness of the interpretation, although most other masses had significant fragments from two or more of the products. The double-peaked nature of the propan-2-01 desorption is an artefact of the heating method: the ramp takes some time to linearise and is only truly linear above ca.400 K.4 Nevertheless the integral under the desorption curves is still proportional to the surface coverage of the zinc oxide by propan-2-01. Monitoring mass 18 during the desorption sequence showed that no water was evolved during heating. The major features of the desorption were similar to the earlier results for ethano1,l namely that intact alcohol was desorbed just after heating was begun (360-430 K), whereas a more strongly adsorbed species appeared to be left after such treatment, which, on desorbing, yielded the alkene and ketone products with a peak rate at 480 K. Although the decomposition peaks for propan-2-01 appeared in the same temperature range, the lineshapes for propene and acetone were different, the latter being much broader (87 K us 57 K for propene at half the peak maximum intensity) and peaking 12 K earlier than propene.The hydrogen desorption appears to be of intermediate lineshape but with a peak maximum temperature coincident with that of propene. Reduction of the catalyst between each desorption experiment was found to be unnecessary, since the desorption was quite reproducible even without such treatment. The coverage of the sample by the adsorbed alcohol was calculated from the time- intensity integral of the desorption spectra using the pumping speed of the chamber and the sensitivity of the mass spectrometer to the various products. The formula used for such a calculation is BLN, I SR n=------- where n is the surface coverage by adsorbate (in molecule cm-2), L is the pumping speed of the system for the product concerned (in dm3 s-l), N, is Avogadro’s number per Torr per dm3, I is the integral under the desorption spectrum measured at the mass spectrometer (in A s) and B is a correction factor to account for the fact that only a part of the material desorbed comes through from the adsorption chamber to the detector.S is the sensitivity of the machine in A Torr-l and R is the total surface area of the sample. The factor B in the present work was 5.3, while L was 1 dm3 s-l. The sensitivity S depends upon the peak analysed, but for mass 43 for acetone, for instance, it has a value of 3 A Torr-l. The surface area of the sample was 0.6 m2 (3 m2 g-l). Using this method the total coverage of the surface by propan-2-01-derived species (ie. propan-2-01, acetone and propene derived from individual adsorbed propan-2-01 molecules) was 2.2 x 1014 molecule cm-2.In product terms this comprised 8 x 1013 molecule cm-2 of propan-2-01 which were desorbed intact, 9 x 1013 molecule cm-2 of acetone and 5 x 1013 molecule cm-2 of propene. The surface density of Zn and 0 atoms depends on the particular plane exposed, and the analytical-reagent ZnO which we have used can be approximately described as being composed of particles of hexagonal cylindrical shape. The ends of these cylinders are polar (0001) and (0001) planes, while the six coaxial faces are prism planes; from previous SEM/TEM studies the polar planes constitute ca. 20% of the total surface area. The density of Zn and 0 atoms on the prism faces is 1.4 x 1015 atom cm-2, while on the polar faces the density is 1.1 x 1015 atom cm-2. Thus, the averaged surface coverage by adsorbate is ca.0.2 of a monolayer. In their study of propan-2-01 decomposition Akiba et aL2 proposed that the propanol forms a very strongly bound intermediate which they believe to be the enol tautomer of acetone. Therefore as a test of whether adsorbed acetone would produce3076 DECOMPOSITION OF PROPAN-2-OL ON ZnO 1 I I I 1 400 500 T l K Fig. 2. T.p.d. spectrum for acetone adsorption on ZnO at 310 K [(m/e 43) x 3.31. the enol tautomer or might be responsible for the products seen in fig. 1 at CQ. 490 K, the sample was dosed with acetone at 320 K (40 Torr for 300 s resulted in saturation coverage, 1.4 x 1014 molecule cm-2) and the subsequent t.p.d. spectrum is shown in fig.2. These indicate that a strongly adsorbed complex was not formed, the adsorbed acetone being desorbed intact in the low-temperature region. DISCUSSION The results described above are grossly similar to the previous results obtained for the lower alcohol, ethano1;l the alcohol is adsorbed in a relatively weakly bound state which is desorbed in the low-temperature range as the parent alcohol, and forms a more strongly bound intermediate which only leaves the surface at higher temperatures. The mechanism of the interaction with the surface can be written in a similar way to that for ethanol, i.e. initial adsorption probably takes place as follows: C,H,OH(a) +O, -+ C,H,O(a) + OH(a) C,H,OH(a) + OH(a) --+ C,H,O(a) + H,O(g) + V,, (2) (3) where (a) and (g) refer to adsorbed and gas-phase species, respectively, s refers to the lattice sites and V, is an anion vacancy.The nature of the equipment we have used precludes an analysis of the propanol adsorption process (the mass spectrometer is shut off from the adsorption chamber during dosing, because of the relatively high pressures used) so that there is no direct evidence for the mechanism of steps (2) and (3), i.e. water production. However, several features point to this mechanism being operative. First, no water is produced in the desorption process shown in fig. 1, even though a dehydration product of propan-2-01, i.e. propene, is produced during the desorption. If no water were produced during adsorption the surface would become ‘saturated’ with oxygen after one or two desorption experiments owing to this conversion, and the propensity of the surface for the dehydration reaction would diminish.This was found not to be the case. The surface seems to maintain a constant state during repeated experiments, without the necessity for re-reduction. Thus water must be lost from the surface during the adsorption process to maintain this reproducibility,M. BOWKER, R. W. PETTS AND K . C. WAUGH 3077 and so steps (2) and (3) are included. Furthermore, earlier experiments on single-crystal Cu( 1 10) doped with surface oxygen showed that alcohols interact with the surface in just this way to produce water which is lost from the sample below room temperat~re.~ However, this mechanism must only apply to that fraction of the surface which subsequently produces propene.The acetone production must proceed through a dehydrogenation mechanism which does not evolve water from the surface, probably by the recombination of dissociated hydrogen atoms during the adsorption (Griffin and Yates6 have shown that such a process can occur below 300 K on zinc oxide): C,H,OH(a) --+ C,H,O(a) -+ H(a) (4) These processes leave a surface populated by 2-propoxyl and adsorbed alcohol species. The alcohol desorption at low temperatures is either due to recombination of adsorbed alkoxyl and hydrogen atoms [the reverse of step (4)] or, more likely, to the desorption of intact propan-2-01 molecules bonded to the surface by relatively weak, oxygen lone-pair interactions with the surface :j C,H,OH(a) -+ C,H,OH(g).(6) This then leaves the surface covered with two types of adsorbed alkoxyl species, those with the usual nearest-neighbour oxygens and those with anion vacancies produced by step (3). They then decompose at a similar temperature to produce quite different reaction products. The alkoxyl species, which is associated with an anion vacancy in the zinc oxide lattice, decomposes in a way which removes the vacancy, thus: C,H,O(g) + v o -+ C3Hdg) + H(a) + O(S) (7) and so balances out the loss of surface oxygen shown in step (3). The deprotonation presumably takes place at adjacent sites on the zinc oxide byp-H removal, as proposed for ethano1:l H H I CH3 \c / / ‘C’H .) c\ -Zn-0-Zn-0-Zn-0-Zn-0-Zn-. The dehydrogenation reaction probably takes place at different sites : either at stoichiometric ZnO sites, as proposed by Fahim et aL7 for propan-2-01 decomposition on ceria, or on the Zn-dominated (0001) polar face of the sample (which constitutes ca.10% of the total area): The possibility that this reaction occurs on the more metal-like polar face is attractive, since the noble metals dehydrogenate in this way5 and the amount of material desorbing as acetone (9 x lo1, molecule cm-2) seems to correspond reasonably well with the total Zn polar-face sites (ca. 11 x 1013 atom cma2 of total catalyst area). However, one feature of the data which does not seem to fit well with the simple mechanism outlined above is the similarity of the decomposition temperature for the two products from two such different sites. Usually, in t.p.d.experiments, the evolution of products in coincident peaks is taken to indicate that the products evolve3078 DECOMPOSITION OF PROPAN-2-OL ON ZnO from an adsorbed intermediate which is common to them all. By this token the products we observe would be derived from some kind of dimeric surface complex, but without spectroscopic equipment the existence of such an intermediate cannot be proved. A more likely explanation for the behaviour observed may, however, be found in the work of Ashby et al.,s who studied the decomposition of a variety of magnesium and zinc dialkoxides and alkyl alkoxides. They found that for the dialkoxides, for instance, a generalised mechanism can be written by reference to the decomposition of dicyclohexyloxymagnesium : This, then, seems to explain all the products seen in the above experiment.Furthermore, it could explain both the similarity of peak desorption temperatures for acetone and propene and the fact that acetone slightly precedes the latter. Clearly the whole mechanism depends on the occurrence of step (A) (a-hydrogen elimination) which also produces an hydride species. If this reaction occurs on a particular site of the ZnO surface, this hydridic species must be mobile in order to 'seek out' other alkoxide species with which to react; it is not likely that, for a zinc ion bound to a substrate, two alkoxide species could be attached to the same ion in the way shown above for metal alkoxide molecules. This hydridic, and presumably cation-bound, entity then deprotonates an alkoxide to produce propene in a p-elimination reaction [step (B)].However, this second step does involve a further small activation energy barrier which results in a temperature shift of the alkene product. Furthermore, it is possible that, as described above, some recombination of hydrogen from step (A) can take place which results in < 50% evolution of propene. Indeed, if the decomposition were as shown in steps (7) and (8) above, the hydrogen-desorption envelope should simply be a convolution of the acetone and propene curves, which it clearly is not from the averaged normalised lineshapes shown in fig. 3. (It is skewed towards the propene curve.) Furthermore, if the mechanism were exactly like that of Ashby et aL8 the lineshape should be coincident with the propene, whereas the halfwidth is between the two.This, together with the fact that there are unequal amounts of the two organic products, indicates that some hydrogen-atom surface recombination takes place corresponding with the lower amount of alkene production by hydride-induced /I-elimination. From their transmission infrared measurements Koga et aL9 conclude that an enol- type intermediate is formed from either propan-2-01 adsorption and heating to 363 K, or acetone adsorption at room temperature, yet in the present work the same intermediate certainly was not formed from acetone. As fig. 2 shows, there was little evidence of strong adsorption of acetone at all. The source of the difference between these results and those of Tamaru and coworkers2? in this respect is not clear, though it is possible that residual OH groups on the ZnO in Tamaru's work contributed to acetone adsorption and reaction. However, several points of similarity between the present results and Tamaru's are noteworthy.First, the specific rate of acetoneM. BOWKER, R. W. PETTS AND K. C. WAUGH 3079 I I I I I I 400 450 500 550 600 T I K Fig. 3. Normalised desorption peaks for propan-2-01 decomposition on ZnO. The solid line is acetone desorption, the dashed line is propene and the dotted line is hydrogen. The curves are the averaged result of four experiments for each product, and mean deviations of signal are indicated on the figure. formation from the latter work2 at saturation surface coverage (1.5 x 1O1O mol- ecule cm-2 s-l at 363 K) is very similar to the rates of acetone formation (at the leading edge of the acetone desorption, ca.1.2 x 1 O 1 O molecule cmW2 s-l at 370 K) shown in fig. 1. Secondly, at the low temperatures which they used, these authors observed the build-up of a species (which they designate as an enol) which is inactive for acetone production in the absence of gas-phase alcohol; this species may well correlate with the propene production shown in fig. 1, the rate of which, at the low-temperature desorption edge, is very much slower than that of acetone. This appears to support the two-species aspect of the model described above. The major difference between these two works is the large amount of propene production seen in the present work. The difference seems to be explicable in terms of (i) the much faster production of acetone at low temperatures, even though the total temperature- integrated product ratio is very similar, and (ii) the fact that the ‘inactive’ species of Tamaru’s work can be converted into acetone by the presence of propan-2-01 in the gas phase, thus preventing its build-up and subsequent blockage of acetone production.In conclusion then, temperature-programmed desorption experiments have shown that propan-2-01 interacts strongly with the ZnO surface to produce adsorbed alkoxyl-type species. These decompose to produce acetone and propene with a peak rate at 480 K. The source of the two products is considered to be two different adsorption sites, those with and those without adjacent anion vacancies, respectively. The alkene is produced by the interaction of hydridic species (generated from the alkoxyl decomposing to yield the ketone) with the more strongly bound anion vacancy associated alkoxyl, which results inp-hydrogen elimination/abstraction. However, the generality of these observations will be examined by further experiments using the adsorption of other alcohols on ZnO in combination with these techniques and will be reported upon completion. We are grateful to the referees for their careful reading of the original manuscript.3080 DECOMPOSITION OF PROPAN-2-OL ON ZnO M. Bowker, H. Houghton and K. C. Waugh, J. Chem. SOC., Faraday Trans. 1, 1982, 78, 2573. M. Bowker, H. Houghton and K. C. Waugh, J . Chem. SOC., Faraday Trans. I , 1981, 77, 3023. M. Bowker, H. Houghton and K. C. Waugh, J. Catal., 1983, 79, 431. M. Bowker and R. J. Madix, Surf. Sci., 1980, 95, 190; 1982, 116, 549. G. L. Griffin and J. T. Yates, J. Catal., 1982, 73, 396. R. Fahim, M. Zaki and R. Gabr, J. Chem. Tech. Biotech., 1980, 30, 535. E. Ashby, G. Willard and A. Goel, J. Org. Chem., 1979,44, 1221. 0. Koga, T. Onishi and K. Tamaru, J. Chem. Soc., Faraday Trans. 1, 1980,76, 19. * E. Akiba, M. Soma, T. Onishi and K. Tamaru, 2. Phys. Chem., 1980, 119, 103. (PAPER 5/275)

ISSN:0300-9599    

DOI:10.1039/F19858103073

出版商:RSC    

年代:1985

数据来源: RSC

摘要:

J. Clwm. Soc., Faraday Trans. 1 , 1985,81, 3081-3090 Solute-Solvent Interactions in Water-t-Butyl Alcohol Mixtures Part 14.---AGQ, A H 0 and A S 0 of Transfer for Alkaline-earth-metal Cations BY JEAN JUILLARD* AND CLAUDE TISSIER UA 434 du CNRS, Universite de Clermont 2 , B.P. 45, 63 170 Aubiere, France AND JOLANTA BARCZYNSKA, JERZY MOKRZAN AND STEFANIA TANIEWSKA-OSINSKA Department of Physical Chemistry, University of t o d i , Novotki 18, 91-416 t o d i , Poland Received 27th February, 1985 Solubility products of alkaline-earth-metal fluorides and enthalpies of solution of alkaline- earth-metal chlorides have been determined in water and water-t-butyl alcohol mixtures containing up to 40 wt ”/, of alcohol. Using the classical assumption of an equivalent change of solvation with the solvent media of tetraphenylarsonium and tetraphenylboride ions, it was, from these data and data obtained previously, possible to estimate ionic enthalpies, entropies and Gibbs free energies of transfer of alkaline-earth-metal cations from water to the mixtures.The results are discussed in terms of the structure of the water-t-butyl alcohol mixtures. Thermodynamic functions of transfer of solutes from water to mixed solvents give informations about the solvation changes which occur when an organic cosolvent is added to water. Much work has been done on a great variety of electrolytes in water-alcohol mixtures but little attention has been paid to alkaline-earth-metal cations, except in water-methanol by Feakins and Wi1lmott.l This is because the only method previously used for AG studies, determination of IF of amalgam electrodes, is not easy.In the course of a continuing study2-4 of solute-solvent interactions in water + t- butyl alcohol (ButOH) media we report here data on the thermodynamic functions of transfer of alkaline-earth-metal cations from water to water + t-butyl alcohol mixtures containing up to 40 wt % alcohol. A G e of transfer of alkaline-earth-metal fluorides are calculated from their solubility products obtained using a fluoride electrode and AH@ of transfer of alkaline-earth-metal chlorides are calculated from their A H of solution. From previous data,3 which allow the ionic contributions to transfer function of electrolytes in those media and A G e of transfer of KF to be calculated, obtained here from potentiometric measurements, it is then possible to calculate to A c e , A H e and A S e for transfer of Mg2+, Ca2+, Sr2+ and Ba2+ from water to t-butyl alcohol mixtures.EXPERIMENTAL PRODUCTS Water and t-butyl alcohol were purified as before. The fluorides were dried in a vacuum oven after careful grinding. CaC1, was dehydrated following the procedure previously described.j SrC1, and BaC1, were dried at 450 K under vacuum. The molality of concentrated aqueous solutions of MgC1, was estimated through potentiometric titration using a silver nitrate solution. 308 13082 SOLUTE-SOLVENT INTERACTIONS POTENTIOMETRIC MEASUREMENTS Two types of ions selective electrodes were used: a glass electrode selective to K t ions (Tacussel type PMeV) and an electrode, with a lanthane fluoride monocrystal, selective to F- ions (Orion type 94-09).The behaviour of the K+ electrode was improved6 and the F- electrode was checked in every solvent mixture used prior to measurement. A Nernstian response was always found, in agreement with previous studies by Coetzee et af.' in water and alcohols. DETERMINATION OF GIBBS FREE ENERGY OF TRANSFER OF KF The following galvanic cell was used : K +(glass) 1 K+, F- I F--selective electrode. The procedure used has been described previously.6 It consisted of adding to a KF solution of molality rn in water, a KF solution of molality rn in water- t-butyl alcohol (30-70 wt %) and then following the change in the e.m.f. of the galvanic cell; measurements were performed for various molalities and extrapolation to rn = 0 allowed AE* to be obtained (the variation of the standard potential of the cell with the alcohol content in the solvent mixture).The following apriori form of the mean molal activity coefficient was used for the extrapolation. A\ tll l+Bq\ oz logy* = --___ + hrn + log (1 + 0.002 Mrn) where rn is the molality, A , Band q are the DebyeeHiickel and Bjerrum parameters in the solvent and b is an empirical parameter. SOLUBILITY PRODUCTS OF ALKALINE-EARTH-METAL FLUORIDES Saturated solutions were prepared by vigorous shaking of the solid salt with the solvent mixture for one week; this was done in a closed vessel immersed in a bath maintained at 25.0 "C. Part of the saturated solution, u,,cm3, was then poured into a cell with two electrodes, an aqueous saturated calomel electrode and a fluoride-selective electrode.Such a galvanic cell : F--selective electrode 1 F-, M2+ 1 1 calomel reference electrode has a potential given by: Eo = E$? - k log S " yo -- Ercsf - Ei where E$? is the potential of the fluoride electrode, Eref the potential of the saturated calomel electrode, Ei a junction potential, so is the solubility and yo is the corresponding mean molar activity coefficient. In order to prevent any change in the standard potential of such a cell, standardization through addition of sodium fluoride was performed in situ. Then, if u, cm3 of a sodium fluoride solution of molarity c' in the same solvent were added to the solution, the e.m.f. was: with [F-1, = [cv,/(v,,+ u,)] +s,, where s, is the concentration of F- ions which result from the solubilisation of the alkaline-earth-metal fluoride.The change of the e.m.f. was then, assuming no change of E j : (2) Ex = E$? - k log [F-1, yx - E,,., - Ej AE = E,-E, =z - k log (L,(;;L)x+sx)Yx+k ~ogsol'o. ( 3 ) On the other hand, the solubility product of the alkaline-earth-metal fluoride was given by: Combination of eqn (3) and (4) allowed s, to be determined and then the ionic strength; the mean activity coefficient was calculated using :J. JUILLARD et al. 3083 Finally, a value of soyo was obtained for each addition of the sodium fluoride solution. Since the number of such additions was of the order of ten, a mean value of soyo could be calculated with reasonable accuracy (1-5%). The Gibbs free energies of transfer were then obtained from the solubility products: AGP = RT loglo (pwK, -psKs) where w stands for water and s for the solvent mixture.CALORIMETRIC MEASUREMENTS All the calorimetric measurements were performed using a home-made ‘ isoperibol ’ calorimeter which as been described previ~usly.~ Enthalpies of solution, as well as of dilution, were obtained by breaking ampoules containing either the dry solid salt or a concentrated solution. Enthalpies of transfer of Ca, Sr and Ba chlorides were estimated from heat of solution of these salts in water and in the mixtures. The enthalpies of solution were measured for various concentrations from ca. to 2 x lod2 and extrapolated to infinite dilution using a method proposed by Criss and Cobble.8 Enthalpies of transfer of magnesium chloride were estimated from the heat of dilution of a concentrated aqueous solution both in water and in the mixtures. Such a method has been described and discussed previo~sly;~ measurements and calculations were identical, except that enthalpies of dilution were not calculated but measured using a classical method; the fitting and extrapolations required were done using the following form of the Debye-Huckel equation where dL is the relative apparent molal enthalpy, I is the ionic strength and: The parameters using : B, and CH were fitted from the experimental A I variation of dL on dilution, A&, RESULTS The experimental data are reported in tables 1-4. Corresponding standard enthalpies and Gibbs free energies of transfer for alkaline-earth-metal fluorides and chlorides are given in table 5 .From previous determinations of ionic contribution to AG* and AH* of transfer in the same media, using the tetraphenylborate assumption, it was possible to calculate the enthalpies, free energies and thus entropies of transfer of alkaline-earth-metal cations. The results are given in tables 6-8. DISCUSSION Previous data2-4 on alkali-metal ions, halides and tetra-alkylammonium ions have been discussed in terms of the structure of the water-t-butyl alcohol mixtures. The results here for alkaline-earth-metal cations do not contradict the conclusions reached previously. The fluctuating-cages modello which is used here to describe the structure of the water-ButOH mixtures has been discussed previ~usly.~ It is therefore unnecessary to repeat the arguments.To sum up, introduction of an alcohol molecule to water will result in reorganisation of the solvent into a fluctuating cage of water surrounding3084 SOLUTE-SOLVENT INTERACTIONS Table 1. Solubility products of alkaline-earth-metal fluorides (molar scale), pK,, ButOH (wt "/,) 0 10 20 30 40 ~~~ ~~ MgF, 9.06 9.78 10.35 10.86 11.59 CaF, 10.40 11.19 11.82 12.64 13.53 SrF, 8.42 9.39 10.34 11.05 11.86 BaF, 5.84 6.78 7.73 8.49 9.25 Table 2. Standard enthalpies of solution, A H . (kJ mol-') ButOH (mol% ) CaCl, SrCl, BaCl, 0 (- 2.5 5.0 7.5 10.0 15.0 - 8l.O+ 0.3 - 51.13 f0.12 (-51.17", -50.74d, -81 .08,a - 81 .30b) - 5 I. 12e) - 77.6 -45.9 - 72.4 - 39.9 - 68.7 - 35.1 - 68.8 - 35.4 - 69.6 - 36.3 - 13.5 (- 13..5Ib, - 13.35') - 8.3 - 2.3 + 2.6 +2.3 + 1.5 a A.Finch, P. J. Gardner and C. J. Steadman, J . Phys. Chem., 197 1,75,2325. V. B. Parker, P. D. Wagman and W. H. Evans, Natl Bur. Stand. Tech. Note, 270-6, 1971. J . Greyson and H. Smell, J. Phys. Chem., 1969, 73, 3208. A. Dadgar and M. R. Taherian, J . Chem. Thermodyn., 1977,9,711. L. M. Morss, and C. W. Williams, J. Chrm. Therrnodyn., 1983,15, 279. the alcohol molecule. Such cages have a well defined stoichiometry and the apex of the mixture organisation corresponds to the stoichiometry of the cages (1 / 17 to 1 /2 1); after that, the number of free ButOH molecules as well as aggregates of the cages increases, corresponding to a decrease in the structure. Therefore, as previously, three zones are observed in the curves of the variation of enthalpies and entropies against composition (fig.1 and 2). Zone I : For moles fractions up to 0.03, enthalpy and entropy changes are weak. In these water-rich solutions the water structure is thought to be enhanced by adding alcohol. Changes in the enthalpies of solvation are thus slightly endothermic, corresponding to a reinforced structure; there is in fact, no change in the entropy of solvation for Ca2+, as if the restriction of free water molecules is compensed for by a reinforcement of the water-cation interactions. Zone 2: For mole fractions from 0.03 to 0.08, strong variations are observed both in the enthalpies and entropies of transfer, underlined by the existence of large maxima located at an alcohol mole fraction of ca. 0.05. Both enthalpies and entropies are positive, corresponding to a collapse of the local structure, the effects being stronger as the structure of the pure mixture increases.Two effects could be determined the amplitude of the maxima in the entropy or enthalpy curves: (a) the size of the cationJ. JUILLARD et al. 3085 Table 3. Enthalpies of dilution of MgCl, m1 m2 %(m, -, m2> m ( m , -+ 0) /mol kg-l /mol kg-l /kJ mol-l /kJ mol-l 4.776 0.5127 0.5127 0.5127 0.5 127 water 0.0098 17.1 1 0.00 12 17.01 0.00 18 16.88 0.003 1 16.58 0.0038 16.46 0.0053 16.60 0.0082 16.56 0.0 103 16.60 2.5 mol % ButOH 0.0020 4.17 0.0030 4.07 0.0038 4.00 0.0039 3.99 0.005 1 3.87 0.0063 3.69 0.0083 3.60 5mol% ButOH 0.0020 5.05 0.0029 4.75 0.0039 4.62 0.0047 4.49 0.006 1 4.35 7.5 mol % ButOH 0.0021 5.21 0.0026 5.07 0.0039 4.73 0.0040 4.71 0.0049 4.63 0.0050 4.55 0.0062 4.25 10 mol% ButOH 0.0029 4.99 0.0039 4.80 0.0041 4.77 0.0052 4.36 0.0062 4.27 0.0074 4.28 0.30 0.34 0.40 0.52 0.57 0.66 0.79 0.86 0.68 0.83 0.94 0.95 1.09 1.21 1.40 0.86 1.05 1.25 1.38 1.61 1.10 1.24 1.58 1.61 1.82 1.84 2.10 1.36 1.59 1.64 1.86 2.05 2.26 (the larger it is the more it has to disrupt to enter the structure) or (6) the electrostrictive power of the ion (the stronger it is, the more the ion disrupt the structure). As far as the latter property is concerned, electrostrictive powers, as expressed by the difference between the molar volume and crystallographic volume, are almost the same for the alkaline-earth-metal cations (- 29.5 to - 32 cm3)11 and larger than those of the alkali-metal cations.A strong promoting effect of alkaline- earth-metal cations is therefore expected.This is well known in water,l2?l3 but here there is competition between two ways of organizing the water network: hydrophobic3086 SOLUTE-SOLVENT INTERACTIONS Table 4. Enthalpies of transfer of alkaline-earth-metal chlorides from water to water + ButOH (kJ mol-l) MgC1, CaC1, SrCl, BaCl, XU P AHB" AHDd azp 0 0.0 126 0.0250 0.0263 0.041 1 0.0500 0.0573 0.0750 0.0943 0.1000 0.1 157 0.1394 0.1500 - 5.00 9.54 10.00 15.00 17.78 20.00 25.01 30.00 31.57 35.00 40.00 42.06 - - 15.1 - - - 13.9 - - 13.8 - - 14.8 - - - - 17.3 - - 16.0 - - - 15.1 - - 15.4 - - 16.4 - - - 0 0.7 2.0 2.2 5.2 6.9 7.9 9.4 9.4 9.2 0 1.7 3.4 3.8 6.5 8.6 10.8 12.3 12.2 12.2 12.0 11.5 11.4 0 2.7 5.2 5.3 9.1 11.2 12.9 16.1 16.0 15.7 15.6 15.1 14.8 0 2.7 5.2 5.4 9.2 11.2 14.0 16.1 16.0 15.8 15.7 15.2 15.0 Mole fraction of ButOH.Wt % of ButOH. " dH,: enthalpy of dilution from rn, = 4.776 in water to rn, = 0.0039 mol kg-l in the mixture. (mixture): enthalpy of dilution from 0 2 , = 0.0039 mol kg-l to infinite dilution in the mixture. -- = ALHB+AHdil (mixture); Table 5. Gibbs free energy of transfer for fluorides (molar scale, kJ mol-l) ButOH (wt % ) KFu MgF2b CaF,b SrF,b, BaF,* 5" 10 15" 20 25" 30 3Y 40 1.35 2.70 4.08 5.33 6.45 7.53 8.64 9.85 2.3 4.1 5.6 7.1 8.5 9.9 11.6 13.9 2.4 4.4 6.2 7.9 9.9 12.2 14.7 17.3 2.8 5.4 8.1 10.6 12.7 14.8 16.9 19.0 From potentiometric measurements; from solubility products ; interpolated values. hydration by t-butyl alcohol and hydrophilic hydration by M2+ cations.Therefore, the structure of the water-alcohol mixtures is locally disrupted by M2+ ions, owing to their strong attraction for water molecules. The amplitudes of the maxima in the entropy and enthalpy curves, approximately the same for all alkaline-earth-metal cations, is two times higher than for alkali-metal cation, which is in agreement with the differences observed in the electrostrictive powers. Mg2+ and Li+ both have smaller variations, which could correspond to interstitial location of both ions in the pseudo-clathrate solutions. Zone 3: In this region there is an increase in the number of free molecules or aggregates and the structure is progressively destroyed. As a result there is a decrease in both the enthalpy and entropy of solvation of alkaline-earth-metal cations.TheJ. JUILLARD et al. 3087 Table 6. Standard ionic Gibbs free energies of transfer AGP (molal scale, kJ mol-') F-b M g Z f b CaZfb sr2+ Ba2+b ButOH (wt %) K"" 3 5 10 15 20 25 30 35 40 1 .O 0.3 2.1 0.6 2.0 2.0 1.8 3.5 1.4 5 .O 1 .o 6.5 0.9 7.7 1.4 8.5 1.7 2.9 1.6 0.1 - 1.5 -3.1 - 3.8 -3.1 1.8 3.2 2.2 0.9 -0.1 - 0.8 - 0.7 +0.3 2.2 4.2 4.1 3.6 2.7 1.8 1.5 2.0 a From previous data; ref. (4); from data in table 2. Table 7. Standard ionic enthalpies of transfer AH? (kJ rnol-') from water to water+ButOH mixtures ButOH (wt %) C1-" Mg2+') Sr", Ba2+b 5 10 15 20 25 30 35 40 - 0 0 - 6.6 10.1 - 2.5 1.3 4.0 5.6 0.7 1.7 2.2 3.8 18.4 19.7 28.1 31.0 14.4 17.3 6.8 9.6 4.0 0.3 - - 2.7 5.4 22.3 33.0-34.2 21.1 13.4 7.6 3.9 a From previous data, ref.(3); from data in table 5. Table 8. Standard entropies of transfer ASP of alkaline-earth-metal cations from water to water + ButOH mixtures (molal scale, J mol-l) ButOH (wt%) Mg2+ Ca2+ Sr2+, Ba2+ 5 10 15 20 25 30 35 40 - 3.3 -0.3 - 3.0 0.6 56 58 94 101 53 58 33 35 16 0 - - 1.7 4.0 61 99 62 39 20 6.43088 SOLUTE-SOLVENT INTERACTIONS 10 20 30 ButOH(wt%) Fig. 1. Variations ofthe standard enthalpiesoftransfer (kJ mol-l) ofsomecations with the ButOH content in the mixture: (a) Li+, (b) Na+, (c) Rb+, (d) Mg2+, (e) Caz+ and (f) Sr2+ and Ba2'. effects of highly structured media disappear and interactions between cations and ButOH molecules, which cause basicity,14 occur. This effect could explain the stronger decrease observed for alkaline-earth-metal than for alkali-metals cations.Data for the calcium ion are presented in fig. 3. AG* of transfer results from compensation between A H 0 and A S e . The effects thus obtained are therefore, as compared with enthalpy or entropy variations, small but significant. Differences observed in zone 1 between A H 0 and A S 0 cause an initial increase in AGe, then a decrease and finally an S-shaped curve is obtained. Plots of AGe against alcohol content have the same form for all the cations, but the amplitude is larger for alkaline-earth-metal than for alkali-metal ions (fig. 4). Because of the complexity of the curves it is difficult to predict what will happen to the Gibbs free energies when alcohol is added to water. In summary, the effects observed here are similar to those observed with alkali-metal cations but larger ; alkaline-earth-metalcations break the structure of the water + t-butyl alcohol media in the water-rich zone; this corresponds to large endothermic enthalpic effects and large positive entropies of transfer from water to a particular composition of the mixtures where the structure is supposed to be at its greatest.J. JUILLARD et al.3089 9 0 80 30 20 10 I I I I I I I I * 10 20 30 40 ButOH (wt% ) Fig. 2. Plot of standard entropies of transfer (J mol-l K-l) of some cations against the ButOH content in the solvent mixture: (a-f) as in fig. 1. Bu'OH (wt%) Fig. 3. Plot of (a) standard Gibbs free energies, (b) enthalpies and (c) entropies of transfer of calcium from water to water-ButOH mixtures.3090 5 i SOLUTE-SOLVENT INTERACTIONS I 0 e E I I I I I I I I t 10 20 30 L O ButOH ( w t x ) Fig. 4. Plot of the standard Gibbs free energies of transfer of some cations from water to water-ButOH mixtures (molar scale) against the amount of ButOH in the mixture: (a) H+, (b) Lif, ( c ) Na+, ( d ) Mg2+, (e) Ca2+ and ( f ) Sr2+. D. Feakins and A. R. Willmott, J. Chem. SOC. A, 1970, 3121. J. Juillard, J. Chem. SOC., Faraday Trans. I , 1982, 78, 37. J. Juillard, J. Chem. SOC., Faraday Trans. I , 1982, 78, 43. J. Juillard and C. Tissier, Electrochim. Acta, 1982, 27, 123. S. Taniewska-Osinska and J. Barczynska, J. Chem. Soc., Faraday Trans. I , 1984, 80, 1409. Y. Pointud, J. Juillard, J.-P. Morel and L. Avedikian, Electrochim. Acta, 1974, 19, 229. 'I J-F. Coetzee and M. W. Martin, Anal. Chem., 1980, 52, 2412. * C. M. Criss and J. W. Cobble, J. Am. Chem. SOC., 1961, 83, 3223. lo E. K. Baungartner and G. Atkinson, J. Phys. Chem., 1971,75, 2336. l1 F. J. Millero, in Water and Aqueous Solutions, ed. R. A. Horne (Wiley-Interscience, New York, 1972), p. 519. l2 K. P. Mischensko and G. M. Poltoratski Problems of Thermodynamics and Structure of Aqueous and Non-aqueous ElectroIyte Solutions (Plenum Press, New York, 1972). l3 G. A. Krestov, Thermodynamics of Zonic Processed in Solution (Khimia Press, Leningrad, 1973). l4 M. K. Chantooni and I. M. Kolthoff, J. Phys. Chern., 1978, 82,994. Y. Pointud, J-P. Morel and J. Juillard, J. Phys. Chem., 1976, 80, 2381. (PAPER 5/336)

ISSN:0300-9599    

DOI:10.1039/F19858103081

出版商:RSC    

年代:1985

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. I, 1985,81, 3091-3102 Ionic Solvation in Water-Cosolvent Mixtures Part 11 .-Free Energies of Transfer of Single Ions from Water into Water-Urea Mixtures BY GRAHAME S. GROVES AND CECIL F. WELLS Department of Chemistry, University of Birmingham, Edgbaston, P.O. Box 363, Birmingham B15 2TT Received 4th March, 1985 The use of the spectrophotometric solvent sorting method for determining the free energy of transfer of the proton, AG,"(H+), from water into waterxosolvent mixtures has been extended to mixtures formed by adding structure breaking urea to water. Values for AG;(XZ-) are calculated from values for AG,"(H,X) using the experimentally determined values for AG,"(H+), and the derived values for AG:(XZ-) with x = 1.0 are used to calculate values for AG:(M+).Values for AG,"(XZ-) are positive and values for AG:(M+) are negative at mole fractions of urea x < 0.15, as found for mixtures of water with structure-forming cosolvents (excluding BPh,). However, in water-urea mixtures, the influence of structure-forming hydrocarbon ligands in controlling values for AG,"(i) is much reduced, compared with their influence when structure- forming cosolvents are used: for example, AG:(BPh;) is positive with added urea and negative with other cosolvents. Moreover, values for AG,"(i) for i = Ph,As+ and BPh;, which are closely related in size and peripheral hydrocarbon structure, differ considerably in water-urea mixtures, showing the importance of the sign of the charge. Following the application of the spectrophotometric solvent sorting procedure for determining the free energy of transfer of the proton, AG:(H+), from water into mixtures of water with a range of cosolvents,1-8 thereby enabling values for AG,"(X-) to be determined from AG,"(HX) and values for AG:(Mn+) from AG;(MX,), this method has now been applied to mixtures of water with urea.The alcoholic2*4 and ketonic4? cosolvents previously investigated all enhance the structure of water quite considerably when added in low mole fractions; added dioxane,6 dimethyl sulph- oxide,' ethanonitrile3 and ethane-1 ,2-dio18 increase the structure to a lesser extent and glycerol perhaps has little effect at all.4 Moreover, all these cosolvents decrease the dielectric constant of the In contrast, when solid urea is added to water the dielectric constant of the mixture increasesg and all the physical evidence,l07 l1 particu- larly the viscosity B coefficient of urea12 and the decrease in the structural element of the ultrasonic absorption with concentration of urea,13 supports the view that urea is a structure-breaker14 when added to water.It is therefore of considerable interest to compare the order of values and the relative spread of values of AG:(i) among a range of individual ions i in water-urea mixtures with those in which the added cosolvent enhances the structure and reduces the dielectric constant. The experimental determination of AG:(H+) using the spectrophotometric method involving the addition of traces of 4-nitroaniline to water-urea mixtures is now reported. These values for AG,"(H+) are then used to calculate values for AG,"(i) for individual ions, i, in water-urea mixtures.309 13092 IONIC SOLVATION IN WATER-COSOLVENT MIXTURES EXPERIMENTAL Analytical reagent urea was used with the other materials needed for the previously reported1-& procedure for the spectrophotometric determination of the concentration of the unprotonated 4-nitroaniline at 25 OC.' As the contraction on mixing urea and water is so great, solutions were made up to a standard volume using known weights of urea and water. This obviates the need for the use of a large correction to the molar concentration of 4-nitroaniline and to the values for the molar concentration quotient Kc for the contraction in volume on mixing, in contrast to the small correction applied to allow for the small contractions found when known volumes of water and the other cosolvents were mixed.l-* RESULTS AND DISCUSSION DETERMINATION OF AG:(H+) IN WATER-UREA MIXTURES Values for the free energy of transfer of the proton between water and water-urea mixtures have been determined in two parts as defined by AG:(H+) = AG:(H+),+AG(H+),,,,.(1) AG:(H+), for the transfer of the charged sphere (H,O+)(H,O), from water into the mixture is given by eqn (2) on the molar sca1e:l AG,"(H+), = ___ Ne2 (D;l- D;I) = 1 67.6(Dg1 - Dil) kJ m o t 1 6yH20 where N is Avogadro's number, rHZ0 is the radius of the water molecule, D is the dielectric constant and subscripts w and s indicate water and the mixture, respectively. Values for D, were interpolated using the experimental values of Wyman9 and of Kundu and Mazumdar.15 The free-energy change for the sorting of the solvent molecules water and urea in the vicinity of the sphere subsequent to the transfer of the latter into the mixture, AG(H+),,,,, is represented by (H,O),H:Olv + (NH,),CO,,lV + ~ ~ ~ 2 0 ~ , - 1 ~ ~ H 2 ~ 2 ~ 0 ~ ~ , 1 ~ + H2Osolv (3) where z 2 5 to allow for sorting outside but close to the original sphere: it is assumed that the bulk concentration of urea is low enough to limit the exchange to equilibrium (3).The free-energy change for each mole of protons transferred from water into the mixture is given AG(H+),,,, = - [H+(NH,),CO] RT In {Kc[H20] Fc}. (4) H+WH,),CO) = ~ ~ ~ 2 0 ~ , - 1 ~ ~ ~ 2 ~ 2 ~ 0 ~ ~ ~ 0 - , , , ~ Kc = ~ ~ + ~ ~ ~ ~ 2 ~ 2 ~ ~ {where P = (H,O),H,f,,,) and Fc = yH+ureayH,O/yPyurea with yi the activity coefficient for species i in any particular H,O-urea mixture related to the standard state of yi = 1.0 and [I] = 1.00 mol dm-3 with yi + 1 as [z] --+ zer0.l The total free energy of transfer of the proton on the mole-fraction scale is then given by AG,"(H+) = AG:(H+), - [Ht((NH2)2CO}solv] RT In {Kc[H20] Fc> + RT In ___ d s M Y ( 5 ) d w Ms dis the density and the values of ds were obtained by interpolation of the data of Kundu and Mazumdar,15 M , is the molecular weight of water and M, is the molecular weight of the mixture { = 100/[(wt % urea/60.06)+(wt % H,0/18.016)]); values for K , and F, are determined experimentally.The determination of K , and F, is performed spectrophotometrically after theG .S. GROVES AND C. F. WELLS 3093 Table 1. Values of K , (dm3 mol-I) calculated from K, F, and values of K , F,' (dm3 mol-I) calculated from the slopes at an ionic strength of 1.00 mol dmV3 and at 25 "C concentration of urea 5.19 10.0 19.0 27.9 36.4 (wt %) acidity 0.0 162 0.0323 0.0657 0.104 0.147 (mole total added /mol dm-3 fraction) 0.10 1.81 3.64 4.3 8.5 3.4 0.16 1.60 2.55 2.7 3.6 5.4 0.20 1.61 2.60 3.1 3.4 - 0.40 I .52 2.1 1 6.0 3.6 0.5 0.80 1.36 2.57 2.9 - - 9.7 7.9 5.4 4.3 3.5 1.64 f 0.01 1.82 0.02 2.97 & 0.04 3.72 & 0.04 4.60 f 0.06 from slope addition of a trace of 4-nitroaniline (B) to the water-urea mixture containing an added acid, HC1.l B competes for P and H+{(NH)2CO}solv in the equilibria Eqn (8) applies1-* to equilibria (6) and (7): and is obeyed by similar systems containing other added cosolvents.1-8 Kl and K2 are the thermodynamic equilibrium constants for equilibria (6) and (7) using the standard states specified earlier, Co is the total added constant concentration of 4-nitroanilinel and C and CR are the spectrophotometrically determinedl concentrations, [B], without and with added urea for the same added C, and [HCl] at a constant ionic strength of 1.00 mol dm-3 (maintained by adding NaCl).l 4 = y,y,/y,,+yH2, and F2 = yByH+urea/yBH+yurea. For constant [urea]total, C, and temperature, a plot of CCR/(CR - C) against C,/(C, - C,) should be linear for varying [HCl], as found with other added cosolvents.l-* Such linear plots with an intercept = [H20]C,/Kl F,[~rea],,,~,, using the value of Kl F,/[H20] determined in the absence of urea1 and C, = 1.45 x mol dm-3 were obtained for total added urea con- centrations of 5.19, 10.0, 19.0, 27.9 and 36.4 wt % at 25 "C.obtained in this way are given in table 1. However, From the slopes of these plots using eqn (S),l K , Kc = [H + { (NH 2) zCO>l/ { [H+l to tal - [H +{ (NH2) 2CO>1>{ [urea1 t otal - [H+{ (NH 2) 2CO>l3 with [H+{(NH,),CO)] calculated from the values for K , &, which themselves are determined from the ratio slope/intercept of the linear These values for K2 F2 and K, are also given in table 1. In general, the values for Kc with varying [HCl] at3094 IONIC SOLVATION IN WATER-COSOLVENT MIXTURES constant [urea],,,,, agree well amongst themselves and with KC&l taken from the slope alone at the same [urea],,,,,, except for some erratic behaviour when [H+]total z [H+{(NH,),CO)] makes {[H+],,,,, - [H+{(NH,),CO)]) z 0 in the calculation of K, at the high [urea],,,,,, as found in similar conditions with all other added coso1vents.18 From the agreement of K , with K, &l it is concluded that F, = 1 .O using urea, as found with all other added cosolvents used.1-8 Values for AG(H+),,,, can now be calculated from eqn (4) using the above experimentally determined values for K , and F, with [H+{(NH,),CO}] calculated by using the equation+ [H+((NH,),CO)] = 0.5 { A - (A2 -4[urea],,,,,):) (9) and A = ([ureaItotal + 1 + KE~).(10) [H,O] = (1000 ds- [urea],,,,, Mure,) MG' (1 1) [H,O] is calculated usinglP3 where Mure, is the molecular weight of urea.These values for AG(H+),,,, in ' water-urea are plotted in fig. 1 against the composition of the mixture. For any required composition in the mixture, values for AG,"(H+) on the mole-fraction scale can now be calculated using eqn (4) and ( 5 ) with the value for AG,"(H+), calculated from eqn (2) and the value for AG(H+),,,, interpolated for the same composition from fig. 1 . These values for AG,"(H+) are given in table 2. FREE ENERGIES OF TRANSFER FOR SIMPLE ANIONS E" values have been determined for the cell Pt, H, 1 HX, H,O + urea I AgX, Ag (12) for X-= C1- by Kundu and Mazumdar15 and by Ahmed and SalehlG on the mole-fraction scale, from which AG:(HX) can be computed using the equation AG,"(HX) = 96.5(E", - g ) kJ mol-l. (13) When the latter values are combined with the values for AG,"(H+) determined above in the equation (14) AG:(X-) = AG,"(HX) - AGt(H+) values for AG,"(Cl-) are obtained.Kundu and Mazumdarl' have also used cell (12) with X- = Br- and X- = I- and values for AG,"(Br-) and AG,"(I-) are obtained from these values for E" using eqn (13) and (14). Das and Kundu18 have used cell (1 5) to determine values for pK, for the ionization of water with varying concentrations of urea: Pt, H, I NaOH, NaC1, H,O-urea I AgCl, Ag. (15) Values of AG,"(H+) + AG,"(OH-) have been calculated using these pK, values,l-s and by combining these latter values with the values for AG,"(H+) determined above values for AG,"(OH-) have been derived. Dash and PadhilS have used E" values for the cell Ag, AgCNS I KCNS(C) I 1 KCl(C) I AgC1, Ag (16) after allowing for the liquid junction potential, to determine values for AGf(H+)+ AG:(CNS-), from which values for AG,"(CNS-) can be calculated using the values forG.S. GROVES AND C. F. WELLS ion selective glass electrode I MCl, H,O-urea 3095 AgCl, Ag I I 1 0.0 5 0.10 0.l5 X'2 Fig. 1. Plot of AG(H+),,,, against mole fraction of urea x2 in water-urea mixtures at 25 "C. AG,"(H+) in table 2 and eqn (14). A similar procedure has been applied to the values for AG,"(H+) + AG,"(N;) determined from the E' values of Dash and Padhi for the cell Ag, AgCl I N a C W I I NaN3(C) I AgN3, Ag (17) after allowance for the liquid junction potentials.20 N; and CNS- are collected in table 3. These values for AG,"(X-) on the mole fraction scale for X- = C1-, Br-, I-, OH-, FREE ENERGIES OF TRANSFER FOR SIMPLE CATIONS E" values for the cell for M+ = Li+, Na+, K+, Rb+, Cs+ and NH: have been determined by Pointud and JuillardZ1 and for M+ = Rb+ by Smits et These values have been converted toTable 2.Values of AG:(M+)/kJ rnol-' in water-urea mixtures at 25 "C urea (mole (wt "/d) fraction) H+ Li+ Na+ K+ Rb+ Cs+ NH; Me,N+ Bu,N+ Ph,As+ NO,PhNH,+ 8 3 5.0 5.7 10.0 10.0 10.7 11.5 15.0 15.3 19.4 20.0 20.0 20.3 23.1 25.0 26.5 29.6 30.0 30.0 35.0 36.8 37.5 0.01 55 0.0178 0.0323 0.0323 0.0347 0.0375 0.0503 0.0514 0.0673 0.0698 0.0698 0.07 10 0.0827 0.0909 0.0976 0.122 0.114 0.1 14 0.139 0.149 0.153 - 4.56 -5.10 - 7.8 - - 8.2 - 8.5 - 10.0 - 10.0 - 11.2 -11.4 - -11.4 -11.8 - 12.1 - 12.2 - 12.4 - 12.6 - - 12.9 - 13.1 - 13.1 - - - 4.52' - - - 4.gd - - - -6.3' - -6.1d - - - - 5.7d - 5.8' - - - 5.2d - - - 2.66b -4.39' - 4.30b - 4.46d - - - - - 4.93' - 5.4d - 5.2b - - - - 4.62d - 4.08' - - - 4.25" -3.71b - - -4.31' - - -4.3Id - - - - 5.8' - - 5.6d - - - - 5.0d - 5.0' - - -4.51" - - 2.45" -4.34a - 4.30' - 4.74d - 5.4a - - - - - 5.8' - 5.9' - 5.gd - 5.7" - 5.2d -5.1" - 5.4' -5.1" - 4.68' - - - - - - 4.72' - - -4.82d - - - - 5.5" - 6.9d - - - - - 5.6d - 6.0" - - - 5.1" - - - 2.99 - - - 4.97 - - - 6.0 - 6.5 - - - - 6.6 - -6.5 - - - - - -6.1 - 8.9 a Ref (22). Ref. (24). Ref. (21). Ref. (23).G. S. GROVES AND C. F. WELLS 3097 values for AG,"(MCl) using eqn (1 3). The latter values derived from the data of Pointud and Juillard are on the molality scale (m) and have to be converted to the mole-fraction scale using the equation AG;(MCl) = AG:(MCl),, + 1 1.42 log,, - 18'016 kJ mol-1.K Values for AG,"(M+) are then calculated using the equation AGF(M+) = AG,"(MCl) - AG,"(Cl-) with the values for AG:(Cl-) in table 3 derived from the data of Kundu and Mazumdar.15 The values of AG;(MCl), of Smits et al. on the molar scale (C) are first converted to the mole-fraction scale using the equation 18.016ds ( Msdw kJ mol-l AG,"(MCl) = AG,"(MCl), + 1 1.42 log,, before using eqn (20) and the values of AG,"(Cl-) to derive values for AG,"(Rb+). for M+ = Li+, Na+, K+, Rb+ and Cs+: Das and KunduZ3 have determined the E" values for cell (22) on the molality scale M(Hg) 1 MCl, H,O--urea I AgC1, Ag. (22) The values of AG;(MCl),, derived using eqn (13) are converted to the mole-fraction scale using eqn (19), and eqn (20) is then used with the values for AG,"(Cl-) in table 3 derived from the data of Kundu and Mazumdar to produce values for AG,"(M+).Values for AG,"(NaCl) have also been determined on the mole-fraction scale from measurements of the vapour pressure of NaC1-H,O-urea Values for AGF(Na+) have been calculated from these using eqn (20) and the values of AG,"(Cl-) in table 3 derived using the data of Kundu and Mazumdar. All these values for AG:(M+) on the mole-fraction scale are collected in table 2. FREE ENERGY OF TRANSFER OF OXY-ANIONS AND LARGER IONS Kummel and Ziegel~ki~~ have derived values for AG,"(NH,NO,) on the mole fraction scale from solubility measurements. These are combined in eqn (23) with the values of AG,"(NH,f) in table 2 to produce values for AG,"(NO;): AG,"(NO,) = AGP(NH4N03) - AG,"(NH,S).(23) Values of E", for the cell Ag, AgCl I NaCl(C) I Na, X - Ag, X, Ag I (:jl after correction for the liquid junction potential26 have been used in eqn (1 3) to derive values for xAG,"(H+) + AG,"(X"-) on the molar scale for Xx- = WO:-, PO:- and As0,3-. 'These were then converted to the mole fraction scale using the equation 18.016ds AG,"(H, X) = AG,"(H, X), + 2.303(~+ 1) RT log,, ( M s ~ ~ ) kJ mol-l. (25) Values for AG,"(XS-) were then calculated by subtracting values of xAG,"(H+) obtained from table 2 from AG;(H,X). Kundu and Das2; have derived values for AG,"(KBPh,)c, AG,"(KPic), and AG:(Ph,AsPic), (Pic = picrate anion) from solubility measurements. These values 101 F A R ITable 3.Values of AGy(Xx-)/kJ mol-' in water-urea mixtures at 25 "C 11.5 11.5 14.9 20.3 20.3 27.0 29.6 29.6 36.8 37.0 0.0375 0.0375 0.0500 0.07 10 0.07 10 0.100 0.1 12 0.1 12 0.149 0.150 4.48" 3.70b 5.4a 5.6b 4.70" 4.53b 4.1 1" - - - 4.27 4.09 5.2 4.24 - - - - - - __ - 4.20 3.80 5.2 3.65 - 8.0 14.6 24.2 - - - - - - 5.5 - __ 0.5 19.7 31.5 - - - - - - 4.9 - - 1 .0 18.4 31.0 - - - - - - - - - 3.63 3.10 4.82 3.12 11.2 18.7 31.2 - - 1.6 - - - - - ___ _ _ _ ~ a Using data of ref. ( 1 5). Using data of ref. ( 1 6). 21 .o - - 28.6 - - 29.6 30.5 - - 3.42 5.4 - - 1.27 4.07 3.05 - - -0.43G. S. GROVES AND C. F. WELLS 3099 have been converted to the mole-fraction scale using an equation analogous to eqn (21), and values for AG,"(BPh,) and AG,"(Pic-) have been calculated using the values of AG,"(K+) in table 2 with an equation analogous to eqn (23).These values for AG,"(Pic-) have been used in turn in an equation analogous to eqn (20) to produce values for AG,"(Ph,As+) on the mole-fraction scale from the above values for AG:(Ph,AsPic). Wen and Chen28 have determined values for AG,"(MBr) on the mole-fraction scale for M+ = Me,N+ and (n-Bu),N+ from vapour pressure measurements on H20- urea-MBr mixtures. These values have been used directly with values for AG:(Br-) in table 3 in an equation analogous with eqn (20) to produce values for AG:(Me,N+) and AG,"(Bu,N+) on the mole-fraction scale. From the change in the free energy of dissociation of 4-nitroaniline from water to water-urea mixtures, GAGo(BH+), determined from the acid dissociation constants of 4-nitroaniline, De29 has calculated the free energy of transfer of the salt BH+Cl- from water into water-urea mixtures on the mole-fraction scale using the equation AG,"(BH+Cl-) = AG:(B) + AGt(HC1) -GAG"(BH+).(26) De determined values for AG,"(B) from the solubilities of 4-nitroaniline in water and water-urea mixtures and the values of AG,"(HCl) determined by Kundu and Mazumdar15 were used. From these values of AG:(BH+Cl-), values for AG,"(BH+) have been calculated using an equation analogous to eqn (20) with the values for AG,"(Cl-) in table 3. The values of AG: for the anions are all included in table 3 and those for the cations in table 2. COMPARISON OF AG,"(i) IN WATER-UREA MIXTURES WITH AG,"(i) IN MIXTURES OF WATER WITH COSOLVENTS Despite the difference in the structure-breaking effect and in the changes in the dielectric constant between additions to water of urea on the one hand and of the cosolvents on the other,lP8 the spread in the values of AG,"(i) for individual ions i is broadly similar in all cases, with AG,"(i) positive for i = anions and AG,"(i) negative for i = cations (fig.2). For halides and pseudo-halides, the order with added urea, C1- > Br- > I- z SCN-, is the same as for the coso1vents,1-8 but N; is displaced relatively to a more positive value than found with added DMS0.l. Moreover, NO; with added urea is roughly in the same position, NO; Z I- z SCN-, as found for the C10, ion with added ethanol,2 dioxane1T6 or methano11*4 and the picrate ion is in the same relative order, Pic- < I-, as found with the cosolvents (ethan~nitrile~ and ethanol2).With hydroxylic cosolvents the equilibrium OH- + ROH e RO- + H 2 0 (27) is set up, and the order in AG,"(i) found in non-hydroxylic cosolvents, OH- > C1- > Br- > I- (dioxane,l? DMSO,l9 acetonel? 5, has AG:(OH-) displaced relatively to more negative values as equilibrium (27) lies farther to the right. This order for non-hydroxylic cosolvents, or for hydroxylic cosolvents, where equilibrium (27) lies to the left (ethanol,2 propan-2-01,~9~ t-butyl alcoho11t4), is also found with added urea. The low values for AG,"(NO;) and for AG,"(Pic-) show that the high positive values for AG:(i) for multi-charged oxyanions are not due to the oxy-anionic function or to the anionic size but arise from the higher negative charge.This effect can be compared with the relatively high positive values found for AG,"(ReCli-) with added ethanol.2 101-23 100 IONIC SOLVATION IN WATER-COSOLVENT MIXTURES 30 2 5 20 15 ; 10 E 2 \ W O r c2 u Q s 0 -5 -1 0 -1 5 w 0:- 0.05 h* Ph,As' Fig. 2. Plot of the free energy of transfer of individual ions from water into water-urea against mole fraction of urea x2 at 25 "C.G. S. GROVES AND C. F. WELLS 3101 @ I /- 0.1 0.2 x, 0.3 0.4 Fig. 3. Plot of the free energy of transfer of the 4-nitroanilinium ion from water into water-cosolvent mixtures at 25 "C for (a) ethanonitrile, (6) urea and (c) ethane-1.2-diol as the cosolvent against mole fraction of cosolvent x,. The real contrast of added urea with the added cosolvents lies in the effect of substituent hydrocarbon groups on the relative values of AG,"(i).With added cosolvents (ethan~nitrile,~ ethanol,2 DMSO,'? acetone,'. dioxanel9 6), AG:(BPh,) is large and negative; but with added urea, AG,"(PPh,) is positive and lies in the same range as AG,"(i) for most singly charged anions. Moreover, with added urea, AG,"(i) for i = Me,N+, Bu,N+ or NHZ are grouped close together and close also to the values for the singly charged alkali-metal ions, even though the order found with the other cosolvents (ethanol,, acetone,l* dioxane,'? methanol1* ,), uiz. AG,"(i) becoming more negative along the series alkali-metal ions -= Me,N+ c Bu,N+, but with a greater spread of values, still holds with added urea. These observations suggest that the structure-forming influence of the hydrocarbon groups14 in the ion on AGt(i) is less when the structure-breaking urea is added to water than that when a structure-forming cosolvent is added.AG:(NO,PhNH:) is more negative than AG,"(Bu,N+), as expected for a larger hydrocarbon group in the 4-nitroanilinium cation. De et al.293 30 have also derived values for AG,"(NO,PhNH:Cl-) in water-ethanonitrile and in water-ethane- 1,2-diol mixtures in the same manner as described above for water-urea mixtures. The values for AG,"(NO,PhNHi) calculated using the appropriate values for AG,"(Cl-)l, 39 * are compared in fig. 3 with those for AG,"(NO,PhNH$) in water-urea: as expected from the spread of values of AG,"(i) for other ions in these mixtures,1T3*8 AG,"(NO,PhNH;) has lower negative values in water-ethane- 1,2-diol than in the other mixtures.AG,"(Ph,As+) is more negative with added urea than AGF(i) for the other cations, as found with other cosolvents (ethanol,, dioxane,l? DMSOl? 7, except ethanonitrile.3 However, it is relatively not so low with urea, where AG,"(Ph,As+) M AG,"(H+), as found with the above cosolvents, where - AG,"(Ph,As+) > - AG,"(H+). An alternative3102 IONIC SOLVATION IN WATER-COSOLVENT MIXTURES method of separating free energies of transfer of salts into AG;(i~n)~l assumes that AG;(i) will be the same for the large ions Ph,As+ and BPh; irrespective of the sign of the charge. In the method of separation described here at low mole fractions of cosolvent to restrict solvent sorting to equilibrium (3), significant differences were found for cosolvents ethanol,, DMSO at x, < 0.2,l~ dioxanel? and ethanonitrile3 between - AG,"(Ph,As+) and - AG:(BPh,), with the former greater than the latter.This difference is accentuated with added urea, where AG;(Ph,As+) remains large and negative but AG(BPh,) becomes positive like AG,"(i) for other anions. C. F. Wells, Aust. J. Chem., 1983,36, 1739; Trans. Faraday Soc., 1965,61,2194; 1966,62,2815; 1967, 63, 147. C. F. Wells, J. Chem. Soc., Faradq Trans. 1, 1984, 80, 2445. G. S. Groves and C. F. Wells, J. Chem. Soc., Faraday Trans. I , 1985, 81, in press. C. F. Wells, J. Chem. Soc., Faraday Trans. 1, 1973, 69, 984; 1974, 70, 694; 1976, 72, 601; 1978, 74, 636; Adv. Chem. Ser., 1979, 177, 53. C. F. Wells, Thermochim.Acta, 1982, 53, 67. C. F. Wells, J. Chem. Soc., Faraday Trans. 1, 1978, 74, 1569. ' C. F. Wells, J. Chem. Soc., Faraday Trans. 1, 1981, 77, 1515. C. F. Wells, J. Chem. Soc., Faraday Trans. I , 1975, 71, 1868. J. Wyman, J. Am. Chem. Soc., 1933, 55, 41 16. lo R. H. Stokes, Aust. J. Chem., 1967, 20, 2087. l1 H. S. Frank and F. Franks, J. Chem. Phys., 1968, 48, 4746. l2 J. A. Rupley, J. Phys. Chem., 1964, 68, 2002. l3 G. G. Hammes and P. R. Schimmel, J. Am. Chem. Soc., 1967,89,442; K. Arkawa and N. Takenaka, Bull. Chem. Soc., Jpn, 1967,40,2739; D. V. Beauregard and R. E. Barrett, J. Chem. Phys., 1968,49, 5241. l4 H. S. Frank and M. W. Evans, J. Chem. Phys., 1945, 13, 507; H. S. Frank and W-Y. Wen, Discuss. Faraday Soc., 1957, 24, 133; G. Nemethy and H. A. Sheraga, J. Chem. Phys., 1962, 36, 3382, 3401; G. Nemethy, Angew. Chem., Int. Ed. Engl., 1967,6, 195; W. Laiden and G. Nemethy, J . Phys. Chem., 1970, 74, 3501. l5 K. K. Kundu and K. Mazumdar, J. Chem. Soc., Faraday Trans. I , 1973,69, 806. l6 K. S. Ahmed and J. M. Saleh, Iraqi J. Sci., 1979, 20, 385. l7 K. K. Kundu and K. Mazumdar, J. Chem. Soc., Faraday Trans. I , 1975,71, 1422. la A. K. Das and K. K. Kundu, J. Phys. Chem., 1975,79, 2604; J. Solution Chem., 1976,5,431. l9 U. N. Dash and M. C. Padhi, Thermochim. Acta, 1981, 45, 245. 2o U. N. Dash and M. C. Padhi, J. Electroanal. Chem., 1981, 122, 147. 21 Y. Pointud and J. Juillard, J . Chem. Soc., Faraday Trans. I , 1977, 73, 1048. 22 R. Smits, D. L. Massart, J. Juillard and J-P. Morel, Electrochim. Acta, 1976, 21, 425. 23 A. K. Das and K. K. Kundu, J. Solution Chem., 1976, 5, 436. 24 V. E. Bower and R. A. Robinson, J. Phys. Chem., 1963, 67, 1524. 25 R. Kummel and R. Ziegelski, 2. Chem., 1980, 20, 232. 26 U. N. Dash and M. C. Padhi, Thermochim. Acta, 1982, 55, 315. 2' K. K. Kundu and A. K. Das, J. Solution Chenz., 1979, 8, 259. 28 W.-Y. Wen and C. L. Chen, J . Phys. Chem., 1969, 73, 2895. 29 A. L. De, Electrochim Acta, 1984, 29, 683. 30 A. L. De and T. K. De, Can. J. Chew., 1984, 62, 1776; K. K. Kundu, A. L. De and M. N. Das, J. Chem. Soc., Perkin Trans., 1972, 2063. 31 R. Alexander and A. J. Parker, J. Am. Chem. Soc., 1967,89,5549; 1968,90,3313; 0. Popvych, Crit. Rev. Anal. Chem., 1970, 1, 73; B. G. Cox, C. R. Hedwig, A. J. Parker and D. W. Watts, Aust. J. Chem., 1974,27, 477. (PAPER 5/365)

ISSN:0300-9599    

DOI:10.1039/F19858103091

出版商:RSC    

年代:1985

数据来源: RSC

摘要:

J . Chem. SOC., Faraday Trans. 1, 1985, 81, 3103-3108 Concentration Dependence of Fickian Diffusivity in Solutions and Sorption Systems BY REINHARD GUTSCHE* AND KLAUS FIEDLER Central Institute of Physical Chemistry, Academy of Sciences of the G.D.R., 1 199 Berlin-Adlershof, Rudower Chaussee 5, German Democratic Republic AND JORG KARGER Physics Department of the Karl-Marx-University, Division of Experimental Physics, G.D.R.-701 Leipzig, LinnestraDe 5, German Democratic Republic Received 8th March, 1985 Using transition-state theory a general relation for the concentration dependence of Fickian diffusivity is derived. This relation is simplified by means of assumptions which are expected to reflect the behaviour of real diffusion systems. In this way one obtains two special forms of the general relation which render it possible to calculate the concentration dependence of Fick’s diffusivity from experimental data.Diffusion systems are discussed for which the assumptions leading to the special equations can hold, Experimental examples are given by using diffusivity data from the literature. The concentration dependence D(c) = D(0) d In a/d In c determined by other authors for different diffusion systems is shown to be one of the special equations derived in this paper. By regarding molecular transport as a succession of activated jumps the diffusion process can quantitatively be treated on the basis of transition-state theory, as was realized by Glasstone et al.’ with respect to diffusion in liquid and solid solutions, including surface diffusion.By using transition-state theory the authors found the concentration dependence of Fick’s diffusivity, D(c), to obey the relation D(c) = D(0) d In a/d In c D(0) being the diffusivity for c -+ 0. The symbol a indicates the activity at the concentration c of solute or adsorbate. This relation was also derived by Ruthven and Derrah2 in connection with the application of transition-state theory to zeolitic diffusion. The following considerations may be used to show that the mathematical description of concentration dependence mentioned above is only a special form of a more general relation derivable on the basis of transition-state theory. THEORETICAL DEVELOPMENT By taking the jump of a particle of the solute or adsorbate A from the equilibrium position x to a neighbouring one (x+Ax) as a first-order reaction: Ax --+ Ax+Ax one can write for the gross mole flux from x to x + Axly Here cb(x) is the concentration of moleculqs in the basic state at position x, while F denotes a geometric factor which depends oil the system in which the diffusion process occurs.For diffusion in the lattice of a microporous A-type zeolite, for example, 31033 104 CONCENTRATION DEPENDENCE OF FICKIAN DIFFUSION where K* = a+/ab so that we obtain from eqn (2) F = 116. According to transition-state theory the jump rate k (i.e. the specific reaction rate of diffusion) is given by the relation'. (3) (4) fb and f + denoting the activity coefficients of molecules in th basic state and in the activated state, respectively. K+ is an equilibrium constant defined by eqn (4), where a , is the activity of the diffusing species in its activated state and ab that in the basic state.The symbols k, T and h stand, respectively, for Boltzmann's constant, the temperature and Planck's constant. The gross mole flux from x + A x to x can be formulated analogously, i.e. The difference between the two gross mole fluxes yields the corresponding net diffusion flux (7) J = Jx, X + A ~ - J Z + A ~ , 5 ab(X) ab(x + A-x) kT f (x) f+(x+ALx) h AX ~- (8) =F'(Ax)~- K* * Considering the fact that the transition state for the forward gross flux is the same as for the backward flux, one obtains after the limiting process cs being the concentration of molecules in the activated state. Comparison with Fick's law dc dx J = - D(c) ~ provides the Fickian diffusivity, where J is the flux through unit area normal to the direction of flow and (1 1) kT K + dab D(c) z= F ( A x ) ~ __ ____ - .h f*<c*> dc For c -+ 0 the quantities f+(c,) and da,/dc approach unity. Hence the diffusivity under ideal conditions is given by kT h D(0) = Do = F(Ax)~ - fl From eqn (1 1) and (12) it follows thatR. G~JTSCHF~, K . FIEDLER AND J. KARGER 3 105 With regard to the concentration dependence of Fick’s diffusivity, eqn (1 3) is the most general relation which can be derived by treating the diffusion mechanism on the basis of transition-state theory. Using convenient assumptions this general expression may be simplified to eliminate the activity coefficientf+(c+). As a result of this simplification special forms of eqn (1 3 ) are obtained.Special case 1 : Assuming ideal behaviour of the transition phase one obtains from eqn (1 3 ) the expression (14) dab D(c) = Do -. dc Because of the relationships and respectively, the activity ab becomes identical with a if c z cb (a supposition which is expected to be satisfied in general), i.e. D(c) = D,da/dc. (17) The symbol f i n eqn (1 6) indicates the activity coefficient of diffusing molecules. Special case 2 : Regarding the assumption f + (c+) = fb(cb) one can replace f+ (c+) in eqn (13) by the activity coefficientf;,(c,) = ab/Cb, so that we obtain the expression D(c) = Do d In ab/d In c. D(c) = Do d In a/d In c. (18) (19) With the supposition c % cb used above one obtains finally DISCUSSION According to the main ideas of transition-state theory the general relation (13) is applicable to diffusion processes, the transport mechanism of which can be described by a succession of activated jumps.Such a mechanism can occur in solution as well as in adsorbed phases. However, the problem of the application of eqn (13) is to estimate the activity coefficient f+ (c+). Therefore it is necessary to simplify eqn (1 3) by means of suppositions which are assumed to reflect the behaviour of the real transport process investigated. The supposition of an ideal transition state, leading to eqn (1 7), is fulfilled if there are no interactions between the particle in the transition state and the surrounding molecules. Such behaviour can be thought to exist with reference to diffusion processes in adsorbed phases, where the particle is in a sterically separate state owing to the structure of the adsorbent.If the diameter of the admolecule is of the same magnitude as that of the window between two neighbouring cavities in the zeolite lattice (as is frequently the case in A-type zeolites), the particle, while in the window, can represent a sterically separate state. Hence, in the case of diffusion in the micropores of an A-type zeolite crystal a description of the concentration dependence of Fick’s diffusivity by means of eqn (1 7) would be possible. While the model mechanism considering an ideal transition state has been confirmed in the case of self-diffusion of light paraffins in NaCaA ~eolites,~ the corresponding experimental evidence for the validity of eqn ( I 7) for (non-equilibrium) diffusion is not yet available.In many transport systems such a sterically separate state does not exist, as in solutions or in the case of surface diffusion in pores with a diameter essentially greater3 106 CONCENTRATION DEPENDENCE OF FICKIAN DIFFUSION Table 1. Interpretation of diffusion data (in cm2 s-l) for the system Ar/Graphon7 on the basis of eqn (20) T = 9 0 K T = 77.6 K clmmol g-l D(c) D(c) d In c/d In ceq D(c) D(c) d In c/d In ceq 0.446 0.625 0.714 0.803 0.848 0.892 0.937 0.959 0.982 1.026 1.071 1.115 1.16 1.249 1.4 2.3 3.4 7.9 19.4 34.4 34.4 - __ 18.2 13.1 11.6 - 1.5 1.2 1 .o 1.3 1.9 2.3 3.1 - - 3.1 3.9 4.8 - 0.4 1 .o 2.4 5.2 5.9 8.2 6.6 5.8 5.1 3.8 3.5 - 0.2 0.4 0.6 0.6 0.4 0.7 0.4 0.7 0.6 0.5 0.8 - - - than the diameter of the admolecules (X-type zeolites) and at plane interfaces.Supposing that in such systems both the particle in the transition state and that in the basic state are exposed, because of their small diameters, to the same interactions, i.e.f+(c+) %fb(cb), one can assume the validity of eqn (19). In ref. (1) eqn (19) is derived in an apparently general way, i.e. without the assumptions stated here. The fallacy with reference to the general character of the derivation given in ref. (1) becomes obvious at the point at which the activity coefficientf+(x) in the formula representing the gross diffusion flux is replaced by the expressionf,(x) +&Ax dfb(x)/dx. Such a statement supposes the requirement Ax -+ 0 and consequently the indentity of the variation in the activity coefficient with x for the transition state and the basic state.Therefore, the derivation given in ref. (1) also includes the assumption f+(c,) =fb(cb). In the general case the profiles of the activity coefficients mentioned above can differ considerably because the transition state and the basic state represent thermodynamically separate phases. There are several experimental proofs of the availability of eqn (19). With respect to diffusion in liquid solutions the validity of this relation is already confirmed by Glasstone et al.' As far as surface diffusion is concerned articles have been published which can be used to corroborate the applicability of eqn (19). On the one hand Pope6 studied the diffusion of SO, over the surface of Spheron 6(2700) and pointed out that the evaluation of D(c) d In c/d In a yields approximately constant values within the range ofdiffusivitiesmeasured. However, Pope calculated the values ofD(c) d In c/d In a not in the light of the concentration dependence of Fickian diffusivity, but in view of a comparison of diffusivity and self-diffusivity.On the other hand Ash et a[.' measured concentration-dependent surface diffusivities for Ar and SF, adsorbed on Graphon. The authors found a similar variation of the reciprocal slope of the sorption isotherm and the surface diffusivity as a function of the amount adsorbed. A quantitative description of the relationship between the sorption isotherm and surface diffusivity becomes possible by using the relation D(c) = Do d In ce,/d In c (20)Table 2.Interpretation of diffusion data (in cm2 s-l) for the system SF,/Graphon7 on the basis of eqn (20) T = 253 K T = 233 K T = 223 K T = 213 K T = 193 K T = 183 K ____- 0 0.045 0.089 0.134 0.178 0.223 0.268 0.312 0.357 0.402 0.424 0.446 0.468 0.49 1 0.513 0.535 0.558 0.58 0.669 0.77 0.848 6.95 6.25 5.96 6.0 6.0 6.14 6.56 9.21 16.99 - - - - - - - - - - - - 6.95 7.4 7.3 7.2 6.8 5.3 5.3 6.3 6.6 - - - - - - - - - - - - 5.43 3.73 3.22 3.22 3.57 4.16 4.7 5.51 6.74 15.07 32.5 1 - - - - - - - - - - 5.43 4.3 4.2 3.9 4.3 4.8 3.9 3.5 3.0 3.2 6.2 - - - - - - - - - - 4.07 2.67 2.1 1 2.05 2.2 2.43 2.83 3.07 4.56 8.21 10.18 16.64 24.95 24.3 1 21.55 15.05 - - - - - 4.07 3.0 2.7 2.9 2.8 2.9 3.0 2.7 2.3 2.5 2.4 2.7 3.9 3.7 3.6 3.0 - __ - - - 1.8 1.58 1.34 1.29 1.34 1.86 2.46 3.18 4.1 5.82 7.05 9.57 16.93 23.7 18.96 14.37 12.99 - - - - 1.8 1.7 1.6 1.5 1.5 1.8 2.1 2.5 2.2 1.6 1.7 1.3 2.5 2.6 2.8 2.6 3.5 - - - - ~ - - - - - 0.49 - - - - 6.18 12.64 9.03 5.18 1.84 0.49 - - - - - - - - - - - - - 1.04 2.45 - - 10.9 7.48 2.63 1.74 0.58 0.17 - -3108 CONCENTRATION DEPENDENCE OF FICKIAN DIFFUSION with ceq denoting the equilibrium concentration of the species in the non-adsorbed phase.For the case of an ideal non-adsorbed phase eqn (20) follows directly from eqn (19). Analysing the experimental data determined by Ash et aL7 on the basis of eqn (20) one obtains the results summarized in tables 1 and 2. Constant values of D(c) d In c/d In eeq ~ o u l d result if eqn (20) were obeyed; this is approximately fulfilled. We thank Prof. R. M. Barrer and Dr C. G. Pope for valuable comments and for making available the experimental data, respectively. Helpful discussions with Prof. H. Pfeifer are gratefully acknowledged. S. Glasstone, K. Laidler and H. Eyring, The Theory of Rate Processes (McGraw-Hill, New York. 1941), p. 516. D. M. Ruthven and R. I . Derrah, J. Chem. Soc., Furaday Trans. I , 1972, 68. 2332. M. Biilow and J. Karger. Thesis (Karl-Marx-University, Leipzig, 1978). H. Eyring, S. H. Lin and S. M. Lin, Basic Chemical Kinetics (Wiley, New York, 1980), p. 125. J. Karger, H. Pfeifer and R. Haberlandt, J. Chem. Soc., Faraday Truns. I , 1980, 76, 1569. C. G. Pope, Trans. Faraday SOC., 1967, 63, 734. 'I R. Ash, R. W. Baker and R. M. Barrer, Proc. R. Soc. London, Ser. A , 1967, 299, 434. (PAPER 5/397)

ISSN:0300-9599    

DOI:10.1039/F19858103103

出版商:RSC    

年代:1985

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. 1, 1985, 81, 3109-3116 Enthalpies of Adsorption of Non-ionic Surfactants from Aqueous Solutions on to Silica BY ANDREA GELLAN AND COLIN H. ROCHESTER* Department of Chemistry, The University, Dundee DDl 4HN Received 18th March, 1985 Enthalpies of adsorption of three n-alkylpolyethylene glycol non-ionic surfactants (C,,E,, C,,E, and C,E,) on silica immersed in aqueous solutions have been determined by calorimetry. Consecutive stages in the adsorption process are distinguishable by marked changes in the differential enthalpy of adsorption. The results are discussed in terms of the orientation of the surfactant molecules at the silica/water interface as a function of increasing surface coverage. Sudden changes in the enthalpy of adsorption of C,,E, at saturation monolayes coverage and saturation bilayer coverage of the surface suggested that structural reorganisation akin to phase change was occurring in the interfacial region.Isotherms for the adsorption of these 0-n-alkylpolyethylene glycols, C,E,, C12E5 and CI2Es [C,H,,+,(OCH,CH,),OH, designated C,E,], from aqueous solutions on to graphitised carbon'? and silica3 exemplify the significant differences which exist4 between the adsorption behaviour of non-ionic surfactants on non-polar and polar adsorbents. Enthalpy changes accompanying the adsorption of the three surfactants on carbon have been measured112 and gave clearer indications of changes in adsorption behaviour as a function of surface coverage than the Gibbs free energy (isotherm) data. The isotherms, in conjunction with contact angle and X-ray diffraction data, for silica immersed in aqueous solutions of the surfactants enabled, particularly for C12E5, a series of consecutive stages of adsorption to be distinguished.3 A model of adsorption consistent with the stepped isotherm for C12E, invoked C,,E5 molecules lying parallel to the silica surface at low equilibrium solution concentrations, molecules linked by hydrogen-bonding interactions between silanol groups and ethyleneoxy chains in a looped configuration at higher concentrations with the hydrocarbon chains lifted from the surface, and adsorption in a bilayer at concen- trations above the critical micelle concentration (c.m.c.).The present paper reports enthalpies of adsorption for C,E,, C12E5 and C,,E, on silica in water which were determined in order to further test the validity of the model of adsorption and to attempt to distinguish additional features not apparent from the isotherm data.There have been no previous calorimetric studies of these systems and we are not aware of any similar studies of the adsorption of other non-ionic surfactants on silica immersed in water, EXPERIMENTAL Samples of 0-n-octyltetraethylene glycol (C,E,), 0-n-dodecylpentaethylene glycol (C,,E,) and 0-n-dodecyloctaethylene glycol (C,,E,) were used as supplied by Nikko Chemical Co. (Tokyo) and stored under nitrogen at ca. 0 "C. Silica (type TK900, Degussa) had a surface area (nitrogen B.E.T.) of 148.6+ 1.1 m* g-l. Water was deionised prior to triple distillation under nitrogen, once from alkaline potassium permanganate and twice from itself.The method of 310931 10 SURFACTANTS ON SILICA I I I I I I -B I I I I I 0 L6 - 0- 0 0 1 I I I I 2 4 6 8 10 amount ad~orbed/lO-~ mol g-' Fig. 1. Relationships between the amount of C,,E, adsorbed on silica at 298 K and (a) heats of adsorption and (6) equilibrium solution concentrations. measurement of enthalpies of immersion of silica in aqueous surfactant solutions, using an LKB 8700-2 reaction calorimeter with the thermostat bath at 25.00+0.05 "C, was the same as that used in parallel studies involving graphitised carbon as adsorbate.', RESULTS The calorimetrically determined enthalpies of immersion Qimm of silica in aqueous solutions of C,E,, C,,E, and C,,E, were converted to integral enthalpies of adsorption Qads by correction for enthalpies of dilution and dimicellisation in accordance with (1) the equation where n is the residual number of moles of surfactant in solution after adsorption and AHi and A P f are molar enthalpies of of a concentrated solution of surfactant to the initial and final surfactant concentrations, respectively, in the adsorption experiment.The results are shown in fig. 1-3 as plots of - (Qads - Q,), where Q, is the enthalpy of immersion of silica in water, as a function of the number of moles of surfactant adsorbed per gram of silica. Slopes of the plots give direct information about differential enthalpies of adsorption as a function of surface coverage. The previously determined adsorption isotherms" are included in fig.1-3 as plots of equilibrium solution concentration against amount adsorbed in order to provide an easily recognisable correlation between the enthalpy and isotherm data. In each figure curve (a) refers to the left hand ordinate axis only and curve (6) refers to the right hand ordinate axis only. The link between the isotherm and enthalpy data is provided by the abscissa axis which is common to both curves. Qads = Qimm -4ARf-ARi)A. GELLAN AND C. H. ROCHESTER 3111 amount adsorbed/ lop4 mol g-' Fig. 2. Relationships between the amount of C,,E, adsorbed on silica at 298 K and (a) heats of adsorption and (b) equilibrium solution concentrations. amount adsorbed/ lo-' mol g-' Fig. 3. Relationships between the amount of C,E, adsorbed on silica at 298 K and (a) heats of adsorption and (b) equilibrium solution concentrations.The enthalpy data for C,,E, suggest that there are six separate stages in the overall adsorption process which are distinguishable by significant changes in slope of the plot of - (Qads - Q,) against the amount of surfactant adsorbed n,. Differential enthalpies of adsorption AH2 were deduced from the slopes of the graph in accordance with the equation31 12 SURFACTANTS ON SILICA Table 1. Differential enthalpies of adsorption of C,,E, on silica immersed in water at 298 K ~ _ _ _ _ _ _ ~ - - _ _ _ _ ~ - ~ ~ c2 n2 Ai7, l-2 A /lo-& mol dm-3 mol g-l /kJ mol-' /10-lo rnol cm-2 /lop2 nm2 -. - __ 1.1-3.3 0.3-1.2 - 120 0.19-0.83 88 1-20 1 3.3-3.8 1.2-4.2 - 1.3 0.83-2.8 201-59 3.9-4.4 4.9 highly exothermic 3.3 50 4.4-5.0 4.9- 1 0.0 - 1.9 3.3-6.7 50-25 5.0 10.0 highly endothermic 6.7 25 5.0-8.4 10.0-1 1.4 small 6.7-7.7 25-22 The sections of the enthalpy plot drawn as a full line in fig.1 allowed reliable estimates of AR2 to be evaluated for three of the distinguishable stages of adsorption. The remaining sections, drawn as dashed lines, were less well defined and corresponded to one highly exothermic and only highly endothermic stage of adsorption each occurring over a very narrow range of surface concentration, and a final stage at plateau surface concentration with a small value of A p 2 - The AET2 results are given in table 1 which includes the ranges of equilibrium surfactant concentration in solution (c2), amount of surfactant adsorbed (n2), excess surface concentration of surfactant (r2), and area A occupied by each adsorbed surfactant molecule for which each AH value is applicable. The enthalpy results for C,,E, (fig.2) resembled those for C12E5 in that a series of distinct changes in the differential enthalpy of adsorption occurred with increasing surface concentration of surfactant. The AH2 value was highly exothermic at low solution and surface concentrations (c2 < 0.5 x rnol dm-", T2 < 0.03 x 10-lo mol cmP3) and was highlyendothennicat equilibriumconcentrations corresponding to the adsorption plateau [c, = ( 2 1-17) x lop5 mol d m 3 , r2 = (2.0- 2.2) x 10-lo mol ~ m - ~ ] . At intermediate concentrations a section of the plot (fig. 2) corresponding to a positive AH2 value (c, z 1.7 x mol dm-3, T2 M 0.2 x mol cm-2)wasfollowedbyastraightline,givingA172 = -85 kJ rnol-l, over a fairly wide range of equilibrium solution concentrations [c, = (3.2- 1 1.2) x lo+ mol dm-3] and surface excess concentrations of surfactant [I?, = (0.4-2.0) x 10--lo mol cm-2].The adsorption of C,E, was only briefly investigated. However, the integral enthalpies of adsorption (fig. 3) again showed the existence of a series of consecutive stages of adsorption corresponding to different differential heats of adsorption. A highly exothermic AH2 value at low surface concentrations paralleled the results for C12E5 and C12E8, but at plateau adsorption the exothermic A S 2 for C,E, contrasted with endothermic AR2 values for C,,E5 and C,,E,. DISCUSSION The differential enthalpy of adsorption ( - 120 kJ mol-l) of C12E5 on silica at law surface concentrations is compatible with a model of adsorption3 in which the Cl2E5 molecules lie parallel to the solid/liquid interface with both their hydrocarbon chains and the hydrophilic oxyethylene groups interacting with the silica surface.Thermodynamic data for alkane monolayers on water and for liquid alkanes in contact with water have led to the conclusion that alkane molecules oriented parallel. to an alkane/water interface do not cause appreciable structuring of water in theA. GELLAN AND C. H. ROCHESTER 31 13 interfacial r e g i ~ n . ~ The corollary in the present context is that the hydrophobic effect of C,, groups on water structure in bulk solution disappears when C12E5 adsorbs with the C,, chains parallel to the silica/water interface.The resulting contribution to the entropy of adsorption is positive6 and therefore favours the adsorption process. The enthalpy change associated with the removal of alkyl chains from water will be negati~e,~-~ which also favours adsorption and partially accounts for the overall exothermicity of AR2. The An, value for adsorption of C,,E5 in the parallel orientation3 on silica was more exothermic than the (albeit less well defined) corresponding AH2 value for C12E5 on graphitised carbon.2 An additional exothermic component of AH, for silica adsorbent arises from the formation of hydrogen bonds between surface silanol groups and oxygen atoms in ethyleneoxy groups of adsorbed m0lecules.~9 It has been argued from isotherm data3 that ethyleneoxy chains are adsorbed in a looped configuration with alternate oxygen atoms involved in specific hydrogen-bonding interactions with silanol groups, and non-interacting oxygen atoms retaining their water of solvation.This mode! is consistent with the enthalpy data since it precludes a major endothermic contribution7? to AH, arising from the dehydration of ethyleneoxy chains. The availability of two lone pairs of electrons at each oxygen atom in surfactant molecules allows two hydrogen bonds to be formed simultaneously; each oxygen atom interacting with either two silanol groups or one silanol group and one water molecule as hydrogen-bond donors. The latter possibility allows some hydration to be retained even by oxygen atoms linked to surface silanol groups by hydrogen bonds.The adsorption of C,&, at low surface concentrations on silica is therefore accompanied by a significant diminution in solvation effects involving the hydrophobic hydrocarbon chain. However, solute-solvent interactions involving the hydrophilic ethyleneoxy chain and hydroxy end group are largely retained in the adsorbed state. The second stage in the adsorption of C,,E, on silica corresponds to a steep rise in the isotherm. This has been ascribed3 to the displacement of hydrophobic C,, groups from direct interaction with the silica surface by hydrophilic groups of incoming molecules to the interfacial layer. The hydrocarbon groups in adsorbed molecules increasingly protrude from the silica surface into the aqueous phase. The fact that the AR2 value (- 1.3 kJ mol-') for this stage is considerably less exothermic than for the first stage is compatible with the model of adsorption and could arise for a variety of reasons.The displacement of hydrocarbon chains from a configuration parallel to the surface would make endothermic contributions to AR, because of the breaking of surface-adsorbate dispersion forces and possibly also because interactions between hydrocarbon chains and ~ a t e r ~ - ~ are partially revived. The enthalpy of adsorption of the incoming molecule would, in addition, be less exothermic than if it were adsorbed in a wholly parallel configuration. This is because the C,, chain does not itself become involved in dispersive interactions with silica and the ex other mi^)^-^ removal of the C , , chain from the aqueous phase is less complete.A further endothermic7$ l4 contribution to An, might arise because C,, chains protruding into the aqueous phase impede solute-solvent interactions involving adsorbed hydrophilic chains in their looped configuration. The sudden change in Qads of ca. -2.3 J g-l (which occurred at the step in the isotherm for which the amount of C12E5 adsorbed was approximately constant at cu. 4.9 x 10- mol g -l) is unlikely to be explicable in terms of a highly exothermic enthalpy of adsorption for the final few molecules necessary to attain saturation of the surface at the step and negligible enthalpy effects associated with the molecules already present on the surface close to saturation. It would be more plausible to suggest that the final molecules induce a sudden change in the configuration of all the molecules existing31 14 SURFACTANTS ON SILICA on the surface at the step.The effect could be likened to a phase change, possibly from a disordered phase to an ordered phase, in the adsorbed layer with the surface concentration of C1zE5 approximately constant. The difference between the integral enthalpies of adsorption at the beginning and end of the step was -4.7 kJ mol-l which may be taken as a measure of the enthalpy change accompanying reorganisation in the surface layer. The data are also consistent with a significant decrease in the entropy per mole of adsorbed C12E5 when reorganisation occurs in accordance with the suggestion that a sudden ordering process is involved.The complexity of the interfacial region precludes an unambiguous assignment of the dominant source of the enthalpy and entropy effects. The conjectured final state after reorganisation is an ordered layer of C,,E, molecules with the hydrocarbon chains protruding into the water phase. The hydrophilic chains are adsorbed in a looped configuration in which no more than half of the oxygen atoms in each chain3 are linked to the surface via specific hydrogen-bonding interactions with surface silanol groups. The second steep rise in adsorption of ClzE, by silica has been attributed to bilayer f~rmation.~ The differential enthalpy of adsorption (- 1.9 J mol-l) compares with values of 9.9 and 0.6 kJ mo1-I for the enthalpies of micellisation and adsorption at the water/air interface, respectively, for C,,E, at 298 K.14 Rosen et a l l 4 argued that endothermic contributions to AH arising from the breaking of interactions between water molecules and ethyleneoxy segments of hydrophilic chains were greater in the micellisation process than when C12E5 was transferred from aqueous solution to an adsorbed monolayer at the water/air interface.The build-up of the second layer of a bilayer on silica constitutes a similar process to the formation of a monolayer at the water/air interface. Therefore, the similarity between AHz (table 1) and AHWlAl4 is further evidence for bilayer f ~ r m a t i o n . ~ Further encouragement for the validity of the comparison is provided by the similar limiting areas occupied per adsorbed molecule at the water/air interface2 and in the second layer of the bilayer on silica (table 1 ; 0.25 nm2 is the average area of silica occupied by each C,,E, molecule taking into account all the C,,E5 molecules in both layers of the bilayer, therefore each C,,E, molecule occupies 0.50 nm2 in a single layer).The value of AH2 differs from A&,,A in the exothermic direction. Although the reported difference is possibly too small to justify comment, one explanation would be that an additional exothermic contribution to AHz arises from weak dispersion forces between the extremities of the hydrocarbon chains at the junction between the two layers in the bilayer. The value of AHz recorded in table 1 as highly endothermic must correspond to the onset of structural reorganisation involving either the entire interfacial region or one or other of the individual layers in the bilayer.The difference between the integral enthalpies of adsorption at the beginning and end of the step in Qads was 5.6 kJ mol-1 which, as with the exothermic step in Qads at lower surface coverage, is a measure of the enthalpy change (per mole of total surfactant adsorbed) accompanying reorganisation in the surface layer. One interpretation of the data is that they support the proposal15 that ellipsoidal associates akin to micelles are formed when C,E, surfactants are adsorbed on silica. The endothermic step would be attributable to reorganisation in the interfacial region from a bilayer structure to a two-dimensional array of adsorbed ellipsoidal micellar associates.Detailed analysis of the factors which might contribute to the endothermic change would be speculative, although it is relevant to note that the transfer of ClzE, from an ordered layer at the water/air interface to a micellar aggregate in solution is also endothermic (AHmip-AHw,A = 9.9-0.6 = 9.3 kJ molk1).l4 The analogy is not ideal, but is pro- bably the best currently available. If the step in Qads is to be interpreted in this way then the adsorption of C12E5 monomer on silica forms adsorbed micellar associates at an equilibrium Cl2E, concentration (table 1) below the c.m.c.2T14 Despite theA. GELLAN AND C . H. ROCHESTER 31 15 apparent attraction of this model in explaining the enthalpy data, the concept of adsorbed micellar aggregates is not supported by X-ray diffraction results for adsorbed C,,E, on silica at high coverage^.^ The X-ray data support a bilayer model of adsorption at maximum surface coverage.The endothermic step in Qads occurs at a mean-surface-area occupied per adsorbed molecule (0.25 nm2) which is exactly half the area per molecule (table 1) when monolayer coverage of the surface is complete. At monolayer coverage the area is dominated by the size of the ethyleneoxy groups and the nature of their bonding to surface silanol groups. Hydrocarbon groups protruding from the surface are either not close-packed, if they extend at right angles from the surface, or lie in configurations at smaller angles to the surface. In contrast, the area of 0.25 nm2 would be consistent with a structure in the interfacial region in which hydrocarbon chains were close packed at right angles to the solid/liquid interface.A model which achieves this [model Vb in ref. (3)] would involve a bilayer, but with the hydrocarbon chains radiating out from the surface in the first layer and in towards the surface in the second layer. The layers interdigitate to form a vertically orientated layer of close-packed hydrocarbon chains between a layer of ethyleneoxy chains interacting with the surface and a layer of ethyleneoxy chains directed into the aqueous phase. It is proposed that the formation of this structure suddenly becomes favourable when the surface populations of C,,E, in the two layers of the bilayer become identical. Bilayer formation (in the adsorption range for which AH, = - 1.9 kJ mol-l) initially gives a less well orientated structure possibly involving little interdigitation between hydrocarbon chains from the two layers.The endothermic enthalpy change accompanying the transition from the less well orientated to the interdigitated bilayer is ascribed primarily to the total loss of water molecules solvating the oxygen atoms in ethyleneoxy chains adjacent to the silica surface. A reduction in hydration of the outermost ethyleneoxy chains might also accompany the interdigitation process. In general, the modified bilayer model seems more favourable than the micellar aggregate model, although it is not easy to distinguish between conflicting models of this type.4 The highly exothermic AH2 for C&, at low coverages suggests that adsorbed ClZE8 molecules lie parallel to the silica surface with ethyleneoxy groups involved in hydrogen-bonding interactions with silanol group, and C,, alkyl groups not involved in the hydrophobic hydration effect.The endothermic differential enthalpy of adsorption at slightly higher coverages showed that a change had occurred in the mode of adsorption despite the fact that the end of the endothermic range coincided with a mean-surface-area occupied per adsorbed molecule which could still have been explicable in terms of an orientation parallel to the surface. The integral enthalpy of adsorption at this stage was similar to the integral enthalpy of adsorption of C12E5 during the first (parallel orientation) stage of adsorption. A proposal consistent with the data for C,,E, would be that at very low coverages favourably disposed surface silanol groups exist for the multi-attachment of C$, via hydrogen b o d s to silica, but that at higher coverages some of the hydrogen bonds to each molecule are broken and the free silanol groups form new hydrogen bonds with incoming C,,E, molecules. The ethyleneoxy chains are therefore partially lifted from the surface.A corollary of the high concentration of solvated hydrophilic groups in the interfacial layer might be to cause partial desorption of hydrocarbon groups. The resulting endothermic contributions to the enthalpy change arise from loss of the dispersive forces between the surface and segments of the C,, chain and re-establishment of interactions between water and the hydrocarbon segments.’~~ A key question3 in relation to the adsorption of C,,E, on silica at 298 K is whether a monolayer or a bilayer of C,,E, is formed at an equilibrium solution concentration of 1.I x mol dmP3, which is ca. 10% above the c.rn.c., Two alternatives exist for31 16 SURFACTANTS ON SILICA a mon01ayer;~~~ one with the C,, groups interacting with silica and the E, groups at the boundary with the aqueous layer, and the other with ethyleneoxy segments adsorbed and the alkyl chains protruding towards the aqueous phase. The former is not consistent with either the exothermic integral enthalpy of adsorption for complete surface coverage (ca. - 94 kJ mol-l) or the exothermic differential enthalpy of adsorption as complete surface coverage is approached (- 85 kJ mol-l).These figures contrast with endothermic enthalpy changes for the micellisation of ClzE, and for the adsorption of C,,E, at the air/water interface ;14 both processes in which association of hydrocarbon chains occurs and hydrophilic groups remain on contact with the aqueous phase. Similarly, it is unlikely that rearrangement of adsorbed species induced by an increase in surface concentration of C,,E, would give a highly exothermic AH, value if ellipsoidal micellar aggregates15 of C,,E, were being formed in the adsorbed state. The second monolayer model is difficult to rationalise in terms of previous X-ray diffraction and contact angle data.3 However, it would be consistent with the AH, data, particularly if there were a structuring effect involving hydrated ethyleneoxy chainP in the adsorbed layer.The relative lengths of the C, and Em chains are clearly important in influencing the mode of adsorption of C,Em non-ionic surfactants. The effects of the ethyleneoxy chains, which will be twice the length of dodecyl groups, dominate for C,,E, and possibly lead to an interfacial region in which water and hydrophilic groups have a mutual structuring effect through hydrogen-bonding interactions. At equibrium solution concentrations around the c.m.c., this structuring prevents the close-packed association of alkyl groups which apparently occurs in the case of C,,E,. Gathering further enthalpy data for C,E, surfactants on silica as a function of both n and rn would be desirable to test these hypotheses. It could be that the endothermic AH2 value for C,,E, at ca.(2.0-2.2) x mol cm-, reflects the onset of bilayer adsorption. Isotherms for C,E, and C,,E, on silica have been interpreted in terms of similar models for ad~orption.~ The enthalpy data are consistent with the conclusion that C,E, resembles C,,E, in its adsorption behaviour. Bilayer formation for C,E, is apparently accompanied by an endothermic differential enthalpy of adsorption. However, in contrast to the results for C,,E5 the marked change in AH, as the saturated bilayer is attained is in the exothermic direction (fig. 3). We thank the S.E.R.C. for the award of a CASE studentship (A.G.) and Unilever Research for collaboration. M. J. Hey, J. W. MacTaggart and C. H. Rochester, f. Chem. SOC., Faraduy Trans. I , 1984, 80, 699. A. Gellan and C. H. Rochester, J . Chem. SOC., Faraday Trans. I , 1985, 81, 1503. A. Gellan and C. H. Rochester, J . Chem. Soc., Faradaj) Trans. I , 1985, 81, 2235, J. S. Clunie and B. T. Ingram, in Adsorption from Solution at the SolidlLiquid Interface, ed. G. D. Parfitt and C. H. Rochester (Academic Press, London, 1983), p. 105. R. Aveyard, in Surfactants, ed. Th. F. Tadros (Academic Press, London, 1984), p. 153. A. Couper, in Swfuctants, ed. Th. F. Tadros (Academic Press, London, 19841, p. 13. J. M. Corkill, J. F. Goodman and J. R. Tate, Trans. Faraday SOC., 1967, 63, 773. R. Aveyard and R. W. Mitchell, Trans. Faraday Soc., 1968, 64, 1757. M. Ueno, Y. Takasawa, Y. Tabata, T. Sawamura, N. Kawahashi and K . Meguro, J. Jpn Oil Chem. Soc., 1981, 30, 421. lo F. Wolf and S . Wurster, Tenside Deterg., 1970, 7, 140. l 1 N. A. Klimenko and A. M. Koganovskii, Colloid f., 1973, 35, 718. la E. Kokufuta, S. Pujii, Y. Hirai and I. Nakamura, Polymer, 1982, 23, 452. l3 H. P. Seng and P. J. Sell, Tenside Deterg., 1977, 14, 4. l4 M. J. Rosen, A. W. Cohen, M. Dahanayake and Xi-yuan Hua, J. Phys. Chem., 1982,86, 541. l 5 N. A. Klimenko and A. M. Koganovskii, Colloid J.. 1974, 36, 135. l6 R. Heusch, Ber. Bunsenges. Phys. Chem., 1979, 83, 834. (PAPER 5/452)

ISSN:0300-9599    

DOI:10.1039/F19858103109

出版商:RSC    

年代:1985

数据来源: RSC

摘要:

J . Chem. SOC., Furaday Trans. I . 1985,81, 3117-3125 Rutile Growth at the Surface of TiO, Films Deposited by Vapour-phase Decomposition of Isopropyl Titanate BY Y ASUTAKA TAKAHASHI,* HARUHISA SUZUKI AND MINEJI NASU Department of Synthetic Chemistry, Faculty of Engineering, Gifu University, Yanagido, Gifu, 501-1 1 Japan Received 19th March, 1985 Isopropyl titanate has been decomposed in the vapour phase to give an anatase film on glass substrate under conditions of slow deposition. It was found that rutile grew on the film surface as the films become thicker than ca. 10pm and when film growth was rapid at substrate temperatures between 450 and 500 "C. In an extreme case the film surface became pure rutile on top of anatase crystals. The occurrence of this phenomenon is explained by the rapid epitaxial growth of dendritic rutile crystals on the rutile crystallites, which protrude over the columnar anatase crystals and suppress their growth.A similar mechanism can be assigned to TiO, deposition from the other alkyl titanates, especially ethyl titanate, which generates more rutile nuclei than isopropyl titanate in vapour-phase decomposition and so can provide rutile films as thin as 0.5 pm. In the chemical vapour deposition of oxide films with several crystal modifications, it is very important whether control of the crystal modifications is possible, because the physical properties of the modification are different and because the morphology of the film depends on the crystal structure. In the case of titanium dioxide, three modifications (anatase, brookite and rutile) are known.In our previous paper,l concerned with the deposition of titanium dioxide films, it was reported that rutile or anatase films can be obtained selectively using ethyl titanate, decomposed under reduced pressure, as the starting material for vapour deposition. Further studies2* made clear that modifications of TiO, crystals formed on a glass substrate depended on the structure of the titanate. Isopropyl titanate was found to give anatase films under the conditions examined, unlike other titanate complexes such as ethyl and t-butyl titanates: the latter gave brookite films.3 The crystal modification seemed to be controlled kinetically rather than thermodynamically. In order to clarify the origin of the control, some kinetic study was performed using isopropyl titanate as a representative compound.In the course of this experiment, it was found that rutile films could be obtained when films thicker than 10pm were grown at moderate film-growth rate and at suitable substrate temperatures. This paper reports the results of our investigation. EXPERIMENTAL Commercial extra-pure isopropyl titanate (Wako Pure Chemicals Co. Ltd) was used following vacuum distillation. The N, gas used for transport of the complex was purified by passing it over titanium sponges heated to =- 800 "C (deoxygenation) and subsequently over phosphorus pentoxide (dehydration). The 0, gas was dried by passing it over the phosphorus pentoxide bed. The vertical furnace used for the decomposition reaction was a 25 mm inside diameter quartz tube.The substrate (glass, Corning Co. Ltd, no. 7059) was mounted on an 31 173118 RUTILE GROWTH AT THE SURFACE OF TiO, 0 400 4 5 0 500 550 600 substrate temperature/'C Fig. 1. Temperature dependence of TiO, deposition on a glass substrate: reaction time, 30 min; complex saturation temperature, 50 "C; carrier N,, 0.5 cm3 s-l; 0,, 0.5 cm3 s-l; bypass N,, 1.0 cm3 s-l; pressure, 600 Pa. inner quartz plate which was rotated (2&30 r.p.m.) and it was heated from the outside by a Nichrome heater. The temperature of the substrate was estimated from a relationship, determined in advance, between the temperatures of the heater and that of the inner quartz plate. Sometimes Sn0,-coated glass was used as the substrate. It was prepared by spraying on a SnCl, + SbCl, + HCI aqueous solution at 500 "C.The crystal structure of the films obtained was determined by X-ray diffraction (Cu K,, Rigaku Ltd, model 2024) in the same way as reported previously.' In this study the anatase crystals were highly oriented to (220) face and the diffraction peak assigned to the anatase structure was the only one present, even in a mixed deposition state of anatase and rutile crystals. Therefore, the rutile content estimated from the intensity ratio of diffraction peaks assigned to rutile to the total peak height may not be correct. Scanning electron microscopy (Akashi EMAX-8000s) of the surface and fractured cross-section of the films was performed in order to examine the growth morphologies. Transmission electron microscopy (Hitachi Ltd, H-800) was used for the electron diffraction.Organic products generated during the decomposition reaction were frozen out in a liquid-nitrogen trap and analysed as described previous1y.l. RESULTS The temperature dependence of the deposition from isopropyl titanate of a TiO, on a glass substrate is shown in fig. 1. The crystal structures of the as-grown films were also examined by X-ray diffraction and the estimated rutile content is shown. The maximum deposition was obtained at ca. 500 "C. Rutile formation is possible only in this temperature region. Fig. 2 shows the time dependence of the deposition at the optimum deposition temperature (500 "C). The amount deposited is linearly proportional to the deposition time. However, the rutile content in the films increased with deposition time and hence with increasing film thickness.The crystal modification of the film after deposition did not vary in the temperature region examined. Therefore, it is assumed that there is an unexpected mechanism changing the crystal structure above some film thickness during film growth. The concentration of oxygen in the vapour did not have a profound effect on the growth rate of TiO, films nor on the crystal modification, as shown in fig. 3, which indicates a linear relationship between the deposition yield and the deposition timeY. TAKAHASHI, H. SUZUKI AND M. NASU 10 N I Eo ? .- 2 5 - \ c .- WJ 0 '0 8 31 19 - .- r; 4 -a 10 5 0 0 1 5 30 45 60 75 9 0 105 time/min Fig. 2. Time dependence of TiO, and rutile content: deposition temperature, 500 "C; complex saturation temperatures, 50 "C; carrier N,, 0.5 cm3 s-l; 0,, 0.5 cm3 s-l; bypass N,, 1.0 cm3 s-l; pressure 600 Pa.6 % 8/ 7 % A/a 9% I 0 1 5 30 45 60 75 90 time/min Fig. 3. Time dependence of TiO, and rutile content ; deposition temperature, 500 "C; complex saturation temperature, 50 "C; carrier N,, 0.5 cm3 s-l; bypass N,+O,, 1.5 cm3 s-l; pressure, 600 Pa; total gas composition of 0,: A, 0% ; 0, 25% ; 0, 75%. for the three different oxygen concentrations. The rutile content in the film increased with increasing film thickness, irrespective of 0, concentration. The effect of the saturation temperature of the complex on the rate of film growth and the crystal modification is shown in fig. 4. The rate increased exponentially with increasing saturation temperature.A plot of the logarithm of the deposition yield against reciprocal absolute saturation temperature showed a good linear relationship, the slope of which is approximately proportional to the vaporization energy of isopropyl titanate. The estimated value of the vaporization energy from fig. 4 is 60.3 kJ mol-l, which is very close to the reported value (61.4 kJ m ~ l - l ) . ~ This suggests that the rate of the film deposition is linearly proportional to the vapour concentration of isopropyl titanate. Fig. 4 indicates that the rutile content increases with increasing film thickness or deposition rate. The effect of the flow rate of the carrier gas (N,) on the deposition rate is shown in fig. 5. An exponential dependence was observed, even though fig.4 indicates a linear relationship between the flow rate and the deposition rate. A plot of log (deposition rate) against log (flow rate of the carrier gas) gave a good linear relationship with slope3120 RUTILE GROWTH AT THE SURFACE OF TiO, 103 K I T 2 . 9 3 . 0 3 . 1 3 . 2 3 . 3 13 16 1 4 E" 12 .o * I l l 8 e 3 --. .- -0 6 4 2 \ \ \ 'a\ \ \ \ \ \ \ \ \ 3 . 0 - N 2 . 0 E 00 E --. 0 .- + .- 8 a 1 . 0 ; - 0.0 10 30 50 70 saturation temperature/*C Fig. 4. Effect of the complex saturation temperature on TiO, deposition and rutile content: deposition temperature, 500 "C; reaction time, 30 min; carrier N,, 0.5 cm3 s-l; 0,, 0.5 cm3 s-l; bypass N,, 1 .O cm3 s-l ; pressure, 600 Pa. 1.5, indicating that the deposition rate is proportional to the carrier-gas flow rate (i.e.the vapour-phase concentration of complex) to the power of 1.5. This result contradicts the results in fig. 4, but we cannot explain the reason for the discrepancy. However, the tendency of increasing rutile content with film thickness or deposition rate was again observed. The crystal structure of the as-grown films on the glass was determined by surface X-ray diffraction, and so the rutile content is that of the surface part of the attached films on the glass, indicative of increasing rutile content of the surface with film growth. It is strange that rutile crystals can grow more easily near the film surface as the film thickness increases than in the bulk. In order to clarify the origin of this curious behaviour, a cross-section of the fractured films was examined by scanning electron microscopy.Typical examples of images of pure anatase and rutile-con taining films are shown in plate 1. In the case of pure anatase films [plate 1 (a)], only regular, columnar crystals were found, while the rutile-containing films [plate 1 (h)] are composed of both short columnar crystals and dendritic crystals, with the latter located near the surface. Therefore, it seems that the columnar and dendritic crystals can be assigned to anatase and rutile modifications, respectively. Examination of plate 1 (b) suggests another interesting growth habit of rutile crystals: they spread over the columnar anatase crystals and suppress their growth until the surface can be completely covered by the rutile crystals (apparently 10Ox rutile content).This is the main cause of rutile growth near the surface, but it should be possible for anatase crystals to grow on the rutile crystals. Continuous growth of rutile on the rutile crystals can be considered to be due to strong epitaxial growth. In order to examine the epitaxial growth of the TiO, film, following experimentsJ . Chem. SOC., Faraday Trans. I , Vol. 81, part 12 (4 (b) Plates I and 2 Plate 1. Scanning electron microscopy photographs of cross-sections of fractured TiO, thin films grown on a glass substrate: (a) pure anatase film and (b) rutile-containing film (79% rutile). & 1 20ym - 10pm Plate 2. Scanning electron microscopy photographs of cross-sections of TiO, films grown on a glass substrate under the following conditions: (a) anatase condition (500 "C) for 30 min followed by rutile condition (500 "C) for 30 min (29% rutile) and (b) rutile condition (500 "C) for 30 min followed by anatase condition (500 "C) for 61 min (81 % rutile).Y. TAKAHASHI, H. SUZUKI AND M. NASU (Facing p . 3 120)J . Chem. SOC., Faraday Trans. 1, Vol. 81, part 12 Plates 3 and 4 - 20ym Plate 3. Cross-section of TiO, film (42% rutile) grown on a glass substrate by the rutile condition (500 "C) for 30 min followed by the rutile condition (400 "C) for 30 min. Plate 4. Ultrasonic microscope images of a rutile film grown on a glass substrate by the rutile condition for 90 min (film thickness is ca. 60 pm). Y. TAKAHASHI, H. SUZUKI AND M. NASUY. TAKAHASHI, H. SUZUKI AND M. NASU 3121 flow rate of carrier gas (N2)/cm3 s-' Fig.5. Effect of carrier-gas flow rate (N,) on TiO, deposition: deposition temperature, 500 "C; complex saturation temperature, 50 "C; O,, 0.5 cm3 s-l; total N,, 1.0 cm3 sp1; pressure 600 Pa. were performed: (1) a film was grown on the anatase films under rutile growth conditions, (2) a film was grown on the rutile films under an anatase growth condition and (3) a film was grown under rutile growth conditions, on a thin gold film which had been vacuum-deposited on the predeposited rutile film. The rutile and anatase growth conditions were obtained by controlling the concentration of isopropyl titanate in the vapour phase. The deposition conditions and the films obtained are summarized in table 1. With low carrier-gas flow rate (N2, 0.2 cm3 s-l), only anatase films were found at all deposition temperatures.They were highly oriented to the (220) face when surface growth was predominant (around the optimum growth temperatures). On the other hand, higher flow rates of the carrier gas gave rutile films at the optimum temperature conditions (450-500 "C). Therefore, the former deposition condition may be referred to as the 'anatase condition' and the latter as the 'rutile condition'. The time dependence of the growth of TiO, films under anatase conditions and the crystal structure of the film are shown in fig. 6, indicating that only anatase films were grown at longer deposition time. Fig. 7 shows that rutile crystals were found to grow very rapidly under rutile conditions, especially when the amount deposited was > 3-5 mg cm-2 (ca.10 pm). Since the anatase and rutile conditions were well established, cross experiments were performed in order to clarify the effect of predeposited TiO, crystals. Plate 2(a) shows a cross-section of a film obtained from sequential growth, i.e. the anatase condition for 30 min followed by the rutile condition for 30 min. The film is mainly composed of columnar anatase crystals. The rutile content, estimated by X-ray diffraction, is 27%, suggesting that the film thickness has no decisive effect on the crystal form of the film surface. On the other hand, in the film obtained by the rutile condition for 30 min followed by the anatase condition for 60 min, the dendritic rutile crystals grown originally by the rutile condition continue to grow, reaching the film surface even after the anatase condition, indicating strong epitaxial growth. The rutile content in the film is very high, as shown in plate 2(b).The deposition temperature has the most important effect on the crystal structure. However, as shown in plate 3, deposits obtained at 400 "C contained a large amount of rutile. In contrast, deposits obtained by the rutile condition for 30 min on the anatase film are mainly anatase. Therefore, epitaxial growth is more important for crystal structure than the deposition3122 RUTILE GROWTH AT THE SURFACE OF TiO, Table 1. Morphology of deposits obtained at different substrate temperature9 deposition carrier N, . . . 0.2 cm3 s-l 0.5 cm3 s-l temperature bypass 0, . . . 0.3 cm3 s-' 0.5 cm3 s-l /"C bypass N, .. . 1.5 cm3 s-l 1.0 cm3 s-l 400 450 500 5 50 600 anatase 100% [220Ib anatase 100% [220] anatase 100% [220] anatase 1000/, [220] anatase 100% [ - ] anatase 100% [220] anatase 31 % [220] rutile 697; anatase 93% [220] rutile 7% anatase 100% [ - ] anatase 100% [ - ] a Deposition time, 30 min; complex saturation temperature, 60 "C. Observed orientation of crystallites. I 1 100 1 0 N E 0 h 0 0 20 40 60 80 100 deposition time/min Fig. 6. Time dependence of the deposition weight and rutile content of TiO, films grown on a glass substrate: deposition temperature, 500 "C; carrier N,, 0.2 cm3 s-l; 0,, 0.3 cm3 s-l; bypass N,, 1.5 cm3 s-l; pressure, 800 Pa. temperature. Epitaxial growth can be confirmed by the following experiments. A half surface of a predeposited rutile film was covered by a vacuum-deposited thin gold layer and the other half remained uncovered.Subsequent growth of TiO, on the surface by the rutile condition indicated that the Au-covered part had less rutile than the uncovered part (see table 2). When grown on an SnO, film, a rutile film can be obtained more easily than on a glass. The films assigned to anatase were highly oriented to the (220) anatase face and in the X-ray-diffraction pattern only one diffraction peak was observed at CuK, of 70.3". The brookite form has also a (332) diffraction at a similar diffraction angle. Therefore, from this diffraction alone identification of the crystal form is impossible. The dendritic crystallites were tentatively assigned to the rutile form from the observation of cross-sections of films.In order to confirm these assignments, the columnar and dendritic crystals were collected on a micromesh by crushing a film over it, and they were examined by a transmission electron microscopy and an electron- diffraction method. Conclusive results were not obtained. From the corrected angle of the diffraction assigned to anatase (220), the lattice constant a was calculated toY. TAKAHASHI, H. SUZUKI AND M. NASU 3123 100 n E c) C Q) n 2 n 30 20 1 0 n N 0 30 60 90 deposition time/min Fig. 7. Time dependence of the deposition weight and rutile content of TiO, films on a glass substrate: deposition temperature, 500 "C; carrier N,, 0,0.2 and 0.5 cm3 s-l; 0,, 0.5 cm3 s-l; bypass N,, 1 .O cm3 s-l; pressure, 800 Pa.Table 2. Effect of coating the surface with a thin gold film on the rutile content in the films grown on predeposited TiO, films deposition condition" rutile content first second after first after second deposition deposition deposition deposition rutile for 20 min rutile for 30 rnin 27 % with Au coating 67% rutile for 20 min anatase for 30 min 47 % with Au coating, 10% without Au, 92% without Au, 68% a Deposition temperature, 500 "C; complex saturation temperature, 60 "C. be 379.0 pm, very close to the reported5 value of 377.7 pm. Many thin films grown on KBr crystals were treated with water and the KBr substrates were dissolved to separate and to collect the TiO, films. The powdered films had the same diffraction pattern as the anatase. The dendritic crystals had a very similar electron-diffraction pattern to that of brookite rather than to that of rutile.We suspect that the surface layer of the rutile has a different crystal form. 'The organic compounds generated during the decomposition reaction were identified as mainly propene with a small quantity of acetone. Isopropyl alcohol, which might be expected as a byproduct, was not found, probably because of its easy dehydration. When the complex was decomposed in the presence of an organic compound such as isopropyl alcohol, acetone or toluene, the maximum decomposition condition shifted to a lower temperature, probably because of the assisted decomposition by water vapour derived from the organic compounds. The effect of isopropyl alcohol is particularly remarkable, as expected from the rapid dehydration reaction.In the presence of these organic compounds brookite films, identified by the characteristic (012) and (004) diffractions, were grown at ca. 300 "C. A contamination of growth surface with organic residue or water may modify the subsequent film-growth mechanisms. The fact that t-butyl titanate gave brookite films3 may be closely related to this contamination effect.3124 RUTILE GROWTH AT THE SURFACE OF TiO, DISCUSSION The most interesting result of this study is that the rutile crystals grow at the film surface with increasing film thickness until the surface is covered completely by the rutile layer when the vapour concentration of the complex is sufficient to maintain the moderate growth rate of TiO, films.In the case of slow deposition only anatase films were obtained. Apart from the deposition rate, film thickness is considered to be another important factor controlling the surface crystal modification. Above a critical thickness (10-15 pm), only rutile crystals grow. Therefore, even 100% rutile films, identified by the ‘surface’ X-ray diffraction, contain anatase layer beneath the surface rutile layer. As shown in fig. 2 and 3, even when the rutile content increased with the deposition time, the film growth rate was constant through the runs. Thus the growth rate of anatase crystals is equal to that of rutile crystals. However, the epitaxial growth rate of rutile on rutile is more rapid than the growth rate of anatase. It can be assumed, therefore, that a mass-transfer-controlled transport of the vapour to the film surface is responsible for the results of fig.2 and 3. The rutile might grow more rapidly than the anatase if the concentration of the vapour was enough to grow rutile crystals, and once the dendritic rutile crystals were grown to some extent their rapid growth rate and their characteristic morphology might lead to suppression of anatase growth. In the case of isopropyl titanate, anatase growth may be usual. In other works, rutile nuclei may not be generated. However, once they are generated, they grow rapidly and rutile growth becomes predominant. Because of this lower capability of rutile deposition from isopropyl titanate, rutile growth at the surface could be followed. When ethyl titanate was used as the starting material, rutile films < 1 pm thick could be grown.This may be attributed to the ease of formation of rutile nuclei.6 Even in this case, however, very thin TiO, films (< 0.5 pm) had a tendency to be composed of anatase crystals irrespective of the growth condition. There is no reason why the ethyl complex gives more rutile nuclei, but its higher thermal stability4 and/or strong aggregation4 may play an important role in the vapour decomposition mechanisms. Unlike the halide process (TiCl, + H, system),’ in which continuous growth gives a rutile film and intermittent growth an anatase film, in the present study this intermittent effect was not found, probably because of the strong epitaxial effect. The existence of the intermittent effect in the halide process suggests the presence of both epitaxial and surface-contamination effects during film growth. Thus the OMCVD process for the growth of TiO, films is very similar to the halide process.The exponential increase of film growth rate with increasing concentration of the complex in the vapour (rate = k[c~mplex]~.~), as shown in the fig. 5, might be explained by acceleration of growth by the development of a rutile surface. If this is the case, a similar acceleration effect is expected in the fig. 4, which shows the effect of the saturation temperature of the complex. From the Clausius-Clapeyron equation, the vapour concentration [C] of the complex is given by (1) where H , is the vaporization energy, R is the gas constant and T is the absolute temperature of the complex.On the other hand, log [C] = - H,/RT+ constant log (rate) = 1.5 log [C] +constant. (2) From eqn (1) and (2) log (rate) = 1.5 H,,/RT+constant. (3)Y. TAKAHASHI, H. SUZUKI AND M. NASU 3125 Therefore, the vaporization energy estimated from fig. 4 must be 1.5 times lower than the value estimated above, i.e. ca. 40 kJ mol-l, which is different from the observed value. It is not clear at present which is wrong, the result of fig. 4 or the value of the vaporization energy. As a method of non-destructive analysis of films, an ultrasonic microscopic examination was performed. A typical image of a rutile-containing film (thickness ca. 60pm) is shown in plate 4. Round islands with a swirl like annual tree-rings are found, corresponding to the dendritic rutile crystals.Plate 4 suggests that the islands extend with film growth (with decreasing 2 value, where 2 is the depth under the surface) until they penetrate each other and cover all the surface. The images for the anatase films had no characteristic pattern, suggesting that the films are composed of very homogeneous layers. These findings are in good agreement with the scanning electron microscopy observation described above. So far the differences in electric or electrochemical properties of rutile and anatase have been described,8-9 but the differences found might be ascribed to crystal modifications as well as to preparation methods. From this work, however, rutile and anatase films can be grown under very similar deposition conditions, including the deposition temperature, which has the most remarkable effect on the electric properties of TiO,. An examination of the electrochemical properties of films obtained by our method has been made, and it is found that the anatase and rutile crystals have different electrical and electrochemical properties. The results of this study will be reported elsewhere.'O We thank Dr Kiyoshi Ishikawa, Hitachi Central Research Laboratory for identifi- cation of the rutile films by ultrasonic microscopy (Hitachi HSAM-100) and Mr Masahito Ota for identification of highly oriented anatase films by powder X-ray diffraction. I Y . Takahashi. K . Tsuda, K. Sugyama, H. Minoura, D. Makino and M. Tsuiki, .I. Chem. Soc., Faruday Trans. I, 1981, 77, 1051. T. Takahashi, A. Ogiso, R. Tomoda, K. Sugiyama, H. Minoura and M. Tsuiki, J . Chem. SOC., Furuday Truns. I, 1982, 78, 2563. Y. Takahashi. S. Wakayama, A. Ogiso and K. Sugiyama, Kinzokuhyoumen Gijutsu, 1984.35, 584. D. C. Bradley, Adc. Inorg. Chem. Rudiochem., 1972, 15, 259. C. Legrand and J. Delville, C.R. Acad. Sci., 1953, 236, 944. L. Springer and M. F. Yan, in Ultrastructure Processing of Cerumics. Glasses and Composites, ed. L. L. Hench and D. R. Ulrich (Wiley-Interscience, New York, I984), p. 464. S. Hayashi and T. Hirai. J . Cryst. Growth, 1976, 41, 41. T. Kawaguchi, C. Izawa and K. Kaseda, Bull. Tokyo Gakugei Univ.. Sect IV, 1969, 21, 78. B. Kraeutler and A. J. Bard, J . Am. Chem. SOC., 1978, 100, 5985. lo H. Minoura, M. Nasu and Y. Takahashi, Ber. Bunsenges. Phys. Chem., in press. (PAPER 5/459)

ISSN:0300-9599    

DOI:10.1039/F19858103117

出版商:RSC    

年代:1985

数据来源: RSC

摘要:

J. Chem. Soc., Faraday Trans. 1, 1985, 81. 3127-3141 Multicomponent Ion Exchange in Zeolites Part 2.-Prediction of Exchange Equilibria over a Range of Solution Concentrations BY KEVIN R. FRANKLIN AND RODNEY P. TOWNSEND* Department of Chemistry, The City University, London EClV OHB Receitled 6th February, 1985 Using a recently derived thermodynamic model for ternary ion exchange, equations are formulated which facilitate the prediction of ternary exchange equilibria in zeolites at a given constant temperature. The implicit equations (which allow predictions over a range of solution concentrations and compositions) can be solved iteratively, and the conditions and constraints which must be applied to the iteration procedures are described. The prediction method is then tested using experimental data on the ternary Na/Ca/Mg-zeolite A system which were reported in Part 1 of this series.The prediction procedure is shown to work satisfactorily. In addition, it is shown that normalisation procedures (widely used for cases of partial exchange in binary systems) are irrelevant for the prediction of either binary or ternary equilibrium exchange behaviour. Finally some shortcomings in the fitting procedures which are currently employed to fit experimental data are discussed. In recent years interest has increased in models for the accurate prediction of binary and multicomponent exchange equilibria over a range of conditions. Barrer and Klinowskil showed that equilibrium isotherms at any concentration below ca. 0.5 normal? may be generated using only data from one experimentally measured isotherm.Methods for predicting ternary equilibrium data from binary results alone have also been developed for use with clays2T and resin exchanger^.^^ However, these have been found to be of limited utility when applied to ion exchange in zeolites.6 This present study was carried out in order to investigate the possibility of using a recently developed and rigorous thermodynamic model for ternary e ~ c h a n g e ~ - ~ to predict ternary equilibria after the manner that Barrer and Klinowskil employed for binary exchanges. This paper describes the procedures which must be employed and shows that it it possible to predict ternary equilibrium data successfully for zeolites using this thermodynamic even when partial exchange precludes the determination of the standard free energies.lo Experimental data reported previously1° for the Na/Ca/Mg exchange in zeolite A are used to demonstrate the prediction procedures.It is, however, necessary first to describe briefly how the ideas of Barrer and Klinowskil for binary exchange have been developed. BASIC THEORY For a binary exchange with ions A Z ~ + and Bzs+ initially in solution and zeolite, respectively, the exchange reaction is t Throughout this paper, the term 'normal solution' refers to a solution containing zi ci mol of charge, where zi and ci are the valency and molarity (in mol dmP3) of ion i, respectively. 3 1273128 MULTICOMPONENT ION EXCHANGE IN ZEOLITES where the overbar refers to the zeolite phase. The thermodynamic equilibrium constant for this reaction may be expanded to give K, = (Ep cEp/Eif c p ) r ( g p / g $ ) = KGT(giB/g%) (2) where r is the correction factor for non-ideality in the solution phase and r = [@/y2] ; y A and yB are then the 'activity coefficients'll of the single ions A Z ~ + and B z ~ + in solution, respectively. Ei and ci refer to the cationic equivalent fraction and concen- tration in solution (mol dmV3) of ion i, respectively.The function [ g p / g 2 ] is a non-ideality correction for the zeolite phase. KGT is thus seen to be the corrected selectivity quotient of Gaines and Thomas. l2 From considerations based on the low levels of salt imbibition which are observed and the small changes in zeolitic water content which occur when the bathing electrolyte solutions are dilute, Barrer and Klinowskil suggested that the ratio g p / g # should be constant for concentrations < 0.5 normal.If this is a reasonable assumption, then only those quantities involved in KGT need be considered further for the purpose of making predictions. If TN is the total normality of the bathing electrolyte solution, then since CA = E A TN/zA, CB = E, TpJ/zB (3) K G T ( E 2 / E p ) = (EP/EP) (r/Q) (4) where Q = ( z ~ / z ~ ) T$-'A. ( 5 ) eqn (2) may be transformed to give Eqn (4) may be used to predict ion-exchange equilibria over a range of solution concentrations at a given temperature. All that are required are isotherm data accurately obtained at one constant value of TN over the whole composition range from EA = 0 to 1. Using the KGT data so obtained, the value of E , corresponding to any combination of EA and TN may be calculated by iteration.For a binary exchange involving only one co-anion, r may be calculated for any EA, TpyT combination using Glueckauf s equation.13 Otherwise, the equations of Fletcher and Townsend are appr~priate.~ 8* lo each of which comprises two of the conjugate binaries, viz. For a ternary exchange, three reaction equations may be 22, zw uzci + ZlJ zw V*v+ + zu z\7 W " W + - 22, Zw UzIJ+ f Z u Zw v * V + + Z u Zv w z + W (6) where By analogy with the binary case there are then three thermodynamic equilibrium constants (although a knowledge of the values of only two is necessary to define the third, by the thermodynamic closure rule)3 : where ,YwT and ,:,Q are the corrections for non-ideality for the ternary compositions of the solution and zeolite phases, respectively. If the same assumption is made asK.R. FRANKLIN AND R. P. TOWNSEND 3129 for the binary case (uiz. that the latter non-ideality correction is near-invariant for external solution concentrations < 0.5 normal) then the variation of vywKG with over all EA,EB need be known at one solution concentration only for the prediction of ternary equilibrium compositions to be possible. Then by analogy with the binary case, eqn (7) may be transformed to give where and (9) where and kCIA and kclB are ratios of the ternary corrected selectivity quotients? kC/A = (AYBKGT/BfCKGT)1/3 (12) and kC/B = (A:BKGT/CBAKCT)1/3* (13) Although more complicated, eqn (8)-(13) are exactly analogous to the binary eqn (4) and (5).Given therefore an adequate quantity of data [obtained under isonormal, isothermal conditions], describing the variation of kCIA and kCiB with EA, EB, then E A , EB can in principle be predicted for any TN, as for the binary case. r C / A , r C / B may be evaluated using the equations of Fletcher and T~wnsend.~ PREDICTION PROCEDURES BINARY EXCHANGES Eqn (4) can be used directly to predict binary exchange equilibria. Given experi- mentally determined KGT data corresponding to a chosen constant value of TN, and for the whole range of EA from 0 to 1, the magnitude of E A corresponding to any given values of EA and TN (0 < TN < 0.5 equiv. dm-3) can be determined by iteration. An iteration procedure is required since I? = JTEA), so that eqn (4) cannot be expressed as an explicit function of E A .Since there could be more than one set of E A , EA values that satisfy eqn (4) for any given value of TN, it is necessary to apply some constraints to the iteration procedure. First, therefore, the iteration is initiated with the value of E A which corresponds to the chosen value of EA at the experimental normality T-(exptl). Secondly, conditions must be defined to determine the direction of iteration (i.e. whether the iteration should proceed towards higher or lower values of EA). This is easily decided for a given pair of zA, zB values using eqn (4). Since' the I' correction is relatively small for low values of TN, it follows that dEA/dTN $ dT/dTN (0 < T N < 0.5 equiv.dm-3). Also, if the assumption of Barrer and Klinowski holds (i.e. that the value for the ratio of zeolite phase activity coefficients remains constant as the external electrolyte solution concentration is changed), then for a given value of EA the left-hand side of eqn (4) will not change as TN is varied. It then follows that if Q increases, EA must decrease, and the converse. However, whether Q increases or decreases depends not only on the direction in which TN is changed, but also on the relative magnitudes of 102 FAR 13130 MULTICOMPONENT ION EXCHANGE IN ZEOLITES Table 1. Iteration conditions for the binary prediction procedure : (a) ccnditions for prediction of EA for a constant value of ,!?A and (b) conditions for prediction of EA for a constant value of E A (4 (b) < O > O < O < O > O > O < O < O > O < O < O < O > O > O > O > O > O < O > O < O < O > O < o > O 0 > o r < O 0 0 < or>O 0 zA and zB [see eqn (5)].Table l ( a ) summarises the different iteration conditions therefore in terms of the relative magnitudes of zA and zB. When zA = zB, no iteration is seen to be required: although r varies very slightly with TN, and consequently EA must also change a little, this change is almost always negligible.l9 l4 Thus for zA = zB, the isotherm shape is nearly independent of TN. The studies reported here have shown that there is some advantage in slightly modifying the above procedure, thus allowing instead the prediction of EA for a fixed value of EA at any given solution concentration.Table 1 (b) contains the conditions for iteration if this approach is adopted. There are two main reasons for preferring this approach. First, as the calculation of r values is by far the most lengthy part of the iteration procedure, considerable computer time may be saved by calculating r only once for each prediction, rather than for each iteration step. Secondly, for ternary equilibria, the prediction of the crystal phase composition from a given solution composition is much the more convenient approach, since the standard ternary equilibrium diagram is one on which the grid lines which represent the solution phase composition are distorted with respect to the crystal phase,6f10 rather than the converse. (This entails expressing EA as the dependent variable and EA as the independent variable, which makes the prediction of EA from EA the preferred procedure: for further comments on this see below.) TERNARY EXCHANGES The iteration procedure here is similar to that employed for binary exchanges, although the existence of an additional compositional degree of freedom within each phase for the ternary case (viz.the composition variables EB and &) means that both the iteration conditions and the computer procedure employed need to be more strictly controlled than for the binary case. Simply scanning all possible values of EA, EB over the surface of the Na/Ca/Mg-A ternary isothermlo invariably gave erroneous results. With a given ternary solution composition EA, EB and for a given total solution normality TN, the values of rc,A and rC,B can be determined using procedures described in detail el~ewhere,~ and using eqn (9) and (1 1).Iteration then yields the predicted EAA,EB values (and hence by difference Ec). The criterion used to end the iteration loop was that be a minimum. However, eqn (8) and (10) are not independent of each other, and the iteration procedure must take this fact into account. As for the binary case, the direction of iteration must be governed by both d TN and the valencies of the ions (nowK. R. FRANKLIN AND R. P. TOWNSEND 3131 Table 2. Iteration conditions for the ternary prediction procedure iteration iteration requirement requirement ion charges for dTN > 0 for dTN < 0 z,, Z, and zc), and the starting point for the iteration must be the composition value EA,EB, which was found by chemical analysis to correspond to EA,EB on the experimental isotherm at the chosen total normality T,(exptl).All ternary systems must fall into one of three types. For zA = zB = zc, the crystal phase composition is near-independent of the external solution concentration1. l4 and no iteration is required. For zA # zB = zc the two ions with the same valency may be treated as indistinguishable for the purposes of iteration, and hence the conditions shown in table 1 (b) may be used. Thus if, for example, [(z,, zc)-zA] > 0 and dTN > 0 then from table 1 (b) it follows that d[(& + &)/EA] < 0, or dEA > 0. (Note that no conditions are placed on the individual behaviour of either EB or EC.Note also that when zA # Z, = zc then will change markedly as TN is changed, but since zB = zc, Qc/B is a constant.) For the case of zA # Z, # zc, a convenient approach is to group together the two ions with the highest valencies and then use a similar procedure to that adopted for the case of Z, # zB = zc. The full set of iteration conditions for three ions (having any combination of valencies) is given in table 2. EXPERIMENTAL Experimental data for the Na/Ca/Mg-A system were obtained at 298 K and for TN = 0.1 equiv. dm-3. The methods employed and the results obtained are given in Part 1 of this series.l0 Computing was carried out in the laboratory using an Apple I1 microcomputer. The iteration procedure was straightforward for binary predictions, but (as mentioned above) for the ternary cases the procedure had to be more strictly controlled.The method employed was as follows. Taking the ternary composition diagram [fig. 5 in ref. (lo)], a semicircular region was defined on the surface of the triangular diagram such that the base of the semicircle lay at right angles to the direction of iteration (defined by table 2). The starting point for the iteration [viz. EA, EB at TN(exptl)] lay at the mid-point of the base, with the arc of the semicircle extending in the direction of iteration. Values of EA and EB within this semicircle were scanned systematically until a point was found within this semicircular region for which D [eqn (14)] was a minimum. A second semicircle with its base parallel to the first was then defined as before.The procedure was repeated until successive scanning yielded no significant further improvement in the minimal value of D. Values of EA, EB corresponding to this composition were then taken as the final prediction. The procedure is shown in diagrammatic form in fig. 1. 102-23132 MULTICOMPONENT ION EXCHANGE IN ZEOLITES A Fig. 1. Diagrammatic representation of the ternary prediction iteration procedEre : &, iteration starting point [EA, EB at TE(ex_ptl)]; *, final prediction. (1) EA, EB; (2) EA, EB+O.l; (3) EA, EB -0.1; (4) E A + O . l , EB -0.05. Fig. 2. Predicted isotherms and experimental points for the Na/Ca-A system. Solid lines are predicted isotherms; experimental points are measured at A, 0.025, 0, 0.1 and ., 0.4 normal.K.R. FRANKLIN AND R. P. TOWNSEND 0 3133 I 0.2 0.4 - 0.6 0.8 1 0.8 0.6 E M g 0.4 0.2 RESULTS AND DISCUSSION BINARY PREDICTIONS Using the experimental data obtained for the systems Na/Ca-A and Ma/Mg-A at a total solution normality of 0.1 equiv. dm-3, isotherms corresponding to solution normalities of 0.025 and 0.40 were generated. These predicted isotherms (together with a few experimental points measured at these same concentrations in order to check the predictions) are found in fig. 2 and 3. Since in the case of magnesium the exchange had been found to terminate at a level of 85.2% ,lo normalisedl5 data were used to make the predictions in this case. The final isotherms were then produced by reversing the normalisation procedure.The good agreement between the experimental and predicted data shows that the basic theory1 regarding the prediction of binary data is sound, and that the iteration procedures work well. The importance of estimating correctly the magnitude of the non-ideality correction was next investigated by assuming that r = 1 for all concentrations, and then repeating the predictions. In the examples given in tables 3 and 4 it is seen that there was no consistent loss of accuracy for predictions made at the lower concentration for either the Na/Ca or the Na/Mg pair. However, at T, = 0.4 equiv. dm-3 all predictions were in greater error when the solution phase non-ideality correction was included. Nevertheless, it is not valid to assume therefore that r may be regarded as equal to unity for all systems if TN < 0.1 equiv.dm-3. Pairs of ions with high valenciesl'j such as the La/Mg couple17 can show very marked deviations from ideality, even in very dilute solution. Similarly, the r correction may be significant for ions such as cadmium or copper, where even in dilute solution a significant degree of association with water can occur.18* l9 Correctly determining the value of the maximum level of exchange has been shown3134 MULTICOMPONENT ION EXCHANGE IN ZEOLITES Table 3. Experimental and predicted data for the system Na/Ca-A at TN = 0.4 and 0.025 equiv. dm-3 crystal phase predicted solution phase experimental predicted r = i at TN = 0.1 TN ECa 'Na 'Ca 'Na 'Ca 'Na 'Ca E N a ECa 0.4 0.220 0.780 0.136 0.864 0.134 0.866 0.158 0.842 0.080 0.920 0.4 0.518 0.482 0.252 0.748 0.259 0.741 0.295 0.705 0.160 0.840 0.025 0.673 0.327 0.144 0.856 0.129 0.871 0.138 0.862 0.210 0.790 0.025 0.9813 0.0187 0.352 0.648 0.373 0.627 0.391 0.609 0.553 0.447 Table 4.Experimental and predicted data for the system Na/Mg-A at T, = 0.4 and 0.025 equiv. dm-3 crystal phase solution phase experimental TN ENa 'Na 'Mg predicted predicted r = 1 predicted EMg(max) EMg(max) EM,(max) = 85.2% = 85.2% = 100% at TN = 0.1 E N a 'Mg 'Na 'Mg 'Na 'Mg 'Na 'Mg 0.4 0.205 0.795 0.447 0.553 0.4 0.498 0.502 0.558 0.442 0.025 0.473 0.527 0.395 0.605 0.025 0.845 0.155 0.512 0.488 0.446 0.554 0.437 0.563 0.444 0.556 0.407 0.593 0.549 0.451 0.540 0.460 0.546 0.454 0.473 0.527 0.401 0.599 0.410 0.590 0.401 0.599 0.470 0.530 0.504 0.496 0.514 0.486 0.501 0.499 0.578 0.422 Table 5.Coefficients of polynomials for in kMg,Na and In kWglCa used to make the ternary crystal phase predictions shown in fig. 4 and 5 m, n dependent variable 0 1 2 3 4 5 coefficient of Qa for rn - 22.5 1 30.89 224.0 -788.7 968.1 - 398.6 - In kMg/Na In kMg/Ca - 20.61 71 3 6 - 148.2 187.2 -91.43 coefficient for EEa for n In kMg/Na - 6.887 108.3 - 377.0 493.7 -216.2 In kMg/Ca - - 11.00 93.76 -164.5 98.43 -K. R. FRANKLIN AND R. P. TOWNSEND 3135 Na Fig. 4. Ternary experimental and predicted points for the system Na/Ca/Mg-A at 0.4 normal. 0 , Solution phase compositions and ., crystal phase composition at 0.1 normal; 0, Experimental crystal phase composition at 0.4 normal; 0, predicted crystal phase composition at 0.4 normal. to be very important if standard free energies of exchange are to be accurately measured, since quite small changes in EA(max) have a large effect on AG*.18120 Accurate evaluation of A G e itself is irrelevant for the prediction of selectivities, as it is not necessary to evaluate the crystal phase non-ideality factor [see eqn (2) and (4)].However, an incorrect value for EA(max) could in principle distort the isotherm shape when predictions are made. To estimate the importance of this possible effect, a further set of predictions were made for the Na/Mg-A system assuming that FM,(max) = 1. Comparison of these results with those obtained assuming that EM,(max) = 0.852 (table 4) confirms that although the predictions are marginally worse if EM,(max) is assumed to be unity, nevertheless the normalisation procedure is not really necessary.Barri and Rees21 also found this to be the case when they used a similar procedure to predict data for the same system. [Their procedure differed in the method of evaluating r: see ref. (17) for comments on this matter.] The fact that normalisation is not a significant factor for successful prediction of data is an important result, as it means that good predictions should be possible for ternary or multicomponent systems which exhibit partial exchange, and for which normalisation is not a practical 8 y lo This consideration is therefore pertinent to the Na/Ca/Mg-A system, which exhibited very marked partial exchange be havi our. O3136 MULTICOMPONENT ION EXCHANGE IN ZEOLITES N a Mg Ca Fig.5. Ternary experimental and predicted points for the system Na/Ca/Mg-A at 0.025 normal. Symbols as fig. 4. TERNARY PREDICTIONS Equilibrium compositions were determined experimentally for the Na/Ca/Mg-A system at total normalities of 0.025 and 0.40 equiv. dm-3 and 298 K. These data were used to test the ternary prediction procedures outlined above. Predictions were made using only the experimental data obtained at TN = 0.1 equiv. dm-3. Fig. 4 and 5 show that although there is some discrepancy between experimental and predicted values, these are relatively small compared with the overall change in crystal phase composition which occurs as TN is varied. Previous work has shown that the polynomial orders chosen to fit In kcjA and In kCiB as a function of EA and EB have a marked effect on the calculated values of the standard free energies of exchange.The best-fitting polynomial has been previously determined6 using a sum of residuals function R, where R = ( c N In [kC/A(predicted)/ln kC/B(rneasured) N-P-Q-1 N is the number of experimental points, and P and Q are the orders of the polynomial with respect to EA and EB, respectively. The best fit is chosen on the basis of R being a minimum. For the Na/Ca/Mg-A system, the best-fit orders were found to be P = Q = 5 and P = Q = 4 for the In kMgiNa and the In kMglCa plots, respectively. The coefficients of these polynomials (which were used to make the predictions seen in fig. 4 and 5 ) are given in table 5.9SZ'O OEE'O 282'0 SI9'0 OLL'O E16'0 9SE'O I€o:o E6E.O €09'0 88€'0 €6€'O 6LZ'O z 12'0 880'0 OSO'O S8P'O I9S'O 19€'0 S6Z'O 9PZ'O ZZ€'O ZL 2.0 SI9'0 OLL'O E16'0 9L'i.O I€O'O €L€'O €09'0 86E'O SOP'O 682'0 ZIZ'O 880'0 OSO'O SLP'O 19S'O I8€'0 S6Z'O 9ZZ'O 062'0 ZLZ'O S€9'0 OhL '0 €16'0 99E'O I PO'O E6F.O $29'0 809'0 €ZP'O 682'0 ZOZ'O 860'0 OSO'O S 8P'O I SS'O I L E'O S8Z'O 9ZZ'O OZ€'O 282'0 S 19'0 OPL'O &I 6'0 90P'O I PO'O €6€'0 E6S'O 81 P'O € I P'O 662'0 z 12'0 880'0 OSO'O SSP'O I es.0 I8E'O SO€'O 9€Z'O OZE'O ZLZ'O Sf 5'0 089'0 €89'0 9LC.O ILO'O €9€'0 €8S'O 8ZP'O EEP'O 60€'0 ZEZ'O 860'0 OPO'O S6P'O IZS'O I IP'O SZ€'O 9LZ'O O€€'O 262'0 S6S'O OLL'O E16'0 96E'O I€O'O EOP'O €09'0 8L€'O €6€*0 6LZ'O ZZZ'O 880'0 OSO'O SOS'O 19S'O I9€'0 S6Z'O €LI'O PI€'O 6PZ'O IPS'O 6EL'O 668'0 s IP'O I LO'O S IP'O LZ9'0 9ZP'O 8ZP'O OO€'O OZZ'O €0 I '0 9PO'O PSP'O PZS'O PL€'O 982'0 I LOO'O 2800'0 P600'0 1 SO'O ZLO'O L9.tr.O €80'0 2800'0 680'0 L€Z'O 61L8'0 SZO'O 6L€6'0 SZO'O LZ8P'O SZO'O 819'0 SZO'O 8€0'0 P'O IPO'O 9'0 ZZ8'0 P'O 8ZIS'O P'O ZOS'O P'O 90S'O P'O 01 6 8 L 9 S P € Z I3138 MULTICOMPONENT ION EXCHANGE N a IN ZEOLITES -7.7 Fig.6(a) and (6). For description see opposite.K. R. FRANKLIN AND R. P. TOWNSEND 3139 Fig. 6. Plots ofln kMglNa as a function ofcrystal phase compositions: (a) ' by eye' fit, (b) third-order polynomial fit and (c) fifth-order polynomial fit. N a Fig. 7(a). For description see over page.3140 MULTICOMPONENT ION EXCHANGE N a IN ZEOLITES Fig. 7. Plots of In kMMglCa as a function of crystal phase composition: (a) ‘by eye’ fit, (b) third-order polynomial fit and (c) fourth-order polynomial fit.K.R. FRANKLIN AND R. P. TOWNSEND 314€ To assess the effect of polynomial order on the accuracy of prediction, fitting equations having various orders were used to predict the crystal phase composition for each point. Table 6 shows (unexpectedly, bearing in mind our earlier conclusions)6 that the closest predictions (those underlined) were obtained not when R was a minimum, but for lower orders. The reasons for this are not clear at present, especially as examinations of the contour diagrams which depict kMelNa and kMelca as a function of crystal phase composition do seem to confirm that the higher-order polynomials fit the data better (fig. 6 and 7). This problem is currently under study. K. R. F. gratefully acknowledges an S.E.R.C. CASE award with Unilever Research. R. M. Barrer and J. Klinowski, J. Chem. SOC., Faraday Trans. 1, 1974, 70, 2080. A. M. Elprince and B. L. Babcock, Soil Sci., 1975, 120, 332. S.-Y. Chu and G. Sposito, Soil Sci. SOC. Am. J., 1981, 45, 1084. W. J. Brignal, A. K. Gupta and M. Streat, The Theory and Practice of Zon Exchange (SOC. Chem. Ind., London, 1976), p. 1 1 . 1 . R. K. Bajpai, A. K. Gupta and M. Gopala Rao, J. Phys. Chem., 1973,77, 1288. P. Fletcher, K. R. Franklin and R. P. Townsend, Philos. Trans. R. SOC. London, Part A, 1984, 312, 141. 'I P. Fletcher and R. P. Townsend, J. Chem. SOC., Faraday Trans. 2, 1981, 77, 955. a P. Fletcher and R. P. Townsend, J. Chem. SOC., Faraday Trans. 2, 1981, 77, 965. P. Fletcher and R. P. Townsend, J. Chem. SOC., Faraday Trans. 2, 1981, 77, 2077. lo K. R. Franklin and R. P. Townsend, J. Chem. SOC., Faraday Trans. I , 1985,81, 1071. l1 A. Dyer, H. Enamy and R. P. Townsend, Sep. Sci. Technol., 1981, 16, 173. l2 G. L. Gaines and H. G. Thomas, J. Chem. Phys., 1953, 21, 714. l3 E. Glueckauf, Nature (London), 1949, 163, 414. l4 G. D. R. Parakrama and R. P. Townsend, paper in preparation. l5 R. M. Barrer, J. Klinowski and H. S. Sherry, J. Chem. SOC., Faraday Trans. 2, 1973, 67, 1669. R. A. Robinson and R. H. Stokes, Electrolyte Sofurions (Butterworths, London, 2nd edn, 1970). P. Fletcher and R. P. Townsend, J. Chem. SOC., Faraday Trans. 2, 1983, 79, 419. R. P. Townsend, P. Fletcher and M. Loizidou, Proc. 6th Znt. Con$ Zeolites 1983, Reno, Nevada (Butterworths, London, 1984), p. 110. I9 P. Fletcher and R. P. Townsend, J. Chem. SOC., Faraday Trans. I , 1985,81, 1731. 2o K. R. Franklin and R. P. Townsend, in preparation. 21 S. A. I. Bam and L. V. C. Rees, J. Chromatogr., 1981, 201, 21. (PAPER 5 / 132)

ISSN:0300-9599    

DOI:10.1039/F19858103127

出版商:RSC    

年代:1985

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. I , 1985, 81, 3143 CORRIGENDUM Liquid Structure and the Excess Volumes of Cyclohexane + Normal- and Branched-alkane Mixtures BY SAILENDRA N. BHATTACHARYYA? AND DONALD PATTERSON* Department of Chemistry, McGill University, 801 Sherbrooke Street West, Montreal, Quebec H3A 2K6, Canada J. Chem. SOC., Faraday Trans. I , 1985,81, 375-385 In the Redlich-Kister equation, eqn (l), x, and x, should be interchanged. 3143

ISSN:0300-9599    

DOI:10.1039/F19858103143

出版商:RSC    

年代:1985

数据来源: RSC

摘要:

Reviews of Books Le Mdthanol. RdalitQ et Perspectives. By R. DUMON, J-C. GUIBET and J-Y. PORTAS. (Masson, This compact survey of current and prospective methods for the manufacture and utilisation of methanol consists of eight chapters of unequal depth and merit. The first of these lists data from the open literature: molecular weight, vapour pressure and liquid density as a function of temperature, reactions of methanol (oxidations and ester formation reactions), ending with information on the toxic effects of methanol upon inhalation, ingestion or absorption through the skin. This short chapter is followed by two substantial ones dealing with the manufacture of methanol. Chap. 2 presents a review of processes for the preparation of synthesis gas for methanol production. The whole spectrum of starting materials is covered: steam reforming of natural gas, partial oxidation of heavier feedstocks and the gasification of coal and biomass.The qualitative comparison of process complexity and costs presented at the end of the chapter is particularly helpful, although processes such as the flash-pyrolysis of biomass (in an advanced state of development in France) and air-gasification systems to produce an almost N,-free gas (e.g. the Batelle system or that of the French company Gaz IntCgral, who recently installed a 40000 tonnes per year plant in Concepcion, Chile) deserve mention, even if only as ‘perspectives’. Suggesting the use of the classical electrolysis of water as a future source of supplementary hydrogen for hydrogen-poor synthesis gas is puzzling at a time when research in hydrogen generation from more energy-efficient cycles has more than completed its stage of infancy.Chap. 3 opens with a description of synthesis-gas purification systems, followed by a lengthy section on the thermodynamics of methanol formation and allied reactions. Although the economic advantages of low-pressure catalysts are frequently emphasised, little information is available on reaction kinetics and catalyst design. The short survey of reactor architecture would have gained much by being placed within the framework of the competing effects of reaction rate and equilibrium conversion with increasing temperature. Aptly entitled ‘Chemical Applications’, the next chapter is a short standard text on the organic technology of methanol utilisation in the chemical industry.The survey briefly covers some novel routes, such as the ICI monocellular protein process. Nonetheless, the promise of ‘perspectives’ along with ‘rtalitts’ lead at least one reader to expect a great deal more on the chemistry, and possible impact on the chemical industry, of carbonyl compounds. Indeed, an analysis of possible process routes, catalysts and reaction mechanisms, and above all an evaluation of critical constraints, would have been very valuable. Paradoxically, the authors conclude their slender section on this subject by recalling the possibly enormous implications of carbon monoxide addition reactions for the chemical industry. Chap. 5 presents an in-depth and balanced technical evaluation of the possible uses of methanol as a fuel or fuel additive for internal combustion engines.The low price and perceived strategic accessibility of methanol in Europe is carefully weighed against technical difficulties encountered in the use of methanol as a transport fuel. Problems encountered and practical experience accumulated over the past few years, mainly in Western Germany, in using 1-5% concentrations of methanol as a petroleum additive are presented in detail. Consequences of introducing two higher tiers of methanol concentration in transport fuels, 10-25 % and 85-95%, are examined separately, along with their implications for engine design, fuel distribution networks, reduction of tetraethyl lead content and engine performance. The first part of chap.6 (the last in the main descriptive body of text) presents some non-startling data on the thermodynamics of the combustion of methanol, followed by a valuable survey of work on the possible uses of methanol as a fuel in the electricity-generation industry. Apparently based on a limited number of EPRI reports, this section outlines two sets of trials for burning methanol in full-sized power stations, the first set using normally oil-fired Paris, 1984). Pp. vii + 240. Price Fl70. 31453 146 REVIEWS OF BOOKS burners, the second in gas-turbine installations normally fired by natural gas and heavy fuel-oil. The attractiveness of using methanol as a fuel in gas-turbine-powered installations is meticulously explained : the use of the methanol decomposition reaction as a low-temperature heat-recovery device makes fascinating reading.The last part of the book presents a short inventory of uses, prices and costs. Almost three quarters of the price of methanol arises from its manufacturing costs alone, making it too expensive (for the time being) as anything other than a raw material in the chemical industry. Readers with research interests in areas covered by this book may be surprised by the lack of attention given to citing original sources. A total of 35 works are listed in the bibliography, much as a suggested reading list, very few of them referred to in the main text. Specialist readers would also have been interested to see some attention paid to fundamental phenomena underlying the processes which have been so meticulously catalogued, possibly expecting the authors to pinpoint critical technical and/or economic constraints to be removed before new advances could be expected.Finally a brief word about editorial standards: apart from the (perhaps controllable) number of spelling mistakes and repetitions of material in different chapters (generously acknowledged in the well constructed introduction). the reader does face an unexpected number of repetitions within individual chapters, often within individual pages: for example, on p. 68 the water-gas shift reaction is presented twice, once with the heat of reaction and once without. This and the few graphs printed with degrees Fahrenheit as one of the axes (no reference cited) must surely be considered as signs of hasty editing.R. KANDIYOTI Received 10th May, 1985 Surfactant Science Series, Volume 15. Electrical Phenomena at Interfaces: Fundamentals, Measurements and Applications Ed. by A. KITAHARA and A. WATANABE. (Marcel Dekker, New York, 1984). Pp. xi+463. Price $95.25, Sfr209. Despite the rather general title of this book it should first be made clear that it is focussed primarily on the problems of dispersed systems and as such covers roughly the same ground as the two volumes of Kuyt and Overbeek, albeit in a modernised and condensed form. Like the older book it is a collective work which has been well edited so that this is not too evident to the reader. As the subtitle indicates, the book is divided into three sections. Part 1 on Fundamentals has five chapters. The first is a very brief account of potentials at interfaces while the second is a conventional account of the electrical double layer (Thermodynamics, Gouy-Chapman, Stern and ion-exchange models).The third discusses DLVO theory and colloid stability including the question of constant potential or charge. The fourth chapter describes the basis of electrokinetic theory, while the fifth discusses the special problems of non-aqueous systems when these are of low dielectric constant. Part 2 on Measurements describes experimental techniques, but almost inevitably overlaps into the discussion of fundamental aspects. This is particularly true of chap. 6, entitled ' Electrocapillary Measurements', which includes a fair amount of interpretation.Besides material conventionally included under such a title, there is a useful account of the oil-water interface reminding us of the contributions of one of the editors to the early development of this subject.Chap. 7, on electrokinetic measurements, includes classical techniques as well as those more recently introduced (such as the Doppler method), and chap. 8 considers the coalescence of droplets and the stability of colloids, concluding with the preparation and study of the monodisperse latex as a model system. Part 3, on Applications, consists of ten rather short chapters on subjects where electrokinetic effects play a role : Detergency, Flotation, Fibres (mainly dyeing), Paper, Electrocapillary Emulsification (depending on the lowering of interfacial tension by the electrocapillary effect), Pigments and Paints, Cosmetics, Antirusting, Electrophoretic Phenomena in Biological Systems and Reproduction in Copying and Electrophoretic Display. These range from a well referenced survey in the style of an Annual Report to a rather personal account of research in progress.REVIEWS OF BOOKS 3147 As a whole, they provide the reader with a useful entry to the subject with references to the original literature as well as to other, more comprehensive reviews.The only slight disadvantage is the small number of recent references, presumably as the result of the origin of this book from a Japanese book published in 1972. As a whole, the book deserves a warm welcome since it covers a field for which no equally comprehensive work is available. Despite some minor flaws in exposition and language, the exposition is generally clear and idiomatic. It is a worthy memorial to Professor Akira Watanabe.R. PARSONS Received 26th April, 1985 Techniques in Organic Reaction Kinetics. By P. ZUMAN and R. PATEL. (Wiley-Interscience, New The word ‘techniques’ in the title is used in a broad sense. Kinetic studies are considered as a means to the understanding of mechanisms of organic reactions in solution. As the authors say (p. 50), ‘the study of the kinetics of a reaction usually forms the backbone of a thorough mechanistic investigation’. The aim of the book is to show young researchers and advanced undergraduates how to obtain reliable rate data, how to treat them and how to use variations in the composition of reaction mixtures in the elucidation of mechanisms.The book has three main sections of differing length and character. Forty pages, out of 340, are devoted to ‘conventional’ techniques of kinetic measurements; this section would be salutary reading for research students, though much of it should be familiar from laboratory manuals. Another eighty pages (the last section) are given to techniques used in the study of fast reactions. This section is a sort of mini-introduction to the range of methods available, so that the student at least becomes aware of the possibilities when he is faced with the problem of a reaction too fast for ordinary techniques (though he may not realise which are the most versatile and relatively simple methods, and which require sophisticated apparatus). More illustrative examples would probably help the reader; in parts the text strikes one as somewhat too schematic.The electrochemical techniques are particularly well handled, as would be expected from the authorship. Between these two sections dealing mainly with experimental methods comes a long middle section of some two hundred pages, entitled ‘Analysis of kinetic data’. This covers two very different kinds of operation. One is empirical, and has as its first phase the treatment of experimental rate data including the derivation of the stoichiometry of reaction, the empirical rate equation, and the values of rate constants; and as its second phase the study of the effects of varying the temperature, pressure, solvent medium, and structures of the reactant molecules, leading to the evaluation of kinetic quantities such as activation energies or solvent parameters. The other operation is the interpretation of these effects in terms of a mechanism, i.e.the path followed by the molecules in passing from the initial to the final product. Such a mechanism is hypothetical, and remains open to revision in the light of further experiment (pp. 3 and 47); successful postulation of mechanisms ‘ requires great experience, ability of logical deduction, and a seasoned chemical intuition’. This central section of the book is not a survey of organic reaction mechanisms; its emphasis is on the methods. Its use will be to help the researcher to recognise the mechanism that may lie behind (say) a negative entropy of activation, or a peculiar dependence of rate on pH, or a curved Hammett plot.Many of the methods are illustrated by applications to actual reactions; a wider range of examples would make the treatment more attractive. In a book such as this, touching on many topics, much depends upon the authors’ judgment on the space allotted to each, and this must depend to some extent upon their own research interests. Thus we are given a full treatment in 47 pages of the effects of pH on reaction rates, and 21 pages on electrochemical methods of determining rate constants, but no more than a mention of the mechanistic use of activation volumes derived from pressure effects. Solvent effects are well handled in 23 pages, but activation entropies and steric effects receive little attention. There are wise and useful remarks on the planning of a mechanistic investigation, York, 1984).Pp. x+340. Price E57.25.3 148 REVIEWS OF BOOKS with emphasis on the time required for its various phases. Computerised data-handling is commended; some analogue computer circuits are described, and for digital computing the reader is referred to other sources, but the use of a mini-computer linked on-line to the detection system (now common) receives no more than a mention. A welcome feature of the book is that there are hundreds of references to original experimental papers, ranging in date over several decades. The illustrations are plentiful and clear, but many of the graphs are schematic and few show experimental points. The writing is clear, if occasionally idiosyncratic.The production is good. The price, none the less, is exorbitant. Presumably the publishers intend to bring out a paperback edition. The sooner the better, for the book will be helpful both to students and to their teachers and supervisors. E. F. CALDIN Received 1st May, 1985 Specialist Periodical Reports: Mass Spectrometry, Volume 7. Senior Reporter R. A. W. JOHNSTONE (The Royal Society of Chemistry, London, 1984). Pp. xii+427. Price €51.50 (members f30), $92. This is the latest volume in the by now well established series of biennial reviews of mass spectrometry and covers the literature published between July 1980 and June 1982. It is unfortunate that Dr Johnstone’s laudable efforts to cut costs and speed up production by using modern technology should have delayed publication because of technical difficulties.Nevertheless, although the subject has progressed substantially since mid-1982, I am sure that the volume will be warmly welcomed by all interested in mass spectrometry and gas-phase ion chemistry. The ten articles cover topics ranging from fundamental theory to applications in many areas of chemistry and biochemistry and bring together over 3000 references. In an article on ionisation processes and ion-dynamics, particular attention is paid to spectroscopic properties of ions, including the calculation of photoionisation cross-sections and multiphoton ionisation (a technique which developed rapidly during the period reviewed). Developments in theoretical and experimental aspects of unimolecular dissociation processes are then considered, including photodissociation and coincidence experiments.The relatively short article on structures and dynamics of ions takes mass spectroscopists to task for seeking very qualitative rationalisations of fragmentation processes, but also advises them that ‘The more one forgets about molecular orbitals when discussing fragmentation processes in a mass spectrometer, the better’. A review of developments in ion and molecular-beam chemistry describes the basic features of crossed- beam experiments (explained by reference to magnetic darts and a magnetic dartboard!) and after a brief discussion of experimental methods a number of specific reactions are discussed, particularly those involving three-, four- and five-atom systems. The review of the structure and reactivity of organic ions emphasises the need to use various types of thermochemical data in conjunction with collision-induced decomposition experiments to try to separate structural and internal energy effects.A short review on the mass spectrometry of organic negative ions concentrates on bimolecular processes, such as the use of negative CI reagent gases, and studies of the mechanisms of ion-molecule reactions in flowing afterglow and ICR instruments. The chapter on instrumentation deals primarily with developments in commercial instruments or in techniques associated with them. A number of desorption ionisation techniques, including fast atom bombardment (FAB), are reviewed and it seems that a chapter devoted to FAB was planned but, unfortunately, did not appear.Developments in magnetic sector and quadrupole mass-analysers are described, with particular emphasis being given to linked scans and to the emergence of tandem mass spectrometry. The two chapters on the use of mass spectrometry with chromatographic methods and its application to pharmacokinetic and drug metabolism studies account for one third of the volume, despite the fact that each contributor has been selective in choice of material. Instrumentation for the use of metastable transitions with chromatographic techniques was available, but at the time had been applied to few real problems. In contrast, LC-MS was still very much in the early stages of development and many problems were (and still are?) awaitingREVIEWS OF BOOKS 3149 its commercial availability as a routine technique.The first of these reviews is concerned primarily with the application of GC-MS to a wide range of chemical, biochemical and environmental problems. A major section of the second review describes the considerable activity in the field of drug metabolite identification and discusses the problems associated with the quantitation of drugs by mass spectrometric methods. The final two chapters, on the mass spectrometry of natural products and of organometallic, coordination and inorganic compounds, follow the pattern set in previous volumes. The increasing use of FAB and LC-MS in structural studies of natural products is evident. In contrast, field desorption ionisation and secondary-ion mass spectrometry played a much more important role in studies of inorganic compounds during the period under review, and FAB had yet to become important in this area.The diversity of subject matter within this volume is proof enough of the very wide range of problems to which mass spectrometry can be applied as a result of its unique blend of sensitivity and specificity, An author index would have helped one in finding one’s way around the many valuable references. Because of the rapid developments in the subject, especially in the use of FAB for the investigation of polar compounds, some of the reviews are already in need of updating. Nonetheless, they give a good picture of the state of the subject in mid-1982, and apart from this reservation, the volume can certainly be recommended. Might one hope that modern technology will deal more kindly with Volume 8? K. R.JENNINGS Received 16th April, 1985 Thermodynamic Properties of Nonelectrolyte Solutions. By W. E. ACREE JR. (Academic Press, One third of the contents of this somewhat unconventional text is devoted to the formulation of the usual thermodynamic relationships concerning the equilibrium properties of binary and ternary mixtures of non-electrolytes. The derivations of these relationships are unexceptionable and can be found in many other established books in the field. The rest of the book is taken up with a reasonably detailed presentation of a variety of currently fashionable semi-empirical treatments of liquid mixtures. Most can be included under the general heading of group contribution methods. The present reviewer is not entirely convinced of the absolute utility, still less the basic scientific merit of such approaches, but a certain number of chemical engineers seem to require to understand and make use of this type of interpolative treatment.As far as can be judged, the coverage of ASOG, AGSM, UNIQUAC and the like, together with the comparable models for- associated mixtures, is reasonably complete and well presented. Each chapter concludes with examples (and solutions) to test whether the reader has comprehended fully the substance of that particular section. The contents are further constrained in that, almost without exception, the discussion and examples quoted are restricted to room temperature and, even more unfortunately, to pressures of around one bar or below.The term gas-liquid critical point does not appear in the text and so the vast array of thermodynamic phenomena that occur in even binary mixtures in the neighbourhood of the gas-liquid critical line are not mentioned. This is particularly unfortunate and misleading as, in the Preface, the author states that one of the reasons for writing the present volume is to assist in calculating . . . ‘such phase-equilibrium data as are needed for either engineering design or laboratory applications’. Many of the interesting phenomena not mentioned, retrograde condensation, supercritical extraction, etc. are of great immediate concern in many industrial processes. In addition, experimental methods are virtually excluded from the text.Although the bulk of the text is free from error and well produced, there are several places where there is obvious lack of attention to detail. Several of the figures are badly drawn and may confuse the undergraduate or recently-graduated reader. The bubble-point and dew-point curves do not coincide at a maximum in fig. 3.1 1 and 3.13, the correct infinite limiting slopes are absent from fig. 1 1.6 and there should be double-minima on two but not three of the curves on fig. 11.7. The use of the terms superheated and supercooled on fig. 3.8 and 3.9 is either Orlando, 1984). Pp. x+308. Price $65.3150 REVIEWS OF BOOKS curiously quaint or just plain wrong depending on your point of view ! Finally, one of the most annoying features is the lack of consistency in the spelling of authors’ names. Brsnsted is spelt correctly on p.124, but wrongly, and in two different ways, five times on p. 110. Pimentel appears as Pimentil on p. 150 and as Pimental on p. 292, van der Waals and van der Waal appear to be used indiscriminately, Berthelot occurs twice as Barthelot and the R. J. Swinton on p. 293 is the same person as the F. L. Swinton on p. 289! Not a book to be recommended to the inexperienced. F. L. SWINTON Received 8th May, 1985 Electrochemical Detectors: Fundamental Aspects and Analytical Applications. Ed. by T. H. This slim volume contains 11 papers given at the 5th Anglo-Czech symposium in Electro- chemistry which was held in September 1981. Five are from the U.K., five from Czechoslovakia and one from the Netherlands.The principal subject is electrochemical detectors in HPLC and seven of the papers deal with this problem. Two of the others could be considered to be loosely connected with the main theme since they deal with the voltammetry of organic compounds at solid electrodes and the technique known as tensammetry. The remaining two on the im- miscible electrolyte interface and on the impedance of Li I CuO primary cells have little or nothing to do even with the more general title subject. The general character of the papers is that of a review, but of course one which is scarcely up-to-date since three years have elapsed between the meeting and the appearance of the book. This material would have been more rapidly published in an appropriate journal, where it should also benefit from a proper refereeing process.Since no individual is likely to be able to afford this book at nearly 23 pence per page, its main sale must be to the hard-pressed libraries. I am sure that there are many better ways in which they could spend their limited resources. R. PARSONS RYAN. (Plenum Press, New York, 1984). Pp. viii+ 172. Price $39.50. Received 13th June, 1985 Spectroscopy of Biological Molecules. Ed. by C. SANDORFY and T. THEOPHANIDES. (D. Reidel, This book is the product of a NATO Advanced Study Institute and as such it illustrates the twin shortcomings of these events. The organisation of a NATO conference is, or appears to be, completely in the hands of one or two individuals both as regards the definition of the topics to be covered, and the composition of the tutorial team.Few organisers can resist the temptation to cast their net too wide, and the volume under review is no exception. The title : ‘ Spectroscopy of Biological Molecules’ is in itself quite ambitious, but it is supplemented by: ‘Theory and Applications - Chemistry, Physics, Biology and Medicine ’. The scope is now widened beyond what is useful. Indeed, any one of the six sections of which this volume is composed could have been made the subject of a NATO conference. The first two chapters respectively deal with ab initio calculations on homopolynucleotides (J. J. Ladik), hardly biological molecules, and charge-relay processes involving haem groups (del Re). While not disputing the importance of these topics, their connection with spectroscopy strikes me as tenuous, at best.Then follow two chapters on various aspects of hydrogen bonding (Hadii). Once again, spectroscopic measurements receive only cursory mention. Raman and f.t.i.r. spectroscopy of nucleotides receive pride of place (seven chapters); this also happens to be the interest of one of the organisers. The review of the interactions between nucleic acids and alkylating agents (Bertoluzza) is particularly informative, as is the chapter on drug-target interactions (Manfait et al.) which compares in vitro and in vim studies, with emphasis on chemotherapy and immunotherapy. Of a more specialist note is a well presented review of the structural properties of inorganic phosphate glasses (Bertoluzza) which find application as biological implants.Dordrecht, 1984). Pp. x + 646. Price Dfl 230, E58.50.REVIEWS OF BOOKS 3151 The important contributions of modern n.m.r. techniques to studies of the structural and dynamic properties of biopolymers receive scant attention in only three chapters. No mention is made of their application to proteins. Of educational value is the introduction to the density matrix formalism of multipulse n.m.r. (Mateescu and Valeriu). A section of nine papers on the mechanism of action of visual and plant pigments constitutes the other major topic covered by this volume. The chapter on plant tetrapyrroles (Scheer) beautifully illustrates how a thoughtful combination of different spectroscopic techniques (u.v., visible, i.r., n.m.r., e.p.r., fluorescence and ENDOR) has resulted in the structure determination of the phytochrome Pfr chromophore and can provide detailed information on the biosynthesis and structure of pigments and energy transfer in situ.An introductory chapter on two-photon spectroscopy (Birge) is followed by the application of this technique to the study of ‘forbidden’ excited states of a series of pigments (Birge et al.) demonstrating the particular value of the NATO Advanced Study Institutes : an authoritative presentation of an advanced technique followed by one or several illustrations of its application. The concluding sections of the volume are devoted to the spectroscopy of membranes and to recent developments of some spectroscopic techniques, e.g. time-resolved Raman (Delhaye) and near-millimetre-wave spectroscopy (Genzel et al.). This volume will be of particular interest to those active in the general area of Raman spectroscopy of complex biological systems. The n.m.r. connoisseur must look elsewhere, despite the claims of the editors. Fortunately, his interests are well catered for by a wealth of monographs and reviews. This reviewer found of particular value the section on spectroscopic techniques applied to studies of photoreceptors. The book is well produced and does not suffer unduly from the usual shortcomings of edited conference volumes with contributions from a polyglot group of authors. It tries to cover too much ground. Any one of the topics identified by the organisers could usefully have been the subject of a NATO Advanced Study Institute that would have made for a more valuable publication, better justifying the high price. Finally, on a chauvinistic note, one may wonder by what criteria the group of invited tutors was assembled. The list of contributors suggests that, scientifically, the centre of gravity lies in the U.S.A. This is hardly surprising. What is perhaps unexpected is that the U.K. is not represented at all, despite the international standing enjoyed by U.K. research groups in several of the areas covered by this book, as witnessed by the bibliographies accompanying some of the chapters. F. FRANKS Received 17th June, 1985

ISSN:0300-9599    

DOI:10.1039/F19858103145

出版商:RSC    

年代:1985

数据来源: RSC