摘要:

J . Chern. Soc., Furuduy Trans. I, 1989, 85(6), 1267-1278 Influence of Pretreatment on the Properties of Ag/a-Al,O, Catalysts containing Large (& 1 pm) Pure and Cs-promoted Silver Particles Part 2.-CO Oxidation Measurements Garmt R. Meima,*? Mees Hasselaar, Adrianus J. van Dillen, Frederik R. van Buren$ and John W. Geus Department of Inorganic Chemistry, State University of Utrecht, Croesestraat 77A, 3522 AD Utrecht, The Netherlands The catalytic activity for the oxidation of CO with molecular oxygen of silver catalysts containing large silver particles (ca. 1 pm) has been studied. Both pure and caesium-promoted silver catalyst were investigated. The various pretreatments to which the catalysts were subjected, i.e. reduction or oxidation at different temperatures, influenced their activity.The promoted catalyst proved to be somewhat less susceptible to the various pretreatments. Arrhenius plots of the catalysts often exhibited a sharp break, giving rise to two sets of kinetic parameters. An apparent activation energy of 40 kJ mol-I was found at low temperatures, whereas an activation energy of 60 kJ mol-' was observed at high temperatures. It is argued that the activation energy of 40 kJ mol-1 can be ascribed to the reaction of CO with oxygen atoms in the neighbourhood of defects and the activation energy of 60 kJ mol-' to the reaction of CO with oxygen penetrated into the bulk at the same sites. The variation in activity could be attributed to changes in the pre-exponential factor. A separate discussion is devoted to a comparison of the present results with the findings on silver catalysts with a smaller mean particle size we reported on previously.Introduction In Part 1 of this publication we have reported on the interaction of oxygen and hydrogen with silver catalysts containing large silver particles (ca. 1 pm) both with and without a caesium pr0motor.l It was established that the effect of the pretreatment on the extent of sorption of both oxygen and hydrogen was small compared with the amounts taken up by the smaller silver particles studied previously.2,3 It was also found that the fraction of oxygen taken up as adsorbed oxygen atoms did not vary significantly with the pretreatment. Moreover, even after oxidation at elevated temperatures the amounts of oxygen that could penetrate into the silver lattice were relatively small.The overall results indicated that the effect of the pretreatment on the surface structure of large silver particles is restricted to the near surface layers. In this paper the influence of the state of these catalysts and of the various pretreatments on the activity for the oxidation of CO by molecular oxygen will be dealt with. In previous studies on catalysts containing much smaller silver particle^,^? we have shown that this reaction is very sensitive with respect to the surface morphology of the t Present address: Dow Benelux N.V. P.O. Box 48, 4530 AA Terneuzen, The Netherlands; to whom all $ Dow Benelux N.V. correspondence should be addressed. 12671268 Influence of Pretreatment on the Properties of Ag/a-Al,O, silver particles.The pretreatment turned out to have a very strong influence on the activity. The results could be explained satisfactorily by the fact that on the most abundant plane in the silver particles, uiz. the (1 1 1)-plane, oxygen is adsorbed exclusively at defects. The large variation in the activity for the oxidation of carbon monoxide could be ascribed to : (i) a change in the morphology leading to a larger or smaller fraction of (1 1 1)-surfaces, and (ii) a varying number of defects in Ag (1 1 1)-surfaces. Taking into account that with large silver particles complete structural rearrangements are more difficult and with our findings in Part 1 that the effect of the pretreatment is restricted to the near surface layers, the influence of defects and/or added caesium on the activity for CO oxidation can be studied.Our previous results have shown that oxygen penetrated into the silver lattice exhibits an activation energy of 60 kJ mol-', whereas adsorbed oxygen gives rise to an activation energy of 40 kJ mol-1.2T Therefore, CO oxidation measurements will also provide additional information on the influence of the pretreatment on the state of the sorbed oxygen species. Experiment a1 Catalysts The preparation procedure of the catalysts is described in Part 1.l Both catalysts consisted of 19.9 wt % silver supported on a-Al,03. The promoted catalyst contained 190 ppm Cs. The catalysts are designated Ag-Cs and Ag, respectively. CO Oxidation Measurements The pre-shaped catalyst particles were crushed into smaller particles and a sieve fraction of 0.5-1.0mm used in the experiments.A sample of 1.38 g of the Ag and 1.25 g of the Ag-Cs catalyst were placed into a tubular glass reactor with an internal diameter of 9 mm. The volumes of the catalyst bed were 1.55 cm3 and 1.40 cm3, respectively. The pretreatment of the catalysts (reduction or oxidation) was performed overnight at various temperatures, either in a 10% H,/N, flow or a 100% 0, flow. Unless otherwise stated, the catalyst was cooled down in the same atmosphere as that of the pretreatment. To study the influence of the pretreatment on the catalytic activity, the following sequence was decided on. First pretreatments in inert and oxidizing atmospheres would take place, whereafter the influence of reduction and subsequent oxidation would be studied. As shown in Part 1, oxidation of the catalyst is necessary to remove the carbonaceous impurities present on the fresh catalyst.' During the experiments a gas-flow of 50 cm3 min-l was passed downwards through the catalyst bed.The gas composition by volume was 1 YO CO, 1 YO O,, N, balance. All gasses were dried using standard methods. The temperature was measured and controlled by a thermocouple placed on top of the catalyst bed. The amount of carbon dioxide formed during passage through the reactor was determined conductimetrically . The conversion was calculated by comparing total combustion with the measured conversion. Total combustion was achieved by passing the effluent gas of the reactor through a bed of a supported copper catalyst kept at a temperature of 673 K. Details of this very accurate technique for measuring the concentration of carbon dioxide have been published el~ewhere.~ Results and Discussion The activity measurements for the oxidation of CO with molecular oxygen of both the promoted and unpromoted catalyst showed the same tendency towards the various pretreatments. However, the influence of consecutive pretreatments was smaller for the Cs-promoted catalyst.Only the conversion curves of the unpromoted catalyst will be shown here as they illustrate the general effects.1269 100 80 60 h 5 Y GO 20 c 323 423 523 T/K Fig. 1. CO conversion-us.-temperature curves for the non-promoted silver catalyst after consecutive pretreatments: (a) N,, 523 K; (b) 0,, 523 K; (c) 0,, 598 and 673 K.After loading, the catalysts were subjected to various (subsequent) pretreatments, first in inert and oxidizing atmospheres, thereafter in reducing atmospheres. Stable activities as a function of time were always measured over the complete temperature range, as long as the measuring temperature remained lower than about 520 K. Higher measuring temperatures sometimes led to lower conversion levels than previously measured at lower temperatures. To illustrate the strong influence of the pretreatment on the catalytic activity, the conversion of CO to CO, is plotted as a function of temperature in fig. 1. The following sequence was employed (measurement after each pretreatment): (1) N,, 523 K; (2) 0,, 523 K ; and (3) O,, 598 and 673 K. As can be seen, the catalyst clearly became more active during the sequence. After oxidation at the more elevated temperatures (pretreatment 3, curve 3), additional pretreatments in inert or oxidizing atmospheres, at the same or lower temperatures, did not alter the activity of the catalyst further.Ultimately, the catalyst was about a factor of four more active than the freshly loaded catalyst. The results will now be presented and discussed in more detail. Arrhenius plots as derived from the conversion us. temperature data provide interesting information. The experimental data often indicate a distinct break (giving rise to two activation energies). The pretreatments are only affecting significantly the pre-exponential factors of the Arrhenius equation. This has also been found in our previous investigation^.^? The data, therefore, were analysed using two fixed activation energies, uiz.40 and 60 kJ mol-1 and varying pre-exponential factors, k,. To facilitate comparison, in all Arrhenius plots the In k , values are given. In the figures the values are given in relative units because no correction has been made for the mass of silver. However, all measurements have been performed on the same samples, making comparison of the obtained values between the various pretreatments of each catalyst valid. First, we will present the Arrhenius plots obtained for the Cs-promoted catalyst, as they clearly illustrate the general observations and also will render information on the1270 InJEuence of Pretreatment on the Properties of Ag/cz-Al,O, Fig.2. Arrhenius plot for the fresh Cs-promoted catalyst after thermal treatment in oxygen at 523 K. At low temperatures oxidation with an activation energy of 40 kJ mol-I is exhibited. Pre-exponential factor in relative units. influence of the caesium promotor. Thereafter, we will return to the measurements on the pure silver catalyst of which some of the conversion curves are presented in fig. 1. As much of the same reasoning holds for both catalysts, the results obtained on the pure catalyst will be discussed more briefly. Fig. 2 shows the Arrhenius plot obtained for the fresh Cs-promoted catalyst after treatment in oxygen at 523 K. The catalyst exhibits an activation energy of 60 kJ mol-1 over almost the complete temperature range with only a very small part displaying an activation energy of 40 kJ mol-' at lower temperatures.When the Ag-Cs catalyst was subsequently thermally treated in hydrogen at 523 K, the Arrhenius plot represented in fig. 3 was obtained. The first consecutive pretreatment in nitrogen after the pretreatment in hydrogen gave rise to an identical plot and is also shown in the figure. Again the plot clearly exhibits a distinct break. Interestingly, now a much larger part is exhibiting an activation energy of 40 kJ mol-' at the lower temperatures and only a small part is displaying the activation energy of 60 kJ mol-' at more elevated temperatures. Also, it can be seen that the pre-exponential factors have substantially increased. The oxygen chemisorption and thermal desorption experiments described in Part 1 have unambiguously shown that the fresh catalyst contained carbon.Formation of carbon dioxide during interaction with oxygen and thermal desorption was demonstrated. As established previously,2 and from the it has been demonstrated that the presence of carbon in the surface layer of silver promotes the penetration of oxygen into the bulk. Also, in previous work2'* we have argued that the activation energy of 60 kJ mol-1 for the oxidation of CO with molecular oxygen is due to the more tightly bound penetrated oxygen atoms and that the activation energy of 40 kJ mol-1 can be ascribed to oxidation by adsorbed oxygen. The results presented in fig. 2, therefore, indicate that the presence of carbon causes the oxygen to be initially taken up mainly as bulk oxygen atoms. The thermal pretreatment after the first oxidation run has removed the carbon virtually completely from the surface.As a result, the activation energy of 40 kJ mol-'G. R. Meima et al. 0 . -Y 5 -2 -4- 1271 - T/K 500 400 350 2 I I I 1 1 1 1 1 1 1 l 1 1 I I I t 2.0 2.5 3.0 lo3 K / T Fig. 3. Arrhenius plot for the Cs-promoted catalyst after thermal treatment in hydrogen at 523 K (+) followed by pretreatment in nitrogen 673 K (0) subsequent to the pretreatment mentioned in fig. 2. Pre-exponential factor in relative units. is mainly exhibited in fig. 3. Furthermore, the pre-exponential factors of the reactions displaying the activation energy of 40 and 60 kJ mol-1 have increased also. The rise in the pre-exponential factor of the latter reaction is unexpected.When carbon is removed from the surface, it must be anticipated that the number of sites where oxygen can penetrate into the surface will be diminished. However, a thermal treatment of silver particles in hydrogen has been found to lead to metal particles of an almost spherical shape,8 whereas a thermal treatment in oxygen causes the particles to assume a more faceted shape.’-’l Consequently, thermal treatment in hydrogen will cause an increase in the proportion of crystallographic planes which are atomically more rough than the smooth (1 1 1)-plane in the silver surface. Since dioxygen can adsorb dissociatively on atomically more rough silver surface^,^ the number of sites where oxygen can be adsorbed in a state capable of rapidly reacting with CO will increase by thermal treatment in hydrogen.Kinetic and thermodynamic factors are governing the interaction of oxygen atoms with silver surfaces. On essentially pure silver surfaces, oxygen is dissociative1 y adsorbed onto atomically rough surfaces and defect sites on the silver (1 1 1)-surface. Penetration of oxygen atoms into the surface proceeds slowly at lower temperatures, and more rapidly at higher temperatures. At still higher temperatures, penetration of oxygen into the silver surface is thermodynamically unfavourable. As a result, only adsorbed atoms are present at high temperatures, and oxygen atoms dissolved into deeper layers of the silver are slowly released. As has to be expected, the reactivity of adsorbed oxygen atoms is higher than that of penetrated oxygen atoms.Accordingly, penetrated oxygen atoms are exhibiting an activation energy of 60 kJ mol-l, and adsorbed oxygen atoms of 40 kJ mol-l. The exact state of the penetrated oxygen is not exactly clear. However, it presumably pertains to a dissolved atomic form without substantial formation of Ag,O. This is especially well supported by the fact that prolonged pretreatments at high temperatures do not eliminate completely the penetrated oxygen. Moreover, Ag,O is thermodynamically unstable at these temperatures. If the adsorbed oxygen atoms would continue to contribute to the oxidation of carbon1272 InJEuence of Pretreatment on the Properties of Agla-Al,O, T/K 2.0 2.5 3 .O lo3 KIT Fig. 4. Arrhenius plot of the Cs-promoted silver catalyst after a second pretreatment in nitrogen at 673 K.Pre-exponential factor in relative units. monoxide, the sharp break in the Arrhenius plot of fig. 3 would not be evident. We therefore, must assume that the oxygen atoms adsorbed at defect sites on the (I 11)- surfaces and on atomically rough planes penetrate much more rapidly into the surface at temperatures above about 450 K. The rapid conversion of adsorbed atoms to atoms penetrated into the surface causes the sharp break in the Arrhenius plot. As stated above, it has been observed that the presence of carbon in the silver surface facilitates appreciably the penetration of oxygen atoms into the surface. The more rapid penetration explains the transition to the activation energy of 60 kJ at already lower temperatures, uiz.at about 370 K. This raises an interesting question about the reversibility of the conversion-us.-temperature plots which we will deal with next. In fig. 4 an Arrhenius plot is presented for the oxidation of carbon monoxide over the Ag-Cs catalyst after a second pretreatment in nitrogen at 673 K. It can be expected that a substantial fraction of the oxygen atoms dissolved in the near surface layer of the silver has been removed by these subsequent treatments in reducing and inert atmospheres. As can be seen, the activation energy of 40 kJ mol-' is displayed up to about 450 K as in fig. 3. However, the range of temperatures where the activation energy of 60 kJ mol-' is exhibited is now very small. At temperatures above about 470 K the conversion predicted by the Arrhenius equation with an activation energy of 40 kJ mol-1 shows up again.The explanation is that at higher temperatures penetration of oxygen atoms into the silver surface is thermodynamically unfavourable. As long as penetrated oxygen atoms are present, sites are provided where more oxygen atoms can still penetrate. When the oxygen content of the silver surface is, however, low, the penetrated oxygen is exhausted fairly rapidly and oxygen adatoms can no longer penetrate into the silver surface. Consequently, CO is oxidized again by adsorbed oxygen atoms. As a result, the course of the Arrhenius plot depends on the amount of oxygen atoms being dissolved into the silver before the start of the catalytic reaction. The small effect of the pretreatment on the catalytic activity of the caesium promoted catalyst, which could be anticipated from the sorption and thermal desorption experiments described in Part 1,l is evident from fig.5. Pretreatment in oxygen or in hydrogen at elevated temperatures followed by cooling in nitrogen leads to the same Arrhenius plot. The fact that no constant activity is obtained at high temperaturesG. R. Meima et al. 1273 500 40 0 350 2 I , , t l l l l I l l l 2.0 2.5 3 .O lo3 KIT Fig. 5. Arrhenius plots for the oxidation of CO over a Cs-promoted silver catalyst. The effects of thermal pretreatment in oxygen at 598 K, (+, x ) and 673 K (u), and in hydrogen at 673 K (0) followed by cooling down in nitrogen are shown. Pre-exponential factors in relative units. ly S -2 - - 1 1 1 1 1 I I I I I I 10’ KIT 2 .o 2.5 3.0 Fig.6. The effect of thermal pretreatment in hydrogen at 673 K on the Arrhenius plot for the oxidation of CO over the Cs-promoted silver catalyst. ( x , 0) cooled in H,; (0, +, a) cooled in N,. Pre-exponential factors in relative units. illustrates the variation in the amount of penetrated oxygen in the different experiments. In some runs the activity at high temperatures returns to the Arrhenius plot of 40 kJ mol-l, whereas in other experiments the penetrated oxygen is not yet completely consumed during the measurement. This is similar to the effects also previously observed in fig. 4. With catalysts containing much smaller silver particles, it was observed that thermal treatment in hydrogen leads to a rise in activity. However, if the catalyst was cooled down in nitrogen after treatment in hydrogen, an appreciably lower activity was Though the effects are smaller with the caesium-promoted catalyst containing12% Injluence of Pretreatment on the Properties of Aglcc-Al,O, Fig.7. Arrhenius plot for the oxidation of CO over the Cs-promoted silver catalyst. The catalyst was thermally pretreated in hydrogen at 673 K and cooled down in hydrogen, which led to a relatively small amount of dissolved oxygen. 0, increasing temperature ; 0, decreasing temperature. Pre-exponential factors in relative units. larger silver particles, qualitatively the same results were obtained (fig. 6). Cooling in nitrogen led to a significantly lower activity than cooling in hydrogen after pretreatment at 673 K.Since the removal of oxygen continues during cooling in hydrogen, the faceting at the edges and corners will be less extended. At the atomically rough surfaces created during the pretreatment in hydrogen and still remaining after cooling in hydrogen, slightly more adsorption and penetration of oxygen proceeds leading to higher pre- exponential factors. The drop in the pre-exponential factor at more elevated temperatures and the transition to the lower activation energy during some experiments is also evident from fig. 6. The irreversible character of the activity at higher temperatures is well illustrated by the results presented in fig. 7, where the Arrhenius plot obtained after a subsequent pretreatment in hydrogen is shown. The points indicated by filled dots were measured on decreasing the measuring temperature.As to be expected, the activity becomes lower after removal of the penetrated oxygen. The effects of hydrogen-hydrogen, or hydrogen-nitrogen pretreatment on the non- promoted catalyst are evident from fig. 8. As in the experiments dealt with in Part 1 of this paper, the variations are small. The Cs-promoted catalyst exhibited a similar behaviour, though the effects on the pretreatment were generally still smaller. With the non-promoted catalyst the drop in activity at low temperatures, which was observed with the catalyst with silver particles of 70 nm,4 is also found, as can be seen from the marked drop at about 350 K in fig. 9. Arrhenius plots after pretreatments in hydrogen and cooling in hydrogen or nitrogen are collected in fig.10. The lower activity after cooling down from 673 K in nitrogen with respect to the cooling down in hydrogen (fig. 9) can be noted again. The data reproduce well at low temperatures, whereas at high temperatures the drop in pre-exponential factor due to exhaustion of penetrated oxygen and the transition to the lower activation energy is again exhibited. To illustrate the relative stability of the Cs-promoted catalyst, it is interesting to consider the range of pre-exponential factors exhibited by both the promoted and non- promoted catalyst. With the non-promoted catalyst the pre-exponential factor (for the reaction with adsorbed oxygen exhibiting an activation energy of 40 kJ mol-l) varies1275 0' I I I 323 423 523 TIK Fig. 8.CO conversion-us.-temperature after consecutive pretreatments (after prior treatments in oxygen; see fig. 1) for the non-promoted catalyst. (a) H, 523 K, (b) H, 673 K, (c) H, 673 K and subsequent cooling down in nitrogen. TIK 500 400 350 2 .o 2.5 3.0 lo3 KIT Fig. 9. Arrhenius plot for the oxidation of CO over a silver catalyst with particles of about 1 pm after pretreatment in hydrogen at 673 K. No caesium promotion. Note the deactivation of the catalyst at low temperatures. Pre-exponential factors in relative units. from 1.7 x lo6 to 28.5 x lo6 gii min-l, and with the promoted catalyst from 3.85 x lo6 to 9.97 x lo6 g;: min-l. Whereas with the non-promoted catalyst the pre-exponential factor varies by a factor of 20, that of the Cs-promoted catalyst varies only by a factor of 2.5.The reaction with an activation energy of 60 kJ mol-' displays pre-exponential factors ranging from 1 x lo9 to 17 x lo9 g;: min-l for the non-promoted catalyst. The variation1276 Influence of Pretreatment on the Properties of Agla-Al,O, 2 0 -Y 5 -2 - 4 TIK 500 LOO 35 0 I I I I I I l l 1 I 2.0 2.5 3.0 lo3 KIT Fig. 10. Arrhenius plots for the oxidation of CO over a non-promoted silver-on-alumina catalyst containing silver particles of about 1 pm. The catalyst was thermally pretreated in hydrogen at 523 K and cooled either in hydrogen or nitrogen (0, +), and at 673 K and cooled in nitrogen (0). Pre-exponential factors in relative units. with both the promoted and the non-promoted catalyst involves a factor of 17. This factor is of the same order of magnitude for the reaction with adsorbed oxygen (40 kJ mol-l) for the non-promoted catalyst.We will now compare the present results obtained on catalysts containing large silver particles with our previous findings on catalysts with much smaller particles. In table 1 we have collected the data measured for the different catalysts (also see ref. 2 and 4). The data refer to the reaction of CO with the adsorbed oxygen species (40 kJ mol-l). Clearly, the pre-exponential factors are not related to the specific surface area of the silver particles. Being proportional to the specific surface area of the silver present, the activity of the catalyst with silver particles of 150-200 nm should have been higher than that of the catalyst containing silver particles of 1000 nm by a factor of about six; the pre- exponential factor of the catalyst with the 70 nm particles by a factor of no less than 16.The remarkably small difference observed for the various catalysts in activity per gram of silver can only be explained if the reaction is highly structure-sensitive and thus largely dominated by surface defects. The influence of defects has already been well established with catalytic reactions over oxides. For instance, experimental evidence has been obtained that the catalytic activity per unit surface area of copper oxide decreases with the particle size. This phenomenon was attributed to an easy annihilation of lattice defects in small crystallites.12 With metallic particles the influence of defects on the catalytic activity has been yet less well documented.However, it can be expected that with metallic particles, such as silver, an effect will only be observed at relatively large particle sizes. Generally, the high mobility of metal atoms over metal surfaces will bring about the rapid migration of grain boundaries out of clusters of metal particles, even with particles of, e.g. 50 nm. PashleyG. R. Meima et al. 1277 Table 1. Comparison of the pre-exponential factors of different silver-on-alumina catalysts pre-exponential factor/ lo6 min-l g;: particle size /nm maximum minimum ratio 70 50.4 3.6 16 15&200 10.4 0.5 20 1000 non-promoted 28.5 1.7 17 1000 Cs-promoted 10.0 3.9 2.5 and coworkers13 have demonstrated this to be true for gold particles, where the migration of grain boundaries out of coalesced particles was already observed to proceed smoothly at 573 K.Hence, it can be expected that lattice defects in pure silver particles will anneal rapidly, especially at high temperatures. Interestingly, it has been found that foreign atoms can decrease the rate of anneal of lattice defects in metal surfaces severely. Consequently, we envisage the effect of foreign atoms in silver particles to be twofold, i.e. the production of defects and the decrease in the (surface) mobility of the metal atoms. As a result, a relatively small amount of foreign atoms can severely affect the catalytic properties. This has been demonstrated for the catalyst containing the large silver particles onto which the Cs promotor was added. With silver particles larger than about 100 nm, annihilation of surface defects can be expected to proceed more slowly in view of the number of metal atoms involved and the distances the metal atoms have to cover.We therefore attribute the astonishingly high specific activity of large silver particles to the (surface) stabilization of defects. The minimum specific activity displayed by the particles of 150 to 200 nm is most probably due to the fact that lattice defects can still be annealed with particles of this size. The smaller specific silver surface thus causes the activity to be slightly lower than that of particles of 70 nm. With the particles of 70 nm, defects are clearly much less stable than with the very large silver particles. A break in the Arrhenius plot is only displayed with the fresh catalyst which contains carbonaceous imp~rities.~ When the carbon has been removed, the break is no longer exhibited.With these small particles, atomically more rough surfaces are dominating the catalytic activity, and not defects present on extended (1 1 1)- surfaces. As has to be expected, the break in the Arrhenius plots of the 70 nm particles is much less sharp. The atomically rough surfaces continue to contribute to the catalytic reaction. Consequently, adsorbed oxygen still contributes to the catalytic reaction, in contrast to the behaviour of the 1 pm particles. As soon as the carbon is removed, the activity due to defects disappears. This is apparent from a drop in the pre-exponential factor, also at lower temperatures. The activity of the freshly oxidized catalyst was higher than that of the catalyst thermally treated for prolonged periods of time.Also the observed instability in the activity at temperatures where annealing of surface defects can be expected to proceed is explained by the presence of defects. This also holds for the irreversible character of the conversion plots observed with measurements at more elevated temperatures. For a good understanding of the activity of silver in oxidation reactions, it is thus highly important to envisage that an undisturbed silver (1 1 1)-surface does not adsorb oxygen dissociatively. Albers et a1.l' have made this important observation on single silver crystals using ellipsometry to assess both the defect density and the oxygen coverage of the surface.Somewhat similar observations have also been made by1278 Influence of Pretreatment on the Properties of Agla-Al,O, Campbell,15 who studied the role of caesium as a promoter in the ethylene epoxidation reaction over a Ag (1 11) single crystal. Under reaction conditions the unpromoted Ag (111)-plane only showed a coverage of a few per cent of oxygen, whereas after Cs-promotion the surface coverage could approach a full monolayer. The promoting effect of caesium on the oxygen uptake is also evident from our results. As discussed above, the fraction of atomically rough surfaces where oxygen can be adsorbed and can penetrate into the silver lattice varies with the pretreatment. Since caesium stabilizes defects and, hence, enhances the ability of the surface to adsorb oxygen, the variation in activity is much smaller with the Cs-promoted catalyst. Any effect of the presence of caesium on the activation energies could not be established.This work indicates that only the pre-exponential factor is affected. Conclusions This study indicates that the activity of silver catalysts for the oxidation of CO with molecular oxygen is governed by the number defects in the silver surface. The pretreatment can influence the surface structure of the silver particles and hence the number of defects. More or less stable defects are established by carbon or the addition of caesium. The latter is not removed during oxidation, rendering the Cs-promoted catalyst more stable towards variations in the pretreatment. References I G. R. Meima, L. M. Knijff, R. J. Vis, A. J. van Dillen, F. R. van Buren and J. W. Geus (part I of this study), J. Chem. SOC., Faraday Trans. 1, 1989, 85, 269. 2 G. R. Meima, L. M. Knijff, A. J. van Dillen, J. W. Geus, J. E. Bongaarts, F. R. van Buren and K. Delcour, in New Developments in Selective Oxidation, European Workshop Meeting, 17-1 8th March 1986, Louvain-la-Neuve, ed. B. Delmon and P. Ruiz; Catal. Today, 1987, 1, 117. 3 G. R. Meima, R. J. Vis, M. G. J. van Leur, A. J. van Dillen, F. R. van Buren and J. W. Geus, J. Chem. SOC., Faraday Trans. 1, 1989, 85, 279. 4 G. R. Meima, L. M. Knijff, A. J. van Dillen, F. R. van Buren and J. W. Geus, J. Chem. Soc., Faraday Trans. I , 1989, 85, 293. 5 A. J. van Dillen, Ph.D. Thesis, Utrecht, 1977, chap. 4. 6 J. J. F. Scholten, J. A. Konvalinka and F. W. Beekman, J. Catal., 1973, 28, 209. 7 A. I. Boronin, V. I. Bukhtiyarov, A. L. Vishnevskii, G. K. Boreskov and V. I. Savchenko, Kinet. Katal., 1984, 25, 1510; Engl. transl., 1984, 25,1301. 8 B. E. Sundquist, Acta Metall., 1964, 12, 67 and 585. 9 G. E. Rhead and M. Mykura, Acta Metall., 1962, 10, 843. 10 R. Shuttleworth, R. King and B. Chalmers, Proc. Roy. SOC. (London), 1948, A193, 465. I1 A. J. W. Moore, Acta Metall., 1958, 6, 293. 12 J. W. Geus, in Preparation of Catalysts UZ, ed. G. Poncelet, P. Grange and P. A. Jacobs (Elsevier, 13 D. W. Pashley, Adv. Phys., 1965, 14, 327. 14 H. Albers, W. J. J. van der Wal, 0. L. J. Gijzeman and G. A. Bootsma, Surf. Sci., 1978, 77, 1. 15 C. T. Campbell, J. Phys. Chem., 1985, 89, 5789. Amsterdam, 1983), p. 1. Paper 8/006 12A ; Received 7th November, 1988

ISSN:0300-9599    

DOI:10.1039/F19898501267

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

J. Chem. Soc., Faraday Trans. I , 1989, 85(6), 1279-1290 Spectroscopic Characterisation and Photochemical Behaviour of a Titanium Hydroxyperoxo Compound G. Munuera, A. R. Gonzalez-Elipe, A. Fernandez, P. Malet and J. P. Espin6s Instituto de Ciencias de Materiales (Centro Mixto CSIC- Univ. Sevilla) and Dpto. de Quimica Inorganica, P.O. Box 1115. 41071-Sevilla, Spain In order to obtain more information about the generation of peroxide species during water photo-cleavage using M/TiO, (M = noble metal) systems in which an 0, evolution is normally not observed, a titanium hydroxyperoxo compound has been prepared by the reaction of TiCl, with H,O,. The structure and the thermal behaviour of this compound have been examined using TG-MS, XRD and TPD techniques. XPS, i.r., e.s.r. and u.v.-visible reflectance spectroscopy have been used to characterise the types of peroxo species present in this sample.The photochemical decomposition of this compound, which leads to 0, evolution, has been studied compared to a TiO, (Degussa P25) sample. A model is proposed to explain the lack of 0, evolution during water photo-cleavage on M/TiO, systems where peroxo species, similar to those observed in the titanium hydroxyperoxo compound, become stabilised against photodecomposition. Introduction During irradiation of aqueous suspensions of metal-loaded TiO, powders or colloids (M/TiO,, M = Pt, Rh, Ni, etc.), evolution of H, is always observed due to water cleavage though 0, could not be detected even after long periods of irradiation (60 h). To explain this behaviour, we have proposedl that peroxo-species are formed according to the stoichiometric reaction 2h' + 20H; + 20Hi,, + H202,ds where OH; refers to basic surface OH groups on TiO,., The resulting peroxide is retained by the TiO, support due to reactions such as / o A-0, Ti=O+H,O, + Ti, I +H,O (2) and 2Ti-OH + H,O, + Ti Ti + 2H,O.(3) 0 Detection of such peroxo species in these systems has been carried out by Kiwi and Gratze13 by the reaction with o-toluidine, a compound which reacts specifically with such species. However, though the amount of peroxo species was quantitatively equivalent to the evolved H,, even after such long periods of irradiation (90 h), most of the peroxo species are expected to be incorporated into the bulk of the TiO, colloidal particles according to reaction (2), as we have recently sh0wn.l Though Kiwi and Morrison4 and Augustynski et aL5 have suggested some models for the actual structure of these peroxo species, such models are considered only as tentative since they are not supported by spectroscopic characterisation.We have previously shown, using XPS,' that Rh/TiO, photocatalysts, used for water photocleavage, show the growth of a new O( 1s) peak at 532 f 0.1 eV which was ascribed 12791280 Photochemical Behaviour of a Titanium Hydroxyperoxo Compound to peroxo- species by comparison with the XPS spectra recorded for a titanium hydroxyperoxo compound, prepared by the reaction of TiCl, with H,O, as reported in the literature.' The aim of this work is to examine in more detail the nature of such peroxo species in this reference compound as well as their photochemical behaviour, in order to explain the lack of oxygen photogeneration during water photo-cleavage when using M/TiO, systems.Experimental A titanium hydroxyperoxide sample [hereafter referred to as TiO(O,)] was prepared as described by Schwarzenbach et al.' by the reaction of a H,O, acidified solution with TiCl,. From TG-MS analysis, the formula of the resulting yellow solid was established as Ti,O,~,(OH)o~,~xH,O, thus corresponding to ca. 1 f0.2 peroxo species per Ti ion, while the B.E.T. surface area, after evacuation at 473 K during 1 h, was 83 m2 g-l, decreasing to 64 m2 g-l after decomposition at 723 K during 4 h. A TiO, sample (anatase P25 from Degussa, SBET = 50 m2 g-') was used as reference in this work.Colloidal rhodium particles, used in some experiments, were prepared by flowing H, at room temperature through a 0.1 mol dm-, RhCl, aqueous solution in 1 mol dm-, NaOH. Portions of the resulting Rho suspension were immediately injected into the irradiation flask. XPS spectra were recorded with a Leybold-Heraeus LHS- 10 spectrometer working in the AE = cte. mode with a pass energy of 50 eV using Mg Ka radiation as an excitation source. The samples, in the form of pellets, were mounted on a tantalum holder which can be heated resistively while temperature was measured by a thermocouple placed at its rear. Gases evolved were monitored with a 4-200 mass quadrupole. The binding energy reference level was taken to be the O(1s) line of the oxide species in TiO, at 530.0 eV, which gives a value for C( 1s) of 284.6 & 0.1 eV due to carbon impurities which are always present in the samples.An HP 1000-E computer on line to the spectrometer was used for data handling (background subtraction, area calculation and fitting). Atomic percentages were estimated using sensitivity factors supplied with the spectrometer, which agree well with those reported in the literature.8 1.r. spectra were carried out on a self-supported disc of the sample using a Perkin- Elmer 684 spectrometer with a data-station on line which allows subtraction of the spectra in the absorbance mode. E.p.r. spectra were recorded, under previously reported condition^,^ with a JEOL JES-3P-X spectrometer working in the X-band (9.6 GHz).XRD diffractograms and u.v.-visible reflectance spectra were recorded with a Philips PW 1060 XR diffractometer and a Perkin-Elmer 554 UV-V spectrometer (equipped with an integration sphere and using BaSO, as reference), respectively. The irradiation experiments were carried out at ca. 313 K in a pyrex flask, described elsewhere,lO using 25 cm3 of a 1 mol dm-3 NaOH solution. Aqueous suspensions (ca. 50 mg of the sample) were irradiated with a 200 W Osram HBO bulb, with a total energy output at the flat window of the flask of 470 mW ern-,. Analyses of H, and 0, evolved into the flask free volume (ca. 17.6 cm3) were made by GC.1° Prior to each irradiation period, the suspension was deaerated with argon until N, and/or 0, were not detected. In some of the experiments, measured doses of 0, were injected in the cell before irradiation.Results Sample Characterisation Fig. 1 shows TPD results recorded for two portions of the sample. For the first one, the evolved gases were passed through a trap at 77 K to remove water (and any other condensable gases), while for the second portion the trap was removed; thus all theG. Munuera et al. 1281 40 - I 7 30 E 2 8 .a" 1 b) ; 20 Y * 10 350 450 550 650 T/K Fig. 1. TPD profile of the TiO(0,) sample. (-) Total signal; (....> H,O; (----I 0,. desorbed products were measured. MS-analysis showed that H,O and 0, were the products desorbed during TPD with only traces of condensable impurities (mainly hydrocarbons and CO,). The TPD profile of water evolution (obtained by subtraction of the two TPD spectra) showed that most of the molecular water was evolved at T < 473 K while the second peak at ca.550 K was probably due to the condensation of OH groups. Meanwhile, oxygen evolution gave two peaks, one broad at 540 K and a sharp one at ca. 600 K, decomposition being completed at 673 K. XRD spectra depicted in fig. 2 for the sample heated in flowing N, at different temperatures, show that the original material was amorphous but a crystalline phase was generated when it was heated at T > 573 K (coinciding with the sharp peak of loss of 0, in fig. 1). This crystalline phase corresponds to the anatase form of TiO, which remains stable without any vestige of rutile, even after heating at 623 K for 4 h. In order to characterise the peroxo species present in the original TiO(0,) sample, XPS spectra were recorded. The original compound was quite stable at 300 K in the pretreatment chamber of the XPS spectrometer (under P < lo-' Torr?)), except for a small loss of H,O and 0,, but heating at T > 473 K readily produced large amounts of H,O and 0,, as expected from the TPD results and the changes in the photoelectron spectrum shown in fig.3 for the O( 1 s) level. After heating at 673 K, the O( 1 s) spectrum, in the figure, coincided with that of the TiO, P25 sample used as reference, which agrees with the total decomposition to form TiO, suffered by this compound after this thermal treatment . The best-fitting O(1s) spectra, shown in fig. 3, were obtained for two bands at 530.0 eV (01) and 532.1 eV (011).The small shoulder at higher binding energy in the spectrum of the sample heated at 673 K, which also appeared in TiO, P25, can be attributed6 to OH groups and/or oxide ions on the surface of the sample. From the area of the peaks, the atomic percentages and O/Ti ratios (given in table 1) were obtained. Besides the changes induced in the O(1s) spectrum by the thermal treatment, a shift to higher binding energies (ca. 0.3 eV) of the Ti(2p) level was recorded, suggesting a higher density of electrons on the titanium ions in the original TiO(0,). The final value t 1 Torr = 101 325/760 Pa.1282 Photochemical Behaviour of a Titanium Hydroxyperoxo Compound I I I I I 1 70 60 50 40 30 20 28 Fig. 2. X-Ray diffractograms of the TiO(0,) sample heated in a N, flow. (a) Original; (b) 523 K, 8 h; (c) 573 K, 4 h; ( d ) 623 K, 4 h.at 458.5 eV (after decomposition at 673 K) corresponds to that of the Ti(2p) level for TiO, P25. Further information about the nature of the 011 peak at 532.1 eV was obtained by assessment of the O(2s) and O(2p) valence photoelectron regions. Fig. 4 shows the spectra in this region for the hydroxyperoxide sample and P25 together with their difference spectrum, after complete subtraction of the 01 peaks at 530.0eV. The resulting O(2s) and O(2p) features in the difference spectrum are ascribed to species 011. In addition to the XPS data, i.r. spectra in the range 110MOO cm-l showed some new vibration modes in the TiO(0,) sample when compared with the corresponding spectrum of P25. The difference spectra in fig.5 showed that heating up to 623 K produced loss of bands at 1040,890,770 and 710 cm-l which may be ascribed to the 011 species in the original sample and growth of a broad band centred at ca. 950 cm-'. E.p.r. spectra showed a very intense signal due to 0; species bonded to Ti4+ ions (gl = 2.026, g , = 2.009 and g , = 2.002)11 which was not affected by a high oxygen pressure (P > 10 Torr) though it vanished after heating at ca. 473 K. Fig. 6 shows that the main difference in the u.v.-visible reflectance spectra of TiO, and TiO(0,) was the presence of a weak band at ca. 420 nm (see inset in the figure) which completely disappeared when the sample was heated at 673 K, thus indicating that this band was characteristic of the 011 species detected by XPS.On the other hand, both samples showed a similar absorption pattern at R < 400 nm. Accordingly, a comparative study of their photochemical behaviour under band-gap irradiation is feasible.G. Munuera et al. 1283 0 536.0 532.0 528 .O binding energy/eV Fig. 3. The O(1s) XP spectra of the TiO(0,) sample outgassed (P = Torr) at: (a) 300, (b) 473 and (c) 673 K. (....) Experimental curve; (----) fitted spectrum. Photochemical Behaviour Suspension of 50 mg of the TiO(0,) sample in 25 cm3 of 1 mol dm-3 NaOH did not give any oxygen evolution in the dark. However, band gap irradiation in the U.V. (1 < 360 nm) led to a fast evolution of oxygen which immediately stopped when the lamp was switched off, thus indicating that the TiO(0,) compound was photochemically1284 Photochemical Behaviour of a Titanium Hydroxyperoxo Compound Table 1.Atomic percentages and O/Ti ratios obtained from XPS analysis sample 0 Ti O/Ti OJTi O,,/Ti ~~ ~ TiO,(P25) 67.2 32.8 2.05 1.98" - Ti0(0,)/300 K 73.4 26.5 2.80 1.85 0.95 Ti0(0,)/473 K 69.8 30.2 2.31 1.94 0.37 Ti0(0,)/673 K 67.2 32.8 2.05 1.95 - a Corrected for OH/H,O at the surface. (61 ( C ) 1 : : : : : : : I 40 32 24 16 8 0 binding energy/eV Fig. 4. The O(2s) and O(2p) XP spectra of (a) TiO, and (b) TiO(0,). (c) Difference spectrum b -0.66~ [a factor of 0.66 was used to obtain complete subtraction of the 01 peaks of both samples in the O(1s) region]. decomposed leading to 0, evolution under illumination, while it remained stable in the dark (or under vacuum), though the initial rate of 0, evolution (ca.38pmol h-l) slowly decayed with time. In order to show whether oxygen photo-uptake occurred simultaneously or not, experiments were carried out after introducing 1 cm3 of 0, in the cell (giving 40pmol in the gas phase) before irradiation. The initial rate of 0, photogeneration became 28 pmol h-l, indicating that two simultaneous processes (photoadsorption and photodesorption) are occurring in this sample. This was confirmed by the results, depicted in fig. 7, using portions of the sample which were partially decomposed in N, at different temperatures and/or times. A change, from a net oxygen photodesorption to photoadsorption, was observed when the TiO(0,) samples were irradiated in presence of 40 pmol of 0,. It is worth noting that 0, photo-uptake observed for the TiO(0,) sample which had been decomposed at 623 K was of the same order as that observed for the P25 reference," which agreed with theG.Munuera et al. 1285 1 I 1 1000 800 600 Wcm- ' Fig. 5. (A) 1.r. spectra, in transmittance mode, of the TiO(0,) sample outgassed (P < 10-5 Torr) at: (a) 298, (b) 423 and (c) 623 K. The dashed lines indicate base lines in each case. (B) Difference spectra in absorbance mode. \\ J n 30 0 400 500 Alnm Fig. 6. Diffuse reflectance spectra of (a) TiO, (P-25), (b) TiO(O,), (c) Difference spectrum b-a. decomposition of the TiO(0,) sample to TiO, (anatase). Moreover, irradiation of the latter sample in argon did not give any 0, evolution as previously observed for the P25. Since the presence of a metal (Pt, Rh, etc.) is required to generate H, under u.v.- irradiation of TiO,, experiments were carried out with two TiO(0,) samples. The original and the sample decomposed at 623 K were irradiated in the presence of colloidal metallic rhodium.As shown in fig. 8, a slow increasing evolution of H, was observed for the sample decomposed at 623 K. The presence of Rh neither modified the 0, evolution in the original sample nor did it generate H,.1286 Photochemical Behaviour of a Titanium Hydroxyperoxo Compound 100 - 80 - - - 8 60- =t 2 - LO- - 20- I I 1 L ;i ;1 10 tirnelh Fig. 7. Gas-phase 0, evolution during the irradiation in 1 mol dm-3 NaOH of the TiO(0,) sample heated at the following temperatures and times in a N, flow: (a) original; (b) 523 K, 2 h; (c) 523 K, 4 h; ( d ) 573 K, 4 h; (e) 623 K, 4 h.0 2 4 6 8 timelh Fig. 8. Gas-phase evolution during the irradiation in 1 mol dm-3 NaOH of (a) the TiO(0,) original sample, (b) the TiO(0,) sample in the presence of colloidal Rh, and (c) the TiO(0,) sample decomposed at 623 K in the presence of colloidal Rh.G. Munuera et al. 1287 Discussion Sample Characterisation To our knowledge, a detailed spectroscopic study of titanium hydroxyperoxide compounds of the type used here has not been previously reported. The two XPS bands at 530.5 (01) and 532.1 eV (011) may be ascribed to 0x0 (02-) and peroxo (0;-) species in agreement with the binding energies reported for similar species in other systems.12 Nevertheless, some contribution of H,O/OH- species (present in our compound) to the 011 peak cannot be ruled out.6 As shown in table 1, the excess of oxygen (O/Ti ca.2.80 compared to 2.05 for TiO,) must be associated with such 011 species (OII/Ti ca. 0.9), corresponding to ca. 0.45 peroxo species per Ti ion. The difference from the 1 0.2 value obtained from the TG-MS analysis may be ascribed to partial decomposition at 298 K of the peroxo species present at the surface of the TiO(0,) sample under the ultra-high vacuum conditions required for XPS recording. The assignment in the valence band region of the molecular orbital levels of the 011 species to a peroxide species (&, a::, oip, nip, n2*,") agrees with the UPS spectra of diatomic oxygen adsorbed on different metals reported by Kamath and Rao.13 The pattern of these M.O. levels is similar to that ascribed to the Si- species by Endo et a1.l4 in the XPS spectra of TiS, and TiS,.Several i.r. bands in the range 1040-890 cm-l, which disappeared upon heating, may be ascribed to different types of peroxo species. l5 Thus, q2-peroxo (T-shaped) complexes show vibrations ascribed to the uo-o mode in the range 932-800 cm-l l6 while solid-state peroxides give bands in the region 770-700 cm-l l 7 normally ascribed to peroxo species bonded to several metal cations (p-peroxo compounds). Our data in fig. 5, clearly suggest the existence of q2-peroxo and p-peroxo complexes in the TiO(0,) sample as proposed by Gratzel et al.3 Heterogeneity in the 0-0 bond order is also suggested by the strong band at 1040cm-' which does not correspond exactly to reported wavenumbers of titanium peroxo-complexes.15 On the other hand, the band at 950 cm-l, which appeared during the thermal decomposition of the TiO(0,) sample, may be associated with vTip0 modes of 0x0 Ti=O groups at the TiO, surface, in agreement with the values reported for mononuclear titanium-oxo complexes.'' A similar set of bands in the region 1180-900 cm-l and 900-730 cm-' have been found by Davidov et all9 upon adsorption of oxygen on reduced TiO, samples. We have reported the growth of the same set of i.r.bands, which were ascribed to peroxo species, during 0, photoadsorption on a TiO, P25 sample,,' which agree with the results of Augustynski et aL5 who showed the formation, during 0, photoreduction or H,O photo-oxidation, of the same set of peroxo species bonded to T14+ ions in TiO, photoelectrodes.The shift to lower binding energies observed for the Ti(2p) level in our sample, compared to TiO,, may be related to a degree of donation to the titanium ions from the 71 and n* molecular orbitals of the peroxide, as has been observed in the case of peroxo-complexes in solution. 21 The study of the stability of the Ti(p-0,) and Ti(q2-0,) species in our TiO(0,) compound, examined by TPD and XRD, indicates that they start to decompose only at ca. 423 K. It is worth noting that under all conditions used both types of peroxo compounds were identified, indicating a similar stability. Photochemical Behaviour The identical optical absorption at A < 400 nm of TiO(0,) and P25, allows a comparative photochemical study of both samples. The results shown in fig.7 clearly indicate that the net observed 0, evolution (i.e. 0, photodesorption or photoadsorption), during irradiation of partially decomposed TiO(0,) samples, corresponds to differences1288 Photochemical Behauiour of a Titanium Hydroxyperoxo Compound Fig. 9. Diagram of the modulated potential barriers expected for an M/TiO, particle irradiated under alkaline conditions. between two simultaneous opposite processes, i.e. photodecomposition of peroxo species and 0, photoadsorption. These two processes probably occur during water photo-cleavage on M/TiO, (M = Pt, Rh, etc.) where only a net 0, photo-uptake is normally observed10*22 as in pure TiO,. This indicates the incorporation of peroxo species into the bulk of the TiO, support in these samples, as previously suggested,' to account for the lack of 0, evolution during water photo-cleavage.It is important to understand why photodecomposition of the peroxide does not occur in such M/TiO, systems. Using TiO, with preadsorbed H,O,, we have a Fenton-like reaction for the decomposition of the peroxide upon irradiation which involves photoelectrons according to (4) 3H,O, + 2e;i,2 + 2H,O + 0, + 20H-. The presence of metallic particles of Rh or Pt on the TiO, seems to prevent this reaction, probably due to trapping the photoelectrons by the metallic particles, which then act as efficient electron scavengers." Nakato and TsubomuraZ4 have concluded that M/TiO, photoelectrodes containing small metallic particles ( < 5 nm) widely dispersed on TiO,? show a modulation of the semiconductor bands.Meanwhile, Aspnes and Heller25 have found that Rh particles on TiO, develop ohmic contacts with the support in a H, atmosphere. From these two observations, a Schottky barrier might exist at the liquid/TiO, interface under our alkaline conditions. An ohmic contact could be set up between the metallic particles and TiO, in the presence of gaseous H,, as shown in the scheme in fig. 9. This latter situation has been recently confirmed26 using e.s.r. which t A situation fulfilled by the Rh/TiO, sample previously used by us in water photocleavage experiments.'OG. Munuera et al. 1289 detected a non-activated and reversible electron transfer (i.e. Ti3+ generation) from weakly adsorbed H, to the TiO, indicating the ohmic nature of the Rh-TiO, contact.According to this model, during irradiation photoelectrons should be trapped at the potential well generated by the metallic particles (where the H+ reduction takes place), while peroxide is expected to be formed [according to eqn (l)] at the free TiO, surface, where photoholes are driven by the Schottky barrier at the liquid/TiO, interface. This would imply the oxidation of the basic OH; groups covering the free TiO, surface under these alkaline condition^.^' The H,O, generated in this way would then react with the TiO, surface according to eqn (2) and (3). This process seems to progress far into the bulk thus leading to corrosion of the TiO,. Moreover, the unchanged 0, evolution and the lack of H, evolution during irradiation of the original TiO(0,) sample in the presence of Rh (fig.8) seems to suggest that such peroxo species prevent a good electric contact between the TiO(0,) surface and the metallic particles, contrary to the situation for the sample which had been decomposed into TiO,, where H, slowly evolved. These results clearly suggest that the photostationary situation normally reached for H, evolution after long irradiations during water photocleavage experiments using M/TiO, systems,lo* 28 might be due to a progressive loss of ohmic contact between the metal particles and the TiO,, even in the presence of hydrogen. This effect occurs when the peroxide is incorporated into the TiO, close to the metallic particles and explains that thermal decomposition of such species restores H, evolution to its original Conclusions The XRD and TPD study of a TiO(0,) sample prepared from TiC1, and H,O, shows that the resulting amorphous material is rather stable at 298 K even under ultra-high vacuum, while it' decomposes by heating at T > 423 K to give 0,, H,O and TiO, (anatase).The characterisation of the peroxo species in this compound, carried out by the analysis of the XPS [O(ls), O(2s) and O(2p) levels], e.p.r. and i.r. spectra, confirm the existence of 0;- (and 0,) species bonded to Ti4' ions as ~ ~ - 0 , and p - 0 , ligands which are readily decomposed evolving 0, under band gap (A < 400 nm) irradiation. Experiments with added colloidal Rh particles to the TiO(0,) sample, either in its original state or after decomposition to TiO,, indicate that such peroxo compounds prevent a good electric contact between the metal and the photosupport.According to these results a mechanism is proposed for water photocleavage on M/TiO, which involves modulation of the semiconductor bands by the metal, as proposed by Nakato and T s ~ b o m u r a , ~ ~ leading to the formation of potential wells at the metallic particles which act as electron traps. This fact prevents the photodecomposition into 0, of the generated peroxide, through a Fenton-like process, and allows the incorporation of peroxide into the TiO, destroying the electrical contact at the metal/TiO, interface and stopping H, evolution. Authors thank Dr A. Navio for the preparation of the TiO(0,) sample, the CAICYT, the 'Junta de Andalucia ' and the ' Fundacion Ramon Areces ' for financial support and the referees for their suggestion to improve the manuscript.References 1 A. Fernandez, A. R. Gonzalez-Elipe, J. P. Espinos and G. Munuera, 6th Int. Conf. on Photochemical 2 G. Munuera and F. S. Stone, Faraday Discuss. Chem. Soc., 1971, 52, 205. 3 J. Kiwi and M. Gratzel, J . Mol. Catal., 1987, 39, 63. 4 J. Kiwi and C. Morrison, J . Phys. Chem., 1984, 88, 6146. 5 M. Ulmann, N. R. de Tacconi and J. Augustynski, J . Phys. Chem., 1986, 90, 6523. Conversion and Storage of Solar Energy, Paris (1986), Book of abstracts D-133.1290 Photochemical Behaviour of a Titanium Hydroxyperoxo Compound 6 G. Munuera, A. R. Gonzalez-Elipe, J. P. Espinos and A. Navio, J. Mol. Struct., 1986, 143, 227. 7 J. Miihlebach, K. Muller and G.Schwarzenbach, Inorg. Chem., 1970, 9, 2381. 8 C. D. Wagner, L. E. Davies, M. V. Zeller, J. A. Taylor, R. M. Raymond and L. H. Gale, Surface 9 J. C. Conesa, P. Malet, A. Muiioz, G. Munuera, M. T. Sainz, J. Sanz and J. Soria, Proc. 8th Inter. 10 G. Munuera, J. Soria, J. C. Conesa, J. Sanz, A. R. Gonzalez-Elipe, A. Navio, E. J. Lopez-Molina, 11 P. Meriaudeau and J. Vedrine, J. Chem. Soc., Faraday Trans. I , 1976, 72,472. 12 L. M. Moroney, R. St. C. Smart and M. W. Roberts, J. Chem. SOC., Faraday Trans. I , 1983,79, 1769. 13 P. V. Kamath and C. N. Rao, J. Phys. Chem., 1984, 88, 464. 14 K. Endo, H. Ihara, K. Watanabe and S-I. Gonda, J. Solid Stat. Chem., 1982, 44, 268. 15 M. Che and A. J. Tench, Ado. Catal., 1983, 32, 1. 16 R. D. Jones, D. A. Summerville and F. Basolo, Chem. Rev., 1979, 79, 139. 17 J. C. Evans, Chem. Commun., 1969, 682. 18 M. Che and A. J. Tench, Adu. Catal., 1982, 31, 77. 19 A. A. Davidov, M. P. Komarova, V. F. Anufrienko and N. G. Maksimov, Kinetika i Kataliz, 1973,14, 20 G. Munuera and J. A. Navio, 4th National Meeting on Adsorption, Sevilla (1979), Book of abstracts 21 D. Schwarzenbach, Inorg. Chem., 1979, 9, 2391. 22 A. Mills and G. Porter, J. Chem. SOC., Faraday Trans. I , 1982, 78, 3659. 23 (a) G. Munuera, V. Rives-Amau and A. Saucedo, J. Chem. Soc., Faraday Trans. I , 1979, 75, 736; (b) A. R. Gonzalez-Elipe, G. Munuera and J. Soria, J. Chem. SOC., Faraday Trans. I, 1979, 75, 748; (c) G. Munuera, A. R. Gonzalez-Elipe, J. Soria and J. Sanz, J. Chem. SOC., Faraday Trans. I , 1980,76, 1535. Interface Anal., 1981, 3, 211. Cong. on Catal., Berlin (1984), V-217. A. Muiioz, A. Fema'ndez and J. P. Espinos, Stud. Surf. Sci. Catal., 1984, 19, 335. 1519. p. 25. 24 Y. Nakato and H. Tsubomura, J. Photochemistry, 1985, 29, 257. 25 D. E. Aspnes and A. Heller, J. Phys. Chem., 1983, 87, 4919. 26 J. C. Conesa, G. Munuera, A. Muiioz, V. Rives, J. Sanz and J. Soria, Stud. Surf. Sci. Catal., 1983, 17, 27 G. Munuera, A. R. Gonzalez-Elipe, V. Rives Arnau, A. Navio, P. Malet, J. Soria, J. C. Conesa and 28 J. Kiwi and M. Gratzel, 8th Inter. Cong. Catal., Berlin (1984), 111-545. 149. J. Sanz, Stud. Surf. Sci. Catal., 1985, 21, 113. Paper 8/00745D; Received 27th June, 1988

ISSN:0300-9599    

DOI:10.1039/F19898501279

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

J. Chem. SOC., Furaday Trans. I, 1989, 85(6), 1291-1302 Photocatalytic Oxidation and Adsorption of Methylene Blue on Thin Films of Near-ultraviolet-illuminated TiO, Ralph W. Matthews CSIRO Division of Fuel Technology, Lucas Heights Research Laboratories, Private Mail Bag 7, Menai, NSW 2234, Australia Aqueous solutions of methylene blue are totally mineralized when recirculated over thin films of titanium dioxide illuminated with near-u.v. light. The rate of destruction obeys first-order kinetics with reasonable precision but the apparent first-order rate constant, k', decreases with increasing initial concentration of solute. 1 dm3 of 10 pmol dmP3 methylene blue solution illuminated with a 20 W lamp, is decreased to 5 pmol dm-3 in 1 1.8 min. Sunlight from a 1 m2 parabolic trough is capable of destroying the methylene blue at 6.4 times this rate. The decrease in k' values with increasing concentration is consistent with curves calculated using the integrated form of the Langmuir expression.The adsorption parameter determined in the analysis of the kinetics of the photocatalytic data agreed with the adsorption affinity parameter determined using the classical Langmuir adsorption isotherm for the dark equilibrium data ; this indicated the key role played by adsorption in photocatalytic oxidation with titanium dioxide. The maximum quantum yield for methylene blue destruction at high flow rates with a 10 pmol dm-3 initial concentration was calculated to be 0.0092. Introduction Photocatalytic oxidation is a comparatively recent technique for the destructive removal of organic impurities from water.It is receiving increasing attention as a means of water purification. '-17 Titanium dioxide powder suspensions illuminated with bandgap light have generally been used as the photocatalyst, but supported TiO, may also be used as a stationary phase.'8-20 Since the photo-oxidation reaction takes place at the surface of the photocatalyst, the adsorption characteristics of the solute are expected to be quite important and experimental evidence that supports this view21 has been obtained for a number of compounds. In some cases, compounds that are normally quite resistant to oxidation were oxidized in dilute solutions more rapidly than less-resistant compounds at the same concentrations, apparently because of the strong adsorption of the former on the reactive surface of the catalyst.This was not, however, a definitive explanation since the parameter indicating the degree of adsorption was obtained from an analysis of the data in terms of Langmuir-Hinshelwood kinetics and not from equilibrium studies. An attempt is made to relate Langmuir parameters from equilibrium experiments to parameters obtained from kinetic studies. Experimental Materials Degussa P25 titanium dioxide was used as a photocatalyst. This material is mostly in the anatase form, and has a B.E.T. surface area of 50 m2 g-l and a mean particle size of 12911292 Adsorption of Methylene Blue on Thin Films 30 nm. The photocatalyst was attached in a thin film to the inner wall of borosilicate glass tubing wound in a ~pira1.l~ Methylene blue laboratory reagent used as the solute was obtained from Aldrich.It was selected because it has strong adsorption characteristics on many surfaces, good resistance to light degradation, a well defined optical absorption maximum at 660 nm and is a common dye. Salicylic acid analytical reagent was from BDH. Solutions were prepared using water from a Millipore Waters Milli Q water purification unit. Perchloric acid was used when pH adjustment of solutions was required. Apparatus The borosilicate glass spiral was wound from 7 m of 6 mm 0.d. tubing in 65 turns so that it fitted closely around a 20 W NEC blacklight blue fluorescent tube (TlO), 32.5 mm diameter, 588 mm long. The spiral and fluorescent tube could be mounted in a standard domestic lamp holder.In some experiments, the spiral was mounted at the focus of a small parabolic trough. The inner surface of the spiral was coated with a layer of ca. 75 mg of TiO, as described elsewhere.,O Solutions were pumped through the spiral using a Cole-Parmer peristaltic pump. Unless otherwise stated, solutions were unbuffered and at their natural pH. pH measurements were made with an Orion model 801 meter. Analysis Changes in methylene blue concentration were measured spectrophotometrically using a Cary model 16 instrument. Calibration curves were prepared using solutions of known concentration with an analytical wavelength of 660 nm. For concentrations up to 10 pmoldm-3 the Beer-Lambert law was obeyed with good precision and the molar extinction coefficient was found to be 66 700 & 350 dm3 cm-l mol-l.Changes in salicylic acid concentration were determined by spectrophotofluorimetry using a Perkin-Elmer LS-5 luminescence spectrometer. The excitation and emission wavelengths were 294 and 41 2 nm, respectively. Carbon dioxide analyses were done by conductivity detection using the apparatus and method described elsewhere. ( b ) Procedure Generally, 500 cm3 of solution was pumped from a reservoir (an 800 cm3 beaker open to the atmosphere) through the spiral and back to the reservoir continuously at a flow rate of 100 cm3 min-'. In some experiments different flow rates and volumes were used. The solution in the reservoir was sampled every 2 min and returned to the reservoir after measurement.After 10 min the lamp was switched on. (In the solar experiments a black cloth was removed.) Between runs, water was circulated through the spiral, with the lamp on for ca. 1 h, to oxidize traces of adsorbed decomposition products photocatalytically. Separate experiments were run for the CO, determinations using 40 cm3 of solution at each concentration. The apparatus allowed continuous monitoring of the CO, formation during illumination. Results and Discussion Solute Concentration The effect of initial solute concentration on the rate of disappearance of methylene blue from the reservoir is shown in fig. 1. At all concentrations there was a small, rapid decrease in concentration that occurred in the dark, plateauing to a constant value after ca. 10min. When the lamp was switched on, a further decrease in concentration occurred that followed approximately first-order kinetics after 2 min of illumination.R.W. Matthews 1293 time/min Fig. 1. Disappearance of methylene blue from solutions recirculated through the TiO, photoreactor. Solution volume: 500 cm3, flow rate 100 cm3 min-', 20 W NEC T10 blacklight fluorescent tube. The Dark Reaction The small rapid decrease in methylene blue concentration that occurred on circulating the solution through the spiral without the light on is attributed to adsorption of methylene blue on the surface of the TiO,. When the change in concentration during this 10min period was plotted against time the results in fig. 2 were obtained. The concentrations at 10min were taken to be equilibrium concentrations and the concentration of methylene blue adsorbed on the surface was calculated from the difference between the initial and equilibrium concentrations.The plot of methylene blue adsorbed us. the equilibrium concentration shown in fig. 3 has the shape of a typical Langmuir adsorption isotherm and may be described by the expression : where [MB],,, is the concentration of methylene blue adsorbed, [MB] is the equilibrium concentration of methylene blue in solution, and k , , k , are constants for the given system, k , being related to the adsorption affinity and k, being related to saturation coverage of the surface. Eqn (1) may be rearranged to the standard reciprocal form and reciprocal values of [MB],,, plotted against reciprocal [MB] values (fig.3 insert) or solved by the method of non-linear least-squares to give the Langmuir parameters. The latter method was used to obtain values for k, and k, of 0.0298 0.0064 dm3 pmol-' and 44 FAR I1294 Adsorption of Methylene Blue on Thin Films 0.0 0 2 4 6 8 10 Fig. 2. Adsorption of methylene blue from solutions of different concentration onto the non- illuminated, supported TiO, of the photoreactor. Initial concentrations (pmol dm-3) : 0, 1 ; a, 2; 8, 5 ; (>, 10; 0, 20; 0, 50; 0, 100. timelmin 5.0850.45 pmol dm-3, respectively. The curve shown in fig. 3 was obtained by substitution of these values in eqn (1). The Photocatalytic Reaction The discontinuity between the rate of decrease in concentration during the first 2 min after the lamp was switched on and the rest of the illumination time was presumably due to the lamp warming up to full power.Regression analysis for the linear portion of the results gave the apparent first-order decay constants, k , listed in table 1. The approximate obedience to first-order kinetics, together with a decrease in the apparent rate constant with increasing initial concentration, has been noted for other solutes.11. 1 9 9 2o In one of those reports2' the concentration dependence of the apparent rate constant was explained in terms of the integrated form of the Langmuir adsorption isotherm : t = -ln--+-([S]o-[S]) where t is the time in minutes for the initial concentration of solute, [S]', to decrease to [S] and k,, K are constants for the given system related to the adsorption and reaction properties of the solute.It follows from eqn (2) that at (2) 1 [S]O 1 klK [SI K when [S]/[S]' = 0.5, (3) 0.693 0.5[SI0 0.693 +- K k , K ' - -- '0.5 = ~ k' Therefore, if eqn (2) is a reasonable approximation to the data, a plot of to.5 values against the initial concentration values should yield a straight line whose slope is 0 S / K andR. W. Matthews I I I I I 1295 0 0 20 40 60 80 100 [methylene blue]/mol dm-3 x lo6 Fig. 3. Methylene blue adsorbed versus equilibrium concentrations, from data of fig. 2. Insert : Plot of reciprocal concentrations. The curve was calculated by the method of least squares. Table 1. Apparent first-order decay constants for the photocatalytic destruction of methylene blue at different initial concentrations 1 0.0830 f 0.0005 8.35 2 0.0785 f 0.0005 8.83 5 0.0728 f 0.000 1 9.52 10 0.0700 0.0002 9.90 20 0.0645 f 0.0005 10.7 50 0.0453 f 0.0005 15.3 100 0.0286 f 0.0009 24.2 whose intercept is 0.693/(k1K).This plot, shown in fig. 4, gave the values of 0.0258 & 0.001 3 dm-3 pmol for k , and 3.22 0.09 pmol min-l dmP3 for K . It is noted that the value for k , obtained from these kinetic data agrees, within experimental error, with k, obtained from the equilibrium data, and supports the view that the k,.parameter obtained from the photocatalytic destruction rates is a genuine reflection of the adsorption characteristics of the molecule. The numbers obtained for k, and K were substituted in eqn (2) for different values of the initial methylene blue concentration to calculate the curves passing through the data points of fig.1. The agreement between the k, values in the two analyses begs the question in regard 4421296 Adsorption of Methylene Blue on Thin Films I I I I I ' I 24 - 22 - 20 - 18 - 4 \ $ 16 - 14 - 12 - 10 - [methylene blue] /mol dm-3 x 1 O6 Fig. 4. Half-life of photocatalytic destruction rate us. initial methylene blue concentration from data of fig. 1. to the relationship between k , and K . The k, parameter is related to the maximum coverage at the surface at equilibrium; K depends on the reaction rate of molecules adsorbed at the surface. It is suggested that K = k2q5NT, (4) where q5 is the quantum yield for the destruction of the solute, N is the total number of photons absorbed by the photocatalyst, and is the rate of transport of solute molecules to the surface.Flow Rate It was previously observed that increasing the flow rate through the illuminated spiral markedly increased the photocatalytic oxidation rate of several solutes. 2o Fig. 5 shows the effect of increasing the flow rate with methylene blue as the solute. A similar marked increase in the rate of disappearance with increasing flow rate also occurred. The apparent first-order decay constants obtained by regression analysis of the linear portion of the curves are given in table 2 together with the corresponding values calculated for 500 cm3 of solution. These calculated apparent first-order constants are shown plotted against flow rate in fig. 6 . It is noted that the increased k' values appear to be approaching a limiting value at high flow rates suggesting that the relationship may be described by an expression of the form where k* is the limiting value of k' at high flow rates, FR is the flow rate, and /3 is a constant for the system.The line drawn through the experimental points was obtainedR. W. Matthews 1297 1 I I I I illumination time/min Fig. 5. Disappearance of methylene blue from 500 cm3, 10 pmol dmV3 solutions recirculated through the TiO, photoreactor at different flow rates (cm3 min-I): 0 , 3 0 ; a, 60; @,90; a, 120; 0, 180; 0, 240. (20 W lamp.) Table 2. Effect of flow rate on apparent first-order decay constant of methylene blue. 200 cm3 of 10 pmol dm-3 solution, 20 W lamp flow rate /cm3 min-’ k’lmin-l k’lmin-l t,.,a/min 30 0.090 f 0.0009 0.036 19.2 60 0.152 & 0.0008 0.061 11.4 90 0.173 f0.0006 0.069 10.0 120 0.2 1 9 f 0.00 1 6 0.088 7.9 180 0.236 f 0.0021 0.094 7.4 240 0.275 f 0.0006 0.1 10 6.3 300 0.292 & 0.0004 0.1 17 5.9 a Calculated for 500 cm3 of 10pmol dm-3 solution.It was confirmed experimentally that k’ was directly proportional to the ratio of solution volumes. from eqn ( 5 ) by the method of least squares, with numerical values of 0.0101 +0.0012 min ml-l and 0.154 0.007 min-’ for /3 and k*, respectively. It is evident from eqn (2) and the apparent first-order decay approximation that, at low initial concentrations of PI0 P I solute, In- x k, Kt x k‘t. The approximate equivalence of k’ and k , K has been noted elsewhere.22 Therefore,1298 Adsorption of Methylene Blue on Thin Films 0.12 0 .lo 0.08 - I .s E 0.06 0.04 0.02 0.00 I I I I I I 100 2 00 300 flow rate/cm3 min-I Fig.6. Apparent first-order destruction rate us. flow rate. Data from fig. 5 +data for 300 cm3 min-'. k, K z k' or k* at high flow rates. The limiting value of K for methylene blue at low concentrations and high flow rates is therefore 5.97 f 0.56 pmol dm-3 min-l (0.154/0.0258). One notable difference between the present and previous results with salicylic acid20 is the absence of any downward curvatures in the semilog plots in fig. 5 . A pronounced downward curvature occurred when salicylic acid was the solute, especially at high flow rates. Those curves are well described by an expression of the same form as eqn (2). The reason for the absence of downward curvature with methylene blue is probably related to the adsorbed intermediate decomposition products competing more strongly with methylene blue for the primary oxidizing species than in the case of salicylic acid.Quantum Yield Actinometry on the lamp and spiral was done by circulating 0.006 mol dmP3 potassium ferrioxalate through the uncoated spiral from the reservoir. The analytical method of Hatchard and Parker23 was used to measure the rate of ferrous ion formation, 833 pmol min-l dm-3. According to specifications given by the lamp manufacturer a broad band of radiation is emitted with a maximum at 350 nm; very little radiation is emitted at a wavelength greater than 400 nm (ca. 5%). The quantum yield $(Fe2+) was therefore taken to be 1.2824 and the quantum yield for methylene blue disappearance was calculated to be 0.0092 for the limiting value k*, assuming that the same number of photons were absorbed by the TiO, layer as by the ferrioxalate solution.Solar Illumination The disappearance of methylene blue from 500 cm3 of an initially 10 pmol dm-3 solution with solar illumination in the parabolic trough is shown in fig. 7. The solution wasR. W. Matthews 1299 illumination time/min Fig. 7. Comparison of 20 W fluorescent lamp and solar illumination for methylene blue disappearance and salicylic acid disappearance with 20 W illumination. Flow rate 100 cm3 min-'. 500 cm3 of solution. (>, methylene blue, 20 W lamp, no TiO, in spiral; 0, methylene blue, 20 W lamp; a, methylene blue, 0.25 m2 parabolic trough, sunlight; 0 , salicylic acid, 20 W lamp.circulated through the spiral at 100 cm3 min-'. The k' value from the first-order plot was 0.109+0.002 min-' and the corresponding to.5 value was 6.35 min. For the purposes of comparison, the first-order plot for the disappearance under the same conditions is also shown but using the 20 W U.V. fluorescent tube as the illumination source. In this case the k' value was 0.068 k0.002 min-l and the 20.5 value, 10.2 min. Since the effective area of the parabolic trough illumination was 0.25 m2, it follows that illumination from 1 m2 at the same intensity at the surface of the TiO, would be equivalent to 128 W (electrical) illumination for a lamp having the same radiant output as the NEC lamp (20 YO emitted as radiation of wavelength < 400 nm). That is, the photoactivating power from 1 m2 of direct sunlight is approximately equivalent to 128 W of ' blacklight ' radiation.Assuming that light of a wavelength shorter than the bandgap of anatase (380 nm) photoactivates the catalyst, the result indicates that of the total 900 W m-2,19 ca. 2.8 YO has a wavelength shorter than 380 nm, in reasonable agreement with the data given by Ohta.25 Comparison with Salicylic Acid Also shown in fig. 7 is the destruction of 500 cm3 of 10pmol dm-3 salicylic acid circulated through the spiral at a flow rate of 100 cm3 min-' and illuminated with the 20 W lamp. The k' value was determined to be 0.117 f 0.001 min-', and 5.92 min. That is, under these conditions, salicylic acid was decomposed 1.72 times faster than methylene blue.This experiment also provides the nexus between methylene blue and the1300 Adsorption of Methylene Blue on Thin Films 100 90 80 - 70 5 60 g 50 40 30 20 10 0 E 3 i! P) P) I illumination time/min Fig. 8. Comparison of methylene blue disappearance and CO, appearance from 40 cm3 10 pmol dm-3 solution at pH 3. Flow rate 90 cm3 min-', 20 W lamp. other solutes in ref. (20). Measurements taken with methylene blue as the reference solute in other thin layer TiO, photoreactors could be used to estimate rates of disappearance of these other solutes. Carbon Dioxide Formation The formation of carbon dioxide was studied using solutions at pH 3. The results for 40 cm3 of 10 pmol dm-3 methylene blue circulated through the spiral at ca. 90 cm3 min-' and illuminated with the 20 W lamp are shown in fig.8. Also shown in fig. 8 is the disappearance of methylene blue from a 10 pmol dm-3 solution at pH 3 measured by the change in optical absorbance at 660nm. The solution volume was 500cm3 and the points shown in fig. 8 were the calculated times for the same decrease in concentration to occur in 40 cm3 of solution. It is seen that the rate of CO, formation lags considerably behind the rate of methylene blue disappearance. The to.5 value for CO, formation is ca. 4 min, four times slower than the rate of methylene blue disappearance (to.5 = 1 min). The importance of intermediate decomposition products as major precursors of CO, is thus clearly established. It is noted that total mineralization occurs at ca. 14 min. The observation of 100% conversion to CO, is consistent with reports of total mineralization for other organics photocatalytically oxidized by u.v.-illuminated TiO,.l, 3-9,12,14-17,26,27 Th e present results are consistent with the reaction stoichiometry of eqn (7). C,,H,,N,SCl+ 25.50, --+ 16C0, + 6H,O + 3HN0, + H,SO, + HC1.(7) It can be seen that in a closed system, with no air space, the total oxidation of a 10 pmolR. W. Matthews 1301 dmP3 methylene blue solution would exhaust the ambient oxygen concentration of an initially air-equilibrated solution (ca. 250 pmol dm-3).28 In the present experiments on the destruction of methylene blue, the solution was circulated through a reservoir open to the atmosphere and the oxygen depleted on passage through the reactor replenished with atmospheric oxygen. One of the referees has suggested that the increase in destruction rate may be caused by the possibly higher steady-state concentration of oxygen present in the solution at higher flow rates which would give rise to a higher concentration of adsorbed oxygen.The oxygen concentration is an important parameter in determining the rate of CO, formation in the photocatalytic oxidation of some solutes in free suspensions of Ti0,21(a) and would also be of importance in the present experiments but the extent of the inhibition by oxygen depletion is unknown. The transport of both oxygen and methylene blue to the photocatalyst surface are expected to be important factors in explaining the flow rate dependence in photocatalytic reactors of this type. Conclusions Aqueous solutions of methylene blue can be totally oxidized photocatalytically by contact with a thin layer of titanium dioxide illuminated with near-u.v.light. The rate of disappearance of methylene blue from the solution followed by spectrophotometric measurements at 660 nm obeys first-order kinetics. The decrease in apparent first-order rate constant, k’, with increasing solute concentration can be explained in terms of the integrated form of the Langmuir expression. One of the two parameters extracted from the analysis of the kinetic data agrees well with the adsorption affinity parameter obtained from equilibrium data on the non-illuminated system. The flow rate has a marked effect on k’; a near linear increase in k’ with flow rate occurs at low rates, but a limiting value is approached at high rates.Light from a ‘ blacklight’ fluorescent tube is a good photoactivating source for the titanium dioxide photocatalyst, but sunlight may also be used. The data obtained allow a quantitative comparison between the electrical power requirements using artificial light with direct sunlight. Measurements on the rate of disappearance of salicylic acid in the same photoreactor allow the methylene blue results to be related to other solutes. Hence methylene blue solutions could be used as a reference for other photoreactors of this type. The rate of CO, formation is ca. four times slower than the rate of methylene blue disappearance indicating the importance of intermediate decomposition products as precursors for CO, formation.Thanks are due to Cuno Pacific Pty. Ltd for financial assistance and to Mr Fred J. Fryer for technical assistance. References 1 J. H. Carey, J. Lawrence and H. M. Tosine, Bull. Environ. Conram. Toxicol., 1976, 16, 697. 2 B. G. Oliver, E. G. Cosgrove and J. H. Carey, Environ. Sci. Technol., 1979, 13, 1075. 3 A. L. Pruden and D. F. Ollis, Environ. Sci. Technol., 1983, 17, 628. 4 C-Y. Hsiao, C-L. Lee and D. F. Ollis, J. Caral., 1983, 82, 418. 5 A. L. Pruden and D. F. Ollis, J. Card., 1983, 82, 404. 6 D. F. Ollis, C-Y. Hsiao, L. Budiman and C-L. Lee, J . Curd., 1984, 88, 89. 7 S. Ahmed and D. F. Ollis, Solar Energy, 1984, 32, 597. 8 T. Nguyen and D. F. Ollis, J . Phys. Chem., 1984, 88, 3386. 9 M. Barbeni, E. Pramauro, E. Pelizzetti, E. Borgarello, M. Gratzel and N.Serpone, Nouv. J. Chim., 1984, 8, 547. 10 H. Kawaguchi, Environ. Tech. Lett., 1984, 5 , 471. 1 1 K-I. Okamoto, Y. Yamamoto, H. Tanaka, M. Tanaka and A. Itaya, Bull. Chem. SOC. Jpn, 1985, 58, 12 M. Barbeni, E. Pramauro, E. Pelizzetti, E. Borgarello and N. Serpone, Chemosphere, 1985, 14, 195. 13 H. Hidaka, H. Kubota, M. Gratzel, N. Serpone and E. Pelizzetti, N o w . J . Chim., 1985, 9, 67. 2015; 2023.1302 Adsorption of Methylene Blue on Thin Films 14 M. Barbeni, E. Pramauro, E. Pelizzetti, E. Borgarello, N. Serpone and M. Jamieson, Chemosphere, 15 R. W. Matthews, Water Res., 1986, 20, 569; J. Catal., 1986, 97, 565. 16 C. K. Gratzel, M. Jirousek and M. Gratzel, J. Molec. Catal., 1987, 39, 347. 17 K. Harada, T. Hisanaga and K. Tanaka, New J. Chem., 1987, 11, 597. 18 N. Serpone, E. Borgarello, R. Harris, P. Cahill, M. Borgarello and E. Pelizzetti, Solar Energy Mat., 19 R. W. Matthews, Solar Energy, 1987, 38, 405. 20 R. W. Matthews, J. Phys. Chem., 1987, 91, 3328. 21 R. W. Matthews, (a) Aust. J. Chem., 1987, 40, 667; (b) J. Catal., 1988, 111, 264. 22 D. F. Ollis, Environ. Sci. Technol., 1985, 19, 480. 23 C. G. Hatchard and C. A. Parker, Proc. Roy. SOC. A , 1956, 235, 518. 24 J. N. Demas, W. D. Bowman, E. F. Zalewski and R. A. Velapoldi, J. Phys. Chem., 198 1, 85, 2766. 25 T. Ohta, Solar-Hydrogen Energy Systems, ed. T. Ohta (Pergamon Press, Oxford, 1979). 26 E. Pelizzetti, M. Barbeni, E. Pramauro, N. Serpone, E. Borgarello, M. A. Jamieson and H. Hidaka, 27 M. Barbeni, M. Morello, E. Pramauro, E. Pelizzetti, M. Vincenti, E. Borgarello and N. Serpone, 28 Handbook of Chemistry and Physics (The Chemical Rubber Publishing Co., 42nd edn, 1960) p. 1707. 1986, 15, 1913. 1986, 14, 121. Chim. Ind., 1985, 67, 623. Chemosphere, 1987, 16, 1165. Paper 8/00921J; Received 23rd May, 1988

ISSN:0300-9599    

DOI:10.1039/F19898501291

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

J . Chem. SOC., Furudaj) Trans. I , 1989, 85(6), 1303-1313 Excess Molar Enthalpies of Steam-n-Hexane and Steam-n-Heptane up to 698.2 K and 12.6 MPa Nabil Al-Bizreh, Charles N. Colling, Neil M. Lancaster and Christopher J. Wormald* Department of Physical Chemistry, University of Bristol, Bristol BS8 1 TS Flow calorimetric measurements of the excess molar enthalpy, H:, of [xH,O + (1 - x)C,H,,] and [xH,O + (1 - x)C,H,,] are reported. The meas- urements extend over the range 448.2-698.2 K at pressures up to 12.6 MPa. Most of the measurements are at x = 0.5 but the composition dependence was investigated under selected conditions. Whereas previous measurements of HZ for steam with C,-C, n-alkanes at pressures up to 12 MPa could be fitted using the virial equation of state truncated after the third term, the new measurements can only be fitted at lower pressures.This is shown to be because the residual molar enthalpies of n-hexane and n-heptane are much larger than those of the fluids previously studied. Graphs of H: against pressure for (0.5H20 + 0.5C,H,,) exhibit maxima at pressures well below the saturation pressure. It is demonstrated that the maxima are due to the shape of the residual molar enthalpy curves of steam and n-heptane and not to any unusual molecular interaction. Residual molar enthalpies of (0.5H20 + 0.5C7H,,) calculated from the H: measurements lie on smooth monotonic curves. Comparison of the (0.5H20 + 0.5C7H,,) HZ measure- ments with the Peng-Robinson and Patel-Teja equations of state show that neither cubic equation fits the measurements.In previous papers we reported measurements of the excess molar enthalpy HE of mixtures containing steam made using two flow-mixing calorimeters of different design. A low pressure differential flow-mixing calorimeter' has been used to make measurements in the range 353.2-423.2 K at pressures around atmospheric. The mixtures studied include steam-hydrogen,' -nitrogen,, -argon,3 -carbon m~noxide,~ -carbon d i ~ x i d e , ~ C,-C, n-alkane~,~-~ +thene,5 -propene,6 -benzene,, xyclohexane,' -chl~romethane,~ -chloroethaneg and -trichloromethane.'o Analysis of the measure- ments showedg that for pairwise interactions, where there are no specific forces, cross- term second virial coefficients B,, can be calculated using the Stockmayer potential parameters E/kB = 233 K, 0 = 0.312 nm and t* = 1.238 for water, together with the combining rule (1) E l , = r ( E I ' E 2 , ) ~ where 2 ( 4 rJ$ (I' 1,): <=---------- 4 2 (I' +I,> and I is the ionisation energy.In the analysis of HE measurements on steam-argon it was noted3 that B12s calculated using Stockmayer parameters for fluoromethane are not very different from B12s calculated using the above parameters for water. Fluoromethane has the same dipole moment (1.85 Dt) as water, but a smaller reduced dipole moment Using fluoromethane as a homomorph for water in non-associative interactions we suggested a combining rule' which allows the calculation of cross-term third virial t 1 D = 3.33564 x C m. t* = 1.044. 13031304 H: of Steam-n- Hexane and Steam-n-Heptane coefficients from the corresponding states correlation of Orbey and Vera." The combining rule was tested using HE measurements on steam-n-pentane at tem- peratures up to 698.2K and pressures up to 14.0MPa obtained using a high-pressure flow-mixing calorimeter.The experimental HZs were compared with H:s calculated from the virial equation of state using virial coefficients B, C and D for steam, which are consistent with the 1984 NBS/NRC Steam Tables'' of Haar, Gallagher and Kell (HGK), the virial coefficient B for n-pentane calculated from the Kihara potential and C calculated from the Orbey-Vera correlation, the cross-term B,, from eqn (1) and (2), and cross-terms Cll2 and C,,, from our suggested combining rule. Agreement with experiment was to within the uncertainty on the measurements over the full range of temperature and pressure.The virial equation of state also fits all the other HE measurements made at temperatures up to 698.2 K and pressures around 12 MPa. These include mixtures of ~team-nitrogen,'~ -hydrogen,14 -C,-C, n-alkanes,15-17 <thene,I6 -carbon monoxide," and +arbon dioxide.', That the truncated virial equation of state fits the H: measurements on these mixtures so well deserves some comment. It might be expected that the equation would fit the residual volume and enthalpy of a fluid up to moderate densities, but it evidently fits the HZ measurements up to ca. 90% of the saturation pressure of water. This is best explained by examining each of the three terms to which Hg is related: Here HZ is the residual molar enthalpy of the mixture; Hil and H:, are the residual molar enthalpies of steam and n-pentane. For all the mixtures studied so far the maximum pressure to which we have been able to make vapour-phase HE measurements is the saturation pressure of water.Fig. 1 ( a ) shows the vapour pressure of water and of the C,-C, n-alkanes; the curves terminate at critical points. The figure shows that for all measurements on C,-C, n-alkanes and for measurements on (0.5H,O + 0.5C,H1,) above 473.2 K the n-alkane was a supercritical fluid. The figure also shows that at 448.2 K the saturation pressure of water is 0.9 MPa and that of n-pentane is 2.3 MPa. An H: measurement made at 90 % of the saturation pressure of water (0.81 MPa) is therefore at a pressure which is (0.81/2.3) x 100 = 35 Yo of the saturation pressure of n-pentane.For steam, Gallagherg has calculated virial coefficients B, C and D, which are a good fit to the 1984 HGK steam stable residual enthalpies at pressures up to ca. 90% of the saturation pressure, so that H:, in eqn (3) is not significantly in error. For n-pentane at 448.2 K only B and C are known, but at pressures up to 35 YO of saturation two virial coefficients give HZ2 with adequate accuracy. Comparison of H:, for n-pentane at 448.2 K calculated using B and C terms with H:, calculated from the BWRS equation of state shows that up to 2 MPa the difference is negligible. For (0.5H,O + 0.5C,H1,) at 448.2 K the saturation pressure is the dew curve, and this may be crudely approximated to the mean saturation pressure (1.6 MPa) of n-pentane and water.The mixture at 0.8 1 MPa is therefore at ca. 50 YO of the saturation pressure, and H: can be calculated from the mixture virial coefficients with good accuracy. In the temperature range 473.2-698.2 K n-pentane is a supercritical fluid, and comparison with the BWRS equation shows that the virial equation of state with coefficients B and C fits H:, up to pressures around 10 MPa. Using the equation of state described in the following paper the maximum density at which HE measurements on (0,5H,0+0.5C,H1,) were made was calculated to be 0.255 g cm-3 at 598.2 K and 14 MPa. This is ca. 2/3 of the critical density of the mixture. It is worth noting that over the whole 448.2-698.2 K temperature range, HZl of steam is very close to that of n-butane.This can be seen from fig. 1 (b) where HZ, is compared with H:, for the C,-C, n-alkanes at 598.2 K. At this temperature HZs are large and have been measured at pressures close to the saturation pressure 12.06 MPa of steam.N . Al-Bizreh, C. N . Colling, N . M. Lancaster and C. J. Worrnald 1305 573 473 T/K 3 73 0 4 a p l W a 12 Fig. 1. (a) Vapour pressure curves for C,-C, n-alkanes and for water, (-) the vapour pressure of n-alkanes,lS (----) the vapour pressure of water.12 Arrows at the top of the figure indicate temperatures at which the H: measurements were made. (b) Residual molar enthalpies H: at 598.2K of C,-C, n-alkanes and of water. (-) H: of n-alkane calculated from the BWRS equation of state, (----) H: of water." Fig.l ( a ) shows that for steam-n-heptane the maximum pressure at which vapour-phase HZ measurements can be made is the saturation pressure of n-heptane. At 498.2 K and 1.25 MPa steam entering the calorimeter is at 50% of its saturation pressure, but n-heptane is at 85 % of saturation. Fig. 1 (b) shows that H:* of n-heptane is ca. four times bigger than HZ1 of steam. We might therefore expect measurements of HE for steam-n-heptane to be fitted using virial coefficients B and C for n-heptane up to such pressures as the virial equation fails to fit H:2 of n-heptane. We now report HE measurements on steam-n-hexane and steam-n-heptane. Experimental The n-hexane, initially 95 mol% pure n-C,H,,, was treated with oleum to remove benzene, washed with alkali, dried with molecular sieve, and distilled twice.The density1306 HE of Steam-n- Hexane and Steam-n- Heptane Table 1. Experimental excess molar enthalpies H: of (0.5H20 + 0.5C6H1,) and (0.5H20 + 0.5C,H16) measured over a range of pressure 448.3 473.2 498.2 523.2 548.2 573.2 598.2 623.2 648,2 673.2 698.2 448.2 473.2 498.2 523.2 548.2 573.2 598.2 648.2 698.2 HZ HZ HZ HZ HZ p/MPa /J mol-' p/MPa /J mo1-I p/MPa /J mol-1 p/MPa /J rno1-I p/MPa /J mol-' - 0.60 0.76 0.73 1.10 1.18 1.79 1.28 1.98 2.07 2.86 2.01 2.66 3.47 1.13 1.98 2.67 1.03 1.79 2.67 1.05 1.45 1.74 1.12 1.86 2.74 0.4 1 0.38 0.40 0.43 0.79 0.69 1.38 0.86 1.44 0.79 1.44 1.13 2.13 1.10 2.17 466 546 399 65 1 603 1061 574 953 853 1326 748 1037 1427 372 659 908 30 1 536 786 234 334 414 255 365 603 382 323 294 259 507 369 859 389 773 302 649 404 805 288 617 0.76 0.95 1.46 2.62 2.72 3.38 3.64 4.34 4.20 4.98 5.48 3.45 4.1 1 4.91 3.45 4.10 4.87 2.55 3.36 4.1 1 3.39 4.1 1 4.78 0.79 0.79 0.76 1.48 2.17 2.48 3.02 2.84 4.24 2.79 4.24 4.13 6.41 4.27 6.36 (0.5H20 + 0.5C6H1,) 605 695 996 1964 1563 2259 1888 2512 1865 2329 2637 1199 1534 1844 1034 1287 1513 654 870 1115 79 1 987 1160 1.16 1.79 3.17 4.00 4.25 4.9 1 5.40 6.14 6.95 5.58 6.17 6.65 5.75 6.07 6.47 4.87 5.54 6.18 5.42 6.15 6.94 8 60 1336 2926 3080 3443 3086 3449 3082 3459 2168 2388 2666 1862 2022 2171 1352 1584 1746 1328 1526 1757 (0.5H20 + 0.5C,H16) 13515 1.0 13400 792 572 1.13 1000 1160 2.48 9453 2509 2.82 9430 2206 3.55 7242 4729 4.58 7654 2070 5.62 6271 5014 6.00 6259 1576 5.62 4784 3180 7.06 5279 1977 8.53 3937 3218 1373 8.44 2815 2160 1.31 2.26 3.45 4.56 5.00 6.1 1 6.79 7.47 8.34 7.06 7.64 8.29 7.12 7.89 8.58 6.89 7.61 8.25 7.62 8.41 9.14 1.65 3.20 4.58 7.10 7.68 8.44 10.6 10.6 1004 204 1 3879 3979 4548 4019 4352 3717 4013 2857 3056 3309 2373 2646 2830 1996 2191 2363 1901 2097 227 1 9304 9259 7657 6083 5301 5177 4066 3173 2.43 3.79 5.45 7.51 8.13 8.93 9.29 9.02 9.7 1 9.42 10.2 10.1 8.95 9.71 9.97 10.4 10.7 2.13 3.54 5.62 8.44 9.82 11.2 12.5 12.6 2484 5376 5065 4567 4716 4100 41 13 3509 3625 3683 3017 3 143 2536 2666 2807 2402 2520 10865 91 19 7246 5623 4890 4617 3923 3333 of the purified material was 654.98 kg mP3 at 298.1 5 K (literature :19 654.84 kg m-3).The purified material was not less than 99.9mo1°/0 n-C,H,,. The n-heptane, initially 99 mol% pure n-C7H16, was dried with molecular sieve and distilled twice.After purification the density was 697.25 kg m-3 at 298.15 K (literature?' 697.51 kg m-3). The purified n-heptane was not less than 99.8 mol% n-C7Hl,. Steam was generated from ordinary distilled water. Measurements of HZ were made using the apparatus describedN . Al-Bizreh, C. N . Colling, N . M. Lancaster and C. J. Wormald 1307 Table 2. Experimental excess molar enthalpies H: of [xH,O + (1 - x)C,H1,] and [xH,O + (1 - x)C,H,,] measured over a range of composition x HZ H; H; HZ H; x /Jrnol-' x /J mol-' x /Jmol-' x /Jmol-' x /J mol-l [xH,O + (1 - x)C,H1,1 T = 548.2 K 0.299 4296 0.402 4467 0.497 4249 0.602 3709 0.702 2950 T = 598.2 K 0.297 3732 0.399 4165 0.499 4200 0.600 3937 T = 648.2 K 0.303 2925 0.400 3249 0.500 3306 0.599 3136 0.702 2672 p = 4.93 MPa p = 9.41 MPa p = 11.48 MPa txH,O + (1 - x)C,H,,I T = 548.2 K T = 573.2 K T = 598.2 K p = 4.58 MPa p = 6.00 MPa p = 7.68 MPa 0.301 8428 0.400 8351 0.498 7672 0.600 6616 0.701 5318 0.301 6296 0.402 6531 0.499 6263 0.601 5587 0.701 4583 0.300 4780 0.398 5310 0.502 5298 0.601 4833 0.700 4061 4 L E 2 W E 2 2 n 523.2 I--" I 0 0 0 4 8 0 4 8 p/MPa Fig.2. (a) Excess molar enthalpies H; of (0.5H20+0.5C,H,,) plotted against pressure. 0, A, table 1. (-) calculated from the truncated virial equation of state as described in the text. (b) (HE/p) of (0.5H20 + 0.5C,H1,) plotted against pressure. The intercepts are values of the excess isothermal Joule-Thomson coefficient of the mixture given by eqn (6). 0, A, table 1. (-), calculated from the truncated virial equation of state.previo~s1y.l~ Hydrocarbon removed from the apparatus was analysed by g.1.c. to check for decomposition. For each mixture most of the HE measurements were made at x = 0.5, at temperatures up to 698.2 K, and at pressures up to ca. 12 MPa. These measurements are listed in table 1. Measurements were also made over a range of mole fraction x at selected pressures, and these are listed in table 2. Results of the1308 Hz of Steam-n- Hexane and Steam-n- Heptane 8 ‘-473.2 ‘-498.2 p 5 2 3 . 2 6 1 p 698.2 2 I 1 I 1 1 1 C 4 8 12 p/MPa Fig. 3. (a) Excess molar enthalpies H: of (0.5H,0+0.5C7H1,) plotted against pressure. 0, table 1 . (-), drawn with a flexicurve. (b) Excess molar enthalpies HZ of (0.5H,0+0.5C7H1,) at temperatures above the critical temperature 540.2 K of n-heptane.0, table 1 . (-), drawn with a flexicurve. (----), calculated from the truncated virial equation of state. I I X r I I I I 1 0 0.2 0.4 0.6 0.8 1 X Fig. 4. (a) Excess molar enthalpies H: of [xH,O + (1 - x)C6HI4]. (b) Excess molar enthalpies HZ of [xH,O+(l -x)C,H,,]. 0, A, 0, table 2. (-) fitted to the measurements by plotting H:/4x( 1 - x) against x. (----) calculated from the truncated virial equation of state.N . Al-Bizreh, C. N . Colling, N . M. Lancaster and C. J. Wormald 1309 measurements on (0.5H20+0.5C,H,,) are plotted against pressure in fig. 2, and on (0.5H20 + 0.5C,H1,) in fig. 3. Results of the measurements on [xH20 + (1 - x)C,H,,] are plotted against x in fig. 4(a), and on [xH,O+ (1 -x)C,H,,)] in fig.4(b). The accuracy of the steam-n-hexane HE measurements is 2 % at temperatures above the critical temperature 540.2 K of n-heptane, and between 2 and 4% below this temperature. Comparison with the Virial Equation of State It was shown previouslyg that the residual enthalpy H: of a fluid is given by H;(V, T) = I0*[ T($,)v-p]dV+pV-RT. (4) Using the virial equation of state in powers of the density truncated after the fourth term, HE is given by9 where $', w and A are related to the mixture virial coefficients B, C and D and their temperature derivatives. The virial coefficients Bll, C,,, and D,,,, for steam obtained by Gallaghers were fitted to seventh-order polynomials in powers of T-l. Coefficients of the polynomials have been published.20 Differentiation gave #yl, ylll, and Allll for steam.For n-hexane B,, and #i2 were calculated from the Kihara potential with the parameters a = 0.1329 nm, CT = 0.5458 nm, and ElkR = 932 K. For n-heptane the parameters a = 0.1526 nm, CT = 0.5769 nm and ElkB = 1057 K were used. These parameters fit class I virial coefficient measurements on the hydrocarbon.21 Third virial coefficients for the hydrocarbons were calculated from the correlation of Orbey and Vera." Cross-term second virial coefficients B,, and isothermal Joule-Thomson coefficients #;2 were calculated by combining the Kihara potential parameters for the n-alkanes with the parameters of the Stockmayer potential D = 0.3 12 nm, &/kB = 233 K and t* = I .238 for water.g Cross-terms E ~ , were calculated using eqn (1) and (2), a12 and o12 were taken to be the mole fraction weighted arithmetic mean of the pure-component parameters.Cross-term third virial coefficients C,,, and C,,, and their temperature derivatives were calculated from the Orbey-Vera'' correlation together with the combining rules given previously.20 In these combining rules critical parameters for fluoromethane were used, instead of water critical parameters, for those interactions where there was no hydrogen bonding. No cross-term fourth virial coefficients could be calculated. The first test of any set of HE measurements made at high pressures is to check that they are consistent with HEs obtained using our low-pressure flow-mixing calorimeter. At low densities eqn ( 5 ) can be written in the form' limp -+ 0 (HE/p) = x( 1 - x ) (2#y2 - #yl - #&).(6) Values of (HE/p) for (0.5H20 + 0.5C,Hl,) are plotted against pressure in fig. 2(b) where circles and triangles are used alternately to make the isotherms clear. Solid curves shown in the figure should be ignored for the moment. For the Hzs measured over the range 498.2-698.2 K extrapolation of (H:/p) to zero pressure was done graphically. The measurements at 448.2 K and 473.2 K are over too small a pressure range for meaningful extrapolation to be made, and were neglected. Values of limp -+ O(HE/p) are plotted against temperature in fig. 5, where they are compared with values obtained using the low-pressure mixing calorimeter. The upper points in fig. 5 were obtained from a similar analysis of the (0.5H20 + 0.5C,H1,) measurements.Solid curves shown in the figure were calculated from eqn (6) using the Kihara potential for the n-alkane and the Stockrnayer1310 H: of Steam-n- Hexane and Steam-n- Heptane 400 500 600 700 TIK Fig. 5. Zero pressure intercepts of graphs of (H:/p) against p plotted against temperature. The lower curve is for (0.5H,0+0.5C,Hl,) and the upper curve is for (0.5H,0+0.5C7Hl,). 0, from H: measurements obtained using our low-pressure flow mixing cal~rimeter.~-~ 0, from H z measurements listed in table 1. (-) calculated from the truncated virial equation of state. potential for steam with the parameters given above. Fig. 5 shows that the measurements from the low- and high-pressure calorimeters are consistent. The solid curves shown in fig. 2 are H:s for (0.5H20 + 0.5C,H14) calculated from eqn (5).Fig. 1 (a) shows that for this mixture n-hexane vapour entering the calorimeter was a supercritical fluid for the measurements in the range 523.2-698.2 K. The measurements at 498.2 K extend up to 2.43 MPa, 92% of the n-hexane saturation pressure, and these are quite well fitted by eqn (5). Fig. 2(b) shows that the fit to the measurements at pressures below 1.5 MPa is within the uncertainty on the measurements. At higher pressures the calculated HEs are smaller than the experimental values. Fig. 2 shows that the inadequacy of eqn ( 5 ) is most marked at 523.2, 548.2 and 573.2K. At higher temperatures the calculated HE values agree with experiment more closely. This is because the density of the mixture diminishes with increasing temperature, and the non- ideality becomes less.The Hzs for (0.5H20 + 0.5C7H,,) are shown in fig. 3. In fig. 3 (a) the solid curves were drawn with a flexicurve. The smooth curves between 548.2 and 698.2 K are all above the critical temperature 540.2 K of n-heptane. Below T, measurements were made at pressures up to and beyond the saturation pressure. At 448.2 K the saturation pressure is 0.6 MPa, and above this pressure n-heptane enters the calorimeter in the liquid phase. The vertical line drawn in the figure corresponds to the enthalpy of evaporation. The isotherms at 498.2 and 523.2K show smaller vertical sections as the enthalpy of evaporation diminishes with increasing temperature. The HZ values obtained by mixing supercritical n-heptane with subcritical steam are shown on a larger scale in fig.3 (b). The solid curves were again drawn with a flexicurve, and the broken curves were calculated from the virial equation of state as described above. As with the n-hexane mixtures the inadequacy of the virial equation is shown most clearly by the measurements at 548.2-648.2 K. At 698.2 K the fit is slightly better and extends to ca. 6 MPa, whereas at 548.2 K the measurements are fitted up to ca. 3 MPa. The HE measurements on [xH,O + (1 - x)C,HI4] are shown in fig. 4(a) and those on [xH20+(1-x)C7Hl,] are shown in fig. 4(b). Solid curves through the points were constructed by plotting the measurements on graphs of HE/4x(1 -x) against x and drawing the best line through the points. The broken curves were calculated from eqn (5).For both mixtures the calculated curves lie below those obtained experimentally, though the shape is correctly given.N . AI-Bizreh, C. N . Colling, N . M . Lancaster and C. J . Wormald 131 1 24 0 2 L 6 plMPa 0 2 4 6 0 - 0 4 8 12 plMPa Fig. 6. (a) The residual enthalpies H:, of water, HZ2 of n-heptane, and H: of (0.5H20 + 0.5C,H1,) at 548.2 K. (----) the mean of HZl and H:2. 0, calculated from H: listed in table 1, HZl from ref. (12), and HZ2 from the BWRS equation of state. The above terms are related by eqn (3). The figure shows the origin of the maxima in the H: against p graphs in fig. 3. (b) Residual molar enthalpies H: of (0.5H20 + 0.5C,Hl,) calculated using eqn (3). The striking feature of the HE values of (0.5H20+0.5C7Hl,) shown in fig. 3 is the maxima at pressures well below the saturation pressure of n-heptane.These are not due to unusual molecular interactions but are a consequence of the shape of the residual enthalpy curves of n-heptane and steam. This is best made clear by rearranging eqn (3) and calculating the molar residual enthalpy HZ of the mixture from HE and the pure- component residual enthalpies. In fig. 6(a) the broken line is the mean of H:, for steam calculated from the HGK equation of state at 548.2 K and HZ, for n-heptane calculated from the BWRS equation. Circles shown in the figure are values of HZ of (0.5H20 + 0.5C7H,,) calculated by adding the experimental HKs at 548.2 K listed in table 1 to the mean of the pure-component residual enthalpies. The circles lie on a smooth curve of similar shape to H:, for steam.The solid line through the points was drawn with a flexicurve. The difference between this line and the broken line is HE, and the origin of the maximum at 548.2 K in the HE against p graph shown in fig. 3 should now be apparent. HZ for the (0.5H20 + 0.5C7H1,) mixture at other temperatures is shown in fig. 6(b). Cubic Equations of State Robinson2, has used the Peng-Robinson (PR) equation of to calculate liquid/vapour equilibria of binary mixtures containing water and methanol. When their equation is used to calculate vapour phase HEs of steam-hydrocarbon mixtures poor agreement with experiment is found. The combining rule used for the calculation of the cross-term critical temperature q12 is Eqn (7) was used to obtain the cross-term parameter aI2 and hence the value of a for the mixture.Comparison of HEs calculated from the PR equation using < = 0.1 with measurements on (0.5H,0+0.5C7H,,) are shown in fig. 7(a). The value of was chosen to give a reasonable fit to the measurements at 598.2K, but this value does not fit the1312 HE of Steam-n-Hexane and Steam-n-Heptane Fig. 7. Excess molar enthalpies of (0.5H20+ 0.5C,H,,) compared with cubic equations of state. 0, table 1, (-) drawn with a flexicurve. (a) Broken curves calculated from the Peng-R~binson~~ equation of state using 5 = 0.1. (b) Broken curves calculated from the Patel-Teja2* equation of state using < = 0.4. measurements at 548.2 and 698.2 K, although the calculated curves have similar shapes to those obtained experimentally.The failure of the PR equation is partly due to the poor fit to the residual enthalpy of n-heptane and steam, and partly due to the failure of the combining rules to adequately reflect the nature of the water-hydrocarbon interaction. A cubic equation which gives a better fit to the residual enthalpy of the pure components is that of Pate1 and Teja24 (PT) who modified the PR equation by adding a parameter c. It has been shown that the PT equation fits recent measurements of the enthalpy of acetone25 to within a few per cent. Comparison with the HGK tables and the BWRS equation of state shows that the fit to the residual enthalpy of steam and of n-heptane is similar to that for acetone. The fit to HZ, of n-heptane is worst between the critical temperatures 540.2 K and 600 K, and improves with increasing temperature.HEs calculated from the PT equation using c = 0.4 are shown in fig. 6(b). The isotherms are close to those obtained experimentally, although the shape is clearly wrong. The value = 0.4 suggests that the water-hydrocarbon interaction is weaker than that corresponding to the calculated value of the cross-term a,, used in the PT equation. Any equation of state which is to fit the experimental water-hydrocarbon HEs must fit H: of the pure components and take the weak water-hydrocarbon interaction into account. References 1 C. J. Wormald, J. Chem. Thermodyn., 1977, 9, 901. 2 P. Richards, C. J. Wormald and T. K. Yerlett, J . Chem. Thermodyn., 1981, 13, 623. 3 P. Richards and C. J. Wormald, 2. Phys. Chem.N.F., 1981, 128, 35. 4 G. R. Smith and C. J. Wormald, J. Chem. Thermodyn., 1984, 16, 543. 5 N. M. Lancaster and C. J. Wormald, J. Chem. Thermodyn., 1985, 17, 295. 6 N. M. Lancaster and C. J. Wormald, J. Chem. Thermodyn., 1986, 18, 545. 7 G. R. Smith, M. J. Fahy and C. J. Wormald, J . Chem. Thermodyn., 1984, 16, 825.N . Al-Bizreh, C. N . Colling, N . M . Lancaster and C . J . Wormald 1313 8 C. J. Wormald and N. M. Lancater, J. Chem. Thermodyn., 1985, 17, 903. 9 C. J. Wormald and N. M. Lancaster, J. Chem. Soc., Faraday Trans. I , 1988, 84, 3141. 10 N. M. Lancaster and C. J. Wormald, Z . Phys. Chem. N.F., 1981, 128, 43. 11 H. Orbey and J. H. Vera, AZChE J., 1983, 29, 107. 12 L. Haar, J. S. Gallagher and G. S. Kell, NBSINRC Steam Tables (Hemisphere, New York, 1984). 13 C. J. Wormald and C. N. Colling, J . Chem. Thermodyn., 1983, 15, 725. 14 C. J. Wormald and C. N. Colling, J . Chem. Thermodyn., 1985, 17, 437. 15 C. J. Wormald and C. N. Colling, AIChE J., 1984, 30, 386. 16 N. M. Lancaster and C. J. Wormald, J . Chem. Thermodyn., 1987, 19, 89. 17 N. M. Lancaster and C. J. Wormald, J . Chem. Thermodyn., 1987, 19, 1001. 18 C. J. Wormald, N. M. Lancaster and A. J. Sellers, J . Chem. Thermodyn., 1986, 18, 135. 19 API Research Project No. 44., Selected tlalues of properties of hydrocarbons and related compounds 20 N. M. Lancaster and C. J. Wormald, J . Chem. Soc., Faruday Trans. 1, 1988, 84, 3159. 21 J. H. Dymond and E. B. Smith, The virial coeficients of Pure Gases and Mixtures (Clarendon Press, 22 D. B. Robinson, D-Y. Peng and S. Y-K. Chung, Fluid Phase Equilibria, 1985, 24, 25. 23 D-Y. Peng and D. B. Robinson, Znd. Eng. Chem. Fundam., 1976, 15, 59. 24 N. C. Pate1 and A. S. Teja, Chem. Eng. Sci., 1982, 37, 463. 25 T. K. Yerlett and C. J. Wormald, J . Chem. Thermodyn., 1986, 18, 371. (Texas A & M University, 1976). Oxford, 1980). Paper 8/00923F; Receiced 7th March, 1988

ISSN:0300-9599    

DOI:10.1039/F19898501303

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

J. Chem. Soc., Faraday Trans. 1, 1989, 85(6), 1315-1326 A Cubic Equation of State for Mixtures Containing Steam Christopher J. Wormald* and Neil M. Lancaster Department of Physical Chemistry, University of Bristol, Bristol BS8 ITS A cubic equation of state and combining rules for mixtures containing steam has been developed. The equation is based on the earlier equations of Clausius, Martin and Kubic, and contains two temperature-dependent parameters, which give it enough flexibility to fit the residual properties of both steam and non-polar fluids with sufficient accuracy. Mixture properties are calculated using pseudo-critical parameters for steam and a single temperature-independent interaction parameter { in the combining rule Tl2 = c(ql q,):. For mixtures such as steamxarbon dioxide, where there is a specific interaction between the unlike components, { must be obtained from cross-term second virial coefficients.Where there is no specific interaction is given by < = 2(K1 V,,): Kl2-l(I1 12)~(Il+12)-1. The equation of state fits measurements of the excess molar enthalpy H: of mixtures containing steam at pressures of 0.1-10 MPa, at temperatures up to at least 700 K, and allows the calculation of ( p , V,, T ) properties. Attempts to measure (p, Vm, T ) properties of mixtures containing steam using a compression apparatus have generally yielded results of low accuracy. Large adsorption errors spoil the measurements, particularly at low densities. An alternative to ( p , Vm, T ) is direct measurement of the excess molar enthalpy HZ using a flow mixing calorimeter.The calorimeter is sensitive only to changes in the enthalpy of the fluid passing through it; adsorbed molecules become part of the apparatus and have no effect on the measured enthalpy change. A low pressure differential flow-mixing calorimeter for HE measurements on dilute gases has been described.' The calorimeter has been used to make measurements on binary mixtures of steam-hydrogen,2 -nitrogeq3 -argon,4 -methane,2-C2-C8 n-alkane~,~~' -ethene,5 -propene,6 -benzene,' -cyclohexane,* -carbon mon~xide,~, -carbon d i ~ x i d e , ~ -chloromethane, lo -chloroethane, lo and -trichloro- methane. l1 Analysis of the measurements yielded cross-term second virial coefficients B12. The only other B,, values for mixtures of water-gaseous substance have been obtained from measurements of the solubility of water in compressed gas.l2?l3 For the few mixtures on which measurements have been made, agreement with B12s from the Hz measurements is within the combined experimental error. As the enthalpy of mixing work progressed and experience was gained, better analysis of the HE measurements became possible.As this paper draws heavily on conclusions reached during the course of the low-pressure measurements it is useful to summarise the relevant points before proceeding. For low density gaseous mixtures it has been shown1 that where #" = B - T(dB/dT), A$") is a function of B and $" for the mixture, and fly) is a function of third virial coefficient terms. If the pair potentials of components 1 and 2 are known B,, and &'2 can be obtained by adjusting the parameter in the combining rule E l 2 = &11&22>~ (2) 13151316 Equation of State for Steam Mixtures until the right-hand side of (1) agrees with the experimental HE.When the Stockmayer parameters ElkB = 233 K, CT = 0.312 nm and t* = 1.238 are used for water it was found’’ that 5 was given by where Z is the ionisation energy. Eqn (3) fits all the HZ measurements except those for steam-carbon dioxide, -ethene, -propene and -benzene. For these mixtures there are specific interactions between the water lone-pair electrons and unfilled orbitals of the other component. Eqn (3) has been shown to give a better fit to cross-term B12s for mixtures of n-alkanes than six other similar formulae.“ 3 l5 Experimental and calculated values of 5 for mixtures containing steam are listed in ref. (10). Where there are specific interactions the experimental values of 5 are bigger than those calculated from eqn (3). An alternative to pair potentials is to use a corresponding states correlation to obtain second virial coefficients and their temperature derivatives. For mixtures of n-alkanes it has been shown14* l5 that the experimental Hg values can be fitted using the correlation of McGlashan and Potter16 together with the combining rule where (5) As pair-potential parameters fitted to second virial coefficients vary according to the quality of the data, values of calculated from eqn (5) are slightly different from those calculated from eqn (3). For the 20 substances for which steam mixture measurements have been made, the average value of 5 calculated from eqn (3) is only 1.5 O/O greater than that calculated from eqn (5).For mixtures of water-Cl-C, n-alkanes the experimental HE values can be fitted using eqn (4) and (5) together with the McGlashan-Potter correlation and the pseudo-critical temperature T, = 230 K for water. V , for water was set equal to the true critical volume 55.5 cm3 mol-’ and the effective chain length was N = 1 . For other choices of pseudo- critical temperature, and N fit the Hgs equally well, our choice of 230 K for T, is not unique. In the analysis of the HZ measurements on water-argon it was noted4 that the HE values could be fitted to within experimental error using fluoromethane as a homomorph for water.The second virial coefficient of fluoromethane is close to that calculated from the McGlashan-Potter correlation using = 230 K for water, and that calculated using the Stockmayer potential with the parameter ElkB = 233 K, CT = 0.312 nm and t* = 1.238 for water. Fluoromethane has the same dipole moment 1.85 Dt as water, although a smaller reduced dipole moment t* = 1.044 and stronger dispersion forces. By using the fluoromethane parameters (T, = 318 K, pc = 5.87 MPa, = 124 cm3 mol-l, w = 0.19) as pseudo-critical parameters for water in the third virial coefficient correlation of Orbey and Vera17 it was shown’* that high-pressure HZ values for water-n-pentane could be fitted to within experimental error. The function fly) in eqn (1) was calculated in the same way.The use of fluoromethane as a homomorph for water was also a key step in the development of our first modell’ to fit the high-pressure enthalpy of mixing of steam-hydrocarbons. Enthalpies of mixing of several binary mixtures containing steam at high pressure have been measured using a flow-mixing calorimeter enclosed in a pressure vessel. Most of the measurements extend from 448.2 to 698.2 K at pressures up to ca. 12 MPa. The mixtures studied include steam-hydrogen,20 -nitrogen,21 -methane,22 -ethene,23 -pro- pane,24 -butane,24 -n-pentane,’, -n-he~ane,~~ -n-he~tane,~~ -carbon monoxide,26 xarbon dioxide.26 Small corrections to some of the measurements listed in ref. (20)-(24) t 1 D = 3.33564 x C m.C. J. Wormald and N . M. Lancaster 1317 and (26) have been published." Most of the measurements were made at x = 0.5, though some measurements of the composition dependence of HE were made under selected conditions for each mixture.At pressures up to ca. 12 MPa the HE values for mixtures containing n-alkanes up to n-pentane can be fitted using the virial equation of state." For steam-n-hexaneZ5 the virial equation fits the HE values only up to ca. 5 MPa and for ~team-n-heptane'~ it fits only up to ca. 2.5 MPa. Cubic equations of state, such as that of Peng and Robinson, are worse than the virial equation, failing to fit HZ measurements at both low and high den~ities.'~ It is clear that for steam mixtures another approach to the problem is needed. Previous Equations for Mixtures Containing Water Attempts to construct cubic equations of stzLc to fit the properties of mixtures containing water are aimed at fitting either liquid/vapour equilibria or at fitting vapour phase properties only.Robinson et a1." applied the Peng-Robinson equation to phase equilibria of binary mixtures containing water and methanol. When their equation is used to calculate HE values for steam-hydrocarbon mixtures poor agreement with experiment is found. Baumgartner et al. proposed an association model" based on the cubic equation of Schmidt and W e n ~ e l . ~ ~ It requires the simultaneous solution of equations for the chemical equilibria of water oligomers containing up to 14 molecules, and is almost intractable for mixture calculations. Ghemling et aL3' extended the perturbed hard-chain equation of Donohue and Prausnitz3' to include dimerisation equilibria for polar molecules.Their equation is used for industrial calculations of liquid/vapour equilibria at high pressures. The equation is not cubic and calculations are complicated. In common with the other equations described above, the adjustable parameter kij must be fitted to experimental data. Cubic equations of state constructed specifically for water vapour have been proposed by de Santis et al.,33 Nakamura et al.,34 Oellrich et Nishida et al.,36 and Wormald." The perturbed hard-sphere equations of Nakamura et al., Oellrich et al., and Nishida et al. give a good representation of the ( p , Vm, T ) properties of steam at high temperature and pressure by using the Carnahan-Starling expression for the hard-sphere compressibility factor and making the covolume b temperature dependent.The first two of this set of equations of state give a poor fit to B(T) for steam; the latter gives B(T) accurately but requires three temperature-dependent parameters. Only a few fluids have been correlated with the equations and they have not been used for calculating mixture properties. The equation of de Santis et al. gives a reasonable fit to B(T) and to steam densities at pressures up to 220 MPa, but the form of a ( T ) is non-analytic. Enthalpy calculations require a( T ) to be a continuous function. Their equation can be used for the calculation of mixture volumes and has the advantage that, given the parameters for steam, no further information is needed to determine the mixture volume, except when there is a specific interaction.Our previous equation of statel' for steam mixture properties was a modification of the Peng-Robinson equation in which the a(T) term was separated into a part due to London and Keesom forces and a part due to hydrogen bonding. This gave a reasonable fit to the residual enthalpy of steam at moderate densities, and provided a way of calculating cross-terms similar to the association treatment of Woolley3' and Lambe~t.~' To improve the fit to both low-pressure and high-pressure HE measurements a temperature-dependent interaction parameter calculated from the second virial coefficient of the mixture was used. To fit the properties of mixtures in which there are specific interactions an additional parameter was needed.The equation fits the H:s of water-C,-C, n-alkanes up to 10 MPa quite well, but fails at higher pressures. We have now developed an improved equation of state for the vapour phase properties of steam mixtures which largely overcomes these limitations.1318 Equation of State for Steam Mixtures A Cubic Equation for Mixtures Containing Steam Any equation of state designed to fit the properties of steam mixtures must first fit the residual properties of the pure components adequately. As the volumetric properties of steam are very different from those of say n-hexane, much is demanded of an equation if it is to fit the properties of both fluids. The Soave-Redlich-Kwong, Peng-Robinson and Patel-Teja equations of state were developed primarily for liquid/vapour equilibrium calculations, and do not give as good a representation of volumetric properties of the vapour as does the Martin3’ equation which was developed specifically for this purpose.Kubic4’ has modified the Martin equation for the calculation of liquid/vapour equilibria. After much experimentation with recent cubic equations we have also taken the Martin equation as a starting point, and have used some of Kubic’s modifications in the development of our equation for steam mixture properties. Martin made a detailed analysis of the ability of two-term cubic equations to predict densities of liquids and gases. He concluded that the Clausius41 form was both the best and the simplest, and modified it by making the parameter a temperature dependent: RT a(T) P=(V-b)-(V+c)2. The virial expansion of eqn (6) is 2ca(T) 1 3c2a(T) 1 Z = l + ( b - - s)) ++ ( b2 +-&) v2+ ( b3 -T) v3.(7) Kubic’s modification was to make c temperature dependent and to equate the second virial coefficient [b - a( T)/ RT] to that given by the Tsonopoulos c~rrelation,~~ which gives adequate values of B(T) for non-polar and slightly polar fluids and so makes the parameter a temperature dependent. For mixtures, the term [b - a( T)/ RT] is equated to the second virial coefficient of the mixture where cross-terms are calculated from combining rules for critical parameters. Kubic’s formulae for the coefficients of eqn (6) are RT, b = - (0.082 - 0.07 1 3 ~ ’ ) Pc (9) The coefficients a’, al, yo and yl are polynomials in powers of the reduced temperature a’= -0.1514T,+0.7895+0.3314~1+0.029~2+0.0015~7 (1 1) ul= -0.237T,-0.7846c1 + 1.0026c2 +0.019c7 (12) yo = 4.275 - 8.879c’ + 8.509c2 - 3.481 c3 + 0.576c4 (1 3) yl= 12.86-34.74c1+ 37.43c2- 18.06r3+3.51C4.(14) (15) T* The parameter m’ is related to Pitzer’s acentric factor o through W’ = 0.000756 + 0.90980 +O. 16230~ +0.1455w3. The residual molar enthalpy H: is given by [a- T(da/dT)] aT(dc/dT) H: = pV- RT- - (V+C) (V+ c)2 -C. J. Wormald and N. M. Lancaster 1319 2 * E X -1 21 1 I + € X - 0 4 8 12 16 20 0 4 8 12 plMPa p l W a Fig. 1. (a) The fit to the residual molar enthalpy HE of steam obtained using eqn (18) and (16). (-), HZ from the 1984 HGK43 Steam Tables. (----) calculated from eqn (1 8) and ( I 6). (b) The fit to the residual molar enthalpy HZ of n-hexane obtained using eqn (10) and (16).0, experimental measurements of Wonnald and Yerlett.45 (-) HZ from eqn (10) and (16). (----) calculated from the BWRS equations of state. (- .---) calculated from the Patel-Teja46 equation of state. Making the a and c terms of eqn (6) temperature dependent gives the equation considerable flexibility. To fit the properties of steam we calculated a( T ) from the second virial coefficient B(T) of steam a(T) = RT[b-B(T)]. (17) B(T) was calculated from a polynomial in powers of T1 fitted to second virial coefficients calculated by Gallagher'' which are consistent with the Haar-Gallagher-Kell (HGK) equation of These virial coefficients agree very closely with those given by the correlation of L e F e ~ r e . ~ ~ The temperature dependence of c(T) was chosen to make saturated vapour densities calculated from eqn (6) agree with those obtained from the HGK equation for steam from the normal boiling temperature to the critical temperature : With these modifications, eqn (6) fits the molar volume of steam with good accuracy up to 20 MPa.It is not possible to fit the molar volume and the saturated vapour pressure of steam simultaneously. When c( T ) was fitted using saturated vapour pressures, poor values of Vm and Wm were obtained. The fit to the residual enthalpy HE of steam obtained using (1 7) and (18) is shown in fig. 1 (a). The fit to H z for n - h e ~ a n e ~ ~ given by Kubic's formulae is shown in fig. 1 (b). At low densities Kubic's equations are superior to the Patel-TeJa46 equation and the multi-parameter BWRS equation. c(T) = 1220.7- 3656.5T1 +4043.8T2- 1847.5T3+252.3T4.(18) Combining Rules for Mixtures We use combining rules for a( T ) and b that are consistent with the statistical-mechanical expression for the second virial coefficient of a mixture of components 1 and 2: a( T ) = xt a,, + 2x, x, al, + xi a,, b = X; b,, + 2x, X, b,, + xi b,,. (19) (20)1320 To calculate c for the mixture we used Equation of State for Steam Mixtures c = x , c, + x, c,. (21) Eqn (21) was recommended by Peneloux and Rauzy4’ for three-parameter cubic equations of state based on the concept of volume transition. Kubic proposed a linear rule for b though this is apparently inconsistent with the virial interpretation of his equation. The difference between linear and quadratic rules for b is, however, unimportant in the density region of interest to this work.Cross-term parameters a12 and b,, were calculated from eqn (8) and (9) using cross- term critical constants. q12 was obtained from equations (4) and (5). Other cross-terms were calculated from the equations (22) CU,, = 0.5 (CO, -t 0,) yC12 = 0.125 ( VEl + V!2)3 ZCl2 = 0.291 - O.08Ul2 (24) Pc12 = z c 1 2 R L , / v c l 2 . Pseudo-critical Constants for Steam From their analysis of cross-term 4 , s for mixtures of water-non-polar-fluid, Smith et al.’ concluded that water in its interaction with a non-polar substance behaved like a small molecule intermediate between methane and argon. Cross-term 4 , s could be fitted by the Potter-McGlashan correlation using a pseudo-critical temperature T, = 230 K for water together with the true critical volume = 55.5 cm3 mol -l.It was also suggested4 that fluoromethane might be a useful homomorph for water in its interaction with a non-polar component. However when fluoromethane critical constants were used as pseudo-critical constants for water in the above equations, agreement with high- pressure experimental HE values was not as good as that obtained by choosing the constants differently. For many non-polar and slightly polar fluids the critical constants are related by the equation (26) Eqn (26) is not exact, but is a good approximation. Using the pseudo-critical temperature T, = 230 K for steam, yC = 55.5 cm3 mol-l, and o = 0.19 (the value for fluoromethane), we obtain a pseudo-critical pressure p , = 9.5 MPa.Many other sets of pseudo-critical constants for water are possible, but this set is adequate for present purposes. Substitution into eqn (4) and (9, (22)-(25), and finally (8) and (9) yields cross-terms a12 and b12, and hence the residual molar enthalpy H: of the mixture. The residual molar enthalpy H:, of the non-polar component was calculated from Kubic’s equations. The residual molar enthalpy H:, of steam was calculated using eqn (17) and (1 8). The excess molar enthalpy HE is given by (27) pc = (0.29 1 - 0.08~) RT,/ K. HEW, P , x ) = H 3 T , P , x ) - Xlll*,,(T, P ) - X,H$,(T’ PI. Comparison with Experiment Excess molar enthalpies H i ( x = 0.5) calculated from the above equations are compared with experimental results for steam-n-alkane mixtures in fig.2. The fit to the HE ( x = 0.5) measurements on mixtures containing ethane, propane, butane and n-pentane is to within experimental error at all temperatures and at pressures up to 14 MPa. The fit to the measurements on (0.5H,0+0.5C,H14) is to within experimental error up to 10 MPa, but above this pressure some of the experimental HEs lie below theC. J . Wormald and I?. M. Lancaster 1321 "0 4 8 12 p l W a Fig. 2. Comparison of calculated and experimental excess molar enthalpies HZ of (0.5H20 + 0.5C,H2,+,) for n = 2-7. Solid curves were calculated from eqn (27) as described in the text. Experimental measurements are listed in ref. (18) and (23H25). (a) (0.5H20 + 0.5C2H,),23 (b) (0.5H2 + 0.5C3H,),24 (c) (0.5H20 + 0.5C4H,,),24 (d) (0.5H20 + O.5C5Hl2),ls (e) (0.5H20 + 0.5C,H,4),25 (f) (0.5H20 + 0.5C,H1,).25 Measurements on all mixtures were made at 448.2,473.2,498.2, 523.2, 548.2, 573.2, 598.2, 648.2 and 698.2 K.Additional measurements on (0.5H20 + 0.5C6H,,) were made at 623.2 and 673.2 K.0 Equation of State for Steam Mixtures 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.0 1 0 0.2 0.4 0.6 0.8 1 Fig. 3. Comparison of calculated and experimental excess molar enthalpies HZ of [xH,O + (1 - x ) C ~ H ~ ~ + J for n = 5-7. Solid curves were calculated from eqn (27) as described in the text. (a) [xH,O + (1 - x)C,H,,],l* all measurements at 4.50 MPa, (b) [xH,O + (1 - x)C,H,,],~~ measurements at 548.2 K and 4.93 MPa, 598.2 K and 9.41 MPa, 648.2 K and 11.48 MPa, (c) [xH,O + (1 --x)C,H,,],~~ measurements at 548.2 K and 4.58 MPa, 573.2 K and 6.00 MPa, 598.2 K and 7.68 MPa.X calculated line by an amount greater than the uncertainty on the measurements. The (0.5H20 +0.5C,H1,) measurements at 573.2 and 598.2 K at pressures above 6 MPa also lie below the calculated line. The vertical sections of the isotherms at 448.2, 473.2 and 523.2 K shown in fig. 2(f) are at the saturated vapour pressure of n-heptane. Above the saturation pressure n-heptane enters the calorimeter as a liquid which evaporates to form a vapour mixture. The height of the vertical sections corresponds to theC. J . Wormald and N . M. Lancaster 1323 p l W a Fig. 4. Comparison of calculated and experimental excess molar enthalpies H z of (a) (0.5H20 + 0.5C,H4) and (b) (0.5H20 + OSCO,).Experimental measurements are listed in ref. (23) and (26). Solid curves were calculated from eqn (27) as described in the text using < = 1.21 for (0.5H20 + 0.5C,H4) and < = 1.33 for (0.5H20 + 0.5C0,). evaporation of n-heptane. Agreement with the calculated HE values at these temperatures is good. For the six n-alkane mixtures shown in fig. 2, < was calculated from eqn (5). The fit to HZ measurements on mixtures of steam-hydrogen, -nitrogen, -methane and -carbon monoxide obtained using 5 from eqn ( 5 ) is as good as that for the C,-C, n-alkanes. The composition dependence of HZ(x) for steam-n-pentane, n-hexane and n-heptane is shown in fig. 3. The solid curves calculated from our equations fit the measurements on [xH,O + (1 - x)C5H12] to within experimental error.The calculated curves are slightly above the experimental points for [xH,O + (1 - x)C6H14] and slightly below for Where there is a specific interaction between water and the other component of the mixture, is greater than that given by eqn (3) or (5) and must be obtained by fitting either cross-term second virial coefficients or excess enthalpies at low pressures. Measurements of HZ 0, = 101.325 kPa) for (0.5H20 + 0.5C,H4) and (OSH,O + OSCO,) yielded values of < of 1.21 and 1.33, respectively. Eqn (5) gives 0.97 and 0.99. High- pressure H z s calculated using these experimental 5 values are shown in fig. 4. For (0.5H20 + 0.5C,H4) the HZ values calculated from our equations agree with experiment at pressures up to 5 MPa. At higher pressures the calculated curves lie slightly below the experimental points.Fig. 4(b) shows the fit to measurements on (0.5H20 + 0.5C0,) *at pressures up to 14 MPa. The fit is very good up to 648.2 K. Only the measurements at 698.2 K are not fitted well. Our equations are a good fit to the H: measurements on all mixtures up to (0.5H20 + 0.5C,H1,). For (0.5H20 + 0.5C6H14) and (0.5H20 +0.5C7H16) at high pres- sures some of the calculated HE values are smaller than those obtained experimentally. The fit to the HZ values for these mixtures is, however, to within the uncertainty on the measurements at pressures up to p , for the hydrocarbon. The lack of agreement at higher pressures is probably due to the inability of our simple equation of state to fit the residual enthalpy of the hydrocarbons at temperatures just above T,. At temperatures well above T, the fit is better, and better agreement with the isotherms at 648.2 and 698.2 K is observed, as shown in fig.2(e) and (f). We expect our equations to work well at [XH,O 4- (1 - X)C7HI6].1324 Equation of State for Steam Mixtures 1 1 1 I I I 360 3 80 400 420 TIK Fig. 5. Comparison of calculated and experimental excess molar enthalpies H: of (0.5H20 + 0.5CflH,,+,) for n = 1-7 at p = 101.325 kPa. Solid curves were calculated from eqn (27) as described in the text using values of 5 calculated from eqn (5). The low pressure H: measurements are listed in ref. (5)-(7). temperatures above 698.2 K, and above this temperature it is likely that they will give good thermodynamic properties of mixtures at pressures up to 20 MPa.The success of our equations in fitting mixture properties over a wide range of pressure is highlighted by the fit to the HEs at p = 101.325 kPa for steam-C,-C, n-alkanes shown in fig. 5 . These HE measurements were made using our low-pressure flow-mixing Agreement with experiment is to within the uncertainty on the measurements at all pressures. The fit to low-pressure HE measurements on mixtures containing ethene and carbon dioxide is just as good. In the comparisons with experiment shown in fig. 2-5 HZ, for steam was calculated using eqn (16)-(18), which were fitted to HGK residual enthalpie~.~~ In the calculation of HE from eqn (27) an alternative procedure is to calculate HZl for steam directly from the HGK equation of state. For the pressure range covered by our measurements there is little point in doing the extra computation this involves, but for the calculation of mixture properties at high pressures it is essential to have an accurate value of the residual properties of steam.Many mixtures of interest contain steam and simple gases such as oxygen, nitrogen or methane. For such mixtures the largest term in eqn (27) is Pml. H:, for the gas is smaller, and as the gas is usually well above its critical temperature its residual enthalpy will be given with good accuracy by eqn (16). Under these conditions the residual enthalpy of the mixture calculated using eqn (19H21) will also be smaller than H:l, and the value calculated from eqn (16) should not be too much in error. The above remarks about residual enthalpies apply equally well to residual volumes. In fig.6 (a) we show the fit to the molar volume Vm of [xH,O + (1 - x)A,] at 673.2 K and at 50 and 100 MPa. The measurements were made by Lentz and Fran~k.~* The curves calculated from our equations fit the measurements well. Fig. 6(b) shows the fit to excess molar volumes V'g of (0.35H2O+0.65N2) at 673.2 K and at pressures up to 250 MPa obtained from the measurements of Japas and F~anck.~' In the 50-250 MPa region our equations are a good fit to the measurements. At 32 MPa our equation has a maximum in VE. Such maxima in supercritical region excess functions are to be expected50 and have been seen before.51 We have shown that experimental H t s for mixtures of steam with fluids as diverse as hydrogen, carbon dioxide and n-heptane can be well fitted using a simple two-termC.J. Wormald and N. M. Lancaster 1325 X plMPa Fig. 6. Comparison of calculated and experimental molar volumes at high pressures. Solid curves were calculated from our cubic equation of state together with HGK molar volumes of steam as described in the text. (a) molar volumes V, of [xH,O+ (1 -x)Ar] at 673.2 K measured by Lentz and F r a n ~ k , ~ ~ (b) excess molar volumes V: of (0.35H2O+0.65N2) at 673.2K from the measurements of Japas and F ~ a n c k . ~ ~ equation of state and a single temperature-independent parameter c. The calculation of densities, compressibilities and other vapour phase thermodynamic functions is now straightforward. References 1 C. J. Wormald, J. Chem. Thermodyn., 1977, 9, 901.2 G. R. Smith, A. Sellars, T. K. Yerlett and C. J. Wormald, J. Chem. Thermodyn., 1983, 15, 29. 3 P. Richards, C. J. Wormald and T. K. Yerlett, J. Chem. Thermodyn., 1981, 13, 623. 4 P. Richards and C. J. Wormald, 2. Phys. Chem. N.F., 1981, 128, 35. 5 N. M. Lancaster and C. J. Wormald, J. Chem. Thermodyn., 1985, 17, 295. 6 N. M. Lancaster and C. J. Wormald, J. Chem. Thermodyn., 1986, 17, 545. 7 G. R. Smith, M. J. Fahy and C. J. Wormald, J. Chem. Thermodyn., 1984, 16, 825. 8 C. J. Wormald and N. M. Lancaster, J. Chem. Thermodyn., 1985, 17, 903. 9 G. R. Smith and C. J. Wormald, J. Chem. Thermodyn., 1984, 16, 543. 10 C. J. Wormald and N. M. Lancaster, J. Chem. SOC., Faraday Trans. 1, 1988, 84, 3141. 11 N. M. Lancaster and C. J. Wormald, 2. Phys.Chem. N.F., 1981, 128, 43. 12 M. Rigby and J. M. Prausnitz, J. Phys. Chem., 1968, 72, 330. 13 C. R. Coan and A. D. King, J. Am. Chem. SOC., 1971, 93, 1857. 14 D. J. Hutchings, E. J. Lewis and C. J. Wormald, J. Chem. Thermodyn., 1978, 10, 559. 15 C. J. Wormald, E. J. Lewis and D. J. Hutchings, J. Chem. Thermodyn., 1979, 11, 1. 16 M. L. McGlashan and D. J. B. Potter, Proc. R. SOC. London, Ser. A , 1962, 267, 478. 17 H. Orbey and J. H. Vera, AIChE J., 1983, 29, 107. 18 N. M. Lancaster and C. J. Wormald, J. Chem. SOC., Faraday Trans. 1, 1988, 84, 3151. 19 C. J. Wormald, Ber. Bunsenges. Phys. Chem., 1984, 88, 826. 20 C. J. Wormald and C. N. Colling, J. Chem. Thermodyn., 1985, 17, 437. 21 C. J. Wormald and C. N. Colling, J. Chem. Thermodyn., 1983, 15, 725. 22 C. J.Wormald and C. N. Colling, AZChE J., 1984, 30, 386. 23 N. M. Lancaster and C. J. Wormald, J. Chem. Thermodyn., 1987, 19, 89. 24 N. M. Lancaster and C. J. Wormald, J. Chem. Thermodyn., 1987, 19, 1001. 25 N. Al-Bizreh, C. N. Colling, N. M. Lancaster and C. J. Wormald, J. Chem. Soc., Faraday Trans. I, 26 C. J. Wormald, N. M. Lancaster and A. J. Sellars, J. Chem. Thermodyn., 1986, 18, 135. 27 C. J. Wormald and N. M. Lancaster, J. Chem. Eng. Data., 1989, in press. 1989, 85, 1303-1313. 45 FAR I1326 Equation of State for Steam Mixtures 28 D. B. Robinson, D-Y. Peng and S. Y-K. Chung, Fluid Phase Equilibria, 1985, 24, 25. 29 M. Baumgaertner, R. A. S. Moorwood and H. Wenzel, ACS Symp. Ser., 1980, 133, 415. 30 G. Schmidt and H. Wenzel, Chem. Eng. Sci., 1980, 35, 1503. 31 J. Ghemling, D. D. Liu and J. M. Prausnitz, Chem. Eng. Sci., 1979, 34, 951. 32 M. D. Donohue and J. M. Prausnitz, AIChE J., 1978, 24, 849. 33 R. de Santis, G. J. F. Breedveld and J. M. Prausnitz, Ind. Eng. Chem. Proc. Des. Dev., 1974, 13, 374. 34 R. Nakamura, G. J. F. Breedveld and J. M. Prausnitz, Ind. Eng. Chem. Proc. Des. Dev., 1976, 15, 557. 35 L. R. Oellrich, H. Knapp and J. M. Prausnitz, Fluid Phase Equilibria, 1978, 2, 163. 36 N. Nishida, M. Ohba and Y. Arai, Fluid Phase Equilibria, 1980, 4, 303. 37 H. W. Woolley, J. Chem. Phys., 1953, 21, 236. 38 J. D. Lambert, G. A. H. Roberts, J. S. Rowlinson and V. J. Wilkinson, Proc. R . Soc. London, Ser. A, 39 J. J. Martin, Ind. Eng. Chem. Fundam., 1979, 18, 81. 40 W. L. Kubic, Fluid Phase Equilibria, 1982, 9, 79. 41 R. Clausius, Am. Phys. Chem., 1881, 9, 337. 42 C. Tsonopoulos, AJChE J., 1974, 20, 263. 43 L. Haar, J. S. Gallagher and G. S. Kell, NBSINRC Steam Tables (Hemisphere, New York, 1984). 44 E. J. LeFevre, M. R. Nightingale and J. W. Rose, J. Mech. Eng. Sci., 1975, 17, 243. 45 C. J. Wormald and T. K. Yerlett, J. Chem. Thermodyn., 1985, 17, 1171. 46 N. C. Pate1 and A. S. Teja, Chem. Eng. Sci., 1982, 37, 463. 47 A. Peneloux and E. Rauzy, Fluid Phase Equilibria, 1982, 8, 7. 48 H. Lentz and E. U. Franck, Ber. Bunsenges. Phys. Chem., 1969, 73, 28. 49 M. L. Japas and E. U. Franck, Ber. Bunsenges. Phys., Chem., 1985, 89, 793. 50 C. J. Wormald, Fluid Phase Equilibria, 1986, 28, 137. 51 C. J. Wormald and J. M. Eyears, J. Chem. Soc., Faraday Trans. 1, 1988, 84, 3097. 1949, 196, 113. Paper 8/00924D; Received 7th March, 1988

ISSN:0300-9599    

DOI:10.1039/F19898501315

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

J. Chern. SOC., Faraday Trans. I, 1989, 85(6), 1327-1335 Reactivity of A1P04-5 and the Origin of its Hydrophilic Property Akira Endoh, Kiyoshi Mizoe, Kazuo Tsutsumi* and Tetsuo Takaishi Toyohashi University of Technology, Toyohashi, 440 Japan The reactivity of A1P04-5 was measured through the isotope exchange reaction of "0 between P O , and the framework oxygen, and its hydrophilicity was investigated by measuring the heat of adsorption of water. About 25% of the framework oxygen atoms were exchanged, and one quarter of them (ca. 6 YO of the total oxygen atoms) was highly reactive. The differential molar heat of adsorption, calorimetrically determined, was ca. 86 kJ mol-' at the initial stage and suddenly dropped to ca. 50 kJ mo1-'. This tendency is characteristic of hydrophobic zeolites, and indicates the existence of special hydrophilic sites. It is concluded, through quantitative analyses, that the highly reactive oxygen atoms and hydrophilic sites originate from defects, and that the concentrations of the defects, estimated from each of the above two, are in good agreement. Modern zeolite science started from studies of the molecular sieving action of zeolites controlled by the regular crystal structure, and defects in zeolites were not a subject of main concern, in contrast to those in semiconductors. Catalytic properties of zeolites, in particular acidic properties,' have focused our attention on defects in zeolites, but the study of defects is standing still at an elementary stage compared with the semiconductor case.The elucidation of defects is an important issue for the characterization of zeolites and other molecular sieves.Effects of defects must be investigated over a wide temperature range, since molecular sieves are used, in science and industry, at a variety of temperatures in the range 4-800 K. The representative techniques for the study of defects, available at present, are adsorption of gases,2 infrared spectroscopy3 and ,'Si magic-angle-spinning n.m.r. spectroscopy* at low and moderate temperatures, and the 180-exchange between oxygen-containing gaseous molecules and the framework of molecular sieves at higher temperatures5, For the clear determination of the properties of defects, combined use of these techniques is essential. In a series of studies on defects, we first investigate A1P04-5.Aluminophosphate compound contains no exchangeable cation to act as an active site for water adsorption, and might be predicted to be hydrophobic. Contradicting this expectation, alumino- phosphate molecular sieves are hydrophilic to some extent.' We anticipate that the hydrophilicity stems from defects in the crystal, and study the relation between the concentrations of the defects and hydrophilic points. The concentration of the hydrophilic sites is determined from the adsorption calorimetry, and that of defects is determined from the exchange reactivity of l80 between Cl80, and the alumino- phosphate crystal. Experimental Materials Crystals of A1P04-5, supplied by Dr A. W. Chester, Mobil Research and Development, had a composition of (AlP0,),,.8.3 H,O per unit cell, with no residual template molecule.They had a high crystallinity and purity, according to powder X-ray diffraction, 31P and 27Al magic-angle-spinning n.m.r. and scanning electron-microscopy. 1327 45-21328 Hydrophilic Property of A1P04-5 - - - r, n 5 8 50 c) 0 0 - 0 50 100 reaction time/h 150 Fig. 1. Temperature dependence of l8O exchange reaction. Each run contains different amounts of oxygen atoms in gas and solid phases. n,, amount (moles) of oxygen in gas; n,, amount (moles) of oxygen in solid. 0, 973 K, n, = 13.74 x ng = 2.07 x a, 1023 K, n, = 14.53 x ng = 2.21 x @, 1073 K, n, = 15.35 x ng = 2.11 x lop4; 0, 1123 K, n, = 1 5 . 9 0 ~ ng = 2.11 x 10-4. Prior to measurements, the sample was calcined in air at 773 K to ensure the elimination of template molecules.Carbon dioxide (supplied from Prochem Ltd) contained 99 % enriched l80. Apparatus The apparatus used in the 180-exchange reaction was described in a previous paper.6 The calorimeter used was of the twin conduction type, equipped with sensors of semiconductive (Sb, Bi)-telluride, and had a sensitivity of 0.12 ,uV/pW. About 1 g of zeolite sample was used after baking-out in the sample cell at 723 K under a vacuum of Pa for 5 h or longer. The adsorbed amount of gas was determined by the volumetric method with an apparatus combined with the calorimeter. The pressure was measured with a diaphragm gauge (MKS Baratron, type 315BHS and 227HS). Further details were described in a previous paper.8 Results and Analysis Isotope Exchange Reaction It is known that A1P04-5 has a high stability and that no deterioration of its crystallinity takes place up to 1273 K.ll2 About 60 mg of A1P04-5 was installed in the reaction vessel and completely dehydrated at 1073 K for 20 h under a vacuum of Pa; cn. 1850 Pa of Cl80, was introduced in the system, and the "0-exchange reaction was measured at temperatures in the range 373-1 123 K.At first, the effect of temperature on the reaction rate was surveyed roughly, and then the more detailed data shown in fig. 1 were collected.A . Endoh, K. Mizoe, K. Tsutsumi and T. Takaishi 1329 From fig. 1, it is concluded that only about 25 O h of the framework oxygen atoms are exchanged after a long reaction time of 150 h. Since A1P04-5 has four crystallo- graphically different kinds of the framework oxygen atoms in equal amounts, it is considered that only one kind of them is reactive and that the other three kinds of them with low reactivities do not participate in the present reaction.Let us divide the exchangeable oxygen atoms into two groups, one (specified by the subscript 1) lying on normal lattice sites, and another being related to defects with an abnormally high exchange reactivity. There may be various kinds of defects having different reactivities, and we sub-divide oxygen atoms related to defects into two sub- groups, less reactive and more reactive ones specified by subscripts 2 and 3, respectively. Denoting by y the fraction of "0 in a given kind of oxygen atom, and by the subscript g the gas-phase carbon dioxide, we have the equation of the exchange reaction,6 n d-l?g = - n g y , C k , ( l - y , ) n , + n , ( l - y , ) 2 kiyini (1) dt i = 1 , 2 , 3 i = 1 , 2 , 3 where n denotes the number of moles of oxygen, and k the rate constant.The condition of material balance gives an auxiliary equation, n,(Y:-Yg> = c niyi i = 1 , 2 , 3 in which the amount of adsorbed CO, was small and has been neglected. oxygen atoms related to defects, i.e. If there are large differences between the reactivities of the normal lattice oxygen and k , 4 k , 4 k3 (3) then in the latter stage of the reaction, the second and the third kinds of oxygen are in equilibrium with the gas phase, and the rate of exchange is determined by the first kind of oxygen. This situation is expressed by Yg = Y2 = Y3 # Y l (4) by which eqn (1) and (2) reduce to ngY:-(ng+n,+n,)Yg = nlY1.(6) These equations are solved with the condition that eqn (4) must be satisfied even at the beginning of the reaction. This means that yg effectively has an initial value of ngy:/(n,+n3+ng) instead of yi, since only the former value satisfies eqn ( 5 ) at t = 0 and y , = 0. The solution becomes n, +- 22, n g +n3 + ng Y:] On the other hand, in the intermediate stage of the reaction the third kind of oxygen is in equilibrium with the gas phase, and the first kind does not participate in the exchange; hence the rate of reaction is determined by the second kind of oxygen. This situation is expressed by y 3 = y g and y , = 0 (8)1330 E:! Hydrophilic Property of AlPO,-5 f -1.0 s o P + - E: reaction time/h Fig.2. Plots of the degree of exchange us. time, based on eqn (7). (a) 973 K; (b) 1023 K ; (c) 1073 K; ( d ) 1123 K. by which eqn (1) and (2) reduce to with, ngY;-(ng+n3)Y, = n,y,* (10) These equations are solved with the condition that eqn (8) must be satisfied, i.e. y , effectively has an initial value of ng yg/(n3 + ng) instead of y:. The solution becomes which has a similar form to eqn (7). the reaction, and one has approximately, In the initial stage of the exchange, only the first kind of oxygen atom takes part in (12) Yl = Y2 = 0 by which eqn (1) and (2) reduce to, with Solving these equations, one has ngy' = n3Y3 + ngYg' n n y :] = In [ 3 y:] - k3(n3 + ng> t. n3 + n,A . Endoh, K. Mizoe, K. Tsutsumi and T. Takaishi 1331 0 5 1 0 0 S 10 1s reaction timefh Fig.3. Plots of the degree of exchange us. time, based on eqn (11). (a) 973 K; (b) 1023 K; ( c ) 1073 K; ( d ) 1123 K. If the correct value of (n, + n2 + n,) is introduced into eqn (7), the plot of against t becomes linear, apart from its initial portion near t = 0. The best-fit value for (n, + n, + n,) was chosen by iteration, and results are shown in fig. 2. The intersect of the tangent to the curve with the ordinate gives the value for In In, n,Y:/(n,+ n2 + n g ) (n, + n2 + n3 + ng>l and hence values for (n,+n,) and n,. The value of (n,+n,) thus determined was introduced into eqn (1 l), and then plots of In b, - n g Y : m 2 + n3 + ng)l against t become linear, apart from its initial portion near t = 0, as shown in fig. 3. The intersect of the tangent of the curve with the ordinate gives the value for In [a, ngu3(fl, + n g ) (n, + n, + ng)l and hence values for n2 and n3.If the value of n, thus determined is introduced into eqn (15), then k, can be calculated. However, the initial change took place too fast for its profile to be followed with the present measuring system. The values of n,, n2 and n3 at various temperatures are summarized in table 1, where they are shown to be self- consistent. Plots of In k, and In k2 against l/T in fig. 4 give values of 1.38 eV and 1.63 eV for the respective activation energies of the reactions. The activation energy is higher for the reactive kind of oxygen, but the frequency factor is much larger resulting in a higher exchange rate.1332 Hydrophilic Property of A1P04-5 Table 1.Amounts of n,, n2 and n3 determined" ~~ 973 24.3 4.4 1.5 1023 22.6 4.8 1.4 1073 20.0 4.2 1.6 1123 23.3 4.4 1.4 a Concentrations expressed as percentage of the total framework oxygen atoms. T/"C 0s 0.9 1D 1.1 lo3 KIT Fig, 4. Temperature dependence of the rate constants of 180-exchange in A1P04-5. (a) k , ; (b) k,. Heat of Adsorption The calorimetrically determined differential molar heats of adsorption of water at 303 K are plotted against the adsorbed amount in fig. 5 . Details of initial regions of adsorption are inserted in the figure. The calorimetric heat curve reflects the energy distribution of adsorption site' and becomes an indication of hydrophilic sites in the case of water adsorption. The heat curve shows a very high value at an extremely low coverage, say 86 kJ mol-l, which is similar to heats of adsorption on Na-type faujasite.1° The heat value decreases very sharply down to the heat of liquefaction, and then remains almost constant up to theA .Endoh, K. Mizoe, K. Tsutsumi and T. Takaishi 1333 H20 adsorbed/mmol g-' 2 4 6 8 I 1 - 1 I I I I 90 I 00 I 90 1 0 1 2 3 4 H 2 0 molecules adsorbed/u.c. 0 0 0 0 0 0 O 0 0 I I I I I 0 2 4 6 8 10 12 H20 molecules adsorbed/u.c. Fig. 5. Differential molar heats of adsorption of water on A1P04-5 at 303 K. adsorption saturation. The hydrophilicity of aluminosilicate zeolites is clearly ascribed to the interaction of the water dipole with an electrostatic field on a zeolitic surface, the field being produced by cations and negative framework charges. However, in the case of aluminophosphate, there is neither cation nor framework charge.Therefore, the initial high-heat value must be attributed to some lattice defects, which are expected to present to a small extent. Discussion The exchange of oxygen atoms takes place rather easily between CO, and aluminosilicate zeolites; namely, it occurs at much lower temperatures than 773 K, and all framework oxygen atoms are l1 AlPO,-5 is far less reactive than aluminosilicates, and only a limited portion, ca. 25 YO, of the framework oxygen atoms can be exchanged even at a high temperature such as 1123 K. It is proved in zeolite A that the activated complex contains, as its constituents, exchangeable cations in the zeolite, and hence the poor reactivity of A1P04-5 may be attributed partly to the absence of such cations.There remains, however, the question of why only one of the four kinds of specified framework oxygens can be exchanged. Fig. 6 shows the wall of the cylindrical pore channel of A1P04-5 developed on a sheet. There are, on the wall, three kinds of oxygen, crystallographically denoted as 01, 011, and OIII.12'The fourth kind of oxygen OIV is located under the surface of the wall and not exposed to a visiting ClSO, molecule. It is evident that the reactive oxygen atom belongs to a given species of these three on the wall, but a further detailed assignment requires some speculation. 01 and 0111 are located in similar geometrical environments, different to 011, and thereby 01 and 0111 may have similar exchange reactivities.It is thus highly probable that 011 is the unique species with the high exchange reactivity. The large divergence in the strengths of A1-0-P bonds is an important and interesting problem in the crystal chemistry of aluminophosphate molecular sieves, and quantum- chemical studies are awaited.1334 Hydrophilic Property of AlPO,-5 (4 (b) - -60 0 60 + e caxis - 60 0 60 Fig. 6. Configuration of 01, 011, 0111 and OIV on the wall of main channel in A1P04-5. The cylindrical wall is developed on a sheet. (a) Vacancy of T-site atom. (b) Vacancy of P-O-A1 complex. 0, ., vacant site; 0, direct neighbour oxygen atom; a, next-neighbour oxygen atom. We now consider the problem of defects. If one of the A1 or P atoms is lost from a normal site, oxygen atoms around that vacant site may become reactive for the l 8 0 - exchange.The three nearest-neighbour oxygen atoms become most reactive, and the six next-nearest neighbours might also become reactive. If a complex P-0-A1 is lost and two OIVs are exposed as shown in fig. 6(6), then six direct neighbours and eight next- neighbours become reactive. Thus, the number of the reactive oxygen atoms introduced by a defect may range between three and 14. If a defect was formed through either mechanism mentioned above, T-OH linkage may appear in the framework to retain electrical neutrality. In fact the stretching vibration of hydroxyl groups was detected at 3695 cm-l in i.r. spectra of dehydrated A1P04-5. Such linkage becomes a hydrophilic site, similar to the case of Si-OH or ME-OH in aluminosilicates.Water molecules may interact with several, say one to four, T-OH to form a hydrogen bond, e.g. / H 'H T a H . . . . . . . . . . 0 or depending on the population of T-OH groups.A . Endoh, K . Mizoe, K. Tsutsumi and T. Takaishi 1335 The calorimetric result indicates that ca. 0.6 water molecules per unit cell interact with hydrophilic points. On the other hand, the highly reactive oxygen atoms, n, and n3, amount to 2.1 and 0.7 per unit cell, respectively. Thus 0.6 hydrophilic points contain 2.8 (= n, +n,) easily exchangeable oxygen atoms. If we assume that one defect produces one hydrophilic point, the number of easily exchangeable oxygen atoms per defect, v, amounts to 4.7(= 2.8/0.6). This value satisfies the structure condition 3 < v < 14 given in the above.At the present, we cannot discuss further details of the defect. Only a small number of defects, amounting to lo-'' part of a host crystal, play decisive roles in commercial semiconductors, while a much larger number is required to influence chemical properties of oxides. Defects in the semiconductors are easily detected by electric measurements, but the techniques cannot be applied to oxides in spite of an overwhelmingly large concentration of defects. Twin-boundaries in zeolites are observed with an image-electron-microscope13 but point defects are not. If the concentration of defects in zeolites is of the order of lo-,, they can be detected by the ,'Si magic-angle- spinning n.m.r. spectr~scopy.'~~ l5 The concentrations of various defects thus determined can explain well the observed acid strength distribution of (H, K)-L zeolites.16 These techniques, however, cannot be applied to aluminophosphate compounds, and so the "0-exchange technique is the only means, available at present, to measure the concentration of defects within these compounds.Recently, Gelsthorpe and Theocharis presented a speculation, for N, adsorption isotherms on AlPO,-5, that there exist point defects on the wall of the channel pore. Our present results quantitatively support their view.17 The consistency between "0-exchange and H,O-adsorption data encourages us to apply these techniques to various problems or to characterizations of molecular sieves. An example is the application to ZSM-5 and -1 1.18 We sincerely thank Dr A.W. Chester, Mobil Research and Development, for the supply of AlPO,-5. The present work was supported by a Grant-in-Aid for Scientific Research of the Ministry of Education of the Japanese Government, contract no. 6260200 1 . References 1 D. Barthomeuf, Zeolites: Science and Technology (NATO AS1 Ser., 1984), p. 317. 2 K. Tsutsumi, Y. Mitani and H. Takahashi, Colloid Polymer Sci., 1985, 263, 838. 3 E. M. Flanigen, Zeolite Chemistry and Catalysis (ACS Monograph 171, 1979), p. 80. 4 J. Klinowski, Prog. Nucl. Magn. Reson. Spectrosc., 1984, 16, 237. 5 R. V. Ballmoos, "0-Exchange Methods in Zeolite Chemistry (Saller and Sauerlander, 1981). 6 T. Takaishi and A. Endoh, J. Chem. Soc., Faraday Trans. I , 1987, 83, 411. 7 S. T. Wilson, B. M. Lok, C. A. Messina, T. R. Cannan and E. M. Flanigen, J . Am. Chem. Soc., 1982, 8 K. Tsutsumi, H. Q. Koh, S. Hagiwara and H. Takahashi, Bull. Chem. Soc. Jpn, 1975, 48, 3576. 9 K. Tsutsumi, Y. Mitani and H. Takahashi, Colloid Polymer Sci., 1986, 264, 445. 104, 1146; ACS Symp. Ser., 1983, 218, 79. 10 0. M. Dzhigit, A. V. Kiselev, K. N. Mikos, G. G. Muttik and T. A. Rahmanova, Trans. Faraday Soc., 11 C. Gensse, T. F. Anderson and J. J. Fripiat, J . Phys. Chem., 1980, 84, 3562. 12 J. M. Bennett, J. P. Cohen, E. M. Flanigen, J. J. Pluth and J. V. Smith, ACS Symp. Ser., 1983, 218, 13 J. M. Thomas, G. R. Milward and S. Ramadas, ACS Symp. Ser., 1983, 218, 181. 14 T. Takaishi, J. Chem. Soc., Faraday Trans. I , 1987, 83, 2681. 15 T. Takaishi, J. Chem. Soc., Faraday Trans. I , 1988, 84, in press. 16 K. Tsutsumi, K. Nishimiya, A. Shiraishi and T. Takaishi, to be published shortly. 17 M. R. Gelsthorpe and C. R. Theocharis, Catal. Today, 1988, 2, 613. 18 A. Endoh, K. Nishimiya, K. Tsutsumi and T. Takaishi, Proc. Inter. Symp. Zeolite, Wurzburg, 1988. 1971, 67, 458. 109. Paper 8/01 I80J; Receiued 18th October, 1988

ISSN:0300-9599    

DOI:10.1039/F19898501327

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. I , 1989, 85(6), 1337-1349 Interactions between Metal Cations and the Ionophore Lasalocid Part 6.-Potentiometric and Electron Spin Resonance Study of the Complexation of Gd3+ in Methanol by Lasalocid and Simpler Carboxylic Acids Madeleine Tissier, Guy Mousset and Jean Juillard* Laboratoire d ’Etude des Interactions Solutks-Solvants et Laboratoire d’Electrochimie Organique de 1’UA CNRS 434, Universite‘ Blaise Pascal (Clermont 11), 631 77-Aubih-e Cedex, France Complexes formed in methanol between the ionophore lasalocid and the Gd3+ ion have been studied using pHmetric, e.s.r. and u.v.-visible titration against tetrabutylammonium methoxide solution. The same has been done for salicylic and benzoic acids. In all cases the formation of the successive complexes AGd2+, A,Gd+ and A3Gd between the acid AH and the Gd3+ ion has been observed and the corresponding formation constants were determined.More or less strong variation of the height, width and intensity of the e.s.r. line has been observed with the formation of the complexes; individual e.s.r. parameters could be had for most of the complexes. Comparison of data obtained for lasalocid and salicylic or benzoic acid suggests that in lasalocid anion-gadolinium cation complexes in methanol, the carboxylate group is the main and possibly the sole binding site of the gadolinium ion. Other complexes are formed in more basic media in methanol; these may involve the salicylate dianion, as shown for salicylic acid-gadolinium complexes, or, more probably, result from successive substitution of a methoxide ion for a lasalocid anion in A,Gd.Lanthanide cations are sometimes used, by virtue of their specific spectroscopic properties, as probes for calcium in its reactions with biorganic ligands. Lasalocid (fig. 1) has been shown to mediate their transport across both natural and artificial membranes.l Studies of the binding of lasalocid to lanthanide cations in various solvents (dimethylformamide, chloroform and especially methanol) have recently been carried out?’ The main issues addressed have been the nature and structure of the complexes formed, the binding sites involved and the conformation of the lasalocid anions thereby implied. Formation of three types of complexes between lasalocid AH and lanthanides Ln3+ has generally been assumed; ALn2+, A,Ln+ and A3Ln.The neutral one, A3Ln is formed in aprotic media such as c h l ~ r o f o r m . ~ ~ ~ In methanol, the presence of all three species was suspected in the case of Pr3+ by Chen and Springer3 who proposed approximate values for the stepwise association constants. Other studies in this s o l ~ e n t ~ . ~ suggested the formation of only 1-1 and 2-1 complexes : ALn2+ and A,Ln+. In order to settle this issue and obtain reliable values of the formation constants, we undertook a systematic study of lasalocid-lanthanide interactions in methanol using pHmetric titration as previously carried out for lasalocid-monovalent’ and lasalocidaivalent cati~ns.~-ll On the basis of their n.m.r. data, Chen and Springer3 suggested that Pr3+ was, in its lasalocid complexes, only bound to the salicylate moiety; the same conclusion was reached by Hanna et ~ 1 .~ for Gd3+ in another polar solvent, DMF. Conversely, Richardson et al.4 comparing the influence of lasalocid and salicylic acid on the 13371338 Complexation of Gd3+ by Lasalocid '2 \ Fig. 1. Structure of lasalocid showing numbering-scheme for the oxygen atoms. electronic excitation and absorption spectra of various lanthanide ions concluded that other oxygen sites of lasalocid participate in the bonding to the cation. In this work, stepwise formation constants were determined for lasalocid and model carboxylic acids : salicylic, o-methoxybenzoic and benzoic, in order to document the participation of the salicylate moiety and to assess the relative involvement of hydroxyl and carboxylic groups in the complexation of lanthanides by lasalocid. By virtue of its paramagnetic properties which permit a simultaneous e.s.r.study, as performed previously for copper(I1)" and manganese(Ir),lO gadolinium(II1) was chosen for a first approach. E.s.r. spectral data for Gd3+ in solution are s c a r ~ e . ~ ~ - ~ ~ An investigation of its potential uses as a structural probe using a wide variety of complexes in aqueous solution, was carried out by Geraldes and Williams.14 This study stressed the difficulty in attributing the changes observed, mostly in the amplitude and linewidth of the first-derivative X-band spectra, to precise structural features. However, specific information regarding both the nature of the species formed and the involvement of the salicylate moiety could be obtained in the present study, from the variation of the e.s.r.parameters with the pH of the solution, comparing lasalocid with simpler carboxylic acids. Experiment a1 Chemicals Methanol and tetrabutylammonium methoxide solutions were as stated el~ewhere.~ Lasalocid was prepared from its commercial sodium salt (Sigma Co.) as described el~ewhere.~ Salicylic, benzoic and o-methoxybenzoic acids were pure commercial products, recrystallized when necessary. Three salts of gadolinium(Ir1) were used ; GdCl, - 6H20 ' rectapur ' grade from Prolabo ; Gd(NO,), * 5H,O ' gold label, 99.999 % ' grade from Aldrich Chemical Co. and Gd(ClO,), * 6H20 from Alfa-Ventron. No attempt was made to remove the bound water; stock solutions were prepared in methanol and titrated in the usual way using EDTA.This being the case, and given the various other sources of water, data reported here are, as stated before,lO for methanol solutions containing 0.05-0.1 wt % water. Potentiometric Measurements The potentiometric measurements were made as described previouslyl0Y l6 by titrating against tetrabutylammonium methoxide solution in methanol a mixed solution of the ionophore and the gadolinium salt in methanol. The concentration of the salt was 2.5 x lop3 mol dm-, and that of the ionophore or the model acid varied from 2.5 to 7.5 x lop3 mol dm-, according to the stoichiometric ratio chosen (1, 2 or more frequently 3). Electron Spectroscopic Resonance Data E.s.r. spectra were obtained using a Bruker model ER 200 D spectrometer at band X (v = 9.21 GHz).100 kHz frequency modulation was used with a phase of 90". All the measurements were made at a room temperature of ca. 20 "C.M . Tissier, G. Mousse? and J . Juillard 1339 The spectra were recorded for various steps of the titration of the metal-ionophore or model acid mixture by the tetrabutylammonium methoxide titrant in concentration conditions strictly identical to those used for the corresponding potentiometric titration. This was achieved using a fluid circulating device; the reaction was carried out in a vessel by addition through a burette of the base solution to the acid-salt mixture; after each addition, the solution obtained was pushed into the e.s.r. cell using a peristaltic pump; after 5 min circulation, the solution in the vessel, the cell and the connecting tubes was homogeneous and the e.s.r.spectrum was recorded. Preparation of successive solutions is easy and no change occurs in the orientation of the e.s.r. cell in the magnetic field during a series of measurements made in this way. Electronic Spectra Electronic spectra were recorded, between 200 and 400 nm, with Pye-Unicam P.U. 8600 u.v.-visible spectrophotometer. Results Association of the Inorganic Gadolinium Salts The gadolinium perchlorate obtained commercially showed residual acidity presumably due to the presence of perchloric acid ; the titration by tetrabutylammonium methoxide indicated a content of ca. 10% of perchloric acid. It was therefore decided to use the other salts, nitrate or chloride.Surprisingly, gadolinium nitrate was found to be more strongly associated in methanol than gadolinium chloride. This was observed in the titration curves both of the gadolinium salt alone and of the gadolinium salt-lasalocid mixtures. Formation of both gadolinium methoxides and lasalocid complexes was markedly restricted for nitrate solutions as compared to chloride solutions ; neglecting the association of the inorganic salt, the calculated formation constants of the various lasalocid anion or methoxide complexes were smaller by 2 to 4 units in nitrate than in chloride solutions, denoting a stronger association of the gadolinium nitrate. This was confirmed by conductivity measurements. The corresponding phoreograms (equivalent conductivity as a function of the square root of the concentration) showed that gadolinium chloride is, in methanol, a stronger electrolyte than gadolinium nitrate.This had already been observed, but to a lesser extent, in water. The gadolinium ion first derivative e.s.r. signal is only slightly disturbed by adding chloride ions, but strongly disturbed by adding nitrate ions (table 1 of Geraldes and William~'~); adding nitrate brought about a narrowing of the signal; the same was observed in DMF by Poupko et al. who noticed13 that Gd(ClO,), solution gave signals that were three or four times broader than those obtained from Gd(NO,), at the corresponding temperatures. Stronger association of the nitrate salt in methanol probably also occurs with other lanthanides. Enhancement of terbium luminescence in the presence of lasalocid or salicylic acid was found to be, at identical concentration, stronger in TbCl, than in Tb(NO,), solutions [see fig.4 and 11 in ref. (4)]. E.s.r. X-band derivative spectra obtained here in methanol with the three salts at the same concentration are given in fig. 2. Like the spectrum of the gadolinium(rI1) ion in water14 these three spectra contain only a single broad line centred at g = 1.958 (g = 1.992 in water'*). Gadolinium signals in its chloride and perchlorate solutions are not greatly different but marked narrowing and enhancement of the height of the signal were obtained in the gadolinium nitrate solution. This again indicates substantial complexation of the gadolinium ion by the nitrate ligand.Gadolinium chloride was accordingly used for all our experiments. It is, as shown by conductivity, slightly associated, and so the formation constants reported in this paper must be considered as specific to gadolinium chloride solutions. They may be expected to be close to their absolute values.1340 Complexation of Gd3+ by Lasalocid I I I I I I I I I I I I I // 1000 G Fig. 2. E.s.r. X-band first derivative spectra of mol dm-, solutions of GdC1, (-), Gd(NO,), (----) and Gd(ClO,), ( + HClO,) (- - -) in methanol at identical spectrometer settings and at room temperature (x 20 "C). Potentiometric Determinations Typical potentiometric titration curves of 3-1 mixtures of acid ligand-gadolinium chloride against a tetrabutylammonium methoxide solution in methanol are given in fig.3. Four acids were studied: lasalocid, salicylic acid, benzoic acid and o-methoxy- benzoic acid. For all these acids, a first pH jump was observed for two equivalents of base (twice the gadolinium ion concentration) ; this presumably corresponds to succes- sive and partially competitive formation of 1-1 and 2-1 complexes, AGd2+ and A2Gd+, by reactions (I) and (2): A- + Gd3+ s AGd2+ (1) (2) The corresponding formation constants Dl and p2 were calculated using Bjerrum formation functions as was done previously for the complexation of the divalent Good fits were only obtained if the first third of this portion of the curves was neglected, this departure being probably due to a small remaining association of GdCl, at the beginning of the titration.Activity corrections were also applied as before using the Debye-Huckel extended law and Bjerrum's q as distance parameter. 2A- + Gd3+ e A2Gd+.M. Tissier, G. Mousset and J . JuiIlard 1341 14 12 10 % 8 6 4 I I I I I I 4 1 2 3 4 5 6 n Fig. 3. Typical titration curves of various solutions in methanol against tetrabutylammonium methoxide solution. (pH of solution us. number of base equivalents added, n.) GdCl, 2.5 x mol dm-, (0), lasalocid 7.5 x lo-, mol dm-, (l), GdCI, 2.5 x lo-, mol dm-,+AH 7.5 x lop3 mol dm-,: AH = lasalocid (2), salicylic acid (3) or benzoic acid (4); $ precipitation and f redissolution of a white precipitate. and p2 were obtained from the calculated apparent formation constants /I; and using, in the concentration conditions of our experiments: log& = logp’, + 0.4 and logp, = logp;+0.7 with a mean ionic strength of lo-, mol dmW3 throughout the titr- ation curve.Another step, attributed to the formation of the 3-1 complex according to reaction (3) was observed in all the acid titration curves, 3A- + Gd3+ -r) A3Gd. (3) This step was strongly marked for lasalocid, but less so for the other acids. Formation constants p3 of A,Gd calculated from this part of the curve, taking into account p1 and p2 values previously obtained, and corrected in the usual way (log& = logpi+O.9) are also reported in table 1. At least in one case, for salicylic acid, other titrations were performed using various initial stoichiometric AH-GdCl, ratios, 1-1, 2-1 and 6-1. Formation constants calculated from these titration curves were in good agreement with the previous values.After this step, i.e. after adding 3 equivalents of base, the acid having then been completely neutralised any given titration curve of the acid-gadolinium chloride mixture might have been expected to be identical to the titration curve of the corresponding acid alone at the same concentration. This was never the case, as seen for lasalocid, comparing curves (2) to (1) in fig. 3. Hence other complexes must be formed. Titration of gadolinium chloride alone [curve (O)] showed the formation of the two first methoxides Gd(CH,0)2+ and Gd(CH,O)i and then the third one Gd(CH,O), which partially precipitates. Formation constants for reactions (1 ’), (2’) and (3’) analogous to reactions (l), (2) and (3) but with CH,O- instead of A- were determined for the 1-1 and 2-1 complexes and roughly estimated for the 3-1 complex (table 1).These values did not1342 Complexation of Gd3+ by Lasalocid Table 1. Logarithms of the overall formation constants of gadolinium complexes of lasalocid and simpler acid anions and of methoxide ions in methanol according to reactions (l), (2) and (3) (from GdC1, solutions, molar scale of concen- tration mol dm-,, 25.0 "C, accuracy kO.2) lasalocid 8.3 15.3 18.7 salicylic acid 7.6 14.0 16.9 benzoic acid 8.4 15.1 18.6 o-methoxybenzoic 8.2 14.8 17.6 methoxide ion 8.9 18.4 (23) acid fit the second part of the titration curves of 3-1 AH-GdCl, mixtures (beyond three equivalents of base). The species formed in basic media are thus probably not simple met hoxides.E.S.R. Data Titration of the gadolinium chloride-acid ligand 1-3 mixture was carried out against a tetrabutylammonium methoxide solution for concentrations identical to those used in the pHmetric titrations, which yields by comparison and interpolation in the potentiometric curve, the pH values at which the e.s.r. spectra were recorded. Typical titrations involved recording spectra at about thirty different pHs. An example of the results is given for lasalocid in fig. 4; for clarity, two-thirds of the recorded spectra are omitted in this figure. Although the most abundant isotope of Gd3+ has a ground state, a single broad line was always, as is usual, observed in the X-band spectra. This line can be characterised by its width, the peak-to-peak horizontal distance in the derivative spectrum 6, in Gauss, its height taken here as half the peak-to-peak vertical distance in the derivative spectrum, h, arbitrarily expressed in cm and its intensity hd2 assuming the curve to be lorentzian.A slight asymmetry of the derivative signal is generally observed, as the curve is not exactly lorentzian and so ha2 is not the true intensity but only an approximation of it. hd2 is then expressed in cm G2, an arbitrary unit admittedly, but used here only for comparing the spectra obtained at constant gadolinium concentration after normalization at the same spectrometer parameter setting. No noteworthy variations of the g factor was observed but very significant variations in the height of the signal were obtained, as shown in fig.4 for the lasalocid-gadolinium mixture titration. Normalized heights of the signals are reported as a function of the pH for the five titrations in fig. 5. Except at the beginning and the end of the titrations, variations in the height of the signal mostly corresponded to a variation in its apparent intensity, its width being roughly constant. For the formation of the salicylic acid complexes, for example, the width measured is equal to 155 f 5 G at pH 3.2-8.9. A strong enhancement of the signal height was observed for the titration of the salicylic acid-gadolinium mixtures. Comparatively, lasalocid behaved rather analo- gously to benzoic or methoxybenzoic acid except that, because of its stronger acidity, the height of the signal began to increase at lower pH.A strong enhancement of the signal was also observed here for the formation of the successive methoxide complexes. However, no signal was detected in the basic part of the other titration curves at the pHs for which formation of basic complexes can be assertedM. Tissier, G. Mousset and J. Juillard 1343 Fig. 4. E.s.r. X-band first derivative spectra of Gd3+ for a lasalocid-GdC1, 3-1 mixture (concentration of gadolinium 2.5 x lop3 mol dm-3) in methanol with various amounts of base added; corresponding pH: 1, 2.91 (no base added); 2, 3.12; 3, 3.43; 4, 4.58; 5, 5.32; 6, 7.70; 7, 9.77; 8, 10.21; 9, 10.56; 10, 10.95; 11, 11.50. from the shape of the potentiometric curves. This thus confirms that these basic species are not simple methoxides. U . V .-Visible Data U.v.-visible spectrometric titration of 3-1 ligand-gadolinium mixtures was also performed but at lower concentrations, 1.2 x mol dm-3 of lasalocid or salicyclic acid, owing to the absorption of the salicylic group.The spectra of various forms of lasalocid and salicylic acid had been previously analysed.8* l7 The three bands corresponding to the n -+ n* transition of the benzoic group were always present, more or less shifted according to the species involved. They were designated17 A, B, C in the order of the increasing wavelengths. Data obtained here are reported in fig. 6. Discussion will be restricted to the bands B and C respectively at 24&250 nm and 30&3 10 nm. Adding the base solution to lasalocid-gadolinium mixtures [fig. 6 (a)] brought about as successive complexes form, both hypsochromic shifts and hypsochromic effects on the C-band.Such variations were again observed after the addition of the first 3 equivalents of base which correspond roughly to formation of AM2+, A2M+ and A3M. Thus formation of basic species is also observed here. Progressive reduction of the peak corresponding to the B-band to a shoulder was concurrently observed. The spectra finally obtained after adding 6 equivalents of base was different from the anion spectrum1344 E 2 25 -r: Complexation of Gd3+ by Lasalocid I 1 I I I I I I I I I I i I I I I I I I \ \ \ \ '1 \ \ \ \ \ \ \ \ I / \ PH Fig. 5. Variation of height of signal (h in cm) normalized at identical spectrometer settings with titration of gadolinium chloride-lasalocid (-), -salicylic acid (----), -benzoic acid (- - -) or -0-methoxy benzoic acid ( e .. - - ) 1-3 mixtures and gadolinium chloride alone (- ----) against a tetrabutylammonium methoxide solution as a function of the pH of the solution (concentration of gadolinium 2.5 x mol dm-3, solvent methanol, room temperature ca. 20 "C, curves plotted using 25-35 experimental points). 250 300 350 A/nm Fig. 6. Evolution of B and C bands of u.v.-visible spectrum of lasalocid (a) or salicylic acid (b) in 3-1 ligand-gadolinium chloride mixtures in methanol with amount of tetrabutylammonium methoxide added. The number of base equivalents added is stated in the corresponding curves. A- stands for the anion spectrum.M. Tissier, G. Mousset and J . Juillard 1345 at the same concentration.The C-band is centred at longer wavelength and its intensity is weaker. As shown in fig. 6(b), the variations of the salicylic acid spectrum upon adding the first three equivalents of base were weak. Even the spectrum corresponding to A,M is not very different from the AH spectrum. The C-band has shifted from 303 to 301 nm. However thereafter, from 3 to 6 equivalents of base, a new band appears here as a shoulder centred at ca. 325 nm. Such a shift was also observed in a very basic solution of salicylic acid alone in methanol; in a 3 mol dmP3 solution of CH,ONa, for example, the C-band was totally shifted to 316 nm. This must be due to the formation of the salicylate dianion through ionization of the phenol function. Therefore the band observed here at ca.325 nm could be due to the formation of successive complexes involving the OC,H,COi- anion. Such formation might be incomplete. The location of the maximum of the band could only be roughly estimated here, but the spectra of the salicylate dianion involved in the gadolinium complexes seem shifted slightly to longer wavelengths than the free dianion. Such a bathochromic shift of the C-band has not been observed for lasalocid alone in basic media and here, in the presence of gadolinium, only a weak remaining absorption is observed in the area 340-360 nm. Evidently then, any formation of complexes involving the lasalocid dianion must be weak. Most of the species observed in basic media after A3Gd would then be formed either by substituting methoxides or by adding methoxides to lasalocid anions.Processing of the potentiometric titration data using the program MINI QUAD-^^'' enabled various assumptions to be tested concerning the nature of the basic species. Formation of substitution complexes was found to be more probable than formation of addition complexes. The overall stability constant of the first one, A,CH,OGd, was estimated. For the following reaction : 2A- + CH30- + Gd3+ * A,CH,OGd (4) an approximate value of logp, = 19 2 was thus obtained. E.S.R. Spectra of the Various Species Combining Potentiometric and E.S.R. Data From the formation constants of complexes AGd2+, A,Gd+ and A,Gd obtained by pHmetry and given in table 1, the concentration of each species can be calculated at any pH. Distribution. curves of these three complexes as a function of the pH thereby obtained can be directly compared with the variations with pH in the height of the e.s.r.signals reported in fig. 5, as potentiometric and e.s.r. titrations were done in the same concentration conditions. This would give access, assuming some sort of additivity of the individual contributions of these three complexes to the overall e.s.r. spectra, to an individual e.s.r. spectrum of each species. Such processing of the experimental results was attempted. An example for lasalocid, given in fig. 7, shows that this is possible, though a limitation appears for the lower pH; h does not increase as strongly as expected at the beginning of this curve. This could be due to the non-negligible association of the gadolinium chloride which delays the formation of the first lasalocid complex.At pH values above 9, calculation of A,Gd concentrations becomes hazardous owing to the formation of the basic species. However, as shown in fig. 7, at least a rough fit can be obtained which allows the height, h, of the derivative signal to be calculated for 100% of each species in methanol. Such an estimation can, given the limitations at the ends of the curves mentioned earlier, be achieved acceptably for A2Gd+, only roughly for A,Gd and only very roughly for AGd2'. Accordingly, height, width and intensity of the lines are only reported in table 2 for gadolinium in 100% A,Gd+ or (CH,O),Gd+ complexes at the concentration of the experiment 2.5 x lop3 mol dm-,. Such results are not absolutely comparable from one species to another, because the orientation of the cell in the magnetic field, which is constant during a given titration, varies slightly from one series of measurements to another.1346 10 2 5 0 100 % 0 Complexation of Gd3+ by Lasalocid ~~~~~ 3 4 5 6 7 8 9 1 0 1 1 PH Fig.7. Tentative separation of individual contributions of the various lasalocid gadolinium complexes to experimental height of first derivative X-band gadolinium spectrum as a function of pH. Top: (----) e.s.r. signal height, ( - . - - - .) estimated contribution of each species; bottom: percentage of each species (as related to gadolinium content) in solution calculated using formation constants given in table 1. Table 2. Height, width and intensity of gadolinium e.s.r.X-band signal in the A2Gd+ complexes of lasalocid and simpler carboxylic anions and (CH,O),Gd+ complex in methanol (concentration 2.5 x lo-, mol drn-,, room temperature x 20 "C) h/cm 6/G hS2/105 cm G2 lasalocid 10 270 7.5 salicylic acid 48 150 11 benzoic acid 6 330 6.5 o-methoxybenzoic acid 8 300 7 met hoxide 21 130 3.5 For gadolinium chloride: h x 0.7 cm, 6 x 860 G, hd2 5 x lo5 cm G2. The height of the signal, roughly estimated for 100 O h of each of the other complexes, was shown to decrease with the formation of the higher complexes; for example the enhancement of the height h* = h/h,, where h, is the height observed in pure gadolinium chloride solution at the same concentration, would be for AGd2+, A2Gd+ and A,Gd: 21, 14 and 12 for lasalocid and 1 1 , 9 and 3 for benzoic acid.However, as said before, these figures must be considered as very rough.M. Tissier, G. Mousset and J. Juillard 1347 Discussion Stoichiometry of the Lasalocid-Lanthanide Complexes From the data reported in this study it can be asserted that in methanol, successive 1-1, 2-1 and 3-1 complexes are formed between the lasalocid anion, as well as simpler carboxylate A-, and the Gd3+ ion. In addition, other species involving methoxide ions have been shown to be formed in basic media particularly in the case of lasalocid. The discussion will be restricted here to complexes AGd2+, A,Gd+ and A3Gd, the mixed methoxide complexes having little or no biological importance. The occurrence in methanol of three such successive complexes of lasalocid and lanthanide ions Ln3+ had been suggested by Chen and Springer3 for Pr3+ ion on the basis of fluorescence, circular dichroism and n.m.r.data; they proposed for the logarithm of the stepwise formation constants rough values of 7, 6 and 5 as compared to: 8.3, 7.0 and 3.4 found here for Gd3+ ion. Consideration of the distribution curves (bottom of fig. 7) calculated using data in table 1 shows that the formation of A,Gd+ is favoured over the formation of AGd2+ and that formation of A3Gd is strongly disfavoured as compared to the formation of A,Gd'. Preliminary measurements have shown us that the same is true for other lanthanide Ln3+ ions. It is not then surprising that some previous authors4g7 using mostly spectroscopic data obtained at low ionophore and metal ion concentrations indicate only the formation of ALn2+ and A,Ln+ with various lanthanides. However it is clear from both the potentiometric titration curves and the variation of the e.s.r.signal parameters as a function of the pH that, although weak, the formation of A3Gd and more generally A3Ln does occur in methanol. Ligand Sites Involved As regards the species formed, lasalocid does not differ from the other carboxylic acids, at least those studied here. Surprisingly, its complex formation constants are close to those obtained with benzoic acid, for example. This raises the question of the part played by the carboxylate group in the complexation of Gd3+ by lasalocid. Sole consideration of these formation constants suggests that the carboxylate group is the only one implicated in Gd3+ bonding, the other oxygen sites not being involved, which would correspond to an unfolded conformation of the lasalocid anions in the gadolinium complexes.Caution is required in drawing such conclusions from the Gibbs functions alone ; previous studies on divalent cation-lasalocid interactions', have shown us that identity of complex formation AG could well be due to a good self-compensation of strongly non-identical complex formation AH and A S and so to structural differences. Nevertheless, e.s.r. signal parameters, reported here in table 2, for A2Gd+, AH being lasalocid, benzoic acid or o-methoxybenzoic acid also suggest an analogous neigh- bourhood for the Gd3+ cation in these three complexes. 'H and 13C n.m.r. studies of lanthanide complexes in solvents such as chloroform have shown that Gd"' and bind to 0, or 0,, 0,, 0, and 0, and possibly 0,, at least some of the three lasalocid anions thus having a folded conformation.However, in a solvent such as methanol3 or dimethylf~rmamide~ involvement only of the oxygens of the salicylate group is suggested by the shift and broadening of the neighbouring 13C or 'H resonance lines. Data reported here would be in agreement with these conclusions. The contrary opinion of Richardson and Gupta4 or Albin et aL7 that other sites on lasalocid are involved in lasalocid ion coordination in methanol was not supported by direct experimental evidence. No proof of specific binding of the Ln3+ ion by one or several of the oxygens from 0, to 0, was obtained. However, the value of the optical activity of the lanthanide-lasalocid complexes or the quenching of fluorescence of the1348 Complexation of Gd3+ by Lasalocid cation by lasalocid, or other spectroscopic properties were thought to suggest a multidentate chelation. Therefore, in the present state of our knowledge, lasalocid anion might well only bind in methanol, and probably in other protic solvents, to a lanthanide cation by its salicylate group.Oxygens of the Salicylate Group Involved in the Bonding Such binding could occur through 0,, 0, and 0,. Comparison of lasalocid with benzoic acid and salicylic acid provides interesting results in this regard. As underlined previously, a marked analogy exists between benzoic acid and lasalocid-gadolinium complexes both in the formation constants from the respective anions and in the e.s.r.parameters related, for example, to A,Gd+ complexes. Greater differences are observed between salicylic acid and lasalocid. The formation constants are 10 to 100 times lower, the height of the derivative e.s.r. signal is at identical concentrations five times greater, and the apparent intensity 1.5 times greater. This would suggest that the neighbourhood of the Gd3+ cation in lasalocid complexes is analogous to its neighbourhood in benzoic acid complexes of the same type but somewhat different from that occurring in salicylic acid complexes. Analogy with benzoate complexes would suggest then that lasalocid anion binds to gadolinium by 0, or 0, or both but not by 0,. 0, could be involved with 0, or 0, in the ligation to Gd3+ in the salicylate complexes ; such bidentate chelation would favour formation of further complexes involving, in more basic media, the salicylate dianion resulting from the ionization of the phenol group.Such formation is not favoured, as underlined before, for lasalocid. In the salicylate complexes, the carboxylate group is generally in the plane of the aromatic group. Calculations of distances from neighbouring carbon relaxation times enabled Hanna et aL5 to propose a description of the respective locations of the Gd3+--salicylate group in the lasalocid complexes in dimethylformamide : Gd3+ only binds to 0,, the carboxylate group is not coplanar with the aromatic ring (30" dihedral angle) and hydrogen bonding between 0, and 0, is not appreciable.An analogous situation in methanol would be consistent with our present results. In any case, binding of Gd3+ to the phenol oxygen 0, can be ruled out. We are indebted to Claude Tissier of our laboratory for helpful discussions and valuable suggestions and to Anne-Marie Albrecht-Gary and Sylvie Blanc-Parasote of the Louis Pasteur University in Strasbourg who kindly made the calculations involving the MINIQUAD-75 program. References 1 J. Granjean and P. Laszlo, J. Am. Chem. SOC., 1984, 106, 1472. 2 C. V. Krishnan, H. L. Friedman and C. S. Springer Jr, Biophysical Chem., 1978, 9, 23. 3 S-T. Chen and C. S. Springer Jr, Bioinorg. Chem., 1978, 9, 101. 4 F. S. Richardson and A. Das Gupta, J. Am. Chem. SOC., 1981, 103, 5716. 5 D. A. Hanna, C. Yeh, J. Shaw and G. W. Everett Jr, Biochemistry, 1983, 22, 5619. 6 G. W. Everett, S. B. Parker and R. J. P. Williams, Biochemistry, 1983, 22, 6149. 7 M. Albin, B. M. Cader and W. D. Horrocks Jr, Znorg. Chem., 1984, 33, 3054. 8 J. Pointud and J. Juillard, J. Chem. SOC., Furaday Trans. I , 1988, 84, 959. 9 J. Juillard, C. Tissier and J. Jeminet, J. Chem. Soc., Furuday Trans. 1, 1988, 84, 951. 10 P. Laubry, C. Tissier, G. Mousset and J. Juillard, J. Chem. SOC., Furuday Trans. I , 1988, 84, 969. 11 P. Laubry, G. Mousset, P. Martinet, M. Tissier, C. Tissier and J. Juillard, J. Chem. Soc., Furuduy 12 L. Burlamacchi, G. Martini and M. Romanelle, Mol. Phys., 1972, 24, 227. 13 R. Poupko, A. Baram and Z. Luz, Mol. Phys., 1974, 27, 1345. Trans. 1, 1988, 84, 3175.M. Tissier, G. Mousset and J. Juillard 1349 14 E. C. N. F. Geraldes and R. J. P. Williams, J. Chem. SOC., Dalton Trans., 1977, 1721. 15 J. Reuben, J . Phys. Chem., 1971, 75, 3164. 16 C. Tissier, J. Juillard, M. Dupin and G. Jeminet, J . Chim. Phys., 1979, 76, 61 1. 17 Y. Pointud, E. Passelaigue and J. Juillard, J. Chem. SOC., Faraday Trans. 1, 1988, 84, 1713. 18 P. Gans and A. Vacca, Talanta, 1974, 21, 45; P. Gans, A. Sabatini and A. Vacca, Inorg. Chim. Acta, 1976, 18, 237. Paper 81021441; Received 31st May, 1988

ISSN:0300-9599    

DOI:10.1039/F19898501337

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

J. Chew Soc., Faraduy Trans. I, 1989, 85(6), 1351-1356 In situ Scanning Tunnelling Microscopy of a Platinum { 11 l} Surface in Aqueous Sulphuric Acid Solution Shizuo Sugawara and Kingo Itaya" Department of Engineering Science, Faculty of Engineering, Tohoku University, Sendai 980, Japan An in situ scanning tunnelling microscope was applied, for the first time, to a single-crystal platinum { 1 1 l} surface both before and after electrochemical potential cycling in aqueous sulphuric acid solution. The single-crystal Pt{ 11 l} was annealed in a flame near 1100 "C for 1 min and then quickly brought into contact with pure water. It was not possible to see any particular structures on the fire-annealed single crystal. It is shown that the fire-annealing procedure can produce an almost completely atomically flat surface on a single crystal.The flat { 11 l} surface of the Pt crystal was roughened by the electrochemical potential cycling. Semi-spherical domains have predominantly been observed on the single crystal. These domains seem to be randomly distributed over the surface. The diameter and heigbt of the semi-spherical domains were in the ranges 20-30 and 5-10A, respectively. The electrochemical properties of platinum, one of the most important electrode materials, have long been discussed by many workers.' Great efforts have been made to characterize the role of surface structure in electrocatalysis, using single crystals with different orientations as well as polycrystalline Pt. 1-5 The characterization of the adsorption and desorption of hydrogen on clean Pt surfaces is a central issue in efforts to establish the effect of surface structure.To define accurately a Pt surface at the atomic level, low-energy electron diffraction (LEED) electrochemical systems have recently been employed.6-'2 In these systems, Pt surfaces have been prepared and examined by electron spectroscopies such as LEED and Auger in ultra-high vacuum (UHV). Such systems, however, cannot be applied directly to in situ solid-liquid interfaces. Recent papers describing scanning tunnelling microscopy (STM) for samples immersed in aqueous solutions have given persuasive evidence that STM is a powerful new method for in situ electrode surface characterization with atomic resolution. 13-217 24 There have been successful demonstrations of the applicability of STM in electro- chemical cells for the observation of Ag,16 Au15 and Pt20 islands deposited electrochemically on a highly ordered pyrolytic graphite (HOPG).This new method has also been employed in various environments with Pt e1ectr0des.l~~ 21-22 Vazquez et al. described ex situ STM in air to determine the topography of electrochemically highly activated Pt surfaces.22T23 Fan and Bard have recently examined the topography of different Pt surfaces obtained with various pretreatments in air and water.19 In our previous papers, in situ STM of polycrystalline Pt electrodes was carried out, for the first time, both before and after moderate electrochemical activation in an aqueous solution of sulphuric acid.21* 24 Large changes in the surface structure were induced by the electrochemical activation procedure.The appearance of a very regular parallel-terrace structure suggested that the migration of adatoms of Pt produced by the oxidation and reduction cycles would occur in a particular crystallographic direction. 21* 24 Semi-spherical domains were also observed at different positions in the polycrystalline Pt. The appearance of such different morphologies at different points was possibly 13511352 STM of a Platinum (1 11) Surface caused by the different orientations of the individual single crystals of which the polycrystalline Pt sample was composed. In this paper, we report the results of in situ observation of a Pt(ll1) surface as well as the effects of potential cycling for the electrochemical activation procedure in a 0.05 mol dm-3 H,SO, solution.Experimental A single Pt crystal was prepared by the method of Clavilier et ~ 1 . ~ The single crystal Pt bead obtained after melting was ca. 2 mm in diameter. The crystallographic axes of the crystal were determined by X-ray diffraction. The crystal was cut with a diamond cutting wheel to give the { 11 1) surface. The Pt{ 11 1) surface was mechanically polished with successively finer grades of alumina down to 0.05 pm. The mechanically polished single crystal Pt(ll1) surface was annealed in a gas + oxygen or hydrogen + oxygen flame at 1200-1 400 "C for 1 h. A final treatment for STM and electrochemical measurements was performed according to the methods of Clavilier et ~ 1 . ~ and Motoo and F u r ~ y a .~ The single-crystal Pt, annealed in a gas + oxygen or hydrogen + oxygen flame near 1 100 "C for 1 min, was quickly brought into contact with ultra-pure water (Millipore-Q) saturated with hydrogen. Exactly the same method was employed in our previous work for polycrystalline Pt.,l The STM apparatus used was the same as that detailed p r e v i o u ~ l y . ~ ~ , ~ ~ A glass-coated Pt electrode was used as a tunnelling tip. Although the details are described elsewhere,,, it should be mentioned that the electrochemical residual current for well sealed tips (ca. 0.1 nA) was only a few percent of the total tunnelling current (ca. 5 nA) in a 0.05 mol dm-3 H,SO, solution. Results and Discussion Fig. 1 shows typical cyclic voltammograms of the single-crystal Pt{ 1 1 1) electrode obtained in a 0.05 mol dm-3 H,SO, solution after the final treatment.Fig. 1 (a) shows a voltammogram for hydrogen adsorption4esorption, in which positive potential was limited in the double-layer region. The range of potential cycling employed here avoids the electrochemical oxidation of Pt surfaces. The shape of curve (a) is quite similar to those reported by Clavilier et ~ 1 . ~ and Motoo and F ~ r u y a . ~ The total charge for the hydrogen adsorption was calculated as 260 pC ern-,, which is in agreement with the calculated value, 243 pC ern-,, obtained by assuming a coverage of one hydrogen atom per Pt surface atom of an ideal Pt(ll1) surface (1 x l).3 The above result strongly indicates that the final surface treatment described above could give atomically flat surfaces.Fig. l(b) was recorded after the completion of 50 potential cycles between the potentials, -0.25 and 1.05 V versus a standard calomel electrode, SCE, at a scan rate of 0.2 V s-l. Two peaks at ca. -0.2 and -0.05 V us. SCE were gradually formed during successive potential cycles and the shape of the voltammogram approached that for the polycrystalline Pt. This behaviour is essentially the same as that reported previ~usly.~.~ It is interesting to note that the total charge for the hydrogen adsorption was increased by ca. 20 % after 50 potential cycles. This might be due to an increase in the surface area caused by the oxidation-reduction cycling of a Pt oxide layer. Based on the above electrochemical result, in situ observation of a Pt{ 1 1 1 ) surface was carried out before and after the electrochemical potential cycling in the 0.05 mol dm-3 H,SO, solution described above.Fig. 2 shows in situ STM images obtained with different magnifications on a virgin Pt{ 1 1 1) surface (unactivated) in a 0.05 mol dmP3 H,SO, solution. The same final treatment of the Pt crystal, described earlier, was applied for the STM study. The tunnelling current and the tip bias voltage were 4 nA and 50 mV, respectively.S. Sugawara and K. Itaya 1353 I I I I I I l l 1 I l l 1.0 0.5 0 Vvs. SCE Fig. 1. Cyclic voltammograms of a single-crystal Pt{ 11 1) surface obtained in a 0.05 mol dm-3 H,SO, solution. The scan rate was 50 mV s-l. (a) The electrode potential was scanned between 0.4 and -0.23 V us. SCE. (b) after 50 potential cycles between 1.05 and -0.25 V us.SCE at a scan rate of 200 mV s-'. - 50 8, - 20 A Fig. 2. STM images of a virgin Pt{ 1 1 I } surface obtained in a 0.05 mol dm-3 H,SO, solution. The tunnelling current and the tip potential were 4 nA and 50 mV, respectively. As can be seen in fig. 2, the virgin Pt{ 1 1 1) surface is microscopically smooth and nearly atomically flat. In the Fase of well crystallized polycrystalline Pt, a few shallow multiatomic steps (1-20 A in height) were found in our previous study.21.24 However, it was not possible to see any particular structures on the single crystal even when views were taken from different positions. The tunnelling current was reasonably stable during the scan of the tip in the x direction. The above result is the first clear evidence that a final annealing procedure can produce an almost completely or completely atomically flat surface on a single crystal.Wagner and Ross have recently shown that a well defined reconstructed surface of Pt prepared in UHV is changed to a ( 1 x 1) structure on contact with an aqueous solution.'' The lack of surface structures as shown in fig. 2 seems to be an1354 STM of a Platinum { 11 l} Surface 50 8, 50 ‘A Fig. 3. STM images of a single-crystal Pt{ 1 1 l } surface after 50 potential cycles between 1.05 and -0.25 V us. SCE. The tunnelling current was 4 nA and tip potential was 50 mV. indication that the surface is atomically flat; this surface is designated the Pt{ 11 1)-( 1 x 1) surface.’l However, it is noteworthy that the final treatment in a hydrogen+oxygen flame needs to be carried out under certain conditions, both the temperature of the flame and the transfer of the Pt to the hydrogen-saturated water seem to be critical.We have sometimes observed an appreciable degree of roughening for a ‘virgin’ Pt{ 1 1 l} surface prepared by cooling in an atmosphere containing oxygen. The effect of annealing and subsequent cooling on the voltammogram for hydrogen adsorption-desorption has been investigated in detail by Motoo and F u r ~ y a . ~ They reported a quite different voltammogram from that shown in fig. 1 (a) for a Pt{ 11 l} surface cooled in an oxygen atmosphere. The above result strongly suggests that a contact with oxygen must be avoided in the transfer of the Pt to the hydrogen-saturated water in order to obtain an atomically smooth, clean surface.After the observation of the unactivated Pt(l1 l} face by STM as shown in fig. 2, the Pt electrode potential was cycled for 50 times at a scan rate of 0.2 V s-l, (the same procedure used in the electrochemical study shown in fig. 1). During the potential cycling, the tunnelling tip was retracted from the surface and disconnected from the z piezo feedback loop. Fig. 3 shows typical STM images obtained after the electrochemical potential cycles. Fig. 3(a) and 3(b) were obtained at different positions of the same sample with the same magnification. In fig. 3(a) it is clearly seen that nearly all the area of the almost completely atomically flat surface has been roughened by tbe electrochemical treatment employed here.However, there were a few areas (ca. 200 A in diameter) which seemed to be still atomically flat even after 50 potential cycles, as shown in fig. 3(b). Nevertheless, the percentage of these flat regions was estimated from many STM images obtained in this study to be < 10% of the total surface area. As shown in fig. 3, semi-spherical domains have predominantly been observed. TheseS. Sugawara and K. Itaya 1355 domains seem to be randomly distributed over the surface, with no particular direction of orientation preferred. Fig. 3(c) is a typical image obtained on a roughened surface with a larger magnification. Tbe diameter and height of the semi-spherical domains are in the ranges 20-30 and 5-10 A, respectively. Note that within a short range there seems to be an ordering of the semi-spherical domains.The electrochemically activated Pt{ 1 1 l} surface seems to consist of hexagonal cells, each of which is semi-spherical in shape, as is shown in fig. 3(c). Many saw-toothed lines hape also been observed, as shown in fig. 3 (c). The height of each tooth is in the range 3-6 A, which seems to correspond to mono- or di-atomic steps. These sharp features might be attributed to the existence of adatoms or kinks on the surface. We have observed similar semi-spherical domains, as well as regular parallel- terrace structures, on polycrystalline Pt after applying the same electrochemical 24 The result shown in fig. 3 suggests that the crystallographic orientation of each of the small single crystals, which comprised the polycrystalline sample examined previously, were such that Pt{ 11 l} surfaces were exposed where the semi-spherical domains have predominantly been observed.It will now be of particular interest to observe different single-crystal surfaces such as Pt{ 100) and Pt{ 1 lo). Finally, it is noteworthy that structures formed by similar potential cycling for Pt( loo} and Pt{ 1 1 11 surfaces have recently been reported in detail by Wagner and ROSS." Using LEED spot-profile analysis, they found that excursions to higher anodic potentials induced the formation of broadly distributed mono-atomic-height up-and-down steps, stating that such a 'randomly stepped surface' was perhaps better visualized as a ' random mesa structure '.'I As pointed out by these authors, a unique transformation from reciprocal-space to real-space structure in the LEED analysis is impossible in the case of the randomly stepped surface.However, the result shown in fig. 3 is a real-space physical structure caused by the oxidation-reduction cycling. We believe that the semi- spherical domains discussed above must correspond to the random mesa structure proposed by Wagner and Ross. A more detailed study changing both the number of potential cycles and the upper potential limits is now under investigation. The authors thank Professor N. Furuya (Yamanashi University) for his guidance in the preparation and treatment of Pt single crystals. References 1 B. E. Conway, Progress in Surface Science (Pergamon Press, 1984), vol. 16, pp. 1-138, and references 2 F.G. Will, J. Electrochem. Soc., 1965, 112, 451. 3 J. Clavilier, D. Armand and B. L. Wu, J . Electroanal. Chem., 1982, 135, 159. 4 C. L. Scortichini and C. N. Reilley, J. Electroanal. Chem., 1982, 139, 247. 5 S. Motoo and N. Furuya, J . Electroanal. Chem., 1984, 172, 339. 6 E. Yeager, J. Electrochem. Soc., 1981, 128, 160C. 7 A. S . Homa, E. Yeager and B. D. Cahan, J. Electroanal. Chem., 1983, 150, 181. 8 A. T. Hubbard, Acc. Chem. Res., 1980, 13, 177. 9 J. L. Stickney, S. D. Rosasco, G. N. Salaita and A. T. Hubbard, Langmuir, 1985, 1, 66. cited therein. 10 P. N. Ross Jr and F. T. Wagner, Advances in Electrochemistry and Electrochemical Engineering (John 11 F. T. Wagner and P. N. Ross Jr, Surf. Sci., 1985, 160, 305. 12 K. Yamamoto, D. M. Kolb, R. Kotz and G. Lehmpfuhl, J. Electroanal. Chem., 1979, 96, 233. 13 R. Sonnenfeld and P. K. Hansma, Science, 1986, 232, 21 1. 14 J. Schneir, R. Sonnenfeld, P. K. Hansma and J. Tersoff, Phys. Rev. B, 1986, 34, 4979. 15 B. Drake, R. Sonnenfeld, J. Schneir and P. K. Hansma, Surf. Sci., 1987, 181, 92. 16 R. Sonnenfeld and B. C. Schardt, Appl. Phys. Lett., 1986, 49, 1172. 17 H. Y. Liu, F. F. Fan, C. W. Liu and A. J. Bard, J . Am. Chem. SOC., 1986, 108, 3838. 18 C. W. Liu, F. F. Fan and A. J. Bard, J . Electrochem. SOC., 1987, 134, 1038. Wiley, Chichester, 1984), vol. 13, pp. 69-1 12.1356 STM of a Platinum ( 1 1 1 ) Surface 19 F. F. Fan and A. J. Bard, Anal. Chem., 1988, 60, 751. 20 K. Itaya and S. Sugawara, Chem. Lett., 1987, 1927. 21 K. Itaya, K. Higaki and S. Sugawara, Chem. Lett., 1988, 421. 22 L. Vazquez, J. Gbmez, A. M. Barb, N. Garcia, M. L. Marcos, J. G. Velasco, J. M. Vara, A. J. Arvia, 23 J. Gbmez, L. Vazquez, A. M. Bard, N. Garcia, C . L. Perdriel, W. E. Triaca and A. J. Ariva, Nature 24 K. Itaya, S. Sugawara and K. Higaki, J. Phys. Chem., 1988, 92, 6714. J. Presa, A. Garcia and M. Aguilar, J. Am. Chem. Soc., 1987, 109, 1730. (London), 1986, 323, 612. Paper 8/02146E; Received 31st May, 1988

ISSN:0300-9599    

DOI:10.1039/F19898501351

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. I, 1989, 85(6), 1357-1363 Measurement of Activity Coefficients, Mass-transfer Coefficients and Diffusion Coefficients in Multicomponent Liquid Mixtures by Reversed-flow Gas Chromatography? Pericles Agathonos and George Karaiskakis" Physical Chemistry Laboratory, University of Patras, Patras, Greece Reversed-flow gas-chromatography sampling is a new method for studying heterogeneous catalysis, diffusion, adsorption and evaporation. We report its application to the simultaneous determination of mass-transfer coefficients for the evaporation of multicomponent liquid mixtures and the diffusion coefficients of vapours from these liquid mixtures into the carrier gas. Using suitable mathematical equations the gaseous equilibrium concentration of each component under study, in the pure state and in the mixture, has been determined. Using these equilibrium concentrations, together with the mass-transfer coefficients and diffusion coefficients already determined, activities and activity coefficients were calculated.The liquid mixtures used were : n-hexane-n-heptane, n-hexane-methanol, n-hexane-ethanol, n- hexane-butan-2-01, n-hexane-acetone, n-hexane-methanol-butan-2-01, n- hexane-ethanol-butan-2-01 and n-hexane-methanol-butan-2-01-water. All the activity coefficients found were compared with those calculated by the UNIFAC (universal quasichemical functional-group activity coefficients) method. Most problems in chemical-engineering design are concerned with separation operations (e.g. distillation and extraction). For a rational design of such separation processes, we require quantitative information on phase equilibria and on interface transport in multicomponent liquid mixtures.Satisfactory experimental equilibrium and interface transport data are seldom available for the particular conditions of composition, temperature and pressure required in a particular design problem. A new method, termed reversed-flow gas chromatography (r.f.g.c.), which gives information not only on phase equilibria,. but also on interphase transport in multicomponent liquid mixtures, is presented in the present work. This method is applied to various physico-chemical measurements. It consists of reversing the direction of flow of the carrier gas from time to time. If other gases are contained in the carrier gas and their concentration depends on a rate process within the chromatographic column, each flow reversal creates a perturbation in the chromatographic elution curve in the form of extra peaks (termed 'sample peaks').Then, by repeatedly reversing the flow of the carrier gas, a repeated sampling of this rate process is performed. and then to the dehydration of alcohols3 and the deamination of primary arnine~.~ The method has been extended to the determination of adsorption equilibrium con~tants,~ rates of catalyst dehydration,6 Lennard-Jones parameters,' molecular diameters and critical volumes in gases8 and diffusion coeffi~ients.~-'l The method has recently been reviewed.12 Recently r.f.g.c. has been used to determine activity coefficients for the alcohol component in binary liquid mixtures of alcohol and water at a constant temperature and (Greece).R.f.g.c. was first applied to kinetic studies in heterogeneous Presented in part at the 1988 World Chromatography and Spectroscopy Conference, May 16-17, Corfu 46 1357 FAR 11358 Thermodynamics of Solutions oven 2 I I ' I Fig. 1. Experimental set up for measuring mass- transfer coefficients, diffusion coefficients and activity coefficients in multicomponent liquid mixtures by the reversed-flow gas-chromatography method. various alcohol mole fractions, and also to estimate excess partial molar thermodynamic functions of mixing for alcohols in water.13 In addition, r.f.g.c. has been used to determine mass-transfer coefficients for the evaporation of the alcohol component at various mole fractions from alcohol-water mixtures and diffusion coefficients of the alcohol vapour into the carrier gas.l* The objective of this work was to determine activity coefficients, together with mass- transfer coefficients, for the evaporation of multicomponent (with two, three or four substances) liquid mixtures.The difference from previous studies is that, while the flame ionization detector used previously did not detect water, now all the vapour components are detected and so a separation chromatographic column, filled with an adsorbent, is necessary. The extension of the reversed-flow gas-chromatography technique to multicomponent mixtures is extremely attractive because, while the number of pure fluids of interest in chemical technology is large, the number of different mixtures is much larger.Experimental The solutes used were n-hexane, n-heptane, methanol, ethanol and acetone (all Uvasol grade from Merck A.G.) and butan-2-01 (laboratory-reagent grade from B.D.H.). The carrier gas was helium of 99.99% purity from Linde (Athens). The apparatus used and the experimental procedure followed have been described elsewhere ;I5 only slight modifications were made. A conventional gas chromatograph (Pye Unicam, series 104) contained in its oven (fig. 1, oven 1) two sections of lengths I' and 1 of a stainless-steel chromatographic column [(loo+ 100) cm x 4 mm i.d.1 containing no chromatographic material. A stainless-steel diffusion column of length L (100 cm x 4 mm i.d.) was connected perpendicularly at its upper end to the middle of the column I' + 1.At the lower end of column L a 4 cm glass tube containing 0.5 cm3 of the pure liquid or liquid mixture was connected. The end D, of the column of length I'+Z was connected to the carrier-gas supply while the other end D, was connected to the separation column of length L', which was placed in oven 2 in another gas chromatograph. The end of this column was connected to the flame ionization detector via a six-port valve. The length of the separation column L' for the experiments with theP. Agathonos and G. Karaiskakis 1359 66 60 54 ro Imin Fig. 2. A reversed-flow chromatogram, showing two sample peaks for the diffusion of butan-2-01 (a), ethanol (b) and n-hexane (c) vapours into helium at 333.2 K and 1.2 atm.n-hexane-n-heptane mixture was 51 cm and it was packed with y-Al,O, (100-120 mesh from MCB), while for the experiments with alcohols and other substances it was of length 43 cm and was packed with 20% Carbowax 20 M on Chromosorb P, 60-85 mesh. At a given time after the liquid or the liquid mixture was placed in position, an asymmetric concentration us. time curve for the vapours of the pure liquid or the liquid mixture was recorded, and was seen to rise continuously and approach a limiting plateau. This concentration us. time curve consists of one, two or more numerous substances. During the rise period, and also when the plateau was reached, flow reversals for 12 s [from the x direction (fig. 1) to the opposite one and vice uersa] were effected by means of the six-port valve.This time period was shorter than the gas hold-up time in column sections Z’, I and L’. When the gas flow was restored to its original direction, sample peaks (one, two or more) were recorded (fig. 2). This reversal of the flow of the carrier gas was repeated several times with the same duration of backward flow. This gave rise to a series of peaks corresponding to various times to from the beginning of the experiment. The pressure drop along column I’ + 1 + L’ was ca. 0.2 atm.? The working temperature range was 298.2-338.6 K for the evaporation of the mixtures (oven 1) and 343.2- 473.2 K for the chromatographic material (oven 2). The volumetric carrier-gas flow rate, V, at ambient temperature was 0.55 cm3 s-l. t 1 atm = 101 325 Pa.46-21360 Thermodynamics of Solutions Results and Discussion In a previous paper15 it was shown that each sample peak produced by a short flow reversal is symmetrical, and its maximum height h from the final baseline is given by 2k, Dc, v(k, L + 0) h = 2c(l', to) = { 1 - exp [ - 2(k, L + D ) t,/L2]) where c ( t , to) is the vapour concentration at x = I' (cf. fig. l ) , the time to is measured from the moment of placing the liquid at the bottom of column L to the last backward reversal of gas flow, kc is the mass-transfer coefficient for solute evaporation, D is the diffusion coefficient of the solute vapour into the carrier gas, c, is the concentration of the vapour in equilibrium with the bulk liquid phase at the working temperature and u is the linear velocity of the carrier gas.Using eqn (1) the values of D and k, can be determined for the substances under study, as described in detail e1~ewhere.l~. l5 As has been shown in a previous paper13 the equilibrium concentration, c,, of a solute is given by c, = vh, [(LID) + (1 / k c ) ] (2) 2 where h, is the maximum value (at infinite time) for the peak height. This concentration can be calculated using the above equation, as all quantities on the right-hand side are known. If experiments are performed with liquid mixtures giving c, and with pure solutes leading to c:, the ratio C,/C: is equal to p / p * which is the activity a of each component in the liquid mixture, assuming that the deviation of the solute vapour from ideal behaviour is small. Thus where co, h,, D and k, refer to the component in the mixture and c t , h z , D* and k,* to the pure component.We can also calculate the activity coefficients, y , for the components of the mixture using the relationship Several experiments were performed with various binary liquid mixtures, in order to determine mass-transfer coefficients for the evaporation of the compounds of the mixtures, diffusion coefficients of the vapours of these compounds into helium, as well as activities and activity coefficients of the components of the mixtures. These results are summarized in table 1. All values of the diffusion coefficients were calculated using the Hirschfelder- Bird-Spotz (HBS) equation,16 but the diffusion coefficient for butan-2-01 was calculated using the Fuller-Schettler-Giddings (FSG) method.17 A comparison of the diffusion coefficients found with those calculated theoretically permits the calculation of the method's accuracy, which is defined as If we disregard the high value of the mass-transfer coefficient for n-hexane in solution with methanol and the high value of the diffusion coefficient of methanol vapour into the carrier gas helium in the same solution, both of which can be attributed to accidental errors, the following conclusions hold for all other systems.The mass-transfer coefficients for the liquids in solution are smaller than those in pureP. Agathonos and G. Karaiskakis 1361 Table 1. Mass-transfer coefficients of pure liquids and liquid mixtures, diffusion coefficients into a carrier gas (helium) and activity coefficients (experimental, calculated and literature values) at various temperatures and a pressure of 1.2 atm pure substance or mixture D/ 1 OP3 cm2 s-I accuracy kc this T/K .'c /lop4 cm s-I work calcd (YO) Y,,,~, ycalcd ylit" n-hexane methanol butan-2-01 n- hexane acetone n-hexane ethanol n-hexane- methanol n-hexane- butan-2-01 n-hexane- acetone n-hexane- ethanol ethanol- n-hexane n- hexane- n-heptane 333.2 333.2 333.2 318.2 3 18.2 298.2 298.2 333.2 333.2 3 18.2 298.2 298.2 318.2 1 1 1 1 1 1 1 0.054 0.946 0.094 0.906 0.1 0.9 0.1 0.9 0.1 0.9 0.097 0.903 162 355 49 67 123 50 44 497 279 79 16 66 29 38 21 31 38 34 25 320 317 0.9 794 604 23.9 337 317 5.9 305 292 4.3 402 397 1.2 247 261 5.7 392 398 1.5 308 317 2.9 1246 604 51.5 460 317 31.1 424 317' 25.2 295 292 1.0 393 397 1.0 246 261 6.1 346 398 15.0 394 398 1.0 249 261 4.8 297 294 1.0 262 251 4.2 12.46 13.06 12.21 1.88 3.27 3.34 6.20 3.76 3.40 4.83' 6.01 5.35 5.92 9.20 8.01 7.13 0.64 0.97 - 0.88 0.96 - ~ " Literature values given by Fredenslund et a l l 8 other values using the HBS method.values using the UNIFAC method. Theoretical values using the FSG method; all Values calculated from the empirical eqn (6); all other liquids, as observed previously.14 The values of the diffusion coefficients of the vapours are very close to the theoretical ones. All the diffusion coefficients show that the effective diffusion coefficient in a multicomponent mixture is approximately equal to the diffusion coefficient of each component in the pure carrier gas. The presence of chromatographic materials (y-Al,O, and 20 % Carbowax 20 M on Chromosorb P) in column L' does not seem to influence the values of the diffusion coefficients found.The values of the activity coefficients for the liquids in solution (especially the values for the solutions of n-hexane and simple alcohols) are close to those determined theoretically by the UNIFAC (universal quasichemical functional-group activity coefficients) method based on the quasichemical theory of liquid solutions, and described by Fredenslund et a1.l' and Oishi and Prausnitz.lg These values, ycalcd, and those that Fredenslund et al.la present as typical experimental values, ylit, are given in the last two columns of table 1 . Our experimental values (especially those for n-hexane in solutions with methanol and ethanol) are closer to the literature experimental values than the theoretical ones.Apart from the UNIFAC method, there are several other methods and many empirical equations for predicting activity coefficients in binary mixtures. For instance, the activity coefficients at infinite dilution, 7 7 , of alkanes in ketones can be estimated by the empirical equation given by Reid et al.*O1362 Thermodynamics of Solutions Table 2. Mass-transfer coefficients of liquids in multicomponent mixtures, diffusion coefficients into a carrier gas (helium) and activity coefficients (experimental and calculated values) at constant temperature (333.2 K) and pressure (1.2 atm) D/ I 0-3 cm2 s-' kc this accuracy Yexptl Ycalcd mixture X cm s-l work calcd ( O h ) butan-2-01- ethanol- n-hexane butan-2-01- methanol- n-hexane butan-2-01- methanol- n-hexane- waterb 0.807 0.172 0.02 1 0.807 0.172 0.021 0.360 0.077 0.009 0.554 103 229 20 1 112 222 140 135 383 230 320 493 315 322 717 318 30 1 60 1 320 317" 484 317 317" 604 317 317" 604 317 0.9 1.8 0.6 1.6 15.8 0.3 5.3 0.5 0.9 0.678 1.047 2.143 0.789 1.053 2.337 1.427 1.121 0.999 1.084 4.167 1.002 1.185 4.018 1.557 0.8 19 - " Theoretical values using the FSG method; all other values using the HBS method.detectable by a flame ionization detector. Not In this equation N , and N2 are the total numbers of carbon atoms in the alkane and ketone, respectively, Ni and N i are the numbers of carbon atoms in respective branches of the ketone, and E , n and 0 are constants depending on the system and the temperature, which are given by Reid et aL2' The application of this equation to the system n-hexane-acetone (x, = 0.1 and T = 318.2 K) resulted in a value of y? of 4.83, instead of 3.76 calculated by the UNIFAC method, which is closer to our experimental value (cf. table 1).Following the experiments with binary mixtures, experiments were performed with two ternary mixtures, and with one quaternary mixture. The results of these experiments are listed in table 2. With respect to the results of table 2 the following tentative conclusions can be drawn. (i) While in binary mixtures the mass-transfer coefficients of the two components are lower than those of the same compounds in the pure state, in ternary and quaternary mixtures the mass-transfer coefficients are in some cases higher than those of the pure compounds.This is probably due to the irreversible adsorption of these components on the solid material contained in column L' in the presence of the other components of the mixtures. (ii) The diffusion coefficients of the vapours from these liquid mixtures into the carrier gas helium are very close to those determined theoretically for the diffusion of the pure compounds into helium. (iii) The activity coefficients for the alcohol compounds in the multicomponent mixtures found by the present method are relatively close to those determined by the UNIFAC method. The abnormally large divergences for the activity coefficients of n- hexane might be attributed again to the separation material in column L'. For instance, while the adsorption of pure n-hexane on the separation material might be a reversible process, the adsorption of n-hexane in the presence of the other components of the mixtures might not be reversible, because of the appearance of competitive effects.Clearly, further work is necessary regarding the simultaneous study of the properties of the two interfaces, e.g. liquid-vapour and vapour-solid, related to this problem. It is hoped that the present paper, together with previous ones,13-15 will help to introduce workers in various fields of separation processes (e.g. distillation andP. Agathonos and G. Karaiskakis 1363 extraction) to the technique of reversed-flow gas chromatography, which is a new tool for studying the thermodynamic properties of solutions.We acknowledge the help of Mrs M. Barkoula. References 1 N. A. Katsanos and I. Georgiadou, J. Chem. SOC., Chem. Commun., 1980, 242, 640. 2 N. A. Katsanos, J. Chem. SOC., Faraday Trans. I , 1982, 78, 1051. 3 G. Karaiskakis, N. A. Katsanos, I. Georgiadou and A. Lycourghiotis, J. Chem. Soc., Faraday Trans. 4 M. Kotinopoulos, G. Karaiskakis and N. A. Katsanos, J. Chem. SOC., Fararday Trans. I , 1982, 78, 5 G. Karaiskakis, N. A. Katsanos and A. Niotis, J. Chromatogr., 1982, 245, 21. 6 G. Karaiskakis, A. Lycourghiotis and N. A. Katsanos, Chromatographia, 1982, 15, 351. 7 G. Karaiskakis, J. Chromatogr. Sci., 1985, 23, 360. 8 G. Karaiskakis, A. Niotis and N. A. Katsanos, J. Chromatogr. Sci., 1984, 22, 554. 9 G. Karaiskakis, N. A. Katsanos and A. Niotis, Chromatographia, 1983, 17, 310. I, 1982, 78, 2017. 3379. 10 N. A. Katsanos and G. Karaiskakis, J. Chromatogr., 1982, 237, 1. 11 N. A. Katsanos and G. Karaiskakis, J. Chromatogr., 1983, 254, 15. 12 N. A. Katsanos and G. Karaiskakis, Adv. Chromatogr., 1984, 24, 125. 13 N. A. Katsanos, G. Karaiskakis and P. Agathonos, J. Chromatogr., 1986, 349, 369. 14 G. Karaiskakis, P. Agathonos, A. Niotis and N. A. Katsanos, J. Chromatogr., 1986, 364, 79. 15 G. Karaiskakis and N. A. Katsanos, J. Phys. Chem., 1984, 88, 3674. 16 R. B. Bird, W. E. Stewart and E. N. Lightfoot, Transport Phenomena (Wiley, Chichester, 1960), p. 51 1. 17 E. Fuller, P. Schettler and J. C. Giddings, Znd. Eng. Chem., 1966, 58, 19. 18 A. Fredenslund, J. Gmehling, M. Michelsen, P. Rasmussen and J. M. Prausnitz, Ind. Eng. Chem., 19 T. Oishi and J. M. Prausnitz, Ind. Eng. Chem., Process Des. Dev., 1978, 17, 333. 20 R. C. Reid, J. M. Prausnitz and T. K. Sherwood, The Properties of Gases and Liquids (McGraw-Hill, Paper 8/02186D; Received 1st June, 1988 Process Des. Dev., 1977, 16, 450. New York, 1977), p. 336.

ISSN:0300-9599    

DOI:10.1039/F19898501357

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. I, 1989, 85(6), 1365-1372 The Structure of Concentrated Aqueous Ammonium Nitrate Solutions P. A. M. Walker, D. G. Lawrence and George W. Neilson* H. H. Wills Physics Laboratory, University of Bristol, Bristol BS8 ITL J. Cooper I.C.I. Ltd, D.R.G., Ardeer Site, Stevenston, Ayrshire KA20 3LN Neutron diffraction studies have been carried out on heavy water solutions of ammonium nitrate at two salt concentrations, 18 and 12 mol/kg-'. The first-order difference method of isotopic substitution was applied to both ND: and NO;, thus enabling the determination of detailed structural information regarding the hydration of these ions. Both ions show relatively weak interactions with water. At the highest concentration, ca. 18 mol kg-', there is evidence which demonstrates a direct interaction between the ND,' and NO; ions.Ammonium nitrate, AN, solutions are of considerable technical and scientific interest. They are essential components in fertilisers, and non-military explosives where they are fabricated in the form of emulsions or as ternary mixtures of AN/alkaline-earth-metal nitrate/aqueous solution. The scientific interest in these systems stems from the complex phase diagram of AN, which has at least five phases at atmospheric pressure.'*2 Furthermore, mixtures of AN and certain metal nitrate salts at certain concentrations can form glasses in a reversible way.3 Despite the extensive studies of AN solutions very little is known of their micro- structural properties. One reason for this is that X-ray diffraction methods applied to aqueous solutions of AN are unable to resolve the local ionic structure because both NH,+ and NO; are only weakly observed compared with the solvent water molecules.In fact, several authors conclude that, based on X-ray diffraction results alone, NH,+ is indistinguishable from a water m01ecule~~~ and it is a matter of debate whether the nitrate ion can be considered as being hydrated at this condition being almost totally dependent on the type of cation pre~ent.~ One objective of the study described below is to establish the validity of such claims. The main aim of our work was to determine the ionic structure of AN solutions as a function of concentration. This was done by application of the first-order isotopic difference method of neutron diffraction* to nitrogen atoms (15N and ") in both the ammonium ions and the nitrate ions.The results of this study enabled us to focus on the local hydration of these two types of ions and draw inferences regarding their relative hydrating strengths. Experimental and Data Analysis Neutron diffraction experiments were carried out on a series of AN-heavy-water solutions (table 1). Each sample was prepared under glove-box conditions by standard procedures. The ratio of heavy water to water was determined from infrared spectrometry and the isotopic content of samples was measured using a mass spectrometer. 13651366 Structure of Concentrated Aqueous ND,NO, Table 1. molality molecular ratio, numberodensity 15N abundance sample label /mol kg-' R (D,O:AN) (Yo) solution 1 (10 mol kg-') "ND,"O, I 17.88 2.8 0.10 - 15ND,NN0,11 17.97 2.8 0.10 95.1 "D,15N0, I11 17.99 2.8 0.10 99.0 "ND,"NO, IV 11.99 4.2 0.1 99.8 'SND,"NO, V 11.99 4.2 0.1 99.8 solution 2 (12 rnol kg-') '5ND,'5N0,VI 11.99 4.2 0.1 - 0 .0 8 ) 1 - '2 0.04. !i e Y n 0.00. Q -0.04 I 1 I 0 4 8 12 16 k /A- Fig. 1. First-order difference function A,,(k) for ND; ion in 18 mol kg-' ND,NO, in heavy water. Crosses represent raw data; full curve back Fourier transform of G,,(r) shown in fig. 3. 0.08 - IL. 0.04 v1 EJ d" e Y h 0.00 -0.04 I I 1 4 8 12 16 klA-' Fig. 2. First-order difference function AN2(k) for NO; ion in 18 mol kg-l ND,NO, in heavy water. Crosses represent raw data; full curve back Fourier transform of GN,(r) shown in fig. 4. The diffraction data were gathered on the D4B diffractometer at the ILL.All experiments were carried out in vacuo under ambient conditions. The data were corrected for absorption, multiple and incoherent scattering effects, and put on an absolute scale of barns sr-l nucleus-' by reference to a vanadium standard.lO The structure factors, F(k), calculated in this way were used to calculate first-order difference functions A,(k) (fig. 1 and 2), and their Fourier transformations G,(r) [fig. 3 and 5 (later)] from which structural information concerning the hydration of both NDZ and ND; is inferred.P. A . M. Walker, D. Lawrence, G. W. Neilson and J . Cooper I367 I I I I 1 0 1 2 3 4 5 6 r / A Fig. 3. The ammonium ion total radial distribution function GN,(r), for 18 mol kg-' ammonium nitrate in heavy water.0.10' If we use the following notation for ammonium nitrate in heavy water: N,,,D,N,,,O, - D,O then : F(k) = C i bi[Soo(k) - 11 + CD 6i[S,D(k) - I] + 2 ~ 0 C , 60 6D[SoD(k)- 13 + 2cNl co 60 6Nl[SoN,(k) - 11 + 2cKl CD 6, 6 N l [ S D N l ( k ) - 11 + 2c0 'N, '0 6N,[SON2(k) - '1 + 2cLl 'X2 'D 6N2[SDNp(k) - '1 + +'if 6i2[SN2Ny(k) - '1. 6kl[SN1 Nl(k)- 'I2 + 2cN1 'N, 'XI 6N2[SN1 N2(k> - '1 c, is the atomic concentration of atom ' a ' whose neutron-scattering length is 6,, and Sclp is the partial structure factors of the atom pair ap*. The Fourier transformation of SZp(k)- 1 is [g,&r) - 11 = [Sclp(k) - 13 k sin kr dk 2n ' s pr where p is the total number density of the solution. A first-order difference ANi(k) can be formed from the F(k)s of two isotopically distinct but chemically identical solutions which have had the nitrogen atoms of their ammonium ions (or nitrate ions) changed from NN to I5N.In particular A,i(k) = W ) - m k ) = AISNio(k>- Il+B[S,p(k)- lI+C[SNiNj(& ll+DISNiNi(k>- 11 where B = 2cD C N t bD A6N2, D = cKt(6iz - b;z) and A6Nz = (6Nz - bX). It is straightforward to show that the Fourier transformation of ANi(k) can be written as: where E = -(A+B+C+D). GNz(r) = AgNtO(r) + BgiYl D('> + 'gN, N,('> + DgNt N,(') +1368 Structure of Concentrated Aqueous ND,NO, 0.01 - 0.00 L E e z -0.01 h v 'U -0.02 Fig. 4. As fig. 3, but the ordinate scale expanded. Full curve, GN,(r) for ND: in 18 mol kg-l ammonium nitrate heavy water solution; dashed curve, G,,(r) for ND,f in 12 mol kg-' ammonium nitrate heavy-water solution.Fig. 5. The nitrate ion total radial distribution function GN,(r), for 18 mol kg-l ammonium nitrate in heavy water. The truncation of the Fourier sum in k-space produces spurious ripples in the trans- form at short distances. Before the first intra-molecular peak, we have, as is usual, set G,(r) = E, as r + 0. Clearly it would be wrong to do so after this point, though it can be seen from the fact that the values appear to drop below the limit, E, that the low r oscillation still produces a small distortion. We have chosen to present the unadjusted data (fig. 3-6). However, techniques of back transformation and truncation were used to identify real features from those produced artificially. The uncertainties involved were taken into account when the errors in table 3 were calculated. Structural analysis follows from the calculated GN(r)s where peak positions can be determined directly and coordination numbers from integration over ranges in r .For example the number of atoms of type x in the range rl < r ,< r2 around N, say is given by: nGl = pc, 1: 4nr2g",,(r) dr.P . A . M. Walker, D. Lawrence, G. W. Neilson and J . Cooper 0.10 0.00 - L $ e h -0.01 t u " -0.02 2 3 5 1369 Fig. 6. As fig. 5, but the ordinate scale expanded. Full curve, GN2(r) for NO; in 18 rnol kg-' ammonium nitrate heavy water solution; dashed curve, GN2(r) for NO; in 12 rnol kg-' ammonium nitrate heavy-water solution. Table 2. Coefficients of G(r) in barns/lO-, sample differences A B C D E solution 1 (18 mol kg-' AN/D,O) "ND4NN0,-'5ND,NN0, 6.148 1 1.61 8 1.70 1 1.459 - 20.926 NND4NN0,-"ND4'5N0, 6.395 12.102 1.768 1.524 - 21.789 solution 2 (12 rnol kg-' AN/D,O) "ND4NN0,-'5ND4NN0, 5.15 10.20 1.15 0.18 - 16.68 '5ND,NN0,-'5ND,'5ND, 5.15 10.20 0.80 0.18 - Table 3.Intramolecular parameters of ND,' and NO; ND; NO; concentration/ rnol kg-' rN,/A Ar1I2/A n rNo/A AF/A n 18 1.04 (2) 0.31 (3) 4.04 (10) 1.23 (2) 0.28 (4) 3.05 (10) 12 1.02 (2) 0.29 (5) 3.9 (1) 1.23 (2) 0.28 (4) 3.0 (1) 7.8 (NaNO,) 1.23 (2) 0.36 (5) 3.0 (3) - 5 (ND4C1) 1.05 (2) 0.29 (3) 4.0 (1) - - - - - Table 2 shows that the G(r)s are dominated by N-0 and N-D correlations. Consequently, the interactions between the ions themselves can only be inferred from the first-order difference functions. However, as we shall see below it is possible to obtain information about the possibility of direct ND; NO, contacts by comparing results as a function of concentration.1370 Structure of Concentrated Aqueous ND,NO, Results and Discussion The first-order difference functions A,(k) and the total radial distribution functions GNi(r) for ND,+ and NO, are shown in fig.1-6. The results are in broad agreement with earlier studies of both these ions in other aqueous solutions, where we found the G(r)s to be relatively featureless except for the presence of the strong intra-molecular peak. ND,f Coordination (fig. 3, 4) In boLh solutions the ND,+ structure exhibits the same strong intra-molecular peak at ca. 1.04 A (table 3) in good agreement with that found in our previous studies1'* l2 and in a diffraction study of the p 0 ~ d e r .l ~ Integration of this peak gives values of about 4 and gives credence to our data normalisation procedures. The rest of the G(r) is, as expected, relatively featureless. However, at the highest concentration 18 mol kg-' where there are effectively less Jhan three D,O molecules to each ND,NO, molecule, a small peak centred at ca. 2.15 A is observed (fig. 3 and 4). If this peak is assumed to contain 0 atoms a coordination number of 1 is obtained. There are two possible explanations for this observation. It could arise from an oxygen a$om of a near-neighbour water molecule whose D atoms are sited under the peak at 3 A, or it could arise from an oxygen atom belonging to an NO; anion. Because it is not present at the lower concentration we feel that the latter explanation is more probable and our result provides the first experimental evidence of ion pairing in aqueous ammonium nitrate solutions.One possible configuration for this interaction is sbown in fig. 7. As we will show below, the nitrate-deuterium distance is estimated to be 3 A, a result consistent with the G,(r)$or the nitrate ion shown in fig. 6, where the closest distance of approach would be 2.3 A. The longer-range structure of both solutions is broadly similar, and tbeir analysis can be carried out by integrating the two broad peaks centred at 3.0 and 3.4 A. Although our interpretation is no+ unique y e suggest the configuration for the 18 mol kg-l solution is one in which the peak at 3.0 A contains one N and about 6( 0.1) 0 atoms which belong to nearest-neighboyr water molecules.The peak centred at 3.4 A when integrated over the range 3.2-4.1 A can readily accommodate the two remaining 0 atoms of the NO, ion, the 12 D atoms of the nearest-neighbour hydration shell and the 5 or 4.5 other oxygen or deuterium atoms, respectively. It is not possible to be more precise because of the complexity of the solution. Interpretation of GN(r) for the 12 mol kg-l AN solution is also subject to the same uncerJainties as those of the 18 rnol kg-l solution. The results do not exhibit a peak at 2.15 A. We note that in the 12 mol kg-: solution no hydration structure is observed within a distance of approach of ca. 2.4 A. The structure beyond this distance is moved only slightly closer at the hjgher concentration.However, there is a relative increase in the size of the peak at 3.2 A and this increase we ascribe to an oxygen atom belonging to an NO, ion. The rest of the peak can accommodate 6 or 7 other oxygen atqms forming a near-neighbour weakly defined hydration shell. The peak centred at 3.4 A is similar to that for the 18 mol kg-l case and there is sufficient area under it to contain 14 D atoms, and one N atom, two 0 atoms and a remaining four other atoms. The general features of the NDI coordination are similar to those found in our earlier study of 5 mol kg-l ammonium chloride. However, there is one significant diffcrence in that, whereas the ND,+ in ammonium chloride shows no resolved peak at 2.9 A, one is clearly evident in the AN study described here.This result agrees well with a recent molecular dynamics study by Walker and Allen.'* Furthermore, these results are in better agreement with an earlier molecular dynamics study by Heinzinger and Szasz15 who found at infinite dilution a relatively strong correlation between NDZ and 10 water molecules. The present study at higher concentrations gives coordination numbersP . A . M. Walker, D. Lawrence, G. W. Neilson and J . Cooper 1371 D 1.04A D D 2.15 A - - - - - 0 1.23A 0 1.23A A L 0 Fig. 7. A two-dimensional projection of a possible ND,+ ... NO; configuration consistent with experimental results (fig. 4 and 6). Clearly, from the distances between the atoms, as shown, the ND; and NO, ions cannot be coplanar. appreciably less than this, and it is clearly of interest to determine the extent of concentration and counter-ion effects on ND,+ hydration.NO; Coordination (fig. 5, 6) In contrast to the ND: coordination, there is no obvious dissimilarity between the G(r)s for the two solutions. In both cases there appears to be no short-range structure. However, there is 9 small but appreciable difference in the shape of the G(r)s in the region around 2.8 A where a shoulder appears in the results at high coacentration and a peak is visible at lower concentration. If the region 2.14 < r < 3.0 A is assigned to N-D nearest-neighbour correlations, integration over this range gives coordination numbers of 3.3 at 12 mol kg-l and ca. 5 at 18 mol kg-l. This latter result is consistent with the above discussion concerning direct cation-anion contact.It is also possible that the remaining three deuterium atoms belong to water molecules weakly coordinated to the NO; ion in a configuration suggested originally by Caminiti et al. (ref. 16, fig. 2a) and subsequently identified by us17 in an earlier stgdy of NO; in 7.8 mol kg-' NaNO,. In contrast to our earlier paper,17 a peak at 2.05 A, which had been assigned to the D above the NO; plane, is not observed in this study. It will be interesting to see whether this observation is a result of the stronger Na+ cation influencing the NO, coordination. Experiments are in hand to investigate the effect of strong cations such as Ni2+ on the NO; hydration by studying AN/Ni(NO,), heavy water solutions using the difference methods described in this paper." The recent computer simulation by Walker and Allen14 of NO; hydration shows fair agreement with the above results.However, more realistic potentials are being developed in order to obtain a better fit to the experimental results. Conclusions and Future Prospects The first-order difference method has been applied to aqueous solutions of ammonium nitrate. The results show that at the highest concentration there is strong evidence for direct cation-anion contacts. Furthermore, as expected, the general form of the cation and anion coordination is relatively weak compared with ions such as Li'or Na+ on the one hand and Cl- on the other. There is a clear need for a more definitive picture before we can resolve the detailed ionic structure.One possible means is by elimination of the hydrogen correlations as could be undertaken using a 'null' water mixture,lg i.e. a mixed solution of water and heavy water with concentrations chosen such that correlations between nitrogen and hydrogen atoms are not present in either ANl(k) and ANz(k) or in their Fourier trans- forms, GN,(r) and GNP(r).1372 study AN solutions at even higher concentrations. Structure of Concentrated Aqueous ND,NO, The use of temperature is also expected to help in our analysis, and will enable us to The authors thank Professor J. E. Enderby for his helpful suggestions during the course of these investigations. We are also grateful to Mr P. Gullidge for assistance with the sample preparations, and Dr P. Chieux for his help with the diffraction experiments.The financial support of S.E.R.C. and I.C.I. is greatly appreciated. References 1 P. W. Bridgman, Proc. Am. Acad. Arts Sci., 1915, 51, 5999. 2 S. B. Hendricks, E. Posnjak and F. C. Kracek, J. Am. Chem. SOC., 1932, 54, 2766. 3 I. V. Vasilokova, M. P. Snarev and T. A. Znchenco, Servia Fiziki Khimiya (University of Leningrad), 4 A. H. Narten, J. Phys. Chem., 1970, 74, 765. 5 P. M. Vollmar, J. Chem. Phys., 1963, 39, 2236. 6 R. Caminiti, G. Licheri, G. Piccaluga and G. Pinna, J. Chem. Phys., 1977, 19, 371. 7 R. Caminiti, G. Licheri, G. Piccaluga and G. Pinna, J. Chem. Phys., 1978, 68, 1967. 8 A. K. Soper, G. W. Neilson, J. E. Enderby and R. A. Howe, J. Phys. Chem., 1977, 10, 1793. 9 J. E. Enderby and P. Gullidge, in Methods of Experimental Physics, ed. K. Skold and D. L. Price 10 J. E. Enderby and G. W. Neilson, Water, a Comprehensive Treatise, ed. F. Franks (Plenum Press, New I I N. A. Hewish and G. W. Neilson, Chem. Phys. Lett., 198 1, 84, 425. 12 S. Cummings, J. de Physique Colloq., 1984, C7, 131. 13 B. W. Lucas, M. Ahtec and A. W. Hewat, Acta Cryst., 1979, B35, 1038; Handbook of Chemical Physics 14 P. A. M. Walker and M. P. Allen, Molec. Simulations, 1989, 2, 307. 15 Gy. I. Szasz and K. Heinzinger, Nuturforsch, 1979, A340, 840. 16 R. Caminiti, G. Licheri, G. Paschina, G. Piccaluga and G. Pinna, J. Chem. Phys., 1980, 72, 4522. 17 G. W. Neilson and J. E. Enderby, J. Phys. C: Solid State Phys., 1982, 15, 2347. 18 D. G. Lawrence, unpublished report, University of Bristol, 1986. 19 D. H. Powell, G. W. Neilson and J. E. Enderby, J. Phys. Condensed Matter, submitted. 1971, 1, 74. (Academic Press, 1987) vol. 23B, p. 471. York, 1979) vol. 6, chap. 1. Edu. 1985 (C.R.C. Press Inc., Florida, 1985). Paper 81022081; Received 2nd June, 1988

ISSN:0300-9599    

DOI:10.1039/F19898501365

出版商:RSC    

年代:1989

数据来源: RSC