|
11. |
Acid–base equilibria in polyelectrolyte systems |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 82,
Issue 8,
1986,
Page 2353-2365
Hans Vink,
Preview
|
PDF (781KB)
|
|
摘要:
J . Chem. SOC., Faraday Trans. I , 1986, 82, 2353-2365 Acid-Base Equilibria in Polyelectrolyte Systems Hans Vink Institute of Physical Chemistry, University of Uppsala, P.O. Box 532, S-751 21 Uppsala 1, Sweden Acid-base equilibria in carboxylic polyelectrolyte solutions and gels have been investigated by studying ion-exchange equilibria involving hydrogen and alkali ions. The investigations were carried out with solutions of poly(acry1ic acid), poly(methacry1ic acid), carboxymethyl cellulose and a carboxymethyl cellulose gel. The acid-base dissociation constants of the polyacids were determined, and acid strength was found to increase in the order poly(methacry1ic acid), poly(acry1ic acid), carboxymethyl cellulose. In the latter case a marked decrease of acid strength with increasing degree of substitution was observed, which was attributed to increased fluctuations in the concentration of carboxylic groups at high degrees of substitution.In the presence of simple salt the ionisation of the polyacids increased with increasing salt concentration. The effect is analogous to the increased ionisation of low-molar-mass weak acids, caused by the salt-induced depres- sion of the ionic activity coefficients. However, the effect is more pronounced in polyelectrolytes and depends on the nature of the alkali-metal counterion, the ionisation being higher for K+ and Cs+ ions than for Na+ and Li+ ions. Acid-base reactions involving polyelectrolytes have been extensively studied in the past. These investigations have mainly been concerned with the potentiometric titration behaviour of weak polyacids, and the experimental results have been interpreted in terms of semiempirical relations, such as the extended Henderson-Hasselbalch equation,’ or by series expansion of the apparent dissociation constant of the polyacid.2 However, the interpretation of the potentionietric titration data is complicated by the fact that pH is not a rigorously defined thermodynamic quantity, as it is related to single-ion activities. This also deprives the corresponding dissociation constant of an exact thermodynamic definition.Alternatively, acid-base equilibria in polyelectrolytes can be studied by means of ion-exchange reactions. Thus, by determining the ion-exchange equilibrium involving hydrogen ions and univalent cation (an alkali-metal ion), between the polyacid phase and an ambient solution, the acid-base equilibrium of the polyacid can be studied.The method can be applied to polyelectrolyte solutions contained by a membrane, or to crosslinked polyelectrolyte gels (ion-exchange resin^).^ In the latter case ion exchange represents the only possible approach, since the pH of a gel is not measurable. In the present work the ion-exchange method has been used to study acid-base equilibria in solutions of the carboxylic polyacids poly(acry1ic acid) (PAA), poly(meth- acrylic acid) (PMA) and carboxymethyl cellulose (CMC), and a carboxymethyl cellulose gel. Ion-exchange equilibria between hydrogen ions and different alkali-metal ions were studied. A few experiments involving Ca2+ ions were also performed. Theoretical The thermodynamics of ion-exchange equilibria in polyelectrolyte systems have been considered in detail in an earlier a r t i ~ l e .~ Using the same formalism in the present work 23532354 Acid-Base Equilibria in Polyelectrolytes we express the condition for ion-exchange equilibrium between two univalent ions i and j in terms of the separation factor (1) ci c; - rj y; cjc; l/i y; where unprimed and primed quantities refer to the polyelectrolyte phase and ambient solution, respectively, ci and cj denote the total concentrations and yi and y j denote the stoichiometric activity coefficients. The bar over the activity coefficient for the polyelectrolyte phase indicates that it is corrected for the pressure difference between the phases.wherep andp’ are the pressures of the respective phases and is the partial molar volume. As the pressure difference is negligible in the present experiments we may delete the bar over the activity coefficient in eqn (1). Furthermore, since at moderate concentrations y; = y i (because the activity coefficients in a simple salt solution only depend on the ionic strength), we have To apply eqn (3) to specific ion-binding phenomena we have to introduce a model specifying the ion-binding reaction. For acid-base equilibria in polycarboxylic acids we have the reaction and the corresponding mass action law CO,H +CO;+H+ (4) where the parenthesis denote activities and K is the thermodynamic equilibrium constant In polyelectrolytes a complicating factor is the possibility that all acidic groups are not equivalent.To account for this possibility we assume that a number of different reaction equilibria exist : (7) Introducing concentrations (square brackets) and activity coefficients we may write [CO;Il [H+] rGrft = [CO,HIz Kl. Y l Summing eqn (8) over all 1 we may rearrange it in the form [COY] [H+] r - y f t = [CO,H] K (9)H . Vink 2355 Eqn (9) may be rearranged Thus it is possible to express to define the concentration dissociation constant Kc : the mass-action law in terms of the total concentrations of the reactants, the average activity coefficient 7- and the average equilibrium constant i% In the former quantity the activity coefficients y; and y, of ionised and un-ionised groups, respectively, are not separable.However, as the activity of uncharged groups is very little affected by the ionisation process we may, for a given polyelectrolyte, conventionally put y, = 1. Differences between the dissociation constants Kz arise from differences in the chemical environment at the binding sites of the carboxy groups. These comprise permanent structural differences of the binding sites and induced environmental effects arising from the ionisation of neighbouring carboxy groups. However, by choosing the same reference states in eqn (6) for structurally similar groups, only permanent structural differences give rise to differences in K,, the induced environmental effects only affecting the activity coefficient 7-. Note that in the presence of groups with different Kz values the average dissociation constant K is not a constant, but decreases with increasing neutralisation of the polyacid, as the distribution function [COOHI,/ [COOH] is shifted in favour of lower dissociation constants at high degrees of neutral- isation (because the more strongly acidic groups are neutralised first).For most polyelectrolytes considered in the present work (except CMC with high degrees of substitution) all carboxy groups can be considered as being equivalent. The average quantities K and 7- then reduce to the corresponding single quantities. In combining the ion-exchange equilibrium condition, eqn (3), with the mass action law, eqn (lo), we denote in eqn (3) hydrogen ions with the indexj = 1 and an arbitrary univalent cation with the index i = 2.Observing that y , in eqn (3) is the stoichiometric activity coefficient whereas y+H in eqn (10) refers to unbound (free) ions, we find that these quantities are related by the equation4 where ci = [H+] is the free hydrogen ion concentration and yf = y$ is the corresponding free-ion activity coefficient. We may also write cf f f 71 = - 71 = a, Y1 C1 where a, is the fraction of free hydrogen ions in the polyelectrolyte phase. Thus, we obtain from eqn (3) Considering all counterions of species 2 as free, y2 represents the free-ion activity coefficient. Therefore, if only electrostatic interactions are important and we have Yf = Y2 Although literature data in support of eqn (14) exist5-’ it should be considered an approximation, valid only at low polyelectrolyte concentrations and low degrees of ionisation of the polyelectrolyte. Otherwise effects depending on ionic size will become important and the ratio yfi/y2 may differ appreciably from unity.A check of the validity of eqn (1 5 ) may be obtained from measurements carried out with different alkali-metal ions.2356 Acid-Base Equilibria in Polyelectrolytes From stoichiometric considerations we obtain [H+] = C! = ale, [C02H] = C , - C: = (1 -a,) C, [CO,] = cp - (1 -a,) c, where cp is the stoichiometric equivalent concentration of the polyelectrolyte. Inserting eqn (16)-( 18) into eqn (10) we obtain where is the total degree of dissociation of the polyacid. [Note that according to eqn (12), a, is the degree of dissociation with respect to the unneutralised part of the polyacid.] Theconcentrations el, c2, c;, ci and cp areexperimentally determinable (see Experimental section).Thus the quantities a,, q, and K , are also experimentally determinable. Experimental Materials The following polyelectrolytes were used in the measurements. Poly(acry1ic acid) (PAA) was from B.D.H. Chemicals, Poole, with a weight-average molecular weight M , = 230000. Poly(methacry1ic acid) (PMA) was from the same manufacturer, molecular weight unspecified. In the case of carboxymethyl cellulose (CMC) the original sample (CMC 1) was obtained by carboxymethylating degraded linters of cellulose with sodium chloroacetate and concentrated alkali in a medium of propan-2-01. The reaction was carried out at 60 "C, in a nitrogen atmosphere. A second sample (CMC2) was obtained by recarboxymethylating a part of sample CMC 1 twice by the same procedure.The two samples had a degree of polymerisation of 360, and a degree of substitution of 1.07 for CMC1, 1.69 for CMC2. CMC gel was obtained by carboxymethylating cellophane (from Union Carbide, Chicago) as described in ref. (8). The gel formed a foil 0.2-0.3 mm thick and had a degree of substitution of 0.152. The polyelectrolytes were purified by prolonged dialysis first against a 0.1 mol dm-3 HC1 solution and later against distilled water. Performance of Measurements With the polyelectrolyte solutions ion-exchange equilibrium was established in a dialysis cell, essentially similar to the cell in ref. (9). The cell was made of Teflon, each half-cell having a volume of 12.5 cm3.A cellophane membrane was used. The equilibration was carried out in an air thermostat at 25 "C, where the cell was kept in slow non-uniform rotation for ca. 20 h. With the polyelectrolyte gel equilibrium was established in a tightly stoppered 100 cm3 Pyrex bottle containing ca. 9 g gel in the form of 3 x 3 cm leaflets and ca. 90 g solution. Also in this case the bottle was rotated in an air thermostat at 25 "C for ca. 20 h. Before each measurement the gel was regenerated in pure acidic form by leaching with a 0.1 mol dmP3 HC1 solution and distilled water. All relevant data concerning ion-exchange equilibria were obtained by analytically determining the concentrations of hydrogen ion and co-ion in the ambient solution, and using the equations for electroneutrality and mass balance, according to the following procedure.H.Vink 2357 The equations of mass balance are mc,+m’c; = C, (21) mc,+m’ci = C, (22) mc, + m’ch = C, (23) mcp = Cp (24) where m and m’ are the masses of polyelectrolyte phase and ambient solution, respectively, and c,, c,, c, and cp are the stoichiometric concentrations of hydrogen ion, alkali-metal ion, co-ion and polyacid, respectively (because all sampling operations were carried out by weighing, weight concentrations, mol per kg solution, are used in the present treatment). C,, C, C, and Cp are the total amounts (in mol) of the respective components. They satisfy the relations c, = (1 -p)G (25) c2 = c3 + $P (26) where 93 is the degree of neutralisation of the polyacid (p is determined by the amount of alkali-metal hydroxide added to the system).The electroneutrality condition yields c,+c, = c3+cp (27) c; + c; = c;. (28) In these equations c; and cb are determined analytically, while C,, C,, C,, Cp, m, m’ and p are known from initial conditions. By straightforward computations we obtain m’c; c1 = (1 -Q7)cp-m C, + m’(c; - ch) cz=pcp+ In experiments with the polyelectrolyte gel only the total mass A4 = m+m’ of the system was directly measurable. In this case the swelling curve of the gel for different degrees of neutralisation and different salt concentrations was established in separate experiments, and the value of m’ was obtained from the difference m’ = M-m. The swelling of the gel in a given solution was determined by equilibrating a piece of gel in the solution, wiping it gently with filter paper and weighing.The analysis of the ambient solution was carried out titrimetrically. A weighed part of the solution was titrated first with a standard 0.01 mol dm-, solution of NaOH to the phenolphthalein end-point and the same solution (or part of it) was then titrated potentiometrically with a standard 0.01 mol dm-3 solution of AgNO,. Either C1- or Br- was used as co-ion, the latter being preferred for its sharper end-point in the titration curve. Results and Discussion Polyelectrol yte Gel Acid-base reactions in a freely swelling gel are in general accompanied by volume changes. Volume changes also occur when the amount of simple salt in the system changes. The swelling characteristics of the gel are therefore of importance in the study of acid-base equilibria.The CMC gel used in the present investigation had a low degree of substitution (0.152), well below the solubility limit for CMC (0.5). No crosslinking of2358 A cid- Base Equilib Y ia in Poly elec t r oly t es 0.1 0.2 0.3 c$/rnol kg-' Fig. 1. Relative mass of the CMC gel in NaBr solution at varying degrees of neutralisation: + , p , = O ; O,g:=0.49and @,q= 1. the gel was therefore necessary, and the volume changes were in general small. This is shown in fig. 1 , where the relative change in the mass of the gel is represented as a function of salt concentration in the ambient solution for different degrees of neutral- isation of the gel. The mass of the gel in the acid form in pure water is taken as reference.We find that the swelling of the unneutralised gel is little affected by the addition of salt, whereas the effect of salt increases with increasing neutralisation. The ion-exchange experiments were carried out in different sequences to study the dependence of the acid-base equilibria on the degree of neutralisation, on salt concen- tration and on the nature of the counterion species. The first series of measurements were carried out at zero degree of neutralisation, p = 0. The results are shown in fig. 2, where the total degree of ionisation a, and the dissociation constant K , are plotted against the square root of the co-ion concentration in the ambient solution (which equals its ionic strength). We find that the dissociation constant increases with increasing ionic strength, the curve being linear in the region of low ionic strength.Thus, according to eqn (10) the activity coefficient y-y& of the polyacid decreases with increasing ionic strength, which is normal behaviour for electrolytes. The results differ for different alkali-metal ions. Thus the dissociation constant is higher for K+ ions than for Naf ions. A few measurements with CsBr and LiCl indicate that Cs+ and K+ ions and Li+ and Na+ ions behave similarly. These results clearly indicate that specific interactions between the polyelectrolyte and counterions are important. Similar effects were observed with the polyelectrolyte solutions, and they will be discussed in more detail below. In one series of measurements the degree of neutralisation of the gel was varied, while the ionic strength of the ambient solution was kept constant.The results are shown in fig. 3. We find that the dissociation constant K , is almost constant at low degrees of neutralisation but increases sharply as the degree of neutralisation approaches unity. AlsoH . Vink 23 59 0 0 0.1 0.2 0.3 (c:/mol kg-' )+ 0.2 at 0.1 0 Fig. 2. Dependence of the dissociation constant K,: (upper curves) and the total degree of ionisation % (lower curve) on simple salt concentration in the ambient solution; 0, NaBr; +, KBr; A, CsBr and 0, LiC1. I I 1 I I I I I 0.5 cp Fig. 3. Dependence of the dissociation constant K, on the degree of neutralisation 9, at constant concentration of NaBr (cj = 0.045 mol kg-l).2360 Acid-Base Equilibria in Polyelectrolytes 0.4 0.3 at 0.2 0.1 Fig.4. Dependence of the dissociation constant K, (0) and the total degree of ionisation %( +) on NaBr concentration in the ambient solution, at constant degree of neutralisation (q = 0.224). this effect has to be attributed to a decrease of the activity coefficient of the polyacid. Since the degree of substitution is very low, all carboxylic groups may be considered a.s equivalent and no change of the average thermodynamic equilibrium constant K is therefore expected. Finally, a series of measurements was performed in which the degree of neutralisation had a constant non-zero value, while the ionic strength of the ambient solution was varied within wide limits. The results are shown in fig. 4. In this case the behaviour is more complex, the dissociation constant passing through a minimum as the ionic strength increases.Consequently the activity coefficient of the polyacid passes through a maximum. The lowering of the activity coefficient at low ionic strength is probably due to an insufficient screening of the polyelectrolyte charges by the salt, whereas the depression of the activity coefficient at high salt concentrations is a normal ionic strength effect. The existence of a minimum in the dissociation constant seems to corroborate the assumption that the alkali-metal ions can be considered as free. A dominant binding reaction between the carboxy groups and Na+ ions would cause a steady increase in K, when the concentration of Na+ increases. Finally, some ion-exchange experiments were carried out with bivalent counterions (added salt CaCl,).The results, together with corresponding data for Na+ counterions, are displayed in fig. 5, where the separation factor (which equals a, for Na+) is plotted as a function of the square root of the co-ion concentration in the ambient solution. The curves illustrate the difference between the distribution laws for uni-univalent and uni-bivalent ion exchange. The steep upturn in the curve for Ca2+ ions reflects the fact that the ion-exchange equilibrium in the gel shifts in favour of the counterion with the highest valence when the ambient solution is d i l ~ t e d . ~ Polyelectrolyte Solutions Compared to polyelectrolyte gels measurements with polyelectrolyte solutions had the advantage that the masses m and m' of the polyelectrolyte and ambient solution phasesH.Vink 236 1 0 .I a1 0.05 n I + + + ‘t - \ U 0 0 .I 0.2 0.3 (c / equiv . kg - )+ Fig. 5. Dependence of the separation factor (a,) on simple salt concentration for uni- and di-valent counterions: 0, NaBr and +, CaCI,. were exactly known (from the volume of the cell and the density of the respective solution). It was also possible to vary the polyelectrolyte concentration, and study its effect on the acid-base equilibrium. All measurements with polyelectrolyte solutions were carried out at zero degree of neutralisation, 43 = 0. The results of the measurements with solutions of PAA, PMA and CMC are shown in fig. 6-9. For all polyelectrolyte solutions the results are similar to those obtained with the polyelectrolyte gel, the dissociation constant K, increasing linearly with the square root of the ionic strength of the ambient solution.The dependence of the dissociation constant on polyelectrolyte concentration, as indicated in fig. 8 and 9, is small, although a decrease of K , with decreasing polyelectrolyte concentration is discernible. Here also the ionisation is higher with K+ counterions compared to Na+ counterions (differences in the ionisation of carboxylic polyacids have also been observed in potentiometric titration experiments, although only at high degrees of ionisation6). Thus the assumption in eqn (14) that the activity coefficient of free hydrogen ions is equal to the activity coefficient of alkali-metal ions is not valid at finite electrolyte concentrations, and a correction factor has to be incorporated in the expression for a,, eqn (15).However, the difference between the curves obtained with the two alkali-metal ions decreases with concentration and seems to disappear at infinite dilution. The extrapolated values of the dissociation constant may therefore be considered to represent the true dissociation constant of the pure polyacid. still contains the activity factor ~ - y & , which differs from unity The constant2362 Acid-Base Equilibria in Polyelectrolytes 5 4 - I C1D Y .i 3 5 d) I 0 u? c -._ kV 2 1 0 01 0.2 0.3 (ci/mol kg-')t Fig. 6. Dependence of the dissociation constant K , of PAA on simple salt concentration : 0, NaBr +, for KBr. Concentration of the polyacid cp = 0.1907 equiv. kg-l. because of the ionisation of the pure polyacid.However, as the degree of ionisation of the polyacid is quite low, the charged groups may be assumed to be randomly distributed within the polyelectrolyte phase. The value of the activity factor F - y A = y? may therefore be estimated from the Debye-Huckel theory for a uni-univalent electrolyte and the thermodynamic dissociation constant K determined. All relevant data about the acid-base equilibria are listed in table 1. We find substantial differences in the dissociation constants for the carboxylic polyacids, the acid strength increasing from PMA to PAA and CMC. The low strength of PMA probably reflects the influence of the hydrophobic environment of the carboxy groups of that acid, whereas the higher acid strength of CMC is due to the influence of the hydroxy groups in cellulose.Contrary to some earlier investigations11912 the data for CMC display a marked dependence on the degree of substitution, the dissociation constant decreasing as the degree of substitution increases. This effect can to some extent be due to differences in the environment of the carboxy groups, which makes the dissociation constant an average quantity according to eqn (9). This is to be expected especially when the degree of substitution exceeds unity, since then the dehydroglucose units having two or three carboxymethyl substituents may differ considerably from the dehydroglucose units having only one substituent. However, there is a more specific polyelectrolyte effect which affects the dependence of K, on the degree of substitution.It arises from the inherent inhomogeneity ofH. Vink 2363 2 I I I 0.1 0.2 0.3 (c;/mol kg-')+ Fig. 7. Dependence of the dissociation constant K , of PMA on the concentration of NaBr. Concentration of the polyacid cp = 0.1024 equiv. kg-l. Table 1. Data for acid-base equilibria extrapolated to zero concentration of salt (NaBr) sample PAA PMA CMCl CMC2 CMCgel degree of substitution c,/equiv. kg-' g/ lop5 equiv. kg-l 4 c&+/mol kg-l Y% a 105 A- PK 1 0.190 7 1.67 0.009 31 0.001 78 0.912 1.52 4.82 1 0.102 4 1.02 0.009 93 0.001 02 0.932 0.95 5.02 1.07 0.052 3 0.049 0 0.002 56 0.897 3.93 13.2 11.8 1.69 0.084 2 7.50 0.029 4 0.002 48 0.899 6.74 4.17 0.152 0.203 0.031 3 0.006 35 0.850 3.76 20.5 17.4 ___-- a Computed from the equationlo 0.509(c&+$ log '' = - 1 + 1.84(&+)5' polymeric systems and reflects the difficulty of gauging the concentration of acidic groups in eqn (19) in terms of the average concentration cp of the polyelectrolyte. To illustrate this we consider the distances between carboxy groups in a solution of CMC with a degree of substitution of 1 .The average distance (in nm), obtained from the cell model, is d= 1oycN)-w (31) where c is the equivalent concentration (in mol dm-3) and N is Avogadro's constant. Thus in a 0.1 mol dmP3 solution of CMC the average distance between carboxy groups2364 Acid- Base Equilib Y ia in Po ly elec t r oly t es I I I I 0.1 0.2 0.3 (c;/mol kg-')+ Fig. 8. Dependence of the dissociation constant K , of CMCl (degree of substitution 1.07) on the concentration of NaBr.Concentration of the polyacid cp = 0.52 34 equiv. kg-l (0) and cp = 0.016 42 equiv. kg-' (0). I 1 I 0.1 0.2 0.3 (c\/mol kg-')+ Fig. 9. Dependence of the dissociation constant K, of CMC2 (degree of substitution 1.69) on simple salt concentration; 0, NaBr and +, KBr, both at cp = 0.0842 equiv. kg-l; 0, NaBr at cp = 0.037 31 equiv. kg-l.H . Vink 2365 is d = 2.55 nm. On the other hand, the distance between the carboxy groups along the polymer chain is ca. 0.5 nm (the length of an anhydroglucose unit is 0.515 nm),13 which corresponds to a concentration of the order of 10 mol dm-3. Thus CMC solutions are far from homogeneous, and large fluctuations in the concentration of carboxy groups occur. The relative magnitude of the concentration fluctuations increases when the average concentration cp decreases or when the degree of substitution increases.A simple model treatment, where the polyelectrolyte phase is considered to consist of high- concentration regions and voids with zero polyelectrolyte concentration, indicates that concentration fluctuations reduce the determined value of the dissociation constant K,. Although more detailed investigations, both experimental and theoretical, are required to give a quantitative account of the concentration fluctuation effect, it gives a plausible explanation of the behaviour of the various polyacids investigated. In the CMC system the effect is probably quite small for the CMC-gel, where the degree of substitution is low and the concentration cp high, but it may be considerable for the CMC solutions, where it can explain the dependence of Kc on DS, and the decrease of K, with decreasing cp. Similar considerations apply to the PAA and PMA solutions, where the carboxy groups are closely spaced along the polymer chains and concentration fluctuations are high. Note that the concentration fluctuation effect does not influence the definition of the thermodynamic dissociation constant K by eqn (5) and (6), since the chemical potentials and activities are uniform throughout an equilibrium system. References 1 A. Katchalsky and P. S. Spitnik, J. Polym. Sci., 1947, 2, 432. 2 M. Mandel, Eur. Polym. J., 1970, 6, 807. 3 D. K. Hale and D. Reichenberg, Discuss. Faraday Soc., 1949, 7, 79. 4 H. Vink, J. Chem. SOC., Faraday Trans. I , 1985, 81, 1677. 5 H. Noguchi, K. Gekko and S. Makino, Macromolecules, 1973, 6, 438. 6 M. Rinaudo and M. Milas, Macromolecules, 1973, 6, 879. 7 J. A. Marinsky, A. Wolf and K. Bunzl, Talanta, 1980, 27, 461. 8 H. Vink, Acta Chem. Scand., Ser. A , 1979, 33, 547. 9 H. Vink, Acta Chem. Scand., 1963, 17, 2524. 10 H. S. Harned and B. B. Owen, The Physical Chemistry of Electrolytic Solutions (Reinold, New York, 11 K. Gekko and N. Noguchi, Biopolymers, 1975, 14, 2555. 12 M. hnaudo, in Polyelectrolytes, ed. E. SClCgny (D. Reidel, Dordrecht, 1974). 13 K. H. Meyer and L. Misch, Helv. Chim. Acta, 1937, 20, 232. 3rd edn, 1958), p. 508. Paper 511490; Received 30th August, 1985
ISSN:0300-9599
DOI:10.1039/F19868202353
出版商:RSC
年代:1986
数据来源: RSC
|
12. |
Radical cations of organic carbonates, trimethyl borate and methyl nitrate. A radiation–electron spin resonance study |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 82,
Issue 8,
1986,
Page 2367-2376
Nader S. Ganghi,
Preview
|
PDF (571KB)
|
|
摘要:
J . Chem. SOC., Faraday Trans. I , 1986,82, 2367-2376 Radical Cations of Organic Carbonates, Trimethyl Borate and Methyl Nitrate A Radiation-Electron Spin Resonance Study Nader S. Ganghi, D. N. Ramakrishna Rao and Martyn C. R. Symons" Department of Chemistry, The University, Leicester LEI 7RH Exposure of a range of dialkyl carbonates as dilute solutions in trichloro- fluoromethane at 77 K gave rearranged cations in which hydrogen is trans- ferred from carbon to the carbonyl oxygen. Similar species were obtained from the cyclic derivatives, ethylene and propylene carbonates. No evidence has been found for complex formation with solvent molecules such as occurs for methyl formate cations. For the dimethyl derivative the rearrangement appeared to occur even at ca. 4 K. For trimethyl borate a similar reaction occurred, giving either H,cOB(OMe)(O( )+or H,COB(OMe) + MeOH.However, methyl nitrate gave 'NO,H+ radicals, thought to be derived from H,CONO,H+. These results are compared with those for ester and lactone cations and also with those for trimethyl phosphate and dimethyl sulphate cations. H Me ~~ ~~ ~ ~~ When dilute solutions of a range of substrates in certain rigid halogenated matrices are exposed to ionising radiation the substrates are converted into their radical cations, or into unimolecular break-down products there0f.l For small molecules, the matrix of choice is generally trichlorofluoromethane (CFCl,), provided the ionisation potential of the substrate is less than ca. 11.8 eV. This is because solvent radicals trapped in this matrix give remarkably broad lines so that the e.s.r.spectra are generally dominated by features for the substrate radical cations.One disadvantage is that solvent adducts are sometimes generated in which weak a-bonds are formed with one chlorine ligand, such that the SOMO (semi-occupied molecular orbital) becomes the o* orbital, and relatively large hyperfine interactions with 35Cl and 37Cl are manife~ted.l-~ This can be avoided by use of matrices such as C,F,. In many cases, primary cations are stable under these conditions. However, despite early claims that n-cations of esters can be detected at 77 K in CFCl,,,. 5 9 opinion73 has now moved in favour of the suggestion of Iwasaki et alS9 that rearrangement to give carbon-centred radicals as in reaction (1) is facile even for methyl esters.// O'+ //OH+ R-C -+ R-C 'OCH, OCH, The issue was confused by the fact that INDO level calculations for methyl formate cations implied that the n-cations (I) should be more stable than the a-cations (11) favoured in photoelectric spectroscopic studies,lo and that they should display hyperfine coupling to only two of the three methyl protons, the estimated coupling being almost exactly equal to that found for the species now thought to be (111). Our results for MeCO,CHD, seem to be definitively in favour of the rearranged species, even on exposure at ca. 4 K8 Thus for this cation, there was no large proton splitting, implying that the unique hydrogen atom must have transferred to give D,COCMe(OH)+ only. 23672368 E.S.R.of Radical Cations MeO’ 6 ‘H ( 1 ) (11) (111) In related studies of lactone cations, we have combined with Sevilla and his coworkers to show that, except for the four-membered ring lactones, similar rearrangements occur, although they are less facile.ll For the four-membered ring cations, ring-opening reactions appear to be favoured. Thus, despite the stability of the cations of aldehydes and ketones,l none of the simple ester or lactone cations are stable with respect to rearrangements such as that in reaction (1). The only clear exception to this is the solvent complex formed between the cation of methyl formate and CFC1,. However, this is stabilised, and considerably modified by bonding to chlorine, and rearrangement (1) occurs concurrently with breakage of this bond.Other radical cations which exhibit very similar behaviour are those of trimethyl and triethyl phosphate,l2? l3 and dimethyl s~1phate.l~ Both (MeO),PO+ and (MeO),SOi form adducts with CFCl,, and these complexes readily decompose to give H COP( OH)+ (OM e) and H CO SO( OH)+ (OMe) . We are not aware of any similar studies on cations derived from trialkyl borates, dialkyl carbonates or alkyl nitrates. This study was undertaken in the expectation that reactions of type (1) would also be facile for these cations. We also expected that solvent complexes might be formed by these cations, especially MeONOi. Experimental Organic carbonates and trimethyl borate were of the best available reagent grades and were used without further purification. Methyl nitrate was prepared by a standard procedure from methanol and nitric acid and its purity was checked by p.m.r.spectroscopy. Trichlorofluoromethane (freon) was obtained from Fluka and was purified by passing through an alumina column. Dilute solutions of substrates in freon were cooled to 77 K. Samples were irradiated in a 6oCo Vickrad y-ray source to doses of up to 1 Mrad. E.s.r. spectra were measured with a Varian El09 spectrometer, calibrated with a Hewlett- Packard 5246L frequency counter and a Bruker B-H 12E field probe, which were standardized with a sample of diphenylpicrylhydrazyl (DPPH). Samples were annealed above 77 K by decanting the liquid nitrogen from the insert Dewar flask and recooling to 77 K, whenever significant changes were obtained. Results and Discussion All the primary cations are expected to have SOMOs confined to oxygen.For the carbonates, this is probably similar to that for ketone cations [i.e. (IV)] and, by analogy with NO;, radical cations of alkyl nitrates are expected to have the in-plane SOMO indicated in (V). However, for trialkyl borates, we expect the SOMO to be one of theN . S. Ganghi, D. N . R . Rao and M . C. R. Symons 2369 Table 1. E.s.r. parameters for radical cations derived from various carbonates together with those for relevant neutral radicals proton hyperfine coupling/G" radical /OH+ OMe H,kOC, /OH' 'OEt H ~ ~ C H , O C II I 31 16 __ - (77 K) - (130K) - is0 21 20.6 22.0 1 7 t 4 11.0 16 34 13.8 34.6 21 30+36 22 27 + 36 a 1 G = Hussain, J . Chern. SOC. B, 1969, 793. T. g values close to free-spin.A. Hudson and H. A. Ref. (18). Ref. (19). two degenerate n-orbitals (VI) or (VII). These are still effectively non-bonding on oxygen but, in this case, had these cations been detected, we would have expected to observe large proton hyperfine couplings to the a-alkyl protons. Unfortunately, spectra of the type expected for these primary cations were not detected. Furthermore, despite the high ionisation potentials (table 1) and, for the carbonates, the expected high degree of locali~ation,~ no solvent adducts were detected. Instead, rearrangement reactions similar to reaction (1) dominated.2370 E.S.R. oj- Radical Cations - - -- -1 0 + I (a ) Fig. 1. First-derivative X-band e.s.r. spectrum for dimethyl carbonate in CFC1, after exposure to 6oCo y-rays at 77 K and annealing to ca.135 K. The major triplet (a) is assigned to H,COC(OH)+ (OMe) radicals and the minor quartet (b) to methyl radicals. Carbonate Cations Dimethyl Carbonate The spectra obtained at 77 K comprised an anisotropic triplet typical of H,C-X type radicals (table 1). On careful annealing, especially for more concentrated systems (ca. 0.1-1.00/, mole fraction) features for methyl radicals,appeared in the region of 135 K (fig. 1). It is possible that these are formed from Me,CO; radical anions as in Me,CO, + e- -+ Me,COg + Me + MeCO; (2) but this seems to us to be unlikely for the following reasons. (i) There was no evidence for methyl radicals at77 K. Work on pure dimethyl carbonate and its solutions in solvents such as CD,OD shows that under these conditions methyl radicals are formed directly, in competition with radical anion formation at 77 K.15916 (ii) In our experience, electron capture by solvent molecules dominates at these concentrations in CFCI,.If this is correct, then it seems that there may be a pathway for the formation of 'CH, from H,COC(OH)+OMe radicals. We suggest that this is triggered by proton transfer to a base, B, fortuitously close to the cation, as in OH+ // \ H,COC + B + H,CO+CO,+'CH,+BH+. (3) OGH, In these solutions B could be a second dimethyl carbonate molecule. Alternatively, proton transfer to the formaldehyde molecule might occur during the reaction. We have not been able to detect 'CH,CH, radicals from diethyl carbonate underN . S. Ganghi, D. N . R. Rao and M .C. R. Symons 237 1 1 3160G 1 3160G U Fig. 2. First-derivative X-band e.s.r. spectrum for diethyl carbonate in CFCl, after exposure to ‘j0Co y-rays (a) at 77 K and (b) at ca. 150 K, showing the change in conformation giving two equivalent B-protons. similar conditions, but this is not surprising since a reaction comparable to reaction (3) cannot be formulated in this case. Diethyl Carbonate As with ethyl esters, hydrogen-atom transfer occurs from the -CH, group rather than the -OCH2- group, giving (VIII). There must be a fine balance as to which hydrogen moves. Thus structure (VIII) is favoured because there is a shorter distance for hydrogen to move in the ideal conformation, whilst (IX) is expected to be the more stable radical owing to conjugation with oxygen n-electrons and hyperconjugation with the methyl group.It is noteworthy that for isopropyl acetate cations, we obtained Me,COC(OH)Me radicals only.6 There were marked, irreversible changes in the e.s.r. spectra on annealing (fig. 2) and the thermally stable structure, with two equivalent /?-protons, must have structure (X) in which the cationic unit (-R+) occupies the extreme out-of-plane site. Exactly the same /?-proton splittings (1 1 G) were obtained for the ethyl esters, the first-formed species again having two inequivalent P-protons, both with small hyperfine coupling constant^.^? This is curious, since the alternative conformation, with the two B-protons giving maximum2372 E.S.R. of Radical Cations 1 3250 G I * H 20G , t1 ’ t1 0 0 -1 -1 +y2 - Y 2 + Y Z -Y2 4 % -% Fig.3. First-derivative X-band e.s.r. spectrum for ethylene carbonate in CFC1, after exposure to 6oCo y-rays at 77 K and annealing to ca. 110 K, showing features assigned to radical (XII). (Lines a are due to an impurity species.) coupling, is usually found for such l7 We suggest that the transition-state structure resembles that shown in (XI). Initially, this ‘ring’ is buckled, making the 8-protons inequivalent. The structure then relaxes to the symmetrical conformation (X). We cannot be certain that this is actually the thermodynamically preferred structure, but there was no evidence for conversion into any alternative structure on annealing to the melting point of the matrix (ca. 160 K). H Ethylene Carbonate The expected features for the rearranged cation (XII) were not detected at 77 K, but well defined features characteristic of such a radical grew in on annealing to ca.110 K (fig. 3). It may be that (XII) is actually formed at 77 K, but that several conformations are present, resulting in very broad lines. However, in our view, the primary cation, having broad features, is probably present. (Some curious features were sometimes observed at 77 K which as yet remain unidentified. They are almost certainly not due to the primary cations and may have been due to impurity centres). Data for (XII), given in table 1, are close to, but not identical with, those for the unprotonated radical,18 also given in table 1. In this case, the 8-protons have no option but to give large coupling constants for both the protonated and unprotonated species, in contrast with ethyl carbonate cations.N .S. Ganghi, D. N . R. Rao and M. C. R . Symons 2373 1 3180G a Fig. 4. First-derivative X-band e.s.r. spectrum for propylene carbonate in CFC1, after exposure to 6oCo y-rays at 77 K, showing features assigned to radical (XIII). Propylene Carbonate In this case, the rearranged cation (XIII) was clearly The e.s.r. parameters once again closely resemble (table l).19 detected directly at 77 K (fig. 4). those for the neutral analogue One important aspect of the results for these cyclic carbonates is that a transition state such as that depicted in (XI) is not a necessary requirement for rearrangement. However, absence of features due to the cation (XII) from ethylene carbonate at 77 K may be taken as evidence for the relative inefficiency of this change, which involves far greater movement of the hydrogen atom.Unfortunately, our inability to recognise clear features for the parent cation at 77 K means that we cannot draw this conclusion definitively. Vinylene Carbonate Finally, we also made a brief study of the radical cation of this related carbonate, in the expectation that the 7t-cation would be strongly stabilised by conjugation with the double bond. Such a cation is expected to exhibit a 1 : 2: 1 triplet arising from this delocalisation. In fact, a triplet with Aiso(lH) = 12.5 G was obtained, but we are puzzled by the fact that the central line appears to be broadened, the appearance being more like 1 : 1 : 1 triplet (fig. 5). However, on annealing, a better defined 1 : 2: 1 spectrum appeared reversibly and we consider that the 7t-SOMO is correct for this species.2374 E.S.R.of Radical Cations 1 3239 G I " c_ I -1 0 * 1 Fig. 5. First-derivative X-band e.s.r. spectrum for vinylene carbonate in CFCI, after exposure to V o y-rays at 77 K, showing features assigned to the parent TZ radical cation. Trimethyl Borate Once again, the only species clearly detected is the rearranged cation giving a I : 2: 1 triplet characteristic of an H,CO- group (similar to that in fig. 1). In this case, since there is no free oxygen to accept the proton and since the solvent is an extremely poor proton acceptor, we suggest either the cations (XIV) or (XV) as the most probable species responsible for this spectrum : ,OMe '0-H' H,CO - B - H,COBOMe + MeOH I Me ( X I V ) ( X V I The second-derivative spectrum for this species showed a poorly defined quartet splitting on the 10) line with a splitting of ca.4.5 G. This is tentatively assigned to hyperfine coupling from IIB. [IlB has I = $ and is 81.17% abundant; features for loB ( I = 3,18.83% abundant) would not be detectable under these conditions.] Unfortunately, this does not help us to distinguish between structures (XIV) and (XV), both of which are expected to give rise to hyperfine coupling to boron. Methyl Nitrate In this case, we had hoped that the primary cation would have been less ready to undergo proton migration, since the SOMO is expected to be delocalised equally on the two oxide ligands. This should have the effect of reducing the tendency for hydrogen transfer, both because of a stabilising effect on the parent cation, and because of the need to modify the electronic structure as the hydrogen atom moves.Indeed, our work on diester cationsN . S. Ganghi, D. N . R. Rao and M . C. R. Symons 2375 I I I X Y 2 + 1 1 32206 n 206 , I + I I I Fig. 6. First-derivative X-band e.s.r 6oCo y-rays at 77 K, showing I t Y (2 + X I -1 spectrum for methyl nitrate in CFC1, after exposure to features assigned to 'NO, or 'NO(OH)+ radicals. - strongly supports this contention.20 Nevertheless, the expected features, which should resemble those for 'NO, radicals, were not observed. Furthermore, features for the rearranged cation, H,CONO(OH)+, were also not detected. Instead, features closely resembling those characteristic of 'NO, radicals appeared directly at 77 K (fig.6). We suggest that these radicals are formed from H,CONO(OH)+ according to reaction ( 5 ) possibly followed by reaction (6): H,CONO(OH)+ --+ H,CO + 'NO(OH)+ ( 5 ) H,CO+'NO(OH)+ + H,COH+ +'NO,. ( 6 ) The data given in table 2 are compared with results for 'NO, in the gas phase and with recent results for radicals resembling 'NO, obtained from irradiated nitroalkanes in CFC1,.21 It is noteworthy that although the general form of the I4N hyperfine coupling and g-tensor components are quite close to those for 'NO,, there are small differences, especially in that A, and A, are significantly greater than expected. These increases cannot simply be caused by slight librational averaging since A, is not concomitantly reduced.In fact, a similar trend was observed for the species obtained from nitroalkanes.21 Two2376 E.S.R. of Radical Cations Table 2. E.s.r. data for ‘NO, type radicals obtained from methyl nitrate and nitromethane g-values ~ 14N hyperfine coupling constants/Ga radical A , A , A, Aiso g, g, g.2 NOb 46.14 44.8 66.76 52.57 2.0062 1.9910 2.0020 53 48 66 55.7 2.0045 1.9935 2.0020 /O c *N ‘OkH3 52 49 67 56 2.0045 1.992 2.0020 * N /’ d \OH+ a 1 G = lop4 T. Gas-phase. From MeNO, [ref. (21)]. From MeONO, (ionisation potential MeONO, = 11.53 eV). radicals were postulated, that formed on annealing, whose data arc included in table 2, being assigned structure (XVI). It therefore seems most probable that ‘NO, in the present study remains protonated, so that reaction (6) is redundant.Conclusions We conclude that, in all cases, hydrogen-atom transfer occurs with great facility once the parent cations are formed. In the case of methyl nitrate, the resulting radical dissociates further giving ‘N02Hf and for dimethyl carbonate evidence for loss of methyl radicals was obtained. Clearly, the behaviour of these cations is very similar to that of ester5-lo and lactone cationsll for which H-atom transfer is also facile. This is also true of trimethyl phosphateJ2* l3 and dimethyl sulphate cations,14 except that in these cases solvent adducts were precursors to rearrangement. References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 M. C. R. Symons, Chem. SOC. Rev., 1984, 12, 393. D. Becker, K.Plante and M. D. Sevilla, J. Phys. Chem., 1983, 87, 1648. G. W. Eastland, D. N. R. Rao, J. Rideout, M. C. R. Symons and A. Hasegawa, J. Chem. Res. (9,1983, 258. T. Clark, A. Hasegawa and M. C. R. Symons, Chem. Phys. Lett., 1985, 116, 79. M. D. Sevilla, D. Becker, C. L. Sevilla and S. Swarts, J . Phys. Chem., 1984, 88, 1701. D. N. R. Rao, J. Rideout and M. C. R. Symons, J . Chem. SOC., Perkin Trans. 2, 1984, 1221. M. D. Sevilla, D. Becker, C. L. Sevilla and S. Swartz, J . Phys. Chem., 1985, 89, 633. J. Rideout and M. C. R. Symons, J . Chem. Sac., Perkin Trans. 2, submitted. M. Iwasaki, H. Muto, K. Toriyama and K. Nunome, Chem. Phys. Lett., 1984, 105, 586. D. A. Sweigart and D. W. Turner, J. Am. Chem. SOC., 1972, 94, 5592. J. kdeout, M. C. R. Symons, S. Swarts, B. Besler and M. D. Sevilla, J . Phys. Chem., 1985, 89, 5251. X. Z. Qin, B. W. Walther and F. Williams, J. Chem. SOC., Chem. Commun., 1984, 1667. G. D. G. McConnachie and M. C. R. Symons, J . Chem. Res. (S), 1985, 54. R. Janes and M. C. R. Symons, J . Chem. Res. (9, 1986, 108. R. L. Hudson and F. Williams, J . Phys. Chem., 1981,85, 510. N. Ganghi and M. C. R. Symons, unpublished results. See for example, D. J. Edge and J. K. Kochi, J . Am. Chem. SOC., 1972, 94, 7695; R. Livingston and H. Zeldes, J . Chem. Phys., 1966, 44, 1245. H. Zeldes and R. Livingston, J . Mugn. Reson., 1976, 21, 109. E. A. Shade, Can. J. Chem., 1973, 51, 2492. J. Rideout and M. C. R. Symons, unpublished results. D. N. R. Rao and M. C. R. Symons, J . Chem. SOC., Furuday Trans. 1, 1985, 81, 565. Paper 5 / 149 I : Receitled 30th August, 1985
ISSN:0300-9599
DOI:10.1039/F19868202367
出版商:RSC
年代:1986
数据来源: RSC
|
13. |
The hydrophobic behaviour of Orange IV in water and in aqueous electrolyte solutions |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 82,
Issue 8,
1986,
Page 2377-2383
Michel De Vijlder,
Preview
|
PDF (363KB)
|
|
摘要:
J . Chem. SOC., Faraday Trans. I , 1986, 82, 2377-2383 The Hydrophobic Behaviour of Orange IV in Water and in Aqueous Electrolyte Solutions Michel De Vijlder Rijksuniversiteit Gent, Laboratorium voor Fysische Scheikunde, Krijgslaan 281, B-9000 Gent, Belgium Unlike the stepwise manner of self-aggregation shown by homologues of the surface-active azo dye Methyl Orange in aqueous solution, cooperative self-aggregation is shown in electrolyte solutions of Orange IV ( C , H , - N H - - ~ N = = N ~ SO; Na+). Evidence of a critical micellar concentration (c.m.c.) appears from the breaks in the plots of absorbance and surface tension values against the dye concentration. The quantitative treatment of the effects of NaCl and SrC1, reveals that the equation log (c.m.c.) = - Kcsalt + constant, derived for zwitterionics by Mukerjee (J.Phys. Chem., 1965, 69, 4038) on the basis of the salting-out of their hydrocarbon groups, is only obeyed at lower salt concentrations. An equation log (c.m.c.) = - K log Csalt +constant fits our experimental results over the whole range of salt concentrations considered and remains valid when 1.0 mol dm-3 urea, a well known disturber of hydrophobic bonds, is introduced. It is known that amphiphilic azo dyes [especially Methyl Orange (I) and its homologues] do not behave in aqueous solution like other ionic surfactants; e.g. their self-aggregation may proceed in a stepwise manner without cooperativity and without evidence of a critical micellar concentration (c.m.c.). However, the addition of foreign ions to such dye solutions, favouring the self- aggregation processes,2 may affect their mechanisms or the nature of the aggregated species.Since we observed recently2 that the amounts of electrolyte required for this purpose are small in the case of Orange IV (R, = C,H,, R2 = H), thus permitting surface-tension measurements, we focused our attention on the behaviour of this dye, referring to analogies with Methyl Orange where possible. Experiment a1 Methyl orange was purchased from B.D.H. and Orange IV from Fluka. Both were purified by a twofold recrystallization from water. Further purification through the foaming method recommended by Giles3 did not introduce any changes in the results of the surface-tension measurements in low concentrated dye solutions and was abandoned because of hastening precipitation in more concentrated ones.We used a Varian Techtron spectrophotometer for the light absorption measurements and a Philips PW 9504/00 bridge for the conductance measurements (the cell was equipped with blacked platinum electrodes). The surface tension measurements occurred with the du Nouy ring method: the force 79 2377 F A R 12378 Hydrophobic Behauiour of Orange IV / 100 /'/ I // 5 10 15 C/ 1 0-4 rnol dm-3 e I P z f ._ + j o y 1- :- n 60 50 10-3 I I I I I I I I I I I I 1 Clrnol dm-3 Fig. 1. Plots of the absorbance ( A ) , the conductance ( K ) and the surface tension (7) of aqueouc solutions of Methyl Orange (0) and Orange IV (+) as a function of concentration, C, at 25 "C. Absorbances are measured at A, = 465 nm for Methyl Orange and 443 nm for Orange IV (water as reference, cell path length 0.2 cm).exerted on the immersed ring by gently lowering the sample holder was recorded automatically by a R-G Cahn electrobalance. Calibration tests undertaken on pure solvents reveal an accuracy in the region of ca. 0.25 mN m-I. Results Spectrophotometric measurements in solutions where no salt is added apparently reveal a deviation from ideal behaviour at the same concentration for both dyes (5.5 x mol dm-3). Conductance measurements, however, show a rather marked difference: 6.0 x mol dm-3 for Orange IV. As will be seen further, extrapolating c.m.c. values of these dyes in electrolyte solutions to an electrolyte concentration equal to zero will confirm the values obtained by conductance measurements.Such a discrepancy associated with the method of observation is not unusual, and has been observed by other workers on Pentyl and Hexyl Orange as well.4 The higher critical concentration of Methyl Orange could be due to a more important hydrati~n.~ We note in passing that the value for Orange IV is fairly higher than those of homologues carrying aliphatic groups on th.= amino nitrogen: 0.5 x for Butyl Orange,6 0.35 x for Pentyl Orange4 and 0.3 x lop4 for Hexyl O ~ a n g e . ~ Apparently the aromatic group confers on Or IV a smaller tendency to aggregate than aliphatics do. As expected, surface-tension measurements fail to show any indication of c.m.c. neither at the above concentrations nor at higher ones. So, if it is true that a c.m.c. may rnol dm-3 for Methyl Orange and 5.0 xM .De Vijlder 2379 A * I E z E * Q .. Fig. 2. Plots of the absorbance ( A ) and the surface tension (7) of aqueous solutions of Orange IV in the presence of increasing amounts of NaCl. Dye concentrations are 3 x lop4 (+), 4.5 x (0) and 7.5 x ([I]) mol drnb3. be determined from the break in the Beer’s law curves of solutes absorbing in the visible or U.V. region^,^ each such deviation does not necessarily give evidence of the existence of a c.m.c. Repeatable results are all plotted in fig. 1.t From a critical amount dependent on the dye concentration, added electrolytes cause a decrease in the absorbance of both dyes. Critical electrolyte concentrations have already been reported for several azo dyes.2 On comparing the course of the decrease in absorbance and in the surface tension of Orange IV as a function of electrolyte concentration, we found that the absorbance values start to decrease precisely at those electrolyte concentrations for which surface-tension values remain unchanged, thus giving evidence of a c.m.c.(fig. 2). This should probably be valid for Methyl Orange as t Note that the degree of absorbance decrease as reported in the figures is not absolute; it depends as well on the surface-volume ratio of the recipient in which the dye solution is kept. The larger the ratio the less dramatic the decrease. The values reported in this paper are those of solutions in equilibrium in conventional 50 x dm3 flasks. 79-22380 Hydrophobic Behaviour of Orange IV A 2 3 4 5 I 20 1 5 1 I 4 / 2 3 C(0range IV) Fig.3. Graphic determination of the breaks in the Beer's curves of Orange IV, associated with a c.m.c. of the dye in the presence of NaCl (left, in mol dm-3) and SrCl, (right, in mol dmp3). well, but attempts at determining surface-tension values remain unsuccessful : the results drifted because of the much higher salt concentrations required. The change in c.m.c. of Orange IV as a function of added NaCl and SrCl, has been determined further only from breaks in the absorbance curves. The results are plotted in fig. 3. In the case of zwitterionics (which Methyl Orange and its homologues indeed are) and non-ionic surfactants it is expected that the logarithm of the c.m.c. vary linearly with the salt concentration.8 Fig. 4 and 5 reveal that such a relationship is only applicable here at lower electrolyte concentrations, and that a better fit is obtained by plotting log c.m.c.us. log Csalt. The same relationship has been reported in the case of Methyl Orange.2 The former equation is derived because of the effect of salts on the monomer-micelle equilibrium interpreted in terms of the salting-out of the hydrocarbon groups. Some deviations from this theory have been discussed,sc but as far as we know, a log-log relationship has not yet been submitted. It may be tempting to associate our results with the equation which is usually applicable to ionic surfactants : log c.m.c. = - K log Ci + constant where Ci is now the total counterion concentration in mole dmp3 (= Cc.m.c. + Csalt for univalent ions).This application may hold in the case of added NaCl, where Csalt % CCarnaC., but is inappropriate in the case of SrCl,. As seen above, extrapolating the c.m.c. values in fig. 4 to zero salt concentration permits us to recover the value of the critical aggregation concentration (c.a.c.) obtained from conductivity measurements (log c.a.c. = - 3.30, c.a.c. = 5.0 xM . De Vijlder 238 1 C(NaCl)/mol dm-3 0.0 5 01 0 0.1 5 Q20 I I I - 3 2 5 3 1 2 3 4 5 6 I I I 1 C (SrC12)/ 1 0-4 mol dm-’ Fig. 4. Plots of the log c.m.c. from fig. 3 us. the electrolyte concentrations. Table 1. Effect of urea on the critical aggregation concentration of Orange IV (all concentrations in mol dm-3) without urea 1 mol dm-3 urea 4 mol dmP3 urea in H,O 5.0 x 10-4 5.0 x 10-4 5.0 x 10-4 0.12 mol dmP3 NaCl 1.0 x 10-4 2.1 x 10-4 > 4.2 x 10-4 0.20 mol dm-3 NaCl 0.9 x 10-4 1.7 x 10-4 > 4.2 x 10-4 0.06 mol dm-3 NaCl 2.0 x 10-4 3.0 x lop4 not determined 0.8 x lop4 mol dm-3 SrCl, 1.3 x 10-4 2.8 x not determined 1.6 x mol dm-3 SrC1, 0.9 x 10-4 2.3 x lop4 not determined2382 Hydrophobic Behaviour of Orange IV A - 3.5 - 3.75 - L.0 - 1.5 -1 0 log [C(NaCl)] 2.0 1.5 1.0 0.5 Fig.5. Plots of log c.m.c. from fig. 3 vs. the log of the electrolyte concentrations. - 2 3 4 5 C(0range IV)/ 1 0-4 mol dmd3 Fig. 6. Graphic determination of the breaks in Beer’s curves of Orange IV, in aqueous 1 mol d ~ n - ~ urea solutions in the presence of NaCl (left) and SrCl, (right).M . De Vijlder 2383 C(SrC12)/10-4 mol d ~ n - ~ (+) 1 2 3 4 1 2 3 4 C(NaCI)/ lo-' mol dm-3 (0) 1 I I -4.5 -4.0 -3.5 -1.2 -1D -0.8 -0.6 1 1 Fig.7. Plots of the log c.m.c. from fig. 6 us. the electrolyte concentrations (left) and the log of the electrolyte concentrations (right) (both in mol dmP3). It also seemed of interest to check the influence of urea on the behaviour of the dye solutions. It is well known that urea lowers the tendency to aggregate through hydrophobic interaction^.^ The results of these experiments confirm that the aggregation of Orange IV in water and in electrolyte solutions are two different processes : apparently, urea has no effect on the value of the critical aggregation concentration in the former case, but lowers markedly the tendency to aggregate in the latter (table 1). The presence of 1 mol dmA3 urea does not affect the validity of the log-log relationship we found for the effect of the electrolyte concentration on the assumed c.m.c. of Orange IV (fig. 6 and 7). References 1 R. Reeves and Sh. Harkaway, J . Colloid Interface Sci., 1978, 64, 342. 2 M. De Vijlder, J . Chem. SOC., Faraday Trans. I , 1985, 81, 1369. 3 C. H. Giles and A. H. Soutar, J . SOC. Dyers Colour., 1971, 87, 301. 4 T. Takagishi, S. Fujii and N. Kuroki, J. Colloid Interface Sci., 1983, 94, 114. 5 R. L. Reeves, R. S. Kaiser, M. S. Maggio, E. A. Sylvester and W. H. Lawton, Can. J. Chem., 1973,51, 6 R. Burkhard, B. E. Buergert and J. S. Levitt, J . Am. Chem. SOC., 1953, 75, 2977. 7 D. G. Duff and C. H. Giles, J . Colloid Interface Sci., 1972, 41, 407. 8 (a) K. Shinoda, T. Yamaguchi and R. Hori, Bull. Chem. Soc. Jpn, 1961, 34, 237; (b) P. Mukerjee, 9 M. Schick, J . Phys. Chem., 1964, 68, 3585. 628. J. Phys. Chem., 1965,69,4038; (c) A. Ray and G. NCmethy, J . Am. Chem. SOC., 1971,93, 6787. Paper 5 / 1548; Receiued 9th September, 1985
ISSN:0300-9599
DOI:10.1039/F19868202377
出版商:RSC
年代:1986
数据来源: RSC
|
14. |
Polyaniline, a novel conducting polymer. Morphology and chemistry of its oxidation and reduction in aqueous electrolytes |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 82,
Issue 8,
1986,
Page 2385-2400
Wu-Song Huang,
Preview
|
PDF (1526KB)
|
|
摘要:
J . Chem. SOC., Faraday Trans. 1, 1986,82, 2385-2400 Polyaniline, a Novel Conducting Polymer Morphology and Chemistry of its Oxidation and Reduction in Aqueous Electrolytes Wu-Song Huang, Brian D. Humphrey and Alan G. MacDiarmid" Department of Chemistry, University of Pennsylvania, Philadelphia, Pennsylvania 19 104, U . S. A . The emeraldine salt forrn of polyaniline, conducting in the metallic regime, can be synthesized electrochemically as a film exhibiting a well defined fibrillar morphology closely resembling that of polyacetylene. Cyclic volt- ammograms of chemically synthesized and electrochemically synthesized polyaniline are essentially identical. Probable chemical changes which occur and the compounds which are formed when chemically synthesized poly- aniline is electrochemically oxidized and reduced between - 0.2 and 1 .O V us.SCE in aqueous HCl solutions at pH values ranging from -2.12 (6.0 mol dm-3) to 4.0 have been deduced from cyclic voltametric studies. These are shown to be consistent with previous chemical and conductivity studies of emeraldine base and emeraldine salt forms of polyaniline. It is proposed that the emeraldine salt form of polyaniline has a symmetrical conjugated structure having extensive charge delocalization resulting from a new type of doping of an organic polymer-salt formation rather than oxidation which occurs in the p-doping of all other conducting polymer systems. The emeraldine salt form of polyaniline, synthesized by the chemical or electrochemical oxidative polymerization of aniline in aqueous acid sol~tionl-~ is currently arousing considerable interest since it exhibits conductivity in the metallic regime (ca.5 W1 cm-l).l-14 [For a full description of the nomenclature used in describing the various forms of polyaniline see ref. (14).] Chemical synthesis results in a powder which can be compressed into pellets. Electrochemical synthesis gives cohesive films which have been reported to show a fairly smooth featureless topography by scanning electron micr0scopy.~9 The emeraldine salt is believed to have the composition + + {[-(C6H,)-N(H)-(C6H4)-N(H)* (C6H,)-N(H)=(C6H,)=N(H)~}~ A- A- where A- is an anion. It may also be synthesized by 'doping' the emeraldine base forrn of polyaniline [-(C6H4)-N(H>-(C6H4)-N(H) (C6H4)-N=(C6H4)=N-1 1, which consists of equal numbers of reduced [-(C6H4)-N(H)-(C6H4)-N(H)-], (IA), and oxidized [-(c6H4)-N=(c6H,)=N-], (2A), repeat units with aqueous 1.0 mol dm-3 HCl.149 l5 This results in an increase in conductivity of ca.lolo. The aqueous electrochemistry of polyaniline films polymerized on electrodes has been studied previously.6*9-13 A variety of effects have been observed depending on the conditions used in preparing the films. Cyclic voltammograms of films prepared by potential cycling between -0.2 and 0.8 V us. SCE or by potentiostatic oxidative polymerization at 0.9 V us. SCE exhibit redox processes associated with degradation products of the polymer.6 On the other hand cyclic voltammograms of films prepared 23852386 Electroactivity of Polyaniline at lower oxidation potentials, e.g.(0.65 V us. SCE) only exhibit degradation products when scanned to higher potentials ( > 0.7 V us. SCE).6, l2 The oxidation potential of the first redox process (I$ = 0.13 V us. SCE in 1 .O mol dm-3 HCl) does not change in the presence of different cations and varies little in the presence of different anions, but does change with the pH of the electr01yte.l~ In the range pH 1 to -2 the first redox peak shifts 59 mV per pH unit to higher potential as the pH is decreased; however, the first redox peak does not shift between pH 1 and pH 4. No study of the effect of pH on the second redox process occurring at a higher potential has been reported.6* 12* l3 The electroactivity of the film ceases at pH >4.13 No detailed interpretation of these observations in terms of the chemical transformations which occur during oxidation and reduction has been given.The present investigation was carried out for the purpose of (i) elucidating the chemical reactions which occur and the materials which are formed during the electrochemical oxidation and reduction of polyaniline in aqueous solutions of varying pH values and (ii) relating these observations to a proposed structure for the highly conducting emeraldine salt form of polyaniline. Experimental Reagents Reagent-grade C,H,NH,, NH,OH and concentrated HCl were purchased from MCB Manufacturing Chemists, Inc. A.C.S.-certified NaCl and (NH,),S,O, were purchased from Fisher Scientific Co. Perchloric acid (70 % ) was purchased from Baker Chemical Co. and HBF, (48%) was purchased from Aldrich Chemical Co.The C,H,NH, was further purified by distillation. All other chemicals were used as purchased. Chemical Synthesis of Polyaniline The method149 l5 employed for synthesizing the polyaniline used in this study was based, in part, on previously described procedure^.^? An aqueous solution of (NH,),S,O, was added slowly to a solution of aniline dissolved in 1 .O mol dmP3 aqueous HC1 at ca. 5 "C. After 1 h the precipitate which had formed was removed by filtration, washed repeatedly with 1 .O mol dm-3 HCl and dried under dynamic vacuum for ca. 48 h. The material thus obtained was identified as emeraldine hydrochloride as described below. The emeraldine hydrochloride was converted into the emeraldine base by stirring with a ca.0.1 mol dm-3 solution of NH,OH for several hours. The material was dried under dynamic vacuum for ca. 48 h. If it was considered necessary to remove the lower-molecular-weight species (ca. 20% by weight) the emeraldine base was extracted with CH,CN until the extract was colourless. Complete elemental analyses1, of both the emeraldine base and its hydrochloride salt were consistent with the compositions proposed above.15 The sum of the individual percentage compositions by weight of various samples fell in the range 99.57-100.88 % , 1 4 9 l5 Elemental analyses are not a reliable criteria for unequivocally establishing the presence or absence of hydrogen atoms in the emeraldine base or in the emeraldine salt. However, the infrared spectra of both types of materials showed the presence of the expected types of N-H stretching ~ibrati0ns.l~ More important still, the amount of charge required to reduce electrochemically an analytically pure sample of chemically synthesized emeraldine base to the colourless leucoemeraldine base, [-(C,H,)-N(H)-(C,H,)-N(H)-],,, in aqueous acid solution (ca.pH 4) was consis- tent with the proposed composition of the emeraldine base.17W-S. Huang, B. D. Humphrey and A . G. MacDiarmid 2387 Electrochemical Synthesis of Polyaniline Polyaniline was synthesized electrochemically in the form of a high-quality smooth cohesive film on a platinum-foil electrode (ca. 1 cm2) by cycling (ca. 45 times) the platinum electrode between -0.20 and +0.75 V vs. SCE at a rate of 50 mV s-l in a solution consisting of 1 cm3 of aniline and 20 cm3 of ca.1 rnol dm-3 HCl. This procedure produced a ca. 0.2 pm thick film and took ca. 30 min. The polymerization was completed at a potential of ca. 0.4 V, characteristic of the emeraldine oxidation state of polyaniline since in this form the polyaniline is not oxidized by air. The film was then washed in 1.0 mol dm-3 HCl for ca. 2 min to free it from traces of aniline. Cyclic Voltammetry Studies Cyclic voltammetry studies of chemically synthesized polyaniline were performed by one or the other of two methods. (1) The polymer was ground to a fine powder and ca. 1 mg was impregnated into glass filter paper (ca. 1 cm2) either by a spatula or by a finger (rubber glove required). The filter paper was then shaken to remove an excess of polyaniline powder and was then placed on platinum mesh which was folded so as to encase the filter paper on both sides.(2) Finely ground polyaniline powder was suspended in acetone or chloroform and a few drops of the suspension were poured on to a platinum foil (ca. 1 cm2) electrode and allowed to dry in the air. Both methods yielded identical cyclic voltammograms. Cyclic voltammetric studies of chemically prepared polyaniline requires a precondi- tioning in order to obtain reproducible cyclic voltammograms. This may involve complete permeation of the electrolyte into the polymer powder particles. All results reported below were obtained with polymer powder/Pt electrodes after they had been cycled 10-20 times in 1 .O mol dm-3 HCl between - 0.20 and 0.40 V us.SCE. During this preconditioning the intensity of the anodic peak at 0.21 V increased and then became constant. For HCl concentrations < 1 mol dm-3, NaCl solution was added to maintain a C1- concentration of ca. 1.0 mol dm-3. All cyclic voltammetry studies were performed using a PAR model 173 potentiostat in conjunction with a PAR model 175 universal programmer. A standard three-electrode configuration involving a saturated calomel electrode (SCE) was used. The pH of the electrolyte was varied by changing the HC1 concentrations. Morphology Studies A smooth green film of emeraldine hydrofluoroborate was electrochemically deposited for ca. 16 h on a platinum anode at ca. 0.7 V vs. SCE in an electrolyte consisting of 5 cm3 of 48% HBF, and 1 cm3 of aniline in 10 cm3 water as described previously.8 Another film was electrochemically deposited on indium oxide conducting glass (Practical Products Co.) at a constant applied current of 50 pA cm-2 for 90 min in an electrolyte consisting of 4 cm3 of HClO, (70%) and 2 cm3 of aniline in 20 cm3 of water.The films were coated with ca. 100 A of gold and were examined with a Philips PSEM 500 scanning electron microscope. Results Cyclic voltammograms of chemically and electrochemically synthesized polyaniline were recorded in aqueous HC1 solutions of different concentrations at a sweep rate of 50 mV s-l. Both the emeraldine base and its hydrochloride salt are insoluble in all the aqueous electrolytes employed in the study. The ‘pH values’ given for solutions more acidic than pH 1.0 are actually Hammett acidity functions, which reflect the proton2388 Electroactivity of Polyaniline mA J.V I 1 I I I I I -0.2 0.0 0.2 0 . 4 0.6 0.8 1.0 EIV us. SCE Fig. 1. (a) Cyclic voltammogram (50 mV s-l) of chemically synthesized emeraldine hydrochloride in 1 .O mol dm-3 aqueous HCl. (b) Cyclic voltammogram (50 mV s-l) of electrochemically synthe- sized emeraldine hydrochloride in 1 .O mol dmP3 aqueous HCl. donating ability of highly acidic solutions more accurately than pH.l8, l9 In dilute solutions the Hammett function and pH are identical. A cyclic voltammogram (Pt mesh method) of chemically synthesized emeraldine hydrochloride in a 1 .O mol dm-3 aqueous HC1 electrolyte (pH - 0.2) is given in fig. 1 (a). This, as expected, is identical to a cyclic voltammogram of chemically synthesized emeraldine base, since the base is converted into the hydrochloride salt when placed in the 1.0 mol dm-3 HCl.*? l4 The cyclic voltammogram is essentially identical to the cyclic voltammogram of electrochemically synthesized polyaniline given in fig.1 (b), except that it has a small peak at +0.85 V us. SCE which is absent in the cyclic voltammogram of the chemically synthesized polymer. The colour changes observed during the oxidation and reduction process were identical for both the chemically and electrochemically synthesized materials. Cyclic voltammograms between -0.20 and 1.0 V us. SCE of the above chemically synthesized polyaniline using the same polyaniline electrode were recorded in a number ofdifferent electrolytes having pH values in the range - 0.2 to 4.0.Typical voltammograms are given in fig. 2. The potential of the first anodic peak (and the corresponding cathodic peak) is almost independent of the pH of the electrolyte at pH 1 and 2 and changes only slightly at a pH - 0.2 (1 .O mol dm-3 HC1). The potentials of the second anodic peak (and the corresponding cathodic peak) are strongly dependent on pH in the range of -0.2W-S. Huang, B. D. Humphrey and A . G . MacDiarmid 2389 I 1 I 1 I I 1 I I -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 EIV vs. SCE Fig. 2. Cyclic voltammograms of chemically synthesized emeraldine hydrochloride in electrolytes of (a) pH 2.0, (b) pH 1.0 and (c) pH -0.2 (1.0 mol dm-3 HCl). 0.8 L 0.0 1 .o 2.0 3.0 4.0 PH Fig. 3. Relationship between E; of the second redox process between pH -0.20 (1.0 mol dm-3 HC1) and pH 4.0.Slope = - 120 mV per pH unit. to 4.0. The relationship between Ej of this second redox process and pH in this range is given in fig. 3. On cycling between -0.2 and 1.0 V us. SCE in the pH range -0.2 to 4.0 degradation is observable after a few cycles, especially in the more acidic electrolytes. Note that no significant degradation is observable after 5 x lo3 cycles between - 0.2 and 0.6 V at ca. pH -0.2.6 Cyclic voltammograms of chemically synthesized polyaniline (Pt mesh method) were also recorded at pH values of -2.12 (6.0 mol dm-3 HCl), - 1.05 (3.0 mol dm-3 HC1) and -0.20 (1.0 mol dm-3 HCl) between -0.20 and 0.50 V us. SCE. They are given in2390 Electroactivity of Polyaniline 1 -0.2 0.0 0.2 0.L 0.6 EIV vs. SCE Fig. 4. Cyclic voltammograms (50 mV s-l) of the first redox process of chemically synthesized emeraldine hydrochloride in electrolytes of (a) pH -2.12 (6.0 mol dm-3 HCl), (b) pH - 1.05 (3.0 mol dmP3 HCl) and (c) pH - 0.2 (1 .O mol dm-3 HCl). 0 0 -2.0 -1.0 0.0 1.0 2.0 PH Fig. 5. Relationship between Ei for the first redox process between pH -2.12 (6.0 mol dm-3 HCl) and pH 2.0. Slope = -58 mV per pH unit.W-S. Huang, B. D. Humphrey and A . G. MacDiarmid 239 1 Fig. 6. Proposed resonance forms for the emeraldine salt form of polyaniline consisting of equal numbers of reduced, f(C,H,)-N(H)-(C,H,)-N(H)+, and oxidized and protonated, f(C,H,)-N=(C,H,)=N+, repeat units. H H + + fig. 4. Continuous cycling in 1 .O and 6.0 mol dmP3 HCl electrolytes in this voltage range gives no observable degradation.The relationship between the potential and electrolyte pH between pH -2.12 and 2.0 is given in fig. 5. However, oxidation to potentials > ca. 0.7 V in these electrolytes causes significant irreversible degradation of the polymer, especially when the acid strength is > ca. 1 mol dmP3. Degradation was so rapid that it was not possible to determine accurately the relationship between potential and pH in solutions much more concentrated than ca. 1 mol dm-3 (pH -0.2). Discussion The extremely large increase of electronic conductivity brought about by treating the emeraldine base form of polyaniline with aqueous acid involves a new type of doping of a conducting polymer. It occurs by proton addition to the polymer rather than by partial oxidation of the polymer 7~ system, as is the case in the p-doping of other conducting The formation of a nitrogen base salt rather than a potentially highly reactive carbonium ion is believed to be responsible for the high chemical stability of the material in the environment.Unlike all other conducting polymers, the conductivity of polyaniline depends on two variables instead of one, viz. the degree of oxidation of2392 Electroactiuity of Polyaniline the polyaniline and the degree of protonation of the material. The proton addition results in partial depopulation of the n system. It is proposed that the emeraldine salt form of polyaniline shows high conductivity because of extensive 7c conjugation in the polymer chain as shown by the four identical resonance forms in fig.6.14 If this should be the case then all nitrogen atoms, all C-N bonds and all C,H, rings would be identical. Each nitrogen atom would bear a +0.5 charge, all nitrogen atoms would bc intermediate between a single and double bond and all C,H, rings would be intermediate between benzenoid and quinoid, viz. The resulting highly conjugated n system, in addition to contributing to the high conductivity, would also be expected to impart extra stability to this form of polyaniline. It would also be expected to increase the strength of the emeraldine base to a value greater than that found in amines containing phenyl-nitrogen bonds. Such amines are known to be very weak bases [pK, of (C,H,),NHl = 1.0].21 Since amine nitrogen atoms are stronger bases than imine nitrogen atoms it might be expected that treatment of emeraldine base with HCl would lead to preferential protonation of the amine nitrogen atoms to give a material such as + + {€(C6H,)-N(H)2-(C,H4)-N(H)2 (c6H,>-N=(c6H,)=N~),.C1- C1- It is proposed here that when there are equal numbers of reduced, f (C,H,)-N(H)-(C,H,)-N( H) t, and oxidized, +( C,H,)-N=( C,H 4)=N +, repeat units in the polymer that protonation of the imine nitrogen is preferred, resulting in the formation of the completely symmetrical, resonance stabilized polymer having one hydrogen atom and an identical positive charge on each nitrogen atom. This is consistent with the well known properties of guanidine,, H N II which, except for quaternary ammonium hydroxides, is the strongest organic base known (pK, of [C(NH,),]+ = 13.6). The high basicity is believed to result from the large resonance stabilization (resulting from preferential protonation of the imine nitrogen atom) of the quanidinium ion relative to free guanidine.A large resonance stabilization is to be expected because the contributing resonance structures t C C C H,N/ ‘NH, H,N/ ‘NH, H,N/ N N h 2 + are identical.J . Chem. SOC., Furuduy Trans. I , Vol. 82, part 8 Plute 1 Plate 1. Scanning electron micrograph of a smooth emeraldine hydrofluoroborate film grown electrochemically on a platinum anode at a constant potential of ca. 0.7 V us. SCE for 16 h. W-S. Huang, B. D. Humphrey and A. G. MacDiarmid (Fucing p . 2392)J. Chem. SOC., Faraday Trans. 1, Vol.82, part 8 Plate 2 Plate 2. Scanning electron micrograph of an emeraldine hydroperchlorate film grown electro- chemically on indium oxide conducting glass at a constant current of 50 pA cmP2 for 90 min. W-S. Huang, B. D. Humphrey and A. G. MacDiarmidW-S. Huang, B. D. Humphrey and A . G. MacDiarmid I 1 I 2393 1 1 T u U 0 C a .- d e c r e a s i n g acidity peak position m A o l e t 1 Fig. 7. Cyclic voltammogram (50 mV s-l) of chemically synthesized emeraldine hydrochloride powder in 1 .O mol dm-3 HC1 (pH - 0.20) with schematic representations of the change in potential of the peaks as a function of the pH of the electrolyte. Approximate colours of the polyaniline observed at different stages of oxidation are also included. A recent investigation’ has shown that minimum resistivity of polyaniline is attained when it is electrochemically converted in aqueous solution into an oxidation state approximately midway between the completely oxidized and the completely reduced forms.This is consistent with the formulation of the emeraldine salt given above. However, no studies have been reported which indicate the nature of this ‘half-oxidized’ form, the nature of the polyaniline species which are formed or the reactions which are involved when this half-oxidized form is oxidized or reduced. It has also been shown23 that polyaniline can be used as an anode or cathode in rechargeable batteries in aqueous electrolytes. The chemical changes which the polyaniline undergoes during the charge and discharge processes have not, however, been formulated.Morphology A scanning electron micrograph (plate 1) of a smooth emeraldine hydrofluoroborate film grown electrochemically on a platinum anode at a constant potential of ca. 0.7 V us. SCE for ca. 16 h showed a relatively even compact microspheroid surface morphology. This may be compared with the scanning electron micrograph (plate 2) of an emeraldine hydroperchlorate film grown electrochemically on indium oxide conducting glass at a constant current of 50 pA cm-2 for 90 min. It is interesting to observe clearly defined fibrils on a compact microspheroid underlayer of polymer. The fibrillar morphology of the polyaniline resembles that of (CH),23 very closely, but the fibril diameter (ca. 2000 A) is greater than that found in (CH), (ca. 200 A).The fibrillar morphology reported for p~lyparaphenylene~~ and poly(3-methylthi0phene)~~ is not as similar to that of (CH), as is that here observed for the emeraldine hydroperchlorate. Note that no fibrillar morphology has been reported for polypyrrole26 and p~lythiophene.~~. 27 It is not yet known if the difference in morphology of the two emeraldine salt samples is due to the2394 Electroactivity of Polyaniline different anions employed, the nature of the electrode or the difference in the electrochemical procedures used. The dependence of the morphology upon these variables is being investigated. Electrochemical Studies A cyclic voltammogram of chemically synthesized emeraldine hydrochloride powder in 1.0 mol dm-3 HC1 (pH -0.20) together with the approximate colours at various potentials, which are identical to those observed with the electrochemically polymerized films, is given in fig.7. The approximate composition corresponding to a given colour together with the classical names28 are listed below: leucoemeraldine base (1A repeat units only; fully reduced material) [-(C,H,)-N(H)-(C,H,)-N(H)-I,, (Pale Yellow) (protonated form : pale yellow) protoemeraldine base (1A and 2A repeat units) (protonated form: light green) emeraldine base (1A and 2A repeat units) [[-(C,H,)-N(H)-(C,H,)-N(H) $& (c,H,)-N=(c,H,)=N-],], (dark blue) (protonated form : green) nigraniline base (1A and 2A repeat units) [[-(C,H,)-N(H)-(C&,)-N(H) +E-(C6H4)-N=( C,H,)=N-],], (blue-black) (protonated form: blue) pernigraniline base (2A units only ; fully oxidized material) [-(c,H4)-N=(c,H4)=N-]4x (violet) (protonated form: violet).The above colours are those observed with thin films or layers of the material on a highly reflective metal surface such as Pt. They are presumably dominated largely by the absorption spectra of the material associated with light reflected from the underlying metal surface. Bulk powders of the polymers are various shades of black in their oxidized forms. There is actually a continuum of oxidation states ranging all the way from the completely reduced leucoemeraldine to the completely oxidized pernigraniline forms. The intermediate oxidation states given above are those which have been reported as being synthesized by chemical oxidizing or reducing agents having characteristic standard reduction potentials.28 Since the degree of protonation is not known for all these species in acids of varying strengths only the composition of the free unprotonated base is listed.All species can, however, be expected to be protonated to some extent depending on the pH of the solution and the extent of oxidation of the polymer, the extent of protonation decreasing with increasing extent of oxidation. For the completely oxidized pernigraniline it is highly probable that there is very little or essentially no protonation in solutions having pH > 1. For a given oxidation state, protonation will increase with increasing acidity of the electrolyte. The smallest number of -(C6H4)-N(H)- and/or =(C6H4)=N- repeat units which can be used which will permit interconversion between the above five compositions is eight.The formulae given above for the three partly oxidized forms of polyaniline basesW-S. Huang, B. D. Humphrey and A . G . MacDiarmid 2395 do not necessarily represent the relative arrangements of 1A and 2A repeat units14 in a given polymer chain; indeed, it might be expected that different repeat units will be distributed uniformly throughout a polymer chain. It is clearly apparent that the protoemeraldine, emeraldine and nigraniline bases intermediate between the fully reduced leucoemeraldine base and the fully oxidized pernigraniline base represent only a small number of the possible combinations of 1A and 2A repeat units which could be used to represent polymers having increasing degrees of oxidation ranging all the way from the fully reduced leucoemeraldine to the fully oxidized pernigraniline polymers.For the reaction the Nernst equation is A -+ An+ +ne- (1) Eqn (2) shows that the reduction potential for the reaction is independent of pH. However, for the reaction the Nernst equation is AH, +A+nH++ne- (3) RT [A][H+], RT [A] RT n F [AH,] n F 0.059 Ered = E;ed +- In n F [AH,] = ged+- In -+- ln[H+]n [*I +0.059 log [H+]. = g e d + y l o g p n [AH,] (4) It shows that the reduction potential is clearly dependent on pH. If the pH is varied in a system consisting of a given fixed ratio of a compound, AH, and A, then a plot of E us. pH will give a straight line of slope 0.059 V per pH unit if the numbers of protons and electrons involved in the oxidation reaction are equal.However, in a reaction such as AH:-+A+2H++e- (7) where the number of protons liberated is not equal to the number of electrons, e.g. where the number of protons liberated is twice the number of electrons, an analogous plot would have a slope of (2 x 0.059) = 0.1 18 V per pH unit. These simple concepts can be used conveniently to determine the chemical reactions which occur during the electrochemical oxidation and reduction of polyaniline. It will be shown that the oxidation processes occurring for peaks 1 and 2 (fig. 7) in the cyclic voltammetry studies of polyaniline represent oxidation processes which differ in the number of protons which are lost per electron transferred. Oxidation Reactions associated with Peak 1 in the Cyclic Voltammetric Studies Oxidation Reactions between ca.pH I and 4 (Peak 1) In this pH range the potential of the initial oxidation reaction which occurs (first portion of peak 1, fig. 2 and 7) is found experimentally to be essentially independent of pH ; hence no protons are involved in the oxidation reaction. It is believed to involve the con- version of pale yellow leucoemeraldine base to the light green protonated form of pro toemeraldine :2396 Electroactivity of PolyaniIine The second oxidation reaction occurs during the second half of peak 1 to give the dark green protonated form of emeraldine : + + €(C6H,)-N(H)-(C6H4)-N(H)~ (C6H4)-N(H)=(C6H4)= N(H)+x + 2x c1- 1 c1- c1- 1 L L The electrochemical reduction reactions are the reverse of those given above. These reactions occur during the time taken for a single cyclic voltammetric scan (ca.24 s for the oxidation step). This study does not indicate whether the protonated units15 would spontaneously undergo deprotonation with the loss of HCI and be converted to f- (c6H,)-N=(c,H4)=r;(H)~ (2s’) or [-(C6H4)-N=(C,H4)=N-] (2A) c1- units if the polymer were permitted to stand in the electrolyte for an extended period in the absence of an applied potential. However, our previous investigationsA* l4 based on elemental analyses of material dried under dynamic vacuum for ca. 48 h show clearly that equilibration of emeraldine base with aqueous HC1 for 55 h results in protonation of ca. 25% of the nitrogen atoms at a pH of 1 and ca. 50% of the nitrogen atoms at a pH of - 0.2 (1 .O mol dm-3).At pH values > 4 there is essentially no protonation. Hence spontaneous deprotonation cannot be extensive in the more acidic solutions, even on standing for an extended period in the absence of an applied potential. Note that the highly conducting form of polyaniline, i.e. the emeraldine salt, can be formed in two different ways as depicted below: f N(H)-(C6H4>-N(H>-(C6H4)-N(H)-(c6H4)-N(H)-(c6H*) 3 2 2 leucoemeraldine base (insulator) oxidation (no protonation or deprotonation) I-42 e - T x A-I (10) + f- N(H)-(C6H4)-N(H)=(C6H4)=N(H)-(c6H4)-N(H)-(c6H4) 3 2 2 A- A- emeraldine salt (conductivity in metallic regime) protonation (no oxidation) T [+4x H+A-] f N(H)-(C6H4)-N=(C6H4)=N-(c6H4)-N(H)-(c6H4) t225 emeraldine base (insulator). In reaction (10) the emeraldine salt is formed by oxidation involving no change in the number of hydrogen atoms attached to nitrogen atoms.In reaction (1 1) it is formed by protonation with no accompanying change in the formal oxidation state of the polymer.W-S. Huang, B. D. Humphrey and A . G. MacDiarmid 2397 Oxidation Reactions between ca. pH -0.2 and -2.12 (Peak 1) In this pH range (1.0-6.0 mol dm-3 HCl) the potential of peak 1 (fig. 4) and also its corresponding reduction peak, l’, move to lower values at a rate of ca. 58 mV per pH unit (see fig. 4, 5 and 7) as the pH is increased, indicating that equal numbers of protons and electrons are lost in the oxidation process. In these more acidic solutions it is believed that the amine nitrogen atoms in the [-(C6H4)-N(H)-(C6H4)-N(H)-] (1A) units in leucoemeraldine will be completely or partly protonated to give + P l - [-(C6H4)-N(H>2-(C6H4>-N(H)-1 ( ”>? or units,I4 No degradation occurred on recycling in this range.Since an iminium nitrogen atom can have only one attached proton, deprotonation must occur when an ammonium nitrogen atom is oxidized to an iminium nitrogen atom, uiz. + + This is then followed by further oxidation and deprotonation during the second part of peak 1, to give a protonated form of emeraldine: + + + + -+ + + +2xH++2xe-. (13) The product of reaction (13) apparently loses HCl from the protonated amine nitrogen atoms on drying under a dynamic vacuum for ca. 48 h, since elemental analysis of the resulting material corresponds to that of the emeraldine hydrochloride given by product of reaction (9).The observation that the polyaniline is oxidized more easily in less acidic solutions, i.e. Ered becomes less positive as the pH increases, is consistent with the above reactions. The reactions associated with the reduction process (peak 1’) are the reverse of the above. Oxidation Reactions between ca. pH -0.2 and 1 (Peak 1) As described above, the potential of peak 1 is essentially independent of pH in the pH range 1-4 and decreases at the rate of ca. 58 mV per pH unit with increasing pH in the pH range ca. -0.2 (1.0 mol dm-3 HCl) to ca. -2.12 (6.0 mol dm-3 HCl). In the intermediate pH range, - 0.2 to 1, where reactions (8), (9), (12) and (1 3) are all occurring simultaneously to varying extents, the slope of the curve relating potential to pH changes from 0 mV per pH unit (for pH values 1-4) to 58 mV per pH unit (for pH values -0.2 to - 2.12) as shown in fig.5.2398 Electroactivity of Polyanilitw Note that no irreversible degradation of the redox process associated with peak 1 (see fig. 7) occurred on cycling between - 0.2 and 0.5 V us. SCE at any pH in the range - 2.12 (6.0 mol dm-3) to 4. Oxidation Reactions associated with Peak 2 in the Cyclic Voltammetric Studies Oxidation Reactions between ca. pH -0.2 and 4 (Peak 2) In this pH range the potential of peak 2 (see fig. 2, 3 and 7 ) and also its corresponding rer~uction peak, 2’, move to lower values at a rate of ca. 120 mV per pH unit as the pH is ncreased from ca. -0.2 to 4. This is consistent with a reaction of the type given by reaction (7), where the number of protons involved is twice the number of electrons.The chief reaction associated with the first part of peak 2 is therefore believed to be of the tY Pe + + [-(C6H4)-N(H)-(C6H4)-N(H) (C6H4)-N(H)=(C6H4)=N(H) (C6H4) c1- c1- -N=(C6H4)=N-],, + 4X H+ + 2~ e- + 2x c1- (14) to give a partly protonated nigraniline polymer. The reaction associated with the second part of peak 2 is believed to involve the final oxidation and deprotonation of the product given by reaction (14), viz. [-(C,H4>-N(H)-(C6H4)-N(H) kd (c6H4)-~(H)=(C6H4)=~(H) (C6H4) c1- c1- -N=(C6H4)=N-],, - [-(c6H4)-N=(c6H4)=N-]4z + 4x H t + 2x e- + 2x c1- (1 5 ) to give the fully oxidized pernigraniline. The observation that the polyaniline is oxidized more easily in less acidic solutions, i.e.Ered becoming less positive as the pH increases, is consistent with the above reactions. Some irreversible degradation of peak 2 is observed after several cycles, especially in more acidic solutions. This is believed to be associated with the hydrolysis of the imine nitrogen-carbon bond in -N=(C,H,)= groups, which become more abundant at higher levels of oxidation of the polymer, viz. -N=(c6H4)=+ H 2 0 --+ -NH, + O=(C6H4)=. (16) Oxidation Reactions between pH -0.2 and -2.12 (Peak 2) At pH values below - 0.2 the chief reaction associated with the second peak is probably of the type + + + + [-(C6H4>-N(H)2-(C6H4)-N(H), (C6H4)-N(H)=(C6H4)=N(H)-12~ - c1- c1- c1- c1- [-(C,H,)-k(H)=(C,H,)=N(H)-],, +4xH++4xe- (1 7) c1- c1- since protonation of both amine and imine nitrogen atoms is expected to be more extensive in these highly acidic solutions.On cycling between - 0.2 and 1 .O V in this pH range the intensities of peaks 2 and 2’ decreased rapidly, indicating irreversible decomposition, probably hydrolysis of the more highly oxidized material. Because of theW-S. Huang, B. D. Humphrey and A . G . MacDiarmid 2399 very great irreversible decomposition which occurred, which was significant even at a pH of ca. -0.2 for ca. 20 s it was not possible to obtain an accurate value for the dependence of potential on pH down to pH values of ca. -2.12 (6.0 mol dm-3 HCl). Hence the reactions given by reaction (17) should be regarded as only tentative. Effect of pH on the Electroactivity of Polyaniline If it is assumed that all the base forms of polyaniline, regardless of their extent of oxidation are insulators, then at pH values sufficiently high that no significant protonation can occur, all such materials would be electrochemically inactive.If however, the electrolyte were sufficiently acidic to permit some protonation of a partly oxidized form of polyaniline so that it exhibited some conductivity, then that part of the polymer in contact with the platinum electrode surface could undergo electrochemical redox reactions and would serve as a conducting medium for electron transport to the remainder of the polyaniline. Indeed, recent studies7 have shown that on electrochemically oxidizing polyaniline in 0.5 mol dm-3 NaHSO, solution (ca. pH 1) that the resistance falls to a minimum when the polymer is approximately half oxidized (emeraldine salt form using the present nomenclature) and then rises as oxidation proceeds further with presumably extensive deprotonation.These results are in agreement with those obtained in the present study. For example, the oxidation and reduction process described in the previous section did not occur readily at pH > 4, and it was necessary to increase markedly the current gain in order to obtain a meaningful reading. This is believed to be due to the fact that essentially no protonation of the polymer occurred in these very slightly acidic electrolytes and hence very little conductivity could be imparted to the polymer film. This was substantiated by the observation that at pH 6 for example, the polyaniline was essentially electrochemically inactive, but regained its electrochemical activity in a completely reversible manner when placed in a more acidic electrolyte.Conclusions The present study presents evidence to show that the term ‘polyaniline’ describes a class of compounds composed of species which differ in their degree of oxidation (ratio of imine nitrogen atoms to amine nitrogen atoms). It is proposed that the polymer base f(C,H,)-N(H)-(C,H,)-N(H) (C6H,)-N=(C6H4)=Njb can exist in principle in a continuum of oxidation states ranging from the completely reduced polymer where b = 0 to the completely oxidized polymer where a = 0. Each oxidation state, defined by a given fixed ratio of imine to amine nitrogen atoms, can exist in forms which differ from each other in their extent of protonation which depends on the experimental conditions to which the polymer base has been subjected.The electrochemical oxidation and reduction of polyaniline in aqueous electrolytes of varying pH values shows two classes of redox processes occurring at different potentials which differ from each other by the extent to which the processes are accompanied by deprotonation (during oxidation) or protonation (during reduction). The emeraldine salt of polyaniline can be synthesized in a form having a fibrillar morphology similar to that of polyacetylene. Its conductivity lies in the metallic regime and is qualitatively consistent with a proposed symmetrical conjugated structure having extensive charge delocalization. This results from a new type of doping of an organic polymer-salt formation instead of oxidation, and as such suggests possible hitherto unexpected classes of conducting polymers.These studies were supported in part by the University of Pennsylvania Materials Research Laboratory through N.S.F. grant no. DMR-82-16718 (W. S.H.) and the Office of Naval Research (B. D. H.).2400 Electroact ivity of Polyan iline References 1 J. Langer, Solid State Commun., 1978, 26, 839; G. Mengoli, M. T. Munari, P. Bianco and M. M. Musiani, J. Appl. Polym. Sci., 1981, 26, 4247; R. Noufi, A. J. Nozik, J. White and L. F. Warren, J. Electrochem. SOC., 1982, 129, 2261; E. M. Genies, A. A. Syed and C. Tsintavis, Mol. Cryst. Liq. Cryst., 1985,121, 181; J. P. Travers, J. Chroboczek, F. Devreux, F.Genoud, M. Nechtschein, A. Syed, E. M. Genies and C. Tsintavis, Mol. Cryst. Liq. Cryst., 1985,121, 195; D. W. DeBerry, J . Electrochem. SOC., 1985, 132, 1022; C. M. Carlin, L. J. Kepley and A. J. Bard, J. Electrochem. SOC., 1985,132,353; M. Kaneko and H. Nakamura, J. Chem. Soc., Chem. Commun., 1985,346; W. R. Salaneck, B. Liedberg, 0. Inganas, R. Erlandsson, I. Lundstrom, A. G. MacDiarmid, M. Halpern and N. L. D. Somasiri, Mol. Cryst. Liq. Cryst., 1985, 121, 191; A. G. MacDiarmid, J. C. Chiang, W. S. Huang, B. D. Hum- phrey and N. L. D. Somasiri, Mol. Cryst. Liq. Cryst., 1985, 125, 309. 2 A. G. MacDiarmid, S. L. Mu, N. L. D. Somasiri and W. Wu, Mol. Crysr. Liq. Cryst., 1985, 121, 187. 3 M. Jozefowicz, J. H. Perichon, L. T. Yu and R. E. Buvet, Br. Patent no.1216569 (1970). 4 R. DeSurville, M. Jozefowicz, L. T. Yu, J. Perichon and R. Buvet, Electrochim. Acta, 1968, 13, 1451. 5 A. F. Diaz and J. A. Logan, J . Electroanal. Chem., 1980, 111, 1 1 1. 6 T. Kobayashi, H. Yoneyama and H. Tamura, J. Electroanal. Chem., 1984, 161, 419. 7 E. W. Paul, A. J. Ricco and M. S. Wrighton, J. Phys. Chem., 1985,89, 1441. 8 A. G. MacDiarmid, J. C. Chiang, M. Halpern, W. S . Huang, S . L. Mu, N. L. D. Somasiri, W. Wu and 9 A. Kitani, J. Izumi, J. Yano, Y. Hiromoto and K. Sasaki, Bull. Chem. SOC. Jpn, 1984, 57, 2254. S. I. Yaniger, Mol. Cryst. Liq. Cryst., 1985, 121, 173. 10 T. Ohsaka, Y. Ohnuki, N. Oyama, G. Katagiri and K. Kamisako, J . Electroanal. Chem., 1984,161,399. 11 N. Oyama, Y. Ohnuki, K. Chiba and T. Ohsaka, Chem. Lett., 1983, 11, 1759. 12 T. Kobayashi, H. Yoneyama and H. Tamura, J. Electroanal. Chem., 1984, 177, 293. 13 T. Kobayashi, H. Yoneyama and H. Tamura, J. Electroanal. Chem., 1984, 177, 281. 14 J. C. Chiang and A. G. MacDiarmid, Synth. Met., 1986, 13, 193. 15 N. L. D. Somasiri, A. R. Richter, J. C. Chiang and A. G. MacDiarmid, unpublished observations (1 984-85). Detailed synthetic procedures, elemental analyses and infrared spectra will be published elsewhere. 16 Schwartzkopf Microanalytical Laboratory, Woodside, N.Y. 1 1377, U.S.A., unpublished results. 17 N. L. D. Somasiri and A. G . MacDiarmid, unpublished observations (1984). 18 C. H. Rochester, Acidity Functions (Academic Press, London, 1970), p. 39. 19 M. A. Paul and F. A. Long, Chem. Rev., 1957, 57, 1. 20 A. G. MacDiarmid and A. J. Heeger, Synth. Met., 1979180, 1, 101. 21 J. B. Hendrickson, D. J. Cram and G. S . Hammond, Organic Chemisfry (McGraw-Hill, New York, 3rd edn, 1972), p. 306. 22 M. Liska, B. Stehlik and A. Tkac, Chem. Zvesti, 1951, 5, 31; C. R. Noller, Chemistry of Organic Compounds (W. B. Saunders, Philadelphia, 1966), p. 343; J. B. Hendrickson, D. J. Cram and G. S . Hammond, Organic Chemistry (McGraw-Hill, New York, 3rd edn, 1972), pp. 3 15-317. 23 C. R. Fincher Jr, D. Moses, A. J. Heeger and A. G. MacDiarmid, Synth. Met., 1983, 6, 243. 24 P. Pradere, A. Boudet, J-Y. Goblot, G. Froyer and F. Maurice, Mol. Cryst. Liq. Cryst., 1985,118,277. 25 G. Tourillon and F. Gamier, Mol. Cryst. Liq. Cryst., 1985, 118, 221. 26 A. F. Diaz, W-Y. Lee and A. Logan, J. Electroanul. Chem., 1980, 108, 377. 27 R. J. Waltman, J. Bargon and A. F. Diaz, J. Phys. Chem., 1983,87, 1459. 28 A. G. Green and A. E. Woodhead, J. Chem. SOC., 1910,97, 2388; 1912, 101, 1117. Paper 51 1554; Received 10th September, 1985
ISSN:0300-9599
DOI:10.1039/F19868202385
出版商:RSC
年代:1986
数据来源: RSC
|
15. |
Fourier-transform infrared spectroscopy of colloidalα-,β- andγ-ferric oxide hydroxides |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 82,
Issue 8,
1986,
Page 2401-2410
Tatsuo Ishikawa,
Preview
|
PDF (638KB)
|
|
摘要:
J. Chem. Sac., Faraday Trans. I, 1986,82, 2401-2410 Fourier-transform Infrared Spectroscopy of Colloidal a-, p- and y-Ferric Oxide Hydroxides Tatsuo Ishikawa,* Satomi Nitta and Seiichi Kondo School of Chemistry, Osaka University of Education, 4 4 8 Minamikawahori-cho, Tennoji-ku, Osaka 543, Japan Fourier-transform infrared spectra and surface properties of colloidal a-, 8- and y-ferric oxide hydroxides [Fe(O)OH], which were synthesized and aged in this laboratory, have been studied. The spectrum of the a phase has a strong absorption band at 3150 cm-l and two weak ones at 3485 and 3661 cm-'. That of the B phase has a strong band at 3480 cm-l, a weak band at 3659 cm-l and two very weak bands at 3686 and 3723 cm-l. That of the y phase has a strong band at 3160 cm-l, a weak band at 3624 cm-l and a very weak band at 3528 cm-l.The changes of these bands by the adsorption of heavy water, water and methyl iodide molecules and cupric ions have been studied in detail and the various bands were assigned to various OH groups with respect to the crystal structures of these materials. There are many studies on the properties of colloidal a-, /3- and y-ferric oxide hydroxides, in relation to industrial applications such as starting materials for manufacturing pigments, colour materials and ferromagnetic ferrites and to soil properties and corrosion behaviour. However, their surfaces are still not well characterized. 1.r. spectroscopy was first carried out on the a phase by Russell et al., who identified two absorption bands at 3485 and 3661 cm-l.' Parfitt et al.studied the i.r. spectra of the a phase after adsorption of phosphate and sulphate ions from aqueous solutions and discussed the adsorption sites2* Inouye and coworkers studied the gas-adsorption isotherms of various molecules on the a, p and y phases in relation to the structures of these materiak49 The purpose of this report is to assign the v(0H) band associated with adsorption on the well crystallized a, /3 and y phases to surface sites by means of FTIR spectroscopy and thermodynamic methods. Experimental Materials a-Ferric oxide hydroxide was prepared by adding 1 .O an01 dm-3 sodium hydroxide solution to 0.1 mol dmP3 ferric nitrate solution up to pH 12 at room temperature. The precipitates were aged at pH 12 at room temperature for 10 days, then washed with distilled and deionized water and dried in air at 373 K for 5 h.P-Ferric oxide hydroxide was prepared by boiling 0.1 mol dm-3 ferric chloride solution containing 6 wt % urea. The precipitates were washed with water and dried in air at 373 K for 5 h. y-Ferric oxide hydroxide was obtained by adding hexamethylenetetramine to 0.3 mol dm-3 ferrous chloride solution and oxidizing the resultant ferrous hydroxide suspension with sodium nitrite at 333 K for 45 min. The precipitates were washed with water and dried at 333 K for 10 h. Crystal structures, crystallinity and purity of these samples were examined by means of high-intensity X-ray diffraction (Cu K,, 45 kV, 120 mA) and the absence of residual anion impurities used in the preparations was checked by i.r.All three modifications are non-porous crystalline powders. Their specific surface areas obtained 240 12402 FTIR of Colloidal Ferric Oxide Hydroxides Table 1. Specific surface areas (A,) and particle sizes of a-, j?- and y-ferric oxide hydroxides - -~ mean particle size/nm - sample As/m2 g-' length width thickness a phase 82 310 43 7 p phase 33 260 40 34 y phase I10 160 35 5 cn m W L1 * . . . . . . . I . - . * . . . 3800 3600 3400 3200 3000 2800 3750 3700 3650 w avenumber/ cm -' Fig. 1. The i.r. spectra of a-, /I- and y-ferric oxide hydroxides, treated in U ~ C U O at 348 K for 30 min. (-) a phase; ( . . a ) p phase; (---) y phase. by the nitrogen B.E.T. method are listed in table 1. The particle shapes of the a, p and y phases observed by TEM are close to rectangular thin plates, rectangular rods and rectangular thin plates, respectively.The thickness of these crystals in table 1 was calculated from the densities and specific surface areas of the corresponding materials. Methods Near4.r. spectra were measured mostly in the range 500&2000cm-l with 2cm-l resolution using an FTIR spectrometer (DIGILAB FTS 15E) with a PbSe detector with very high sensitivity in the region of interest. Quartz windows were used in a vacuum cell in which the temperatures of samples were controlled in situ. Samples were pasted on thin glass plates to form a thin layer of material for measuring bands of high absorbance. Otherwise, ordinary disc samples were used for low-absorbance samples. Pressures of water and methyl iodide vapour introduced into the sample cell were measured by a Baratron manometer.lHJH isotope exchange of surface OH groups was carried out by 30 to 40 cycles of adsorption of heavy water on to the sample at 1.33 x lo3 Pa at 298 K for 5 min, followed by desorption. Cupric ions were adsorbed by immersing the samples in a 0.1 mol dm-3 solution of cupric chloride at pH 5 at room temperature and then by drying at the temperatures indicated above for each sample. The amount of cupric ions adsorbed was determined by atomic absorption spectroscopy from the difference in the concentrations of the solutions before and after sample immersion.T. Ishikawa, S . Nitta and S . Kondo Table 2. The influence of adsorption of heavy water, water, methyl iodide and cupric ions on v(0H) bandsa 4OH) adsorption band assignment sample /cm-' D20 H 2 0 CH,I Cu2+ v(0H) 3480 (s) L3723 (vw) 0 00 00 0 00 00 00 0 00 00 X X 00 00 00 00 X X 00 00 00 00 X X X X 00 00 X 00 00 X 00 X X 00 ~- bulk OH surface OH surface OH bulk OH surface OH surface OH surface OH bulk OH ? surface OH a The active, less active snd much less active bands are marked by 00, o and x, respectively.(s), (w) and (vw) indicate strong, weak and very weak bands, respectively. 2403 2 7 3 323 373 4 2 3 4 7 3 323 373 423 47 TlK Fig. 2. The relation between the absorbances of v(0H) bands and temperatures of heating in uacuo. 0,3661 cm-l (aphase); 0,3485 cm-'(aphase); 0,3150 cm-l(aphase); A, 3659 cm-l @phase); A, 3480cm-' (D phase); 0, 3624cm-I ( y phase); W, 3528 cm-l ( y phase); 0, 3160cm-l (Y phase).Results and Discussion 1.r. spectra of a-, 8- and y-ferric oxide hydroxides which were treated in vacuo at 348 K for 30 min are shown in fig. 1. They have various strong and weak absorption bands, together with a very weak set from the 8 phase, as is shown in the right-hand side of this figure with an expanded absorbance scale. The positions of all these bands are listed in table 2 with their relative absorbances marked s, w and vw for strong, weak and very weak bands, respectively. All the intensities of absorption bands of the a and y phases decrease on heating and disappear at ca. 473 K with the accompanying decomposition of these materials to anhydrous oxides,6 as is seen from fig. 2, which shows the change2404 8.C 6.C a T .fl 4 .c 9 2 2.0 0' FTIR of Colloidal Ferric Oxide Hydroxides -1 0.4 I I I I ' J o 5 10 15 20 ageing time/days Fig. 3. The change of the absorbances of v(0H) bands of a-ferric oxide hydroxide by ageing in water. 0, 3661 crn-l; a, 3485 crn-l; 0, 3150 cm-l. I I I 3500 3000 2500 wavenumber/cm-' Fig. 4. The change of the spectrum of y-ferric oxide hydroxide following lH-2H exchange. (-) Before 1H-2H exchange; (---) after 1H-2H exchange (6 times); ( * - - ) after 1H-2H exchange (15 times). of intensity of the weak and strong bands. The intensity of the weak absorption band of the /? phase increases on heating in vacuo up to 423 K and then decreases. This is probably because water strongly adsorbed in the structural channels discussed elsewhere' desorbs on heating. As is seen in fig.3, the intensities of three bands of the a phase increase proportionally with the time of ageing in water at 373 K and are maximal at ca. 10 days. Also, the X-ray diffraction patterns showed similar behaviour. These results indicate that crystals of theT. Ishikawa, S. Nitta and S. Kondo 2405 0 0 10 20 30 40 50 cycles of H-2H exchange Fig. 5. The change of the absorbances of v(0H) bands by 1H-2H exchange. 0, 3150 cm-l (a phase); A, 3480 cm-l (D phase); 0, 3160 cm-l (y phase). Ln 0 3700 3600 3500 w avenum ber/ cm - Fig. 6. The effect of water adsorption on the spectrum of y-ferric oxide hydroxide. (-) Before adsorption; (---) after adsorption at 26.6 Pa at 298 K; (-.-) after adsorption at 106 Pa at 298 K.2406 FTIR of Colloidal Ferric Oxide Hydroxides W N In , p? I 37 00 3600 3500 wavenumber/cm-' Fig.7. The influence of methyl iodide adsorption on the spectrum of y-ferric oxide hydroxide. (-) Before adsorption; (---) after adsorption at 1.33 x lo4 Pa at 298 K ; (..-) after desorption at 323 K. a phase grew sufficiently after 10 days ageing. It was possible to obtain well developed crystals of the and y phases by the methods of preparation used here, with almost no further growth of the crystals by ageing the materials. The behaviour of the absorption bands of the y phase is described below in detail. In fig. 4, both the 3528 and 3624 cm-l bands change rapidly to the two corresponding bands at 2610 and 2670 em-' by 1H-2H exchange, while the strong band at 3160 cm-l changed to a band at 2356 cm-l, but more slowly compared with the former bands.All the ratios of positions (in wavenumbers) of new bands after 1H-2H exchange to those of the original bands are from 1.34 to 1.36, equal to the isotope ratio vOFT/vOD. Hence, these three bands at 3 160, 3528 and 3624 cm-l are without doubt due to OH stretching vibrations. This behaviour is more or less similar to that of the a and p phases. All the weak bands of the a and p phases can be replaced by v(0D) bands very easily, compared with the strong bands, which can be exchanged only very slowly. These results indicate that the OH groups responsible for these weak bands exist on the surface, while the strong bands lie within the bulk structure. Fig. 5 shows the 1H-2H exchange rate of strong bands of each modification. The exchange rate of these modifications is in the order of y $= a > p phases, because OH groups between the layers of the y phase constitute linear hydrogen- bonded chains, so that migration of protons might be easier than in other cases, as is suggested by the crystal structures shown schematically in fig.9, 10 and 11 (later). Gallagher investigated the 1H-3H exchange of OH groups of the a and phases by means of a radioactive tracer method and found that OH groups of the a phase are less exchangeable than those of the a phase,s consistent with the present result. Fig. 6 illustrates the spectra of the y phase before and after water adsorption. The 3624 cm-l band disappeared after adsorbing water vapour and a corresponding perturbed v(0H) band appeared at 3550 crn-l, but, on the other hand, the 3528 cm-l band seemed almost unchanged.T.Ishikawa, S . Nitta and S. Kondo 2407 I I I I I 37 00 3 600 3500 wavenumberlcm-’ Fig. 8. The effect of cupric ion adsorption on the spectrum of y-ferric oxide hydroxide. The amount of adsorbed cupric ion is 5.68 pmol g-l. (-) Before adsorption; (-..) after adsorption. As is seen in fig. 7, the 3624 cm-l band disappeared on introducing methyl iodide gas and a small perturbed band appeared at 3565 cm-l. After evacuation at 323 K, the former band recovered ca. 70430% of its original intensity and the latter band disappeared. This may be because methyl iodide gas has reacted with OH groups to some extent. This reaction was confirmed by measuring the weight change of this material before and after adsorption and desorption of methyl iodide gas.The 3528 and 3 160 cm-l bands showed no change on water or methyl iodide adsorption. The 3661 and 3485 cm-l bands of the a phase and the 3659 cm- band of the p phase disappeared after water and methyl iodide adsorption, but the strong bands of the a and p phases did not change. All of these v(0H) bands showed behaviour similar to those mentioned above on methanol adsorption, with the exception of the a phase. When methanol was adsorbed on this material, not only did the 3485 and 3661 cm-l bands shift to lower wavenumber but also a very small and somewhat broad band appeared at 3640 cm-l which did not disappear on evacuation. When the vacuum system was contaminated by, for instance, vacuum grease, this band appeared more clearly together with a v(CH) band.Therefore, this adsorption site may possibly be some kind of chemically active Lewis acid site which would then be the same site as that adsorbing nitrogen oxide and sulphur dioxide gas,* and as that found by Rochester and T ~ p h a m . ~ $ lo As is seen in fig. 8, the 3624 cm-l band of the y phase almost disappeared on adsorption of cupric ions. This result indicates that the protons of these OH groups are ion-exchanged by cupric ions. The intensity of the 3528 cm-l band was also decreased to some degree by the same procedure, although this band showed no change on water or methyl iodide adsorption. This indicates that this site is located on the surface, but has less hydrogen-bond and dipole interaction with adsorbates than ordinary OH groups do.It is more likely that this is a hydroxide ion site. The results of 1H-2H isotope exchange and adsorption of water, methyl iodide and cupric ions on each band described above are summarized2408 FTIR of Colloidal Ferric Oxide Hydroxides 2 3 4 0 : O @:OH O I F e - - - - : H bond @ : surface OorOH Fig. 9. (a) The crystal structure of a-ferric oxide hydroxide. Double lines show hydrogen bonds. a, b and c are the three crystal axes. (b) The crystal plane (OOl), perpendicular to the predominant plane (010). Dotted lines show hydrogen bonds. bb 0 : O @:OHO:Fe 8 : surface OorOH Fig. 10. (a) The crystal structure of /?-ferric oxide hydroxide. (b) The crystal plane (OOl), perpendicular to the predominant planes [(OlO) and (loo)].T.Ishikawa, S. Nitta and S. Kondo 2409 - - - - : H bond 8 : surface 0 or OH Fig. 11. (a) The crystal structure of y-ferric oxide hydroxide. ( h ) The crystal plane (IOO), perpendicular to the predominant plane (0 10). in table 2, in which the letters 00, o and x show the bands as active, less active and much less active, respectively, for interactions with the molecules or ions mentioned above. The origins of these absorption bands are discussed in some detail on each modification as follows. The predominant crystal planes of the a and y phases, determined by selected-area electron diffraction patterns, are (100) and (0 10) planes, respectively,ll and those of the /? phase are (010) and (100) planes.I2 The fractions of the areas of predominant surfaces to the total surface areas of these particles which can be calculated from the particle sizes and B.E.T.surface areas of these samples, are 84, 93 and 84% for the a, /? and y phases, respectively. Therefore, most of the OH groups which are active to adsorbates are on these surfaces, with those less active being inside the substrates. The crystal structures of the a, p and y phases are schematically illustrated in fig. 9 (a), 10 (a) and 1 1 (a), respectively. The perpendicular sections of the predominant surface planes are also shown in fig. 9(h), 10(b) and 11 (b). The crystal structure of the a phase in fig. 9 (a) consists of strips of condensed octahedra containing ferric ions at each centre and oxygen ions or OH groups at each corner as well. There are channels surrounded by these strips.These strips share 0 atoms on octahedral edges to each other, and the hydrogen bonds are shown by double lines.I4 In fig. 9(b), the a phase can probably have five kinds of OH sites, i.e. 1, 2, 3 and 4 on the surface and site 5 in the channels. The strongly perturbed and intense v(0H) band at 3150 cm-l can be assigned to the bulk OH groups on site 5 , which interact with neighbouring oxygen atoms with hydrogen bonds and whose number is largest compared with other sites, since this band changes to a v(0H) band only slowly and is inactive to adsorption of molecules. The 3485 cm-l band is that of a perturbed v(0H) and can 80 FAR 12410 FTIR of Colloidal Ferric Oxide Hydroxides probably be assigned to OH groups and/or ions on site 1 and/or 3, since these OH groups are coordinated to three ferric ions and bear more negative charge than those which have a higher coordination number.The 3661 cm-l band, which is less perturbed than 3485 cm-' band, may be assigned to OH groups on sites 2 and/or 4 which coordinate to one or two ferric ions and have correspondingly less negative charge. In the crystal structure of the p phase in fig. 10(a),12 octahedra linked to each other form strips parallel to the c-axis. The channels surrounded by these strips are said to contain chloride ions and water molecules alternately in line, with these ions and molecules more or less stable below 423 K.14 The p phase can have five types of OH sites 1, 2, 3, 4 and 5 as is shown in fig. lO(b). The strong 3485 cm-l band can be assigned to hydroxide ions on site 4.This is because the number of sites of this type is the largest. This site lies inside the channels and interacts with water molecules and chloride ions, hence its low wavenumber and also the very slow 1H-2H exchange rate of this band. The OH groups on site 1, which lies on the concave part of the surface, and probably site 5 , which lies inside the empty channels, seem to be responsible for the weak 3659 cm-1 band, i.e. these sites coordinate to three ferric ions and are active to 1H-2H exchange, as well as to adsorbates and cupric ions. The very weak 3686 and 3723 cm-l bands can be assigned to OH groups on site 2 coordinated to two ferric ions and site 3 coordinated to one ferric ion, respectively, in that both of these OH groups are active to 1H-2H exchange, the adsorption of water molecules and the fact that the OH bond of the latter OH group would be more covalent than that of the former site.In the structure of the y phase in fig. 1 1 (a), octahedra constitute two-dimensional layers by sharing edges of each octahedron, with linear chains of hydrogen bonds shown as double lines and dotted lines in fig. 11 ( a ) and (b), respectively, and form crystals of thin ~ 1 a t e s . l ~ Only sites 1 and 2 are present in this structure, in which the number of former sites is much less than that of the latter, as is shown in fig. 11 (b). The strongly perturbed 3160 cm-l band can be assigned to OH groups on site 2 between layers. Site 1 on the surface is the origin of the weak 3624 cm-1 band. Finally, it was not possible to assign the 3528 cm-l band in the present investigation. We thank Dr Yoshiko Nakahara of the Government Industrial Research Institute, Osaka, for her experimental help in transmission electron microscope and X-ray diffraction experiments. References 1 J. D. Russell, R. L. Parfitt, A. R. Fraser and V. C. Farmer, Nature (London), 1974, 248, 220. 2 R. L. Parfitt, J. D. Russell and V. C. Farmer, J. Chem. Soc., Faraday Trans. I, 1976, 72, 1082. 3 R. L. Parfitt and R. S. C. Smart, J. Chem. SOC., Faraduy Trans. 1, 1977,73, 796. 4 K. Kaneko, M. Senzawa, T. Ishikawa and K. Inouye, Bull. Chem. Soc. Jpn, 1975,48, 1764. 5 T. Ishikawa and K. Inouye, Prog. Colloid Polym. Sci., 1983, 68, 152. 6 T. Ishikawa and K. Inouye, Bull. Chem. SOC. Jpn, 1972,45, 2350. 7 T. Ishikawa, K. Kaneko and K. Inouye, Nippon Kagaku Kaishi, 1975, 1635. 8 K. J. Gallagher and D. N. Phillips, Chimia, 1969, 23, 465. 9 C. H. Rochester and S. A. Topham, J. Chem. SOC., Faraday Trans. 1 , 1979, 75, 591. 10 C. H. Rochester and S. A. Topham, J. Chem. Soc., Furaduy Trans. I , 1979, 75, 872. 11 G. W. van Oosterhaut, Acra Crystallogr., 1960, 13, 932. 12 A. L. Mackay, Miner. Mag., 1960, 32, 545. 13 F. J. Ewing, J. Chem. Phys., 1935, 3, 203. 14 T. Ishikawa and K. Inouye, Bull. Chem. SOC. Jpn, 1975, 48, 1580. 15 F. J. Ewing, J. Chem. Phys., 1935, 3, 420. Paper 511557; Received 10th September, 1985
ISSN:0300-9599
DOI:10.1039/F19868202401
出版商:RSC
年代:1986
数据来源: RSC
|
16. |
Small-angle neutron scattering studies of microemulsions stabilised by aerosol-OT. Part 3.—The effect of additives on phase stability and droplet structure |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 82,
Issue 8,
1986,
Page 2411-2422
Andrew M. Howe,
Preview
|
PDF (878KB)
|
|
摘要:
J. Chem. SOC., Faraday Trans. I , 1986,82, 2411-2422 Small-angle Neutron Scattering Studies of Microemulsions Stabilised by Aerosol-OT Part 3.-The Effect of Additives on Phase Stability and Droplet Structure Andrew M. Howe,"? Christ0 Toprakcioglu,$ John C. Dore and Brian H. Robinson Chemical Laboratory and Physics Laboratory, University of Kent, Canterbury, Kent CT2 7NH The droplet structure of water-in-oil microemulsions formed with the surfactant AOT in various hydrocarbon solvents has been studied by small-angle neutron scattering (SANS). As the temperature of the single-phase microemulsion is raised towards that at which a phase transition occurs, an increased scattering at low Q, characteristic of critical scattering, is exhibited, which is a result of the attractive interactions between the droplets.The effect on the microemulsion of small quantities of additives (toluene, octan- 1-01 and benzyl alcohol) has been investigated for the n-alkane solvents heptane, decane, undecane and dodecane. Both a systematic increase or decrease in the transition temperature and a corresponding change in the magnitude of the critical scattering component of the SANS intensity profile are observed, which depend on the nature of the additive. These results are coupled to a change in the water droplet radius, consistent with a model in which amphiphilic molecules such as benzyl alcohol and octanol are preferentially adsorbed into the water-surfactant interfacial region, whereas toluene remains predominantly in the bulk hydrocarbon phase and probably does not come into contact with the aqueous core of the droplet.The droplet radius and the phase stability of the microemulsion are also affected by addition of short-chain alcohols and show a systematic variation as the chain length is increased. These changes have been interpreted in terms of the partitioning of alcohols between the water core and the interface and their effect on the surfactant shell. Microemulsions are transparent, thermodynamically stable dispersions of water and oi1.l Usually both a surfactant and a cosurfactant are necessary to stabilise these dispersions. The structure and stability of microemulsions are the subject of much experimental s t ~ d y l - ~ and theoretical Changes in the phase behaviour and inter-droplet interactions as a result of a change in the composition (e.g.the oil, cosurfactant or addition of salt) have been reported for four- (or more) component ~ystems.~~ Microemulsions are potentially of great commercial interest, for example in enhanced oil recover^,^ solar-energy conversionlo. l1 and as media for novel, enzyme- cat a1 y sed synthetic processes. 2-1 * In this study we use AOT [sodium 1,4-bis(2-ethylhexyl)sulphosuccinate] as the surfactant. It is not necessary to use a cosurfactant to stabilise these microemulsions and it is therefore possible to study the effects of temperature, oil and cosurfactant separately or together. In AOT-stabilised microemulsions the droplet radius, r, increases in an approximately linear fashion with the mole ratio of water to AOT, R.We have previously t Permanent address: AFRC Institute of Food Research, Colney Lane, Norwich NR4 7UA. Present address : Cavendish Laboratory, Madingley Road, Cambridge CB3 OHE. 241 1 80-22412 Microemulsions stabilised by Aerosol-0 T reported small-angle neutron scattering (SANS) investigations of the droplet str~cture,'~ the polydispersity16 and the inter-droplet interactions on approaching a phase transition1' in AOT-stabilised microemulsions. In the present paper we report a SANS and photon correlation spectroscopy (PCS) study of the effect of addition of alcohol and changing the oil or temperature on the droplet sizes and phase stability of these systems. The studies are limited to low volume fractions (< 0.1) of disperse phase (water and AOT) where the microemulsion is thought to consist of discrete AOT-coated drops in an oil- continuous medium.Experimental Materials The AOT was obtained from Sigma as the sodium salt of dioctylsulphosuccinate and was used without further purification. It had a quoted purity of >99% and was found to contain only minimum amounts of the alcohol and carboxylic acid impurities, which form as a result of se1f-hydrolysis.l8 Purity was confirmed by interfacial tension measurements at the heptane-water interface. The sample showed similar phase behaviour to a pure sample of AOT, kindly provided by Prof. L. Magid. The n-alkane solvents (heptane, decane, undecane and dodecane) and the additives (methanol, octanol, benzyl alcohol and toluene) were Aldrich or BDH materials. Heavy water (zHH,O) of 99.9% isotopic purity was purchased from Prochem (BOC).The microemulsions were prepared by adding small quantities of heavy water with the aid of a microsyringe to a solution of AOT in the appropriate solvent and shaking the mixture vigorously until a visually clear, single-phase liquid was obtained. The additives, which in most cases had a volume fraction of only a few percent, were also introduced by means of a microsyringe. Upper-temperature Phase Separation The extent of the stable single-phase region of the microemulsion may be studied by varying the temperature until the clear dispersion acquires a cloudy appearance and separates into two phases. This defines a cloud point, T,, which may be studied as a function of composition for a fixed AOT concentration.Results for the upper-temperature phase transition Ty have already been presented17 for the systems 2H,0-AOT-undecane and 2H,0-AOT-dodecane at an AOT concentration of 0.1 mol dm-3 with varying values of the molar ratio of water to surfactant, R. Additives effect a shift in the cloud point. There is also a strong isotope effect;17 2H,0 has been used for many of the measurements, so that the results can be directly related to the corresponding SANS profile for each sample. The 7': values are reproducible, but since visual observation is used, they have an uncertainty of k0.5 "C in most cases. At 7': the liquid separates into two phases. If the temperature is raised further, it may be possible to form three distinct phases. A more detailed study of the different coexisting phases above 7':J is in progress and will be presented in a later paper.lg Apparatus Neutron-scattering measurements were performed on the Pluto small-angle facilityz0 at AERE, Harwell, and also on the D17 instrument21 at the Institut Laue-Langevin (ILL), Grenoble.The wavelength of the neutron beam was 6 A for the AERE experiments and lOA for the ILL experiments; the scattering vector covered a range from 0.013 to 0.18 A-l. The samples were contained in stoppered quartz cells with a neutron pathlength of 1 .O mm and the temperature was maintained constant to f 0.5 "C throughout each run. The SANS intensity was recorded with a two-dimensional multidetector and theA . M . Howe et al. 241 3 usual data reduction procedures22 were used to produce the normalised intensity distributions Iobs(Q). For studies on the effect of short-chain alcohols, it was found convenient to use photon correlation spectroscopy (PCS) to determine the hydrodynamic radius, rh.The apparatus consists of a 2 W Spectra Physics model 168 argon-ion laser with a K7025 Malvern correlator and an EM1 9863 photomultiplier tube in a Malvern RR109 housing system. The equipment and data analysis procedure have been described previo~sly.~~ Measurements were made with solubilised H 2 0 (instead of 2H20) for a fixed value of R = 20, for a range of alcohols (n, = 1-8) for which the phase stability results have already been obtained. Theory For a monodisperse microemulsion the observed intensity lobs(Q) of scattered neutrons may be expressed as where n is the number of droplets in the neutron beam and F(Q, r ) is the droplet form-factor for spherical droplets Zobs(Q9 r ) = nF(Q, r ) S(Q) (1) sin (Qr) - Qr cos (Qr) Q3r3 F(Q, 4 = 9 W 2 P ( with a radius, r, volume, V, contrast, Ap, and where S(Q) is the structure factor for the droplet distribution.In the case of a dilute dispersion with a volume fraction of water and AOT < 0.1 the structure factor for these particular droplet dispersions, which have a relatively soft interaction potential, approximates to unity (except near the phase- transition region). Under these circumstances the extrapolated value at Q = 0 may be written as I(0) = nF(0) = K n P (3) where K is a calibration constant. For small values of Qr (< 1) the radius of the particles may be obtained from the linear portion of a Guinier plot [In I(Q) us.Q2] at low Q2 where the Guinier law applies: (4) The microemulsion droplets are subject to a weakly attractive interaction which increases as the temperature is raised to the phase-stability limit. At 7'y there is extensive clustering of the droplets, the system becomes turbid and eventually separates into two phases. Above Ty three distinct phases may be observed. For temperatures less than 10 K below 7':J there is an incipient clustering of the droplets, which produces critical scattering effects. Under these circumstances the contribution to the SANS intensity can be described by the Ornstein-Zernike In I(Q) = In I(0) - r2Q2/5. 25 as where t is the correlation length and K is a scaling factor.Both 5 and K obey scaling laws with respect to the reduced temperature: 5 = and K = K ~ E - Y ( 6 4 & = [ s q with critical exponents v and y for a critical temperature qrit. Note that Grit cannot, in general, be identified with 7':.2414 Microemulsions stabilised by Aerosol-0 T Fig. 1. Phase in heptane in 101 I I I I I I 1 0 10 20 30 40 50 60 70 TI0C diagram of the water-in-oil microemulsion system stabilised by 0.1 mol dm the presence of additives: (a) 0.1 mol dmP3 benzyl alcohol; (b) no additive; 10% v/v toluene. -3 AOT and (c) Results Phase Behaviour and PCS The effect of additives on the stable microemulsion region with respect to temperature is shown in fig. 1 for the H,O-AOT heptane system. The data for pure heptane are typical of the general shape of the stability curve for a wide range of alkane solvents.The behaviour for benzyl alcohol (systematic decrease in phase transition temperatures with increasing concentration) is qualitatively different as the displacement is in the opposite sense to that for toluene and octanol, as shown in fig. 1 and 2 and table 1. The effects due to addition of different short-chain alcohols (C,,H,,,+,OH), to 0.05 mol dmP3, to the 0.1 mol dm-3 AOT-H,O (R = 20)-undecane system are also of interest in this context. The upper transition temperature is reduced by the addition of methanol and ethanol, which are freely soluble in water, whereas longer-chain alcohols (nc > 3) cause an increase in the Ty value. The variation is as shown in fig. 3(a). The lower phase transition exhibits the opposite behaviour for the short-chain alcoholsA .M. Howe et al. 60 40 4 ’ 241 5 ( C 1 / - / 9’ / 0 0 0 - / 0 / , 9’ / 0’ / 7’ I I 1 I 40 I I + I 0.1 0.2 [ benzyl alcohol]/mol dm-3 55 i 0.1 0.2 [ benzyl alcohol]/mol dm-3 / *’ 4 0 1 0.05 0.1 0 [octanoll/mol dm-3 Fig. 2. The variation of the upper-transition temperature of microemulsions, stabilised by 0.1 mol dm-3 AOT, as a function of additive concentration for (a) R = 40 (H,Otheptane-benzyl alcohol, (b) R = 30 (H,O)-heptane-benzyl alcohol (c) R = 20 (H,O)-dodecane-toluene and (d) R = 20 (H,O)-dodecane-octanol. methanol and ethanol, showing an increase in T‘t with n,. There is then a reduction in Tk, with a minimum at n, = 4, followed by a further increase for hexanol and octanol. The measured values of rh for the R = 20 (H,O) microemulsion in heptane stabilised by 0.1 mol dm-3 AOT in the presence of 0.1 mol dm-3 alcohol are given in fig.3 (b). For small n, the alcohol is partitioned mainly into the aqueous core of the droplets, causing an effective increase in the ratio of volume to interfacial area and hence the droplet size. Increasing nc further increases the molar volume of the alcohol and hence the droplet size. Corresponding effects are observed on the phase behaviour. The increase in r h is larger than expected from consideration of the molar volume of alcohol alone. There may be desolvation of AOT from the less-polar droplet core, or there may be an increase in the attractive interactions between the droplets. The latter may be unlikely as 7‘: > 60 “C for the R = 20 (H,O) microemulsions in heptane in the presence of 0.1 mol dm-3 alcohol.As the alcohol chain length is increased, more alcohol will partition into the surfactant shell, both decreasing the droplet size and changing the nature of the interface.2416 Microemulsions stabilised by Aerosol-0 T Table 1. Upper phase-transition temperatures, TF, for R = 20 (2H20) microemulsions stabilised by 0.1 mol dm-3 AOT with various solvents and additives ~~~~~~~~ ~ ~~~ T;/"C additive concentration decane undecane dodecane 0 2.5 5 12.5 0 0.05 0.10 0.15 0 0.025 0.050 0.075 toluene (% v/v) - 49.0 k 1 .O - 67.0 & 1 .O > 80 - benzyl alcohol/mol dmP3 60.0 f 0.5 49.0 f 1 .O 59.5 0.5 47.0 & 1 .O 58.5 k0.5 42.5 k0.5 57.5 k0.5 31.0k0.5 octanol/mol dm-3 36.0+ 1 46.0 -t 1 .O 52.5 0.5 69.0+ 1 36.0+ 1 58.0+ 1 68.0 f 0.5 77.5 k0.5 80 60 40 20 0 1 2 3 4 5 6 7 6 n C / \ \ \ \ I I 1 1 I I 1 2 3 4 5 6 nC Fig.3. Effect of alcohol (C,,H,,,+,OH) on R = 20 (H20) 0.1 mol dm-3 AOT in alkane microemulsions. (a) The variation in upper and lower transition temperatures, T;' and Tk as a function of n, for 0.05 mol dm-3 alcohol in undecane. (b) The variation in the hydrodynamic radius, rh, with n, for 0.1 mol dmP3 alcohol in heptane at 35 "C.A . M . Howe et al. 241 7 5 5 5 - n 9 , " 4 D c - 3 2 ( A ) \ . '... f** --. I * . . . . ' , . I '* . . f ' 0 . * ** 0. ' . *. . . '* . . 0 . . . . . 0 . ' . . - * . . ' . ( b ) . .. (C) I 1 1 -. . . . 0. . . . . . ' ( a ) . . . . . . . . . . . ( b ) . . * * . . . (c) I 1 I 0 5 10 15 0 5 10 15 Fig.4. Guinier plots of the SANS intensity, I@), for various R = 20 (2H20) microemulsion systems stabilised by 0.1 mol dm-3 AOT in the presence of: ( A ) benzyl alcohol in undecane at 24 "C; (a) 0.05 mol dm-3, (6) 0.10 mol dm-3 and (c) 0.15 mol dmP3 benzyl alcohol; and (B) toluene in dodecane at 33 "C; (a) 2.5% ; (b) 5% and (c) (36 "C) 10% v/v toluene. SANS Measurements All measurements were made with a 2H20 core and an oil phase consisting of the normal hydrogenated material. Therefore the SANS intensity profile effectively characterises the inner-core radius rw ; the contrast profile for 2H20-AOT-n-alkane microemulsions has been described previously. l6 Hydrogen-containing additives can influence the shape of this profile in different ways, depending on their location.For additives which prefer- entially reside in the continuous oil phase, the effect on the shape and intensity is negligible because of the relatively small volumes of additives involved. In contrast, if a water-soluble additive is present in the inner core, there is little change in p, but the increased core volume will produce some change in the scattering intensity pattern. Of the additives which might be preferentially located at the interface, only benzyl alcohol has a p value significantly different from that of the surfactant chain. However, it can readily be shown that even if all the benzyl alcohol molecules were located at the interface, the effect on the contrast profile would be minimal for the volume fraction used in this study and such behaviour would have a much more significant effect on the droplet size.Therefore for the purpose of this study, it may be concluded that the presence of additives has little effect on the contrast profile of microemulsion system. Typical results for toluene as additive, with R = 20 2H20-AOT-dodecane and for benzyl alcohol as additive, with R = 20 2H20-AOT-undecane are shown in the form of a Guinier plot in fig. 4. At sufficiently low Q the plot is expected to be linear in the absence of significant inter-particle interactions. The upward curvature shown in the data clearly indicates the presence of a significant critical scattering component, the magnitude of which correlates directly with the expected variation due to the shift in 7'y as the additive concentration is increased.In the case of additives such as toluene, which induce241 8 Microemulsions stabilised by Aerosol-0 T 0 0.025 0.075 0.125 0.175 Q/A-' Fig. 5. Experimental points (x) and least-squares fit (solid curve) for the system R = 20 (2H20)-0.1 mol dm-3 AOT-dodecane with 5% v/v toluene at 33 "C. a positive shift in 7': with respect to the additive-free microemulsion, the deviation from linearity diminishes with increasing additive concentration. For benzyl alcohol, which gives a decrease in TF, there is a corresponding increase in the critical scattering component as a function of additive concentration. Thus the SANS results are in complete accordance with the phase-diagram studies and the evidence suggests that the data in the phase-transition region can be interpreted in terms of an essentially unchanged form factor plus critical scattering.17 It is possible to attempt a quantitative analysis using the relations given in the Theory section [eqn (1) and (2)].The SANS results are particularly important in this respect since the effects of critical scattering and droplet size are effectively separated in the observed I(Q) function, whereas the PCS data give an overall description characterised by an increase in a single parameter, which defines an effective correlation length for the critical density fluctuations. A least-squares fit to the experimental SANS data sets was made with the four variables r,, I(O), and K and a typical fit to the experimental data is shown in fig. 5 . The complete results are given in table 2.A constant, experimentally determined, calibration and background level was kept for comparative values for different additive concentrations. The absolute scattering intensities were not used as it was the relative effects of the presence of additives on the microemulsion that are of interest in this study. No attempt was made to introduce a polydispersity function and the value for rw should therefore be regarded as a weighted mean, 7,. Although polydispersity effects might be identified in the high-Q region of the measurements, this feature is not expected to have a major influence on the comparative values for 7, and has been omitted to simplify the analysis procedure. The values of I(0) and r, given in table 2 confirm the findings of our previous work,17 which indicate a slight but systematic decrease in both parameters as (7': - T ) is reduced.A .M . Howe et al. 2419 Table 2. The fitted parameter values obtained from the SANS data for AOT-stabilised micro- emulsions [R = 20 (2H,0), 0.1 mol dm-3 AOT] for various additives ~ amount of additive solvent 5 10 2.5 5 10 0.025 0.050 0.075 0.05 0.10 0.15 0.05 0.10 0.15 0.3 0.3 0.3 - -_ - - - Cll Cl 1 c12 Cl2 c 1 2 c, 2 c12 c12 CIO ClO ClO Cl 1 CI 1 Cll C,/R = 30 C,/R = 30 C,/R = 40 c,o* Cll* c12* C,/R = 30 C,/R = 40 4 0 ) %VIA K toluene (% v/v) 65+6 33f 1 1.6 f0.5 46+5 31+1 2.2 20.5 124f 15 32f0.5 7.2f 1.0 119f 15 32k0.5 3.2f0.5 121 & 15 32f0.5 1.5f0.5 102+ 10 31 f0.5 2.0k0.3 91f10 30f0.5 1.0f0.2 72f.8 29 f 0.5 1 .o & 0.2 86f10 31f0.5 1.3f0.2 7 3 f 8 30k0.5 2.6f0.3 4 3 f 5 26f0.5 4.7k0.5 55+6 30 + 0.5 3.4 f 0.3 37&5 26k0.5 11.4f 1.0 2 7 f 3 25f0.5 19+2 7 9 f 8 30 & 0.5 5.6 f 0.5 79+8 30 f 0.5 6.7 & 0.5 190 f 20 36 f 0.5 7.2 f 0.5 - 34+2 - - 33f.1 2.3 f0.4 - 33f 1 4.8 f 1 .O - 4 8 f 2 - - 59+4 - octanol/mol dm-3 benzyl alcohol/mol dmP3 no additive * From ref. (17). 5 15+5 18+5 96+_ 10 6 2 f 5 4 0 f 5 42f 5 20+5 12+3 36f.3 48+5 48+5 57&5 81 f 8 94f 8 48k5 60+ 5 7 0 f 6 ~ 5626 6 5 f 8 - - T/"C 21 38 33 33 36 33 31 31 21 21 21 24 24 24 21 38 21 25 21.5 21.5 21.5 21.5 This behaviour suggests that there is a decrease in the mean droplet size as the system approaches the upper-temperature transition point. The R = 20 2H,0-AOT-undecane microemulsion shows a much larger drop in P, for the addition of 0.15 mol dm-3 benzyl alcohol, from 33 to 25 A.Since I(0) is only dependent on the total volume of water and the number of droplets [eqn (3)], the combined reduction of P, and I(0) is indicative of an increased number of droplets and a correspondingly greater interfacial area. This suggests that the benzyl alcohol plays an active role in the interface and presumably shows partitioning between the water and the surfactant regions. It is possible that at lower temperatures not all of the AOT surfactant is bound at the interface and there is some evidencez6 that some AOT is present in the form of almost-dry reversed micelles which coexist with water-containing droplets. The results for toluene show comparatively little dependence of Pw on additive concentration, indicating that partitioning of toluene is much less pronounced.The parameter values obtained from a fit to the critical scattering contribution are also instructive, but it is difficult to make an assessment of the critical constants v and y for these microemulsions as the appropriate scaling laws apply only to variation of the correlation length, <, along a properly defined critical path to the critical temperature,2420 Microemulsions stabilised by Aerosol-0 T &. For a given temperature there is a unique critical composition.li Therefore, although quantitative data can be extracted from the observations, the system is not sufficiently well characterised to permit an evaluation of the critical scaling laws. Discussion Microemulsions exhibit maximum solubilisation, and hence stability, at the 'phase inversion temperature' (the temperature at which the surfactant exhibits equal affinity for oil and water) for a water-non-ionic surfactant-oil system.2i Shinoda has reported that this is also the case for AOT-stabilised microemu1sions28* eg and that the trend in phase behaviour with temperature for oil-water systems stabilised by ionic surfactants, such as AOT, is the opposite of those stabilised by non-ionic surfactants.2i, 30.31 The systems also respond to pressure in different ways (P. D. 1. Fletcher and B. H. Robinson, unpublished results). There are various interpretations of the phase behaviour of surfactant-stabilised systems. As the temperature is raised, the effective hydrophilic- lipophilic balance of the AOT is changed2' and the spontaneous curvature of the surfactant shell becomes increasingly convex to the aqueous phase.:" Thus at high temperatures the relative solubility of AOT in the oil and water domains changes and in practice the miscibility gap between the AOT and water is decreased, while that between AOT and oil is increased.33 The lower-temperature phase transition is essentially from separate water and AOT-oil (microemulsion) phases to a single-phase, droplet-containing microemulsion.The size of the droplets in the separated microemulsion phase is dependent on the temperature of the two-phase system (fig. 1) as has been reported in studies of the phase separation of this system on addition of The transition is thought to be entropically driven by the formation of a semi-diffuse electrical double-layer within the droplet.35 The upper-temperature phase transition, which is a transition from a single-droplet phase to a concentrated droplet or liquid-crystalline (lower) phase coexisting with a dilute droplet (upper) phase, is driven by attractive interactions between the drop1ets.j.l i , 36-38 This transition may be associated with desolvation of oil from the alkyl tails of the s ~ r f a c t a n t . ~ ~ ~ 40 The attractive interactions between droplets are reported to increase with droplet size,8! 41 thus large droplet systems (higher R) are less stable with increase in temperature (fig. 1). The behaviour of the short-chain alcohols (methanol to propanol) is consistent in both phase behaviour and droplet structure with the partitioning of the alcohol into the aqueous core of the droplet, and the magnitude of the observed effects on droplet size reflect, in part, the different molar volumes of the additives.Longer-chain alkanols partition mainly into the surfactant shell and this has three effects: (i) the droplet size decreases as a result of the increase in interfacial area; (ii) the spontaneous curvature of the surfactant shell is also more convex to the oil (alkanols are probably less hydrophilic than AOT); and (iii) the distance of closest approach of the droplets is also likely to be increased when the alcohol tail is longer than that of AOT, which may be expected to reduce attractive interactions between the droplets. Thus, on addition of long-chain alkanols, there is a decrease in the size of the droplet core and an increase in the temperature range over which the single droplet-phase is stable (fig.3). Addition of benzyl alcohol is found to decrease the droplet size and, at a given temperature, to increase the critical scattering observed at low Q. Therefore benzyl alcohol is partitioned into the surfactant shell region and its effect is to lower the upper temperature at which the single phase microemulsion system is stable. Benzyl alcohol is relatively soluble in water, less soluble in alkanes, but very soluble in AOT-alkane mixtures, while butanol is only sparingly soluble in water.42 Thus in the presence of benzyl alcohol the spontaneous curvature of the surfactant shell is relatively less convex to the oil than in the absence of additive, in contrast to the behaviour on addition of long-chain alkanols.A .M . Howe et al. 242 1 Toluene has little effect on the droplet size, but it causes a shift of the phase diagram to higher temperature, and a corresponding decrease in the critical scattering. Toluene is expected to be partitioned mainly into the bulk oil phase. We previously reported a decrease in the temperature at which the single droplet-phase is formed in microemulsions stabilised by AOT with an increase in alkane chain 1ength.l' In the presence of toluene the phase diagram is shifted to much higher temperatures. Oils of small, polar or polarisable molecules have been reported to lower the phase-inversion temperature for water-non-ionic surfactant-oil mixture^.^' The phase behaviour of surfactants in different oils may be interpreted in terms of increased penetration by oils of small, polar and polarisable molecules between the hydrophobic tails of the surfactant molecules which makes the surfactant shell more convex to the oil."2 For the surfactant AOT, a 13C n.m.r.study by Martin and Magid43 indicated that there is more penetration between the surfactant molecules by benzene than cyclohexane and a proton n.m.r. study by Maitra et suggests there is an interaction between benzene (but not iso-octane) and the head group of the AOT. Thus toluene may 'solvate' the AOT molecules more effectively thaE alkanes, but there is insufficient toluene in the surfactant shell to significantly change the size of the droplet core.Conclusion Since microemulsions formed in the presence of AOT do not need a cosurfactant for stability they are ideal systems for a systematic study of the effect of a fourth component. By means of such additives the stable region for a single-phase microemulsion can be shifted to higher or lower temperature, depending on the nature of the additive. Hence spherical-droplet microemulsions can be studied over an extended temperature range for a given alkane solvent. SANS is a useful technique to investigate the influence of additives, both on droplet structure and phase instability (critical phenomena). Parti- tioning of the additives between the water core, the surfactant shell and the bulk oil may be investigated. The relative amphiphilic properties of a range of additives have been demonstrated.The SANS work was carried out with the support of the Neutron Beams Research Committee of the S.E.R.C. The S.E.R.C. is acknowledged for a fellowship (C.T.) and a studentship (A.M.H.). We are grateful to Andrea Fraser for help with the PCS experiments, and to Dr J. Mead for the interfacial tension measurements. We thank the S.E.R.C. (Biotechnology) for a grant (to B. H. R.) to purchase the photon correlation spectrometer. References 1 L. M. Prince, Microemulsions (Academic Press, London, 1977). 2 Microernufsions, ed. 1. D. Robb (Plenum Press, New York, 1982). 3 A. M. Bellocq, J. Biais, P. Botherel, B. Clin, G. Fourcht, P. Lalanne, B. Lemaire, B. Lemanceau and 4 J. Jouffroy, P. Levinson and P. G. de Gennes, J.Phys. (Paris), 1982, 43, 1241. 5 M. J. Grimson and F. Honary, Phys. Lett. A , 1984, 102, 141. 6 B. Widom, J . Chem. Phys., 1984, 81, 1030. 7 A. M. Cazabat, D. Chatenay, D. Langevin and J . Meunier, Furadu.y Discuss. Chem. Soc., 1982,76, 29 1 . 8 A. M. Cazabat and D. Langevin, Jr Chem. Phys., 1981, 74, 3148. 9 D. 0. Shah, Surface Phenornenu in Enhanced Oil Recovery (Plenum Press, New York, 1981). D. Roux, Adv. Colloid Interface Sci., 1984, 20, 167. 10 I. Willner, J. E. Ford, J. W. Otvos and M. Calvin, Nature (London), 1979, 280, 823. 1 1 R. Hilhorst, C. Laane and C. Veeger, Proc. Natl Acad. Sci. USA, 1982, 79, 3927. 12 S. Barbaric and P. L. Luisi, J . Am. Chem. Soc., 1981, 103, 4239. 13 K. Martinek, A. V. Levashov, N. L. Klyachko, V. I . Pantin and I . V. Berezin, Binchim.Biophys. Acta, 198 1, 657, 277.2422 Microemulsions stabilised by Aerosol-0 T 14 P. D. I. Fletcher, R. B. Freedman, B. H. Robinson and R. Schomacker, Biochim. Biophys. Acta, in press. 15 P. D. I. Fletcher, B. H. Robinson, F. Bermejo-Barrera, D. G. Oakenfull, J. C. Dore and D. C. Steytler in Microemulsions, ed. I. D. Robb (Plenum Press, New York, 1982), p. 221. 16 B. H. Robinson, C. Toprakcioglu, J. C. Dore and P. Chieux, J. Chem. SOC., Faraday Trans. 1, 1984, 80, 13. 17 C. Toprakcioglu, J. C. Dore, B. H. Robinson and A. M. Howe, J. Chem. Soc., Faraday Trans. I , 1984, 80,413. 18 P. D. 1. Fletcher, N. M. Perrins, B. H. Robinson and C. Toprakcioglu in Biological and Technological Relevance of Reverse Micelles and other Amphiphilic Structures in Apolar Media, ed.P. L. Luisi (Plenum Press, New York, 1984). 19 J. A. McDonald, A. M. Howe and B. H. Robinson, in preparation. 20 D. I. Page, Harwell Technical Report AERE-R9878, 1980. 21 B. Maier, Neutron Research Facilities at the High Flux Reactor (Institut Laue-Langevin, Grenoble, 22 R. E. Ghosh, ILL Report No. 78 GH 247T (Institut Laue-Langevin, Grenoble, 1978). 23 P. D. I. Fletcher, M. F. Gala1 and B. H. Robinson, J. Chem. SOC., Faraday Trans. I , 1984, 80, 3307. 24 L. S. Ornstein and F. Zernike, Proc. Acad. Sci. Amsterdam, 1914, 17, 793. 25 L. S. Ornstein and F. Zernike, 2. Phys., 1918, 19, 134; 1926,27, 761. 26 A. N. North, J. C. Dore, J. A. McDonald, B. H. Robinson, R. K. Heenan and A. M. Howe, Colloid 27 K. Shinoda and H. Kunieda in Encyclopaedia of Emulsion Technology, ed. P. Bccher (Marcel Dekker, 28 H. Kunieda and K. Shinoda, J. Colloid Interface Sci., 1979, 70, 577. 29 H. Kunieda and K. Shinoda, J. Colloid Interface Sci., 1980, 75, 601, 30 R. N. Healey, R. L. Wade and G. Stenmark, SOC. Petr. Eng. AIME SPE5565, 1975. 31 M. Bourrel, J. L. Salager, R. S. Schechter and W. H. Wade, J. Colloid Interface Sci., 1980, 75, 451. 32 D. J. Mitchell and B. W. Ninham, J . Chem. SOC., Faraday Trans. 2, 1981, 77, 601. 33 C. U. Herrmann, G. Klar and M. Kahlweit in Microemulsions, ed. 1. D. Robb (Plenum Press, New 34 R. Aveyard, B. P. Binks, S. Clark and J. Mead, J. Chem. Soc., Faraday Trans. I, 1986, 82, 125. 35 H. F. Eicke and R. Kubik, Faraday Discuss. Chem. Soc., 1983, 76, 305. 36 M. Kotlarchyk, S. H. Chen and J. S. Huang, Phys. Rev. A , 1983, 28, 508. 37 M. Kotlarchyk, S. H. Chen, J. S. Huang and M. W. Kim, Phys. Rev. A, 1984, 29, 2054. 38 S. A. Safran and L. A. Turkevich, Phys. Rev. Lett., 1983, 50, 1930. 39 B. Lemaire, P. Botherel and D. Roux, J. Phys. Chem., 1983, 87, 1023. 40 J. S. Huang, J. Chem. Phys., 1985, 82, 480. 41 S. Brunetti, D. Roux, A. M. Bellocq, F. G. Fourche and P. Botherel, J . Phys. Chem., 1983, 87, 1028. 42 C.R.C. Handbook of Chemistry and Physics (C.R.C. Press, Boca Raton, Florida, 56th edn, 1975). 43 C. A. Martin and L. J. Magid, J . Phys. Chem., 1981, 85, 3938. 44 A. N. Maitra, G. Vasta and H. F. Eicke, J. Colloid Interface Sci., 1983, 93, 383. 1983). SurJ, in press. New York, 1983), p. 337. York, 1982), p. 1. Paper 5/ 157 1 ; Received 12th September, 1985
ISSN:0300-9599
DOI:10.1039/F19868202411
出版商:RSC
年代:1986
数据来源: RSC
|
17. |
Correlation between hydrodesulphurization activity and reducibility of unsupported MoS2-based catalysts promoted by group VIII metals |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 82,
Issue 8,
1986,
Page 2423-2434
Sandor Göbölös,
Preview
|
PDF (947KB)
|
|
摘要:
J. Chem. SOC., Faraday Trans. I, 1986, 82, 2423-2434 Correlation between Hydrodesulphurization Activity and Reducibility of Unsupported MoS,-based Catalysts promoted by Group VIII Metals Sandor GobolosJ Qin WuJ Olivier Andre', Francis Delannay and Bernard Delmon* UniuersitP Catholique de Louuain, Groupe de Physico-Chimie MinPrale et de Catalyse, Place Croix du Sud I , 1348 Louuain-la-Neuve, Belgium Unsupported FeMo, CoMo and NiMo sulphide hydrodesulphurization (HDS) catalysts have been prepared by one or several different methods: homogeneous sulphide precipitation (HSP), inverse HSP (IHSP), comacer- ation (CM) and coprecipitation (CP). They have been characterised by (i) differential thermal analysis (DTA) during temperature-programmed reduc- tion/sulphidation (TPR/S) and subsequent air oxidation (TPO), (ii) temperature-programmed sulphur extraction (TPSE), (iii) XPS and (iv) XRD.For catalysts prepared by the HSP, IHSP and CP methods the onset temperature of TPR/S decreases, and the threshold temperature of oxidation, the amount of released H,S in TPSE and the HDS activity all increase in the sequence Mo < FeMo < CoMo < NiMo. XPS reveals a change in the Co 2p,,, and Ni 2p,,, binding energies for promoted catalysts with respect to pure Cogs, and NiS, respectively. The results demonstrate a strong correlation between catalytic activities in HDS of thiophene and hydrogen- ation of cyclohexene on the one hand, and the reducibility of the catalysts on the other. The industrial importance of hydrosulphurization (HDS) and hydrotreating and the vigorous expansion of these processes are encouraging numerous attempts to correlate the HDS activity of both supported and unsupported catalysts with their physicochemical properties.Characterisation of catalysts by Mossbauer spectroscopy,' 0, chemisorption,2 NO chemisorption combined with i.r. ~pectroscopy,~ e.s.r. ~pectroscopy,~ temperature- programmed reduction, temperature-programmed hydrogen de~orption,~ conductivity measurements6 and magnetic susceptibility measurements' all provide interesting corre- lations with the catalytic activity. The catalytic activity was also reported to be related to the number of sulphydryl groups8 and amount of labile sulphur.5* 9+11 The experimental material available for this type of study is extremely rich, in the sense that samples differing by many variables have been prepared and studied. These variables are (i) the methods of preparation; (ii) presence or absence of carrier and its nature; (iii) composition, defined by r, the atomic ratio of the Group VIII metal to the Group VIII plus Group VI metals; and (iv) the nature of the Group VIII (Co, Ni or Fe) and Group VI (Mo, W) metals.It is surprising that relatively little use has been made of the possibilities offered by series of catalysts containing different Group VIII metal promoters. The similarity in characteristics and catalytic activities of the corresponding catalysts is considerable. A criterion for a good correlation between HDS (or hydrogenation, HYD) activity and Budapest, Pusztaszeri ut 59-61, Hungary.P.B. 2653, People's Republic of China. 7 On leave from the Central Research Institute for Chemistry of the Hungarian Academy of Sciences, 1025 $ On leave from the Beijing Municipal Chemical Industry Research Institute, Cheng Fu, Haidian, Beijing, 24232424 Activity of MoS,-based Catalysts physicochemical properties is that it should hold for all six systems comprising one Group VI metal with one Group VIII metal. The present study is an attempt in this direction. A comparison is made between unsupported FeMo, CoMo and NiMo HDS catalysts. We shall study the correlations between (i) the reduction/sulphidation of catalyst precursors under an H,S-H, mixture, (ii) the oxidation of the catalysts under air, (iii) the reduction (sulphur extraction) of the catalysts under H, and (iv) the HDS activity.XPS and XRD results will also be included. The role of sulphur in unsupported HDS catalysts will be discussed. Experimental Catalyst Preparation Precursors The chemicals used for the preparations were Merck products (pro analysi). The various stages of preparation (precipitation, evaporation and drying) were carried out under an argon atmosphere at 343 K. The unsupported sulphide catalyst precursors were prepared by four different methods. All methods involve the mixing of chosen proportions of compounds of molybdenum and Group VIII metal (M) in a solution of (NH,),S. The S/(M +4Mo) atomic ratio in the mixture was > 1.5 times that needed for the complete sulphidation of M and Mo into MS and MoSi-. These methods were as follows : (a) the homogeneous sulphide precipitation (HSP) method:12 a mixed solution of the Group VIII metal nitrate and ammonium heptamolybdate (AHM) was added to a solution of (NH4),S; (b) the inverse HSP (IHSP) method:', a solution of (NH,),S was added to a mixed solution of the Group VIII metal nitrate and AHM; ( c ) the comaceration (CM) method:'* powdered MOO, and the Group VIII metal oxide (e.g.Co,O,) were allowed to react with a solution of (NH,),S; (d) the coprecipitation (CP) mefhod:l5 a mixture of MS and MoS, (or mixed sulphides) was coprecipitated by adding a solution of the Group VIII metal nitrate into a previously prepared solution of ammonium thiomolybdate (ATM). Details of the experimental procedures have been described elsewhere. l5 Unsupported CoMo sulphide precursors with atomic ratios r = Co/(Co+Mo) = 0.1 and 0.3 were prepared by the HSP, IHSP and CM methods. Unsupported NiMo sulphide catalyst precursors with r = 0.1 and 0.3 were prepared by the HSP method.Unsupported FeMo sulphide catalyst precursors with r = 0.15 and 0.3 were prepared by the CP method using iron(II1) nitrate. In order to guarantee the precipitation of iron sulphide by the sulphur of ATM and thus to provide an intimate contact between iron and molybdenum, only a 3% excess of sulphur was used in this case, i.e. S/(Fe+4Mo) = 1.03 instead of 1.50. A precursor of pure molybdenum disulphide (r = 0) was also prepared by the HSP method. Reduct ion/ Sulphida t ion Treatment The active catalysts were prepared from the above precursors by a reduction/sulphidation treatment.The precursors were heated to 673 K at a rate of 0.33 K s-l under a 1.1 cm3 s-l flow of 15% H,S-H, at atmospheric pressure. After 4 h at 673 K, the samples were cooled to room temperature under a flow of argon. The catalysts will be designated in terms of the metal(s) present, followed by the method of preparation and the composition ratio r (e.g. Mo-HSP-0.0, FeMo-CP-0.15, CoMo-CM-0.3). Catalytic Activity Measurement The activity in HDS of thiophene and hydrogenation (HYD) of cyclohexene was measured on 0.5 g of catalyst (grain size 0.1-0.2 mm) under 3 MPa, at 573 K using aS . Gobolos, Q. Wu, 0. Andrk, I;. Delannay and B. Delmon 2425 feed mixture containing 69.5 wt % cyclohexane, 30.0 wt % cyclohexene and 0.5% thiophene.The mass flow rate of the liquid feed was 48 g h-l. As exposure to air could not be avoided during the loading of the reactor, samples were treated in situ for 1 h under 3 MPa, 4% H2S-H, at 573 K prior to activity tests. Samples of the product were analysed every half-hour for 8 h. Steady-state activities were always obtained after 3-4 h. The intrinsic activities are expressed as follows : mol mP2 s-l wc xc HYD=- Mc mSa where W, and W, are the mass flow rates (g s-l), MT and M , are the molecular weights, XT and X c are the conversions (at steady state after 8 h on stream) for thiophene and cyclohexene, respectively, Sa (m2 g-l) is the surface area of the used catalyst and m is the weight of the catalyst. Other details on activity measurements are described elsewhere fi Characterization Surface Area Measurement B.E.T. surface areas of fresh and used catalysts (after 8 h on stream) were determined by gravimetric measurement of nitrogen adsorption.Temperature-programmed Reactions Differential thermal analysis (DTA) of temperature-programmed reduction/sulphidation (TPR/S) of the dry precursors under a 15% H,S-H, mixture, and of the subsequent temperature-programmed oxidation (TPO) of the sulphided catalysts under air were performed under conditions identical to those described elsewhere.” Temperature-programmed sulphur extraction (TPSE) experiments were carried out at atmospheric pressure in a flow reactor.l8 The catalyst samples (0.2 g) were first resulphided in situ under 15% H,S-H, at 673 K for 2 h. After purging with argon at the same temperature for 1 h and cooling to room temperature, they were heated at a rate of 0.17 K s-l to 773 K in a flow of 0.5 cm3 s-l of pure H,.The H2S produced was measured by a thermal-conductivity detector. XPS After the activity test, samples were collected and stored in iso-octane, and pressed, also under a film of iso-octane, into the cupules used for XPS measurements. This procedure enabled to protect the samples from contact with air. The cupules covered with a meniscus of the solvent were then introduced into the vacuum chamber of the spectrometer where the solvent was evacuated. The effectiveness of this procedure in protecting sulphide catalysts from oxidation has already been proved. l6 The binding energies (Eb) were referred to the contaminant carbon peak (C 1s = 285.0 eV).Differences in Eb between levels Mo 3d5/, and S 2p [AE,(Mo-S)] and M 2p,/, and S 2p [AEb(M-S); M = Fe, Co or Nil were also calculated. S 2p/Mo 3d and M 2p/Mo 3d intensity ratios were converted into atomic ratios S,/Mo (t = total) and M/Mo using sensitivity factors proposed in the literature.ls2426 Activity of MoS,-based Catalysts Table 1. B.E.T. surface area, XRD and catalytic activity results intrinsic activityb / lo-* mol m-* s-l surface area/m2 g-l -~ ha catalyst fresh used /nm HDS HYD HYD/HDS M 0-HSP-0.0 FeMo-CP-0.15 COMO-CM-0. 1 COMO-IHSP-0. 1 COMO-HSP-0. 1 NiMo-HSP-0.1 FeMo-CP-0.3 COMO-CM-0.3 COMO-IHSP-0.3 COMO-HSP-0.3 NiMo-HSP-0.3 102.4 59.2 45.8 22.5 35.8 24.1 40.4 26.0 13.7 20.2 - 26.7 35.9 12.8 3.5 8.3 7.7 27.4 6.9 14.4 9.3 16.6 5.8 7.7 - - 7.6 6.6 5.5 5.2 5.5 0.82 1.78 1.96 5.39 4.63 2.94 2.34 4.71 8.41 7.62 13.6 19.2 24.1 23.8 37.7 29.3 52.5 31.5 37.7 76.8 57.4 59 1 23.4 13.5 12.1 7.0 6.5 17.9 13.5 8.1 9.1 7.5 43.6 a Average crystallite sizes estimated by XRD.Steady state activities after 8 h on stream. XRD X-Ray diffractograms were recorded using Cu Ka radiation and a monochromator to suppress the fluorescence from cobalt. The average crystal sizes were evaluated from the broadening of the (002) (h) lines for MoS,-containing catalysts using a k value of 1.077 in the Scherrer equation and assuming that the crystallites were free from strain. Results Catalytic Activities and B.E.T. Surface Areas B.E.T. surface areas of fresh and used catalysts and intrinsic activities are given in table 1.Surface areas always decrease during catalytic testing, especially in the case of catalysts with r = 0.1. HDS and HYD activities increase by the addition of Fe, Co or Ni. This increase in activity is always higher for catalysts with r = 0.3 than with r = 0.1 or 0.15. The promoting effect of various Group VIII metals for both HDS and HYD increases as follows : FeMo < CoMo < NiMo. The only exception is catalyst NiMo-HSP-0.1, which exhibits a lower HDS activity than catalyst CoMo-HSP-0.1. The HDS activity of catalysts CoMo-HSP-0.1 and CoMo-HSP-0.3 is always higher, by a factor of 2, than that of catalysts CoMo-CM-0.1 and CoMo-CM-0.3. The HYD/HDS ratio (selectivity) increases in the following order: CoMo < FeMo < MoS, < NiMo (the only exception is again catalyst NiMo-HSP-0.1).Physicochemical Characterization Temperat we-programmed React ions Fig. 1 presents the DTA curves for TPR/S and TPO of samples Mo-HSP-0.0 and NiMo-HSP-0.3. The onset temperatures of reduction/sulphidation ( TR,s) and oxidation (To) (T, and &, respectively, as in previous papers)l7? l9 are very much affected by the presence of the Group VIII metal. For more details the reader is referred to our previous papers. Table 2 summarizes the values of TR,s and To for all samples.S. Gobolos, Q. Wu, 0. Andre!, I;. Delannay and B. Delmon 2427 'R / S I I I I I I 3 00 500 700 TI K Fig. 1. DTA curves of temperature-programmed reduction/sulphidation of precursors and subsequent oxidation of Mo-HSP-0.0 and NiMo-HSP-0.3 catalysts.Table 2. Temperature-programmed reductiona and XPSb results ~ ~ ~ ~ ~ ~ TE,s To TPSE area Eb(s 2p) AEb(M-S)" catalyst /K /K (arb. units) St/Moc M / M O ~ . ~ /eV /ev -~ ~~~ ~ ~~ Mo-HSP-0.0 FeMo-CP-0.15 COMO-CM-0. 1 COMO-IHSP-0. 1 COMO-HSP-0. 1 NiMo-HSP-0.1 FeMo-CP-0.3 COMO-CM-0.3 COMO-IHSP-0. 3 COMO-HSP-0.3 NiMo-HSP-0.3 59 1 543 577 515 518 528 533 57 1 51 1 503 505 600 618 598 625 618 642 608 603 625 630 698 1.5 - - 5.0 5.0 8.0 15.0 - 1.93 1.96 - 1.91 1.95 2.17 1.89 2.1 1 2.04 2.09 - 0.22 0.08 0.13 0.10 0.52 0.16 0.37 0.43 0.48 162.8 162.9 162.9 162.5 162.4 162.9 162.6 162.5 162.2 162.3 - 545.9 616.6 616.8 692.4 545.9 61 6.9 617.0 617.1 692.6 a From ref. (17) (19). The precision of such quantitative XPS ratio measurements is & 5 % . The M/Mo atomic ratios calculated from the chemical compositions are 0.11, 0.18 and 0.43 for catalysts with Y = 0.1, 0.15 and 0.3, respectively." AE,(M-s) for CO,S,,~~ NiS20 and Fe,-, S21 are 616.1, 691 .O and 545.8 eV, respectively. Measurements on used catalysts. The threshold temperature T,,, of reduction/sulphidation of molybdenum-containing catalyst precursors is shifted towards lower temperatures in the presence of promoters. This effect is especially important for CoMo and NiMo catalysts with Y = 0.3, and for catalysts prepared by the HSP and IHSP methods. In contrast, less influence on the onset2428 Activity of MoS,-based Cutalysts 1 1 I I 1 I 400 600 800 TIK Fig. 2. Temperature-programmed sulphur extraction curves of catalysts referred to unit surface areas.temperature of reduction/sulphidation is observed in the case of Fe addition and of the CM method of preparation. The onset temperature of oxidation, To, increases as follows : Mo-HSP-0.0 z CoMo- CM-r < FeMo-CP-r < CoMo-HSP(1HSP)-r < NiMo-HSP-r. A change in the concen- tration of promoter markedly affects oxidation only in the case of Ni-containing catalysts. The results of temperature-programmed sulphur extraction measurements are shown in fig. 2. The amount of released H2S increases in the following order: Mo-HSP- 0.0 < CoMo-CM-0.3 z FeMo-CP-0.3 < CoMo-HSP-0.3 z NiMo-HSP-0.3. The main characteristics of the spectra are twofold : a small peak appears between 450 and 550 K for catalysts Mo-HSP-0.0, CoMo-CM-0.3 and FeMo-CP-0.3, and a large one appears with a maximum around 700 K for catalysts CoMo-HSP-0.3 and NiMo-HSP-0.3.The specific TPSE spectral areas are also listed in table 2. The S,/Mo and M/Mo atomic ratios, the binding energy of the S 2p level and the AEb(M-s) binding-energy differences are also reported in table 2. The S,/Mo ratio is slightly higher for catalysts with r = 0.3 than with r = 0.1. The M/Mo ratios are close to the bulk values for catalysts prepared by the HSP(1HSP) and CP methods. The measured Co/Mo ratio is significantly lower than the bulk ratio for catalyst The AE,,(Mo-S) values of 66.9-67.0 eV (not listed in table 2) observed for all catalysts were equal to that of pure M0S2.16 AEb(Ni-s) and AEb(co-s) values are significantly higher (by 1.4-1.6 and 0.5-1 .O eV) than those of pure NiSZ0 and Co,S,,16 respectively.No change in AEb(Fe-s) between FeMo catalysts and Fe,-, SZ1 was observed. COMO-CM-0.3.S. Gobolos, Q. Wu, 0. Andre!, F. Delunnay and B. Delmon 2429 XRD Average crystallite sizes of MoS, in the catalysts calculated from XRD results are also listed in table 1. Crystallite sizes of MoS, seem to be significantly higher in catalysts CoMo-CM-0.3 and FeMo-CP than in other samples. Discussion Textural Properties The crystallite sizes listed in table 1 correspond to higher surface areas than actually measured by B.E.T. techniques. This is due to the fact that MoS, presents a turbostratic structure22 where the particle size is larger than the domain size, with the consequence that a large proportion of the ‘domain’ boundaries is not accessible to the N, molecules in the B.E.T.measurement. The average crystallite size of MoS, decreases only slightly in the presence of Co or Ni in catalysts prepared by the HSP and IHSP methods. Candia et al.12 have found a more significant decrease in the particle size of MoS, by the addition of Co in unsupported catalysts prepared by the HSP method, We feel that this contradiction could be due to a difference in the details of the preparation of the samples. Both cobalt, in CMI4 and CP19 samples, and iron, in CP samples,1g increase the crystallite size of MoS,. This suggests that Group VIII metal sulphides decrease the disorder normally observed in the MoS, structure. Delannay et al.23 and Thakur et al.24 also reported that the addition of a small amount of Group VIII metal (r < 0.02) brings about a marked increase in crystallinity of MoS, in unsupported catalysts prepared by the CM method.They assumed that a small amount of promoter facilitates the growth of MoS, ~rystallites.~~* 24 However, the present results suggest that the same phenomena can occur at even higher promoter concentrations (r = 0.3) in the CMl49 2 3 9 24 and CPl5* l9 catalysts. However, the influence of promoter dispersion on the crystallinity of MoS, still needs further elucidation. Surface-area losses during catalytic tests are probably due to both crystal growth, favoured by a highly reducing atmo~phere,,~ and some blocking of the micropores by coke deposition.26 The higher the initial surface area and molybdenum content, the greater the loss.16 Catalytic Activity It is worthwhile to note the (low) promotion of both HDS and HYD activity of MoS, by iron in catalysts prepared by the CP method. This is in agreement with the results of Thakur et ~ l ., , ~ who have first reported a synergy between Fe and Mo in unsupported HDS catalysts prepared by the CM method. As already reported, the HDS activity of catalysts CoMo-HSP-r is higher than that of catalysts CoMo-CM-r.12 There is no significant difference in the activity of catalysts prepared by the HSP and IHSP methods, probably owing to the similarities of their structure and texture. The promoting effect of various Group VIII metals for both HDS and HYD (Fe < Co < Ni) changes in a similar way for HSP or CP and CM24 catalysts. The behaviour of catalyst NiMo-HSP-0.1 is not very surprising, since no or only slight promotion of HDS and HYD at low promoter concentrations ( r z 0.1) has been reported for both supported21 and unsupportedll? 24 NiMo sulphide catalysts.The HYD/HDS ratios confirmed again that CoMo sulphide or NiMo sulphide catalysts are the best choices when, respectively, HDS or HYD activity should be favoured.2430 Activity of MoS,-based Catalysts Physicochemical Characterization Before discussing in detail the results of physicochemical characterizations, it is worth recalling our present knowledge about the location of promoter atoms in unsupported MoS,-based HDS catalysts prepared by the different methods used in this study. The catalysts prepared by the CM method are essentially biphasic, containing separate MoS, and the thermodynamically stable MS, (e.g.Cogs8, Ni,S,, Felpz S) There is much similarityz8> 29 between the HSPl2 and IHSP13 samples. Topsrae et ~ 1 . ~ 9 30 have shown using Mossbauer spectroscopy and XRD that, in unsupported CoMo sulphide catalysts prepared by the HSP method,12 Co is present in two forms: Cogs, and a MoS,-like phase which was termed CO-MO-S.~*~O Very recently, infrared spectroscopy3~ 31 and analytical electron 32 have proved that, in the Co-Mo-S structure, Co atoms are located on the edges of MoS, slabs, most likely in substitutional or interstitial positions. Therefore, Co atoms cover, at least partly, the Mo sites on the edges of MoS, cry~tallites~~. 32 and modify their catalytic33 and physicochemical properties,l* 34 including, presumably, the metal-sulphur bonding.35 The existence of F~-Mo-S~~ and N~--Mo--S~~ structures in HDS catalysts was also established.In addition to FeS and Fel-,S, Fe-Mo-S was also detected by Mossbauer spectroscopy in unsupported catalysts prepared by the CP method.31 phases.14, 15, 23, 24, 27 Tempera t we-programmed React ions A detailed discussion of the TPR/S and TPO results is presented in ref. (17) and (19). Note that the exothermic effects, observed in the TPR/S experiments between 500 and 700 K, are due to the transformation of MoS, and/or Mo oxysulphides into MoS,, accompanied by the release of elemental sulphur and the crystallisation of MoS,.~~ The highly dispersed M-S species catalyse the reduction/sulphidation of Mo-O,S, species in catalyst precursors prepared by the HSP, IHSP and CP methods.Co and Ni sulphides are particularly efficient in promoting the reduction of Mo-containing species.lg In the TPO experiments a threshold temperature of 600 K is observed for the oxidation of pure MoS, under these ~0nditions.l~ Since the oxidation of MoS, proceeds from the edges,39 the lower reactivity of MoS, towards oxygen in the presence of Ni and Co is probably due to the covering of the edges of MoS, crystallites by promoter 31. 32 The origin of this reduced affinity towards oxygen is difficult to ascertain as long as the mechanism of this oxidation remains to be elucidated. One may speculate that the onset of oxidation corresponds to oxidation of the topmost atom exposed on the edge. In the case of pure MoS,, this atom is sulphur bound to molybdenum, which oxidizes into SO,.In the case of promoted catalysts, the first exposed sulphur atom may desorb before being oxidised into SO, (owing to a lower bond strength of M-S than Mo-S)~~ and the oxidation might thus be controlled by the reactivity of the underlying Me atom towards oxidation into MOO. Another explanation is that, in the presence of promoter, surface oxidation takes place at low temperatures and the protecting oxysulphide layer thus formed on the edges of MoS, prevents the bulk oxidation of MoS, (which gives rise to the exothermic effect detected by DTA) up to temperatures >600 K. Peuski et ~ 1 . ~ ~ observed the same phenomena for the oxidation of Cogs,. The TPSE experiments show that the amount of H,S released up to 773 K increases in the presence of highly dispersed Co and Ni species on the edges of MoS, crystallites.This suggests that the strength of binding of S on the edges of MoS, slabs is weaker in the promoted catalysts than in pure MoS,.~~S. Gobolos, Q. Wu, 0. AndrP, F. Delannay and B. Delmon 243 1 t 1 I I 500 550 6 00 TR/S IK Fig. 3. Correlation between HDS activity and the onset temperature of reduction/sulphidation of catalyst precursors: x , Mo-HSP-0.0; 0, 0, FeMo; e, D, CoMo; @, 0 , NiMo catalysts with Y = 0.1 or 0.15 (circles) and 0.3 (squares), respectively. XPS One can distinguish schematically four possible locations of the promoter atoms with respect to MoS,: (a) the homogeneous dispersion of promoter and molybdenum sulphides, (b) the segregation of promoter sulphide in the form of a layer covering the surface of the MoS, crystallites, (c) the presence of separate promoter and molybdenum sulphide particles and ( d ) the segregation of MoS, covering the surface of promoter sulphide particles.The XPS intensity data do not provide much insight in enabling us to differentiate between these models. Indeed, the escape length for the 2p electrons can be estimated to be 0.8-1.1 nm using the equation of Szajman and Le~key.~l As shown in table 1, this value is fairly large compared with the size of the elementary catalyst particles. For catalysts prepared by the HSP, IHSP and CP methods, the M/Mo ratios estimated from XPS peak intensities are close to the bulk composition ratios. The only safe conclusion that can be drawn is that complete segregation of either MoS, or MS in the centres of particles is unlikely.In catalysts CoMo-CM-0.1 and CoMo-CM-0.3 the presence of separate phases is established by the appearance of diffraction lines of Cogs, in the XRD spectra. The S 2p peak in the XP spectra is mostly due to sulphur associated with molybdenum in MoS,. This is confirmed by the constant value of AE,(Mo-S) = 66.9-67.0 eV. The significant shift Of AEb(N1-S) and AEb(Co-S) towards higher values as compared with the binding-energy differences in NiS and Co,S,, respectively, suggests various possible phenomena: (i) some electron transfer from Ni or Co to molybdenum, (ii) more ionic Co-S and Ni-S bonds than in Cogs, and NiS, respectively, or (iii) a larger average charge on Ni and Co than in Cogs, and NiS, respectively. These phenomena are not observed in the FeMo sulphide catalysts. Correlation between Catalytic Activity and Physicochemical Properties The correlation between HDS activity and the threshold temperature of temperature- programmed reduction/sulphidation is shown in fig.3. In general, the HDS activity of sulphided catalysts increases with decreasing TR,s. One curve is obtained for the catalysts2432 Activity of MoS,-based Catalysts - ‘WY 10 E N I I I 0 area (arb. units) Fig. 4. Correlation between HDS activity and TPSE spectral area for the promoted catalysts with r = 0.3: 0, FeMo-CP; ., CoMo-CM and HSP; 0, NiMo-HSP and x , Mo-HSP-0.0. prepared by the HSP, IHSP and CP methods; another curve, of similar shape but at a higher temperature, may be drawn for catalysts CoMo-CM-r.Note also that the same curve fits the results obtained with catalysts with r = 0.1 and 0.3. The existence of different curves for the ‘HSP-type’ (HSP, IHSP and CP) and the essentially biphasic CM catalysts1’ could be due to differences in the location and dispersion of the promoter in the two types of samples. This suggests again the importance of highly dispersed M-S species in decreasing TRIs. The existence of the correlation between TRIs and HDS activity may reflect the ‘lability’ of sulphur in both the precursor and the catalyst. It seems that there is a correspondence between the reducibility of the precursors and that of the catalysts or, more precisely, between the reducibility of the precursor and the ease of formation of anion vacancies, which are believed to be the active sites6.8 f 42 in the HDS of thiophene. Correlation between the reducibility of unsupported CoMo sulphide catalysts and their HDS activity has already been reported by Hoodless et al.5 As shown in fig. 4, the present work confirms this recent finding. A linear relationship is found between the HDS activity and the specific TPSE spectral area (referred to unit surface area) for pure MoS, and promoted catalysts with r = 0.3. The TPO and TPSE results listed in table 2 indicate that the higher the reducibility, the lower the oxidisability of the catalysts. Thus, To values also correlate roughly with HDS activity. Both the HDS activity and the threshold temperature of oxidation increase in the sequence Mo-HSP-0.0 < FeMo-CP-0.3 < CoMo-HSP(1HSP)-0.3 < NiMo-HSP- 0.3 for catalysts prepared by methods leading to the formation M-Mo-S structures.(For HSP and CP = ca. 30% M in M-Mo-S was found by Mossbauer spectroscopy in these 42) Oxidisability, which corresponds to a bulk reaction, does not necessarily correlate with oxygen adsorption. The present results should thus, in principle, be considered independently from literature data concerning oxygen chemisorption, which do not seem always to correlate with activity. The differences between the efficiency of promoters in increasing the reducibility, To and HDS activity of MoS, may be related to the binding energy difference between M 2p and S 2p levels. Indeed, another correlation exists between HDS activity and the changeS.Gobolos, Q. Wu, 0. Andre, F. Delannay and B. Delmon 243 3 20- - I N 1 w E AE/eV Fig. 5. Correlation between HDS activity and AE,, = AEb(M-S)catalyst - AEb(M-S)sulfide for the promoted catalyst with r = 0.3; 0, FeMo-CP; ., CoMo-HSP(1HSP) and n, NiMo-HSP) in binding energy with respect to pure Cogs, or NiS. Fig. 5 shows that both the activity and the differences in AE,(M-S) values between MMo-HSP(CP)-0.3 catalysts and MS, increase in the sequence FeMo < CoMo < NiMo. This correlation between binding- energy change, HDS activity and reducibility may be related to electronic effects (the charge on M, the covalency of the M-S bond and electron transfer between M and Mo). If such electronic transfers occur at the surface of promoted catalysts, the M-S bond strength may change and thus also the reducibility and HDS activity of the catalysts.This transfer could be created not only by the formation of Co-Mo-S or Ni-Mo-S structures, but also by a sufficiently close contact between separate Co or Ni sulphide and MoS, phases.16 In the latter case, the different reducibilty and HDS activity of Fe-, Co- and Ni-promoted MoS,-based catalysts could also be related to different abilities of the pure promoter sulphides for producing hydrogen s p i l l ~ v e r . ~ ~ Conclusions The following conclusions may be drawn from the present work. (1) The increase of the HDS activity of unsupported MoS,-based catalysts in the sequence Mo < FeMo < CoMo < NiMo is correlated with a decrease in the threshold temperature of reduction/ sulphidation of the precursors.(2) Both the onset temperature of oxidation of sulphi- dized catalysts and the amount of H,S released in temperature-programmed sulphur extraction increase in parallel with the HDS activity of the promoted catalysts. (3) A change in the binding energy of the XPS Co 2p3,, and Ni 2p3,2 peaks as compared with the pure sulphides, is also correlated with the HDS activity of unsupported catalysts prepared by the HSP(1HSP) and CP methods. (4) A general conclusion is that the reducibility of the precursors and the catalysts plays a major role in determining HDS activity . We are grateful to Mr M. Genet for help with the XPS experiments. S. Gobolos thanks the Central Research Institute for Chemistry of the Hungarian Academy of Sciences and Q.Wu thanks the Beijing Municipal Chemical Industry Research Institute, Cheng Fu, for leaves of absence during which this work was accomplished. The financial support of the Services de Programmation de la Politique Scientifique for a fellowship (S. G.) and general expenditures, and FNRS (Belgium) is gratefully acknowledged.2434 References Activity of MoS,-based Catalysts 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 H. Topsse, B. S. Clausen, R. Candia, C. Wivel and S. Mssup, Bull. SOC. Chim. Belg., 1981, 90, 1190. S. J. Tauster, T. A. Pecoraro and R. R. Chianelli, J. Catal., 1980, 63, 515. N. Y. Topsse and H. Topsse, J. Catal., 1983, 84, 386.G. Hagenbach, P. Menguy and B. Delmon, Bull. SOC. Chim. Belg., 1974, 83, I . R. C. Hoodless, R. B. Moyes and P. B. Wells, Bull. SOC. Chim. Belg., 1984, 93, 613. P. R. Wentrek and H. Wise, J. Catal., 1978, 51, 80. J. T. Richardson, Ind. Eng. Chem., Fundam., 1964,3, 154. J. Matemova, Appl. Catal., 1983, 6, 61. N. K. Nag, D. Fraenkel, J. A. Moulijn and B. C. Cates, J. Catal., 1980, 66, 162. C. G. Gachet, E. Dhainaut and L. de Mourges, Prepr. Petr. Chem. Diz;. ACS, 1982, 27, 753. C. Gachet, R. Paulus, L. de Mourges, C. Durand and H. Toulhoat, BUN. SOC. Chim. Belg., 1984, 93, 681. R. Candia, B. S. Clausen and H. Topsse, Bull. SOC. Chim. Belg., 1981, 90, 1225. M. Breysse, R. Frety and M. Vrinat, Prepr. Petr. Chem. Dic. ACS, 1982, 27, 772. G. Hagenbach, Ph.Courty and B. Delmon, J. Catal., 1971, 23, 295; 1973, 31, 264. S. Gobolos, Q. Wu and B. Delmon, Appl. Catal., 1984, 13, 89. S. Gobolos, Q. Wu, J. Ladriere, F. Delannay and B. Delmon, Bull. SOC. Chim. Belg., 1984, 93, 687. Q. Wu, S. Gobolos, P. Grange and F. Delannay, Thermochim. Acta, 1984, 81, 281. C. D. Wagner, L. E. Davis, M. V. Zeller, J. A. Taylor, R. H. Raymond and L. H. Gale, Surf. Interface Anal., 1981, 3, 21 1. Q. Wu, S. Gobolos, F. Delannay and B. Delmon, in 10th Symp. Reactivity of Solids, ed. P. Barret and L-C. Dufour (Elsevier, Amsterdam, 1985), p. 1067. V. I. Zaikovskii, A. P. Shepelin, V. A. Burmistrov, A. N. Startsev and Yu. I. Yemakov, React. Kinet. Catal. Lett., 1984, 25, 17. C. K. Groot, A. M. van der Kraan, V. H. J. de Beer and R. Prins, Bull. SOC. Chim. Belg., 1984,93,707. B. S. Clausen, H. Topsse, R. Candia and B. Lengeler, ‘Catalytic Materials: Relationship between Structure and Reactivity’, ACS Symp. Ser., 1983. F. Delannay, D. S. Thakur and B. Delmon, J. Less-Common Met., 1979, 63, 265. D. S. Thakur, P. Grange and B. Delmon, J. Less-Common Met., 1979, 64, 201. E. Furimsky and C. H. Amberg, Can. J. Chem., 1975, 53, 3567. N. M. Zaidman, D. B. Osechkin, M. F. Gladovskaya and E. N. Martynova, Khim. Tekhnol. Topl. Masel, 1961, 6, 25. Ph. C. H. Mitchell and C. E. Scott, Bull. SOC. Chim. Belg., 1984, 93, 619. M. Vrinat, M. Breysse and R. Frety, Appl. Catal., 1984, 12, 151. M. Breysse, R. Frety and M. Vrinat, Appl. Catal., 1984, 12, 165. H. Topsse, B. S. Clausen, R. Candia, C. Wise1 and S. Msrup, J . Catal., 1981, 68, 433. N.-Y. Topsse, H. Topsse, 0. Ssrensen, B. S. Clausen and R. Candia, Bull. SOC. Chim. Belg., 1984,93, 727. 0. Ssrensen, B. S. Clausen, R. Candia and H. Topsse, Appl. Catal., 1985, 13, 363. C. Wivel, R. Candia, B. S. Clausen and H. Topsse, J. Catal., 1981, 68, 453. H. Topsse, R. Candia, H-Y. Topsse and B. S. Clausen, Bull. SOC. Chim. Belg., 1984, 93, 783. R. R. Chianelli, T. A. Pecoraro, T. R. Halbert, W-H. Pau and E. I. Stiefel, J. Catal., 1984, 86, 226. S. Msrup, B. S. Clausen and H. Topsse, J. Phys. Colloy., 1979, 40, C2-88. S. Gobolos, Q. Wu, J. Ladrikre, F. Delannay, P. Grange and B. Delmon, to be published. E. Ya. Rode and B. A. Lebedev, Russ. J. Inorg. Chem., 1961, 6, 608. 0. P. Bahl, E. L. Evans and J. M. Thomas, Proc. R. SOC. London, Ser. A, 1968, 306, 53. A. V. Peuski, A. R. Bobenko and R. G. Kefer, Fro. Vyssh. Uchebn. Zaved. Tswtn. Metall., 1973, I, 37. J. Szajman and R. C. G. Leckey, J. Electron Spectros., 1981, 23, 83. G. C. A. Schuit, Int. J. Quantum Chem., Suppl., 1977, 12(2), 43. Paper 5 / 1624; Received 19th September, 1985
ISSN:0300-9599
DOI:10.1039/F19868202423
出版商:RSC
年代:1986
数据来源: RSC
|
18. |
Thermodynamic properties of binary alcohol–hydrocarbon systems |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 82,
Issue 8,
1986,
Page 2435-2457
Alf Pettersson,
Preview
|
PDF (1278KB)
|
|
摘要:
J. Chem. SOC., Faraday Trans. I, 1986,82, 2435-2457 Thermodynamic Properties of Binary Alcohol-Hydrocarbon Systems Alf Pettersson, Paul Saris and Jar1 B. Rosenholm" Department of Physical Chemistry, Abo Akademi, SF-20500 Abo 50, Finland Head-space gas chromatography has been used to determine the partial vapour pressures of the components of seven binary alcohol-hydrocarbon systems at 298.15 K. The chain lengths of the alcohols were chosen to record the differences between a long-chain alcohol (decan- 1-01) and intermediate homologues (pentan- 1-01 and butan- 1-01). The hydrocarbon solvents chosen (n-octane, cyclohexane and benzene) offer the possibility to extract the contribution of ring closure and aromaticity. The accuracy of the activities were controlled applying a commonly used thermodynamical consistency test.Theexcess Gibbs freeenergies werecombined with previously determined excess enthalpies to obtain the molar excess entropy of solution and the partial molar entropies of the components. The results suggest that the mixtures can be grouped into three concentration regions with significantly different properties. First, in the dilute alcohol solutions the hydrogen- bonding equilibria determine the properties of the system. The influence of the apolar interaction also contributes significantly to the energetic state of the system. Secondly, in the mid-concentration range the hydrogen bonding seems to be of only secondary importance, while the apolar interaction is gaining enhanced importance. In this range the benzene-hydroxy group interaction seems to deviate drastically from the normal behaviour of alcohol-hydrocarbon systems.Thirdly, in dilute hydrocarbon systems the interaction seems to be purely apolar in nature, erasing the previous difference between the aromatic and saturated hydrocarbon systems. ~~~ -~ ~ -~~ The importance of the hydrogen-bonding equilibria to the properties of a multitude of living and technical systems has for a long time been acknowledged. Alcohols may be considered as favourable models for hydrogen-bonded systems. Many of the properties recorded suggest a division into short-chain alcohols (methanol to propanol) with a large association constant1 and to long-chain alcohols, which on steric grounds show a reduced association tendency. Butanol and pentanol are considered intermediate in character.In many investigations it is fortuitous to compare the influence of a solute or the solvent on the hydrogen-bond network with the effect produced when the temperature is increased. It is possible to convert the shift induced by the solute to a temperature effect on the pure The hydrophobicity of the alcohols offers another common and useful way to study the extent of the hydrogen bond association: when diluting alcohols in inert hydrocarbons it may be assumed that the disruption of hydrogen bonds dominates the properties of the Recently a few authors have indicated that hydrogen bonding makes an important contribution to the thermodynamic properties only in the most dilute region.5* For most solutions the structural compatibility of the alcohol hydrocarbon residue with the solvent molecules is more i m p ~ r t a n t .~ ? The present investigation reports the Gibbs free energy and the entropy as well as the partial molar enthalpies of seven alcohol-hydrocarbon systems. The alcohols were chosen to record the difference between a long-chain alcohol and the intermediate homologues. The hydrocarbon solvents offer the possibility to extract the contribution of aromaticity to the parameters considered. The influence of breaking the cyclohexane 24352436 Thermodynamics of Alcohol-Hydrocarbon Systems ring is also given. The systems investigated should thus provide a picture of the relative importance of the polar and apolar interactions for the state of the systems. The entropy should be successful in recording the structural state of the system.Experimental Chemicals The alcohols used were: butan-1-01 (Merck AG, p.a. grade), pentan-1-01 (Merck AG, p.a. grade), and decan-1-01 (Merck AG, pro synthesis). The hydrocarbons were: benzene (Merck AG, p.a. grade), cyclohexane (Merck AG, p.a. grade), and n-octane (Fluka AG, puriss.). All the alcohols were distilled before use and molecular sieve pellets (Merck AG, 0.4 nm) were added to the mid-fraction used in order to remove residual traces of water. For the same reasons metallic sodium was added to the hydrocarbon liquids. The final water content, as determined by a Karl-Fisher titrator (Metrohm 652 KF-Coulometer), was 0.02% by weight for butan-1-01, 0.01 1 % for pentan-1-01, 0.01 % for decan-1-01, 0.0012% for benzene, 0.001 % for cyclohexane and 0.001 ;< for n-octane.Instrumentation The vapour pressures of the components were analysed with a Hewlett-Packard 5730A gas chromatograph equipped with an f.i.d. detector and an HP 7672A automatic sampler unit. The sampler was controlled with an HP 19400A sampler/event control module and the results were recorded using an HP 3390A integrator. The temperature (298.15 0.10 K) of the closed sampler box was measured at the vial to be analysed. In order to minimize the influence of possible temperature gradients on the gas-liquid equilibrium in the vials, the time between two successive samplings was programmed to be considerably longer than the retention time of the components. Care was also taken to ‘wash’ the syringe between the samplings so that no residual traces from previous injections could be detected.When running the binary solutions containing butanol, a 0.7 m x 1 /8 in? stainless steel column packed with 10% Carbowax 20 M on 80-100 mesh Gas-Chrom Q was used. The oven temperature was 368 K. The other solutions were analysed with Porapak Q, 150-200 mesh packed in a stainless steel column of the same dimensions. In these cases the oven temperature was kept at 503 K, except for the decanol-octane samples, for which a temperature of 513 K was chosen. The injection port was kept at 523 K and the f.i.d. detector at 573 K. Ultra-pure helium was used as a carrier gas with a flow rate of 30 cm3 min-l and a pressure of 3 bar. The pressures of hydrogen and air were optimized to give the largest peak areas and the least tailings possible.The pressures used were 0.85 bar and 1.8 bar, respectively. Measurements The activities of the components were determined as the relative integrated gas chromatogram peak areas of the sample to the reference. Although this procedure neglects some contribution from non-ideality it has been found that the non-ideality of the sample vapour and the pure liquid vapour tend to ~ a n c e l . ~ To ensure maximal accuracy several reference samples of the pure liquids were run in the beginning, the middle and at the end of a sample series. The experimental points were smoothed in order to make the activities approach the Raoult’s slope. When analysing the decanol systems it was found that the vapour pressure of decanol was too low to get reliable activities. Therefore the Gibbs-Duhem law was applied to calculate the activity factor of decanol t 1 in = 2.54 cm.A .Pettersson, P . Saris and J. B. Rosenholm 2437 from the activity factors of the hydrocarbons by integration. In estimating the activity factor for the most dilute solution we utilized the standard method of using a temporary lower limit.1° Originally 14 cm3 of the samples were mixed gravimetrically in order to reduce weighing errors although only 0.5 cm3 was used in the experiments. The gas volume taken from the head space of the 2 cm3 vials was 0.1 cm3. The consistency of the rational activity coefficients calculated were checked using the integral test in the way suggested by Gmehling and 0nken.ll The test involves a comparison of the areas of the plot of In [f(alcohol)/f(hydrocarbon)] us.x(alcoho1) above ( A ) and below ( B ) the abscissa using the following equation The following deviations were found : butanol-benzene (2.3 % ), butanokyclohexane (9.9 % ), pentanol-benzene (1.2:4 ), pentanokyclohexane (2.8 % ), decanol-benzene (0.6% ), decanol-cyclohexane (1 .O% ), and decanol-octane (0.7% ). The test suggests that the activities measured for all the systems may be considered thermodynamically consistent. Some care should, however, be taken when using the activities of the butanol-cyclohexane system owing to the somewhat too high deviation percentage level observed. There are very few publications on the activities of the components of binary alcohol-hydrocarbon systems.Gmehling and Onken have tabulated the vapour pressures of the butanol-benzene, butanol-cyclohexane and decanol-cyclohexane systems. l1 When recalculated to the excess Gibbs free energy of solution our results agree with theirs within 20%. However, since their number of data points was small and the data failed the consistency test, we consider our results to be of higher accuracy. Recently Sjoblom and Henriksson have published activities of pentanol mixed with benzene at 293 K using a similar head-space gas chromatography technique.12 Since they provided no information of the activity of the hydrocarbon component we could not compare the thermodynamic consistency of their activities. Probably the most reliable activities obtained with an alternative technique have been reported by French and Stokes on the butanol- cyclohexane system at 298 K.139 l4 It was found that both our partial molar excess Gibbs free energy of the butanol and our enthalpy function [H~/x(butanol)x(cyclohexane)] agree very well with their results.Since our thermodynamic consistency test indicated some uncertainty for this particular system, the comparison suggests that some doubt may be laid on our activities of cyclohexane for this system. Results Gibbs Free Energy In reporting the results we first compare the properties of the alcohols when mixed with benzene (fig. 1) and cyclohexane (fig. 2), respectively. This provides an opportunity to study the influence of lengthening the chain on the property investigated, keeping the solvent and the functional group the same.Finally we illustrate the effect imposed by the change of solvent, keeping the chain length of the alcohol (decanol) the same (fig. 3). In fig. 1 (a)-3 (a) the molar excess Gibbs free energies are plotted against mole fraction of the alcohols. The figures also include partial molar excess Gibbs free energies (the chemical potentials) for the alcohols (b) and the hydrocarbons (c), respectively, plotted against the mole fraction of the alcohols. As described in the previous section, the partial molar free energies are obtained directly from the measurements. This option reduces considerably the error which would be introduced if they were derived from the molar excess Gibbs free energy of the system! In our case the latter function has been calculated by summation of the contributions of the components.2438 Thermodynamics of Alcohol-Hydrocarbon Systems 200 0 0 0.2 0. 4 0. 6 0. 8 1. 0 x (alcohol) 10000 L 1000 - 800 - 3 - I 2 600 - W E u c, 1 400 - 4 I - 0 6000 --. v c- u w 4000 2000 0 0 0.2 0.4 0.6 0.8 1.0 x(alcoho1) Fig. l ( a ) and (b). For description see opposite.A . Pettersson, P. Saris and J . B. Rosenholrn 2439 3000 1000 0 0 0.2 0.4 0.6 0.0 1.0 x(alcoho1) Fig. 1. (a) The molar excess Gibbs free energies for (a) butanol-, (A) pentanol- and (a) decanol-benzene systems at 298.15 K us. the mole fraction of alcohol and the partial molar excess Gibbs free energies of the alcohols (b) and benzene (c). For all the solutions the molar excess Gibbs free energy is positive with the maximum shifted slightly towards solutions poor in alcohol [x(alcohol) z 0.41. The maximum values vary between the limits of ca.1.2 and 0.78 kJ mol-1 found for the benzene systems. The most important feature of the GZ curves [fig. 1 (a)-3(a)] is that the spacing between them changes at both extremes of concentration. When the alcohols are diluted with the hydrocarbons the partial molar excess Gibbs free energies (the excess chemical potentials) remain close to zero in the x(alcoho1) = 0.8- 1.0 range. For the solutions poorer in alcohol a successive deviation towards positive energies is registered. For the same solvent the deviation is largest for butanol, while octane produces the largest effect when mixed with decanol. The partial molar excess free energies of the hydrocarbons increase monotonically over the whole concentration range, but show a saturation tendency for the dilute solutions.The magnitude of the differences found follow the trend given for the excess Gibbs free energies of the alcohols. The limiting values of the plots are given in table 1. Enthalpy In fig. 4 we have collected the partial molar enthalpies for all the systems discussed. They have been calculated from the molar excess enthalpies published in another context6 by stepwise derivation and smoothed when possible to averages of three consecutive points. The equations used were where A = alcohol and HC = hydrocarbon.2440 Thermodynamics of Alcohol-Hydrocarbon Systems 0 D.2 0. 4 0. 6 0.8 1.0 x (alcohol) 10000 8000 6000 w u 4000 2000 0 0 0.2 0.4 0.6 0.8 1.0 x(alcoho1) Fig.2(a) and (b). For description see opposite.A . Pettersson, P. Saris and J . B. Rosenholm 244 1 1000 0 0 0.2 0.4 0.6 0.8 1.0 x( alcohol) Fig. 2. (a) The molar excess Gibbs free energies for (0) butanol-, (A) pentanol- and (0) decanokyclohexane systems at 298.15 K us. the mole fraction of alcohol and the partial molar excess Gibbs free energies of the alcohols (b) and cyclohexane (c). The influence of the alcohol chain length on the enthalpic state of the components is very small [fig. 4(a) and (b)]. The largest effects are recorded when varying the hydrocarbon liquids mixed with decanol [fig. 4(c)]. Accordingly, these systems also produce the highest maximal molar excess enthalpy of solution of ca. 1.30 kJ mol-l (benzene system), 0.66 kJ mol-1 (cyclohexane system) and 0.50 kJ mol-1 (octane system).6 The corresponding values for the pentanol and butanol systems are only ca.100 (cyclohexane) - 200 (benzene) J mol-l lower. Note, however, that the sequence is opposite as compared with the molar excess Gibbs free energies! The partial molar excess enthalpies are largest for decanol mixed with benzene, while the partial enthalpies observed for the saturated hydrocarbon mixtures are much the same [fig. 4 (c)]. For the latter systems a significant contribution from decanol is obtained only for solutions with x(decano1) = 0-0.1. For the benzene mixtures the corresponding concentration range extends to x(decano1) = 0.3 [fig. 4(a)]. The partial molar excess enthalpy of the hydrocarbons increases smoothly with alcohol concentration. There is, however, a significant exothermal shift in the partial molar excess enthalpy of both benzene and cyclohexane when the mole fraction of decanol exceeds 0.9.For octane this effect is not observed. Since the enthalpies reported by FrenchI4 did not exhibit any exothermal parts in the dilute region a relatively high uncertainty is suggested for the limiting hydrocarbon values given in table 1. Entropy All the molar entropies are expressed as ‘entropy function’ calculated by applying the Gibbs-Helmholtz equation expressed in the form (4) The negative sign was used for the entropy function in order to obtain the energetically favourable and unfavourable contributions to the state of the system directed in the same -TSE = GR-H” 81 F A R 12442 Thermodynamics of Alcohol-Hydrocarbon Systems 1000 800 4 - I 2 600 W E u 400 200 0 0 0.2 0.4 0.6 0.8 1.0 x( decanol) 10000 8000 - 2 6000 13 *.u W W 4000 2000 0 0 0.2 0.4 0.6 0.8 1.0 x( decanol) Fig.3(a) and (6). For description see opposite.3000 - I - E" 2 2000 n w ._ v u 1000 0 A . Pettersson, P. Saris and J . B. Rosenholm r 2443 0 0.2 0.4 0-6 0.8 1.0 x(decano1) Fig. 3. (a) The molar excess Gibbs free energies for (*) octane-, (0) cyclohexane- and (A) benzenedecanol systems at 298.15 K us. the mole fraction of decanol and the partial molar excess Gibbs free energies of decanol (6) and the hydrocarbons (c). Table 1. The limiting excess Gibbs free energies, excess enthalpies and excess entropies of alcohols in hydrocarbons and hydro- carbons in alcohols, respectively, at 298.15 K" mixture GE H E b - TSE /kJ mol-l /kJ mol-l /kJ mol-l alcohols butanol in benzene 10 pentanol in benzene 8 decanol in benzene 5 butanol in cyclohexane 12 pentanol in cyclohexane I1 decanol in cyclohexane 9 decanol in octane 8 hydrocarbons benzene in butanol 4.0 benzene in pentanol 3 .2 benzene in decanol 2.2 cyclohexane in butanol 4.6 2 Q 15 16 17 20 21 23 23 0.3 0.3 0.3 0.8 - 5 -8 - 12 -8 - 10 - 14 - 15 3.7 2.9 1.9 3.8 ciclohexane in decanol 2.8 0.8 2.0 octane in decanol 3.1 1.1 2.0 a The accuracy of the limiting alcohol values are + 2 kJ mole'. The accuracy of the limiting hydrocarbon values are k0.5 kJ mol-l. CJ ref. (6). 81-22444 Thermodynamics of Alcohol-Hydrocarbon Systems 0 0 0 0 (I) d 0 0 0 * 4 0 0 0 2 0 0 0 v 0 0 0 *A .Pettersson, P . Saris and J. B. Rosenhoh 2445 18000 16000 d I - 8 14000 b n -.- .- @ 12000 I0000 8000 6000 4000 2000 0 0.2 0. 4 0. 6 0. 8 1.0 x( decanol) Fig. 4. The partial molar enthalpies of the components of the alcohol-benzene systems (a), the alcohol*yclohexane systems (b) and the decanol-hydrocarbon systems (c) at 298.15 K us. the mole fraction of alcohol. The symbols refer to those in fig. 1 (a), 2 (b) and 3 ( c ) respectively. way. A negative deviation thus indicates an increase in entropy which contributes favourably to the Gibbs free energy (negative AG). In order to make the comparison of the enthalpy and entropy contributions possible we included the appropriate temperature (298.15 K) in the entropy function.Fig. 5-7 present the molar excess entropy functions plotted against the mole fraction of the alcohols. It is very interesting to note that the systems are characterized by both negative (positive entropy) and positive (negative entropy) parts. The molar excess entropy function is mainly positive for the saturated hydrocarbon systems and mainly negative for the benzene solutions. The extension of the chain length of the alcohol deepens the minimum. The maximum value of ca. 0.55 kJ mol-1 is then found for the butanol-cyclohexane system, while the lowest value of -0.61 kJ mol-1 is found for the decanol-benzene system. As found for the excess Gibbs free energies, the spacing between the - TS: curves changes with the concentration. The partial molar excess entropy functions of the alcohols do not deviate much from2446 d I 3 i$ 2 h \ WE I Thermodynamics of Alcohol-Hydrocarbon Systems 0 -100 -200 -300 -400 -500 0 0.2 0.4 0.6 0.8 1.0 x( alcohol) -2000 -8000 -10000 0 0.2 0.4 0.6 0.8 1.0 x(alcoho1) Fig. 5(a) and (b). For description see opposite.A . Pettersson, P. Saris and J . B. Rosenholm 2447 2000 0 - 1000 0 0.2 0.4 0. 6 0.8 1.0 x(alcoho1) Fig. 5. (a) The molar excess entropy (expressed as - 7‘s:) of the (0) butanol-, (A) pentanol- and (m) decanol-benzene system at 298.15 K us. the mole fraction of alcohol and the partial molar excess entropies of the alcohols (b) and benzene (c). zero for alcohol mole fractions > 0.5. Below this limit the entropy function of the mid-chain alcohols departs to positive values.In the presence of benzene the changes are small and may reverse the function to the negative side [fig. 5(b)]. Conversely, the positive ‘hump’ gets narrower when benzene is exchanged for the alkanes [fig. 6(b) and 7 (b)]. In the dilute alcohol range [especially below x(alcoho1) z 0.11 the entropy function of all the alcohols falls rapidly towards very low values (high entropy). The entropy function representing the alkanes is smooth and mainly positive. The dependence on the chain length of the alcohols is largest in the x(alcoho1) = 0.6-0.9 range. For the benzene systems the negative loop is large, while the molar excess entropy is negative (positive entropy function) only in a narrow concentration range,7 x(alcoho1) > 0.7. The limiting values are given in table 1.Discussion Although the Gibbs free energy provides the ‘final’ information about the state of the system, this function is the sum of essential interactions which compensate each other, making it a relatively insensitive parameter. Therefore higher-order derivatives in e.g. temperature and pressure are engaged when investigating in detail the interactions of thermodynamical irnportance.l5* l6 At the first-derivation level we find the entropy and volume, respectively. As has been pointed out the enthalpy may be considered as a practically accessible function needed to calculate the entropy, but which does not relate to the Gibbs free energy in a truly straightforward way.16 When comparing functions of the first-derivative level one may thence expect the entropic and the volumetric properties in the first hand to be mutually comparable.Viewing the relationship between the parameters in this way, the enthalpy can be understood as the part of the total entropy sacrificed to maintain the maximal disorder of the system and which thence is compensated in the Gibbs free energy functi0n.l’2448 0 -2000 -4000 4 F h 5 -6ooo v w -8000 Thermodynamics of Alcohol-Hydrocarbon Systems I 0 0.2 0.4 0.6 0.8 1.0 x (alcohol) -10000 -12000 0 0.2 0.4 0.6 0.8 1.0 x(alcoho1) Fig. 6(a) and (b). For description see opposite.A . Pettersson, P. Saris and J . B. Rosenholm 2449 0 0.2 0.4 0.6 0.8 1.0 x(alcoho1) Fig. 6. The molar excess entropy (expressed as - TS;) of the (0) butanol-, (A) pentanol- and (0) decanokyclohexane systems at 298.15 K us.the mole fraction of alcohol and the partial molar excess entropies of the alcohols (b) and cyclohexane (c). Gibbs Free Energy The general qualitative information obtained from the concentration dependence of the Gibbs free energies is that the cumulative interactions active in the solutions are energetically unfavourable (positive GE values). The interactions which tend to oppose the favourable state obtainable by a complete mixing [Gideal = RTX x(i) In x(i) = nega- tive] may, in the first hand, be considered as being due to preferential interaction of each component with molecules of their own kind. However, the increment of the chemical potential of both the alcohol and the hydrocarbon is negative when mixed into a solution. In agreement with the findings of Benson and coworkers5? the dominant contribution from the hydrogen bonding equilibria is confined to mole fractions of alcohol < 0.2.In this concentration range the consumption of energy for the disruption of the hydrogen bonds of the alcohol associates (fig. 4) is also reflected in a very high limiting Gibbs free energy. Diluted in the same solvent we find the largest limiting Gibbs free energies and the smallest enthalpies (and entropies) for the medium-chain alcohols (table 1). These homologues seem thence to resist most successfully the dissolving effect on the associates upon dilution. This is in accordance with the findings of Treszczanowicz and Benson, who found negative deviations in the molar excess volume (enhanced hydrocarbon chain-solvent interaction) when the homologous series of the normal alcohols were a s ~ e n d e d .~ ~ ~ ~ - ~ ~ Comparing the solvents we note that benzene seems to provide an energetically more favourable environment (lowest GE) than cyclohexane or octane (table 1). When mixed in large amounts, benzene interacts readily with the alcohols, while octane seems not to interact. Indeed, somewhat surprisingly, long-chain alkanes seem to avoid mixing with equal or longer-chain alcohols. This may be the cause for the slightly enhanced association observed for the alcohols in octane.6 The situation changes significantly when one of the components is present in excess over the other. Over the mid-concentration range the Gibbs free energy of the components of the cyclohexane systems resembles those of the benzene systems, but approaches the octane systems at2450 Thermodynamics of Alcohol-Hydrocarbon Systems 300 200 100 0 + - I 0 h E -100 2 -200 \ W E I -300 -400 -50 -600 0 0.2 0.4 0.6 0.8 1.0 x(decano1) 0 -2000 4 -4000 C v -6000 c, 1 w ? -8000 - 10000 - 12000 0 0.2 0.4 0. 6 0.8 1.0 x(decano1) Fig. 7(a) and (b). For description see opposite.A . Pettersson, P . Saris and J . B. Rosenholm 245 1 150.0 irl 0 -500 z I -1000 -1500 1 ~~~ ~ 0 0.2 0.4 0.6 0.8 1.0 x( decanol) Fig. 7. The molar excess entropy (expressed as - TS:) of the (*) octane-, (0) cyclohexane- and (A) benzenedecanol systems at 298.15 K us. the mole fraction of decanol and the partial molar entropy of decanol (b) and the hydrocarbons (c). both extremes of the concentration scale.An important contribution may thus be expected from the structural compatibility on the dilution p r o ~ e s s . ~ ~ * En t h a1 p y If the partial molar enthalpy of the alcohols is considered a sensitive indicator of the hydrogen bonding equilibria the enthalpies recorded suggest that the self association is confined to the mole fraction range below 0.1 (0.2 for the benzene systems). If associates with a larger number of hydrogen bonds would form at higher concentrations of alcohols4 one would expect a more dramatic influence on the partial molar enthalpy of the alcohols in that range. Since a smooth decay is found our partial molar enthalpies do not support this view. Alternatively two or more opposing contributions erase the effect expected.One may suggest, on the basis of the steeper limiting slopes of the decanol systems, that the association is most intense in the long-chain alkane solutions (decanol-octane system).6 In agreement with the above conclusions the associatiqn seem to be considerably retarded in the benzene system (smaller initial slopes) owing to an enhanced intermolecular interaction made possible by the n-electrons. This is also evident from the limiting values (table 1) being close to the expected enthalpy for a hydrogen bond in the alkane systems, but some 20% lower for the benzene systems. Comparing the alcohols diluted in the same solvent we note that only a small difference is found, the partial molar enthalpy being largest for decanol. Exceeding the intense association range the hydrocarbon interactions seem to dominate the enthalpic properties of the system.When comparing our results with the maximal molar excess enthalpies found for binary alkane-alkane systems21 we find that the introduction of the hydroxy group, as a rule, contributes only a fraction of the total endothermal interaction energy.6 The contribution increases in the order benzene < cyclohexane < octane in accordance with the finding of an intensified alcohol association in the latter systems. According to Patterson and coworkers7 a negative contribution to2452 Thermodynamics of Alcohol-Hydrocarbon Systems the enthalpy, entropy and volume is obtained when the segments of the chain are capable of an intensified correlation of molecular order (CMO). The highest enthalpies are recorded for benzene, which indicates a strong benzene-hydroxy interaction, while decanol mixed with octane (strong CMO internally among the octane chains) gives the lowest overall, but endothermic partial molar excess enthalpies.The limiting partial molar enthalpy of the hydrocarbons is independent of the alcohol, which indicates the interaction being of similar nature. The limiting partial molar enthalpy increases in the order benzene < cyclohexane < octane. However, the order is reversed when entering the alcohol-rich concentration range [fig. 4(c)]. There thus seems to be some kind of rapid interruption of the benzene-hydroxy interaction when x(alcoho1) > 0.9. Probably all the alcohol molecules are occupied in alcohol association, giving rise to an interaction with the chains surrounding the alcohol aggregates. Entropy As discussed above, entropy should be especially suitable as an indicator for the structural state of the system.The main shortcoming of this parameter lies in the accumulating error due to the calculation procedure. For the present purpose it suffices to consider the entropy simply as a measure of the order created in the system (positive deviation of - TS”) owing to the intramolecular interaction between each of the components. The picture emerged fits very well with the conclusions drawn in the preceding sections. Consequently the benzene system exerts the highest degree of intermolecular interaction (negative entropy function). The roughly equal spacing in the mid-concentration range suggests that the origin of the entropy increase is not linearly dependent on the chain length or that some other factors also contribute to the structure of the system.The first explanation available is the above mentioned ability of benzene to interact with polar molecules such as alcohols. In the solutions there seems to be a fraction of unassociated alcohol molecules in the benzene mixtures owing to the retardation of the association process. The surprisingly similar functional dependence as compared with volume suggests a comparison can be made with the volumetric behaviour. Since a negative entropy function (positive entropy) corresponds to a positive volume, it should be connected with the chemical (disruption of hydrogen bonds) or the physical (non-specific interactions) contributions or both.5 In accordance with the conclusions drawn on the basis of both the limiting Gibbs free energy and the limiting enthalpy values, the increase in the degree of disorder created upon dilution in hydrocarbon liquids is largest for the long-chain alcohols [fig.7(b), table 11. Mixed in the same solvent the association may be considered retarded for the long-chain alcohols [fig. 5(b) and 6(b)] owing to the increased molecular compatibility of the chain with the solvent molecules. In the case of self-association enforced by the solvent [decanol-octane system, fig. 7(b)] or non-compatibility of the alcohol associates with the solvent [medium-chain alcohol-hydrocarbon, fig. 5(b) and 6(b)] the poor mixing results in a ‘relaxation effect’ shown as a maximum in the - TSE(alcohol) function.In the high-concentration range [x(alcohol) > 0.81 all the hydrocarbons seem to be involved in some new kind of interaction with the alcohol associates, producing a positive partial molar entropy function for the hydrocarbons. If the intermolecular interaction between alcohol and benzene is taken as a basis, the effect observed for x(alcoho1) > 0.9 is probably due to a lack of alcohol molecules engaged in benzene-hydroxy group interaction. Instead the alcohol molecules seem to be consumed by the energetically more favourable hydrogen-bond association. The limiting partial molar entropies are for all hydrocarbons roughly the same when dissolved in the same alcohol (table 1). Once the option of polar interaction is excluded the interaction seems to be purely apolar in nature.Then long-chain alcohol complexes mix most readily with the hydrocarbon solutes, leading to the highest degree of intermolecular mixing (least-negative limiting hydrocarbonA . Pettersson, P. Saris and J. B. Rosenholm 2453 entropies). This conclusion is in accordance with the largest limiting enthalpies and the lowest limiting Gibbs free energies found for the hydrocarbon-decanol systems [fig. 5 (c) and 6(c)]. However, as mentioned before, decanol mixes least readily with octane (fig. 7). The insensitivity of the limiting partial molar entropies to the nature of the hydrocarbon solute may be explained if assumed that the size of the apolar alcohol aggregates determines the magnitude of the interaction.Again a parallel may be found with the volumetric behaviour of hydrocarbon mixtures, for which a positive molar excess volume has been found when a low-molecular-weight alkane is dissolved in a high- molecular-weight alkane.22 Conclusions It has been shown that head-space gas chromatography may be used to obtain accurate and thermodynamically consistent activities of volatile components of alcohol- hydrocarbon mixtures. The results suggest that the mixtures can be grouped into three concentration regions with significantly different properties : first, in dilute alcohol solutions the hydrogen-bonding equilibria determine the properties of the systems. However, the influence of the apolar interaction also influenced alcohol association.Of special importance is the retarding effect of aromaticity on association equilibria. Secondly, in the mid-concentration range the hydrogen bonding seem to be of only secondary importance for the state of the system. Again the 7r-electron interaction plays a central role. The lengthening of the alcohol chain makes intermolecular mixing more probable, but the long-chain alkanes (octane) reject intermolecular mixing. Thirdly, a new situation is created when the hydrocarbons are diluted in the alcohols. In nearly neat alcohol solutions, all the free monomers seem to be consumed by alcohol association and a purely apolar interaction seems to contribute to the solution properties. The functional dependence of the entropy is, in many respects, similar to that which is found for volumetry.It is therefore suggested that these functions are on the first hand mutually comparable, while the enthalpy may be understood merely corresponding to the part of the total entropy sacrificed to counteract the thermal fluctuations to maintain the maximal disorder of the system. Although the Gibbs free energy provides the information of the equilibrium of the system, this function is frequently a relatively insensitive parameter owing to the compensation of essential interactions. This work has been supported by the Finnish Research Council for Natural Sciences. References 1 W. E. Achree Jr, Thermodynamic Properties of' Non-electrolyte Solutions (Academic Press, Orlando, 2 J. Paquette and C. Jolicoeur, J. Solution Chem., 1977, 6, 403.3 Y. DeGrandpre, J. B. Rosenholm, L. L. Lemelin and C . Jolicoeur, Solution Behavior ofSurfuctunts, ed. 4 E. Tucker and E. D. Becker, J . Phys. Chem., 1973, 77, 1783. 5 A. J. Treszczanowicz, 0. Kiyohara and G. C. Benson, J . Chem. Thermodyn., 1981, 13. 253. 6 P. Saris, J. B. Rosenholm, E. Sjoblom and U. Henriksson, J . Phys. Chem., 1986, 90, 660. 7 S. N. Bhattacharyya, M. Costas, D. Patterson and H-V. Tra, Fluid Phase Equilibria, 1985, 20, 27. 8 K. N. Marsh and C . Burfitt, J . Chem. Thermodyn., 1975, 7, 955. 9 D. M. Mohliner, L. M. Bowman, S. J. Freeland and H. Nakadomari, J . Electrochem. SOC., 1973, 120, 1658. 10 I. M. Klotz and R. M. Rosenberg, Chemical Thermodynamics: Basic Theory and Methods (W. A. Benjamin, Menlo Park, California, 3rd edn, 1972), chap.20, p. 342. 1 1 J. Gmehling and U. Onken, Vapor-Liquid Equilibrium Data Collection, (a) Aqueous-Organic Systems, vol. 1, part 1, chap. 2.3, p. XXII, (b) Organic Hydroxy Compounds, Afcohols and Phenols, Part 2b (Dechema, Frankfurt am Main, 1978). 12 E. Sjoblom and U. Henriksson, Surfuctants in Solutions, ed. K. L. Mittal and B. Lindman (Plenum Press, New York, 1984), vol. 3, p. 1867. Florida, 1984), chap. 8, p. 150. K. L. Mittal and E. J. Fendler (Plenum Press, New York, 1982), vol. 1, p. 431.2454 Thermodynamics of Alcohol-Hydrocarbon Systems 13 H. T. French and R. H. Stokes, J. Phys. Chem., 1981,85, 3347. 14 H. T. French, J. Solution Chem., 1983, 12, 869. 15 C . Jolicoeur, Methods of Biochemical Analysis, ed. D. Glick (J. Wiley, New York, 1981), vol. 27, p.171. 16 J. B. Rosenholm, Fresenius' Z. Anal. Chem., 1985, 321, 731. 17 R. Lumry, Bioenergetics and Thermodynamics: Model Systems, ed. A. Braibanti (D. l e d e l , Dordrecht, 18 A. J. Treszczanowicz and G. C. Benson, J . Chem. Thermodyn., 1977,9, 1189. 19 A. J. Treszczanowicz and G. C. Benson, J. Chem. Thermodyn., 1978, 10, 967. 20 A. J. Treszczanowicz and G. C. Benson, J. Chem. Thermodyn., 1980, 12, 173. 21 J. J. Christensen, R. W. Hanks and R. M. Izatt, Handbook of Heats of Mixing (J. Wiley, New York, 22 R. S. Hutchings and A. van Hook, J . Chem. Thermodyn., 1985, 17, 523. 1980), p. 405-423. 1982). Paper 5/1632; Received 20th September, 1985 Appendix Tables of Thermodynamic Data for the Alcohol-Hydrocarbon Systems Table A 1. C,H,OH-C,H, GZ TSZ G~(C,H,OH) TSE(C,H,OH) GE(C6H6) /J mol-' /J mol-I x(C,H,OH) /J mol-1 /J mol-' x(C6H6) /J mol-' 23.3 38.1 77.9 154.3 233.5 342.2 466.7 591 .O 800.5 917.3 1101.8 1187.3 1120.6 970.7 835.2 639.9 515.9 355.6 268.0 115.2 10.3 22.7 51.6 117.0 117.4 140.8 169.0 160.6 91.5 69.8 - 9.3 - 99.2 - 89.4 - 129.8 - 146.4 - 189.6 - 171.1 - 142.2 - 131.4 - 80.3 0.002 26 0.004 25 0.009 53 0.020 44 0.031 07 0.050 48 0.075 96 0.099 91 0.153 31 0.196 18 0.298 32 0.423 71 0.503 58 0.631 07 0.701 59 0.805 91 0.853 91 0.908 79 0.933 98 0.974 31 9 934 8 173 7 091 6 378 6 167 5 187 4 375 4 038 3 312 2 808 1 997 1 307 893 443 279 122 68 33 23 13 4 945 6 146 6 508 5 461 2 632 1372 344 -118 - 752 - 808 - 687 - 507 ~ 393 - 233 - 199 - 122 - 68 - 33 + 23 - 13 0.997 74 0.995 75 0.990 47 0.979 56 0.968 93 0.949 52 0.924 04 0.900 09 0.846 69 0.803 82 0.701 68 0.576 29 0.496 42 0.368 93 0.298 41 0.194 09 0.146 09 0.091 21 0.066 02 0.025 69 1 3 10 24 43 84 145 208 345 455 72 1 1098 1351 1871 2 142 2 787 3 129 3 564 3 732 3 983 T s 4 c 6 H 6 ) /J mol-' -1 - 10 5 36 75 154 191 244 284 278 20 1 218 48 - 22 - 467 - 769 - 1224 - 1662 - 2623 - 3A .Pettersson, P . Saris and J . B. Rosenholm Table A 2. C,H,OH-C,H12 2455 G2 TSg /J mol-I 27.7 53.1 92.9 147.9 230.0 343.1 505.6 564.6 803.0 882.9 982.2 1063.8 1098.1 1070.7 973.4 841.8 738.8 447.1 236.2 115.6 ~~ /J mol-1 ~- ~ 11.3 61.0 88.9 71.5 38.6 -51.9 - 149.3 - 147.9 -265.6 -321.5 -401.9 - 449.9 - 506.3 - 522.6 - 499.1 - 464.1 -416.3 - 282.2 - 167.7 - 90.4 x(C,H,OH) ~ - _ _ 0.002 15 0.004 60 0.009 33 0.0 16 28 0.028 45 0.049 09 0.081 75 0.098 43 0.167 75 0.203 53 0.254 87 0.325 70 0.407 81 0.503 27 0.611 89 0.719 15 0.773 66 0.890 97 0.948 07 0.976 17 ~~ G~(C,H,OH) TS~C,H,OH) E(C6H12) /J mol-l /J molP x(C,H1,) /J rno1-l - 11 043 9 606 8 409 7 672 6 706 5 578 4 543 4 183 3 052 2 647 2 165 1 628 1156 738 388 168 91 37 18 10 7 116 6 553 4 710 - 232 -2 386 -2 938 -2 543 -2 423 -1 932 -1 767 -1 525 - 1068 - 736 - 468 - 228 - 88 -31 - 37 - 18 - 10 0.997 85 0.995 40 0.990 67 0.983 72 0.971 55 0.950 91 0.918 25 0.901 57 0.832 25 0.796 47 0.745 13 0.674 30 0.592 19 0.496 73 0.388 11 0.280 85 0.226 34 0.I09 03 0.051 93 0.023 83 4 8 14 23 40 72 146 169 349 43 1 577 79 1 1057 1407 1895 2564 2953 3792 4206 4422 TSE(C6H12) /J mol-' -4 31 45 76 109 97 63 100 70 48 - 17 - 151 - 347 - 577 - 925 - 1424 - 1733 - 2280 - 2886 - 3362 G g /J mol-I TSg /J molP 20.4 37.4 73.4 129.1 179.4 276.9 375.0 469.5 63 1.4 754.3 905.2 963.0 938.5 845.1 743.4 598.1 463.9 312.7 164.7 88.4 20.0 34.1 69.2 121.7 166.4 227.3 203.3 249.8 300.5 3 14.0 290.6 196.9 136.4 98.8 - 40.4 - 119.4 - 88.0 - 56.2 - 38.4 - 32.9 Table A 3.C,H,,OH-C,H, G~(c,H,,oH) TS~C,H,,OH) x(C5Hl,0H) /J mol-l /J mol-l x(C,H,) 0.002 65 7 612 7667 0.997 35 0.004 89 7 196 7443 0.995 11 0.010 56 6 504 7 015 0.989 44 0.020 23 5 866 6053 0.979 77 0.029 97 5 395 4204 0.97003 0.050 63 4 729 2230 0.949 37 0.070 56 4 063 856 0.92494 0.100 24 3 586 493 0.89976 0.149 59 2 890 269 0.85041 0.200 22 2 441 18 0.799 78 0.299 90 1705 -145 0.700 10 0.400 04 1117 -197 0.59996 0.499 97 714 -194 0.50003 0.598 75 405 -115 0.401 25 0.700 46 218 - 138 0.299 54 0.800 56 113 -113 0.19944 0.848 45 57 -57 0.151 55 0.897 81 21 -21 0.102 19 0.949 69 9 -9 0.050 31 0.973 33 3 -3 0.026 67 GE(C6H6) /J mol-l TSF<C6H6) /J mol-' 1 2 4 10 18 39 75 122 234 33 I 562 860 1162 1500 1972 2544 2738 2873 3101 3210 - 1 -2 -4 0 41 120 174 222 305 388 477 459 467 419 187 - 144 - 258 - 363 - 591 - 11302456 Thermodynamics of Alcohol-Hydrocarbon Systems Table A 4. C,H,,0H-C6H,, GlE TSE G~(c,H,,oH) TSE(C,H,,OH) GE(C6H1'2) TSE(C6Hl'2) /J mol-1 /J mol-' x(C,H,,OH) /J mol-l /J mol-I x(C,H,,) /J mol-I /J mol-I 34.3 75.7 105.3 195.0 253.3 412.7 501.2 582.1 7 16.0 841.6 942.4 1022.3 1003.4 896.9 767.9 624.1 465.2 353.1 195.2 80.2 21.4 35.6 60.2 43.0 15.5 - 109.2 - 135.6 - 148.6 -216.0 - 298.3 - 342.4 -414.3 -428.7 - 388.2 -351.4 - 356.1 -268.5 - 206.4 - 126.3 - 60.5 0.003 03 0.007 11 0,010 46 0.021 23 0.030 23 0.059 85 0.078 58 0.101 96 0.148 57 0.198 90 0.306 30 0.399 94 0.501 95 0.603 74 0.699 53 0.794 52 0.859 54 0.898 10 0.946 21 0.981 09 10 379 8 892 8 138 7 126 6 327 4 968 4 532 3 948 3 187 2 672 1589 I 150 745 41 7 205 84 26 10 4 7 8 020 3 987 2 021 - 1 446 -2 567 -2 568 -2 412 -2 092 - 1 747 -1 592 - 989 - 782 - 505 - 257 - 125 - 84 - 26 - 10 -4 -7 0.996 97 0.992 89 0.989 54 0.978 77 0.969 77 0.940 15 0.921 42 0.898 04 0.851 43 0.801 10 0.693 70 0.600 06 0.498 05 0.396 26 0.300 47 0.205 48 0.140 46 0.101 90 0.053 79 0.018 91 3 12 20 44 63 122 157 199 284 386 659 937 1263 1627 2078 2710 3149 3369 3559 3877 -3 7 39 75 96 47 58 72 51 23 - 59 - 169 - 351 - 587 - 878 - 1406 - 1749 - 1929 - 2279 - 2837 Table A 5.C,,H,,OH-C,H, 15.8 20.5 41.4 91.4 130.0 204.2 302.2 373.0 468.1 582.0 712.6 763.8 755.8 708.0 603.5 445.3 292.3 230.4 105.7 68.7 40.4 56.3 117.9 247.4 273.8 364.8 437.6 512.5 592.5 615.2 562.9 526.0 435.3 337.7 264.5 176.4 106.5 71.0 2.1 - 13.0 0.003 42 0.004 45 0.009 09 0.020 61 0.029 97 0.049 30 0.078 22 0.102 01 0.153 98 0.198 66 0.302 84 0.393 96 0.505 81 0.586 63 0.679 09 0.777 94 0.859 53 0.890 77 0.950 95 0.968 34 4571 4551 4459 4223 4037 3675 3 197 2864 2273 1916 1330 910 547 369 160 41 17 13 2 1 11 908 11 608 10 900 8 896 6 202 4 204 2 962 2 295 1406 883 229 25 - 107 - 193 - 80 -41 - 17 - 13 -2 -1 0.996 58 0.995 55 0.990 91 0.979 39 0.970 03 0.950 70 0.921 78 0.897 99 0.846 02 0.801 34 0.697 16 0.606 04 0.494 19 0.413 37 0.320 91 0.222 06 0.140 47 0.109 23 0.049 05 0.031 66 1 1 1 4 9 24 56 90 174 25 1 444 668 968 1188 1541 1858 1972 2000 2123 2136 -1 4 19 65 90 165 223 309 409 548 707 85 1 99 1 1091 994 94 1 867 759 76 - 376A . Pettersson, P. Saris and J . B. Rosenholm 2457 Table A 6. Cl,H,,OH-C,H,, 18.4 39.5 69.0 135.6 174.9 260.7 329.3 423.4 545.3 640.9 764.8 820.4 819.6 759.1 640.2 498.6 367.7 295.5 136.5 52.1 27.5 51.9 113.3 171.2 137.0 96.9 82.3 48.3 8.7 -42.3 - 83.4 - 119.9 - 154.1 -191.1 - 199.3 -212.1 - 143.5 - 120.9 - 77.4 -35.5 0.002 19 0.004 97 0.009 52 0.021 76 0.030 14 0.051 18 0.070 81 0.102 04 0.152 22 0.203 46 0.303 50 0.402 76 0.497 43 0.601 69 0.703 31 0.789 32 0.854 09 0.886 27 0.950 70 0.981 45 8057 7013 6103 4964 452 1 3 809 3379 290 1 2361 1944 1320 958 713 429 176 83 42 39 10 1 12 942 11 386 8 896 2 395 38 - 529 -715 - 741 -681 - 568 - 360 - 334 - 345 - 253 - 106 - 83 - 42 - 39 - 10 -1 0.997 81 0.995 03 0.990 48 0.978 24 0.969 86 0.948 82 0.929 19 0.897 96 0.847 78 0.796 54 0.696 50 0.597 24 0.502 57 0.398 31 0.296 69 0.210 68 0.145 91 0.113 73 0.049 30 0.018 55 1 4 11 28 39 69 96 141 219 307 522 727 924 1257 1739 2055 2 269 2 294 2 562 281 1 - 1 -4 28 121 140 130 143 138 132 92 37 24 35 - 97 -419 - 695 - 733 - 758 - 1362 - 1883 Table A 7. C1,H,,OH-C,Hl8 TSE /J mol-' 27.2 50.3 83.1 149.9 21 1.6 328.3 446.9 541.2 684.1 807.4 93 1.2 953.4 902.2 799.2 651.7 470.2 365.8 251.6 131.4 78.5 36.9 59.7 76.7 152.7 135.1 75.4 - 33.3 - 113.2 - 150.9 - 256.5 - 330.9 - 422.4 -446.5 - 378.6 - 293.2 -282.3 -219.8 -151.8 -81.3 - 49.2 0.003 68 0.006 92 0.011 84 0.022 70 0.033 64 0.056 86 0.084 74 0.112 56 0.163 51 0.221 85 0.327 99 0.430 46 0.534 14 0.630 69 0.726 90 0.819 35 0.864 81 0.910 92 0.955 69 0.974 52 7282 6890 6479 5889 545 1 4715 3993 3336 2764 2150 1378 865 529 31 1 147 96 49 24 7 3 10 157 7 589 4 520 2 270 548 -1 595 -2 353 -2 056 -1 754 -1 350 -818 - 465 - 269 - 101 -7 - 96 - 49 - 24 -7 -3 0.996 32 0.993 08 0.988 16 0.977 30 0.966 36 0.943 14 0.915 26 0.887 44 0.836 49 0.778 15 0.672 01 0.569 54 0.465 86 0.369 31 0.273 10 0.180 65 0.135 19 0.089 08 0.044 31 0.025 48 1 2 6 16 29 63 118 186 277 424 712 1019 1329 1631 1 994 2 165 2 389 2 571 2 814 2 945 - 1 7 23 93 120 176 181 133 162 55 - 92 - 389 - 649 - 851 - 1054 - 1125 - 1309 - 1451 - 1684 - 1795
ISSN:0300-9599
DOI:10.1039/F19868202435
出版商:RSC
年代:1986
数据来源: RSC
|
19. |
Isothermal and non-isothermal molecular gas transport in model non-homogeneous porous adsorbents |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 82,
Issue 8,
1986,
Page 2459-2471
John H. Petropoulos,
Preview
|
PDF (883KB)
|
|
摘要:
J. Chem. SOC., Faraday Trans. I , 1986, 82, 2459-2471 Isothermal and Non-isothermal Molecular Gas Transport in Model Non-homogeneous Porous Adsorbents John H. Petropoulos Physical Chemistry Laboratory, Democritos Nuclear Research Centre, Aghia Paraskevi, Athens, Greece The evaluation of isothermal and non-isothermal transport parameters for dilute gases in axially or radially non-homogeneous porous adsorbing diaphragms has been considered on the basis of the molecular transport approach of Nicholson and Petropoulos. The behaviour of model porous adsorbents with realistic axial or radial variation in porosity have then been considered and the salient features of the deviations from homogeneous medium behaviour have been investigated and illustrated with suitable examples. Analogies and differences between the effects of non-uniform porosity and of non-ideal pore structure are discussed.Particular attention has been paid to the effect of axially non-uniform porosity on the evaluation of heats of transport by means of integral-type experiments. The results obtained are of particular significance for the interpretation of experimental results. A common approach to isothermal and non-isothermal molecular gas flow in finely porous adsorbents is based on analysis of the relevant phenomena in idealised single pores.l-' The deviation of the observed transport parameters from the ideal values is then attributed to the complicated pore microstructure of real porous media, such as pore tortuosity, variation in pore size and shape etc., which is embodied in a suitable ' structure factor'.In a recent paper,s we pointed out that such interpretations are bound to be incomplete in the vast majority of cases because they ignore the macroscopic non- homogeneity of the porous solid test samples caused by the method of their preparation (compaction or other pelletization p r o c e ~ s ) . ~ - ~ ~ The fact that macroscopic non- homogeneity can contribute materially to the aforesaid deviations of real isothermal flow parameters from the ideal values was then demonstrated by means of suitable model calculationss and experimental data. l3 More particularly, ref. (8) was chiefly concerned with the effect of axially or radially non-uniform porosity on isothermal non-adsorbable and adsorbable gas molecular flow.For the latter case, the conventional surface flow treatment1-3. 6* was used. In the present paper, a recently developed, more refined and rigorous approach1*? l5 is employed for this purpose. The same methods are then applied to the study of the effect of axially and radially non-uniform porosity on non-isothermal molecular gas flow. Theory We consider steady-state permeation in the axial direction, X , through a porous diaphragm in the form of either a slab of dimensions Ix = 1, lq, I, or a cylinder of length 1 and radius 1,. At any position 0 < X < I in the porous medium, we have: (a) Isothermal flux J = - A , PdC,ldX where C,(X = 0) = c,"; C,(X = I ) = cgp 24592460 (b) Non-isothermal flux Gas Transport in Non-homogeneous Media where Cg(X = 0) = cgo; C,(X = I ) = c,, T(x = 0) = 6; T ( X = I ) = q.In eqn (1) and (2), C, denotes gas concentration in the gas phase and T the temperature (the corresponding boundary values Cgo, C,, and To, & being constant in any one experiment); qc = q,/RT, where qc is the (differential) heat of transport and R is the gas constant; P is the isothermal permeability of the gas; and A , = I, I, (slab) or A , = 711; (cylinder). We restrict ourselves to the Knudsen flow regime (assuming perfectly diffuse reflection of the gas molecules at the pore walls) and the Henry law region of adsorption. Under these conditions, P is independent of C, and, for a homogeneous medium, may be determined directly from an experimental measurement over a suitable concentration interval ACg = Cgo-Cgl by straightforward integration of eqn (1).In the case of a non-homogeneous medium, this operation will yield an effective isothermal permeability coefficient, the value of which will depend on the spatial variation of P in the manner indicated by Nicholson and Petropoulos.8 A convenient way of determining 4, is to impose a temperature difference across the diaphragm AT = To- q and then allow a dynamic equilibrium (JT = 0) to be established through the build-up of a corresponding concentration difference AC, = Cgo - Cgl. Since 4, is generally a function of T, integration of eqn (2) for a homogeneous medium yields an integral heat of transport P = Jl/A, ACg (3) which reduces to 4, only when AT--+ 0 (differential-type experiment). In practice 4, is often determined from a series of integral-type experiments in which and C,, are kept constant and T, is varied (cf., e.g. Ash et al.3).Then If the above procedures are applied to a non-homogeneous medium, the resulting effective differential heats of transport, ;,( T,), may reasonably be expected to depend on the spatial variation of P, qc and dT/dX (in the manner indicated below). The spatial variations of P etc. are, in turn, determined by the macroscopic structural non-homogeneity of the porous medium, or, more specifically, by the variation of the local porosity ( E ) and pore As already mentioned, the effect of pore structure is embodied in the appropriate structure factor K , which is defined as the property of interest of the porous solid relative to that of an idealised single pore of hydraulic radius r h equal to that of the porous medium.A cylindrical reference pore is mostly chosen, but other choices are, of course, possible.lg 6 y l6 The local isothermal permeability of a non-absorbed gas (subscript g) is given by8 l1 Pg = I C ~ BErh = I C ~ Bc2/A0( 1 - E ) where A , is the specific surface area exposed to gas within the porous medium per unit volume of the solid material, which may be taken as essentially constant;ll B involvesJ . H . Petropoulos 246 1 the mean molecular gas speed multiplied by a numerical factor depending on the chosen cross-sectional shape and length to radius ratio of the reference pore1$ l6 and E and K~ are local values. The overall isothermal permeability of the non-homogeneous medium is similarly given by Fg = k, BE2/Ao( 1 - 21) where 21 and t, are the experimental overall porosity and gas-phase diffusion structure factors, respectively.In the context of the theory of isothermal absorbable gas flow of Nicholson and P e t r o p o u l o ~ , ~ ~ ~ l5 the conventional distinction between ‘ gas-phase’ and ‘ surface ’ perme- abilities used in our previous paperE is not very meaningful. Accordingly, the local permeability coefficient is formulated here as (6) P = Pg 4 = P, ~4 de = I C ~ I C ~ BE^^^ / A o ( 1 - E ) (7) where eqn ( 5 ) has been used; de is calculated for the appropriate reference pore according to Nicholson and Petropou1os;l4~ l5 and I C ~ is the relevant structure factor, which is a function16 of pore geometry and of Do = Uo/RT; U,, measures the adsorbability of the gas (it represents the depth of the adsorption potential well at a single solid surface for the given gas).The overall observed value of 4 which characterizes a non-homogeneous medium is then given by (8) = PAo( 1 - qpg BE2. Non-isothermal flow of a non-adsorbed gas is characterized by qc = qcg = f, inde- pendently of porosity and pore structure.17 In the presence of adsorption, however, qc contains an additional pore-structure dependent term.18 Hence, the local heat of transport of interest here may be written qc = + + I C ~ ( ~ , , -f), where qce is calculated for the appropriate reference pore according to Nicholson and P e t r o p o u l o ~ ~ ~ and I C ~ is the corresponding structure factor (which, like ?c4, is also a function of Do).It is worth noting in passing that the gas-phase structure factor may be written as = K , R,, where IC, is an ‘orientation’ or ‘ anisotropy ’ factor (equal to for an isotropic medium) and R, contains contributions from all other pore structural features6 The ideal value of zg, rcCI and K~ is unity. Furthermore, it is obvious that I C ~ = 1 for a non-adsorbed gas. The evaluation and properties of kg for radially and axially non-homogeneous model porous media have been described in our previous paper.E Expressions for the calculation of 4 and tc in such media are given below. We note at the outset that, since qcg is independent of E and of pore structure, we must have GCg = qcg = t. Radially Non-homogeneous Medium Here, eqn (1) and (2) are modified tos J = - A c J o l P s d w dx JT = - A c ~ l P C , ( ~ + q c - - - & ) d w d In T 0 where w = y (slab) or w = y 2 (cylinder).(6)-(8), it is easily shown (cf. Nicholson and PetropoulosE) that Bearing in mind that dC,/dx and dT/dx are independent of y and using eqn (3) and f l / fl2462 Gas Transport in Non-homogeneous Media The corresponding expression for the effective heat of transport is Gc = Jol qc dw/jol Pdw 1 jol Kg K+ K Q [&'/(I -&>I 4e(qce -i) dw = -+ (10) Jo1KgK4[E2/(1 -&)14& and the result of a differential-type experiment, or of a series of integral experiments using eqn (4), should be the same. Axially Non-homogeneous Medium For isothermal flow we have8 6 = i' Pi1 dx/jol where eqn (6)-(8) have been used. 0 In the case of non-isothermal flow, one should bear in mind that the thermoconductivity of the porous solid, G, is also a function of its macroscopic structure (cf.De Vries'O). The thermal flux at any position x is given by d T dx JH = - A , G(x) - = const. Integrating this equation first between limits (x, l), then between (0, l), and dividing, we obtain (12) TW- T = (T,- T ) [1 -~(X)/W)l where F(x) = G(x')-ldx'. iox Application of the method of determining qc described previously then yields wherein T(x) is given by eqn (12). Hence, we get which, for a differential-type experiment (T, -+ q), reduces to It is obvious that eqn (1 3) and (14) will, in general, yield different results. Furthermore, the result of eqn (14) is independent of the direction in which the temperature gradient is applied.Reversal of the direction of the temperature gradient is equivalent to conversion of the functions G(x), qc(x, T,) into their mirror images about the mid-section of the diaphragm. Using asterisks to denote the new functions, we have G*(x) = G( 1 - x), q:(x, T,) =qc(l -x, To), P ( 1 ) = F(l), whence it follows that G: = Gc (cf. Petropoulos et ~ 1 . ' ~ ) . In the case of eqn (13), however, the relation between @(x, T) and qc(x, T) is complicated; furthermore21 P ( x ) = F( 1) - F(x). Hence, in general, <:(To) # ic(q). -J. H. Petropoulos 2463 0.7 0.6 0.5 W IL, 0.4 0.3 0.2 1 I I I I 0 0.2 0.4 0.6 0.8 1 U Fig. 1. Types of spatial variation of porosity examined: (a) c(u) = E" = 0.5(A); (b) ~ ( u ) = ~ ( 0 ) (1 + k , u + k , u2) with k , > O(B, D, F) or k , < O(C, E) and k, = O(B, C , D, E) or k , = -k,(F); emax/emin = l(A), 1.6 (D, E) or 2.4 (B, C, F).The computed d, or 4, in subsequent figures are labelled with the letter denoting the particular ~ ( u ) function used, followed (except for A) by a numerical suffix identifying the different cases considered here, namely: (i) $ or 4, for radially variable porosity with u = y (suffix 1) or u = y2 (suffix 2); (ii) 4, or GP derived from differential-type experiments assuming G = constant, for axially (u s x ) variable porosity (suffix 3); (iii) 4, for axially variable porosity (u = x) derived from differential-type experiments assuming G = G(0)(1 - E ) (suffix 4); (iv) tP for axially variable porosity (u = x) derived from integral-type experiments with G = constant (suffix 5) or G = G(O)(l - E ) (suffix 6).Results and Discussion Computations based on the expressions derived in the previous section were carried out along the lines of our previous study.8 The integrals appearing in eqn (9), (lo), (1 l), (1 3) and (14) were evaluated by Gauss quadrature. It was found more convenient to present the heat of transport results in terms of -4, = 1 -4,. In contrast to our aforementioned earlier study, the evaluation of the local values q5e and -qpe = 1 -qce here is time- consuming. We have shown l5 that computer demands may be drastically reduced, without affecting the salient features of the computed flow behaviour, by modelling pores as two-dimensional slits with a ' triangular ' adsorption potential-energy well. This practice was adopted (and further justified) in our subsequent model studies of pore structure effecW9 and is continued here.More specifically, the local q$e and qpe were computed numerically on the basis of eqn (1 2), (1 3) and (1 5 ) of Petropoulos,ls putting R = &/Aozo(l - E ) , with a pore length to radius ratio equal to 20 and all other specifications unchanged. The axial (u = x) or radial (u = y or u = y,) variation of the porosity was represented, as in our previous studys by (1 5 ) where k , = constant and k, = 0 (lines B-E in fig. 1) or k, = - k , (line F of fig. 1). These functions were chosen in order to represent realistically the axial variation of porosity E(U) = E(O)(l +k, U + k , U 2 )2464 Gas Transport in Non-homogeneous Media 1 0.5 1 - r: - 0 -0.5 I 1 I I I 0 2 4 6 vo = UO/(RT) Fig.2. Examples of computed 4 for radially (Bl, D1, F l ) or axially (B3, D3, F3) variable porosity, in comparison with the corresponding 4, for uniform porosity (A), in the medium to low go range. Labels on curves as defined in fig. 1. in porous adsorbents in the form of pellets produced by one-ended (k, = 0), or symmetrical two-ended (k, = -kl), compaction of a powdered l1 The fabrication c process can also proauce raaiai variation in porosity. ine case or u --= y is appropriate for a porous material made in the form of a sheet: E may increase or decrease from one edge of the sheet to the other (k, = 0), or from the middle of the sheet to the edges (k, = -kl). The porosity of pellets made by compaction of a powder in a cylindrical die (case of u = y 2 ) may also vary significantly from the axis to the peripher~.~.l2 In each case, the computed 4 and 6, are compared with the corresponding Je,Gpe values characteristic of a homogeneous porous medium with E = E, where Z. = Jol e(u)du. (16) Since the aim of the present investigation was to study the effect of non-uniform porosity, ideal pore structure was assumed throughout, thus eliminating I C ~ , I C ~ and ‘c& (where relevant) from the expressions for 4 and 6,. Under these conditions, the behaviour of the homogeneous medium with E = E is given by that of the reference pore, i.e. #e = be(Re), cce = #,,(Re), where Re = rh/zO =-E/A,(l -%)zo. Values of ~ ( 0 ) and kl(k,) were chosen yielding e = 0.5 and different degrees of non-homogeneity, corresponding to E , .J E , ~ ~ = 1.6 or 2.4 (cf. fig. 1). We also used A , z, = 0.5 (corresponding to Re = 2). In the case of axial non-homogeneity, computations were performed for either (i) a linear temperature gradient (G = constant) ; or (ii) a non-linear temperature gradient, assuming for the sake of simplicity that the effective local thermal conductivity of the porous medium is proportional to the local fractional volume of solid, i.e. G(E) = G(E = 0)( 1 - E ) .J . H. Petropoulos I 2465 -'t I I I I I I 5 6 7 8 9 1 0 g o = Uo/(R T ) Fig. 3. Examples of computed 6 as in fig. 2, but in the medium to high Uo region: A (0); Bl (O), Dl (A), F1 (V); €33 (W, D3 (A), F3 (V). Isothermal Flow Examples of the behaviour of $ are given in fig.2 and 3. The most ncticeable features in the lower uo range (fig. 2) are a tendency of the minimum in the 3 us. Uo plot to become - shallower in the case of radially variable porosity (lines B 1, D 1, F I), and to shift to lower Uo values in the case of axially variable porosity (lines B3, D3, F3), by comparison with 4, (line A). In the higher go region, (fig. 3), 3 for axially or radially variable porosity tends to lie above or below $, respectively. These data have been plotted in the form appropriate to the analytical expression found to apply to a good approximation, at sufficiently high Vo and low R, to single model pores of the type used here, namely18 4 = 1 + K exp (auo) = 1 +KO R-" exp (ago) 6 = 1 +I? exp (atTo) (17) where KO x 0.045, rn z 2.1 and a x 0.8.Introduction of eqn (17) into eqn (9) (with 7cg = constant, q = 1) yields (174 where Eqn (1 7a) also applies to the case of E = 2. with I? equal to Re % KO [Aozo( 1 -E)/E]m. Since m z 2 in eqn (18) and (18a) it follows [for proof, see Appendix in ref. (S)] that2466 Gas Transport in Non-homogeneous Media I I I 0 1 2 3 L Go = Uo/(RT) or Uo/(RTo) Fig. 4. Examples of computed 4 and 4, for radially (Bl, B2) or axially (B3, B4, B5) non-uniform porosity, in comparison with the corresponding ge and &,e for uniform porosity (A). Labels on curves as defined in fig. 1. Conditions applicable to B5 as specified in fig. 5. Hence, for values of Do for which eqn (17) is applicable to sufficient approximation in the range emin < E < the effect of radially non-uniform porosity is to shift the In (4- 1) us.Do plot downward without changing its slope. The relevant data of fig. 3 conform to this conclusion reasonably well. Comparable simple expressions cannot be obtained in the case of axially variable porosity. As shown in fig. 3, the respective In (4 - 1) us. Do plots exhibit a fairly extensive nearly linear region of slope appreciably les than a. (A tendency for the slope to increase at very high Do is evident and can be inferred analytically; but this is hardly likely to be relevant from the practical point of view, because of breakdown of the dilute gas assumption inherent in the basic treatment.)l59 l6 All the above effects of radially or axially non-uniform porosity closely resemble the pore-structure effects represented by the parallel (P) or serial (S) capillary models, respectively, investigated by Nicholson and Petropoulos.16 Also, we note that the deviation of d, from 4, revealed by the present calculations depends more strongly on the degree of non-homogeneity (cf.lines D1, D3 with lines B1, B3, respectively) than on the functional form of E(U) (cf. lines BI, B3 with lines F1, F3, respectively). This conforms to the conclusions drawn from our earlier study based on the conventional theory of adsorbable gas flow.* Heats of Transport Examples of the behaviour of the computed &,, in comparison with the corresponding tpe are given in fig. 4-6. Here again, the effect of radial non-uniformity in porosity corresponds to that of P-type pore structure; but axial variation in E can also produce effects which have no counterpart in S-type pore structure, as a result of (i) non-linear temperature gradients, or (ii) derivation of 4, from integral-type experiments.Effect (i) is illustrated by lines B3, B4 (fig. 4 and 5 ) and F3, F4 (fig. 6) for Gp derived from experiments of the differential type and by lines B5, B6 (fig. 5 ) for GP determinedJ . H. Petropoulos 2467 1 1 2 1 L TOIT, Fig. 5. Examples of computed 4, for, radially (Bl,B2) or axially (B3, B4, B5, B6) non- uniform porosity (in compariso_n with gpe for uniform porosity denoted by A) demonstrating the temperature dependence of gp, determined by differential (T, + q; lines Bl-B4) or integral- type (& = constant, &/& % 1-1.6; lines B5, B6) for gases of different adsorbability (U,,).Labels on curves as defined in fig. 1. U,,/R&: (a) 0.5, (b) 2.0, (c) 4.0. from experiments of the integral type. In all cases, the Gp us. O,, or TJT plot, corresponding to the linear temperature gradient, suffers an appreciable downward shift. Effect (ii) is the one more worthy of attention, because it introduces important new features. More specifically ip determined at T, via eqn (4) from an integral-type experiment is not a function of Uo/RT, alone; but depends on both Uo/RT, (or U,/RT) and ZJT, as illustrated in fig. 4 and 5 (lines B5, B6). The deviation from the corresponding &, derived from a differential-type experiment at T, (lines B3, B4, respectively), can be marked and increases with increasing TJT. It is accentuated as the degree of non-homogeneity becomes more marked (cf.lines D3, D5 us. B3, B5 or lines E3, E5 vs C3, C5 in fig. 6). Also, both the magnitude and the direction of the said deviation depend on the functional form of E ( X ) (cf. the line pairs B3, B5; C3, C5; F3, F5; F4, F6 in fig. 6). In particular, substantial positive (lines C3, C5) or negative (lines B3, B5) deviations are produced when the porosity increases in the direction of increasing or decreasing temperature, respectively. In the symmetrical function F, these opposing tendencies very nearly cancel out, leaving only a small net deviation (lines F3, F5 or F4, F6). Bearing in mind that reversal of the direction of the temperature gradient is2468 Gus Transport in Non-homogeneous Media o.8 t 1 1.2 1.b TOIT, Fig.6. Examples of computed &,, as in fig. 5, for axially non-uniform porosity, illustrating the effect of the functional form of E(X) and of the degree of non-homogeneity on the discrepancy between ip determined by differential (B3, C3, D3, E3, F3, F4) and integral-type (B5, C5, D5, E5, F5, F6) experiments, respectively. Labels on curves as defined in fig. 1 ; U,/RT = 4. equivalent to conversion of function B(D) into C(E) or vice versa, whilst F is unaffected (cf. theoretical section), we conclude that the difference between heats of transport determined from differential- and integral-type experiments can be expected to be comparatively large (small) and sensitive (insensitive) to the direction of the temperature gradient, in the case of non-homogeneous porous diaphragms formed by one-ended (two-ended) powder compaction.Relation between 5, and 4 The relation between &, and 6 is of considerable interest. Nicholson and Petropo~losl~ showed that for single pores - 1 dlnq5 4 =--- 2 d 1nT' - 4 = 1-- This means that -i&, 2 according as d4/dT 5 0. Insertion of eqn (19) into eqn (10) or (14) and differentiation of eqn (9) or (1 1) (assuming icg = constant, K~ = 1, K& = 1 in all cases) shows that the above relation is recovered, i.e. in the case of radial non-homogeneity, but not in the case of axial non-homogeneity;J . H . Petropoulos 2469 as expected from the corresponding results for P-type and S-type pore structure, respectively, reported in a recent paper.18 In the said paper it was further shown that the breakdown of eqn (19) in S-type model porous media gives rise to a region in no in which -pp < $ is associated with d4/dT < 0, in line with what is found in the extensive data reported by Ash et aL3 The same arguments can be used in the case of axially variable porosity for tp determined by differential-type experiments with linear temperature gradient.A relevant example is afforded by lines B3 in fig. 4, where -<, < h, d$/dT < 0 in the range Uo z 0.7-1. The present calculations further show that this condition may be realized over a considerably wider go range, under suitable circumstances (function B), if tp is determined by integral-type experiments (cf. line B5 for Uo/R& = 2 in fig. 4 and 5). This result is of particular importance, in view of the fact that the porous diaphragms of Ash et aL3 were constructed by a one-ended powder compaction procedure (cf.Savvakis and Petropoulos12 for a fuller discussion of this procedure) and the relevant heats of transport were determined by means of integral-type experiments. One may, therefore, justifiably point to axially non-uniform porosity as an additional cause (possibly even the principal one) for the -tP < 4, d4/dT < 0 correlations shown by the experimental data in question. Conclusion The first task of the present paper was to develop the formalism required for a reasonably rigorous evaluation of isothermal Knudsen gas permeabilities and heats of transport in axially or radially non-homogeneous porous adsorbents. For this purpose, our previous treatment of isothermal permeability for such porous media8 was combined with our fundamental treatment of isothermal Knudsen gas flow in model pores14* l5 and the requisite heat of transport formalism was similarly developed on the basis of our corresponding fundamental approach to non-isothermal Knudsen flow? Particular attention was paid to the complications introduced by axial non-homogeneity into the determination of heats of transport as a result of (i) non-linear temperature gradients and (ii) the use of integral-type (rather than differential-type) experiments.The most important conclusion in this respect is that, GP determined from integral-type experiments not only differs from that derived from experiments of the differential type, but generally also depends on the direction in which the temperature gradient is applied across the diaphragm.Our second task was to assess realistically the nature and magnitude of the effect of non-uniform porosity on 4 and &,, likely to be encountered in practice. This was done by means of appropriate model calculations. The results for 8 indicate that the dependence on Do remains qualitatively the same as in a homogeneous porous medium, although there may be considerable quantitative differences. The nature of the latter corresponds closely to what was found in our study of pore structure;l6? l8 radially or axially non-uniform porosity being the counterparts of P-type or S-type pore structure, respectively. The above conclusions apply also to &,, with the reservation that the additional effects (i) and (ii) mentioned above must be considered in the case of axial non-homogeneity.Effect (ii) is of particular practical importance, because it destroys the simple relation between the dependence of GP on gas adsorbability and on temperature. It is also noteworthy that the correlation -pp 8 t, dd/dT 5 0 deduced for single pores, remains valid in the case of P-type pore structure or radial non-homogeneity, but breaks down in the case of S-type pore structure or axial non-homogeneity. The realisation that the breakdown of this correlation in the latter case can be considerably exacerbated by effect (ii) is important for the interpretation of this aspect of the data of Ash et aL3 On the other hand, there are other aspects of the aforesaid data (cf.the unusually low values of qp reported for a Graphon diaphragm) which cannot be properly explained at present. This is not surprising, in view of the fact that our model calculations are based2470 Gas Transport in Non-homogeneous Media on highly idealised pore and surface geometries15 and that pore structure and axial and radial non-homogeneity are, in practice, combined in complicated ways. l2 We conclude that the results of the present paper, in conjunction with those of our previous related papers, have yielded fundamental new insights into both isothermal and non-isothermal gas transport in porous adsorbents. It seems likely that further work along these lines will provide answers to some at least of the remaining questions. Thanks are due to Dr D.Nicholson for useful discussions and to Mr J. Petrou for help with the computations. Glossary cross-sectional area of porous diaphragm (m2) specific surface area defined in eqn (5) (m-l) parameter defined in eqn (5) (m s-l) gas concentration in the gas phase (mol mA3) upstream and downstream boundary values of C,, respectively function defined in eqn (1 2) thermal conductivity of porous solid (W m-l K-l) isothermal or non-isothermal flux of gas (mol s-l), respectively constants defined in eqn (1 5 ) thickness of porous diaphragm (m) width (slab) or radius (cylinder) of porous diaphragm (m) breadth of slab (m) gas permeability (m2 s-l) differential heat of transport (J mol-l) defined with reference to gas concen- tration or pressure, respectively integral heat of transport (J mol-l) hydraulic radius (m) dimensionless pore radius (in units of zo) gas constant (J mol-1 K-l) temperature (K) upstream and downstream boundary values of T, respectively variable defined in eqn (1 5) maximum depth of adsorption potential-energy well (J mol-l) variable defined in eqn (1 a), (2a) axial coordinate (m) radial coordinate normalised with respect to ly(O < y < 1) maximum width of triangular adsorption potential-energy well (m) porosity structure factors for Pg, 4 and qc, respectively normalised permeability of adsorbable gas XI, (0 < x < 1) value relating to reference pore or medium parameter pertaining to non-adsorbed gas Superscripts - N parameter normalised with respect to RT overall effective value characterising a non-homogeneous porous medium * parameter pertaining to reversed temperature gradientJ .H. Petropoulos 247 1 References 1 R. M. Barrer, Appl. Muter. Res., 1963, 2, 129. 2 C. N. Satterfield, Mass Transfer in Heterogeneous Catalysis (M.I.T. Press, Cambridge, Massachussetts, 1970). 3 R. Ash, R. M. Barrer, J. H. Clint, R. J. Dolphin and C. L. Murray, Philos. Trans. R. SOC. London, Ser. A, 1973, 275, 255. 4 D. Nicholson and K. S . W. Sing, Specialist Periodical Report, Colloid Science (Royal Society o f Chemistry, London, 1979), vol. 3, p. 1. 5 E. R. Gilliland, R. F. Baddour and H. F. Engel, AIChE J., 1962,8, 530. 6 D. Nicholson and J. H. Petropoulos, J . Phys. D, 1971, 4, 181. 7 D. Nicholson and J. H. Petropoulos, J . Phys. D, 1975, 8, 1430. 8 D. Nicholson and J. H. Petropoulos, J . Chem. Soc., Faraday Trans. I , 1982, 78, 3587. 9 C. G. Goetzel, Treatise on Powder Metallurgy (Interscience, New York, 1949), vol. 1, chap. 8 and 9. 10 C. N. Satterfield and S . K. Saraf, Ind. Eng. Chem. Fundam., 1965, 4, 451. 1 1 J. H. Petropoulos and P. P. Roussis, J . Chem. Phys., 1968, 48, 4619. 12 C. Savvakis and J. H. Petropoulos, J . Phys. Chem., 1982, 86, 5128. 13 C. Savvakis, K. Tsimillis and J. H. Petropoulos, J . Chem. Soc., Faraday Trans. 1, 1982, 78, 3121. 14 D. Nicholson and J. H. Petropoulos, J . Colloid Interface Sci., 1979, 71, 570. 15 D. Nicholson and J. H. Petropoulos, J . Colloid Interface Sci., 198 1, 83, 420. 16 D. Nicholson and J. H. Petropoulos, J. Colloid Interface Sci., 1985, 106, 538. 17 D. Nicholson and J. H. Petropoulos, J . Membr. Sci., 1981, 8, 129. 18 J. H. Petropoulos, to be published. 19 D. Nicholson and J. H . Petropoulos, J. Colloid Interface Sci., 1981, 83, 371. 20 D. A. De Vries, Bull. Inst. Int. Froid, Annexe 1952-1, 1952, 115. 21 J. H. Petropoulos, P. P. Roussis and J . Petrou, J . Colloid Interface Sci., 1977, 62, 114. Paper 5/ 165 1 ; Received 23rd September, 1985
ISSN:0300-9599
DOI:10.1039/F19868202459
出版商:RSC
年代:1986
数据来源: RSC
|
20. |
The Dubinin–Radushkevich–Kaganer equation |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 82,
Issue 8,
1986,
Page 2473-2479
Joaquin Cortés,
Preview
|
PDF (534KB)
|
|
摘要:
J . Chem. SOC., Furaday Trans. I, 1986, 82, 2473-2479 The Dubinin-Radushkevich-Kaganer Equation Joaquin Cortes* and Paulo Araya Facultud Ciencias Fisicas y Matematicas, Universidud de Chile, Santiago, Chile The Dubinin-Radushkevich-Kaganer (DRK) model is discussed in general in terms of the replacement of Polanyi’s concept of an adsorption space by a bidimensional concept that leads to a relation between that model and the adsorptive energy distribution. Since the model assumes a Rayleigh distribution, it is found that the usual criteria of the experimentalists, such as the adaptation of the DRK equation to the experimental data, are not sufficient. These general ideas are illustrated with simulations of some representative cases of actual laboratory situations. ~ _ ~ _ _ _ _ ~ ~ _ ~ _ _ ~ ~ ~~ ~~~~ The history of the Dubinin-Radushkevich-Kaganer (DRK) equation dates back to 19 14, when Polanyil proposed his potential theory, which interpreted adsorption phenomenon by means of a model that takes into account the adsorption space in the vicinity of the solid.That space is characterized by a series of equipotential surfaces, each having a definite value of the adsorption potential E , defined as the work needed to transfer a molecule from the gas phase in equilibrium with the solid at pressure P to a volume W of the adsorption space. The value of E , which is therefore a free energy of adsorption, is given by where Po is the saturation pressure of the adsorbate at temperature T. Polanyi also added a postulate that assumes that the function W =f(~), known as the characteristic curve, is independent of temperature. Later, Dubinin2 introduced an affinity coefficient that allows the superposition of the characteristic curves of different gases on the same solid.Additionally, by proposing a relation between the adsorption space located in the solid’s micropores and the adsorption potential, he derived the well known Dubinin-Radushkevich equation3 for the adsorption isotherm : E = RT In (PIP,) (1) where Wo is the total volume of the micropores. Thus, eqn (2) became the basis of the Dubinin-Radushkevich micropore volume-filling theory. Dubinin’s original idea was later modified by Kagane~,~ who suggested the validity of eqn (2) for the case of non-porous solids, where 0 would now be the fraction of covered surface, n,/n,, with n, representing the number of adsorbed moles per gram of solid, and n , those corresponding to the monolayer.This implies implicitly replacing Polanyi’s space of adsorption concept by a less precise bidimensional concept, as will be discussed later. The DRK equation written in a linearized form then becomes Inn, = Inn, - BR2T2 In (Po/P)2. (3) Eqn (3) drew a lot of interest following Hobson’s w ~ r k , ~ ? ~ which showed its applicability to physisorption on non-porous solids in the ultra-high-vacuum region. Later, Cerofolini7 did some work that showed its importance in the study of heterogeneous solids. 82 2473 F A R 12474 The DRK Equation While Dubinin’s model assumes an energetic distribution of the solid’s micropores, Kaganer’s assumes implicitly an energy distribution over the surface.If the heterogeneous surface is thought of as a collection of homogeneous patches, in each of which the covered fraction is O,, the overall adsorption isotherm is given by8 O(P) = O,(P, E)f(E) dE JOE (4) wherefTE) dE is the fraction of the surface having energies between E and (E+dE). If, additionally, the condensation approximation for 0 [ref. (9)] is considered, eqn (4) becomes6? In terms of the adsorption potential E , it can be shown in this case that the DRK equation assumes the Rayleigh distribution given by f(~) = -2Be exp BE^). (6) The DRK equation is used by experimentalists because it is the only method available for the determination of n, from the experimental isotherm in the low relative pressure, PIP, region, where other models, such as the BET.model,1° have no validity. This is important when gases are used at temperatures that imply high Po values. However, since the model is, in general, incorrect (except in fortuitous situations that must be corroborated in every case using criteria that are not normally applied), there is uncertainty as to the contradictory results found in the literature. This has led to involved interpretations, such as certain mathematical explanations designed to show the relation between the B.E.T.lO and the DRK4 values of the monolayer,ll or physical explanations such as Hobson’s, which interprets Kaganer’s nm as corresponding to the ‘number of molecules adsorbed on a surface before adsorbate-adsorbate interactions become significant 7 .5 9 l2 Kaganer’s Model The modification proposed by Kaganer, which involves changing in the model the idea of a volume by one of a surface, carries implicitly, however, a deeper theoretical meaning. In Polanyi’s original theory there is the idea of an attractive field that acts on the adsorbate molecules from the solid’s surface. Thus, an attractive energy as a function of distance leads directly to the idea of an adsorption space and, therefore, to W as a volume. This idea is still retained in the micropore volume of the Dubinin-Radushkevich model, but is lost in Kaganer’s because the latter is valid in the low-adsorption region, where the existence of multilayers can be neglected. This leads to the concept of a bidimensional adsorption space that, after some additional assumptions, can be related to the idea of fractions of the surface having an adsorptive energy distribution.It was stated above that the DRK equation corresponds to the Rayleigh distribution. The obvious consequence is that, in so far as the energy distribution on the surface is not that described by Rayleigh, the isotherm will not be interpreted by the DRK equation. Moreover, as has been pointed out by Marsh and Rand13 in their analysis of the DR method as a theory the filling of micropore volumes, a scan over the full range of equilibrium pressures is required to take into account the whole surface. This is an interesting point since, for experimental reasons, it is improbable that data for the whole range of pressures will be available, introducing, therefore, one more uncertainty in the experimental results.To illustrate these points, a simulated analysis of some specific cases will be made, so that it may become possible to visualize the degree of reliability of the parameters ofJ . Cortks and P. Araya 2475 i - 1 2 H E - z - 0 8 2 - - 0 L - I I I I I I I I I 1 2 L 6 8 1.0 1 L 1 8 e/kcal mol-’ Fig. 1. Solid A. (a) Adsorptive energy distribution curve ( y = 10). (b) Adsorption isotherm. (c) Energy distribution : (I) condensation approximation ; (11) Rayleigh distribution for Kaganer’s isotherm in the range 10-3-10 mmHg; (111) same as (11), but in 10-3-0.88 mmHg range; (IV) same as (11), but in 0.88-10 mmHg range. ( d ) Covered fraction us. adsorption potential: (I) simulated isotherm; (11) Kaganer’s isotherm in 10-3-10 mmHg range; (111) same as (11), but in 1OP3-0.88 mmHg range; (IV) same as (11), but in 0.88-10 mmHg range.Kaganer’s model, such as n,, as well as the degree of validity of the usual criteria (e.g. the adaptation of the model to the experimental data, the range of application etc.) applied by the experimentalists. In order to carry out this analysis, a heterogeneous solid will be simulated having a Gaussian energy distribution of the same form as that proposed by the Ross-Olivier mode114 and obeying the equation 1 fTE) = exp [ - y(E - (7) where y is a measure of the heterogeneity of the distribution, E is the mean value of the adsorption energy, and n is a normalization factor that can be calculated for each y from the expression jomflE)dE = 1.(8) Consider the case ofa solid A having an intermediate heterogeneity with y = 10 and an average energy of 1.5 kcal mol-l, which is shown graphically in fig. 1 (a). If a model is assumed for the local isotherm 01, it is possible to generate, starting from eqn (4) 82-22476 The DRK Equation Table 1. Kaganer’s isotherm (DRK) parameters for the different solids and pressure ranges pressure /mmHg /kcal mol-l n,,(K)/mmol correlationa E = B-112 range m solid A 10-3-10.0 1.062 3.824 0.976 1OP3-0.88 0.916 10.173 0.998 0.88- 1 0 .O 1.558 2.238 0.969 solid B 1 .0&13.74 0.598 1.370 0.99995 l.OG2.98 0.605 1.299 0.999 97 2.98-13.74 0.593 1.406 0.999 996 solid C 10-3-2.103 1.009 2.555 0.986 1 OP3-0 .03 0.896 6.174 0.999 0.03-2.103 1.323 1.482 0.986 ~ a Correlation with the DRK equation.and (7), adsorption isotherms of the form (0, P) at different temperatures, as shown in fig. 1 (b) for the distribution of fig. 1 (a) with 0, given by Langmuir’s isotherm15 P P+& exp(-E/RT) 0, = (9) assuming a value of 8.217 x lo5 mmHgt for 8, which corresponds to nitrogen at 77 K as the adsorbate. Naturally, the analysis and its consequences are similar if other expressions for 0, are assumed, such as the Hill-de Boer equation16 discussed by Ross and Olivier,14 or others. The isotherm generated in the form (0, P) can be expressed in the form ( 0 , ~ ) by means of eqn (1) if the value of Po is assumed to be 680 mmHg, which is the vapour pressure of nitrogen at 77 K. If the condensation approximation is also taken into account it is possible to obtain the formf(E) = [d@/dE, -el from (0, E ) and eqn ( 5 ) and this is shown in curve I of fig.l(c) for the isotherm of fig. l(b). This is a sufficient assumption for the analysis, and it happens to correspond to the first approximation of Adamson’s well known graphical method17 for determining the energy distribution curve. Since in practice it is often only possible to obtain a limited set of these experimental data for an isotherm, the analysis will be made first assuming that the full set of data of 0 us. P is available for the isotherm of fig. l(b), then assuming that only the data for low pressures are available, and finally assuming that only the data for high pressures are available, all within the range recommended4 for the application of the DRK equation.Table 1 shows the DRK parameters obtained in each of the above cases, when the data for the isotherm, expressed as n, us. E (assuming n, = 1.78 mmol g-l) are linearized according to eqn (3). From these DRK parameters and eqn (2), the isotherm can be expressed in the form (a,&) as shown by the corresponding curves 11, I11 and IV of fig. 1 (d). From the condensation approximation and the above curves, the energy distribution curves 11, I11 and IV of fig. l(c) may be obtained, These curves can also be calculated directly from eqn (6). It should be noted that in this analysis a comparison is made of the energy distributions 7 1 mmHg z 133.3 Pa.J. Cortks and P. Araya 2477 I 3.0 mol-’ L LO 0 5 0.7 09 e/kcal moi-’ e/kcal mol-’ Fig. 2.(a) Adsorptive energy distribution curve for solid B. (6) As (a), but for solid C. (c) Covered fraction us. adsorption potential for solid B: (I) simulated isotherm; (11) Kaganer’s isotherm in full range (1-13.74 mmHg). ( d ) Covered fraction us. adsorption potential for solid C : (I) simulated isotherm; (11) Kaganer’s isotherm in 10-3-2.103 mmHg range; (111) same as (11), but in 10-3-0.03 mmHg range; (IV) same as (11), but in 0.03-2.103 mmHg range. obtained using the condensation approximation. Therefore, it would make no sense, for example, to use Adamson’s second approximation for determining curve I, or to apply Hobson’s6 or Cerofolini’s7 asymptotically correct approximation for the determination of curves 11, III and IV of fig. 1 (c). The argument that should be stressed is that the DRK equation would describe the system that is being studied only when the energy distribution curves of the simulated system coincide with those obtained from the DRK equation, making the same assumptions.Another minor point refers to the description of the distribution curve which, in fig. 1 (c), is expressed as a function of the adsorption potential, 1, instead of the energy E of the adsorption site. Since both energy forms differ by an approximately constant factor, the latter may be used as a scale factor for the curves themselves. In fig. I (d) it can be seen that curves 11,111 and IV obtained from the DRK equation for the different pressure ranges described in table 1 for solid A do not coincide with curve I for the simulated isotherm, and therefore the determined parameters n and Em do not agree with the simulated ones.This is due to the fact that the simulated energy distributions are not represented faithfully by the Rayleigh distributions of the isotherms adjusted by the DRK equation. This is clearly seen in fig. I (c) when curve I is compared with the others. However, fig. I(c) shows that curve IV intersects curve I at ca. E = 1.5 kcal mol-l. Obviously, if n, is evaluated by means of eqn (3) for2418 The DRK Equation E = 1.5 kcal mol-1 and the Em value that corresponds to curve IV together with the respective pressure and n, values for the isotherm that correspond to that value of E , the original simulated value will be obtained once again. In fig. 2(a), a solid B is simulated, with a distribution similar to that of fig.1 (a), but in a lower-energy region. In this case the DRK parameters of table 1 have been determined within a pressure range that does not include the energy distribution peak. This.situation is of common occurrence in laboratories. A high value is observed in all cases for the correlation of Kaganer’s straight line, but with a value for recovered n, lower than the one assumed in the simulation and independent of the pressure interval employed. Fig. 2(c) shows the 0 vs. E plot for the simulated isotherm as curve I, and that for Kaganer’s isotherm adjusted over the total pressure range as curve 11. This shows that the invariance of the recovered parameters with the pressure interval in this case is not a sufficient criterion for the reliability of the parameters obtained, in spite of what experimenters sometimes say. Finally, a solid C has been simulated having a two-peak distribution with the same heterogeneity, but centred at El = 1.2 kcal mol-1 and E, = 2.5 kcal mol-l. Each peak satisfies an equation of the type of eqn (7), and both together agree with eqn (8).When only the pressure interval between and 2.103 mmHg is used, as is commonly done, then only the energy region covered by the highest energy peak is being considered, i.e. the peak centred at 2.5 kcal mol-l. In the same way as with the other simulated solids, the DRK parameters were obtained for three pressure intervals within the above mentioned range. The values of these parameters and their corresponding pressure intervals are given in table 1.Fig. 2(d) shows the 0 us. E plots for the simulated isotherm (curve I) and for the isotherms adjusted by the DRK equation (curves 11, 111 and IV). Similarly to observations of solid A, the variation of the DRK parameters with the pressure intervals are remarkable. A coincidence exists between curve I of the simulated isotherm and curve 11 of Kaganer’s isotherm obtained for the whole pressure range at their end points, suggesting that an evaluation of n , by means of eqn (3) at those points would recover the original simulated value of n,. Conclusions In a general analysis of Kaganer’s model, the repIacement of PoIanyi’s original idea of an adsorption space by that of a bidimensional adsorption space has been discussed, leading to the establishment of a relation between Kaganer’s model and the energy distribution on the solid’s surface.Since Kaganer’s model implicitly assumes a Rayleigh distribution, the usual criteria of the experimentalists, such as the adaptation of the model to the experimental data, the independence of the recovered parameters with the pressure range etc., are not sufficient; it would also be necessary to check the distribution curve, thereby making subsequent use of the DRK equation less meaningful. Therefore, Kaganer’s model can supply fortuitously the correct parameters, but this is more likely the exception than the rule. These general ideas are illustrated with some particular simulated examples which are representative of situations found in the laboratory.The results for the cases chosen, in spite of providing proof of the incorrectness of the model, are included as specific examples of the general ideas discussed. The uncertainty in the interpretation of the results by means of semiempirical equations, whose obscure theoretical foundations give them a historical character, should convince the experimentalists that they must intensify their attempts to use the modern tools of statistical mechanics to solve adsorption problems. This has been, in fact, the direction given to all the publications coming from this laboratory in recent l9, ‘ OJ . Cortks and P. Araya 2479 References 1 M. Polanyi, Verh. Dtsh. Physik Ges., 1914, 16, 1012. 2 M. M. Dubinin, in Chemistry and Physics of Carbon, ed. P. L.Walker (Arnold, London, 1966), vol. 2, p. 51; M. M. Dubinin, Quart. Rev. (London), 1955, 9, 101; M. M. Dubinin, J . Colloid Interface Sci., 1967, 23, 487. 3 M. M. Dubinin and L. V. Radushkevich, Proc. Acad. Sci. USSR, Phys. Chem. Sect., 1947, 55, 327. 4 M. G. Kaganer, J . Russ. Phys. Chem., 1959, 33, 352. 5 J. P. Hobson and R. A. Armstrong, J. Phys. Chem.. 1963,67,2000; J. P. Hobson, J . Chem. Phys., 1961, 6 J. P. Hobson, Can. J. Phys., 1965,43, 1934; 1941, 7 G. F. Cerofolini, Colloid Sci., 1983, 4, 59. 8 W. A. House, Colloid Sci., 1983, 4, 1. 9 L. B. Harris, Surf. Sci., 1968, 10, 128; 1968, 13, 377; 1969, 15, 182. 34, 1850. 10 S . Brunauer, P. H. Emmett and E. Teller, J . Am. Chem. Soc., 1938,60,305; S . Brunauer, The Adsorption of Gases and Vapours (Clarendon Press, Oxford, 1945). 11 B. A. Gottwald, Surface Area Determination (Butterworths, London, 1970), p. 59. 12 J. P. Hobson, personal communication. 13 H. Marsh and B. Rand, J . Colloid Interface Sci., 1970, 33, 101. 14 0. Ross and J. P. Olivier, J . Phys. Chem., 1961, 65, 608; 0. Ross and J. P. Olivier, On Physical 15 1. Langmuir, J . Am. Chem. SOC., 1918, 40, 1361. 16 T. Hill, J . Chem. Phys., 1946,14,441; J. H. de Boer, The llynamical Character of Adsorption (Clarendon 17 A. W. Adamson and I. Ling, Adv. Chem. Ser., 1961, 33, 51; A. W. Adamson, Physical Chemistry of 18 J. Cortes, Adv. Colloid Interface Sci., 1985, 22, 151. 19 J. Cortes, H. Puschmann and E. Valencia, J. Comput. Chem., 1984, 5, 104; J. Cortes, H. Puschmann 20 J. Cortes, G. Troconso, L. Alzamora and E. Valencia, J . Chem. Soc., Faraday Trans. 1, 1985,81, 1631 ; Adsorption (Interscience, New York, 1964). Press, Oxford, 1953). Surfaces (Interscience, New York, 2nd edn, 1967). and E. Valencia, J . Chern. SOC., Faraday Trans. 1, 1983, 79, 1833. J. Cortes, G . Troconso and E. Valencia, J . Chem. SOC., Faraday Trans. 1. 1985, 81, 1637. Paper 511695; Received 30th September, 1985
ISSN:0300-9599
DOI:10.1039/F19868202473
出版商:RSC
年代:1986
数据来源: RSC
|
|