- Email: [email protected]
Surface Titrations of Perlite Suspensions
JOURNAL OF COLLOID AND INTERFACE SCIENCE ARTICLE NO.
207, 90 –96 (1998)
CS985694
Surface Titrations of Perlite Suspensions Mahir Alkan*,1 and Mehmet Dogˇan† *Department of Chemistry, Necatibey Education Faculty, †Department of Chemistry, Faculty of Science and Literature, Balıkesir University, 10100 Balıkesir, Turkey Received March 2, 1998; accepted May 26, 1998
ceutical manufacture. As most perlites have a high silica content, usually greater than 70%, and are adsorptive, they are chemically inert in many environments and, hence, are excellent filter aids and fillers in various processes and materials. Miscellaneous uses of expanded perlite include fillers or extenders in paints, enamels, glazes, plastics, resins, and rubber; as a catalyst in chemical reactions, an abrasive, and as an agent in mixtures for oil well cementing (4). Surface acidity is an important property of hydrous solids to which are closely related the mode and extent of interfacial reactions such as adsorption and coagulation. Many studies have been reported on the titration of hydrous solids. Of note are those on Fe2O3(s), by Atkinson et al. and Parks and de Bruyn; on FeOOH(s), by Atkinson et al.; on SiO2(s), by Bolt, Abendroth, Breeuwsma, and Lyklema and Schindler and Kamber; g-Al2O3(s), by Huang and Stumm, Hohl, and Stumm; on TiO2(s), by Berube and de Bruyn and Yates; on ZrO2(s) and ThO2(s) by Ahmead; on Fe3O4(s) by Ahmead and Maksimov; on CaCO3(s) by Huang; and on Ca10(OH)(PO4)6(s) and Ca10F2(PO4)6(s) by Bell et al. (5). The pHzpc (pH at the zero point of charge) of a series of seven silica–alumina oxide supports, previously heat treated to 500°C, were determined using a potentiometric titration method by Schwarz et al. (6). The pHzpc of this homologous series varied with the weight fraction of the pure components. They calculated the pHzpc of the oxide mixtures, assuming that each oxide would contribute to the pHzpc in proportion to the oxide weight in the binary mixture by using either of the relationships
The surface charge behaviour of unexpanded and expanded perlite samples in KNO3 and NaCl solutions were investigated as a function of pH and ionic strength. The solutions of KNO3 and NaCl ranging from 1023 to 1.0 M were used. The potentiometric titration method was used to determine the surface charge of perlite samples. It was confirmed that the perlite samples had no the point of zero charge and was negatively charged in the pH range of 3–10. The double extrapolation method was used for determining the intrinsic equilibrium constants for simple ionization and complex ionization reactions. The values obtained are int int pKint a2 5 2.5 and p*KK1 5 2.3 in KNO3 solutions and pKa2 5 3.0 int and p*KNa1 5 2.4 in NaCl solutions for unexpanded perlite, and int int pKint a2 5 2.6 and p*KK1 5 2.4 in KNO3 solutions and pKa2 5 2.7 int 1 5 2.4 in NaCl solutions for expanded perlite. and pKNa © 1998 Academic Press
Key Words: perlite; surface charge; intrinsic equilibrium constants; surface titrations.
1. INTRODUCTION
Perlite is a glassy volcanic rock, commonly light gray, with a rhyolitic composition and 2 to 5% of combined water. Commercially, the term perlite includes any volcanic glass that will expand or “pop” when heated quickly, forming a lightweight frothy material. The temperature at which expansion takes place ranges from 1400 to 2000°F (760 to 1100°C); a volume increase of 10 to 20 times is common (1). However, this versatile, lightweight material with its low bulk density continues to grow in popularity even though it is by no means the cheapest (2). Along the Aegean coast, Turkey possesses about 70% (70 3 109 tons) of the world’s known perlite reserves (3). Over half of the perlite produced goes into the construction industry, in particular as aggregate in insulation board, plaster, and concrete. In cryogenic (extremely low temperature) applications, perlite is used to insulate storage vessels for liquefied gas. Expanded perlite is used as a rooting medium and soil conditioner, and as a carrier for herbicides, insecticides, and chemical fertilizers. Accurately sized perlite is used as an aid in filtering water and other liquids, in food processing, and in pharma1
pHzpc(mixture) 5 4.1 1 3.08 f Al2O3 pHzpc(mixture) 5 7.18 2 3.08 f SiO2, where fAl2O3 is the weight fraction of Al2O3 in the mixture and SiO2 is the weight fraction of SiO2 in the mixture (6). In this paper we present experimental data on the potentiometric titrations of perlite samples, and surface acidity constants are reported.
Corresponding author. E-mail: [email protected].
0021-9797/98 $25.00 Copyright © 1998 by Academic Press All rights of reproduction in any form reserved.
90
91
SURFACE TITRATIONS OF PERLITE SUSPENSIONS
100 cm3 of the electrolyte solutions of the appropriate concentrations. The temperature of titration suspensions remained constant at 25°C. Titration data were obtained using an Orion 920-A pH-meter with a combined pH electrode (Ross).
TABLE 1 Chemical Composition of Perlite Constituent
Percentage present
SiO2 Al2O3 Na2O K2O CaO Fe2O3 MgO TiO2 MnO2 SO3 FeO Ba PbO Cr
71–75 12.5–18 2.9–4.0 4.0–5.0 0.5–2.0 0.1–1.5 0.03–0.5 0.03–0.2 0.0–0.1 0.0–0.1 0.0–0.1 0.0–0.1 0.0–0.5 0.0–0.1
3. RESULTS AND DISCUSSION
The silicon atoms at the surface tend to maintain their tetrahedral coordination with oxygen. They complete their coordination at room temperature by attachment to monovalent hydroxyl groups, forming silanol groups. Theoretically, it is possible to use a pattern in which one silicone atom bears two or three hydroxyl groups, yielding silanediol and silanetriol groups, respectively. It is stated as improbable that silanetriol groups exist at the silica surface. The type of silanol groups are shown below (7–9): { O SiO OH }
2. MATERIALS AND METHODS
2.1. Materials
Hydroxyl or silanol groups
The unexpanded and expanded perlite samples were obtained from Cumaovasi Perlite Processing Plants of Etibank (I˙zmir, Turkey). The chemical composition of the perlite found in Turkey is given in Table 1 (3). Treating and characterization of the unexpanded and expanded perlite samples, before using in the experiments, were given elsewhere (7). The results are summarized in Table 2.
{ }
Si
}
OH
{
OH Silanediol groups
OH } O Si O OH { OH Silanetriol groups
The hydrous oxide surface groups in alumina are given as (7, 8, 10):
'AlOOH
or
AAl
} {
OH
OH
2.2. Potentiometric Titrations The surface charge density (s0) on hydrous oxides can be determined by potentiometric titration and calculation of the net uptake of protons by the surface
s 0 5 F~G H1 2 G OH2! 5 F~C A 2 C B 1 @OH2# 2 @H1#!/A,
[1a] [1b]
where s0 is surface charge density (C z m22), F is Faraday’s constant (C z mol21), GH1,OH2 are the moles of H1 or OH2 bound to the suspension surface (mol z m22), C A and C B are the concentrations of strong acid or strong base after each addition during the titration (mol z L21), A is the surface area of the suspension (m2 z L21) (6). The zero point of charge pHzpc is defined as the pH of the suspension at which s0 5 0. To determine the pHzpc of perlite samples potentiometric titrations were made in 1023, 1022, 1021, and 1 M solutions of NaCl and KNO3, 1 g of the perlite sample and 100 cm3 of electrolyte solution were titrated in polyethylene vessels. The suspension was mixed with a magnetic stirrer. Prior to the titration the solution was equilibrated under N2 for 24 h; 0.1 M HNO3 and 0.1 M NaOH were the titrants used. Blank titrations were similarly performed using
The dependence of the surface charge of unexpanded and expanded perlite samples for the pH of the solution for different electrolytes is given in Figs. 1 and 2. They confirm that perlite has no zero point of charge and is negatively charged at all pH in the range 3–11 as found by Dogˇan et al. before (7). It should be noted that perlite is a mixed oxide system such as kaolinite. Despite that perlite is, however, a mixed oxide consisting mainly of SiO2 and Al2O3, the results show that pHpzc is not simply related to the pHpzc of the pure components TABLE 2 Some Physicochemical Properties of Perlite Samples Used in the Study Samples
Nomenclature Zeta potential (mV) CEC (meg/100 g) Density (g/mL) Specifice surface area (g/m2) Surface site density (C/m2)
Expanded perlite
Unexpanded perlite
EP 246.8 33.3 2.2422 2.30 21
UP 223.5 25.97 2.3047 1.22 43
92
ˇ AN ALKAN AND DOG
grinding, this mechanism cannot account for the presence of negative charges at low pH. The most likely source of the negative charge is the occurrence of crystal lattice defects of certain types. Silica tetrahedra and alumina octahedra occur in alternate layers. They are, however, of almost the same thickness and the crystal can, therefore, accommodate some degree of interlayering of silica by alumina. Furthermore, if the number of aluminum atoms is less than that required (;0.15%), the observed charge would be produced. Alternatively, if occasional aluminum atoms in the lattice are surrounded by oxygen atoms the disposition of which tends to be tetrahedral rather than octahedral, then the same effect would be produced (12). A surface charge will develop via the amphoteric ionization of the surface hydroxyl groups according to the following reactions (13, 14):
FIG. 1. Surface charge of perlite samples as a function of pH for NaCl: (a) UP; (b) EP.
possibly due to the formation of various alumino–silicate phases. Although the surface charge of perlite is expected to vary, depending on the source of the sample and the procedure used for cleaning the surface prior to study, it carries a permanent negative charge. The negative surface charge on oxides, originates from the acidic dissociation of the surface hydroxyl groups and leads, subsequently, to the adsorption of cations on oxides (11). However, the origin of negative changes on the surfaces of perlite can be explained as discussed by Hunter and Alexander for kaolinite (12). The charge at the surface cannot be completely accounted for by isomorphous replacement. Ionization of surface groups is of importance at high pH. However, this mechanism cannot account for the observed negative charge at low pH. The usual mechanism cited is the formation of broken bonds at the crystal edges. Although this would account for the increase in cation exchange capacity on
FIG. 2. Surface charge of perlite samples as a function of pH for KNO3: (a) UP; (b) EP.
93
SURFACE TITRATIONS OF PERLITE SUSPENSIONS Kaint1
1 2
SOH L | ; SOH 1 H1 s
[2]
sb, which represents the charge in the mean plane of specifically adsorbed counterions,
Kaint2
; SO2 1 H1 SOH | L s ,
1 2 s 0 5 B~@SOH1 2 # 1 @SOH2 2 Cl #
[3]
2 @SO2# 2 @SO2 2 Na1#),
where the subscript s denotes the surface and the equilibrium constants: Kaint1 5
@SOH#@H1 s # @SOH1 # 2
[4]
Kaint2 5
@SO2#@H1 s # @SOH#
[5]
In addition to these reactions involving protons or hydroxyl ions, electrolyte counterions could adsorp to neutralize the surface charge and account for specific adsorption of electrolyte ion pairs as surface complexes at charged surfaces (for example, in the case of NaCl) (15): int
KNa1
| ; SO2 2 Na1 SO2 1 Na1 S L
[6]
int
1 2
2 S
KCl2
2 SOH 1 Cl L | ; SOH1 2 2 Cl .
int
*KNa1
| ; SO2 2 Na1 1 H1 SOH 1 Na L s
[8]
int
1 s
2 S
1/*KCl2
2 SOH 1 H 1 Cl L | ; SOH1 2 2 Cl ,
[9]
so that
and B 5 10 6F/A, where A is the surface area of oxide available in solution and F is the Faraday constant (13). The electroneutrality condition demands that
s 0 1 s b 1 s d 5 0,
5K zK
int Na1
[10]
int int int *KCl 2 5 Ka1 /KCl2 .
[11]
*K
int a2
[14]
where sd is the diffuse layer charge. The surface species are distributed among the total number of sites available (7, 13, 16): 1 2 N s 5 B~@SOH1 2 # 1 @SOH2 2 Cl # 1 @SOH#
1 @SO2# 1 @SO2 2 Na1#!.
The formation of surface complexes readjust the acid– base equilibrium and affects the surface charge. The surface charge determined from the proton balance, s0, represents the net number of protons released as consumed by all surface reactions and not just the formation of the ionized surface species, [SO2] and [SOH1 2 ]. Increases in electrolyte concentration cause additional adding of counterions until equilibrium is reestablished, subject to the effects of the electrical field. In this manner surface complexation provides a mechanism for the development of the surface charge, in addition to the role of protons and hydroxyl ions. With the surface species defined, it is now possible to write equations for s0, surface charge, and
[15]
The surface site density (N s) can be determined experimentally by various methods; the method given by Hohl and Stumm was used in this study (10). Taking into account that the perlite surface has no zero point of charge, s0 may be considered as the result of charged sites: 2s 0 5 B~@SO2# 1 @SO2 2 Na1#!.
int Na1
[13]
[7]
According to Davis et al. these reactions may be written as surface complex ionization reactions (13), 1 S
2 s b 5 B~@SO2 2 Na1# 2 @SOH1 2 2 Cl #!,
[12]
[16]
At low electrolyte concentrations, i.e., when specific ion binding contributes a smaller amount of charge to the total balance, we may approximate the surface charge as being due to simple acid dissociation (reactions [3]). Then [SO2] may be approximated by s0 from the surface proton balance, s0/B: 2s 0 5 B~@SO2#!.
[17]
For the negative surface, the total fractional ionization may be written (13, 14):
a2 5 2
s0 Ns
[18]
ˇ AN ALKAN AND DOG
94
an essential condition for the appropriate use of the graphical extrapolation method. Conversely, neglecting [SOH1 2 ] and 2 2 2 [SOH1 2 NO3 ] relative to [SO ] and [SO K] for s0 , 0 implies that the obtained value of DpK i is large when the graphical method is used, even if the actual value of the DpK i is small. In general, the double-extrapolation technique (14, 20) is the best available graphical determination method to obtain the adjustable constants, if only titration curves are available. In addition, for surface which contain only negative or positive surface groups, such as for instance the polystyrene latex investigated by James et al. (14) and also the case in our study, the double-extrapolation method seems well suited. For those reasons mentioned above, the double-extrapolation technique proposed by James et al. (14) have been used in order to calculate the constants in this study. The acidity quotient pQ int a 2 can be determined from experimental titration data points at different ionic strengths and plotted as a function of a 2 1 (C salt) 1/ 2 as seen in Figs. 3 and 4. For each ionic strength, curves were extrapolated through
FIG. 3. The variation of the surface ionization activity quotient as a function of surface charge and concentration of supporting electrolyte (KNO3). The condition that s 0 5 (C) 1/ 2 gives pK int a 2 ; (a) UP; (b) EP.
and pK aint2 5 pH 2 log 5 pQ a2 1
a2 eC 0 1 1 2 a 2 2.3 kT
eC 0 . 2.3 kT
[19]
Since the term pQ int a 2 5 pH 2 log( a 2 /1 2 a 2 ) is a function of both surface charge and electrolyte concentration, James, Davis, and Leckie introduced a new kind of double extrapolation plot to obtain pK int a 2 (14). In order to determine the constants from titration data, another method, developed by Westall et al., is an optimalization of the constants using a numerical procedure (17, 18). Koopal et al. (19) stated that for 1 2 s0 , 0 the densities [SOH1 2 ] and [SOH2 NO3 ] can be safely neglected only if DpK i is sufficiently large. Therefore, this is
FIG. 4. The variation of the surface ionization activity quotient as a function of surface charge and concentration of supporting electrolyte (NaCl). The condition that s 0 5 (C) 1/ 2 gives pK int a 2 ; (a) UP; (b) EP.
95
SURFACE TITRATIONS OF PERLITE SUSPENSIONS
TABLE 3 Intrinsic Equilibrium Constants for Perlite Samples and Other Oxides Sample
Electrolyte
pK int a2
p*K Kint1
g-Al2O3 ZrO2 Fe3O4 Magnetite Fe2O3 z H2O SnO2 TiO2 U-kaolin C18-kaolin EP EP UP UP
NaCl KNO3 KNO3 KNO3 NaCl KNO3 NaCl KNO3 KNO3 KNO3 NaCl KNO3 NaCl
8.6 9.0 9.00 10.7 7.9 8.9 6 0.2 4.1 6.7 2.6 2.7 2.5 3.0
7.6 7.2 7.07
int 1 p*K Na
References
8.6 6 0.1
21 22 22 23 24 25 16 26 26 This paper This paper This paper This paper
9.0 6.3 5.8 0 2.9 2.4 2.4 2.3 2.4
Again pQ a2 5 pH 2 log( a 2 /1 2 a 2 ) is a function of log[Na1] and s0. So the intrinsic ionization constant for the surface complex formation of charged sites may be written approximately as
S
int p*K Na 1 5 pH 2 log
D
a2 e~C 0 2 C b! 1 log[Na1] 1 . 1 2 a2 2.3 kT [23]
The results obtained are plotted in this form pH 2 log( a 2 /1 2 a 2 ) as a function of a2 2 log[Na1] in Figs. 5 and 6. The double extrapolation technique, i.e. at constant concentration, extrapolation to a 5 0 and then extrapolation to log C 5 0, int 1 values given in Table 3. If yielded an estimation of the p*K Na the electrolyte used in KNO3, all Na1s must be replaced by K 1 in the equations above. The concentration of ionized surface
the points to the condition that a2 5 0. The extrapolation points themselves are extrapolated to zero electrolyte concentration, i.e., a2 5 0, C 5 0. The obtained pK int a 2 values were given in Table 3. The formation of surface complexes is the principal mechanism by which protons are released (or consumed) by many oxide surfaces in aqueous electrolyte solutions. As can be seen from Table 3, K int a 2 values higher than those given in the literature for different oxides were obtained int in this study. For oxides in general, as pK int a 1 and pK a 2 increase, 1 the acidity of both surface species S 2 OH2 and S 2 OH decreases. In this sense pK int a values are considered as a measure of surface acidity. This means that since the pK int a 2 values of perlite samples are less than those of other oxides given in Table 3, the surface hydroxyl groups on perlite are more acidic than the others. When considered, together with the fact that the surface is negatively charged through all ranges of the pH values studied (i.e., pH 3–10), this result shows that the reaction [3] is in favour of the right-hand side at low electrolyte concentrations. It has been proposed that clay minerals have two types of acidic surface sites: a strongly acidic site, SOH, and a weaker acidic site TOH. This model assumes that the clay surfaces do not develop a positive charge, nor a point of zero charge, in the pH range of 3.5 to 11 considered (20). This correlates with the data in this study. So the ionization reactions of surface sites can be represented as Al(OH)(OH) º Al(OH)O2 1 H1 SiOH º SiO2 1 H1.
[20] [21]
At the higher electrolyte concentrations surface complexes dominate the contribution to surface charge formation, i.e. 2s 0 5 a 2N s 5 B~@SO2# 1 @SO2 2 Na1#! 5 B([SO2 2 Na1]).
[22]
FIG. 5. The variation of the surface ionization activity quotient as a function of surface charge and the logarithm of the electrolyte concentration (KNO3). The condition that s0 5 log C 5 0 gives the surface complex acidity constant p*K int K ; (a) UP; (b) EP.
ˇ AN ALKAN AND DOG
96
@SO2 2 Na1# 5
@SOH#@Na1# int C C x*K Na 1exp@~e 0 2 e b !/kT#. @H1# [25]
int int 1 and *K 1 values obtained can be attributed to The high *K Na K the strength of the electrostatic attraction forces between the highly negatively charged oxide surface and positively charged ions (that is, Na1 or K1).
REFERENCES
FIG. 6. The variation of the surface ionization activity quotient as a function of surface charge and the logarithm of the electrolyte concentration (NaCl). The condition that s0 5 log C 5 0 gives the surface complex acidity int constant p*K Na : (a) UP; (b) EP.
sites, SO2 and SOH1 2 , is generally small in comparison to the surface complexes. This is partly a result of the perturbation of surface equilibria by the electrostatic field. As the surface charge and potential become more negative, the release for a proton by the simple dissociative reaction is opposed by a decreasing electrostatic term, exp(e c 0 /kT), i.e., @SO2# 5
@SOH# int K exp~eC 0/kT!. @H1# a2
[24]
However, the formation of surface complexes, e.g., SO2 2 Na1, is not retarded in an equivalent manner because of the attractive electrostatic term for the counterion, exp(eCb/kT), i.e.,
1. Harben, P. W., and Bates, R. L., “Industrial Minerals Geology and World Deposits,” p. 184. Metal Bulletin Inc., London, 1990. 2. Anonymous, Market Anal. 11, May (1969). 3. Uluatam, S. S., J. AWWA 70, June (1991). 4. Chesterman, C. W., “Industrial Minerals and Rocks,” 4th ed., p. 927. AIME, New York, (1975). 5. Huang, C. P., in “Adsorption of Inorganics at Solid-Liquid Interfaces” (M. A. Anderson and A. J. Rubin, Eds.), p. 181. Ann Arbor, MI, 1981. 6. Schwarz, J. A., Driscoll, C. T., and Bhanot, A. K., J. Colloid Interface Sci. 97(1), 55 (1984). ¨ ., J. Colloid Interface Sci. 192, 114 7. Dogˇan, M., Alkan, M., and C ¸ akir, U (1997). 8. Dogˇan, M., M.Sc. thesis, University of Balikesir Faculty of Science and Literature, Balikesir, Turkey, 1997. [Turkish] 9. Karakas¸, R., Ph.D. thesis University of Ege Faculty of Science and Literature, Izmir, Turkey, 1996. 10. Hohl, H., and Stumm, W., J. Colloid Interface Sci. 55(2), 281 (1976). 11. Ahmed, M. S., J. Phys. Chem. 73, 3546 (1969). 12. Hunter, R. J., and Alexander, A. E., J. Colloid Sci. 18, 820 (1963). 13. Davis, J. A., James, R. O., and Leckie, J. O., J. Colloid Interface Sci. 63(3), 480 (1978). 14. James, R. O., Davis, J. A., and Leckie, J. O., J. Colloid Interface Sci. 65(2), 331 (1978). 15. Yates, D. E., Levine, S., and Healy, T. W., J. Chem. Soc. Faraday Trans. 70, 1807 (1974). 16. Sprycha, R., J. Colloid Interface Sci. 102(1), 173 (1984). 17. Westall, J., and Hohl, H., Adv. Colloid Interface Sci. 12, 265 (1980). 18. Westall, J., and Morel, F., “FITEQL: A General Algorithm for the Interpretation of Experimental Data,” Technical Note No. 19, Ralph M. Parsons Laboratory, MIT, Cambridge, MA, 1977. 19. Koopal, L. K., Riemsdijk, W. H. V., and Roffey, M. G., J. Colloid Interface Sci. 118, 117 (1987). 20. James, R. O., and Parks, G. A., Characterization of aqueous colloids by their electrical double-layer and intrinsic surface chemical properties, in “Surface and Colloid Science” (E. Matijevic, Ed.), Vol. 12, p. 119. Plenum, New York, 1982. 21. Sprycha, R., J. Colloid Interface Sci. 127(1), 12 (1989). 22. Regazzoni, A. E., Blesa, M. A., and Marato, A. J. G., J. Colloid Interface Sci. 91(2), 569 (1983). 23. Blesa, M. A., Figholia, N. M., Marato, A. J. G., and Regazzoni, A. E., J. Colloid Interface Sci. 101(2), 410 (1984). 24. Davis, J. A., and Leckie, J. O., J. Colloid Interface Sci. 74(1), 35 (1980). 25. Houchin, M. R., and Warren, L. J., J. Colloid Interface Sci. 100(1), 278 (1984). 26. Braggs, B., Fornasiero, D., Ralston, J., and Smait, S. R., Clays Clay Minerals 42(2), 133 (1994).
207, 90 –96 (1998)
CS985694
Surface Titrations of Perlite Suspensions Mahir Alkan*,1 and Mehmet Dogˇan† *Department of Chemistry, Necatibey Education Faculty, †Department of Chemistry, Faculty of Science and Literature, Balıkesir University, 10100 Balıkesir, Turkey Received March 2, 1998; accepted May 26, 1998
ceutical manufacture. As most perlites have a high silica content, usually greater than 70%, and are adsorptive, they are chemically inert in many environments and, hence, are excellent filter aids and fillers in various processes and materials. Miscellaneous uses of expanded perlite include fillers or extenders in paints, enamels, glazes, plastics, resins, and rubber; as a catalyst in chemical reactions, an abrasive, and as an agent in mixtures for oil well cementing (4). Surface acidity is an important property of hydrous solids to which are closely related the mode and extent of interfacial reactions such as adsorption and coagulation. Many studies have been reported on the titration of hydrous solids. Of note are those on Fe2O3(s), by Atkinson et al. and Parks and de Bruyn; on FeOOH(s), by Atkinson et al.; on SiO2(s), by Bolt, Abendroth, Breeuwsma, and Lyklema and Schindler and Kamber; g-Al2O3(s), by Huang and Stumm, Hohl, and Stumm; on TiO2(s), by Berube and de Bruyn and Yates; on ZrO2(s) and ThO2(s) by Ahmead; on Fe3O4(s) by Ahmead and Maksimov; on CaCO3(s) by Huang; and on Ca10(OH)(PO4)6(s) and Ca10F2(PO4)6(s) by Bell et al. (5). The pHzpc (pH at the zero point of charge) of a series of seven silica–alumina oxide supports, previously heat treated to 500°C, were determined using a potentiometric titration method by Schwarz et al. (6). The pHzpc of this homologous series varied with the weight fraction of the pure components. They calculated the pHzpc of the oxide mixtures, assuming that each oxide would contribute to the pHzpc in proportion to the oxide weight in the binary mixture by using either of the relationships
The surface charge behaviour of unexpanded and expanded perlite samples in KNO3 and NaCl solutions were investigated as a function of pH and ionic strength. The solutions of KNO3 and NaCl ranging from 1023 to 1.0 M were used. The potentiometric titration method was used to determine the surface charge of perlite samples. It was confirmed that the perlite samples had no the point of zero charge and was negatively charged in the pH range of 3–10. The double extrapolation method was used for determining the intrinsic equilibrium constants for simple ionization and complex ionization reactions. The values obtained are int int pKint a2 5 2.5 and p*KK1 5 2.3 in KNO3 solutions and pKa2 5 3.0 int and p*KNa1 5 2.4 in NaCl solutions for unexpanded perlite, and int int pKint a2 5 2.6 and p*KK1 5 2.4 in KNO3 solutions and pKa2 5 2.7 int 1 5 2.4 in NaCl solutions for expanded perlite. and pKNa © 1998 Academic Press
Key Words: perlite; surface charge; intrinsic equilibrium constants; surface titrations.
1. INTRODUCTION
Perlite is a glassy volcanic rock, commonly light gray, with a rhyolitic composition and 2 to 5% of combined water. Commercially, the term perlite includes any volcanic glass that will expand or “pop” when heated quickly, forming a lightweight frothy material. The temperature at which expansion takes place ranges from 1400 to 2000°F (760 to 1100°C); a volume increase of 10 to 20 times is common (1). However, this versatile, lightweight material with its low bulk density continues to grow in popularity even though it is by no means the cheapest (2). Along the Aegean coast, Turkey possesses about 70% (70 3 109 tons) of the world’s known perlite reserves (3). Over half of the perlite produced goes into the construction industry, in particular as aggregate in insulation board, plaster, and concrete. In cryogenic (extremely low temperature) applications, perlite is used to insulate storage vessels for liquefied gas. Expanded perlite is used as a rooting medium and soil conditioner, and as a carrier for herbicides, insecticides, and chemical fertilizers. Accurately sized perlite is used as an aid in filtering water and other liquids, in food processing, and in pharma1
pHzpc(mixture) 5 4.1 1 3.08 f Al2O3 pHzpc(mixture) 5 7.18 2 3.08 f SiO2, where fAl2O3 is the weight fraction of Al2O3 in the mixture and SiO2 is the weight fraction of SiO2 in the mixture (6). In this paper we present experimental data on the potentiometric titrations of perlite samples, and surface acidity constants are reported.
Corresponding author. E-mail: [email protected].
0021-9797/98 $25.00 Copyright © 1998 by Academic Press All rights of reproduction in any form reserved.
90
91
SURFACE TITRATIONS OF PERLITE SUSPENSIONS
100 cm3 of the electrolyte solutions of the appropriate concentrations. The temperature of titration suspensions remained constant at 25°C. Titration data were obtained using an Orion 920-A pH-meter with a combined pH electrode (Ross).
TABLE 1 Chemical Composition of Perlite Constituent
Percentage present
SiO2 Al2O3 Na2O K2O CaO Fe2O3 MgO TiO2 MnO2 SO3 FeO Ba PbO Cr
71–75 12.5–18 2.9–4.0 4.0–5.0 0.5–2.0 0.1–1.5 0.03–0.5 0.03–0.2 0.0–0.1 0.0–0.1 0.0–0.1 0.0–0.1 0.0–0.5 0.0–0.1
3. RESULTS AND DISCUSSION
The silicon atoms at the surface tend to maintain their tetrahedral coordination with oxygen. They complete their coordination at room temperature by attachment to monovalent hydroxyl groups, forming silanol groups. Theoretically, it is possible to use a pattern in which one silicone atom bears two or three hydroxyl groups, yielding silanediol and silanetriol groups, respectively. It is stated as improbable that silanetriol groups exist at the silica surface. The type of silanol groups are shown below (7–9): { O SiO OH }
2. MATERIALS AND METHODS
2.1. Materials
Hydroxyl or silanol groups
The unexpanded and expanded perlite samples were obtained from Cumaovasi Perlite Processing Plants of Etibank (I˙zmir, Turkey). The chemical composition of the perlite found in Turkey is given in Table 1 (3). Treating and characterization of the unexpanded and expanded perlite samples, before using in the experiments, were given elsewhere (7). The results are summarized in Table 2.
{ }
Si
}
OH
{
OH Silanediol groups
OH } O Si O OH { OH Silanetriol groups
The hydrous oxide surface groups in alumina are given as (7, 8, 10):
'AlOOH
or
AAl
} {
OH
OH
2.2. Potentiometric Titrations The surface charge density (s0) on hydrous oxides can be determined by potentiometric titration and calculation of the net uptake of protons by the surface
s 0 5 F~G H1 2 G OH2! 5 F~C A 2 C B 1 @OH2# 2 @H1#!/A,
[1a] [1b]
where s0 is surface charge density (C z m22), F is Faraday’s constant (C z mol21), GH1,OH2 are the moles of H1 or OH2 bound to the suspension surface (mol z m22), C A and C B are the concentrations of strong acid or strong base after each addition during the titration (mol z L21), A is the surface area of the suspension (m2 z L21) (6). The zero point of charge pHzpc is defined as the pH of the suspension at which s0 5 0. To determine the pHzpc of perlite samples potentiometric titrations were made in 1023, 1022, 1021, and 1 M solutions of NaCl and KNO3, 1 g of the perlite sample and 100 cm3 of electrolyte solution were titrated in polyethylene vessels. The suspension was mixed with a magnetic stirrer. Prior to the titration the solution was equilibrated under N2 for 24 h; 0.1 M HNO3 and 0.1 M NaOH were the titrants used. Blank titrations were similarly performed using
The dependence of the surface charge of unexpanded and expanded perlite samples for the pH of the solution for different electrolytes is given in Figs. 1 and 2. They confirm that perlite has no zero point of charge and is negatively charged at all pH in the range 3–11 as found by Dogˇan et al. before (7). It should be noted that perlite is a mixed oxide system such as kaolinite. Despite that perlite is, however, a mixed oxide consisting mainly of SiO2 and Al2O3, the results show that pHpzc is not simply related to the pHpzc of the pure components TABLE 2 Some Physicochemical Properties of Perlite Samples Used in the Study Samples
Nomenclature Zeta potential (mV) CEC (meg/100 g) Density (g/mL) Specifice surface area (g/m2) Surface site density (C/m2)
Expanded perlite
Unexpanded perlite
EP 246.8 33.3 2.2422 2.30 21
UP 223.5 25.97 2.3047 1.22 43
92
ˇ AN ALKAN AND DOG
grinding, this mechanism cannot account for the presence of negative charges at low pH. The most likely source of the negative charge is the occurrence of crystal lattice defects of certain types. Silica tetrahedra and alumina octahedra occur in alternate layers. They are, however, of almost the same thickness and the crystal can, therefore, accommodate some degree of interlayering of silica by alumina. Furthermore, if the number of aluminum atoms is less than that required (;0.15%), the observed charge would be produced. Alternatively, if occasional aluminum atoms in the lattice are surrounded by oxygen atoms the disposition of which tends to be tetrahedral rather than octahedral, then the same effect would be produced (12). A surface charge will develop via the amphoteric ionization of the surface hydroxyl groups according to the following reactions (13, 14):
FIG. 1. Surface charge of perlite samples as a function of pH for NaCl: (a) UP; (b) EP.
possibly due to the formation of various alumino–silicate phases. Although the surface charge of perlite is expected to vary, depending on the source of the sample and the procedure used for cleaning the surface prior to study, it carries a permanent negative charge. The negative surface charge on oxides, originates from the acidic dissociation of the surface hydroxyl groups and leads, subsequently, to the adsorption of cations on oxides (11). However, the origin of negative changes on the surfaces of perlite can be explained as discussed by Hunter and Alexander for kaolinite (12). The charge at the surface cannot be completely accounted for by isomorphous replacement. Ionization of surface groups is of importance at high pH. However, this mechanism cannot account for the observed negative charge at low pH. The usual mechanism cited is the formation of broken bonds at the crystal edges. Although this would account for the increase in cation exchange capacity on
FIG. 2. Surface charge of perlite samples as a function of pH for KNO3: (a) UP; (b) EP.
93
SURFACE TITRATIONS OF PERLITE SUSPENSIONS Kaint1
1 2
SOH L | ; SOH 1 H1 s
[2]
sb, which represents the charge in the mean plane of specifically adsorbed counterions,
Kaint2
; SO2 1 H1 SOH | L s ,
1 2 s 0 5 B~@SOH1 2 # 1 @SOH2 2 Cl #
[3]
2 @SO2# 2 @SO2 2 Na1#),
where the subscript s denotes the surface and the equilibrium constants: Kaint1 5
@SOH#@H1 s # @SOH1 # 2
[4]
Kaint2 5
@SO2#@H1 s # @SOH#
[5]
In addition to these reactions involving protons or hydroxyl ions, electrolyte counterions could adsorp to neutralize the surface charge and account for specific adsorption of electrolyte ion pairs as surface complexes at charged surfaces (for example, in the case of NaCl) (15): int
KNa1
| ; SO2 2 Na1 SO2 1 Na1 S L
[6]
int
1 2
2 S
KCl2
2 SOH 1 Cl L | ; SOH1 2 2 Cl .
int
*KNa1
| ; SO2 2 Na1 1 H1 SOH 1 Na L s
[8]
int
1 s
2 S
1/*KCl2
2 SOH 1 H 1 Cl L | ; SOH1 2 2 Cl ,
[9]
so that
and B 5 10 6F/A, where A is the surface area of oxide available in solution and F is the Faraday constant (13). The electroneutrality condition demands that
s 0 1 s b 1 s d 5 0,
5K zK
int Na1
[10]
int int int *KCl 2 5 Ka1 /KCl2 .
[11]
*K
int a2
[14]
where sd is the diffuse layer charge. The surface species are distributed among the total number of sites available (7, 13, 16): 1 2 N s 5 B~@SOH1 2 # 1 @SOH2 2 Cl # 1 @SOH#
1 @SO2# 1 @SO2 2 Na1#!.
The formation of surface complexes readjust the acid– base equilibrium and affects the surface charge. The surface charge determined from the proton balance, s0, represents the net number of protons released as consumed by all surface reactions and not just the formation of the ionized surface species, [SO2] and [SOH1 2 ]. Increases in electrolyte concentration cause additional adding of counterions until equilibrium is reestablished, subject to the effects of the electrical field. In this manner surface complexation provides a mechanism for the development of the surface charge, in addition to the role of protons and hydroxyl ions. With the surface species defined, it is now possible to write equations for s0, surface charge, and
[15]
The surface site density (N s) can be determined experimentally by various methods; the method given by Hohl and Stumm was used in this study (10). Taking into account that the perlite surface has no zero point of charge, s0 may be considered as the result of charged sites: 2s 0 5 B~@SO2# 1 @SO2 2 Na1#!.
int Na1
[13]
[7]
According to Davis et al. these reactions may be written as surface complex ionization reactions (13), 1 S
2 s b 5 B~@SO2 2 Na1# 2 @SOH1 2 2 Cl #!,
[12]
[16]
At low electrolyte concentrations, i.e., when specific ion binding contributes a smaller amount of charge to the total balance, we may approximate the surface charge as being due to simple acid dissociation (reactions [3]). Then [SO2] may be approximated by s0 from the surface proton balance, s0/B: 2s 0 5 B~@SO2#!.
[17]
For the negative surface, the total fractional ionization may be written (13, 14):
a2 5 2
s0 Ns
[18]
ˇ AN ALKAN AND DOG
94
an essential condition for the appropriate use of the graphical extrapolation method. Conversely, neglecting [SOH1 2 ] and 2 2 2 [SOH1 2 NO3 ] relative to [SO ] and [SO K] for s0 , 0 implies that the obtained value of DpK i is large when the graphical method is used, even if the actual value of the DpK i is small. In general, the double-extrapolation technique (14, 20) is the best available graphical determination method to obtain the adjustable constants, if only titration curves are available. In addition, for surface which contain only negative or positive surface groups, such as for instance the polystyrene latex investigated by James et al. (14) and also the case in our study, the double-extrapolation method seems well suited. For those reasons mentioned above, the double-extrapolation technique proposed by James et al. (14) have been used in order to calculate the constants in this study. The acidity quotient pQ int a 2 can be determined from experimental titration data points at different ionic strengths and plotted as a function of a 2 1 (C salt) 1/ 2 as seen in Figs. 3 and 4. For each ionic strength, curves were extrapolated through
FIG. 3. The variation of the surface ionization activity quotient as a function of surface charge and concentration of supporting electrolyte (KNO3). The condition that s 0 5 (C) 1/ 2 gives pK int a 2 ; (a) UP; (b) EP.
and pK aint2 5 pH 2 log 5 pQ a2 1
a2 eC 0 1 1 2 a 2 2.3 kT
eC 0 . 2.3 kT
[19]
Since the term pQ int a 2 5 pH 2 log( a 2 /1 2 a 2 ) is a function of both surface charge and electrolyte concentration, James, Davis, and Leckie introduced a new kind of double extrapolation plot to obtain pK int a 2 (14). In order to determine the constants from titration data, another method, developed by Westall et al., is an optimalization of the constants using a numerical procedure (17, 18). Koopal et al. (19) stated that for 1 2 s0 , 0 the densities [SOH1 2 ] and [SOH2 NO3 ] can be safely neglected only if DpK i is sufficiently large. Therefore, this is
FIG. 4. The variation of the surface ionization activity quotient as a function of surface charge and concentration of supporting electrolyte (NaCl). The condition that s 0 5 (C) 1/ 2 gives pK int a 2 ; (a) UP; (b) EP.
95
SURFACE TITRATIONS OF PERLITE SUSPENSIONS
TABLE 3 Intrinsic Equilibrium Constants for Perlite Samples and Other Oxides Sample
Electrolyte
pK int a2
p*K Kint1
g-Al2O3 ZrO2 Fe3O4 Magnetite Fe2O3 z H2O SnO2 TiO2 U-kaolin C18-kaolin EP EP UP UP
NaCl KNO3 KNO3 KNO3 NaCl KNO3 NaCl KNO3 KNO3 KNO3 NaCl KNO3 NaCl
8.6 9.0 9.00 10.7 7.9 8.9 6 0.2 4.1 6.7 2.6 2.7 2.5 3.0
7.6 7.2 7.07
int 1 p*K Na
References
8.6 6 0.1
21 22 22 23 24 25 16 26 26 This paper This paper This paper This paper
9.0 6.3 5.8 0 2.9 2.4 2.4 2.3 2.4
Again pQ a2 5 pH 2 log( a 2 /1 2 a 2 ) is a function of log[Na1] and s0. So the intrinsic ionization constant for the surface complex formation of charged sites may be written approximately as
S
int p*K Na 1 5 pH 2 log
D
a2 e~C 0 2 C b! 1 log[Na1] 1 . 1 2 a2 2.3 kT [23]
The results obtained are plotted in this form pH 2 log( a 2 /1 2 a 2 ) as a function of a2 2 log[Na1] in Figs. 5 and 6. The double extrapolation technique, i.e. at constant concentration, extrapolation to a 5 0 and then extrapolation to log C 5 0, int 1 values given in Table 3. If yielded an estimation of the p*K Na the electrolyte used in KNO3, all Na1s must be replaced by K 1 in the equations above. The concentration of ionized surface
the points to the condition that a2 5 0. The extrapolation points themselves are extrapolated to zero electrolyte concentration, i.e., a2 5 0, C 5 0. The obtained pK int a 2 values were given in Table 3. The formation of surface complexes is the principal mechanism by which protons are released (or consumed) by many oxide surfaces in aqueous electrolyte solutions. As can be seen from Table 3, K int a 2 values higher than those given in the literature for different oxides were obtained int in this study. For oxides in general, as pK int a 1 and pK a 2 increase, 1 the acidity of both surface species S 2 OH2 and S 2 OH decreases. In this sense pK int a values are considered as a measure of surface acidity. This means that since the pK int a 2 values of perlite samples are less than those of other oxides given in Table 3, the surface hydroxyl groups on perlite are more acidic than the others. When considered, together with the fact that the surface is negatively charged through all ranges of the pH values studied (i.e., pH 3–10), this result shows that the reaction [3] is in favour of the right-hand side at low electrolyte concentrations. It has been proposed that clay minerals have two types of acidic surface sites: a strongly acidic site, SOH, and a weaker acidic site TOH. This model assumes that the clay surfaces do not develop a positive charge, nor a point of zero charge, in the pH range of 3.5 to 11 considered (20). This correlates with the data in this study. So the ionization reactions of surface sites can be represented as Al(OH)(OH) º Al(OH)O2 1 H1 SiOH º SiO2 1 H1.
[20] [21]
At the higher electrolyte concentrations surface complexes dominate the contribution to surface charge formation, i.e. 2s 0 5 a 2N s 5 B~@SO2# 1 @SO2 2 Na1#! 5 B([SO2 2 Na1]).
[22]
FIG. 5. The variation of the surface ionization activity quotient as a function of surface charge and the logarithm of the electrolyte concentration (KNO3). The condition that s0 5 log C 5 0 gives the surface complex acidity constant p*K int K ; (a) UP; (b) EP.
ˇ AN ALKAN AND DOG
96
@SO2 2 Na1# 5
@SOH#@Na1# int C C x*K Na 1exp@~e 0 2 e b !/kT#. @H1# [25]
int int 1 and *K 1 values obtained can be attributed to The high *K Na K the strength of the electrostatic attraction forces between the highly negatively charged oxide surface and positively charged ions (that is, Na1 or K1).
REFERENCES
FIG. 6. The variation of the surface ionization activity quotient as a function of surface charge and the logarithm of the electrolyte concentration (NaCl). The condition that s0 5 log C 5 0 gives the surface complex acidity int constant p*K Na : (a) UP; (b) EP.
sites, SO2 and SOH1 2 , is generally small in comparison to the surface complexes. This is partly a result of the perturbation of surface equilibria by the electrostatic field. As the surface charge and potential become more negative, the release for a proton by the simple dissociative reaction is opposed by a decreasing electrostatic term, exp(e c 0 /kT), i.e., @SO2# 5
@SOH# int K exp~eC 0/kT!. @H1# a2
[24]
However, the formation of surface complexes, e.g., SO2 2 Na1, is not retarded in an equivalent manner because of the attractive electrostatic term for the counterion, exp(eCb/kT), i.e.,
1. Harben, P. W., and Bates, R. L., “Industrial Minerals Geology and World Deposits,” p. 184. Metal Bulletin Inc., London, 1990. 2. Anonymous, Market Anal. 11, May (1969). 3. Uluatam, S. S., J. AWWA 70, June (1991). 4. Chesterman, C. W., “Industrial Minerals and Rocks,” 4th ed., p. 927. AIME, New York, (1975). 5. Huang, C. P., in “Adsorption of Inorganics at Solid-Liquid Interfaces” (M. A. Anderson and A. J. Rubin, Eds.), p. 181. Ann Arbor, MI, 1981. 6. Schwarz, J. A., Driscoll, C. T., and Bhanot, A. K., J. Colloid Interface Sci. 97(1), 55 (1984). ¨ ., J. Colloid Interface Sci. 192, 114 7. Dogˇan, M., Alkan, M., and C ¸ akir, U (1997). 8. Dogˇan, M., M.Sc. thesis, University of Balikesir Faculty of Science and Literature, Balikesir, Turkey, 1997. [Turkish] 9. Karakas¸, R., Ph.D. thesis University of Ege Faculty of Science and Literature, Izmir, Turkey, 1996. 10. Hohl, H., and Stumm, W., J. Colloid Interface Sci. 55(2), 281 (1976). 11. Ahmed, M. S., J. Phys. Chem. 73, 3546 (1969). 12. Hunter, R. J., and Alexander, A. E., J. Colloid Sci. 18, 820 (1963). 13. Davis, J. A., James, R. O., and Leckie, J. O., J. Colloid Interface Sci. 63(3), 480 (1978). 14. James, R. O., Davis, J. A., and Leckie, J. O., J. Colloid Interface Sci. 65(2), 331 (1978). 15. Yates, D. E., Levine, S., and Healy, T. W., J. Chem. Soc. Faraday Trans. 70, 1807 (1974). 16. Sprycha, R., J. Colloid Interface Sci. 102(1), 173 (1984). 17. Westall, J., and Hohl, H., Adv. Colloid Interface Sci. 12, 265 (1980). 18. Westall, J., and Morel, F., “FITEQL: A General Algorithm for the Interpretation of Experimental Data,” Technical Note No. 19, Ralph M. Parsons Laboratory, MIT, Cambridge, MA, 1977. 19. Koopal, L. K., Riemsdijk, W. H. V., and Roffey, M. G., J. Colloid Interface Sci. 118, 117 (1987). 20. James, R. O., and Parks, G. A., Characterization of aqueous colloids by their electrical double-layer and intrinsic surface chemical properties, in “Surface and Colloid Science” (E. Matijevic, Ed.), Vol. 12, p. 119. Plenum, New York, 1982. 21. Sprycha, R., J. Colloid Interface Sci. 127(1), 12 (1989). 22. Regazzoni, A. E., Blesa, M. A., and Marato, A. J. G., J. Colloid Interface Sci. 91(2), 569 (1983). 23. Blesa, M. A., Figholia, N. M., Marato, A. J. G., and Regazzoni, A. E., J. Colloid Interface Sci. 101(2), 410 (1984). 24. Davis, J. A., and Leckie, J. O., J. Colloid Interface Sci. 74(1), 35 (1980). 25. Houchin, M. R., and Warren, L. J., J. Colloid Interface Sci. 100(1), 278 (1984). 26. Braggs, B., Fornasiero, D., Ralston, J., and Smait, S. R., Clays Clay Minerals 42(2), 133 (1994).