- Email: [email protected]
Effects of Pb2+-doping on alkoxide-derived silica sols and gels
}OURNA
Journal of Non-Crystalline Solids 139 (1992) 165-171 North-Holland
L OF
NON-CRYSTALLINESOLIDS
Effects of pb2+-doping on alkoxide-derived silica sols and gels Amitava Patra and Dibyendu Ganguli Central Glass and Ceramic Research Institute, Calcutta 700032, India Received 14 February 1991 Revised manuscript received 13 June 1991
Five sols prepared from tetraethyl orthosilicate (TEOS) with 0 to 2 equivalent tool% PbO displayed sharp and progressive changes in the viscosity of the sols, and an increase in viscosity with time and gelling time. Activation energies of gelation (calculated from gelling times at 10, 24, 30, 40 and 50 ° C) showed similar sharp changes. The acceleration of gelling time with Pb 2+ was found to be much more pronounced than with divalent cations Mg 2+, Ca 2+ and Zn 2+. UV absorption edges of the gels showed significant changes with increase in dopant content, as did the corresponding optical gaps. The acceleration of gelation was considered to be caused by the linkage of Pb 2+ with small polymeric clusters via -=SiOterminal groups, and the consequent aggregation of the clusters.
1. Introduction Preparation of pure silica gel and glass monoliths from alkoxide sols and their characterization accounts for the majority of publications in the modern sol-gel literature. Because of preparative advantages, binary silicate gels and glasses with high percentages of the second oxide have also been prepared in different systems. There is, on the other hand, relatively limited awareness that low level doping (2 mol% or less) of a second cation in an alkoxide-derived silica sol system can in certain cases cause remarkable shifts in the sol-gel transition, as well as in the properties of the corresponding gels and glasses. In a recent work, Ganguli [1] showed that the presence of 0.5-2.0 wt% of oxides of Al3+, Cr3+, Fe3+, Mg2+, Ca 2+ and Sr 2+ in alkoxide-derived silica gels caused striking changes in the surface area and specific pore volume, and in certain cases caused important changes in pore size distribution, e.g., monomodal to bi-modal. Bansal [2] has recently shown shifts in gelling time with the incorporation of a second cation (2-10 mol%) in a tetraethyl orthosilicate (TEOS) sol. In the present work, silica gels containing minor quantities of Pb 2+ (0.2-2.0 mol% PbO) have
been prepared from alkoxide sols. This paper describes the very significant changes that can be caused in (i) the viscosity of the sol, (ii) the sol-gel transition and (iii) the optical character of the gels by such low level doping.
2. Experimental All the gels (0-2 equivalent mol% PbO) were prepared from alcohol-free (initially) alkoxidewater systems [3]. For Pb-doping, a solution of Pb(NO3) 2 (Sarabhai M., GR grade) and HNO 3 (Merck, GR grade) in water was poured in accurately weighed tetraethyl orthosilicate (TEOS; Fluka, Purum grade) under vigorous stirring with the formation of a two-phase solution. The H z O / T E O S / H N O 3 molar ratio was 14:1:0.01. The liquid thus obtained was stirred for a few minutes, during which it changed into a clear, monophase solution. Stirring was continued for 0.5 h. In each case, the pH was adjusted to 3.50 by addition of NH4OH solution, and the sol cast in glass containers. Gelation was studied at 10, 24, 30, 40 and 50 ° C in a constant temperature bath. The gels prepared at 24°C were dried for 4-5
0022-3093/92/$05.00 © 1992 - Elsevier Science Publishers B.V. All rights reserved
A. Patra, D. Ganguli /Alkoxide-deriued silica sols and gels
166
7.0
days at 60 ° C and heated very slowly to 200 and 400 ° C under oxygen with 0.5 h soak. Crack-free samples were obtained in this way. These samples were ground on both sides by alundum powder (600 and 800 mesh) for 20 min, followed by polishing with diamond paste (grade ¼-OS-475) for 25 min to a final thickness of about 5.6 ram. The viscosities of the sols were measured as a function of time with a Brookfield digital viscometer (LVTDV-II). A Hitachi 3210 U V - V I S spectrophotometer was used for obtaining the transmission spectra (190-900 nm). Infrared transmission spectra were recorded (400-4000 cm -1) with a F T I R spectrometer (Nicolet 5 PC).
o
F---
3.5
B
1.4
I
I
I
I
0.2
0.5
1.0
2.0
MOL% PbO
3. Results
Fig. 2. Gelling time vs. equivalent P b O content in sols (gelled at 24 ° C).
3.1. Viscosity of sols and sol-gel transition The sharp changes in viscosity versus time for a set of sols at 2 4 ° C with the addition of very small amounts of Pb 2+ are shown in the viscosity-time curves of fig. 1. With 2 mol% addition,
5
the increase in viscosity was very fast and reminiscent of viscosity-time relationships in sols derived from highly hydrolyzable alkoxides [4], and
3
2
2.7
2.3
1.9
~,
1.5
0.7
i
I
[
1
I
LO
2.0
3.0
4.0
5.0
I 6.0
TiME(h)
Fig. 1. V i s c o s i t y - t i m e curves for p u r e and d o p e d gels. 1, P u r e silica; 2, 0.2PbO; 3, 0.5PbO; 4, 1.0PbO; 5, 2.0PbO.
A. Patra, D. Ganguli / Alkoxide-derived silica sols and gels
167
Table 1 Variations in gelling time as a function of temperature Equivalent mol% PbO (added in sol)
Equivalent mol% PbO (analyzed in gel)
Gelling time (h) 10 o C 24°C
30°C
40°C
50°C
0 0.2 0.5 1.0 2.0
0.192 0.482 0.961 1.931
30 18 7.94 5.0 2.5
4 2.33 1.67 1.16 0.50
1.67 1.13 0.8 0.42 0.3
0.67 0.52 0.4 0.3 0.13
6.58 4.42 2.65 1.66 0.75
also certain special situations during sol-gel transitions in TEOS-derived systems [5]. Figure 2 indicates that, in conformity with the sharp increase in viscosity with doping of lead, the gelling times at 24°C also showed a very sharp decrease. With 2 mol% of the dopant, gelation took place in less than 1 h (a greater than 88% decrease from the gelling time of the undoped sol). This made it clear that, for the Pb-doped sols, gel formation could be expected to take place almost instantaneously when helped by at least one of the gelation-promoting conditions [6], e.g., increased temperature. To examine
the effect of temperature in conjunction with that of doping, gelling times were recorded for a series of pure and doped sols maintained at 10, 30, 40 and 50 ° C (in addition to those kept at 24 ° C). Table 1 shows the variations in gelling time with the addition of dopants at the above temperatures (including the data for 24 ° C, as in fig. 2). The very rapid gelation caused by doping, especially at temperatures of 30-50 o C, is probably unique in the sense that such quick polymerization and aggregation of clusters to form the final spanning cluster is very rare (see ref. [5] for one example).
L50
1
2 1.00
-
3 4
5
0,00
- 1.00
3.00
I
I
3.20
3.40
3.60
I0~/T, I( I
Fig. 3. Log(tgel) vs.
I/T plots for pure and doped sols at different temperatures. 1, Pure silica; 2, 0.2PbO; 3, 0.5PbO; 4, 1.0PbO; 5, 2.0PbO.
168
A. Patra, D. Ganguli / Alkoxide:deriued silica sols and gels
Table 2 Apparent activation energies for the process of gelation
Table 3 Optical absorption edges of gels heated at two temperatures
Equivalent mol% PbO (added in sol)
E* (kJ/mol)
Equivalent tool% PbO (added in sol)
0 0.2 0.5 1.0 2.0
71.25 67.09 56.03 55.40 54.15
0 0.2 0.5 1.0 2.0
Following Bansal [2], activation energies of gelation (which is a combined effect of the rate of hydrolysis and condensation, diffusion of polymeric clusters and' the amount of the dopant, as in the present case) were calculated with the help of the equation log(tgel ) = - l o g A ' + E*/(Z.3026RT),
(1)
where tgel is gelling time, A' is the usual Arrhenius constant; E * ~S an apparent activation energy for the overall gelation process, R is the gas constant and T is the reaction temperature in Kelvin. Log(tgel) was plotted against l / T , and the slope was obtained from a regression analysis. Figure 3 shows the nature of the plots. The values of E * obtained are given in table 2. Activation energies of gelation of silica sols have been calculated for different conditions [7], and the values varying in the range 10-20 K c a l / m o l are not very different from those obtained in this work. However, the noteworthy fact is the sharp drop in the value of E * around 23% on addition of as small as 1-2 tool% of PbO. Such a sharp change has not been observed when 2-10 mol% of MgO and CaO were added in a silica sol composition [2].
3.2. Optical properties of the gels
Optical absorption (0.5% T) edge (nm) 200 ° C
400 o C
273.4 286.0 300.8 314.8 324.8
281.2 288.1 293.6 298.4 327.4
length at which transmission [8] reached 0.5%. The values are presented in table 3. The absorption coefficient in a material is known to be related to the optical gap by the following equation
a(w) = B(hw - Eopt)n/hw,
(2)
where a(w) is the absorption coefficient, Eopt is the optical gap, B is a constant and n = 2 for amorphous materials [9]. Optical gaps for the present samples were estimated from plots of [a(w)(hw)] ½vs. (hw). The data for gels heated at 7.0
5
6.0
>~ ~J
I II
I
4.0
,Y 2.0
--
/
If/
i
UV-visible spectra were obtained of the pure and doped gel samples heated at 200 and 400 ° C under oxygen and ground and polished to an approximate standard thickness of 5.6 mm with a view to examining how the absorption edge of a silica gel changed with low dopings of Pb 2+. The UV absorption edges were defined as the wave-
4321
t 2.0
2.5
/
I
I I !
I
I
3.0
3.5
11 I~ ,/
II
I
4.0
4.5
PHOTONENERGY(eV) ]
Fig. 4. (ah(o)~ vs. (h~o) plots for pure and doped gels heated at 400 ° C. i, Pure silica; 2, 0.2PbO; 3, O.5PhO; 4, l.OPbO; 5, 2.0PbO.
A. Patra, D. Ganguli / Alkoxide-derived silica sols and gels
400 ° C are presented in fig. 4. As is observed in comparable oxide glass systems including P b O SiO 2 glasses [10], the optical gap in the doped gels decreased with increasing dopant content. It is, however, noteworthy that the change in Eopt in the present gels with change in dopant content was found to be much steeper than that reported for P b O - S i O 2 glasses [10]. Unlike UV-visible spectra, the F T I R spectra of the doped gels did not show any important change with dopant content. One interesting change, however, was noticed in case of the SiO H terminal band at 958.7 cm-x for pure silica gel (heated at 400 o C). With increasing dopant content, this band shifted towards higher frequencies and, with 2 mol% PbO, was situated at 969.4 c m - 1.
4. Discussion
4.1. Sol-gel transition Acceleration of gelation of a silicic acid sol (or flocculation of a colloidal silica sol) in the presence of ionic impurities is quite an old observation, dating from the early part of this century [11,12]. The role of alkali and alkaline earth ions in decreasing gelling time was identified in these early publications and later confirmed [13-15]. Further observations were made recently by Bansal [2] who showed that the gelling time of TEOS-derived sols decreased with addition of dopants (2-10 tool%, which was much higher than the range examined for Pb 2+ in this work) such as alkali and alkaline earth ions, acting as network modifiers, but increased with cations expected to act as network formers. However, the dependence of gelling time on dopant concentration reported by Bansal was much less strong than observed here. For systems where there was a decrease in gelling time, the overall activation energy did not show appreciable dependence on the dopant concentration. This leads us to the inference that the effect of Pb 2+ on the gelling time of TEOS-derived sols was specially prominent. Additional experiments were performed to confirm this point so far as gelling time was
169
Table 4 Gelling times of different doped sols at 50 o C Dopant (2.0 mol% equivalent added in sol)
Gelling time (h)
PbO CaO ZnO MgO Undoped
0.13 0.52 0.58 0.50 0.67
T h e sols were prepared and stirred for 0.5 h at 20 o C, after which they were transferred to a constant temperature bath equilibrated at 50 ° C.
concerned, since the preparative method of the present sols was different from that of Bansal. Table 4 presents a comparative picture of the gelation behaviour of silica sols with 2 mol% oxide dopants of various cations. Allen and Matijevid, in a series of papers [16], concluded that unhydrolyzed cations undergo ion exchange with H ÷ of silicic acid monomers in the following way: - S i - O H + M Z + ~ - S i - O M (z-
1)++ H +.
(3)
Since the pH of the sols in this work was adjusted in each case to 3.5, i.e., higher than the isoelectric point of silica, another reaction sequence [17] should also be considered, which is probably more important for the present case (i.e., polymerization pH in the range 2-7): -Si-OH + OH-~
-SiO- + H20 ,
- S i O - + H O - S i --* - S i - O - S i - + O H -
(4) (5)
The former is a fast step, while the latter is slow. Equation (4) should be especially fast in the present case because of the expected completion of the hydrolysis reaction at an early stage [18] and the c o n s e q u e n t generation of silicic acid monomers. Because of the slowness of the reaction shown in eq. (5), only small-sized polymeric clusters can be expected to form at the early stages of aging of the sol. Such small clusters are known to have only small interaction among themselves [19]. If the - S i - O - species in eqs. (4) and (5) can be taken to represent the product of a terminal S i - O H of a small polymeric cluster, one can probably expect such terminal members
A. Patra, D. Ganguli / Alkoxide-derived silica sols and gels
170
of two clusters to react via the dopant cation in the following way to form bridges: - S i O - + MZ++ - O S i -SiO a- ... M(Z-2a)... Oa-Si=(transition state) --+ - S i - O . . . M . . . O - S i and to form larger clusters. Such a process of aggregation via dopant cations should reduce although not completely eliminate) the necessity of going through the slow step of eq. (5), and therefore, reduce the gelling time, and also the activation energy required for gelation (table 2). In fact, for completely screening its charge and also satisfying its ability of coordination, the dopant cation may become associated with several polymeric clusters having - S i O - groups. It must be noted that when the dopant cation is provided through the salt of a strong acid, its dissociation is complete and in a TEOS solution where the H z O / T E O S mole ratio is as high as 14, such cations must be in a hydrated state. This can be a factor hindering the bridge formation proposed above. The degree of hydration being proportional to the charge density, this hindering factor is less for Pb 2+ than for smaller divalent cations. This explains, at least partially, the differences in gelling time shown in table 4. Considering the high H z O / T E O S mole ratio ( = 14) in the doped sols of this work, and their very short gelling time, we suggest that the present situation compares well with the diffusion-limited fractal growth model of Pope and Mackenzie [20]. Significant changes in surface area and porosity are expected in the present gels with change in dopant content, as shown in other systems by Ganguli [1]. This will be examined in a separate study.
[22]. However, the presence of dopant cations and increase in their content in silica gels increases the surface area of the gel [1], which in turn means increase in O H content [23]. Thus the total O H content should be dependent on the dopant content. Therefore, in this work, the optical properties have been described as a function of the independent variable, i.e., the Pb 2+ content. For a given temperature of heat treatment, the optical absorption edge shifted progressively to longer wavelengths with increase in Pb 2+ content (table 3), which was evidence of the direct relationship between the two.
5. Conclusions
(i) Addition of 0.2-2.0 equivalent mol% PbO via Pb(NO3) 2 in a tetraethyl orthosilicate sol showed remarkable changes in viscosity, increase in viscosity with time and gelling times as compared with a dopant-free sol. Activation energies of gelation also showed a sharp drop with addition of as little as 1 equivalent mol% of the dopant. (ii) Quick gelation of silica sols with the assistance of several divalent cations has been recorded, although the effect was the strongest for Pb 2+. (iii) The gels formed by the progressively increasing additions of Pb 2+ up to only 2 mol% showed significant shifts of the UV absorption edge towards higher wavelengths. The corresponding optical gaps showed changes with change of Pb 2--- content sharper than in P b O SiO 2 glasses. (iv) Pb 2÷ has been considered to accelerate gelation by connecting several small polymeric clusters via terminal - S i O - groups.
4.2. Optical properties The presence of Pb 2+ ions in silicate and other oxide glasses is known to influence the position of the optical absorption edge significantly, causing a progressive shift towards longer wavelengths with increasing Pb 2+ content [21]. For gel samples, the effect of the surface hydroxyl on the absorption edge should also be considered
The authors are thankful to Dr B.K. Sarkar, Director of the Institute, for his kind permission to publish this paper. They are also thankful to the Chemical Laboratory of this Institute for their help in the analysis of lead and Mr Samiran Bhattacharjee, Jadavpur University, for his help in obtaining some of the spectra. Thanks are also due to the Department of Electronics, Govern-
A. Patra, D. Gangufi /Alkoxide-derived silica sols and gels ment of India for providing funds under Sanction no. MATE/I/5(18)/89.
References [1] D. Ganguli, J. Non-Cryst. Solids 101 (1988) 117. [2] N.P. Bansal, J. Am. Ceram. Soc. 73 (1990) 2647. [3] D. Avnir and V.R. Kaufman, J. Non-Cryst. Solids 92 (1987) 180. [4] B.E. Yoldas, Appl. Opt. 21 (1982) 2960. [5] E.J.A. Pope and J.D. Mackenzie, J. Non-Cryst. Solids 87 (1986) 185. [6] M. Yamane, in: Sol-Gel Technology for Thin Films, Fibers, Preforms, Electronics and Specialty Shapes, ed. L.C. Klein (Noyes, New Jersey, 1988) p. 200. [7] M.W. Colby, A. Osaka and J.D. Mackenzie, J. Non-Cryst. Solids 99 (1988) 129. [8] J. Wong and C.A. Angell, Glass Structure by Spectroscopy (Dekker, New York, 1976) p. 152. [9] N.F. Mott and E.A. Davis, Electronic Processes in NonCrystalline Materials (Clarendon, Oxford, 1971) p. 249. [10] A. B~rbulescu and L. Sincan, Phys. Status Solidi (a)85 (1984) K129. [11] F. Dienert and F. Wandenbulcke, C. Rendus 178 (1924) 564.
171
[12] E. Laskin, Kolloid Z. 45 (1928) 129. [13] C.B. Hurd, J.W. Rhoades, W.G. Gormley and A.C. Santora, J. Phys. Chem. 62 (1958) 882. [14] J. Depasse and A. Watillon, J. Colloid Interf. Sci. 33 (1970) 430. [15] S.K. Mookerjee and S.K. Niyogi, Cent. Glass Ceram. Inst. Bull. 22 (1975) 1. [16] L.H. Allen and E. Matijevic, J. Colloid Interf. Sci. 31 (1969) 287; 33 (1970) 420; 35 (1971) 66. [17] C.J. Brinker and G.W. Scherer, Sol-Gel Science (Academic Press, San Diego, CA, 1990) p. 103. [18] D.L. Wood and E.M. Rabinovich, J. Non-Cryst. Solids 107 (1989) 199. [19] S. Sakka, in: Sol-Gel Science and Technology, eds. M.A. Aegerter, M. Jafelicci Jr., D.F. Souza and E.D. Zanotto (World Scientific, Singapore, 1989) p. 76. [20] E.J.A. Pope and J.D. Mackenzie, J. Non-Cryst. Solids 101 (1988) 198. • [21] B.M. Cohen, D.R. Uhlmann and R.R. Shaw, J. Non-Cryst. Solids 12 (1973) 177. [22] L.L. Hench, S.H. Wang and J.L. Nogu~s, SPIE, Multifunctional Materials 878 (1988) 76. [23] C.J. Brinker and G.W. Scherer, Sol-Gel Science (Academic Press, San Diego, CA, 1990) p. 624.
Journal of Non-Crystalline Solids 139 (1992) 165-171 North-Holland
L OF
NON-CRYSTALLINESOLIDS
Effects of pb2+-doping on alkoxide-derived silica sols and gels Amitava Patra and Dibyendu Ganguli Central Glass and Ceramic Research Institute, Calcutta 700032, India Received 14 February 1991 Revised manuscript received 13 June 1991
Five sols prepared from tetraethyl orthosilicate (TEOS) with 0 to 2 equivalent tool% PbO displayed sharp and progressive changes in the viscosity of the sols, and an increase in viscosity with time and gelling time. Activation energies of gelation (calculated from gelling times at 10, 24, 30, 40 and 50 ° C) showed similar sharp changes. The acceleration of gelling time with Pb 2+ was found to be much more pronounced than with divalent cations Mg 2+, Ca 2+ and Zn 2+. UV absorption edges of the gels showed significant changes with increase in dopant content, as did the corresponding optical gaps. The acceleration of gelation was considered to be caused by the linkage of Pb 2+ with small polymeric clusters via -=SiOterminal groups, and the consequent aggregation of the clusters.
1. Introduction Preparation of pure silica gel and glass monoliths from alkoxide sols and their characterization accounts for the majority of publications in the modern sol-gel literature. Because of preparative advantages, binary silicate gels and glasses with high percentages of the second oxide have also been prepared in different systems. There is, on the other hand, relatively limited awareness that low level doping (2 mol% or less) of a second cation in an alkoxide-derived silica sol system can in certain cases cause remarkable shifts in the sol-gel transition, as well as in the properties of the corresponding gels and glasses. In a recent work, Ganguli [1] showed that the presence of 0.5-2.0 wt% of oxides of Al3+, Cr3+, Fe3+, Mg2+, Ca 2+ and Sr 2+ in alkoxide-derived silica gels caused striking changes in the surface area and specific pore volume, and in certain cases caused important changes in pore size distribution, e.g., monomodal to bi-modal. Bansal [2] has recently shown shifts in gelling time with the incorporation of a second cation (2-10 mol%) in a tetraethyl orthosilicate (TEOS) sol. In the present work, silica gels containing minor quantities of Pb 2+ (0.2-2.0 mol% PbO) have
been prepared from alkoxide sols. This paper describes the very significant changes that can be caused in (i) the viscosity of the sol, (ii) the sol-gel transition and (iii) the optical character of the gels by such low level doping.
2. Experimental All the gels (0-2 equivalent mol% PbO) were prepared from alcohol-free (initially) alkoxidewater systems [3]. For Pb-doping, a solution of Pb(NO3) 2 (Sarabhai M., GR grade) and HNO 3 (Merck, GR grade) in water was poured in accurately weighed tetraethyl orthosilicate (TEOS; Fluka, Purum grade) under vigorous stirring with the formation of a two-phase solution. The H z O / T E O S / H N O 3 molar ratio was 14:1:0.01. The liquid thus obtained was stirred for a few minutes, during which it changed into a clear, monophase solution. Stirring was continued for 0.5 h. In each case, the pH was adjusted to 3.50 by addition of NH4OH solution, and the sol cast in glass containers. Gelation was studied at 10, 24, 30, 40 and 50 ° C in a constant temperature bath. The gels prepared at 24°C were dried for 4-5
0022-3093/92/$05.00 © 1992 - Elsevier Science Publishers B.V. All rights reserved
A. Patra, D. Ganguli /Alkoxide-deriued silica sols and gels
166
7.0
days at 60 ° C and heated very slowly to 200 and 400 ° C under oxygen with 0.5 h soak. Crack-free samples were obtained in this way. These samples were ground on both sides by alundum powder (600 and 800 mesh) for 20 min, followed by polishing with diamond paste (grade ¼-OS-475) for 25 min to a final thickness of about 5.6 ram. The viscosities of the sols were measured as a function of time with a Brookfield digital viscometer (LVTDV-II). A Hitachi 3210 U V - V I S spectrophotometer was used for obtaining the transmission spectra (190-900 nm). Infrared transmission spectra were recorded (400-4000 cm -1) with a F T I R spectrometer (Nicolet 5 PC).
o
F---
3.5
B
1.4
I
I
I
I
0.2
0.5
1.0
2.0
MOL% PbO
3. Results
Fig. 2. Gelling time vs. equivalent P b O content in sols (gelled at 24 ° C).
3.1. Viscosity of sols and sol-gel transition The sharp changes in viscosity versus time for a set of sols at 2 4 ° C with the addition of very small amounts of Pb 2+ are shown in the viscosity-time curves of fig. 1. With 2 mol% addition,
5
the increase in viscosity was very fast and reminiscent of viscosity-time relationships in sols derived from highly hydrolyzable alkoxides [4], and
3
2
2.7
2.3
1.9
~,
1.5
0.7
i
I
[
1
I
LO
2.0
3.0
4.0
5.0
I 6.0
TiME(h)
Fig. 1. V i s c o s i t y - t i m e curves for p u r e and d o p e d gels. 1, P u r e silica; 2, 0.2PbO; 3, 0.5PbO; 4, 1.0PbO; 5, 2.0PbO.
A. Patra, D. Ganguli / Alkoxide-derived silica sols and gels
167
Table 1 Variations in gelling time as a function of temperature Equivalent mol% PbO (added in sol)
Equivalent mol% PbO (analyzed in gel)
Gelling time (h) 10 o C 24°C
30°C
40°C
50°C
0 0.2 0.5 1.0 2.0
0.192 0.482 0.961 1.931
30 18 7.94 5.0 2.5
4 2.33 1.67 1.16 0.50
1.67 1.13 0.8 0.42 0.3
0.67 0.52 0.4 0.3 0.13
6.58 4.42 2.65 1.66 0.75
also certain special situations during sol-gel transitions in TEOS-derived systems [5]. Figure 2 indicates that, in conformity with the sharp increase in viscosity with doping of lead, the gelling times at 24°C also showed a very sharp decrease. With 2 mol% of the dopant, gelation took place in less than 1 h (a greater than 88% decrease from the gelling time of the undoped sol). This made it clear that, for the Pb-doped sols, gel formation could be expected to take place almost instantaneously when helped by at least one of the gelation-promoting conditions [6], e.g., increased temperature. To examine
the effect of temperature in conjunction with that of doping, gelling times were recorded for a series of pure and doped sols maintained at 10, 30, 40 and 50 ° C (in addition to those kept at 24 ° C). Table 1 shows the variations in gelling time with the addition of dopants at the above temperatures (including the data for 24 ° C, as in fig. 2). The very rapid gelation caused by doping, especially at temperatures of 30-50 o C, is probably unique in the sense that such quick polymerization and aggregation of clusters to form the final spanning cluster is very rare (see ref. [5] for one example).
L50
1
2 1.00
-
3 4
5
0,00
- 1.00
3.00
I
I
3.20
3.40
3.60
I0~/T, I( I
Fig. 3. Log(tgel) vs.
I/T plots for pure and doped sols at different temperatures. 1, Pure silica; 2, 0.2PbO; 3, 0.5PbO; 4, 1.0PbO; 5, 2.0PbO.
168
A. Patra, D. Ganguli / Alkoxide:deriued silica sols and gels
Table 2 Apparent activation energies for the process of gelation
Table 3 Optical absorption edges of gels heated at two temperatures
Equivalent mol% PbO (added in sol)
E* (kJ/mol)
Equivalent tool% PbO (added in sol)
0 0.2 0.5 1.0 2.0
71.25 67.09 56.03 55.40 54.15
0 0.2 0.5 1.0 2.0
Following Bansal [2], activation energies of gelation (which is a combined effect of the rate of hydrolysis and condensation, diffusion of polymeric clusters and' the amount of the dopant, as in the present case) were calculated with the help of the equation log(tgel ) = - l o g A ' + E*/(Z.3026RT),
(1)
where tgel is gelling time, A' is the usual Arrhenius constant; E * ~S an apparent activation energy for the overall gelation process, R is the gas constant and T is the reaction temperature in Kelvin. Log(tgel) was plotted against l / T , and the slope was obtained from a regression analysis. Figure 3 shows the nature of the plots. The values of E * obtained are given in table 2. Activation energies of gelation of silica sols have been calculated for different conditions [7], and the values varying in the range 10-20 K c a l / m o l are not very different from those obtained in this work. However, the noteworthy fact is the sharp drop in the value of E * around 23% on addition of as small as 1-2 tool% of PbO. Such a sharp change has not been observed when 2-10 mol% of MgO and CaO were added in a silica sol composition [2].
3.2. Optical properties of the gels
Optical absorption (0.5% T) edge (nm) 200 ° C
400 o C
273.4 286.0 300.8 314.8 324.8
281.2 288.1 293.6 298.4 327.4
length at which transmission [8] reached 0.5%. The values are presented in table 3. The absorption coefficient in a material is known to be related to the optical gap by the following equation
a(w) = B(hw - Eopt)n/hw,
(2)
where a(w) is the absorption coefficient, Eopt is the optical gap, B is a constant and n = 2 for amorphous materials [9]. Optical gaps for the present samples were estimated from plots of [a(w)(hw)] ½vs. (hw). The data for gels heated at 7.0
5
6.0
>~ ~J
I II
I
4.0
,Y 2.0
--
/
If/
i
UV-visible spectra were obtained of the pure and doped gel samples heated at 200 and 400 ° C under oxygen and ground and polished to an approximate standard thickness of 5.6 mm with a view to examining how the absorption edge of a silica gel changed with low dopings of Pb 2+. The UV absorption edges were defined as the wave-
4321
t 2.0
2.5
/
I
I I !
I
I
3.0
3.5
11 I~ ,/
II
I
4.0
4.5
PHOTONENERGY(eV) ]
Fig. 4. (ah(o)~ vs. (h~o) plots for pure and doped gels heated at 400 ° C. i, Pure silica; 2, 0.2PbO; 3, O.5PhO; 4, l.OPbO; 5, 2.0PbO.
A. Patra, D. Ganguli / Alkoxide-derived silica sols and gels
400 ° C are presented in fig. 4. As is observed in comparable oxide glass systems including P b O SiO 2 glasses [10], the optical gap in the doped gels decreased with increasing dopant content. It is, however, noteworthy that the change in Eopt in the present gels with change in dopant content was found to be much steeper than that reported for P b O - S i O 2 glasses [10]. Unlike UV-visible spectra, the F T I R spectra of the doped gels did not show any important change with dopant content. One interesting change, however, was noticed in case of the SiO H terminal band at 958.7 cm-x for pure silica gel (heated at 400 o C). With increasing dopant content, this band shifted towards higher frequencies and, with 2 mol% PbO, was situated at 969.4 c m - 1.
4. Discussion
4.1. Sol-gel transition Acceleration of gelation of a silicic acid sol (or flocculation of a colloidal silica sol) in the presence of ionic impurities is quite an old observation, dating from the early part of this century [11,12]. The role of alkali and alkaline earth ions in decreasing gelling time was identified in these early publications and later confirmed [13-15]. Further observations were made recently by Bansal [2] who showed that the gelling time of TEOS-derived sols decreased with addition of dopants (2-10 tool%, which was much higher than the range examined for Pb 2+ in this work) such as alkali and alkaline earth ions, acting as network modifiers, but increased with cations expected to act as network formers. However, the dependence of gelling time on dopant concentration reported by Bansal was much less strong than observed here. For systems where there was a decrease in gelling time, the overall activation energy did not show appreciable dependence on the dopant concentration. This leads us to the inference that the effect of Pb 2+ on the gelling time of TEOS-derived sols was specially prominent. Additional experiments were performed to confirm this point so far as gelling time was
169
Table 4 Gelling times of different doped sols at 50 o C Dopant (2.0 mol% equivalent added in sol)
Gelling time (h)
PbO CaO ZnO MgO Undoped
0.13 0.52 0.58 0.50 0.67
T h e sols were prepared and stirred for 0.5 h at 20 o C, after which they were transferred to a constant temperature bath equilibrated at 50 ° C.
concerned, since the preparative method of the present sols was different from that of Bansal. Table 4 presents a comparative picture of the gelation behaviour of silica sols with 2 mol% oxide dopants of various cations. Allen and Matijevid, in a series of papers [16], concluded that unhydrolyzed cations undergo ion exchange with H ÷ of silicic acid monomers in the following way: - S i - O H + M Z + ~ - S i - O M (z-
1)++ H +.
(3)
Since the pH of the sols in this work was adjusted in each case to 3.5, i.e., higher than the isoelectric point of silica, another reaction sequence [17] should also be considered, which is probably more important for the present case (i.e., polymerization pH in the range 2-7): -Si-OH + OH-~
-SiO- + H20 ,
- S i O - + H O - S i --* - S i - O - S i - + O H -
(4) (5)
The former is a fast step, while the latter is slow. Equation (4) should be especially fast in the present case because of the expected completion of the hydrolysis reaction at an early stage [18] and the c o n s e q u e n t generation of silicic acid monomers. Because of the slowness of the reaction shown in eq. (5), only small-sized polymeric clusters can be expected to form at the early stages of aging of the sol. Such small clusters are known to have only small interaction among themselves [19]. If the - S i - O - species in eqs. (4) and (5) can be taken to represent the product of a terminal S i - O H of a small polymeric cluster, one can probably expect such terminal members
A. Patra, D. Ganguli / Alkoxide-derived silica sols and gels
170
of two clusters to react via the dopant cation in the following way to form bridges: - S i O - + MZ++ - O S i -SiO a- ... M(Z-2a)... Oa-Si=(transition state) --+ - S i - O . . . M . . . O - S i and to form larger clusters. Such a process of aggregation via dopant cations should reduce although not completely eliminate) the necessity of going through the slow step of eq. (5), and therefore, reduce the gelling time, and also the activation energy required for gelation (table 2). In fact, for completely screening its charge and also satisfying its ability of coordination, the dopant cation may become associated with several polymeric clusters having - S i O - groups. It must be noted that when the dopant cation is provided through the salt of a strong acid, its dissociation is complete and in a TEOS solution where the H z O / T E O S mole ratio is as high as 14, such cations must be in a hydrated state. This can be a factor hindering the bridge formation proposed above. The degree of hydration being proportional to the charge density, this hindering factor is less for Pb 2+ than for smaller divalent cations. This explains, at least partially, the differences in gelling time shown in table 4. Considering the high H z O / T E O S mole ratio ( = 14) in the doped sols of this work, and their very short gelling time, we suggest that the present situation compares well with the diffusion-limited fractal growth model of Pope and Mackenzie [20]. Significant changes in surface area and porosity are expected in the present gels with change in dopant content, as shown in other systems by Ganguli [1]. This will be examined in a separate study.
[22]. However, the presence of dopant cations and increase in their content in silica gels increases the surface area of the gel [1], which in turn means increase in O H content [23]. Thus the total O H content should be dependent on the dopant content. Therefore, in this work, the optical properties have been described as a function of the independent variable, i.e., the Pb 2+ content. For a given temperature of heat treatment, the optical absorption edge shifted progressively to longer wavelengths with increase in Pb 2+ content (table 3), which was evidence of the direct relationship between the two.
5. Conclusions
(i) Addition of 0.2-2.0 equivalent mol% PbO via Pb(NO3) 2 in a tetraethyl orthosilicate sol showed remarkable changes in viscosity, increase in viscosity with time and gelling times as compared with a dopant-free sol. Activation energies of gelation also showed a sharp drop with addition of as little as 1 equivalent mol% of the dopant. (ii) Quick gelation of silica sols with the assistance of several divalent cations has been recorded, although the effect was the strongest for Pb 2+. (iii) The gels formed by the progressively increasing additions of Pb 2+ up to only 2 mol% showed significant shifts of the UV absorption edge towards higher wavelengths. The corresponding optical gaps showed changes with change of Pb 2--- content sharper than in P b O SiO 2 glasses. (iv) Pb 2÷ has been considered to accelerate gelation by connecting several small polymeric clusters via terminal - S i O - groups.
4.2. Optical properties The presence of Pb 2+ ions in silicate and other oxide glasses is known to influence the position of the optical absorption edge significantly, causing a progressive shift towards longer wavelengths with increasing Pb 2+ content [21]. For gel samples, the effect of the surface hydroxyl on the absorption edge should also be considered
The authors are thankful to Dr B.K. Sarkar, Director of the Institute, for his kind permission to publish this paper. They are also thankful to the Chemical Laboratory of this Institute for their help in the analysis of lead and Mr Samiran Bhattacharjee, Jadavpur University, for his help in obtaining some of the spectra. Thanks are also due to the Department of Electronics, Govern-
A. Patra, D. Gangufi /Alkoxide-derived silica sols and gels ment of India for providing funds under Sanction no. MATE/I/5(18)/89.
References [1] D. Ganguli, J. Non-Cryst. Solids 101 (1988) 117. [2] N.P. Bansal, J. Am. Ceram. Soc. 73 (1990) 2647. [3] D. Avnir and V.R. Kaufman, J. Non-Cryst. Solids 92 (1987) 180. [4] B.E. Yoldas, Appl. Opt. 21 (1982) 2960. [5] E.J.A. Pope and J.D. Mackenzie, J. Non-Cryst. Solids 87 (1986) 185. [6] M. Yamane, in: Sol-Gel Technology for Thin Films, Fibers, Preforms, Electronics and Specialty Shapes, ed. L.C. Klein (Noyes, New Jersey, 1988) p. 200. [7] M.W. Colby, A. Osaka and J.D. Mackenzie, J. Non-Cryst. Solids 99 (1988) 129. [8] J. Wong and C.A. Angell, Glass Structure by Spectroscopy (Dekker, New York, 1976) p. 152. [9] N.F. Mott and E.A. Davis, Electronic Processes in NonCrystalline Materials (Clarendon, Oxford, 1971) p. 249. [10] A. B~rbulescu and L. Sincan, Phys. Status Solidi (a)85 (1984) K129. [11] F. Dienert and F. Wandenbulcke, C. Rendus 178 (1924) 564.
171
[12] E. Laskin, Kolloid Z. 45 (1928) 129. [13] C.B. Hurd, J.W. Rhoades, W.G. Gormley and A.C. Santora, J. Phys. Chem. 62 (1958) 882. [14] J. Depasse and A. Watillon, J. Colloid Interf. Sci. 33 (1970) 430. [15] S.K. Mookerjee and S.K. Niyogi, Cent. Glass Ceram. Inst. Bull. 22 (1975) 1. [16] L.H. Allen and E. Matijevic, J. Colloid Interf. Sci. 31 (1969) 287; 33 (1970) 420; 35 (1971) 66. [17] C.J. Brinker and G.W. Scherer, Sol-Gel Science (Academic Press, San Diego, CA, 1990) p. 103. [18] D.L. Wood and E.M. Rabinovich, J. Non-Cryst. Solids 107 (1989) 199. [19] S. Sakka, in: Sol-Gel Science and Technology, eds. M.A. Aegerter, M. Jafelicci Jr., D.F. Souza and E.D. Zanotto (World Scientific, Singapore, 1989) p. 76. [20] E.J.A. Pope and J.D. Mackenzie, J. Non-Cryst. Solids 101 (1988) 198. • [21] B.M. Cohen, D.R. Uhlmann and R.R. Shaw, J. Non-Cryst. Solids 12 (1973) 177. [22] L.L. Hench, S.H. Wang and J.L. Nogu~s, SPIE, Multifunctional Materials 878 (1988) 76. [23] C.J. Brinker and G.W. Scherer, Sol-Gel Science (Academic Press, San Diego, CA, 1990) p. 624.