Abstract: Effect of heat treatment on the water adsorption and surface acidity of two nano-ball allophane samples with varying Si/Al ratio under different relative humidities (RHs) was studied. The water vapor adsorption of two allophane samples under various relative humidities, decreased with preheating treatment up to 400 °C for 2 h. The decrease in water adsorption at monolayer level (RH≤0.45) was greater for KnP sample than for KyP sample, whereas the decrease in water adsorption due to capillary condensation between allophane unit particles (RH≥0.6) was greater for KyP sample. These indicate that allophane hollow spherical particles in KyP sample were directly connected each other with the preheating, but those in KnP sample were not. Heat treatment caused the enhancement in the surface acidity of nano-ball allophane samples. The enhancement in the surface acidity after heat treatment is attributed to the inductive effect on the Si-OH groups present at the pore region of the hollow sphere. The results showed that surface acidity of the allophane with higher Si/Al ratio (KnP) was stronger than the (KyP) sample having lower Si/Al ratio. This trend was observed under RH between 0 and 75%; then the acid strength for the two samples was the same at RH of 98%. After the heat treatment at lower level of RH, the surface acidity of KnP was higher than KyP. The presence of polymerized silicate tails exposed outside of hollow spherical allophane particles (KnP), causes the enhancement of the BrØnsted acidity and also prevent direct connection between the particles after heating.
INTRODUCTION
Allophane separated from weathered volcanic ashes and pumices has aggregate morphology composed of fine hollow spherules with diameter of 3.5 to 5.0 nm (Ghoneim et al., 2007). The chemical composition of hollow spherules wall varies in Si/Al atomic ratio from 0.5 to 1.0 (Parfitt and Henmi, 1980; Henmi et al., 1981). Wall of the particles has open pores of about 0.35 nm in diameter (Wilson et al., 1986). Because of the particles morphology and resultant high degree of surface activity, gases and liquid including water could be adsorbed easily onto inside, outside and pore region. We determined water vapor adsorption isotherm on nano-ball allophane in relation to detailed chemical structure of the allophane and found that the amount of monolayer water adsorption is close to the amount of functional groups exposed on the allophane structure (Khan et al., 2006a). This indicates that water molecules are firstly connected to the functional groups such as silanol and aluminol groups on the allophane by hydrogen bonding. Molecular orbital calculations also showed that the adsorption affinity or strength of the hydrogen bonding of water molecule was greater for silanol groups than for aluminol groups (Khan et al., 2006a).
Nano-ball allophane has surface acidity that is weaker than kaolinite and montmorillonite but stronger than for immogolite. The functional groups (Si-OH and Al-OH) in the nano-ball allophane structure are able to behave as BrØnsted acidity sites. Elhadi et al. (2001) using molecular orbital calculation determined that BrØnsted acidity of Si-OH groups of allophane near the adsorption sites increased with molybdate adsorption. There is a regular tendency in physiochemical characteristics such as structure stability against grinding, heating, change in CEC, AEC and surface acidity depending on Si/Al ratio of allophane samples (Henmi et al., 1997).
In case of nano-ball allophane, thermal treatment is greatly change its water adsorption characteristics, because heating causes close approach or bonding of each hollow spheres, leading functional groups at inner and outer surfaces and pore region inaccessible to the water molecules (Khan et al., 2006b). Therefore purpose of this research was to investigate the effect of heat treatment on water adsorption and surface acidity behavior of nano-ball allophane in relation to changes in chemical structure, aggregation state and water-accessible functional groups of nano-ball allophane. Further, the effect of preheating on water adsorption and surface acidity may different between allophanes with different Si/Al ratio, because the ratio is known to affect thermal reaction of allophane (Henmi et al., 1981). Therefore two allophane samples with different Si/Al ratio were used in this experiment.
MATERIALS AND METHODS
Weathered pumice grains from two different locations in Japan were used in this experiment during 2007. The first sample with lower Si/Al ratio, KyP, was collected near Mt. Daisen in Kurayoshi, Tottori prefecture. The second sample with higher Si/Al ratio, KnP, was collected near Mt. Aso in Kakino, Kumamoto prefecture. To obtain nano-ball allophane samples without contaminants such as volcanic glasses, opaline silica, imogolite and organic matter, only the inner part of the pumice grain was used and fine clay fraction (<0.2 μm) was separated (Henmi and Wada, 1976). The separation was carried out by centrifugation after ultrasonification at 28 kHz and dispersion at pH 4 for KyP and at pH 9 for KnP (Henmi and Wada, 1976). The collected fine clay fraction was flocculated by NaCl, washed with water, freeze-dried and used as sample. The Si/Al atomic ratios of the samples were 0.67 for KyP and 0.99 for KnP, respectively. Figure 1 shows chemical structure of nano-ball allophane with lower and higher Si/Al atomic ratio: the KnP sample contains much more polymerized silanol groups at inside and pore region.
| Fig. 1: | Chemical structure of nano-ball allophane. A: Full structure, B and C: Atomic arrangement in cross section at pore region |
| Table 1: | List of Hammett indicators used for surface acidity measurement and ranking to express H0 values |
For water adsorption, one gram of freeze-dried sample for each nano-ball allophane samples, KyP and KnP was taken in glass weighing bottles and was heated at 100 to 400°C for 2 h in a thermostated furnace. For each heat treatment new freeze-dried sample was used. After the heat treatment, the bottles were kept in a dessicator containing silica gel for cooling. After cooling, the bottles were placed in desiccators with saturated solutions of different salts for establishing desired humidity levels at constant temperature (20±1°C) and mass was checked every five hours until a constant value had reached. The various salts used for the humidity control were LiCl (relative humidity (RH) = 0.15), CH3COOK (0.20), CaCl2 (0.31), KNO2 (0.45), Na2Cr2O7 (0.52), NaNO2 (0.66), NaClO3 (0.75), (NH4)2SO4 (0.81), ZnSO4 (0.90) and Pb (NO3)2 (0.98). The RH of 0 was achieved through continuous evacuation with P2O5 powder.
The acid strength for each sample after heat treatments and equilibrating at various relative humidity levels for 24 h were then measured using the method of walling (1950) and Benesi (1956). The acid strength of the samples was estimated from the color of Hammett indicators with known pKa values (Hammett, 1940), where suitable indicators are used as base for measurement of the change in strength of surface acidity. The indicators used are shown in the Table 1. The concentration of each indicator was 0.5%, dissolved in benzene solution. The acidity (H0) was expressed with ordinary numbers from rank I to rank VI, as shown in Table 1. After equilibrating the samples under different RHs levels, about two drops of indictors were added to the samples and the change in color was observed by the naked eye. The acid strength was then estimated using Table 1, for example a sample that give a red coloration with butter yellow but was yellow with o-aminoazotoluene was estimated to have acid strength of H0 = 3.3 ~ 2.0 or rank III.
RESULTS AND DISCUSSION
Change in Water Adsorption of Allophane with Heat Treatment
The isotherms indicate that the amounts of water adsorbed by allophane
samples increased with increasing RH or water vapor pressure (Fig.
2). Without heat treatment, the water adsorption isotherms were type
II (sigmoidal function curve) as commonly found for the sorption of water
by clay minerals (Newman, 1987).
| Fig. 2: | Adsorption isotherms of water vapor on two allophane samples (KyP and KnP) as affected by heat treatment |
Water adsorption increased relatively rapidly until RH = 0.3, increased gradually between RH = 0.3-0.6 and again increased rapidly above RH = 0.6. The water adsorption until RH of about 0.5 was attributed to monolayer water adsorption onto functional groups at inner surface (Si-OH), outer surface (Al-OH-Al) and pore region (Al-OH2, Si-(OH)2 and Si-O-Al) of the nano-ball allophane structure (Khan et al., 2006a). From molecular orbital method simulations, the water adsorption at inner surface and pore region were estimated to be stronger than at outer surface and occurred first followed by adsorption at outer surface region (Khan et al., 2006a). However, in allophane structure, the calculated proportion of the inner surface and pore region to the total surface is approximately one sixth and the rest is the outer surface (Abidin et al., 2006). Therefore even at lower RHs, water adsorption might occur almost simultaneously at all the surfaces although inner surface and pore region were more preferred. This is the reason for the relatively gentle increase of the isotherms at lower RHs (Fig. 2) as compared with those of materials with micropore such as zeolites.
The water adsorption beyond RH of about 0.6 is due to capillary condensation at inter-particle region of the hollow spherical allophane unit particles. When compared the two unheated allophane samples, water adsorption at this region was steeper for KnP than for KyP. This means, even for the unheated sample, more aggregation or attachment between hollow spherical particles have occurred for the KyP sample with lower Si/Al ratio, leading water molecules inaccessible to the inter-particle region.
After heating, the amount of water adsorption of the samples decreased with increasing preheating temperature through all RH levels. The decrease was steep until preheating temperature of 200°C and became gradual until 400°C (Fig. 2).
As has been mentioned, there are three different water adsorption sites exist in allophane, i.e., outer and inner surfaces and pore region. By heating, allophane hollow spherical particles come close to each other and therefore the outer surface of some particles is blocked by adjacent particles. In this case water molecules are not able to go near the blocked outer surface. This might be one of the reason for lower water adsorption at monolayer level after the preheat treatment. Another reason for the decrease in water adsorption with the preheat treatment is blocking the pore region of allophane particle, because water adsorption to the inner surface of allophane particle requires penetration of water molecules from the pore. If the pore was blocked by adjacent particles, the pore region and also the inner surface become unavailable to water molecules. As a whole, monolayer water adsorption on allophane decreased due to close approach between allophane particles with the preheat treatment.
However, as had been noticed from isotherms in Fig. 2, the decrease in monolayer water adsorption with the preheat treatment was greater for KnP sample with higher Si/Al atomic ratio and decrease in water adsorption at higher RHs where inter-particle capillary water condensation occurred is clearly greater for KyP sample (Fig. 2). These mean that the blocking effect of the outer and inner surfaces and pore region of allophane particles toward water molecules and way of the close approach between particles with heating are different between KyP and KnP samples.
From the plots (Fig. 3) the difference in the effect of preheating between the two samples became more clear. For KyP sample with lower Si/Al ratio, the decrease in water adsorption with heating continues until the highest RH, indicating both monolayer adsorption and capillary condensation decreased with the preheating. The decrease in the amount of capillary condensation means close contact between allophane particles and loss of outer surface available for water adsorption. Because the decrease in monolayer water adsorption at outer and inner surfaces and pore region is relatively small for KyP sample (Fig. 2), it can be concluded that the decrease in monolayer water adsorption for KyP sample is assumed to be almost due to blocking of the outer surface and inner surface and pore region remained nearly unaffected by the preheating.
On the other hand, for KnP, the decrease finished within the range of monolayer adsorption and capillary condensation was not affected by the preheating, indicating inter-particle region of hollow spheres in KnP sample remained unaffected by the preheating. This further indicates that accessibility of water molecules toward outer surface of the particle is also not affected: the allophane particles in KnP are not directly contacted each other after the preheating. Therefore in KnP, decrease in water adsorption in monolayer level is assumed to be ascribed almost only to the blocking of inner surface and pore region.
The distinct difference between the two allophane samples in the mode of contact between hollow spherical particles is supposed to be due to Si tetrahedra attached accessorily on the allophane structure. It is already shown that fundamental structure of allophane is that with Si/Al = 0.5 and allophane samples with higher Si/Al ratio, such as KnP, contain polymeric Si tetrahedra in their structure (Shimizu et al., 1988). The accessory Si attach not only to the inner Si-OH, but also to Si-(OH)2, Al-OH2 and Si-O-Al groups at the pore region to create polymerized Si tetrahedra. If the polymerized Si was formed at the pore, the hollow spherical allophane particles become possess tails projected outward.
With heating, for KnP sample, the tailed polymeric Si tetrahedra may be connected each other, leaving space between two allophane particles (Scheme 1). This way of contact between allophane particles in KnP sample keeps outer surface accessible to water molecules. On the other hand, in case of KyP with lower Si/Al ratio, little tails are projected and connection between particles may occur directly, irrespective of the pore region (Scheme 1). As has been described, the direct connection between particles is already seen in the unheated KyP sample (Fig. 2). In the natural allophane sample preparation processes, it was noticed that allophane samples with lower Si/Al ratio tend to coagulate by air-drying and also by freeze-drying (unpublished data). The direct connection in the KyP sample was further enhanced by the preheating, leading to the decrease in water adsorption due to capillary condensation (Fig. 2).
| Fig. 3: | Change in water adsorption of allophane samples with heat treatment at 200°C |
| Scheme 1: | Aggregation state of KyP and KnP samples after heating |
Although the presence of the tails in allophane with higher Si/Al ratios has not been proven by such as electron microscope, difference in flocculation-dispersion behavior between allophanes with higher and lower Si/Al ratios can be interpreted by the tails. As has been reported (Henmi and Wada, 1976) and described in the experimental section of this report, allophane with lower Si/Al ratio and also imogolite (Si/Al = 0.5) disperse well at acidic condition but flocculate at alkaline pH, whereas allophanes with higher Si/Al ratio, such as KnP, disperse well at alkaline condition. Because the `tailes`is composed of polymerized silicates with silanol groups, the tails will have negative charges due to dissociated silanol groups at alkaline conditions. This is the reason for high dispersibility of allophanes with higher Si/Al ratio at alkaline conditions. In case of allophanes with lower Si/Al ratio and imogolite, negative charges or dissociated silanol groups at higher pH exist only at the inner surface and the outer surface, Al-OH-Al, remains electrically neutral: therefore they flocculate at higher pH region. The difference in Si/Al ratio of allophane is known to affect other physicochemical properties of allophane such as CEC and surface acidity (Henmi, 1985) and the effect have to be re-explained by using the tails.
Change in Surface Acidity of Allophane with the Heat Treatments
The result showed that at a relative humidity (RH) between 0 to 75% the acid strength of the unheated allophane sample with high Si/Al ratio (KnP) was stronger than that of its low Si/Al ratio sample counterpart (KyP) (Table 2). However, at RH of 98%, the acidity of both samples was the same. The stronger acidity exhibited by KnP samples can be attributed to the enhancement of BrØnsted acidity on the silanol groups the fact that KnP, has more of the accessory condensed silanol groups attached to the main frame as shown in Fig. 1. These results are in line with Henmi et al. (1997). For the lower Si/Al ratio (KyP) sample, results indicated that with increasing relative humidity (RH) from 31 to 98%, there was no apparent change in the acid strength (H0= II). This result is attributed to the water molecules occupying the same number of acid sites, however heating the allophane samples up to 105°C or evacuation against P2O5, increase acid strength up to H0 value of (+3.3 ~ +2.0) or rank III. It is well known that, when the mineral surfaces become drier the BrØnsted acidity and protons are concentrated in a smaller volume of water resulting in more extreme surface acidity (Haung, 2000). Even very weak bases, i.e., poor proton accepters can be protonated on the surface of such mineral. While on the other hand, at RH of 98%, both unheated allophane samples have the same rank II (H0= 4.0 ~ 3.3) of surface acidity. These results are credited to higher moisture level, because at 98% RH, the water molecules engaged almost all the available sites for the acidity, resulting in very weak BrØnsted acidity. In addition, when the nano-ball allophane samples become wetter, the H+ ions are concentrated in large volumes of the water and subsequently resulting in weak BrØnsted acidity (Ghoneim et al., 2007). Similar results were obtained by Henmi and Wada (1974); they reported that the acid strength of clay is lowered when it has been equilibrated in an atmosphere of higher relative humidity indicating that the adsorption of water on the clay surface changes stronger acid sites into weaker one.
| Table 2: | Measure acid strength (H0) of allophane samples before and after heat treatment |
| Table 3: | Possible origin of surface acidity of allophane before heat treatment |
In general, the acid strength of both KyP and KnP samples was observed to increase after heat treatments as well as with decreasing relative humidity. For KyP, the acid strength increased from rank II to III at relative humidities of 31 to 98% with the increase in heat treatments. However a strong increase in acid strength was observed after heat treatment at relative humidity of 0%. The increase was from rank II (unheated) to rank V (with heat treatment); again the degree of increase was the same among the different heat treatments. These same degree of strengthing observed among the treatments can be attributed to the loss of almost all of the water adsorbed on the surface of the allophane sample, so the acid strength was only determined by exchangeable Na and hydroxyl on the silanol (Si-OH) and (Al-OH2+) groups.
In case of KnP, the allophane sample with high Si/Al ratio, the acid strength increased from rank II to III at relative humidity of 98%. While no change was observed at relative humidity of 75%, but an increase from rank III to rank V was noticed at relative humidity of 31%. The increase in surface acidity was different between heat treatments of 200 to 300°C, on the other hand no change in surface acidity was found between 300 and 400°C. However, at relative humidity of 0%, no change in acid strength was recorded, this may be due to evacuation with P2O5 the sample losses almost all water adsorbed on its surfaces, in addition rank V is regarded is a wider range (e.g., = +1.5 ~ -3.1). The increase in the acid strength after heat treatment for the two nano-ball allophane samples KnP and KyP at the RH of 31% is attributed to the H2O coordinated the exchangeable sodium on the allophane surface. From the (Table 2) it is clear that in KnP having higher Si/Al ratio, the increase in acid strength was higher as compared to KyP, at RH of 31%. This higher increase in the acid strength of KnP sample, as compared to KyP can be correlated to the higher decrease in water adsorption for KnP at lower level of RH after heat treatments (Fig. 2). This higher decrease in water adsorption with the heat treatment for KnP, provide more vacant sites, to the Hammett indicator, where water molecules were not present to neutralize their effect. This leads to high surface acidity in case of KnP, sample at lower level of RH after the heat treatment of 300 to 400°C. The surface acidity of clay minerals arises from two main sources: the BrØnsted acidity and Lewis acidity. The BrØnsted acidity comes from the deprotonation of the H+ on the constant charge, water molecules, which coordinated to surface bound cations and also from the dissociation of functional groups. On the other hand, Lewis acidity originates from un-saturated structural silicon and aluminum, from constituents ions such as aluminum and iron that exposed at the edges of the minerals (Sposito, 1984; Wu et al., 1992; Mcbride, 1994). Table 3 shows the possible origin of surface acidity before heat treatment.