C YCLIC VOLTAMMETRY

The use of laponite clay pillared with the cubane – like [(H3N-Pr)8Si8]⁸⁺ octamers for electrode surface modification and ion - exchange properties of the resulting coatings have been reported [160, 161]. The present electrochemistry results are intended to demonstrate the ability of montmorillonite to bind the octameric [(H3N-Pr)8Si8]⁸⁺ cubes not only at its ability sites but also in the form of physically adsorbed ion - paired [(H3N-Pr)8Si8](X)8 species. The latter can participate in an anion exchange reaction in which X¯ is replaced by some other anionic, redox active species present in solution, e.g. IrCl6

2¯ . As a result of this exchange the redox - active anions are effectively preconcentrated within the clay coating. Figure 3.3.11. compares the cyclic voltammetric response from K2IrCl6, 1mN solution in 0.1 M acetate buffer at pH = 4,

recorded with a bare graphite electrode (a), and with the same electrode modified with a film of sodium montmorillonite loaded with [(H3N-Pr)8Si8

8+, 1 CEC (b) and 3 CEC (c). It is easily seen that the IrCl62

¯ is concentrated within the clay coating, leading to an enhanced current response compared to the unmodified electrode response, only when the quantity of intercalated octamer exceeds the CEC of the mineral, curve (c).

Figure 3.3.11. Cyclic voltammograms (100 mV / s) of K2IrCl6, 1 mM in 0.1M acetate buffer at pH = 4 on (a) a bare graphhite electrode and the same electrode modified with clay - APTEOS compositions, (b) APTEOS loading 1 CEC, and (c) 3 CEC.

The electrode from experiment (c), i.e. coated with IrCl6

2¯ - APTEOS - CME, was then transferred to pure electrolyte solution. The current intensity accounted for about 70% of the previously accumulated electroactive species. Rinsing the electrode in pure water four times and then immersimg it again into the electrolyte solution gave a CV very similar to the response from (b). Obviously, this time leaching of the ion-paired

[(H3N-Pr)8Si8][IrCl6]4 into the aqueous solution deprives the clay coating from most of the electroactive IrCl6

2¯ anions.

Conclusion

The main conclusion of this work is the high tendency of smectite clays to insert into their surfaces three - dimensional organosilicon oligomers in amounts exceeding the CEC of the mineral, the excess being in the form of physically adsorbed ion pairs. The presence of ion – paired silsesquioxanes in the clay galleries influences the free distance between the layers and the surface area and porosity of the silica – pillared compositions.

SUMMARY

The concept of pillaring smectite clays was demonstrated more than 50 years ago by Barrer and McLeod when they utilized tetraalkylammonium ions to induce interlayer porosity in montmorillonite. From this time on, it was extensive interest in pillared clays, since their pore sizes can be made larger than zeolites. Moreover, by varying the size of the pillar or the spacing between pillars, or both, one may adjust the pore size to suit a particular application. Thus pillared clays offer new possibilities for catalysis of larger molecules such as those found in residual crude oils. The remarkable thermal stability of Zr4 and Al13 pillared smectites has been attributed to the formation of metal oxide cluster upon dehydroxylation of the hydroxy cations at elevated temperature. Although polynuclear hydroxy metal ions formed by hydrolysis in aqueous solution can yield pillared clays with interlayer free spacings in the range 5 to 20 Å, the number of metals that form suitable oligomeric species is limited. New approaches to the pillaring of smectite clays promise to extend the number of pillaring species. In general, the pillaring process is carried out with diluted systems: a diluted clay suspension (less than 1wt%) is brought in contact with a diluted pillaring solution (0.5M for aluminum). From an industrial viewpoint, in which large amounts of clay have to be prepared, this process may result uneconomic, since enormous equipments as well as the manipulation of huge volumes of water are required. As recent literature indicates, however, this problem has been overcome, and new processes using 6-10wt% or some trial 40wt% clay suspensions have been developed which use pillaring suspensions with much higher concentrations. They observed that the intercalation occurs mainly during the washing step, where, indeed, the system is highly diluted.

In this research work three types of pillared clays, namely alumina-, iron-oxide, and silica-pillared materials were produced and characterized. A brief review was first given on preparation and physical and chemical characterizations of pillared clays, and

SURSHUWLHVRIVPHFWLWHVHVSHFLDOO\+XQJDULDQEHQWRQLWHV7KHOHVVWKHQ PIUDFWLRQV

were collected from the natural bentonites and also were sodium-saturated by reacting the suspension of the clays overnight with an excess of 1 N NaCl solution, and then washing the solid phase, gained by centrifugation. Among Hungarian bentonites

Na-Mád and Na-Koldu have the greatest difference between the (001) basal spacings of original and ethylene glycol treated montmorillonite samples, what is means the highest swelling ability. Na-Istenmezeje bentonite has a double 001 XRD peak as Wyoming bentonite has but its expandibility was the lower. Among the chosen Hungarian bentonites the Istenmezeje has the biggest CEC value, than coming the Koldu and Mád bentonite. The sodium forms of the present Hungarian bentonites have the same CEC value (95±1). From these results it is clear that all Hungarian bentonites is very good parent clay for any kind of pillaring reactions.

The character of the studied Hungarian bentonites changed by doing the alumina intercalation procedures. The successfully intercalated montmorillonites with aluminium-polyoxycations were settled to the bottom of beaker, and the sediments looked as a velvet and were softly light. On the basis of screening experiences it was concluded that the efficiency of the pillaring reaction was better by adding the solutions of pillaring agents to the clays suspensions. In the intercalated products were detected well discernible color and textural differences for lower (10g/L) and higher (15g/L) concentration clay suspensions. After calcination at 350ºC and 500ºC in air the color of the samples was darker (at higher temperatures deeper), but the texture was the same in the case of intercalated samples.

A basal spacing XRD pattern around 18.5-19.5 Å indicated right away that alumina pillaring occurred since the layer thickness of an anhydrous smectite layer is 9.5 Å and the size of the Al13

7+ Keggin ion is about 9 Å. The lowest d-spacing value of XRD patterns was obtained for the Istenmezeje intercalated montmorillonite. In this case the intercalation was poorer and more unstable, since the change in basal spacing is the highest value, precisely 37% both for original and Na-form clays. These per cent values were in agreement with the SEM morphological data, namely the appearance of Koldu and Mád particles was very similar to each other, while the Istenmezeje bentonite had a bit different features; the most jagged, frilled surface of the adhered bunches and the least grain size.

Considering the change of basal spacing in the case of intercalated structures at 500ºC temperature; Koldu samples have the strongest decrease (29%, 28% and 7%, 6% for sodium form) showing that these pillared structures were the least of stable. Similarly,

the Istenmezeje samples have high loss values in the original Ca-rich form (28%, 22%

however 0%, 2% for sodium form). The d(001) values of Istenmezeje intercalated montmorillonites were the lowest, the calcined samples had line broadening, as well as the basal peaks of clays were partly returned. During calcination at 500ºC the sodium-form Istenmezeje clay samples preserved the interlayer distance, and also the sharp line profile. Istenmezeje pillared montmorillonite samples have lower thermal stability than other clays since the basal spacing of sintered pillared products was below 18Å.

Calcined Mád clay samples displayed the least decrease in d-spacing (3%, 1% and 7%, 3% for sodium form) indicating the best pillared samples for sodium form, both for lower and higher concentrations of clay suspension. Pillared samples prepared by 15g/L concentration of clay suspension disclosed lower d-spacing loss for all clay than 10g/L at 500ºC temperature.

Mád montmorillonite samples had again the lowest decrease in d-spacing (0%, 2% and 3%, 0% for sodium form) at 350ºC temperature showing the best pillared layered structures in all cases. Both the Istenmezeje and the Koldu samples displayed considerable decrease in 001 basal spacing (less for sodium form).

Furthermore, the data mentioned about Hungarian pillared structures also demonstrated the well-known fact for overseas (non-Hungarian) clays that the sodium forms have higher tendency or they are more suitable for good pillaring by producing controlled structures than the original forms of montmorillonite.

The minimum criteria for a material to be called pillared are chemical and thermal stability and molecular distribution of pillars, but no order of the pillars is required in the interlayered region. Applying the same pillaring procedure for all chosen Hungarian clay after heating to 350ºC and 500ºC temperatures the expected d(001) spacing value at about 19Å can be detected for Mád samples but not for the Istenmezeje and Koldu samples. The basal spacings of Istenmezeje samples are below 18Å that demonstrated the not too high thermal stability of these pillared clays. It was a little bit surprising since Istenmezeje clay has the highest (90 meq/100g for original and 95 meq/100g for sodium form) cation exchange capacity (CEC) as compared to the other Hungarian bentonites. Mád and Koldu pillared layer structures absolutely fulfilled the minimum criteria of pillared material according to the X-ray.

Since the final criterion of the pillared structures is accessibility of the interlayer region by molecules at least as large as nitrogen, porosity and surface area investigations were required. The Mád and Koldu materials were studied by these methods. The specific surface area was typically obtained from the adsorption branch by applying the BET equation. The BET surface area of some of original Hungarian montmorillonites was a little bit higher than the average surface area value (50 m²/g) of overseas bentonites. Alumina pillared Hungarian clay minerals (Na-Mád and Na-Koldu) had specific surface areas around 150-200 m²/g that nicely met the accepted requirements.

The isotherm of the alumina-pillared clay mineral was Type I isotherm. A significant increase in N2 adsorption in a lower relative pressure interval was observed for all these samples, indicating the presence of micropores. The mean micropore size of pillared structures can be different from the free interlayer spacing measured by XRD. This results in various pore sizes and accordingly pillared structures reveal pore size distribution (PSD) instead of a fixed value. The mean micropore size is the average of all micropores so that it should be an intermediate value between the pore sizes of the main pore groups. The mean pore size of alumina pillared structures may be higher than the interlayer distance measured by XRD 19 Å for the Koldu and 25 Å for the Mád samples.

The actual distribution of the pillars in the Hungarian alumina pillared montmorillonites requires additional investigation considering their elemental composition. As expected the Si/Al ratio was the highest for the starting Koldu and Mád montmorillonites, 2.14 and 2.83, respectively, while in the pillared clays, this value decreased showing that a high amount of aluminum has been incorporated. The exact structure of the pillars requires additional investigation and careful calculation taking high attention for error range and variation in Si/Al ratio with the number of pillars.

Iron-intercalated clays were prepared using two kinds of iron-pillaring agents, namely iron-chloride and iron-benzoate. Iron-intercalated structures were depended critically on the hydrolysis conditions of iron(II)ion. It was considered, how the base to metal ratio is related to the hydrolysis process, and simultaneously followed up the basal-spacing and surface area of iron pillared products. The pH of the solutions were kept under 3.5, because at this value the iron is formed iron(II)oxide precipitate on the

surface of the silicate layers. By increasing the base to metal ratio it was detected higher surface area and bigger basal spacing values. It was also studied the effect of different kind of preparation ways, namely such as shaking, reflux, and stirring, using X-ray diffraction. The structures of the intercalated clays were collapsed under calcination process at 450 °C with only a few exceptions.

A typical octameric silsesquioxane derived from 3-aminopropyl-triethoxysilane (APTEOS) presented certain special features, which made this material ideal pillaring agents for interactions with sodium-smectites. Particular emphasis was given to the the amount of the adsorbed APTEOS exceeded considerably the coverage imposed by the cation exchange capacity of the clay. In fact, it corresponded to about 1.8 times the CEC. However, because the organosilicon cubic octamers are bulky and robust organophilic cations, there was also the possibility of being adsorbed in the form of ionpairs that were derived from the protonation of the organosiloxane units, [H3N(CH2)3Si8O12][OH]8. In the ion-pair mechanism the excess metal chelates has been loosely held by the surfaces and has been easily removed by washing with water.

In an experiment designed to probe the type of aggregation responsible for the excess adsorption of the amino-functionalized organosilicon cubes by the clay surfaces the intercalation of the (H2N-Pr)8Si8 was performed in a 0.05 M solution of sodium sulfate. The purpose was to demonstrate the presence in the clay layers of sulfates, acting as the counteranion of the ion-paired cubes, by infrared spectroscopy and in addition, by the chemical analysis of sulfur to verify the presence of sulfates in the interlayers and to estimate the amount of the physically adsorbed ion-paired cubes. The IR spectrum of clay sample, which was treated with 0.05 M Na2SO4 and then rinsed once with water, did not show sulfate absorptions. The chemical analyses results offered support for the ion-pair formation. It has seen that washing the clay

composition four times with water bringed a decrease in the sulfur content from 2.57 % to 0.38 % implying that most of the sulfate anions, incorporated in the form [(H3 N-Pr)8Si8][SO4]4 ion-pairs, were desorbed from the clay galleries. Similarly the substantial decrease in the carbon content of the clay – silsequioxane product was in line with the loss of siloxane octamers upon water washing. Based on the sulfur analysis we calculated the maximum quantity of the ion-paired [(H3N-Pr)8Si8][SO4]4 cubes incorporated into the clay. This amounted to 0.20 g per gram of the clay complex. The conclusion was that the clay surfaces, apart from the electrostatically bounded octameric silicate units, also hosted additional cubic units in the form of physically adsorbed ion-pairs, which were easily leached into the aqueous environment by, water washing.

The XRD patterns clearly demonstrated that water treatment of the intercalated samples affected greatly the d(001) values. Successive washing bringed a gradual decrease in the interlayer distance from 18.2 Å to 15.9 Å. The XRD results suggested that the preformed cubic silsesquioxanes were introduced as intact units into the clay layers. The cubes were first fixed at the exchanged sites of the mineral and

In this configuration the distance between two parallel levels encompassing the alkyl chains was estimated to be 6.6 Å, a value very close to the experimentally observed intersheet separation, ∆ = 16.0 - 9.6 = 6.4 Å, 9.6 Å being the thickness of the clay sheet. On the contrary, if the alkylamino chains were directed in an upward position then the ∆ value would be expected to be near to the estimated lateral dimension of 11.2 Å of the cubic octamer.

In the flat orientation of the alkyl chains the distance of 11.2 Å between two diametrically opposite -NH3

+ end groups was close to the calculated free distance between two negative sites, 12.5 Å, in clay framework for a montmorillonite with a CEC of 80 meq / 100 g of clay. Furthermore, in this configuration the empty space between the electrostatically bound cubes can be used to provide the necessary

hydrophobic environment for hosting additional silicate cubes in the form of ion-pairs.

In view of the enhanced d-spacing, observed for the unwashed samples, we proposed that these additional ion-paired cubes have their side alkyl chains directed in an upward position in the interlayer zone. After washed sample was calcined at 500 °C the interlayer distance was slightly decreased from 15.9 Å to 15.7 Å. This small difference favored further the parallel orientation of the side organic chains in a washed sample. In such a geometry, the removal of the alkyl chains by calcination would hardly affect the d-spacing, as was indeed observed. Finally, for the calcined sample the intersheet separation, ∆ = 15.7 - 9.6 = 6.1 Å, indicated that the silica pillars comprised two silicon atom layers, since previous studies have shown that one layer of silicon particular value in assigning the orientation of an alkylamino group in the lamellar space of a clay. In accord with our findings, we associated the bands at 1500 cm^-1 and 1560 orientation where the aliphatic chains were inclined at certain angle to the silicate layer.

This was the orientation taken up by the physically adsorbed ion-paired cubes hosted in the empty space between the electrostatically bound cubic units. Inspection of spectrum, from the washed sample, showed that the band at 1560 cm^-1 was missing.

This was evidence that the band at 1560 cm^-1 was derived from the ion-paired, washable cubic units. It is very important that these vibrations did not disappear when the clay suspensions were washed with water. The infrared spectra therefore did not support the formation of hydrogen bonded cubic polymeric associations in the clay interlayers.

The nitrogen adsorption - desorption isotherms for clay samples of first intercalated with the cubic octamer once washed and thoroughly water washed sample, and then

calcined at 500°C were prepared. The adsorption data were treated with the BET equation in the linear part of the isotherm and with the t-plot method. Both samples showed a large hysteresis loop illustrative of a mesoporous structure with a contribution from micropores. The corresponding t - plots were also representative of micropores in the presence of mesopores. The BET surface area was 142 m² / g for the unwashed and 168 m² / g for the washed sample. The washed sample with the higher surface area had a total pore volume of 0.179 cm³ / g with 0.076 c m³ / g attributable to micropores, whereas the low surface area once - washed sample had a total pore volume of 0.123 cm³ / g with the micropores volume being 0.064 c m³ / g. From the higher surface area observed for the washed sample we may concluded that repeated washing of the material caused the evacuation of the micropores from the ion-paired intercalated species that otherwise stuffed the empty space of the clay galleries. It was likely that ion-paired silicate cubes, if remained in the galleries, could generated during calcination aggregates that hinder the entrance of the N2 molecules into the labyrinth of the pillared structure and thus resulted in lower values of surface area and porosity for the unwashed clay.

KEY – POINTS OF THE THESIS

1. Using Hungarian bentonites, namely such as Istenmezeje, Koldu and Mád, the alumina-pillaring was successfully carried out in the case of highly concentrated clay suspensions (10 and 15 wt%) and highly concentrated Al13-Keggin-ion solutions with simple stirring engineering at room temperature. The efficiency of the pillaring reaction was better by adding the solutions of pillaring agents to the Hungarian clays suspensions. Both original and sodium-saturated forms were showing pillared layered structures with basal spacing around 19Å, indicating the successful pillaring procedure. Furthermore, the data mentioned about Hungarian pillared structures also demonstrated the well-known fact for overseas (non-Hungarian) clays that the sodium forms have higher tendency or they are more suitable for good pillaring by producing controlled structures than the original forms of montmorillonite. Mád and

1. Using Hungarian bentonites, namely such as Istenmezeje, Koldu and Mád, the alumina-pillaring was successfully carried out in the case of highly concentrated clay suspensions (10 and 15 wt%) and highly concentrated Al13-Keggin-ion solutions with simple stirring engineering at room temperature. The efficiency of the pillaring reaction was better by adding the solutions of pillaring agents to the Hungarian clays suspensions. Both original and sodium-saturated forms were showing pillared layered structures with basal spacing around 19Å, indicating the successful pillaring procedure. Furthermore, the data mentioned about Hungarian pillared structures also demonstrated the well-known fact for overseas (non-Hungarian) clays that the sodium forms have higher tendency or they are more suitable for good pillaring by producing controlled structures than the original forms of montmorillonite. Mád and