Función de los materiales didácticos

Silica polymerisation and deposition, on the RO membrane surface, has been researched experimentally (Semiat and Hasson 1996, 1999, 2003, Gill 1996, Bowen 1979) and more recently computational simulations have been performed (Jianjun, et al, 2006), where the molecular mechanism and rate of hydrolysis have been explored through calculation of the reaction barriers and pathways. Both studies showed that the main factors influencing silica polymerisation are pH, temperature, saturation, impurities present in the solution, and the autocatalysis effect of already precipitated silica that accelerates further precipitation. Total silica surface area in solution is also a factor determining the rate of silica polymerisation. Carbonate hardness also accelerates

30

further polarisation and precipitation as does magnesium. The presence of magnesium enhances silica polymerisation and also depresses silica solubility when calcium was present.

Computational simulations found that the siloxane bond, often presented on silica surfaces, is difficult to hydrolyse because of the high reactivation energy barrier, especially with the aid of hydrolysis (Jianjun et al., 2006). However, monomers of silicic acid condense to form larger oligomers, which link together to produce primary particles (nucleation). Depending on process conditions, these particles can either grow by reaction with monomers or grow by aggregation (Iler 1976). Aggregation can lead to gelation of the colloidal suspension, but not necessarily to silica deposition on the membrane surface (Bergna 2006).

The research by Bergna (2006, 1994), Sjoberg (2008, 2001), Iler (1979), Kiselev (1978), Baldyga (2009) and others provide insight into the physical and chemical processes involved. The experimental data are rather debatable, and there is no general agreement about silica hydrolysis and condensation in aqueous solutions and complexation to cations and anions at the particle – water interface. Healy (1994) referred to the behaviour of colloidal silica as “anomalous silica sols” because the Derjaguim-Landau-Verwey-Overbeek (DLVO) theory seem to be unable to cope with silica hydrolysis while it explains satisfactory the behaviour of all other colloidal systems. For example, it is well known that the silica sols are stable at their point of zero charge and that they also coagulate in alkaline solutions, in which their electrical surface charge is high and where their stability should increase. Such behaviour is very unusual. Yates (2000) proposes a thermodynamic approach to replace the failing DLVO theory. According to Healy (1994), on other hand, the DLVO theory should give a coherent description of the aqueous silica behaviour on the condition that all the forces that play a role in the interaction are introduced in the model. Stability of silica at the point of zero charge may be explained by steric stabilisation from oligomers on the surface of the silica sols (Healy 1994) that prevent aggregation.

31

Membrane scaling phenomena are governed by the silica solubility limit prevailing in the CP layer on the membrane surface (Semiat 2003, 2001, 1996). According to Semiat (2003), the rate of change in the silica scale formation during the course of RO processing is dictated by two opposing trends: the concentration effect due to permeate withdrawal which acts to increase the silica scale formation, while the decline in permeate flux due to scaling and osmotic effects acts as to decrease the rate of silica scale deposition. Permeability decline data provides a more accurate characterisation of the silica scaling process (chapter 5 of this thesis). What is not yet clear from the experimental silica studies by Semiat (2003) is the impact of dissolution (hydrolysis) on existing silica deposits and on colloidal silica present in the concentration polarization zone (i.e., close the membrane surface) and how on-going processes of hydrolysis and condensation effect silica polymerisation in this zone. Will silica deposit on the membrane surface as a result of monomer silica groups concentration exceeding the practical solubility limit or will it remain in dispersion and why?

Baoxia and Elimelech (2012) tried to explain silica scaling reversibility in RO process by proposing three steps of both homogeneous and heterogeneous nucleation processes on the membrane surface. However, mechanisms of silica precipitation leading to two different nucleation processes are conflicting to what can be expected for homogeneous silica nucleation. The diagram, figure 2.3, describes indicative silica species distribution in different silica solubility zones, for various pH and concentrations. These silica species arise from published light-scattering experiments and help to define the pH- concentration domain in which multiple ions are present as precursors to silica polymerisation (Stumm and Morgan 1987).

32

Figure 2.4 – Dissolved silica species polymerisation (Q0 _{< Q}1 _{< Q}2 _{< Q}3 _aggregation)

path into amorphous silica structure (mechanism proposed by Iler (1976).

According to Ilter’s (1976) description of silica chemistry and dissolved silica polymerisation pathways (Figure 2.4), increasing dissolved silica concentration well above solubility limit in the CP layer will initiate silica precipitation likely firstly by aggregation of monomeric silica acid in bulk solution before deposition on the membrane surface. This is contrary to what was proposed by Baoxia and Elimelech (2012), who claimed in their RO research that monomeric silicic acid could attach to the membrane surface without the presence of nucleation coupling points. Dietzel (2003, 2001, 1997) also illustrated in his analytical study of dissolved silica polymerisation pathways that dissolved silica species in the polymerisation process follows as Q0 _{< Q}1

< Q2 < Q3 aggregation path until they react with other impurities. It is possible, however, the work completed by Baoxia and Elimelech (2012) considered the presence of other potential coupling points on the membrane surface such as –OH, -COOH groups and coupling points that are a function of the water chemistry from precipitates such as Al(OH)n, Fe(OH)n leading to monomeric silicic acid coupling to the membrane

surface.