Silanol number

The number of -OH groups present on the PCH was determined to be 1.89 mmol/g or 1.03 OH/nm2. For the different concentrations of Al(acac)3 with respect to the silanol number (25%, 50%, 75% and 200%), the parameter R can be calculated. This R-value is defined as follows and gives us further confirmation about the bonding mechanism ... 

Further calculations with respect to the loading capacity on the surface can be performed knowing the surface area of the support (SA= 999 m2/g) and the mean cross-sectional area of the Al(acac)3 complex (0.60 nm ) [8] A full monolayer would therefore correspond to an Al-loading of 2.7 mmol/g. However, the actual maximal Al-loading obtained here is 0.672 mmol/g, which means only 25% of the monolayer capacity. When comparing this value to the silanol number of 1.89 OH mmol/g, it can be concluded that 36% of the available OH-groups on the surface have reacted with the Al-complex. 

Although in a less polarized way (it is now agreed that -if geminols exist- their relative contribution to the total silanol number is relatively small), this ambiguity exists until today. 

Since Kiselev1 discovered the surface hydroxyl groups on silica in 1936, many studies on the quantification of the silanol number (a0H number of hydroxyl groups per nm2) and on the characterization of the different hydroxyl types have been published. These studies can be divided into theoretical calculations, physical methods and chemical methods. 

The authors made no attempt to determine the silanol number, but concluded that the results obtained reveal a necessity to verify the hypotheses, using silica gel samples containing known amounts of different types of OH groups . 

The determination of the silanol number of a fully hydroxylated silica has been reviewed by many authors.26,27,28,29 In this section, we will focus upon a few studies that estimate the silanol number as a function of the dehydroxylation temperature. This is not possible using theoretical calculations, which can only provide a fair estimation of the silanol number of a fully hydroxylated silica. 

Based upon more than 100 samples, with a specific surface area varying from 5 to 1000 m2/g, Zhuravlev found that the silanol number for a fully hydroxylated silica amounts 4.6 0.5 OH/nm2. This constant is claimed to be independent of the origin and structural characteristics (specific surface area, type of pores, pore size distribution,. ..) of the sample. 

The deuterium exchange method was also used to determine the average value of the silanol number as a function of dehydroxylation temperature (473 - 1373 K in vacuo). The experimental results are shown in figure 4.1. 

Figure 4.1 Silanol number as a Junction of the temperature of pretreatment in vacuo for different samples ofSi02 taken from ref (31) with permission.

Temp, of vacuum treatment (K) Silanol number < 0H, v (OH groups/nm2) Degree of coverage with OH groups (0OH)... 

Reaction (A) has been used by various researchers, in order to determine the silanol number. 

Looking at the reaction mechanism of HMDS with silica, the possibility of an in situ determination of the silanol number arises. The on-line determination of the liberated NH3 during the reaction directly yields the number of silanols reacted with HMDS. FTIR spectroscopy is only used to check whether all hydroxyls have reacted. 

At pretreatment temperatures below 673 K, HMDS is no longer able to remove all surface hydroxyl groups, due the previously described sterical hindrance effects. In these cases, quantification of the silanol number is still possible, using the integrated values of the hydroxyl and C-H infrared bands, normalized to the reference band. 

Figure 4.3 Determination of the silanol number on Kieselgel 60 using the HMDS method.

The silanol number can be considered as a physicochemical constant, independent of the silica type. 

Once the silanol number as a function of temperature is known, further distinction between isolated, vicinal and geminal hydroxyls should be made. Since all silanol types exhibit different reactivities, an exact knowledge of the ratio isolated/vicinal silanols is crucial for a thorough study of a modification reaction. 

Random condensation of bridged silanols yields siloxane bridges, but also causes a relative increase in the free silanol number. 

An evaluation of the functional dependence of the activation energy on coverage is performed with the constant coverage, variable heating method of Richards and Rees.30 The results are shown in figure 5.21. In this figure, the relative coverage 6 (x-axis) is replaced by the silanol number (a0H). 

A remarkable feature of this model is the fact that this is the first model, discussed so far, that proposes a dehydroxylation behaviour of silica, dependent on the porosity of the sample. Whereas the total silanol number is a constant (the Zhuravlev constant), the relative distribution of free and bridged silanols is not. 

In 1993, L.T. Zhuravlev35 published a review article of work performed in the former USSR on the surface characterization of amorphous silica. This review article is a very important document in the study of the silanol distribution. Also, the energetical aspects of the dehydration and dehydroxylation processes are discussed in detail. The determination of the silanol number as a function of treatment temperature has already been discussed in chapter 4. 

The components of the silanol number (free and bridged hydroxyls) were determined by the deuterium exchange method and infrared spectroscopic measurements.36... 

The correlation between the silanol number at 673 K and the intensity of the infrared absorption band is based on the fact that for the samples calcined in vacuo at 673 K, there are practically only free hydroxyl groups. This means that the total concentration of silanol groups on the silica surface at 673 K corresponds to the concentration of free hydroxyl groups. 

In the previous chapter, it was concluded that the silanol number (i.e. the total concentration of silanol groups on the silica surface, expressed in OH/nm2) can be considered as a physicochemical constant, within certain margins of error ( 0.5 OH/nm2). 

It is obvious from figure 6.1 that there is no difference in desorption energies in the silanol region from 0 to 4 OH/nm2, confirming Zhuravlev s statement that the silanol number can be considered as a physicochemical constant. 

Figure 6.1. Water desorption energies as a function of the silanol number for Kieselgel 40 (o) and Kieselgel 60 (O).

The effect of solvent type and aminosilane concentration has been evaluated. The third component in the reaction system is the silica substrate. The surface of the silica gel carries the active sites for adsorption. The concentration of these sites varies with varying silica type, its specific surface area and pretreatment temperature. Additionally, surface adsorbed water has a clear effect on the reaction mechanism. Isotherm data, reported in the previous paragraph, only accounted for fully hydrated or fully dehydrated silica. The effect of the available surface area and silanol number remains to be assessed. Information on these parameters allows the correlation of data from studies in which different silica types have been used. In this part the effect of these parameters in the loading step is discussed. Silica structural effects on the ultimate coating, after curing, are evaluated in the next paragraph. 

The above-cited behaviour of the dehydrated silica (pretreatment temp. = 473 - 873 K) is further exemplified by plotting the total coverage as a function of the silanol number (figure 9.16). 

Figure 9.16 Total coverage of APTS modified dehydrated silica gel (pretreatment temperature = 473-873K) as a Junction of silanol number O Kieselgel 60, a Kieselgel 100.

In analogy to the loading step study, the effect of substrate structure has been studied by modifying mesoporous silica gels with a variable mean pore diameter.28 Sample pretreatment and curing (20 h, 423 K) were performed under vacuum. Variation of the pretreatment temperature causes a change in specific surface area and silanol number. 

In order to study the role of surface silanols, three silicas with a variable silanol number (ot0H) were modified. Data on the silanol content are listed in table 9.7. 

Primary interest goes to the surface hydroxyls found after modification. Unreacted Q3 and Q4 sites are detected. Concurrent with the silanol number before reaction, also the number of surface silanols after reaction decreases with increasing pretreatment temperature. From the general reaction scheme, these bands may be attributed either to surface silanols in hydrogen bonding interaction with amine groups, or to unreacted surface groups. 

In these terms, the ammonia adsorption is not entirely reversible. Figure 12.1 shows the total and irreversible ammonia adsorption (at room temperature) on Kieselgel 60, thermally pretreated at 473, 673 and 973 K. The adsorption capacity is exceptionally expressed as /nm2. The figure clearly demonstrates that the total ammonia adsorption on silica is determined by the silanol number, in a 1 1 relationship. This means that ammonia adsorption on silica involves an attachment (by a hydrogen bond) of 1 NH3 molecule to 1 silanol group. Such species produce an infrared band at 3419 cm1, assigned by Peri1 to the v3 valence vibration of ammonia. 

---