Splitting of the CN stretch vibration

An explanation might be related to the formation of surface clusters of different sizes, e.g., Agn with n = 4,5,6...etc. In [3] we pointed out the influence of different cluster sizes on the stretch vibration, discussing the influence of an increasing weight of the Agn particles.

We considered the influence of the mass on the AgnC group using the normal-coordinate analysis. The results of these calculations indicated that an Ag mass increase from n = 1 to n = 4 only caused a shift of the CN" stretching of about 2 cm"1. The computed wavenumber shifts for n to n + 1 decreased drastically for n > 4. Therefore this simple explanation obviously failed.

Furthermore we discussed the possibility that the observed splitting of the CN" vibrations resulted from differences in the density of charge in respect to a variety of silver cyanide complexes. In an early paper Otto et al. [7] explained the trend

VCN~ < VAg(CN)l- < VAg(CN)\~ < VAg(CNy2 < VAg(CN) 2080 cm"1 2097 cm"1 2111cm"1 2143 cm"1 2167 cm"1

by changes in the bonding between silver and the ligands. In analogy to that paper we tentatively suggested the following explanation for the observed splitting of the SER signal: assuming that an increase of n in the Agn(CN) clusters effects a change in the density of charge, and moreover that this change influences the bonding between silver and the ligands, we suggested that with increasing mass of silver the average bond order between silver and the CN" ions via the 5 cr orbital increases. Thus the ligand stretch vibration should appear at higher frequencies, since the 5a orbital is a bonding orbital between silver and the ligands but antibonding with respect to the intraligand bonding [3],

The earlier attempt to explain the splitting did not take into account the results shown in Fig. 9, i.e., the jump of spectral structure with the potential from peaks around v = 2111 cm"1 (-700 mVscE) to v = 2143 cm"1 (0 mVSCE). This behavior is more reminiscent of an adsorption of the complex Ag(CN) 32~ (v = 2111 cm'1) changing with the potential to the complex Ag(CN)j (v= 2143 cm"1). That means we have to discuss the adsorption of Agad(CN)^-Jt complexes on clusters of different size (Agn -Agad(CN)'""*). The splitting of the CN" Raman signal must then be correlated with a change of the charge on the complexed Agad atom with the size of the cluster Agn to which the Agad atom is bound. The partial charge of the Agad atom could be estimated from the width of the splitting: For -700 mVscE 2080 - 2140 cm"1 For 0 mVscE 2110 - 2175 cm"1 by comparing the frequencies with the values for the complex ions in solution. This is carried out in the next section.

If one assumes a linear change of the frequency of the CN" strech vibration with the value of the partial charge per CN" ion, we obtain Fig. 10. The frequencies of AgCN and free CN" were excluded because the frequency of AgCN was measured in the solid state and that of CN lacks the influence of Ag. As indicated in Fig. 10, one gets a for the Agad atom per CN" ligand of 0.16 - 0.47 at the potential -700 mVSCE and 0.32 -0.68 at 0 mVscE- Then the variation of the average of the (positive) charge on the Agad atom is obtained by multiplying the 8+ value for -700 mVscE by the factor 3 and that for 0 mVscE by the factor 2. This gives Ag^8 - Ag^1 at -700 mVscE respectively

Fig. 10. Relation between the frequency of the CN" stretch vibration in the different Ag(CN) x complexes and the partial (positive) charge per CN" ligand; v^gCN and vcn were excluded.

Fig. 10. Relation between the frequency of the CN" stretch vibration in the different Ag(CN) x complexes and the partial (positive) charge per CN" ligand; v^gCN and vcn were excluded.

Ag°f - Ag 'ad36 at 0 mVscE depending on the size of the cluster Ag„ to which the Agad atom is bound. This variation of the positive charge on the Agad atom would be an explanation for the appearance of the observed splitting of the C=N stretch vibration.

0 0

Post a comment