I think the measurements can be understood as follows:

Draw two concentric circles and hash the resulting regions differently as representing the cross section a spherical coated nano glass with a plane going through its centre. Number the differently-hashed regions in the inward direction by 1, 2 and 3. The outside region (region number 1) is occupied by air (or vacuum) for whose refractive index one has n1 ≈ 1. The second region is occupied by Silica (SiO2) and the third region, the interior region, by glass. Denoting the refractive indices of regions 2 and 3 by respectively n2 and n3, evidently n2, n3 > n1 [1]. The light shone on a coated nano particle will be partly totally reflected. Of the amount entering region 2, depending on the thickness of the coating, by Snell's law a substantial part, if not the total amount, can travel solely through region 2: considering n1/n2 = sin(θ2)/sin(θ1), the relationship n1 < n2 implies that θ1 > θ2. So, while one expects the absorption edge to be determined by the material with the smallest optical band gap (the absorption edges 330 nm and 300 nm correspond to optical band gaps of respectively 3.76 eV and 4.13 eV [1,2]), depending on the thickness of the silica coating, the inner glass core can be largely, if not totally, irrelevant to the transmission of the coated nano glass.

It would be interesting to see similar experiments as reported by Alam for different coating thicknesses. This is also interesting (that is aside from the influence of the coating thickness on the optical paths) because optical bang gap proves strongly to depend on the film thickness, this at least in the case of Ge [3] (note that the bandgap of SiO2 in the bulk is around 9 eV).

_____

[1] At the absorption edge, the refractive index is equal to the square root of the relative dielectric function of the material under consideration just below the absorption edge, which I denote by ε'. With ε' in a dielectric scaling like the inverse of the energy gap, it is expected that of two dielectrics the one with the largest (smallest) gap has the smallest (largest) refractive index. In the case at hand where the optical gap of the glass core (4.13 eV) is larger than that of the coating (3.76 eV), on this account one expects that n3 < n2, resulting in n1 < n3 < n2.

[2] Nandanwar et al.

http://www.journalrepository.org/media/journals/ACSJ_16/2014/Sep/Nandanwar512014ACSj10875_1.pdf

report optical gap of 3.54 eV for their SiO2 nanoparticles.

[3] ESM Goh et al., J. Appl. Phys. 107, 024305 (2010).

Article Thickness effect on the band gap and optical properties of g...

Following..

I would recommend you to check with fused silica substrate instead off glass by maintaining same particle size and thickness of the samples. Since the fused silica absorption is below ~160 nm and ~1400 nm as shown in the graph attached.

I understanding is changing the peak shift or band edge shift would be due to the interactions, changes in the particles size (increasing in this case), and changes in the thickness (increasing the thickness).