Eur. Phys. J. B 76, 197–201 (2010) DOI: 10.1140/epjb/e2010-00189-y Regular Article T HE EUROPEAN P HYSICAL JOURNAL B Structural properties of the range-II- and range-III order in amorphous-SiO 2 probed by electron paramagnetic resonance and Raman spectroscopy G. Vaccaro 1, a , G. Buscarino 1 , S. Agnello 1 , G. Messina 2 , M. Carpanese 2 , and F.M. Gelardi 1 1 Dipartimento di Scienze Fisiche ed Astronomiche, Universit` a di Palermo, Via Archirafi 36, 90123 Palermo, Italy 2 ENEA C.R. Frascati, Italy Received 19 January 2010 / Received in final form 2 April 2010 Published online 24 June 2010 – c  EDP Sciences, Societ`a Italiana di Fisica, Springer-Verlag 2010 Abstract. In the present work we report an experimental investigation by electron paramagnetic resonance spectroscopy on the hyperfine structure of the E ′ γ point defect, probing the local arrangement of the network (range-II order), and by Raman spectroscopy on the D1 and D2 lines, probing mean features of the network (range-III order). Our studies, performed on a-SiO2 samples thermally treated at 1000 ◦ C in air for different time durations, show that changes of the hyperfine structure and of the D1 and D2 lines occur in a correlated way. These results give strong evidence that the range-II and range-III order properties are intimately related to each other and that these properties are determined by the history of the material. 1 Introduction Structural modifications of silica induced by mechanical or thermal treatments have been studied from a long time and are still of actual importance due to the technological applications of the material [1–4]. Based on X-ray and neu- tron diffraction studies, four ranges of order have been de- fined in silica glasses [5,6]: range-I, the SiO 4 tetrahedron; range-II, the interconnections of two tetrahedra; range- III, larger features of the network topology (including ring structures) up to ∼2 nm in size; and range-IV, larger-scale density fluctuations [7]. In particular, the properties of the range-II and range-III order can be investigated by various experimental probes . For example, the point defects, in- duced in silica during the manufacturing process or after the exposure to particle or ionizing radiation, are sensi- tive to the nearest neighbor atoms and consequently can be used as probes of the range-II order [5]. Among the point defects in silica, the E ′ γ center is one of the most known and relevant to this aim [5,6]. This defect is an intrinsic paramagnetic center, located at the site of an oxygen vacancy, whose atomic structure is characterized by an unpaired electron highly localized in a sp 3 hybrid orbital of one of the silicon atoms: O≡Si • (where ≡ repre- sents the bonds with the three distinct oxygen atoms and • is the unpaired electron) [5,6,8–10]. The electron para- magnetic resonance (EPR) spectrum of the E ′ γ point de- fect is characterized by a slightly orthorhombic main reso- nance line and a correlated hyperfine structure consisting a e-mail: gvaccaro@fisica.unipa.it of a pair of lines split by ∼42 mT. This doublet originates from the interaction of the magnetic moment of the un- paired electron with that of the 29 Si nucleus (having spin I =1/2 and a natural abundance ∼4.7%) on which the electron is localized [9]. First-principle calculations have shown that the hyperfine splitting depends on the Si-O- Si and O-Si-O bond-angles and Si–O bond-lengths at the oxygen vacancy [11]. Furthermore, it was experimentally found that the hyperfine splitting monotonically increases on increasing the densification induced in silica samples by hydrostatic pressure [1]. In this latter work, the au- thors were able to conclude that densification results in a reduction in the mean O-Si-O tetrahedral bond angle and an increase in the mean Si–O bond length at the O≡Si • sites. All the characteristics discussed above make the hy- perfine splitting of the E ′ γ center an excellent experimental probe to investigate the range-II order properties of silica network [12,13]. At variance, an experimental investiga- tion of the range-III order properties is much more diffi- cult. Indeed, despite the fact that theoretical and simula- tive works have pointed out that from 3 to 10 membered rings may exist in silica, a direct spectroscopic identifi- cation has been possible only for the 3- and 4-membered rings [14]. In fact, it is widely accepted that two sharp Raman lines peaked at 495 and 605 cm -1 , known as D 1 and D 2 lines, are assigned to the symmetric oxygen breath- ing vibrations of 4- and 3-membered rings, respectively, embedded in the silica network [14,15]. It is worth noting that although their statistical population is low (≤1%), these rings play a fundamental role in many macroscopic