I am employed at the ISIS Facility of the Rutherford Appleton Laboratory, where I am responsible for the neutron diffractometer GEM (GEneral Materials). I am responsible for experiments on GEM to study the structure of glasses, liquids and disordered crystals.
My personal research is into the structure of oxide glasses, such as germanate glasses, and chalcogenide glasses, such as arsenic sulphide glasses. As well as developing general techniques for studying glasses by neutron diffraction, I have particular interests in studies of bonding in glasses, and the variation of coordination numbers with composition.
I am an editor of the Society’s journal Physics and Chemistry of Glasses, and I am the chairman of the international organising committee of the International Conference on Borate Glasses, Crystals, and Melts.

Alex Hannon

ISIS Facility, Rutherford Appleton Laboratory, Chilton, Didcot, Oxon OX11 0QX.

The coordination of alkali cations, A+ (A=Li, Na, K, Rb or Cs), in a silicate glass, and the corresponding coordination number, nAO, are a subject of longstanding, fundamental interest from the glass structure point of view. Pure SiO2 glass forms a network in which all oxygen atoms participate in Si-O-Si bridges that connect pairs of corner-sharing SiO4 tetrahedra. It is well known that when alkali oxide, A2O, is added to a silicate it acts as a network modifier so that bridging oxygens (BOs) are replaced by non-bridging oxygens (NBOs). The exact way in which A+ cations bond to the oxygens is, however, not well established. Experimental measurements of the coordination numbers, nAO, have not yielded a highly consistent set of results, and in fact, as will be shown, it is a gross oversimplification to characterise the distribution of A-O bond lengths of varying strength in terms of a single number.
The structure of the full series of alkali disilicate crystals, A2Si2O5, is examined to provide a detailed description of the behaviour of the alkali coordination in silicate solids, which can be used to predict its likely behaviour in alkali silicate glasses, and to interpret experimental results for glasses. As the alkali size increases, the coordination number, nAO, is shown to increase, and furthermore the distribution of A-O bond lengths is shown to extend increasingly towards longer distances. A bond valence method is used to show that fundamentally there is no significant change in the distribution of A-NBO distances, but in contrast there is a steady lengthening of the A-BO bond lengths, with a corresponding growth in the A-BO coordination number, nA-BO. The distribution of A-O bond lengths can be characterised in terms of two coordination shells:
The first coordination shell involves short A-O bonds in a narrow range of distances, and is dominated by A-NBO bonds. This shell has a coordination number of about four, which shows no strong change with alkali size. Typically the number of these short A-NBO bonds is three, but sometimes there are four such bonds, whilst the number of short bonds from an alkali to a BO is one or less.
The second coordination shell involves longer, more widely varying A-O distances, and is dominated by A-BO bonds. This shell has a coordination number that grows steadily from zero to five as the alkali gets larger. The distances contributing to this shell arise mostly from A-BO bonds, but sometimes an alkali may have one NBO in its second coordination shell.
This structural description provides a basic terminology for the discussion of experimental results on the structure of alkali silicate glasses. Furthermore, in fundamental considerations of glass structure (for example, Sun’s original paper on glass formers and modifiers,(1) or modern constraint theory papers(2)) there is a tendency to present a table of coordination numbers for each cation as though this is sufficient to characterise the bonds formed by the cation. This may be sufficiently close to the truth to be useful for a glass former cation, such as Si4+. However, the structure of alkali silicate crystals shows that there is much greater variation in the strengths of the A-O bonds, and the emphasis on a single coordination number is a gross over-simplification.

  1. Sun, K. H., “Fundamental Condition of Glass Formation”, J. Am. Ceram. Soc., 1947, 30, 277-281.
  2. Rodrigues, B. P. & Wondraczek, L., “Cationic constraint effects in metaphosphate glasses”, J. Chem. Phys., 2014, 140, 214501.