Pharma glass compositions – Part 3. Network intermediates

Hello everyone – welcome to Part 3 of an ongoing series that is taking a closer look at glass compositions used for parenteral packaging.? Part 1 considered “glass network formers”, a majority component that provides a structural “backbone” to glass.? In the case of pharma glass, there are only two network formers (SiO? and B?O?) that are relevant.? Part 2 dealt with “glass network modifiers” – these components alter the connectivity of the glass network, thereby changing properties such as viscosity that make glasses easier to manufacture, albeit potentially at the expense of other properties such as chemical resistance.

The subject of today’s post is “glass network intermediates”, meaning those components that fall somewhere in between acting consistently as a network modifier or a network former.? There are many compounds that can act as a network intermediate, but alumina (Al?O?) is the most relevant to modern pharma glass compositions in use today (see Footnote 1).? Alumina is interesting because it provides me with an example to introduce how we consider glass composition/structure relationships in a simplistic way.? As a first pass, we tend to think about the constituents of oxide glasses as existing in a purely ionic state (see Footnote 2).? For example, the idealized structure of fused quartz (a glass that is effectively 100% SiO?) would consist of a silicon cation (Si??) surrounded by four oxygen anions (O2?) to create a SiO? tetrahedron.? If you just consider the SiO? tetrahedron in isolation, you may notice that there is a charge imbalance – i.e., 4 positive unit charges from the silicon ion and 8 negative unit charges in total from the four oxygen ions. ?However, we discussed how these polyhedra link together at the corners to form a glass network in Part 1 of this series.? Each oxygen anion at the corner of a SiO? tetrahedron is therefore ideally being shared between two silicon cations.? One-half of each oxygen anion’s negative charge is being given to the central silicon cations, thereby preserving electrical charge balance (see Footnote 3).? ?The left-hand image in Figure 1 illustrates this concept by showing a SiO? tetrahedron that is charge neutral due to the sharing of oxygen ions with neighboring SiO? tetrahedra.

I mention all of this because Al?O? provides an interesting counter-example.? Just like silicon, let’s start by assuming that the aluminum cation (Al3?) is surrounded by four oxygen anions – the details of why this might be a reasonable assumption are beyond the scope of this post.? The aluminum cation contributes 3 positive unit charges and the oxygen anions contribute 4 negative unit charges to the AlO? tetrahedron – this creates an electrical charge imbalance of -1 (see Footnote 4).? How do we balance this residual charge?? Let’s assume for now that we are working within what’s called a “metaluminous” compositional space – i.e., a glass composition in which the concentration of network modifiers is equal to or greater than the concentration of Al?O?.? For simplicity, let’s also assume that the sodium cation (Na? – formally represented in the glass composition via the oxide component Na?O) is the only network modifier present within the glass.? Given this set of conditions, a single sodium cation having a +1 charge can balance the -1 charge associated with the AlO? tetrahedron (see the right-hand image of Figure 1).

The charge compensation of the AlO? tetrahedron by network modifiers has important implications for glass properties.? Looking back to Part 2 of this series, you’ll recall that network modifiers (sodium, potassium, calcium, etc.) typically reduce the connectivity of the glass network by disrupting links between network polyhedra (also called bridging oxygens or BO) to create non-bridging oxygens (NBO).? The bond strength of NBO is generally lower than BO, which impacts a number of glass properties, including but not limited to: viscosity, ionic conductivity, thermal expansion, and chemical durability.? However, notice in Figure 1(b) how the presence of sodium (a network modifier) does not change the connectivity of the glass network.? The AlO? tetrahedron is still connected to four neighboring network polyhedra by four BO despite the presence of a network modifier.? The introduction of Al?O? into compositions can therefore be used to balance the ratio of BO and NBO within the glass network.? This is why you’ll find Al?O? frequently used to varying degrees in pharma glass compositions.? It acts as a convenient compositional lever for tuning glass properties in a way that cannot be readily achieved through just the use of network modifiers and network intermediates.

Finally, let’s revisit one of our initial assumptions – all of the prior discussion was based on metaluminous glass compositions.? What happens if the concentration of Al?O? is greater than the network modifiers?? The structure of these so-called “peraluminous” glasses is still the subject of active research.? One early hypothesis was that AlO? tetrahedra unable to achieve charge compensation from network modifiers would instead convert to a 6-fold coordination geometry, otherwise known as octahedral coordination (see Footnote 5).? Octahedral coordination introduces two additional oxygen anions that contribute another -1 charge (1/2 x two O2? anions) needed to balance the +3 charge from the aluminum cation. ?This seemed reasonable at the time because aluminum is also known to exhibit octahedral coordination in numerous crystalline oxide compounds including but not limited to corundum, mica, and various clays.? However, structural studies enabled by advanced characterization tools have generally disproven this hypothesis – alternate charge compensation mechanisms are used in the peraluminous regime (see Footnote 6).? All of this discussion is ultimately academic when it comes to pharma glass compositions.? A review of technical data sheets will reveal that all commercially available pharma glasses are within the metaluminous compositional regime in which the network modifier content exceeds the Al?O? content.

Questions or comments? – please leave them below or feel free to contact me directly.

?

Footnotes

1.? I think this has been mentioned before, but as a reminder – we’re following the typical approach of discussing glass components in their oxide form (SiO?, Al?O?, etc.).? This is mostly out of convenience and in recognition of how we like to think about structural polyhedral units such as the SiO? tetrahedron.? Just recognize that these oxide components do not build up the structure of glass in a discrete manner like the assembly of LEGO? bricks.? Instead, the oxide components are dissolved together into a molten liquid that is quenched to the glassy state.

2.? Note that I specified oxide glasses – this was intentional.? As previously mentioned in Part 1 of this series, there are other inorganic non-oxide glasses in which the bonding is much more covalent in nature.? I should also emphasize that the bonding in oxide glasses is not purely ionic, particularly depending on the specific bond of interest.? For example, a silicon-oxygen bond will be more covalent than a sodium-oxygen bond.? It would therefore be more appropriate to describe the bonding within typical oxide glasses as ionocovalent.

3.? You may find reference to SiO?/? tetrahedra in the literature to reinforce the concept that four oxygen anions are shared with neighboring polyhedra.

4.? Similar to Footnote 2, you may encounter the terminology AlO?/? or even [AlO?/?]? to denote a polyhedral structural unit with a residual charge of -1.

5.? You might be wondering why an arrangement of six oxygen ions around an aluminum ion is called octahedral coordination – doesn’t octahedral refer to the number eight?? In this case, the descriptor is referring to the shape of the polyhedron created by the anions (oxygen) surrounding the central cation (aluminum).? Pull up a picture of an octahedron on your search engine of choice and count the number of vertices – you should find the answer is six.

6.? If you’re interested in learning more about this topic, I would recommend checking out a relatively recent article by Welch et al. that provides an excellent review of charge compensation mechanisms in aluminosilicate glasses.

·?????? Welch RS, Astle S, Youngman RE, Mauro JC (2022).? High-coordinated alumina and oxygen triclusters in modified aluminosilicate glasses. International Journal of Applied Glass Science, 13: 388-401.