Let’s talk about some details that might sound small or even ridiculous, but (as you’ll see) they’re just the sorts of things that you have to worry about at the intersection of chemistry, biology, and physics. That makes it sound like I’m going to be going into something really high-tech here, but you be the judge – we’re going to veer off into beer cans in a minute or two.
Here’s a detail for you: how many vials will we need to put all those coronavirus vaccine doses in? What will they be made out of? Are there even that many vials of the right size in the world at the moment? The answer is, well, no there aren’t. Not yet. And the companies that manufacture them are getting orders that are beyond what they are currently able to produce. Here’s a good piece at Wired that goes into the details. In short, if someone waved a wand and made several huge vats of effective vaccine appear tomorrow, we wouldn’t have enough containers to get it rolled out to the general population.
The traditional glass companies (Corning, Schott, etc.) are ramping up as quickly as they can, but as the article notes, timing will be crucial. If we are rolling out a vaccine in the middle of next year, the vial supply should be in pretty good shape. But if we’re trying to do that by the end of this year, well, there might well be a gap. We are, of course, talking about glass vials – if you think back to all the shots you’ve been given in a doctor’s office over the years, the little vial that the syringe goes into is invariably made out of glass.
Why not plastic? That term covers a lot of different materials, of course, but each of them tend to have their own strengths and their own problems. Glass is in many ways a wonderful substance for containers in biology and chemistry, with some advantages – particularly on longer-term storage – that are hard to beat. A big one is that glass is basically impermeable to most gases. You probably don’t spend a lot of time thinking about that as a feature, but believe me, people in packaging sure do. For example, people my age and older will remember the days when carbonated drinks of all sorts were to be found almost entirely in glass bottles – vending machines, larger bottles in the stores, etc. Cans (at first tin-plated steel and then aluminum) moved up in popularity in the 1960s after the invention of the now-reviled pull tab, but I well recall the glass-bottle-dispensing machines.
Now it’s either metal cans or plastic bottles, and the rise of plastic depended on getting polyethylene glycol terephthlate (PET) worked out. Shorter-chain polymers of that sort were already big in the 1950s (Dacron polyester, Mylar, etc.). Resins with longer polymers are now used for water bottles, and even longer-chain ones for carbonated drinks. There’s more to it than that (small amounts of copolymer additives, drying the resin before processing, the heat treatment to control the degree of crystallization in the polymer), but all of these are optimized to produce a plastic that is clear, moldable, and does not let carbon dioxide pass through it. Otherwise the unopened bottled drink would go flat in a few days.
Unfortunately, that sort of PET will let oxygen through. Ever wondered why you don’t see many plastic wine or beer bottles? No one worries too much about oxygen in their bottle of cola, but oxidation will throw off the taste of wine and beer pretty noticeably. It’s especially problematic considering that wine is kept around for years at a time. You can get around this with another oxygen-impermeable polymer (ethylene vinyl alcohol is the standard). But EVOH is expensive to produce, and there are only a few suppliers in the world – what you see in actual use is a thin layer of EVOH sandwiched with cheaper polymers, which still makes the manufacturing harder than good ol’ PET.
This would be a good time to note that even metal cans are more complicated than they might appear (and see the epilogue below if you want even more). They’re pretty impermeable to oxygen and carbon dioxide, but there is that direct metal contact to think about. Cola drinks have a lot of phosphoric acid in them, just to pick one problem, and this is not a good match for a metal surface. Beverage and food cans invariably have a thin plastic layer (generally an epoxy resin) on their inner side to be a barrier. According to that link, beer is actually pretty benign and doesn’t even need the polymer barrier, which explains why beer cans were a thing many years before soda cans appeared (although World War II and its need for metals rather slowed that process down!) Some energy-drink formulations, on the other hand, still can’t be packaged in cans because they’re just too corrosive.
As with cola and wine, so with vaccines and monoclonal antibodies. All of these considerations apply: you want a material that’s gas-impermeable, because oxidation will degrade those complex biomolecules as surely as it will make a bottle of wine taste funny on storage. The cysteine amino acid residues are an obvious starting point for trouble, and there are others. You want as inert an interior surface as you can get, too. That means chemically inert, of course, but also physically inert, since it’s also well-established that some plastic surfaces can adsorb both small molecules and larger protein species in troublesome ways. Glass performs quite well in all these characteristics.
That link back at the top of the post has some interesting details about a company that’s trying to help fill things in, SiO2 of Alabama. They were making plastic milk jugs as those were becoming a thing back in the early 1960s and have been supplying the pharma industry for several years now with an unusual packaging alternative: plastic containers that have extremely thin silica layers inside them, applied by a plasma vapor deposition technique. Glass has all the desirable chemical properties noted so far, and plastics are unbreakable and can be made to more precise physical specifications, which makes it easier for high-speed manufacturing lines. So glass on the inside, plastic on the outside gives you some of the advantages of each substance, albeit at a price – but for high-value products like this it can make a lot of sense.
Epilogue, not relevant to vaccines: I mentioned above that metals are basically impermeable to oxygen and carbon dioxide, but there is a gas that gives them trouble: hydrogen. The permeability varies a great deal between metals (and alloys thereof), and also varies with pressure and temperature, as you’d expect. But it’s a real engineering problem, made worse by the fact that some metals (high tensile strength steel in particular) actually become brittle over time if hydrogen is allowed to permeate them: in practice, tanks and pipelines used for compressed hydrogen are lined with a polymer such as high-density polyethylene (HDPE) which is much more impermeable to the gas than the metal outer layer is. And there’s been a lot of work put into the idea of non-metallic pipelines made out of some sort of fiber-reinforced polymer entirely. Meanwhile, if you need extremely pure hydrogen, one way to get it is to let it diffuse through a membrane made of palladium alloy. Chemists are familiar with how easily palladium and platinum take up hydrogen gas, and such membranes have effectively infinite selectivity: no other gas can get through them at all!