Sounds more like a design issue, as opposed to a material issue. From the information you provided.
Let me see if I understand this correctly...
You are looking for a 2ml container, as opposed to a 2ml cup with wadding. This container having a mechanical or electrical
device which physically moves a drop onto the vaporizer coil/pad.
Here are the possible issues, in the most simple outline, as I understand it...
For liquid to exit the container, air must be able to get inside, unless the container can collapse. (Such as a soft plastic container would allow, like a tooth-paste tube.)
For air to enter, which would not be that hard to manage, you would need a valve of some kind. (Having any hole, other than the one which the liquid is ejected, would cause all the fluid to drain.)
Having a valve to open/close while the fluid ejects, is a lot more complex than it sounds. By design, it is simple. By long-term manufacturing stability through daily use, it is difficult. (Economically, something of that precise of a design, which had near zero failure rate, would be a little costly. This would be fine, if the cartridge were a permanent structure which could be cleaned and thus, used for the life of the device.)
On a plastic level... The government does not like to use plastic containers for more than a year. After a year, all sorts of break-down related to chemical degradation can occur. CO2 interaction and UV and bacterial interactions will rapidly degrade any food-safe plastics.
Glass would be the most ideal alternative, but has obvious issues related to the mechanics of the design.
Metals are possible, but FDA does not like metal food storage. The one year rule might be extended to a three year rule, with the use of expensive processing of the metals with exotic food-safe electroplating.
A mixture of a glass chamber "U" shape, with a plastic dispenser tip, may be the most viable solution. The tips could be disposed after time, while the glass would severely reduce plastic contact exposure. (You could get less exposure if you fastened a glass disk/washer to the plastic side, on the inner part of the tip.)
For the mechanical ejection device, a simple rod or pole would suffice as a liquid plunger, which could also control the air-intake valve. The valve is still needed, so the liquid does not drip out freely. (The valve may be bypassed if a micro-filter insert was used. Air could pass freely into the cartridge, but liquid could not freely drip out, as long as nothing touched the filter. Like touching the bottom side of many umbrellas will pull all the water through the surface.)
Operation of the manual rod or pole plunger, would most-likely have to be done from the outside. This can be done with a "U" shaped plunger. One side would be shorter, while the longer side would extend out past the cap, and have a physical connection to something which can push and pull it. (This could be attached to an air-plunger, and also to the coil activation switch. Using the draw-air to operate the small motion required to eject a 0.02ml squirt onto the coil.)
Mechanical parts are usually not that fun to have, or deal with. But in this case, with minimal parts, it may be an exception. The cost for the long-use pure liquid cartridge, which would reduce break-down refilling, and use-life... might be more than worth it. (I imagine our land-fills would appreciate having a lot less plastic and nicotine in them.)
Just so you are aware... It costs around $60,000.00 to have an iso-certified single ejection, bubbler injection molding mold, designed for a 150-Ton press, manufactured in most places. That is a rather simple design, with only two prototype molds made, and does not include the remanufacture of a new mold, when the mold has become unusable after a few million runs.
However, that mold would easily hold a 8 piece runner, without the need for a hot-runner. Since the parts are internal, the only important issues would be with the plunger hole and the connection ring. (If this was to use the "U" glass container with only the plastic tip and plunger.)
The glass, I am sure could be easily manufactured, if it did not already exist. (I am sure they do, as mini-test-tubes.)
The plunger could be a simple steel rod, that was coated and bent into a "U" or "J" shape.
The plunger activation contact could be a simple snap-on or heat-staked scrap of flat plastic, which could be on the same runner which made the top.
I imagine a wick feeding directly from the tip to the coil-mesh would still be needed, but a little modding to an existing coil-setup would suffice.
I say it is possible. Within a reasonable price.
"Dymotek" plastics. Previously they were called "Truebro injection molding". We made microscopic heart valve gears... so anything is possible there. (All ISO certified. I was one of the leads, and testing technicians.)
