I am working on a device that will spray a small amount of liquid (mist) and I need this thing to be work through our cold winters unattended. The device will have a small pressurized containing for the liquid (perhaps the size of a 12 oz soda can). I am looking for recommendations to keep the liquid from freezing during very low temps. I anticipate that the device will have a solar cell to charge a lithium battery. So, the solution would have to be a rather low power solution and it would also need to provide enough heat to the Arduino to keep it functional. It does not have to be worm, it just can't freeze.
You need some way of measuring temperature and some way of controlling a heater. A simple MOSFET is enough to control the heater. The heater itself can be resistance wire wrapped around the bottle or an immersion heater. Resistance wire is probably best because you can then wrap it around the solenoid and output nozzle to keep those warm too.
What is your power source?
You're going to need to seek the help of somebody who knows some thermodynamics. How cold will it be, and how much wattage is needed to produce the heat? You'll have to heat your spray can, the electronics, and the battery. You might need to size a rather large solar array and batteries for this. And do you have a lot of cloudy weather in winter? Solar cell ratings are usually given for the most optimistic conditions, i.e., high noon and a bright sunny day. If you have a spell of weather where your power supply can't keep up, your batteries might freeze. I don't know of any battery types that would survive that.
What is the full application, even if you keep the fluid liquid, in sub zero will freeze the instant it leaves the spray nozzle.
Being honest, knowing the solution he is spraying will be of great use. Knowing it's melting point could mean heating it is a wasteful task in the first place!
Knowing it would also be useful to gauge the amount of energy required to heat the solution if required as well.
I do not know if the OP will be spraying Sodium Chloride solution, but any ionic salt in solution can drastically change its melting point.
23% NaCl in water by weight (23g salt, 77g water) gives a melting point of -20 deg C...which very few places ever get to.
If the OP does not know the melting point (a freezer may be of use here?) and still wants a heated solution...
Id go Lead Acid batteries (longer lasting and less annoying to charge), a cheap ebay solar charge controller, a 12V immersion heater(low wattage ofc) and a semi-decent solar panel.
The more mass of solution in the first place the better. If the OP can bury the reservior....bonus points.