Although it doesn't neatly fit into any category that the AT community has, I thought that people would at least find this project interesting, and I'm sure that some will have some valuable if indirect experience to contribute. Warning: this will be a text wall, but I promise I'll add pictures, links, and better formatting as time goes along!
So, essentially, my goal is to build a supercritical CO2 chamber. The home use of such a thing is typically for extracting THC, but that's not anything I have any interest in. It, of course, could be used to extract lots of things (and is done commercially in a number of applications). However, my intent is to use it as a reaction chamber, and I don't particularly care about the liquid or solid precipitants left over. I actually want to build this in order to treat wood, eventually if successful building a much larger chamber to treat wood for use in pool cues.
What is supercritical CO2, if you don't already know? Well, we are aware of the 3 phases of matter (solid, liquid, gas), however, there really is a fourth called supercritical fluid. For any matter that is hot enough and at a high enough pressure that both the liquid and gas phases exist, as you increase that temperature and pressure the density of the gas phase will go up. As you increase the temperature, however, the density of the liquid phase goes down. At some point for all matter, there will be a "critical point" wherein the pressure and temperature will result in a density of the gas phase equaling the density of the liquid phase, so the two phases become indistinguishable. It's a really wild effect. If you watch a video of it happening, you start with a distinct line between liquid and gas, and then it starts to get hazy as you approach the critical point. If you agitate the vessel, you'll get swirly clouds. As you increase beyond the critical point, there will no longer be any capacity for a distinct liquid or solid phase to exist, and at that point the chamber looks crystal clear. A supercritical fluid has properties of both phases. Being gas like it can easily permeate any material it is small enough to fit through, and being liquid like it readily forms solutions, with solubility increasing with pressure (and generally temperature as long as you are far enough past the critical temperature). What's great is that, because there are no distinct phases, there is no surface tension, so you aren't bound by capillary action. This means that anything which is small enough to make it through the pore structure of the wood to access the carbohydrate backbone will be able to reach those sites so long as it can be brought into solution with the CO2. It seems to be super-soluble, but for high polarity molecules, you can use a co-solvent such as ethanol or methanol.
For wood treatment, there are two basic applications of supercritical CO2
1. Drying of wood. There is research in drying waterlogged artifact wood and in drying green Eucalyptus wood as a potential commercial application. If it can be used to speed up drying, particularly of exotic species, without having a tendency of inducing anything more than surface checking, then this would be really useful. From what I've researched, it seems to do better than kiln drying although not as good as the fancy methods used for preservation of archaeologic species. I'm not sure, but intend to find out, if it is a reasonable way to avoid years of air drying exotic wood.
2. Esterification of wood. This is really interesting stuff. There is a good deal of research here and some commercial applications, generally using acetic anhydride to add an acetyl ester to the free hydroxyl groups of the carbohydrate backbones comprising the cell wall structure of the wood (typically lignin > cellulose >> hemicellulose). I found one paper that did this using acetic anhydride and supercritical CO2, but as far as I can tell almost all of these treatments of wood are done under normal atmosphere. There are a lot of possible organic esterifying agents out there, but I think I want to use phthalic anhydride and experiment with tung oil either separately or in addition. The great thing about phthalic anhydride is it's much easier to obtain, cheap, and is a solid at normal room temperature and pressure. It is reasonably benign, though I'd still want to use skin protection and ventilation + respiration.
Specs:
Operating pressure: ~1100-2000 PSI
Operating temperature: ~room temp - 130 C
Chamber volume (final design): 3 in diameter x 30in+
Basic Design:
Ultimately, the final chamber will probably be made from 3" schedule 80 316 SS pipe, and I'll try to get anything I can professionally welded. Obviously, at a minimum, I'll need a threaded opening with the full 3" diameter to access the wood. Flanges could also be used, but near as I can figure they tend to be inferior, and I don't want to have to worry about any seals being degraded by the CO2.
Now, I plan to use regular pipe fittings to create a control module that I ought to be able to re-use for the final design. On the inlet side, I'll have a ball valve connected to a liquid CO2 tank (typically pressurized to 5000 PSI), that will fit to a cross with a bimetal thermometer (will use a thermowell to protect the thermometer). I'll hook up another cross that will have a pressure gauge, a safety valve, and a needle valve for output (generally I'd just be depressurizing the chamber...If I needed to capture the contents, I'd want the valve to be draining to gravity).
As far as safety valve goes, I think I use a spring-control valve to essentially ensure the operating pressure doesn't exceed my maximum (e.g. 2000 PSI). I might want to be redundant and add an additional pop valve just in case the pressure builds up too rapidly for the control valve to vent or that the control valve fails. I guess it never hurts to have more layers of safety, although that would require additional fittings so I'm not committed to it. Since it is a closed chamber, the only way for the pressure to rise is by heating the fluid, so I'm not too personally concerned about runaway pressure. My planned method of heating really wouldn't enable much exceeding the max operating temperature even in the event of some kind of catastrophic failure. One relevant consideration though is that the CO2 precipitates on the relief valve and causes it to freeze, so I might be mindful of insulating/heating it.
For prototyping, Pretty easy to find a 2" schedule 80 316 SS nipple up to 12" in length. Apparently you can get a fitting made with a sapphire window to cap one end. It isn't necessary in theory, but in case something is going wrong, it may be valuable to visualize things, and it might be able to allow me to visualize how readily certain chemicals would go into solution with the supercritical CO2. It's an expense, but probably worth it. On the other end, I can get some reducing bushings to attach to the control module.
Heating.... I bought a programmable temperature controller so I can program exactly how the temperature is regulated. It should be very important to slowly increase the temperature/pressure and to slowly decrease the temperature (and therefore pressure) as to avoid damaging the wood. Since it is a closed chamber, as long as I have enough CO2 in there and minimal leaks, increasing the pressure is just a matter of increasing the temperature. I should easily be able to reach operating temperatures and pressures from my research, but I need a means of reasonably evenly heating the chamber. My plan is to submerse the vessel in water. I might use DI water to be extra safe with corrosion, but I think so long as I properly dry the chamber after use, that really isn't a concern, especially in using 316 SS which is probably overkill. I bought a cheap heating element and a cheap aquarium pump. I hooked them up at home and made a sous-vide (technically misnomer) egg to test the setup works in principle. It worked well, but I didn't bother to find tune the program, so the temperature control was very sub-optimal. I think it will work.
Many of you will have noticed a serious problem with my proposed method of heating. My operating temperature is up to 130C, and water at sea level boils at 100C... That means that, essentially, I'll have to build a pressure cooker to get the temperature up to 130C. I don't think that will be a problem, really. I could build a container out of an acrylic cylinder and drill holes for the control module and cords for the temp sensor, water pump, and heating element and seal them. The pressure will be much less than that of the CO2 chamber, so I think that's pretty doable. The acrylic would let me see through to the window of the reaction chamber. One advantage of doing this in water for the heating is I can turn off the water pump and tell easily if their are any leaks for anything that is in the water. Anyway, I'll probably start experimenting at temperatures sub 100C, so if this idea doesn't pan out I am not wedded to designing the whole thing around it.
One thing I'm not sure about is circulating the CO2 in the chamber. I don't know how important that would be, but I would imagine it is needed to make a difference. The most obvious way is agitating the whole chamber. Something like a paint can mixer makes sense, although at scale I don't know about agitating a 3ft tall paint can
. I am also leery of adding any kind of movement to something depending on threaded components to seal in pressure. The worst case would probably be just leaking, but I'm sure there's a better way. It seems that electric fans wouldn't really be designed to operate at 130C and 2000 PSI . And I don't imagine it would be a good idea to subject a battery to that kind of environment. I thought about something like a magnetic stir bar. 316 SS is only slightly affected by magnetic fields, but there would be a lot of distance for that field to cover to work the stir bar, so I am just unsure if that is viable. Curious if others have any thoughts here.
I plan to supply the CO2 via liquid CO2 canister. I think that's just the most straightforward. Some folks have done this with dry ice, which is definitely doable, I don't need to skimp on the budget here. Plus I can clean out the chamber with some acetone to get rid of any water then pull a vacuum on it before adding in some CO2. I don't think it would be absolutely necessary to remove moisture in the air and especially the oxygen, but it really can't hurt. Since the temperature across time is only going to go in one direction until the slow return to atmospheric conditions, I don't need to worry about adding any more CO2 into the chamber at any point. If anything, some will need to vent through the pressure control valve to keep the pressure from going to high.
For the reaction, I'm imagining that just adding some powdered phthalic anhydride into the canister with the wood is all I need to do. I could try to calculate how much would be needed to get the desired amount of esterification, but the desired amount is also as much as possible. It will still be something in the neighborhood of 20% based on what I've researched, but that has to do with the accessibility of the free hydroxyl groups in the wood. Well, if they are not accessible to the CO2 solution, they aren't going to be readily accessible to water, either, so the effect of only a 20% esterification on dimensional stability, etc., is still really profound. So I think I'll just dump what seems to be reasonable as an excess amount into the chamber and see what happens. My mind pictures that any unreacted excess will simply precipitate out as solid once the CO2 is vented. Whatever is on the surface of the wood could be washed off pretty easily, and any unreacted phthalic anhydride retained in the wood itself would still eventually react. There doesn't need to be heat or a catalyst, but raising the temperature definitely accelerates things. If not, I'm not really afraid of unreacted phthalic anhydride. I would take precautions before working with any exotic species anyway as many have sensitizing compounds, and in particular I value my lungs. A byproduct of phthalic acid if produced in some quantity also doesn't sound scary and is solid at normal atmospheric conditions, so it would probably also mostly precipitate out. What I don't know is what reactions might occur between things in the wood other than the lignin/cellulose/hemicellulose. Most of the papers involve a variety of washings of some form of sawdust or wood pulp explicitly to standardize things and look at the chemistry of the effects on those carbohydrates. There is some data (and commercial process) on treating unmodified wood samples of some species, but not a lot, and exotic hardwoods of interest have oils and pigments and such that I'm not really sure what will happen. The CO2 itself will also bring much of that into solution, so it will be interesting to see what effects that has. I don't think a huge amount would end up extracted changing the wood markedly, but who knows. The data on color and UV resistance is generally positive, but lab conditions are who knows what. Still, I am not generally afraid of whatever gets extracted or whatever reacts with the phthalic anhydride. There won't be a lot produced. Anything volatile will get vented into atmosphere, and I'll do this under vacuum hood or outdoors with my own respiratory protection. Otherwise, they don't seem to be highly reactive in a normal environment, and things will probably be pretty high molecular weight. One consideration is that phthalates have been implicated in chronic health concerns especially hormone levels. They are a ubiquitous plasticizer, especially with vinyl plastics, and studies have shown that we all have a high environmental exposure to them. So me producing a little more might not be exactly good, but it'd be a drop in the bucket. But I also don't think there's anything to worry about in reality. The things produced would be high molecular weight, and it seems the worrisome phthalates are those with 6 or fewer carbons. But I wouldn't pretend to be any sort of chemist, so maybe someone else has relevant expertise here.
Hopefully this post provides at least a little curiosity for someone out there. Would love to know your thoughts!
So, essentially, my goal is to build a supercritical CO2 chamber. The home use of such a thing is typically for extracting THC, but that's not anything I have any interest in. It, of course, could be used to extract lots of things (and is done commercially in a number of applications). However, my intent is to use it as a reaction chamber, and I don't particularly care about the liquid or solid precipitants left over. I actually want to build this in order to treat wood, eventually if successful building a much larger chamber to treat wood for use in pool cues.
What is supercritical CO2, if you don't already know? Well, we are aware of the 3 phases of matter (solid, liquid, gas), however, there really is a fourth called supercritical fluid. For any matter that is hot enough and at a high enough pressure that both the liquid and gas phases exist, as you increase that temperature and pressure the density of the gas phase will go up. As you increase the temperature, however, the density of the liquid phase goes down. At some point for all matter, there will be a "critical point" wherein the pressure and temperature will result in a density of the gas phase equaling the density of the liquid phase, so the two phases become indistinguishable. It's a really wild effect. If you watch a video of it happening, you start with a distinct line between liquid and gas, and then it starts to get hazy as you approach the critical point. If you agitate the vessel, you'll get swirly clouds. As you increase beyond the critical point, there will no longer be any capacity for a distinct liquid or solid phase to exist, and at that point the chamber looks crystal clear. A supercritical fluid has properties of both phases. Being gas like it can easily permeate any material it is small enough to fit through, and being liquid like it readily forms solutions, with solubility increasing with pressure (and generally temperature as long as you are far enough past the critical temperature). What's great is that, because there are no distinct phases, there is no surface tension, so you aren't bound by capillary action. This means that anything which is small enough to make it through the pore structure of the wood to access the carbohydrate backbone will be able to reach those sites so long as it can be brought into solution with the CO2. It seems to be super-soluble, but for high polarity molecules, you can use a co-solvent such as ethanol or methanol.
For wood treatment, there are two basic applications of supercritical CO2
1. Drying of wood. There is research in drying waterlogged artifact wood and in drying green Eucalyptus wood as a potential commercial application. If it can be used to speed up drying, particularly of exotic species, without having a tendency of inducing anything more than surface checking, then this would be really useful. From what I've researched, it seems to do better than kiln drying although not as good as the fancy methods used for preservation of archaeologic species. I'm not sure, but intend to find out, if it is a reasonable way to avoid years of air drying exotic wood.
2. Esterification of wood. This is really interesting stuff. There is a good deal of research here and some commercial applications, generally using acetic anhydride to add an acetyl ester to the free hydroxyl groups of the carbohydrate backbones comprising the cell wall structure of the wood (typically lignin > cellulose >> hemicellulose). I found one paper that did this using acetic anhydride and supercritical CO2, but as far as I can tell almost all of these treatments of wood are done under normal atmosphere. There are a lot of possible organic esterifying agents out there, but I think I want to use phthalic anhydride and experiment with tung oil either separately or in addition. The great thing about phthalic anhydride is it's much easier to obtain, cheap, and is a solid at normal room temperature and pressure. It is reasonably benign, though I'd still want to use skin protection and ventilation + respiration.
Specs:
Operating pressure: ~1100-2000 PSI
Operating temperature: ~room temp - 130 C
Chamber volume (final design): 3 in diameter x 30in+
Basic Design:
Ultimately, the final chamber will probably be made from 3" schedule 80 316 SS pipe, and I'll try to get anything I can professionally welded. Obviously, at a minimum, I'll need a threaded opening with the full 3" diameter to access the wood. Flanges could also be used, but near as I can figure they tend to be inferior, and I don't want to have to worry about any seals being degraded by the CO2.
Now, I plan to use regular pipe fittings to create a control module that I ought to be able to re-use for the final design. On the inlet side, I'll have a ball valve connected to a liquid CO2 tank (typically pressurized to 5000 PSI), that will fit to a cross with a bimetal thermometer (will use a thermowell to protect the thermometer). I'll hook up another cross that will have a pressure gauge, a safety valve, and a needle valve for output (generally I'd just be depressurizing the chamber...If I needed to capture the contents, I'd want the valve to be draining to gravity).
As far as safety valve goes, I think I use a spring-control valve to essentially ensure the operating pressure doesn't exceed my maximum (e.g. 2000 PSI). I might want to be redundant and add an additional pop valve just in case the pressure builds up too rapidly for the control valve to vent or that the control valve fails. I guess it never hurts to have more layers of safety, although that would require additional fittings so I'm not committed to it. Since it is a closed chamber, the only way for the pressure to rise is by heating the fluid, so I'm not too personally concerned about runaway pressure. My planned method of heating really wouldn't enable much exceeding the max operating temperature even in the event of some kind of catastrophic failure. One relevant consideration though is that the CO2 precipitates on the relief valve and causes it to freeze, so I might be mindful of insulating/heating it.
For prototyping, Pretty easy to find a 2" schedule 80 316 SS nipple up to 12" in length. Apparently you can get a fitting made with a sapphire window to cap one end. It isn't necessary in theory, but in case something is going wrong, it may be valuable to visualize things, and it might be able to allow me to visualize how readily certain chemicals would go into solution with the supercritical CO2. It's an expense, but probably worth it. On the other end, I can get some reducing bushings to attach to the control module.
Heating.... I bought a programmable temperature controller so I can program exactly how the temperature is regulated. It should be very important to slowly increase the temperature/pressure and to slowly decrease the temperature (and therefore pressure) as to avoid damaging the wood. Since it is a closed chamber, as long as I have enough CO2 in there and minimal leaks, increasing the pressure is just a matter of increasing the temperature. I should easily be able to reach operating temperatures and pressures from my research, but I need a means of reasonably evenly heating the chamber. My plan is to submerse the vessel in water. I might use DI water to be extra safe with corrosion, but I think so long as I properly dry the chamber after use, that really isn't a concern, especially in using 316 SS which is probably overkill. I bought a cheap heating element and a cheap aquarium pump. I hooked them up at home and made a sous-vide (technically misnomer) egg to test the setup works in principle. It worked well, but I didn't bother to find tune the program, so the temperature control was very sub-optimal. I think it will work.
Many of you will have noticed a serious problem with my proposed method of heating. My operating temperature is up to 130C, and water at sea level boils at 100C... That means that, essentially, I'll have to build a pressure cooker to get the temperature up to 130C. I don't think that will be a problem, really. I could build a container out of an acrylic cylinder and drill holes for the control module and cords for the temp sensor, water pump, and heating element and seal them. The pressure will be much less than that of the CO2 chamber, so I think that's pretty doable. The acrylic would let me see through to the window of the reaction chamber. One advantage of doing this in water for the heating is I can turn off the water pump and tell easily if their are any leaks for anything that is in the water. Anyway, I'll probably start experimenting at temperatures sub 100C, so if this idea doesn't pan out I am not wedded to designing the whole thing around it.
One thing I'm not sure about is circulating the CO2 in the chamber. I don't know how important that would be, but I would imagine it is needed to make a difference. The most obvious way is agitating the whole chamber. Something like a paint can mixer makes sense, although at scale I don't know about agitating a 3ft tall paint can
. I am also leery of adding any kind of movement to something depending on threaded components to seal in pressure. The worst case would probably be just leaking, but I'm sure there's a better way. It seems that electric fans wouldn't really be designed to operate at 130C and 2000 PSI . And I don't imagine it would be a good idea to subject a battery to that kind of environment. I thought about something like a magnetic stir bar. 316 SS is only slightly affected by magnetic fields, but there would be a lot of distance for that field to cover to work the stir bar, so I am just unsure if that is viable. Curious if others have any thoughts here.I plan to supply the CO2 via liquid CO2 canister. I think that's just the most straightforward. Some folks have done this with dry ice, which is definitely doable, I don't need to skimp on the budget here. Plus I can clean out the chamber with some acetone to get rid of any water then pull a vacuum on it before adding in some CO2. I don't think it would be absolutely necessary to remove moisture in the air and especially the oxygen, but it really can't hurt. Since the temperature across time is only going to go in one direction until the slow return to atmospheric conditions, I don't need to worry about adding any more CO2 into the chamber at any point. If anything, some will need to vent through the pressure control valve to keep the pressure from going to high.
For the reaction, I'm imagining that just adding some powdered phthalic anhydride into the canister with the wood is all I need to do. I could try to calculate how much would be needed to get the desired amount of esterification, but the desired amount is also as much as possible. It will still be something in the neighborhood of 20% based on what I've researched, but that has to do with the accessibility of the free hydroxyl groups in the wood. Well, if they are not accessible to the CO2 solution, they aren't going to be readily accessible to water, either, so the effect of only a 20% esterification on dimensional stability, etc., is still really profound. So I think I'll just dump what seems to be reasonable as an excess amount into the chamber and see what happens. My mind pictures that any unreacted excess will simply precipitate out as solid once the CO2 is vented. Whatever is on the surface of the wood could be washed off pretty easily, and any unreacted phthalic anhydride retained in the wood itself would still eventually react. There doesn't need to be heat or a catalyst, but raising the temperature definitely accelerates things. If not, I'm not really afraid of unreacted phthalic anhydride. I would take precautions before working with any exotic species anyway as many have sensitizing compounds, and in particular I value my lungs. A byproduct of phthalic acid if produced in some quantity also doesn't sound scary and is solid at normal atmospheric conditions, so it would probably also mostly precipitate out. What I don't know is what reactions might occur between things in the wood other than the lignin/cellulose/hemicellulose. Most of the papers involve a variety of washings of some form of sawdust or wood pulp explicitly to standardize things and look at the chemistry of the effects on those carbohydrates. There is some data (and commercial process) on treating unmodified wood samples of some species, but not a lot, and exotic hardwoods of interest have oils and pigments and such that I'm not really sure what will happen. The CO2 itself will also bring much of that into solution, so it will be interesting to see what effects that has. I don't think a huge amount would end up extracted changing the wood markedly, but who knows. The data on color and UV resistance is generally positive, but lab conditions are who knows what. Still, I am not generally afraid of whatever gets extracted or whatever reacts with the phthalic anhydride. There won't be a lot produced. Anything volatile will get vented into atmosphere, and I'll do this under vacuum hood or outdoors with my own respiratory protection. Otherwise, they don't seem to be highly reactive in a normal environment, and things will probably be pretty high molecular weight. One consideration is that phthalates have been implicated in chronic health concerns especially hormone levels. They are a ubiquitous plasticizer, especially with vinyl plastics, and studies have shown that we all have a high environmental exposure to them. So me producing a little more might not be exactly good, but it'd be a drop in the bucket. But I also don't think there's anything to worry about in reality. The things produced would be high molecular weight, and it seems the worrisome phthalates are those with 6 or fewer carbons. But I wouldn't pretend to be any sort of chemist, so maybe someone else has relevant expertise here.
Hopefully this post provides at least a little curiosity for someone out there. Would love to know your thoughts!
Facebook
Twitter