Hydrocarbons as a Solvent, Chapter 1: Solvent Chemistry

Hydrocarbons as a Solvent, Chapter 1: Solvent Chemistry

February 05, 2021


Continuing Education

By Boris Kogon, Bizzyee

There are three primary solvents that are commonly used in cannabis and hemp extraction: ethanol (EtOH), carbon dioxide (CO2), and hydrocarbons including butane and propane (C4H10 and C3H8). Sure, there are a few others which we’ll touch on briefly, but those are mostly used for postprocessing. In this first of three articles, we’ll discuss the merits of primary extraction solvents used to separate desired phytochemicals from plant biomass and concentrate them. In the two articles that follow, we’ll take a deeper dive into the fundamentals of hydrocarbon extraction and the finishing techniques to create products like shatter, wax, and live resin. 

An Introduction to Alternative Solvents



Ethanol is a fabulous substance. We use it as a residue-free cleaning solution, as a primary extraction solvent, and as a mind-altering entheogen. It dissolves polar and nonpolar compounds and makes them miscible with each other. In fact, it’s so effective that it works too well, dissolving undesirable and target compounds alike: chlorophyll, sugars, and lipids along with the tetrahydrocannabinol (THC), cannabidiol (CBD), and terpenes we are targeting. The answer is to run it cold, in the range of -20 to -40°C, where chlorophyll and lipids become largely insoluble. The result will be a post-winterized crude oil once the ethanol is recovered.

Because ethanol is a liquid at room temperature, it does not require a pressure vessel like the “compressed gas” solvents. But it is flammable and requires a C1D2 environment for its safe handling and use. It’s also subject to standard Maximum Allowable Quantities (MAQ) depending on building occupancy. In a Factory classification (F) this will be a 120-gallon limit, with a doubler to 240 gallons with fire sprinklers.

This lower classification and low volatility make ethanol a useful solvent for large-scale hemp and cannabis processing, allowing us to use continuous or batch-continuous atmospheric processes like centrifuges and screw presses, and Type 6 licensing instead of Type 7 in California. However, the boiling point for ethanol is quite high, roughly 78°C at 1 atm. This is comparable in boiling point to the lightest terpenes, so the recovery of the ethanol is denaturing and removing the very terpenes that give cannabis oil its flavor and smell. Consequently, ethanol extraction is great for bulk processing for CBD and THC, but suboptimal for making fine products that retain their original smell and flavor. Instead, it’s an ideal solvent for making “crude oil”, a full-spectrum extract that’s an input for distillation and other postprocessing methods.

Carbon Dioxide

Next up: CO2, or supercritical CO2 as it’s often called. CO2 is a compressed gas at room temperature at around 800 psi (55 bar). It’s a subcritical liquid at this temperature and

pressure, and in this cold liquid state, it’s only a modest solvent. If it’s cooled to -45°C and 100 psi (~7 bar) vapor pressure, it will become very selective and dissolve only terpenes and little else.

This is a great technique for extracting and concentrating pure terpenes from cannabis and hemp biomass before a secondary solvent is used to extract cannabinoids.

Supercritical CO2 is exactly that: CO2 in the state that exists above its critical

point, at 31.0°C and 1070 psi (~74 bar). The supercritical state of matter is a region in the pressure/temperature triple-point diagram where the distinction between a liquid and a vapor disappears. In other words, there are no boiling or condensing phase-transitions above the critical temperature and pressure (critical point). 

Compounds in this state exhibit properties of a liquid and vapor simultaneously, for instance, the space-filling property of a gas and the dissolving/entraining properties of a liquid. Carbon dioxide in the supercritical state is much more aggressive than the subcritical liquid. Supercritical CO2 will dissolve THC and CBD, albeit more slowly than ethanol. A typical CO2 extraction run may last from 4 to 12 hours to get everything out of the biomass, while -40°C ethanol only needs to be exposed to the biomass for a few minutes. Likewise, the higher temperatures and higher pressures of supercritical CO2 extraction seem to denature delicate terpene molecules. Consequently, supercritical CO2 oil often has an identifiable burnt taste that some people find unappealing. This has made it fall out of favor in the last few years for vape-cartridges, being replaced by pure THC distillate flavored with cannabis-derived, hemp-derived, or other plant-derived terpenes.

Additionally, CO2 is under a lot of pressure. Unlike ethanol, which wants to be a liquid at room temperature, CO2 is physically incapable of being liquid at atmospheric pressure; it must be a solid or a gas. CO2 in the open atmosphere is well below its triple-point pressure of 5.111 atm, or roughly 75 psi. This means that we can only observe it as a frozen solid (dry ice), or a vapor, without pressurizing it. Supercritical CO2 extractions run between 1500 (~103 bar) to 5000 psi (~345 bar), and subcritical CO2 runs are at 400 (28 bar) to 1000 psi (~69 bar) as well. So, even though CO2 isn’t flammable, an often-cited selling point for equipment, it is still dangerous due to its own high pressure. A pressure level of 2000 psi (~138 bar) can cause a lot of damage if there is a mechanical failure, even without ignition.

For these reasons: poor taste, slow extraction runs, and high-pressures, CO2 systems have fallen out of favor over the last five years as other extraction methods have become more popular. The “CO2-taste” makes the oil best suited for crude, but the expensive equipment and slow run-times make it uncompetitive with ethanol for bulk processing.


An Introduction to Hydrocarbons

This brings us to the hydrocarbons (HCs), or as I call them: “The Three Kings of Cannabis Solvation”. First off, don’t let haters tell you that hydrocarbon solvents aren’t natural or organic. They are, in fact, the very definition of “organic”, containing nothing but carbon and hydrogen. They are also as natural as can be, being formed from fossilized living organisms. There are many kinds of hydrocarbons, but for this discussion, we are interested in butane (C4H10) and propane (C3H8). There are other commonly used HCs, namely pentane, hexane, and heptane, but these are all liquids at room temperature, and are mostly used for postprocessing refinement and crystallization.

Butane and propane are considered compressed gases at room temperature, forming a liquid at 20°C with a vapor pressure of about 20 psig for n-butane and roughly 125 psig for propane. Both pressures are an order of magnitude or two below the working pressures of CO2 extraction. Both solvents are also extremely flammable and explosive when volatilized and mixed in the air with oxygen. For this reason, all hydrocarbon extraction must be done with expertly engineered and certified equipment, in a CID1 certified booth or explosion-proof environment. Please note, in this context “explosion-proof” does not mean a concrete bunker that can contain an explosion. Rather, it involves controlling the gasses and keeping them below their Lower Explosive Limits (LEL) with ventilation and gas-testing sensors, as well as using electrical equipment that cannot start a fire.

Butane and propane are both alkanes, meaning that their carbons share only single-bonds (no double-bonds), and all their free valences are saturated with hydrogen atoms. Butane can form two different isomers (molecules that share the same chemical formula but have different structures): the n-isomer, which looks like a straight line, and the isobutane isomer which is bonded in a triangular/tetrahedral shape. Each has 4 carbons and 10 hydrogen atoms, but they are arranged in different shapes. This turns out to have quite an effect on their respective performance as cannabis solvents.

Originally butane hash oil (BHO) extract was made using cans of compressed lighter fluid, in a process called “open blasting”. It turns out that these cans contain a mixture of n-butane with isobutane and propane as propellants. As the industry matured, and open-blasting was replaced with closed-loop systems (where the solvent is recovered instead of being released into the air), practitioners began buying pure forms of these compressed gasses from specialty suppliers, preferably with instrument-grade purity of 99.9% to minimize contaminants in final products. Practitioners began to make mixtures of these three solvents to try to maximize the positive characteristics of each. 

In a nutshell, their characteristics are as follows:

n-Butane is the most noble of solvents and the most unforgiving. It is the most aggressive of the three and needs to spend the least time in contact with the biomass, oftentimes just being washed over the material column briefly. It’s so aggressive that it only needs minutes of contact, instead of the hours it takes with CO2. It is also so aggressive that at room temperature, it will extract chlorophyll and other pigments producing dark-colored oil. To resolve this we run it cold, and I mean very cold. The preferred temperature for running n-butane is -60°C for the solvent, something only possible using dry ice or an expensive low-temperature chiller. At low temperatures, n-butane will not pick up chlorophyll and it will be more selective against pigments and waxes. n-Butane also has the lowest vapor pressure of the three. In fact, when it is this cold, it’s below 0 psig or 1 atm! It’s no longer a compressed gas at all, but a sub-cooled liquid with no vapor-pressure of its own. For this reason, we often use regulated nitrogen (N2) as a propellant to push solvent from vessel to vessel.

Propane is the most forgiving of solvents, but also very finicky. Propane is under a fair amount of pressure, 125 psig at room temperature, and 25-50 psig even at colder operating temperatures. Propane is more selective in what it dissolves than n-butane. It produces a lighter color of extract, even when used at room temperature. It is also less effective at dissolving our target cannabinoids, THC and CBD, so it will often yield a more terpene-rich fraction than

n-butane. It’s sometimes said that propane produces lower yields than butane, but that is just a rookie talking. All solvents yield the same amount when run correctly by a skilled technician.

The yield is in the starting material, not the solvent used to get it out. Yes, propane may yield lower for the same amount of solvent run, but the answer then is to run more solvent (or run it warmer). The more solvent you run over your biomass, the more yield you get out with it. Likewise, the warmer the solvent, the more it will yield per unit solvent mass. But it may also yield more of the impurities you want to avoid like pigments and plant-waxes.

For these reasons, it requires an experienced extraction technician to get the desired end-product with full yield from the plant without sacrificing quality. In general, there is a matrix between yield and quality, with butane on the left and propane on the right of the horizontal axis, and temperature on the vertical. The warmer your solvent is run, the higher the yield with less solvent, but the darker your product will be. 

Likewise, the more butane used in your solvent mix, the higher the yield, but it must be run colder to maintain light color. The more propane in your mix, the lighter (and orange) your extract will be, but at the expense of running more solvent to get a full yield. In general, propane must be run warmer than butane, in the range of -10 to -20°C. Colder than this and you must run lots of it, and you may even drop cannabinoids out of solution when dewaxing. In the cannabis industry, the lightness of color is taken as a proxy for the “quality” of the extracts.

As mentioned, practitioners often mix these solvents to get the best of both worlds: the lighter color of propane with the full extraction of butane. There is another parameter at play: stability. This will come back again in the discussion of finishes below. Pure n-butane makes a product called shatter, which is clear like glass and stays that way (hence “stable”). Pure propane extract tends to spontaneously crystallize over time, turning the shatter cloudy at first, and eventually separating into a mixture of liquid terpenes and fine crystals called sugar-wax.

Practitioners can use one solvent or the other to make different finished products, but a problem arises when using a mixture of them together. A solvent mix that starts out as 50/50 of each, or 70/30 butane-rich, is very popular, but over time this mix will change. Propane, having a significantly higher vapor-pressure, will be recovered first, leaving mostly butane behind in the collection pot at the end of a run. Practitioners often like to “pour” their extract out, removing some of this liquid butane from the solvent mix. The result is that the mixture is constantly changing every run, becoming richer in propane and leaner in butane, resulting in a slow change of the extracts from shattery to sugary even with the same starting material.

Isobutane is the perfect compromise. As mentioned before, isobutane has the same chemical formula as n-butane, but in a different shape. It is a triangular star shape, with a central carbon and three others as methyl groups bound to it. There is a free hydrogen bound to its fourth valence, and that proton pushes the three peripheral carbons away from it to form a shallow tetrahedron. The important feature is that from any one angle, the molecule only has three carbons facing out. In other words, from any angle, isobutane appears like propane to another molecule that is physically interacting with it (the fourth carbon being occluded by the other three). This means that isobutane extracts a lot like propane while having the lower pressures we enjoy with n-butane (roughly 35 psi at 20°C). Additionally, if one uses pure isobutane instead of a mixture of compounds, there are no ratios to change over time because of different vapor-pressures and resulting recovery differentials. This means that its extraction characteristics don’t change over time.

© 2020,Boris Kogon

Leave a comment