Extraction Considerations and Advice from a Leading Researcher

Published on: 
Cannabis Science and Technology, January/February 2019, Volume 2, Issue 1

Columns | <b>Cannabis Voices</b>

Dr. Markus Roggen discusses his experiences in the industry and what extraction methods he has used.

There are many factors to consider when starting an extraction method in cannabis-from flower supply and which method to use to particle size and flower to solvent ratios, among others. With all of these factors to consider, it might be hard to find a starting point. We recently spoke to leading extraction researcher Dr. Markus Roggen, Founder of Complex Biotech Discovery Ventures, about his experiences in the industry and what extraction methods he has used. Dr. Roggen also offers his insight into what to consider first in an extraction as well as his next research projects and plans for the future.

How did you get involved with cannabis extraction?

Dr. Markus Roggen: Pure serendipity! In 2014, I finished up my postdoc in physical organic chemistry at The Scripps Research Institute in San Diego and was looking for what to do next when my triathlon coach introduced me to a group that needed a chemist to run their analytical laboratory for cannabis. Initially, I was hesitant to leave a classical chemical career for the still young cannabis industry. Although, I was intrigued by the research and discovery potential in the field, so I did move into a totally unknown industry and culture (for me). While I worked as a laboratory director, on the side, I was mentoring young startups at Canopy Boulder and San Diego. At one of their networking events, I was explaining the scientific fundamentals of CO2 extraction (you know as you do over a good drink). The CEO of OutCo, a fully vertically integrated medical marijuana producer in San Diego, happened to be listening. He offered me the chance to join his team to put theory into practice. So, I did. I started leading the extraction and production processes in late 2016.

Do you prefer one method of extraction over another?

Roggen: No, my focus is on the required output! If the local regulations only allow ethanol extraction, then this is the best solvent to use. If the concentrate product is focused on solventless connoisseurs, rosin press or ice water hash might be best. But if I consider control over the extract composition paramount, then I highly recommend supercritical CO2 extraction (supercritical fluid extraction [SFE]) equipment. Due to the solvent properties of CO2, one can modulate which compounds are selectively enriched and how the extract properties are set. That allows one to, for example, separate terpene fractions from cannabinoid oils, or even enrich one cannabinoid over the other.

What are the major things to consider before starting an extraction?

Roggen: How much time do you have? The most obvious things to consider are flower supply and extraction method, but there are many, many other factors that influence the extraction outcome. For example, my research has shown that the particle size distribution impacts extraction precision and efficiency in SFE. For ethanol extraction, it’s the ratio of flower to solvent that has an effect on efficiency-and it’s opposite what one might think: More solvent extracts less THC! I could go on and on, so I will just name a few others: cannabinoid concentration, terpene concentration, water content, particle size, temperature, pressure, time, production costs, production volume, and most important product needs.

What is the biggest obstacle to overcome in cannabis extraction?

Roggen: Overcoming fixed mindsets. The debate over which solvent is best feels like a religious war. Although, there are enough deities for everyone, and those represent the regulatory, economical, and customer powers of the market. There likely is space for every technology, but there is no space for stubborn dogmatists.

Is there a common extraction problem that all users face? If so, how do you overcome it?

Roggen: Efficiency (or lack thereof). This simply is yield over costs. Costs include everything from plants to facility or instrumentation and operation. The best way to make your extraction process more efficient is through research, development, and calculations.

For example, I may know from controlled experiments that there are certain extraction rate curves on my particular SFE instrument. Based on these, I can develop a unique product line and optimize extraction for its specifications.  But, if I now want to change the starting material, the instrument, or the outcome, I need to start anew with different experimentation. This is because, right now, there are not enough fundamentals known in the cannabis industry.

In a recent article (1), you discussed how fallacies and ignorance of extraction mispresent the cannabis flower. Can you tell us more about that viewpoint? Are there are any other trends in extraction that you think are doing a disservice to the plant?

Roggen: The point I wanted to make was that the molecular makeup of cannabis flower and extract oils are very different. A cannabis flower has a plethora of cannabinoids, terpenes, flavonoids, carotenes, and others to offer the consumer, while cannabis oils are often refined to a purified single cannabinoid.


For example, vape carts mostly contain clear, off-yellow oil that has THC concentrations of around 90 wt.%. To add flavor, the producers introduce terpene mixtures that were blended to imitate terpene profiles of the cannabis flower. I am not saying that those highly refined products are bad in general, but I refuse to accept that both flower and “clear” oils are interchangeable.

There is an analogy that fits quite well, actually, I see cannabis flower like a red wine and “clear” oils more like a flavored vodka. Each product has its place and validity, but they do not substitute for one another across settings.

In another article (2), you addressed the issues surrounding “winterization” and a call to action for people to think about the reasons behind the extraction steps they are using. Do you think there is a need for more science-based education for newcomers to extraction? Are there better methods than winterization or precipitation to achieve a similar extract?

Roggen: Yes, I am convinced that more scientific education is needed in the industry. For that reason, I will be teaching a course about process optimization at Loyalist College in Canada this spring. This Ontario college has launched an eight-month post-graduate certificate program to offer students in-depth skills to support the burgeoning cannabis industry.

Although, not only do we need to change the mindset of cannabis professionals through education, we also need to overhaul production methods from the ground up. I studied chemistry at universities for over a decade, culminating in a PhD in chemistry. And still, I don’t know what the “best” method is to extract flavonoids from cannabis. The main reason for that is that right now there is basically no data on flavonoids in open-market flower or extracts. And if I cannot “see” them, I cannot optimize for them.

I picked winterization as an easy target in the article you mention. For one, winterization is the wrong term to describe the process-it’s a precipitation. Secondly, it is a bottle neck in production. The step is either unnecessary or can be done more efficiently with other instruments. It just takes knowing about them and how to use them. The point I wanted to make was not about winterization in particular, but about how cannabis production steps are often habitually tackled. Methods that were done in a basement have now been scaled-up and brought into a commercial building. However, this doesn’t mean they are the “best.” Production methods should not evolve into larger scale, but they need to be revolutionized into new practices, if we want to industry to further develop.

What are the next steps your research?

Roggen: I just moved to Vancouver to pursue more fundamental cannabis research. Before my research was within the constrains and goals of commercial production. Now, I am partnering with researchers at the University of British Columbia to work on questions that might not have an instantaneous commercial value, but which push the industry forward in the long run.

We are currently working on a paper about computational studies of the decarboxylation pathway. One might wonder about the value of such research, especially since decarboxylation reactions are already a routine step for extract producers. But I predict that what we discover will facilitate an improvement to the current process that leads to reductions in cost and THC degradation.

Actually, the question of THC degradation is another project that I am pursuing right now.

To sum up, I am quite excited about the work ahead. I’m a big advocate of making my results public through publications and conference lectures, so please be sure to watch this space.


  1. M. Roggen, Cannabis Industry Journal, August 3, 2017,

  1. M. Roggen, Marijuana Venture, October 30, 2018,

How to Cite This Article

M. Roggen interview, Cannabis Science and Technology2(1), 54-55 (2019).