Make some hot chocolate and settle in over winter break as Leafly’s Nick Jikomes, PhD, unshackles your mind from the chains of both THC and terpene scores. We hereby grant you permission to follow your nose into weed’s wilderness.
“What drives the aroma of marijuana, and gives different strains their characteristic smell?”
Ask this question to any budtender or most cannabis scientists–the experts–and you’re likely going to get this answer: terpenes.
It’s logical to suppose that cannabis aroma is driven primarily by terpenes: They are the most abundant class of volatile compounds (those you can easily smell) in cannabis. They’re literally vaporizing at room temperature. They are also abundant in nature, dictating the aroma of many plants. Cannabis contains other volatiles, but they are present in much lower quantities. Terpenes are therefore the natural candidate molecules responsible for marijuana’s aroma.
Not only are terpenes supposed to be the main drivers of cannabis aroma, they are hypothesized (often with great zeal and confidence) to play a key role in the psychoactive effects of cannabis. THC may be the biggest player, but an “entourage” of terpenes shapes the particular and sometimes divergent psychoactive effects that different strains are said to have.
This idea, popularly known as the “entourage effect,” gained traction among cannabis enthusiasts following this influential 2011 paper. It’s not uncommon to hear budtenders in marijuana dispensaries cite it as they explain the unique effects of different strains and what causes them. And sometimes it seems to match.
Problem: Two strains with the same terpenes can smell and feel different
But several cracks in this narrative have emerged in recent years–observations explained away with awkward answers.
For one, no cannabis terpenes have a distinct “marijuana” smell. The most common explanation for this is to double down on the dominant narrative: the smell of cannabis strains doesn’t come from one terpene. It’s the whole entourage. Our sense of smell works in subtle and complicated ways. The aromas we perceive can’t be reduced to specific terpenes. That may be true for other plants (pinene does smell like pine), but not cannabis.
Another puzzling observation: when carefully study terpene entourages in legal weed, strains with similar terps can smell pretty different.
How to order weed delivery online with Leafly
We see this in Leafly’s research: there are distinct terpene entourages consistently found among commercial strains, which can be used to cluster strains into distinct “chemotypes” (chemical phenotypes). Many popular strain names are robustly associated with specific terpene entourages. And yet, the same entourages are often associated with strains reported to have distinct aromas.
The industry has been trying to decode the chemistry of cannabis in order to reliably predict which products will have the effects consumers desire.
“This strain will smell like this and feel like this, because it has this specific entourage of cannabinoids and terpenes,” we want to say. But these decoding efforts have yielded mixed results. Time and again, the people doing the research have told us: “People don’t perceive a strain with one profile as “Fruity” and another as “Gassy” in a way that maps with the dominant terps.
We concur. White Tahoe Cookies and Original Glue both score high in caryophyllene, myrcene, and limonene. However, Leafly users say Cookies has notes of “butter,” “peach,” and “nutty,” while the Glue is “pungent,” “pine,” and “earthy.” And they’re right.
Yes, its’ complicated. Yes, strain names vary. Yes, subjective experiences differ. But something beyond the terp scores is happening. And research conducted by chemists at Abstrax Tech points to more being out there. And it forces us into a world with more than terps.
Meet the non-terpene cannabis volatiles & the aroma of dank
To understand the volatiles present in cannabis beyond terpenes, you need instrumentation more sensitive than what’s commonly used by commercial labs to measure cannabinoid and terpene content.
This is because what’s being measured, non-terpene volatiles, are present at minute levels. Parts per million. And parts per billion. The human nose can smell them.
Using ultra-sensitive analytical techniques, chemists at Abstrax Tech—who are trying to formulate more precise vape flavors—measured both the terpene and non-terpene volatile content of ice hash rosin samples made from various strains. Ice hash rosin was used for several reasons: (1) it’s a concentrated form of cannabis oil, allowing for easier access to lower quantity molecules; (2) it’s produced using mild manufacturing conditions which minimize loss of volatiles; (3) rosin extracts are produced from many plants, providing an average profile across many samples.
To evaluate aroma, a panel of cannabis consumers described the aroma of rosin samples. Panelists were not trained to use a specific lexicon of words in order to emulate the experience of typical consumers. They were asked to describe what they smelled and provide intensity ratings on a 0-100 scale. Raw user inputs were then classified into three broad groups:
- Sweet exotic (samples had a distinctly sweet/fruity aroma compared to average)
- Prototypical (samples with a more or less “classic” marijuana smell)
- Savory exotic (samples had a distinctly savory/chemical aroma compared to average)
Despite natural variation in the aroma descriptors that untrained consumers use, there was a remarkable level of alignment between panelists: they consistently reported that certain rosin samples fell into either the exotic sweet, prototypical, or exotic savory category.
These sensory data were then compared to the terpene and non-terpene volatile contents of each sample. They found that terpene content poorly predicted the aroma descriptors and classifications of the sensory evaluations consumers provided. For example, limonene, one of the most common cannabis terpenes, was the dominant terpene in some of the most exotic sweet and some of the most exotic savory samples.
Instead, smelly chemicals that were not terpenes much more robustly correlated to reviews. Aromas were related to groups of these non-terpene volatiles and in some cases even to individual compounds.
Below is a brief description of some of the major classes of non-terpene volatiles the team detected, together with their associated aroma profile and the cannabis strains that contained them.
Tropical Volatile Sulfur Compounds (VSCs)
- Chemistry: Sulfur-containing compounds similar to, but different from, the VSCs previously found to produce the “skunk” aroma of certain strains.
- Aroma: Tropical nuances with citrus, fruity, and sulfuric aromas.
- Strains: Garlic Cocktail #7
- Chemistry: Volatiles containing indoles–specific ring structures found in many biologically important molecules such as tryptophan and melatonin. This structure is also a key component of psychedelic tryptamines like psilocybin, LSD and DMT.
- Aroma: Floral, mothball-like scent likely contributing to the prototypical cannabis aroma rather than specific exotic scents.
- Strains: Indole is common in cannabis. Skatole trends strongly with “savory exotic” samples like GMO, Garlic Cookies, and 710 Chem.
- Chemistry: A previously unreported family of ester compounds ubiquitous in “exotic” cannabis.
- Aroma: Members of this group share a similar, fruity base note with different nuances (e.g. banana, green apple, etc.)
- Strains: Many strains with high “exotic scores,” such as Starburst 36 #1 and Motornana.
- Chemistry: There are many ester compounds found in nearly every fruit and other plants. Over 30 were identified in cannabis.
- Aroma: A wide range of sweet- and fruit-like aromas.
- Strains: Banana Scream.
Structurally Unique Compounds
- Chemistry: A diverse group of molecules with unique chemical structures compared to other cannabis volatiles. Some of these had not been identified before in cannabis. Many are used as flavorants in food or beverage products, or naturally in various plants and fruits.
- Aroma: Grapelike aromas.
- Strains: Papaya Peach, Juiceman
The exotic score: Terpene content does not predict the cannabis experience & aroma of strains
In the Abstrax study, researchers gave each strain an “exotic score”—a zaza score—based on the ratings provided by the consumer panel. Exotic strains were consistently rated as having especially sweet aromas compared to prototypical cannabis (see below for a full list of strains sorted by exotic score). And there was a wide range of exotic scores across all of the studied strains. This white paper contains more information for those interested in more detail.
The Abstrax team looked at the correlation between the exotic score of strains and three different classes of volatiles:
- flavorants (all of the minor, non-terpene volatiles)
- and two different classes of terpenes: monoterpenes and sesquiterpenes.
Flavorants and monoterpenes both had statistically significant positive correlations with exotic scores. However, the strength of this association was greater for flavorants, indicating that they are likely a stronger driver of the exotic compared to monoterpenes.
Looking at individual terpenes, the Abstrax team found mixed results. Some terpenes, such as pinene, ocimene, and linalool, showed modest positive associations with exotic scores. Interestingly, some of the most abundant cannabis terpenes, like limonene and myrcene, showed little to no correlation.
Another study: The nose beats the scores on the bag
These results echo another recent report that looked at the relationship between the terpene and THC content of strains and the perceived quality of their aroma and psychoactive effects.
In that study, researchers found a positive association between strains’ psychoactive appeal and aroma score: those with more pleasant aroma ratings tended to be those with more pleasant psychoactive effects. The nose knows what’ll get you lifted.
Intriguingly, the THC score and total terp score did not predict which strains produced the most pleasant psychoactive effects. To repeat: There was not a significant correlation between the quality of strains’ perceived psychoactive effects and their total terpene or total THC content.
Pushing beyond THC and terpenes in ‘24
Collectively, these results suggest cannabis terpenes are a part of the aroma equation—but do not comprise the totality of it. Every smell molecule is not created equal. Terpenes are the most abundant class of volatile compounds in cannabis oil, but various flavorants present at very low levels seem to be the bigger drivers of each strain’s sensory “personality.”
If the goal is to crack the chemistry code of cannabis in order to predict the sensory qualities of cannabis for consumers, then the industry has work to do. The thorough chemical dissection performed by Abstrax was made possible by their use of highly sophisticated analytical methods that are not used in most commercial testing labs. The flavorant content of the legal cannabis products on dispensary shelves is simply unknown and it’s unlikely that systematic measurements of these things will be commonplace anytime soon.
For now, we can simply report the exotic scores of the strains tested in the recent Abstrax study (listed below). Beyond that, consumers will simply have to follow their nose.
For more information about the research Abstrax has done on cannabis terpenes and flavorants, check out their, “Science of Exotic” anthology.
The “Exotic” Score of all Abstrax-tested strains
- Grape Pie x Do-Si-Do = 87.4
- Juice Man = 87.1
- Papaya Peach = 86.3
- Cake Crasher = 83.7
- Starburst 36 #1 = 81.7
- Upsidedownfrown #5 = 80.7
- Zkittlez 710 = 80.1
- Banana Scream = 79.7
- Bubblegum Zkittlez = 77
- TK x Butterscotch = 76.3
- Purple Churro = 74.3
- Pure Guava = 74
- Rainbow 2.0 = 71.6
- Fruity Pebbles = 71.4
- OPP x Smarties = 70.9
- Original Glue = 65.1 (Note: the Original Glue sample studied was likely from another strain, mislabeled as Original Glue)
- Bubba Kush = 62.1
- Trainwreck = 43.7
- Bacio Gelato = 43.1
- MAC 1 = 39.3
- Gelato #33 = 34.6
- Kimbo Kush = 34
- Lucky Charms = 30.7
- Pie Hoe = 25.3
- Cookies & Cream = 15.4
- GMO Cookies = 5.7
- OG = 3
- 710 Chem = 1.9
- GMO = 1.7
(As you can see the top exotic scorers really do look like some zaza. But we also love savory GMO, so let’s not ignore everything off the Chem line.)