Water Activity: The Key to Controlling Flavor and Safety in Natural Cannabis Products

Natural products for human consumption have several critical quality attributes (CQAs) that are related to composition. If one is not careful, the composition of natural products can change with time, sometimes for the better, but generally not. “Fresh” eggs and milk are preferred. Only a few products such as cheese and wine improve with age. Many fresh products have “use-by dates” to help the consumer avoid products that are about to degrade or spoil.

The many varieties and uses of cannabis present a new, complex issue for packaging and storage. With mild heating or time, THCA can decarboxylate to THC, a psychoactive cannabinoid. Many of the aroma-generating chemicals are semivolatile organics, which can slowly sublime and be absorbed preferentially into and possibly through nonpolar plastics such as polyethylene and polypropylene. However, water is a major cause of degradation in CQA of cannabis products, including leaves, extracts, and finished formulations. “Warm and wet” is often bad.

Let’s take cannabis leaves and flowers. These are grown until the flower starts to mature. The grow conditions can be optimized by increasing the light intensity and duration. Adding carbon dioxide in enclosed grow rooms also improves photosynthesis and hence yield. Controlling the relative humidity in the grow room can boost yield, up to a point. If the humidity is too high, mold and bacteria can start to grow, and plant health declines rapidly. Growers have responded by restricting access to grow rooms or even entire properties. Some require changing into protective clothing to protect the plants from pathogens carried by streetwear.

At harvest time, the flowers and leaves are separated from the stalks and roots and soon enter the drying rooms. These rooms are designed to quickly (<3 days) remove water (called water activity, aW) from the plant material to prevent spoilage. Generally, dry rooms circulate a lot of cool (15–22 °C), dry (35% RH [relative humidity]) air.1

Drying must be done quickly since the microorganisms (bacteria, yeasts, and molds) are waiting to grow when their microenvironment is just right. Mycotoxins—toxins produced by mold—are a particular problem. Aflatoxin, a potent liver carcinogen, is one such mycotoxin. It can be released during smoking and ingested by humans in first- and second-hand smoke. Patients with HIV/AIDS and cancer are vulnerable to aspergillosis, a potentially fatal lung disease that can result from ingestion of cannabis molds such as Aspergillus. 

Table 1 presents the minimum water activity required by various microorganisms for growth.2 For example, salmonella will not grow when the aW value is lower than 0.93. Cannabis is protected from spoilage from microorganisms when the aW is less than 0.60. However, other factors such as light and warm temperature can lead to degradation of the sample.

Table 1 –Water activities (aW) required to support the growth of representative microorganisms


(reproduced from Ref. 2)

Commerce requires use-by dates

Some production chains can involve multiple steps, possibly with several business entities. The raw input materials need to be traceable to assure suitability and responsibility. Since materials change with age, complying with use-by dates can be critical.

Both medicinal and recreational retail customers will buy cannabis products including leaves and flowers based on a match between their needs and the CQA of the product. Generally, they will purchase more than a single dose; hence there is interest in storage. The CQA profile of the cannabis product is vulnerable to changes with time. Regulators thus often require and track use-by dates, especially if a problem arises.

Use-by dates are determined experimentally. The simplest way is to store representative samples under well-defined conditions and assay them on the prescribed dates. Key assays include potency based on the amount of active pharmaceutical ingredient (API). A change of less than ±10% from the label amount is usually acceptable. Larger percentage tolerance bands are acceptable, especially if the amount is low. The specific tolerance limits should be documented in the certificate of analysis (CoA), manufacturing quality control specifications, and process log.

In some instances, it is impractical to wait months or years for the test times to pass for real-time assays. Thus, accelerated stability-indicating assay protocols have been developed. These typically involve storing aliquots of the sample for days or weeks at exaggerated heat and humidity conditions, typically 40 °C and 75% RH, to accelerate degradation reactions. For example, a patent involves different stabilizers in developing cannabis therapeutic formulation.3 The teachings of the patent show that water degrades THC and CBD quickly at 55 °C.

However, even at room temperature, THC and CBD can be damaged by molds and bacteria caused by excess moisture. But over-drying must be avoided, since dry leaves and flowers become brittle, which makes it difficult to roll for smokes. The aW is about 55–62 for smokable cannabis.

An article from points out that flavor and other CQAs are important in sales:

“Over time, the cannabis trichomes that house those precious terpenes and cannabinoids will break down if your humidity levels are too low. For the consumer, this means you’ll no longer get an overwhelming whiff of oranges from a jar of Tangie, or taste the rich, juicy notes that give Blackberry Kush its name. Losing terpenes can take away from the medical benefits experienced by patients, too; myrcene, for instance, is considered partially responsible for the anti-inflammatory effects attributed to strains like Himalayan Gold, and losing those terpenes can mean losing those positives.”4

Controlling the humidity of smokables like tobacco is not new technology, but the sweet spots for tobacco and cannabis are different. For example, California and Oregon regulations limit flower products to an aW value of less than 0.650.

Laboratory impact

Today, measuring the stability of cannabis products is more an art than a science. A Google search leads to several general methods, but I was unable to find any description that had scientific rigor. So, the lab will need to develop its own protocol to determine use-by dates. The basic experimental design is to use existing protocols approved for lot release run at advancing time points such as 30, 60, 90, 180, 270, and 360 days. A rigorous study of storage conditions seems long overdue. It would be an interesting factorial design experiment.

Let’s look at the key variables affecting stability.

Water activity

Water activity (aW) is the most significant variable, since the target window is narrow and the consequences of failure to control render the product inconsistent, unusable, or even deadly. The target aW is 55–62. Instruments for measuring aW are available.

This all seems complicated until one sees the simple active sample storage system proposed by Leafly. It uses Mason/Ball canning jars for the sample container. Size (100 mL to 4,000 mL) depends on the sample. The wide mouth is ideal for retrieving the shredded plant material for consumption or analysis. To control humidity, they add an Integra Boost Humidiccant packet from Desiccare (Reno, NV). This is a sealed package that uses patent-pending technology that releases or absorbs moisture to maintain an aW of 62% in a closed environment such as a sample jar.

How does the packet work? Recall Raoult’s law from your chemistry course? The vapor pressure of a component of a solution is the vapor pressure of the pure material times its mole fraction in the solution. The Integra Boost controls water vapor composition at 62%, which is achieved by dissolving a water-miscible polymer so that the mole fraction of water is 62%. Thus, the packet absorbs water when the RH is over 62%; water is delivered from the packet when the RH is below 62%. For those who desire a drier product, the company offers packets with an RH of 55%.

The novel part of the design is the membranes and fabrication that permit transfer of water vapor in or out of the packet without wetting the dry contents (sample).

Storage temperature

The choice of storage temperature is controversial for leaf cannabis and flowers. It is generally accepted that cool is better. Warm is definitely bad, since it promotes the growth of microbes listed above. Plus, it increases volatilization and chemical reactions including decarboxylation of THC-A and CBD-A. Temperatures above 70 oF (21 oC) are considered warm. The American Herbal Pharmacopeia lists 18–20 oC.5

Freezing is generally not recommended since the flowers and trichomes become brittle. When they break, they release terpenes with their flavor. If you do freeze the sample, yet want it to be representative, let it warm up to about 60 oF before opening the lid and extracting a sample. This may require several hours for larger containers. Opening the sample jar with cold contents may lead to condensation of water on the contents, which increases the aW. Degradation arising from repeated freeze/thaw cycles may be important and should be recorded as part of the sample and standard log.


The short answer is: keep it in the dark. This can be done by wrapping glass containers in aluminum foil or putting glass containers in a carton closed with a top, in a drawer, etc.

Container materials

Glass seems to be the preferred material for containers. Plastic bags are often used, but web sources do not recommend them. Pay attention to the closure design. Ideal closure should:

  • Be gas-tight under test conditions
  • Prevent metal contamination, especially with As, Cd, Pb, and Cd

The same design should be used for sample and large storage containers. Home growers seem to have focused on Ball or Mason canning jars.

Sample plan

Validating a method should be built around three lots run in parallel. Each time point should be run in triplicate, with a second set of three held in reserve. This is 18 samples per assay and time point. A two-year cycle could require 12 time points for a total of 216 samples/assay. At a minimum, the assays should run aW, potency and degradation profile, and microbiology. With care, these assays could be run on the contents of each 100-mL (4-oz) glass container.

Distribution facilities should carefully design the inventory storage facility and work flow to avoid thermal degradation of the products waiting for sale and testing. Customers should be given a written, illustrated protocol to follow for home storage.

Accelerated testing protocols

A two-year hold time required to establish the real-time performance is not practical for the vast majority of commercial samples. So, accelerated testing protocols, such as reported by Kottayil,3 will need to be developed. One will also need to calibrate the accelerated test results against the real-time assay.


  1. Upton, R.; Craker, L. et al. Cannabis inflorescence Cannabis spp. Monograph; Amer. Herb. Pharm. 2014, 30.
  2. USP 112. Application of water activity on determination of nonsterile pharmaceutical products.
  3. Room-temperature stable dronabinol formulations. Patent no. 8,628,796 B2;
  5. Upton, R.; Craker, L. et al. Cannabis inflorescence Cannabis spp. Monograph; Amer. Herb. Pharm. 2014, 31.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.



Product Enquiry