Hemp Testing 101: Analytical Testing Protocols Explained and Evaluated

Using Proper analytical testing methods is essential for the hemp industry.

Proper analytical testing is vital to the hemp industry. In order to qualify as “industrial hemp” under the 2014 Farm Act (or as “hemp” under the 2018 Farm Bill), cannabis must be tested for delta 9 (∆9) tetrahydrocannabinol (THC) concentrations to ensure that it does not exceed three tenths of one percent (0.3%) on a dry weigh basis. Analytical testing raises a number of issues, including, but certainly not limited to, the specific testing methods that are (and should be) used, at what point in the growth cycle of the cannabis plant testing should occur, what part(s) of the plant should be tested, from which part(s) of the field plant samples should be acquired, how the plant material should be combined to form a test sample, whether the acid form of THC should be decarboxylated as part of the testing process, and the scope of the test (i.e., the degree to which other compounds are measured, if at all, in a particular test).

In this article I will discuss the first issue, testing methods. In particular, I will address the fact that different testing methods produce different results. This can lead to catastrophic legal consequences when an improper testing protocol is used.

Part 1- Different Tests Produce Different Results

A cannabis plant sample can be subjected to several different types of scientific tests to determine its ∆9-THC concentrations. I was initially surprised to learn that different tests produce different results. In fact, since testing requires both separating plant molecules and then measuring them, there are a number of different combinations of tests, all of which can generate different outcomes. In a very real and practical way, using one testing method can result in a different legal outcome than a using a different method. This article addresses the primary types of tests currently used in the industry and the ones most appropriate for hemp.

Part 2- Evaluation of Commonly Used Testing Methods

The cannabis plant has a complicated biological profile. It includes more than 400 compounds, including cannabinoids, terpenoids, alkaloids, flavonoids, fatty acids, etc. Teasing out and measuring one of those compounds, ∆9-THC, can be done using several different methods. I argue that high performance liquid chromatography (HPLC) combined with mass spectrometry (MS) is the most accurate and reliable testing protocol for measuring ∆9-THC concentrations in hemp.

a. Separation techniques

All cannabis testing utilizes chromatographic techniques, which are chemical tests that determine whether the characteristic chemical constituents of the plant (in our case, ∆9-THC) are present. On a fundamental level, this requires separating out the various compounds. Separation is typically done by using chromatography, which is a laboratory technique for separating a mixture. The mixture, a cannabis sample, is dissolved in a fluid or a gas called the mobile phase, which carries it through a structure holding another material, a solid or liquid supported by a solid, called the stationary phase. The various constituents of the mixture travel through the stationary phase at different speeds causing them to separate. In other words, ∆9-THC travels through the stationary phase at a different rate than the other cannabis compounds. It is separated from them by virtue of its unique speed during the test.

The most common chromatographic procedures are thin-layer chromatography, gas chromatography, and high performance liquid chromatography. Each operates differently.

i. Thin Layer Chromatography (TLC). The simplest, and least used, chromatography method is TLC. This technique is performed on a sheet of glass, plastic, or aluminum foil that is coated with a thin layer of absorbent material, such as silica gel, as the stationary phase. The mobile phase is a liquid. The primary advantages of TLC are that it is relatively simple and inexpensive, that it requires less time than other techniques, and that it uses fewer amounts of substances. Its primary disadvantages are that its detection limit is relatively high (ie, it is not very sensitive) and that it is an “open system”, which means factors such as humidity and temperature will affect the result. TLC is not appropriate to hemp because of these limitations.

ii. Gas Chromatography (GC). A commonly used chromatograph technique is GC, in which the mobile phase is a gas (vapor) that passes through a column which is packed with a solid stationary phase (usually silica). With GC the sample must be heated in order to vaporize it. This allows it to pass through the column with a carrier gas. GC is a very good technique for separating volatile organic compounds, such as terpenes, which easily become vapors or gases. However, the fact that GC requires heating the sample is particularly problematic for separating ∆9-THC. GC heats the cannabis sample to a temperature that changes its molecular structure in such a way that higher concentrations of the molecule being measured (ie, ∆9-THC) are created. This can cause a sample which initially contains ∆9-THC concentrations within the statutory limit to test above the limit. In other words, the GC technique can, and often does, give a “false positive” result. This transformation is called decarboxylation.

On a molecular level, all cannabinoids contained within cannabis trichomes have an extra carboxyl ring or group (COOH) attached to their chain. Decarboxylation is a chemical reaction that removes a carboxyl group and releases carbon. It converts a cannabinoid, such as THC-A (the “A” stands for acid), from its acid form to its neutral form. The conversion is caused by heat. THC-A is the precursor to ∆9-THC, which is a neutral form of the THC cannabinoid. THC-A converts to ∆9-THC when heated. The heat used in GC is sufficient to decarboxylate THC-A and convert it to ∆9-THC. In other words, GC literally creates higher levels of the very compound being measured. Using GC to separate ∆9-THC for measurement is somewhat akin to a highway patrol officer using a radar gun that speeds up the vehicle it is measuring. The mere act of testing increases the quantity (or, in the radar gun analogy, the speed) of the thing being tested. For this reason, GC is an inappropriate and unreliable technique for separating ∆9-THC from a cannabis sample.

Given that GC is an inappropriate chromatograph technique for hemp, it is reasonable to ask why is it used at all, much less as the de facto test in much of the USA. Although there is no single answer, its popularity in the marijuana industry is a likely reason.

GC is a popular analytical technique in adult-use (“recreational”) marijuana states. Generally speaking, the average consumer who purchases cannabis in a recreational state wants as much ∆9-THC “bang for the buck” as possible. Cannabis companies know this and understandably focus on marketing the ∆9-THC concentrations in their products. Using GC as part of a testing protocol necessarily involves increasing (some would say “inflating”) the reported ∆9-THC levels of a given cannabis plant. If we assume the average consumer wants more ∆9-THC, then this testing method results in higher sales.

Although at first glance this may appear to be somewhat deceptive from a marketing standpoint, the fact is consumers of flower products in adult-use states almost always decarboxylate their cannabis before or during ingestion by smoking, vaping, or heating it as part of an edibles recipe. So, while the GC technique is technically inaccurate when it comes to separating the actual ∆9-THC in a plant, it is accurate in its practical application since most people in adult-use states heat the cannabis flowers they purchase and decarboxylate the THC-A in them. In this way, GC is a somewhat appropriate method to determine the total “available ∆9-THC” in a marijuana sample. For cannabis testing in recreational states, at least for cannabis destined to be heated by the consumer, this is fine and perhaps even the most appropriate measure. Additionally, since the legal status of a cannabis crop in an adult-use state is not determined by whether or not it exceeds a statutorily defined ∆9-THC limit, the consequences of an inaccurate result are not usually significant. However, for hemp states an inaccurate result can result in the loss of an entire crop. Finally, there are a number of different uses for hemp flower and biomass, most of which do not involve heat and decarboxylation. GC is a decidedly inappropriate and unreliable technique in these situations.

As an aside, I have heard and evaluated the argument that a mathematical step can be inserted into the GC protocol to account for the effect of decarboxylation and determine the true ∆9-THC in a cannabis plant sample. The idea is that while GC increases the ∆9-THC present in a sample, a mathematical step can fix this issue in the final result. Although relatively easy in theory, this does not work well in practice. In order to obtain a true measure of the ∆9-THC concentrations in a given plant you must have a method to determine the concentrations present in the initial sample. Adding a mathematical step to GC necessarily makes an assumption of the amount of ∆9-THC and the amount of other molecules which can convert to ∆9-THC originally present. Although somewhat “circular”, this can be done; however, doing so introduces a new opportunity for uncertainty (ie, by making an assumption of the original amount of ∆9-THC present) and also adds an additional source for potential error. 

iii. High performance liquid chromatography (HPLC). In HPLC a liquid mobile phase is pumped at high pressure through a porous solid, such as glass, silica, or aluminum that is packed into a glass or metal tube or that constitutes the walls of an open-tube capillary. HPLC is the most reliable separation method for ∆9-THC because it does not involve high temperatures and leaves all relevant molecules intact. This means it is highly accurate and does not require the addition of a mathematical step or assumptions about the concentrations of the very thing that it is separating. Aside from not decarboxylating the sample, another advantage to using HPLC is speed. Due to the pressure exerted on the mobile phase, HPLC also produces the fastest results. Its disadvantages include that it can cause coelution, in which two compounds with similar structure and polarities exit the HPLC device at the same time. This makes measuring the relative quantities of the respective compounds difficult or even impossible. Fortunately, coelution of cannabinoids is rare. HPLC is also generally more expensive than GC, though not significantly.

b. Measurement techniques

Once the ∆9-THC has been separated from other compounds in a cannabis sample by using chromatography, the actual amounts must be detected and measured. This can be done by using several different devices, the most common of which are a flame ionization detector, an ultraviolet detector, and a mass spectrometer.

i. Flame Ionization Detector (FID). A commonly used detector for GC is FID. It is a “mass sensitive instrument” because it measures ion mass per unit of time. FID operates by collecting and measuring the amount of analyte (ie, the substance being measured) in a gas stream. The fact that it operates in the context of a vapor makes it particularly useful for GC; however, for this reason it is not used with HPLC, which does not emit a gas. This is its primary disadvantage for use with hemp. Given that GC is inappropriate for separating ∆9-THC for the reasons discussed above, FID is inappropriate by virtue of the fact that it operates with GC. Additionally, of all of the detection and measuring methods described in this article, it is the least specific. (As an aside, it is also the detection method most utilized by the DEA.)

ii. Ultraviolet Detector (UV). This detection device operates on the principal that chemicals absorb UV light at specific wavelengths. The principal at issue, Beer’s Law, states that the concentration of a chemical is directly proportional to the absorbance of a solution. In UV, the concentration of ∆9-THC can be determined by the amount of UV absorption that occurs. Although UV is generally an accurate detecting technique, it can generate a false positive result when UV absorbing molecules in addition to ∆9-THC are present during the detection stage.

iii. Mass Spectrometer (MS). MS detects various molecules based on their distinctive mass and mass fragmentation pattern. It does this in three stages. The first stage, ionization, vaporizes molecules in a sample and then bombards them with an electron beam, which causes them to become charged particles (ie, ions). The ions are then sorted by two processes, acceleration and deflection, and measured. MS only measures the mass of ions. Neutral molecules are not detected. The result is a spectrum. A high quality mass spectrometer detects and measures a fragmentation of each molecule. This removes much uncertainty and generates a very accurate result. For this reason, MS is the most reliable method to detect and measure ∆9-THC in a cannabis sample.

Part 3- Conclusion

Testing protocols may seem abstract. Teasing out their differences can be tedious. However, the testing protocol used to measure ∆9-THC concentrations can mean the difference between a lawful hemp harvest and a “hot” crop. This can translate to having a profitable crop or a total loss.

Although HPLC in conjunction with MS is the most reliable protocol for testing ∆9-THC concentrations in a sample, it is not the most prevalent. Anecdotally, it appears that GC-FID is the most common protocol, despite the fact that GC increases ∆9-THC in the sample being tested, and FID is the least specific detector of the available options. The reason for the prevalence of this testing protocol is unknown. One of the most likely reasons is that GC-FID is commonly used in adult-use marijuana states, where higher ∆9-THC levels often mean higher sales and where measuring “available ∆9-THC” makes sense because most consumers will smoke, vape, or heat it to make edibles. However, the same protocol is inappropriate and unreliable for hemp.

Federal law explicitly states that the operative issue for determining whether a particular cannabis crop is lawful hemp or unlawful marijuana is its ∆9-THC concentration. Any reliable testing method must, at a minimum, measure the actual concentrations of ∆9-THC in order to comply with the statute. Attempting to quantify the total “available THC” using a testing method that distorts the result by creating the very molecule it is measuring is contrary to law.

Additionally, most hemp plants are not destined to be heated (ie, by smoking, vaping, or cooking in a recipe). The THC-A molecules in these plants will never be decarboxylated. For this reason, and aside from conflicting with the statute’s requirements, testing total “available ∆9-THC” makes improper assumptions that do not always, or even often, apply to hemp. The only consistently reliable testing protocol is HPLC-MS.

As the hemp industry continues to consolidate standards it must promote and adopt reliable testing methods. Since the threshold issue in determining a cannabis plant’s legal status is its ∆9-THC concentrations, an error caused by an unreliable method is intolerable and contrary to law. The most reliable testing protocol is HPLC-MS, which should be adopted as the de facto test for measuring ∆9-THC in hemp.

Special thanks and acknowledgement. I am indebted to my friends, Dr. Chris Hudalla of ProVerde Laboratories and Dr. Volker Bornemann of Avazyme Laboratories, for their invaluable information and insights that I obtained in conversations and email exchanges over the past several months. Helping a lawyer and philosophy major gain some small understanding of the scientific issues presented in this post is no small feat. I also thank two brilliant fellow soldiers in arms, attorneys Marshall Hurley and Bob Crumley (of Founder’s Hemp), for traveling with me up the learning curve. And, I appreciate those nice scientists from Waters Corporation that I was fortunate enough to meet at a recent networking event during a hemp conference for their willingness to forego casual conversation to talk these things out with a complete stranger. Most importantly, I thank my wife and office manager, Ashley, for her unwavering support and wicked editorial skillz. I apologize for any inaccuracies in the article, all of which are mine alone.

December 3, 2018

Rod Kight is an attorney who represents lawful cannabis businesses. He speaks at cannabis conferences across the country, drafts and presents cannabis legislation to foreign governments, is regularly quoted on cannabis matters in the media, and maintains the Kight on Cannabis legal blog, where he discusses legal issues affecting the cannabis industry. You can contact him here.

This article appeared in the December 2018 edition of the Cannabis Law Journal. Many thanks to editor Sean Hocking for his support and friendship. 

16 comments on “Hemp Testing 101: Analytical Testing Protocols Explained and EvaluatedAdd yours →

  1. Thank you for this excellent explanation, Rod. Most of us in the industry just know that “testing happens”. Your breakdown is helpful and provides a perfect introduction for further study.

  2. Hi Rod,

    Thank you for the very thorough assessment of hemp laws, the challenges of GC testing, and delta-9 THC levels in industrial hemp. I was hoping you could comment on some text related to testing hemp in the Agricultural Act of 2018. The text is linked here and is as follows:
    https://www.govtrack.us/congress/bills/115/hr2/text/enr#link=X_10113_~Q1_G_297B_a_2_A_ii_~T1&nearest=HE103B75298874A37BADFE847E7FED538

    a procedure for testing, using post-decarboxylation or other similarly reliable methods, delta-9 tetrahydrocannabinol concentration levels of hemp produced in the State or territory of the Indian tribe;

    The farm bill seems to explicitly state they expect post-decarboxylation values to be tested for delta-9 THC, which seems to negate the “dry weight basis” only delta-9 THC rule. Is this your interpretation? And if so, can you comment on how this may impact high CBD hemp production at this time?

    1. Amy- Thank you for reading and commenting. You make a very good point. I’ve discussed it with some of the drafters of the bill. They confirm my view that “other similarly reliable methods” is the key to having HPLC be deemed an appropriate test. The statute cannot simultaneously be read to say that delta-9 THC is the sole defining factor to determine whether a particular cannabis plant is lawful hemp or unlawful marijuana AND post decarboxylation testing is the sole testing method. This is nonsensical. The proper reading of the statute is that a post-decarboxylation OR OTHER SIMILARLY RELIABLE method is appropriate. Since gas chromatography is the most used method in the industry, it makes sense that the statute would reference it (albeit obliquely) as being allowable among other similarly reliable testing methods, such as HPLC. In other words, I read the statute as assuming HPLC but allowing GC. -Rod

      1. The new USDA rule seems to differ from your opinion that the farm bills are only regulating delta-9 THC. It seems, from the USDA rule, that what the farm bills are specifying as “delta-9 THC” in the definition of hemp is actually the post-decarboxylation result for delta-9 THC, ie total THC.

        Why did congress not state this directly in the definition of hemp contained in the farm bills? How can the USDA legally re-interpret this definition to specify total THC? Is there evidence of bad-faith here?

        1. Good questions, Aaron. We do not know why Congress did not state this specifically in its definition. As for your question about the USDA, you are correct that the USDA cannot create rules that contravene a statute. While there are arguments that the USDA overreached, the ambiguity of the statute, which says that a “post-decarboxylation or other similarly reliable method”, is probably sufficient for it to use a method that tests for post-decarboxylation levels of THC.

  3. I have a few questions related to the LC-MS/MS based assay that utilizes MRM (multiple reaction monitoring) step in fingerprinting each known analyte (reference standard should be available).

    It is well known fact that the MS/MS peak intensity depends on the ionization status of the parent ion of selected analyte, and that unfortunately is highly affected by matrix. For instance in bio-analytical assays matrix match is mandatory in the process of sample preparation for standard calibration curve. The blank matrix (blood, plasma, cell or tissue extract, or any other biological fluid) should not contain any trace of the target analyte in order to avoid any interference with spiked reference standard (analyte).

    What kind of matrix (blank plant extract) is employed in your test for the accurate measurement of expected analytes? It is hard to imagine that there are any extracts from hemp materials that is absolutely free from any cannabinoids or terpenoids, to name a few.

    Thanks for your clear answer.

    Levan Darjania, PhD

    1. Dr. Darjania- Thank you for your insightful comments and questions. In order to respond appropriately I reached out to my friend, Dr. Volker Bournemann of Avazyme Laboratories in Durham, NC, for his thoughts. Avazyme is a contract analytical laboratory for the NC Industrial Hemp program. In response to your question Dr. Bournemann states: “Matrix effects in quantitative mass spectrometry based bioanalysis are quite common and frequently overcome by preparing matrix-matched standards, especially in pharmaceutical bioanalytical applications. In natural products bioanalysis of environmental (soil, water) and crop derived samples it is often difficult to obtain relevant matrix samples that are free of the desired analytes (e.g., cannabinoid or terpene free hemp). At Avazyme, we overcome these matrix effects by utilizing state-of-the-art triple quadrupole mass spectrometers with sufficiently high sensitivity levels, that allow us to dilute the analysis sample to a level that essentially eliminates interfering matrix effects. In addition, we apply stable isotope labeled internal standards for the most important analytes (THC, CBD) in cases were the highest degree of measurement accuracy is required, to ensure that our results are not obscured by matrix effects.”

  4. Rod
    Using the testing protocol you are suggesting would essentially legalize high thc cannabis nationally. No freshly harvested high thc cannabis will have a delta 9 thc concentration higher than .3. Using that methodology all the adult use high thc cannabis would technically legally be hemp. I believe the only way to use hplc and be similarly reliable would be to mathematically convert the thca to thc for a max or total theoretical thc value. I understand what you would like to do but it doesn’t pass the straight face test. If we could collectively pull it off in a state plan approved by USDA then all cannabis is hemp.

    1. Thanks for your comment. While I appreciate your perspective, I’m simply following the statute. If Congress had intended that “total THC” be the measure it would have said so. Instead, it said delta-9 THC.

  5. This article is very well written! I have mentioned it on a few blogs of my own.

    A lot of hemp flower sellers, selling the bud intended for smoking, are gravely disregarding THCA, in my opinion. This article helps explain well!

    1. Thanks, Robert. Hemp testing for delta-9 THC concentrations is a major issue. I appreciate you sharing this article.

  6. As the THC is the artificial product of heat, it does not exist originally in the plant. If THCA is not counted, most, if not all, marijuana dried in low temperature will contain < 0.3% THC with a high content of THCA. Then the definition of hemp of < 0.3% THC makes no sense.
    For my understanding, "using post-decarboxylation or other similarly reliable methods" means the method should be post-decarboxylation or similar to post-decarboxylation. No matter the method is GC, HPLC or others, the key is similar to post-decarboxylation.

    1. Jaingnan- Thank you for your comments. Delta-9 tetrahydrocannabinol does, in fact, exist natively in the plant. However, as you mention, it is usually in relatively low concentrations. The issue, though, is not biology; rather, it is law. The statute expressly states that “delta-9 THC” is the operative compound to measure in order to distinguish between legal hemp and illegal marijuana. Congress easily could have included THCA, but it did not. With respect to the post-decarboxylation method that you reference, I have written the following article on that point which I encourage you to read: https://cannabusiness.law/what-is-total-thc-and-does-it-matter/

  7. I appreciated your article. Are you aware of any testing protocols for determining the amount of ‘CBD oil in a honey stick or agave stick? Everything I find seems to deal with initially extracting CBD oil from hemp.

    1. Thanks, Brook. Unfortunately, I do not have the testing protocols for testing CBD oil in a honey or agave stick matrix.

Leave a Reply to Rod Kight Cancel reply

Your email address will not be published. Required fields are marked *