New Colorado Regulations for the Measurement of Heavy Metals in Cannabis Vaping Aerosols, A Realistic and Practical Assessment: Part 3

Part 1 of this series on the new Colorado regulations for heavy metals in cannabis vaping aerosols examined the fundamentals of vaping and the process of converting a vape liquid into an aerosol and the difficulties associated with characterizing the metal content in a hydrophobic liquid. Part 2 focused on the challenges of trapping and collecting the aerosol without contaminating the sample, and how best to validate the process using standard methods developed by the tobacco industry for electronic nicotine delivery systems (ENDS).

The final part of the series will look at the ICP-MS measurement technique and the many potential sources of interferences observed when determining the most common metals found in aerosols generated from organic-rich vaping liquids. The series will offer some final thoughts as to how the data could possibly be used to regulate cannabis vaping devices. Finally, I will wrap up with some suggestions to help the industry better scrutinize the quality of vaping devices before they even enter the marketplace.

Challenges in measuring some elements by ICP mass spectrometry 

So up to now, this series of articles has focused on the problem areas associated with sampling a representative aerosol from the liquid in the tank of the vaping system and how best to ensure that no contamination is occurring from the vapor testing machine. Let’s now take a closer look at some of the major analytical challenges presented by measuring a multielement panel of analytes in vaping aerosols, containing high levels of carbon from the cannabinoid extract and organic diluents used. Note: For an educational primer on the major interferences in ICP-MS please refer to the following references ¹.

The big four

Of the traditional big four heavy metals, arsenic and mercury are considered the most problematic for measurement by ICP-MS, while lead and cadmium are relatively straight forward to determine, because they are easily ionized, have very low detection limits, and, there are multiple isotopes available for quantitation with very few interferences. However, it should be emphasized that lead is by far the most likely to be present in vaping aerosols because of the possibility of leaded-solder used in the battery connectors or the use of lead in some brass components to make them easier to machine ². 

So let’s first take a closer look at arsenic. It is monoisotopic at mass 75 amu, which suffers from a major interference from the ⁴⁰Ar³⁵Cl polyatomic ion at mass 75. So if there are any chloride (Cl^-) ions in the sample, detection at low levels becomes difficult. There are ways to reduce this interference using collision/reaction cell (CRC) technology by either utilizing a collision gas (helium) to reduce the ⁴⁰Ar³⁵Cl interference with kinetic energy discrimination or a reaction gas (such as oxygen) to mass shift ⁷⁵As to ⁷⁵A s¹⁶O at mass 91 for quantitation. 

With regard to mercury, it’s important to point out that when quantifying trace levels, the addition of approximately 1-2 ppm of gold (in the form of the chloride) may be necessary to stabilize the mercury to stop it from being reduced to elemental mercury, which can be absorbed into the walls of the sample container or lost when the cap is removed. The excess gold competes with mercury for reducing substances in the samples, thus maintaining the mercury in oxidized mercuric ions which remain in solution. This is compounded if the solutions are left for extended periods of time before they are analyzed ³. Having hydrochloric acid in the sample matrix is also beneficial to ensure there is an excess of chloride ions for coordination of mercury and moves the form of mercury away from the easily reducible mercuric ion to multichloro mercury complexes.

Different metallic components

As mentioned in Part 1 of the series, if kanthal, brass, nichrome, stainless steel, soldered connectors or ceramic materials are used in the vaping system, it would seem appropriate to include aluminum, iron, chromium, copper, zinc, nickel, lead, tin and silicon as analytes, many of which are extremely challenging to determine in organic matrices by ICP-MS. In particular elements like aluminum, silicon and zinc are so environmentally ubiquitous that they are analytically challenging even using commercially-available ultrapure acids and high purity calibration standards. In addition, it is also very difficult to get clean blanks, which can severely compromise the detection capability. Another thing to consider is that an ICP-MS torch is made from quartz, which makes it very difficult to achieve a low silicon blank. Silicon also forms silicon oxide in the plasma that produces a refractory coating on sampler and skimmer cones. The buildup of the refractory silicon oxide on the cones causes a gradually increasing silicon background. In addition, the major isotope of silicon is at mass 28 amu (92% abundant), where there is massive interference from the nitrogen dimer (¹⁴N¹⁴N) at mass 28. Mass 30, only 3.1% isotopic abundance, has a similar interference from ¹⁴N¹⁶O. Mass 29 could be used, but it is only 4.7% abundant, so detection capability would be severely compromised compared to mass 28. One way around this would be to use “triple or multi quadrupole” ICP-MS technology using oxygen as a reaction gas to mass shift ²⁸Si to ²⁸Si¹⁶O at mass 44 for quantitation. But this technology is relatively new and would be challenging for analysts with very little experience in working with the ICP-MS technique. 

Iron, though not as ubiquitous as aluminum and silicon, can also be a problematic analyte unless utmost care is taken to avoid contamination. Additionally, the major isotope of Fe is at mass 56, where there is a major interference from ⁴⁰Ar¹⁶O at mass 56. This interference can either be minimized using helium in a collision cell with kinetic energy discrimination or removed with ammonia (NH₃) using reaction chemistry. For this reason, iron may also be an element that could be a problematic choice for routine purposes by non-experienced laboratories. Chromium and nickel are not as environmentally problematic as iron, but analysts should remove any external stainless steel ejectors from pipettes or autosampler probes prior to using them with acid solutions, or use pipettes that do not have steel pipette tip ejectors for trace metal analyses to prevent contamination issues. In addition, chromium and nickel are not straight forward to determine by ICP-MS. In particular chromium’s major isotope is at mass 52 (84% abundant), where there is a massive polyatomic interference from ⁴⁰Ar¹²C produced by the argon gas and the high level of organic compounds in the sample (cannabinoids and diluents such as propylene glycol, glycerine and/or MCT oils). A less abundant isotope could be used at either mass 53 or 54, but they are much less sensitive. A collision cell using helium will only have limited success in reducing the interference because of the organic matrices in vaping liquids. For that reason, measuring ultra-trace levels of chromium in an organic matrix might require a triple/multi quadrupole technology using oxygen as a reaction gas to mass shift the ⁵²Cr and use ⁵²Cr¹⁶O at mass 68 for quantitation. Note: Nickel is not as difficult as chromium, but the ⁴⁰Ar¹⁸O polyatomic could potentially pose some problems for the major isotope of nickel at mass 58, especially when measuring trace levels.

Finally, tin presents some unique challenges in organic solvents, particularly when a heated spray chamber is being used. Because of the high volatility of Sn (IV) in the presence of hydrochloric acid, it will be lost as a volatile tetrachloro complex before it enters the plasma and not be accurately measured. However, this can be compensated by adding a few mLs of 1% hydrofluoric acid to the sample, which converts the Sn (IV) to SnF₄, which is less volatile. However, if laboratories are not well trained in the safe handling of hydrofluoric acid, then it should not be used, and either tin should be eliminated from the list of analytes or a Peltier cooled spray chamber should be used in place of the higher sensitivity desolvating sample introduction system.

These are just a few examples of the many analytical challenges in characterizing ECDS aerosols for elemental contaminants by ICP-MS. Without question, carrying out quantitation of low-level impurities requires knowledge and experience of working in the ultra-trace environment, particularly when the sample matrix is organic in nature, which requires specialized sample introduction equipment. These are not insurmountable hurdles to overcome, but in light of the many novice users in cannabis testing labs who are not well versed in the complexities of the technique, these issues can easily lead to errors and inaccuracies that could result in false positive and false negative results.

Final thoughts and future direction

It is hoped that this three-part series of articles has provided a realistic assessment of the new Colorado regulations for the measurement of heavy metals in cannabis vaping aerosols. Its main objective is to help the analytical testing community who are tasked with carrying out this analysis to avoid pitfalls and improve analytical practices. The information is by no means exhaustive, but it is meant to give the reader an overview of what additional equipment is required and how best to optimize its operation to trap, collect and sample elemental contaminants from this complex sample matrix. In addition, it has offered guidance on how best to approach the quantitation of this panel of analytes with a more complete understanding of the strengths and weaknesses of ICP-MS for this type of sample matrix. 

However, without knowing what an “average” vaper typically consumes per day or per week, it is very difficult to correlate the weight of heavy metals per inhalation (ng/number of puff) with the current regulations for inhaled cannabis products (µg/g). However an approximation can be made, using the weight of the liquid in a typical vaping cart, the number of inhalations (puffs) of an average user, and historical heavy metal data derived from electronic nicotine delivery (END) systems. So let’s take lead (Pb) as an example. A typical vaping cartridge contains approx. 0.5-1 g of liquid, which is equivalent to 150-250 puffs depending on the user ⁴. Based on evidence in a study by Halstead and coworkers at the CDC, who characterized 17 different END system for a suite of heavy metals, they found that for the majority of the systems Pb varied between 0.050 ng/10 puffs (Limit of Detection) and 3.3 ng/10 puffs, with a mean of 1.3 ng/10 puffs  (with one flyer of 11.4 ng/10 puffs) ⁵. If we take the average number of puffs in a vaping device containing 1g of liquid to be 200, this correlates to 26 ng/g of Pb in one cartridge or 0.026 µg/g. So based on most states’ inhalation maximum contaminant limits (MCL) for Pb to be 0.5 µg/g, an assessment could be made that a user would need to go through about 20 vapes or vape refills before they reached the MCL. Of course, the timeframe would depend on what is considered a normal frequency of vaping…..someone who uses a vape refill every day would reach this in 20 days, while someone who’s vaping frequency is 1 per week, it would take them 20 weeks. Moreover, irrespective of whether it’s valid to make a comparison of the classic big 4 heavy metals in this way, it would be difficult to do it for the other metals found in vaping devices, because there are no state limits to compare them with. For that reason, perhaps there needs to be separate regulations specifically for vaping devices, because it is not immediately obvious how the comparison can be made for all metals, based on the current state-based regulatory framework.  It’s also important to emphasize that there were large variations in the data generated by the Halstead study, which is confirmation of the wide variability in the type of components and materials used in different vaping technology for both nicotine and cannabinoid delivery systems ⁵.

Unfortunately, we are trying to regulate cannabis vaping devices with very few standards or validated methods for guidance, although there is some preliminary work being carried out by the ISO standards group, as mentioned in Part 2 of the series ⁶. It would therefore be far better to scrutinize their quality before they enter the supply chain by carrying out basic testing procedures to ensure some kind of contaminant assessment. This would save a great deal of time and effort in trying to regulate products that in many cases are flawed when they reach the consumer. One approach is to use standardized toxicity leaching procedures that reproduce the potentially corrosive environment of liquids inside the vaping device over time. This can be something like MCT (medium chain triglycerides) oil which is commonly used as a diluent or weak acids like acetic or citric acid which simulates an approximate pH of the liquid ⁷. The vaping device components are typically wetted, by filling with the appropriate liquid and left for a fixed period of time (up to 24 hours) before they are analyzed for the selected suite of elements by ICP-MS. It would at least be a qualitative indicator of whether heavy metal contaminants are being leached from inside the device, which could also help to get a better understanding of their shelf life. This is particularly relevant for lead, because it is extremely toxic and is the only one of the classic big four heavy metals that we know is used in the manufacture of some vaping components. which could potentially pose a serious long term health hazard for users of vaping systems ⁸.  Ambitious perhaps, but we have to find a way to eliminate these “dubious” and often illicit products from finding their way into the market place. It might even encourage the industry to design vaping devices with more inert materials that are unlikely to corrode or even develop systems that are free of metal components.

In the meantime, it will be interesting to see how Colorado “figures out” these new regulations, because other states are watching closely and possibly looking to add this to their list of regulations in the near future. There might be more clarity over the next twelve months, but as there are currently no standard testing procedures for cannabis vaping devices, it will be interesting to see how the industry responds.

Further reading

1. Practical Guide to ICP-MS: A Tutorial for Beginners, Third Edition, R. J. Thomas, CRC Press, Boca Raton, FL, ISBN 9781466555433, (2014), https://www.routledge.com/Practical-Guide-to-ICP-MS-A-Tutorial-for-Beginners-Third-Edition/Thomas/p/book/9781466555433 
2. Heavy Metal Contamination From Metal Vape Cartridges, J. McKeil, Rx Leaf, May 20, 2019, https://www.rxleaf.com/heavy-metal-contamination-from-metal-vape-cartridges/ 
3. Mercury Preservation Techniques, Environmental Protection Agency  (EPA), March 2003, https://www.inorganicventures.com/pub/media/wysiwyg/files/mercury_preservation_techniques.pdf 
4. CBD Dosages for Vape, Oil, Tincture, and More, Vaping 360 Website: https://vaping360.com/learn/cbd-dosages/ 
5. Analysis of Toxic Metals in Electronic Cigarette Aerosols Using a Novel Trap Design, N. Gray et.al., Journal of Analytical Toxicology, 2020;44:149–155
6. ISO Method, 20768:2018, Vapor Products - Routine Analytical Vaping Machine - Definitions and Standard Conditions, https://webstore.ansi.org/Standards/ISO/ISO207682018?gclid=EAIaIQobChMI2r70-6Ls7AIVHOy1Ch329wcKEAAYASAAEgIQA_D_BwE 
7. Heavy Metals Contamination: Is Cannabis Packaging to Blame? R. Newman, Analytical Cannabis, February 20, 2020, https://www.analyticalcannabis.com/articles/heavy-metals-contamination-is-cannabis-packaging-to-blame-312246 
8. Disposable Cannabis Vape Cartridges May Pose Latent Lead Exposure Risk, A. Cheadle, Analytical Cannabis, August, 27, 2020, https://www.analyticalcannabis.com/news/disposable-cannabis-vape-cartridges-may-pose-latent-lead-exposure-risk-study-finds-312620 

Note: A more comprehensive discussion of this topic can be found in the author’s recently published reference book, Measuring Heavy Metal Contaminants in Cannabis and Hemp, published by CRC Press. More information about the book can be found at the following link:

https://www.routledge.com/Measuring-Heavy-Metal-Contaminants-in-Cannabis-and-Hemp/Thomas/p/book/9780367417376 

In addition, the author of this article organized a 1-day educational workshop on measuring heavy metals in cannabis by ICP-MS last year (2019) in conjunction with the Maryland Medical Cannabis Commission (MMCC) where Dr. Steve Pappas, the leader of the Tobacco Inorganics Group at the CDC gave a talk on the challenges of measuring elemental contaminants  in vaping delivery systems. The entire workshop was videoed including Dr. Pappas’s talk, which can be found at the link below:

https://www.analyticalcannabis.com/articles/workshop-on-heavy-metals-in-cannabis-by-icp-ms-now-available-to-view-312735 

For more information about the importance of testing cannabis and cannabinoid products for elemental contaminants, please refer to the 5-part series the author recently published in this magazine. Part 1 can be found here, which includes links to the remainder of the series:

https://www.analyticalcannabis.com/articles/regulating-heavy-metals-in-cannabis-part-i-what-can-be-learned-from-the-pharmaceutical-industry-312336