NIST researcher Tara Lovestead is working toward a way to determine the chemical signature of marijuana intoxication. Photo: J. Burruss/NIST

Editor's note: This is part two of a two-article series on breathomics, as published in the February print issue of Laboratory Equipment.


Researchers and doctors have long investigated the connection between a person’s state of being and their exhaled breath. In the 1800s, Francis Anstie observed that small amounts of alcohol were excreted in breath—long before Robert Frank Borkenstein used chemical oxidation and photometry to power his breath analyzer for alcohol concentrations, which would become law enforcement’s standard.

Fast forward to 2017 and breath research—or breathomics, as its now called—is just as important as ever. The field is a combination of breath and metabolomics, the study of unique chemical fingerprints that specific cellular processes leave behind. 

Once relegated to the lab due to the mass spectrometric techniques that traditionally power breathomic tests, the field is now emerging for some fresh air, with researchers increasingly developing portable breathomic devices for everything from diabetes and cancer to marijuana intoxication.

“The field is already making headway in early diagnosis of cancer and other medical applications,” Tara Lovestead, a NIST researcher working in the field, told Laboratory Equipment. “The potential for testing for chronic disease, inflammation and cancer is very tangible. As applied to law enforcement, a breath signature for intoxication could eliminate the need to know the actual avenue for becoming intoxicated.”

Breath test for marijuana 

It’s relatively easy to single out an individual that is driving under the influence of alcohol. Swerving between lanes, running red lights and inconsistent speed are signs easily identifiable to passersby and police alike. Once suspicions are raised (or even before that), police can rely on the accuracy of their portable breathalyzer to identify the exact amount of blood alcohol content in said driver—providing unequivocal confirmation of the presence of alcohol.

But how do police spot a driver who is driving “high,” or under the influence of marijuana? He or she may be showing signs similar to a drunk driver, but not necessarily since the effects of marijuana in large doses dramatically differs from the effects of alcohol.

Each state has its own laws regarding the blood alcohol level of drivers in terms of what is considered legal and what is considered legally drunk, or under the influence. And now, with more states legalizing both medical and recreational marijuana, legal limits are being set in this arena as well.
However, they are not as clear as drunk driving laws—which is more a function of the vice than the law itself.

Unlike alcohol, THC—the active compound in marijuana that leads to a high—can stay in a user’s blood, saliva or urine for minutes, days, hours or even months depending on what strain of marijuana is used, how often, and the specific method of ingestion.

“There are just so many questions we need to address, and so much we don’t know,” said Lovestead, a chemical engineer at NIST. “The biggest issue for law enforcement is Δ9-THC in the blood does not correspond to intoxication.”

At the Forensics@NIST conference in November, Lovestead gave a presentation describing her work to identify other chemical markers indicative of marijuana intoxication. She is focused on creating noninvasive, portable breath tests for Δ9-THC that can indicate recent marijuana usage from 30 minutes to 2 hours prior—the only real way to determine if a user is driving under the influence of marijuana. 

Lovestead said her team’s approach incorporates three key areas: fundamental data; materials development; and breatholomics (a subset of breathomics). 

Chemistry-wise, there is still a lot researchers do not know about cannabis, THC and other cannabinoids. So, part of Lovestead’s research is to research. She and her team are measuring vapor pressure, molecular interactions and partition co-efficients to successfully apply it to overall cannabis research.

They are also researching and/or developing new materials to outfit a potential marijuana breathalyzer device, especially in terms of what is best suited to crucial absorption and desorption techniques.

Lovestead is also paving a way toward determining the chemical signature of marijuana intoxication.

To do so, she has thus far relied on a dynamic headspace sampling technique called porous layer open tubular (PLOT)-cryoadsorption that has provided extremely sensitive quantitative recovery of Δ9-THC. The method highly decreases the amount of time expended to identify vapor pressure information. Most importantly, Lovestead and her team have adapted the technology to create a portable version that could ultimately be used at the roadside. 

For breath collection, Lovestead relies on capillary microextraction of volatiles. This method is already ready for in-the-field sampling, and is most suited for the breath collection of cannabis-related metabolites, which indicate if a user actually smoked marijuana or just ingested it secondhand.

“[Fundamental data, materials development and breatholomics] build upon one other and help the other out,” Lovestead said about her lab’s three-fold approach to a marijuana breath test. “In the future, we are going to look at more artificial breath work with the different materials available.”

Hound Labs’ Hound device is the world’s first marijuana breathalyzer to measure recent THC in the breath of drivers at the roadside. Photo: Triple Ring Technologies/Hound Labs

Marijuana intoxication test

Measuring THC in breath requires a scientific method more than a million times more sensitive than those used to measure alcohol. And since a lab’s LCMS can’t leave the building, finding a portable solution is imperative to developing a roadside alternative. 

Hound Labs, Inc. (Oakland, Calif.) is the first company to detect and measure THC in breath. Using only one or two breaths, the patent-pending approach detects THC and measures levels to well below 500 picograms, or parts-per-trillion (ppt). The company’s scientific breakthrough has been optimized to create a prototype of the first dual alcohol and marijuana breathalyzer for use by law enforcement at the roadside.

The device comprises two small pieces: a handheld unit that looks like a smaller version of an alcohol breathalyzer, as well as a docking station.

Once a breath sample is captured, the device performs a series of automated routines that first chemically isolate THC molecules from the breath sample, and then generate a composition for measurement—all of which happens inside a single-use, disposable cartridge. 

The automated process then triggers a measurement mechanism to optically measure the THC level in the generated composition, outputting a measurement value in parts per trillion. The resulting measurement value corresponds to the level of THC present in the sample at the time of breath sample collection.

“We feel that chemistry is truly the only way you can detect and measure [THC] at the level needed to be effective,” Hound Labs co-founder and CEO Dr. Mike Lynn told Laboratory Equipment. “Other technologies people are trying to use, such as ion mobility, was an obvious place to try because it’s been used in airport explosion detection machines. When I first started the company, I said ‘maybe we can take ion mobility and shrink it down,’ but because of the physics of it, you lose the ability to be sensitive and specific enough. Only chemistry offers the preciseness needed to tag just THC in a breath sample, and nothing else.”

Hound Labs has been testing results from its device against laboratory standards from day one, as well as conducting field trials with various police departments. The company has worked with UC Berkeley, UC San Francisco (UCSF) and Stanford University. In fact, a clinical trial with UCSF has already been approved for early this year. According to Lynn, the trial will comprise about 20 human subjects. They will smoke marijuana and have their breath tested at regular intervals. The company will also be conducting track tests. The drivers will smoke marijuana then drive through certain obstacles (at a safe location). 

Lynn and his team specifically designed the cartridge in the Hound device so that it could be retained and extracted at a later date for analysis on an LCMS. 

“Once we get out there and start testing people, whether it’s at a roadside stop by law enforcement or employers, they are going to have to correlate our data with the actual gold standard, and that’s going to require laboratory testing, so we knew it was important that samples could be retested,” Lynn said. “I think there is going to be an explosion in tests that are going to require laboratory instruments and mass spectrometers.”

Data collected from Hound Lab trials can also be used to begin correlating the measured value or ppt-level to actual driving impairment. 

“It’s exactly how the alcohol standards were developed,” Lynn said. “Because we measure and have the objective data to share, we will be really instrumental in determining what levels are actually considered impairment.”

The most important feedback Hound Labs received from law enforcement is that the device needs be rugged enough to meet the needs of a demanding environment—which the company has already worked into the latest iteration of the unit. Field trials also confirmed that current saliva, urine and blood tests are not practical for law enforcement at the roadside, nor do they provide accurate information on current impairment. 

“The officer needs to know if a person behind the wheel is actually stoned, or if their car just stinks like marijuana but they haven’t smoked since yesterday,” Lynn said. “For the officer, it is critical we show only people who have smoked in the last couple hours. They can use that as objective data in conjunction with their regular driving observations and field sobriety tests—all the other things they have to do anyway.”

Hound Labs will be testing its new ruggedized version at the roadside in the next quarter, before the company incorporates feedback into final changes for the device. Lynn anticipates manufacturing the Hound device in the latter part of 2017.