Lawrence Livermore National Laboratory

Matthias Frank


The goal of this project was to demonstrate the feasibility of using exhaled-breath analysis in two important application areas that are largely unexplored: detection of drugs in biomedicine and detection of biological agents in biosecurity. We investigated the potential use of exhaled-breath trace-gas analysis for screening patients for drug abuse, prescription-drug compliance, and variations in individual drug uptake. In addition, we evaluated its use for rapid screening of individuals for recent exposure to biological weapons during assessments and triage after a terrorist attack.

Background and Research Objectives

In biomedicine and biosecurity, there exists a need for new capabilities to quickly and reliably assess individuals for possible disease states and previous exposures to biological agents and illicit materials without using invasive procedures. Exhaled-breath analysis is very promising in this context because the breath (called a "window into the blood”) allows analyses for a wide range of conditions and exposures that traditionally require blood draws or other invasive procedures. Breath analysis is also hard to fool, hence its use to check if someone is driving under the influence of alcohol. Human exhaled breath is very complex, containing many trace gases, including volatile organic compounds (VOCs) in the vapor phase and nonvolatile trace compounds (e.g., proteins) embedded in exhaled aerosol droplets. Both trace gases and non-volatile compounds found in breath hold clues to the potential disease state of an individual (Kim et al. 2012, Lawal et al. 2017). Also, proliferating bacteria emit trace gases (i.e., the smell of bacterial cultures) that result from metabolic processes and are a tell-tale sign of the bacteria's presence. In some cases, these gases are specific to certain bacterial species (Audrain et al. 2015, Rees et al. 2018). This may enable detection and identification of proliferating bacteria present in the lungs of a person who is suffering from an infectious respiratory disease or who has inhaled a biological agent. In the near future, breath analysis will be performed in real-time with small point-of-care or handheld instruments that integrate efficient breath sampling with miniaturized trace-gas analyzers, electronic noses, or micro-fluidic assays for a wide range of applications.

Impact on Mission

By supporting the development of counterterrorism capabilities to detect and promptly respond to biological attacks, our research advances the NNSA goal of preparing for broad national security challenges. Our work addresses the Laboratory’s research-and-development challenges in chemical and biological countermeasures, and also supports the Laboratory’s core competency in bioscience and bioengineering. This research is relevant to the Laboratory’s missions in basic science and biosecurity; it could also have broad impacts for a large range of applications in biomedicine, biosecurity, and counterterrorism.


Our overall approach included designing and performing demonstration experiments, as well as finding and characterizing molecular signatures detectable in breath that could be used for a large range of applications in biomedicine, biosecurity, and counterterrorism.

For breath analysis for drugs detection, we participated in an ongoing UC Davis study on breath analysis of chronic pain patients. As a result of that study, we developed a combined breath vapor/breath condensate sampler from which we obtained and analyzed breath samples (both vapor and condensate) from volunteers including patients taking prescribed opioid-based pain medications and from healthy volunteers (as controls). We collected and analyzed breath samples (both breath vapor and breath condensate collected in parallel) from several patients who were being treated with opioid-based pain medications. In some cases, the sample analysis revealed the presence of the medication and known metabolic products in the breath condensate, but not in the breath-vapor sample. This may be due to the relatively low volatility of these substances (due to their large molecular mass). Therefore, while these compounds may be detected in the liquid-breath condensate, their abundance in the gas phase may be below the detection limit of our current analytical method. Insights from our work resulted in several leads for how to improve study design, sampling, and analytical techniques for similar work in the future.

We determined that our analytical procedure for rapidly and non-invasively determining exposure levels of individuals could be useful for patient triage after a bioterrorist attack. Bacterial agents that may be used for bioterrorism are known to emit characteristic trace gases during their proliferation. Exposure to a biological weapon agent may be detectable in an individual’s breath shortly after inhalation of the agent (as the agent begins to proliferate and interact with human cells in the lungs) long before the individual shows any symptoms. For this and other biosecurity applications, we set up and conducted bacterial sampling of volatile organic compounds (VOCs) emitted by proliferating bacteria from strains of Francisella sp., Bacillus sp., and Yersinia sp. bacteria that are among the Centers for Disease Control’s Category A agents considered to be dangerous biological weapons agents. Initially, we focused on strains of Franciscella tularensis, collecting volatile organic compounds emitted by the proliferating bacteria at multiple time points, then analyzed the samples with a gas chromatography-mass spectrometry system, and identified a number of compounds that were clearly produced by the bacteria.

The methods developed by and results obtained from this work are relevant to several other application areas, such as breath analysis for other biomedical applications (e.g., systemic diseases, infectious diseases, lung injuries, and traumatic brain injuries), as well as analyses conducted for metabolic studies, scientific studies of selected bioterrorism agents, and as a tool to study the health of algal cultures for biofuels. Results from this project have attracted the attention of several external sponsors who may provide support for continued research in some of the above subject areas.


Audrain, B., et al. 2015. "Role of Bacterial Volatile Compounds in Bacterial Biology. FEMS Microbiology Reviews 39(2): 222–233. doi: 10.1093/femsre/fuu013.

Kim, K.-H., et al. 2012. "A Review of Breath Analysis for Diagnosis of Human Health." TrAC Trends in Analytical Chemistry 33(1-8). doi: 10.1016/j.trac.2011.09.013.

Lawal, S., et al. 2017. "Exhaled Breath Analysis: A Review of 'Breath-Taking' Methods for Off-Line Analysis." Metabolomics 13(10):110. doi: 10.1007/s11306-017-1241-8.

Rees, C., et al. 2018. "Comprehensive Volatile Metabolic Fingerprinting of Bacterial and Fungal Pathogen Groups." Journal of Breath Research 12(2). doi: 10.1088/1752-7163/aa8f7f.

Publications and Presentations

Reese, K. L. 2018. “Biosecurity Advances: Characterization and Diagnostics.” Pittcon Conference, Orlando, FL, 2018. LLNL-PRES-746898.

Reese, K. L., et al. 2018. “Characterization of VOCs Emitted from Pathogenic Bacteria Using SPME-GC-MS: Towards Non-Invasive Breath Diagnostics for Biosecurity.” Pittcon Conference, Orlando, FL, Feb. 2018. LLNL-ABS-736588.

Reese, K. L., et al. 2018. “Characterization of VOCs Emitted from Pathogenic Bacteria Using SPME-GC-MS: Towards Non-Invasive Breath Diagnostics for Biosecurity.” International Breath Analysis Research (IABR) Conference, Maastricht, Netherlands, June 2018. LLNL-ABS-749437.

Reese, K. L., et al. 2018.“Volatile Microbial Metabolome: Strategies for SPME-GC-MS Sampling and Data Analysis.” American Society of Mass Spectrometry (ASMS) Conference, San Diego, CA, 2018. LLNL-ABS-745423.