Executive Summary

Toxicology of Diacetyl and 2,3-pentanedione

August 3rd 2014

Diacetyl and 2,3-pentanedione: A critical analysis of the literature as it pertains to electronic nicotine
delivery systems and personal vaporizers.

Abstract: A critical review of the most recent toxicological and epidemiological information available for
Diacetyl (DA) and 2,3-pentanedione (PD) was undertaken in an attempt to quantify the associated risk
level of incorporating DA and/or PD into electronic nicotine delivery devices (e-cigs) and personal
vaporizers (vaporizers). Data gathered for recommended exposure levels (REL) determined by the
National Institute of Occupational Safety and Health (NIOSH) and occupational exposure levels (OEL)
determined by the European Commission’s Scientific Committee on Occupational Exposure Limits for
diacetyl (SCOEL) were critically evaluated. Additional literature review on the associated risks of DA and
PD from cigarette smoking and correlations to exposure from e-cigs and vaporizers were included. Based
on the current understanding of the risks associated with DA and PD inhalation, as well as the inhalation
characteristics of humans utilizing e-cig and/or vaporizer devices, it is not recommended to include DA
or PD into the flavoring solution of e-cig and/or vaporizer devices. This recommendation is carried
further to include volatile organic carbonyls as a chemical class in general until a more thorough review
of the individual components can be ascertained.
Introduction:
Diacetyl is a natural flavoring substance found in a number of food products; e.g. beer, butter,
coffee and others. Diacetyl is also a metabolite of acetaldehyde in mammals. Synthetic diacetyl is also
used as an artificial flavoring in a wide range of snack foods (microwave popcorn, potato/corn chips,
processed cheese, sour cream, salad dressing, icing, sauces, marinades and other processed foodstuffs).
[1, 2] Diacetyl imparts a creamy, buttery flavor to these products and is included on the FDA’s generally
regarded as safe (GRAS) list of food ingredients for ingestion.
In 2000 NIOSH evaluated 135 workers at a microwave popcorn production plant. Air sampling of
the microwave production areas of the facility was performed and a qualitative evaluation of volatile
organic compounds (VOCs) found over 100 different compounds present in the air of the facility. The
predominant compounds identified included the ketones diacetyl, methyl ethyl ketone, acetoin, 2nonanone and acetic acid. Diacetyl was identified as the predominant ketone in the facility. NIOSH
reported that these workers had 2.6 times the expected rate of chronic cough and shortness of breath,
when compared to national data and twice the expected rate of physician diagnosed asthma and
chronic bronchitis. Overall the workers had 3.3 times the expected rate of airway obstruction, with
“never-smokers” having 10.8 times the expected rate. Further investigations by NIOSH showed that
workers who were employed longer (i.e. longer exposure) and those exposed to higher concentrations
of the flavoring agents (near isolated tanks of oils and flavorings), had a higher frequency of airway
obstruction and respiratory symptoms. These NIOSH investigations included cases where workers mixing
butter flavors with heated oil were diagnosed with bronchiolitis obliterans. [2] References [2] and [3]
describe additional studies performed in the workplace since 2000.

1

 Executive Summary

Toxicology of Diacetyl and 2,3-pentanedione

August 3rd 2014

Bronchiolitis obliterans (also named constrictive bronchiolitis or obliterative bronchiolitis) is a
condition characterized by the irreversible fixed airways obstruction caused by the narrowing of the
bronchiolar lumen by submucosal fibrosis or fibrous tissue in the adventia or adjacent alveolar septa.
Essentially bronchiolitis obliterans is a deep lung disease that significantly restricts breathing due
to the formation of fibrous tissue around the alveoli. The symptoms can be described from mild to
severe and in some cases it has proven fatal. It is a rare, permanent disease and as such can be
potentially mis-diagnosed by physicians as, bronchitis, asthma or pneumonia. In this regard, it is
important to consider the possibility of some number of workers exposed to diacetyl and diagnosed
with airway obstruction of some form (emphysema, COPD, asthma etc.) could in fact have bronchiolitis
obliterans. [2]
In 2011, based on the epidemiology and available toxicology data, NIOSH provided a draft
document on exposure to diacetyl and 2,3-pentanedione. NIOSH proposed a 5 ppb recommended
exposure level (REL) based on an 8 hour time weighted average. And a 25 ppb short-term exposure level
(STEL) for fifteen minute acute exposures. [3] That same year, the National Toxicology Program (NTP)
published its most recent findings from animal inhalation studies of diacetyl and based on the data
collected arrived at a 60 ppb REL. [4] NIOSH reviewed the NTP data and arrived at the conclusion that
since the animal data was within an order of magnitude of the epidemiology data and due to the
severity of risk associated with the inhalation of diacetyl, the original determination of the REL and STEL
of 5ppb and 25 ppb respectively for diacetyl exposure was appropriate. [5]
OSHA has yet to define an occupational exposure level (OEL) for diacetyl, but a thorough review
of the available toxicology and epidemiology data was performed by the European Commission, which
published in 2014 an OEL of 20 ppb and a STEL of 100 ppb. [2]
2,3,-pentanedione is currently being used as an alternative to diacetyl in the food and flavoring
industry as well as other compounds that provide the buttery flavor characteristics in food. [6] However
quantitative risk assessments described by NIOSH, determined that preliminary evidence from animal
based inhalation toxicological studies suggest that the relative toxic potency of 2,3-pentanedione may
be of equal or greater magnitude than diacetyl with regard to respiratory disease. [5] Further discussion
within this document relating to diacetyl specifically, should also be interpreted to include 2,3pentanedione.

Discussion:
OELs and RELs: What they mean and how they correlate to e-cigs/vaporizers
From a toxicology standpoint, an OEL or REL is based on an 8 hour time weighted average, 5
days a week for a total of 40 hours of exposure. This means that while exposure levels can spike to
above 5 ppb (but not above 25 ppb), there must be periods throughout the day where exposure is less
than 5 ppb such that, over the course of 8 hours, the average exposure is 5 ppb or less.
2

 Executive Summary

Toxicology of Diacetyl and 2,3-pentanedione

August 3rd 2014

One could argue that for an e-cig/vaporizer device, the consumer will not be exposed
continuously over the course of 8 hours and therefore exposures of up to 25 ppb over the course of
fifteen minutes may be acceptable. This argument is flawed for several reasons as described in the
following paragraphs.
Time frames for OELs and RELs were developed to determine the potential risk of employees to
toxic substances in the workplace. What is embedded within the OEL or REL, but may not be readily
apparent, is clearance time associated upon removal of the exposure. Clearance times are essential to
the understanding of OELs and RELs because they essentially provide 1/3 of a day to exposure
considerations and 2/3 of the day for the person to clear the toxin from the body through a variety of
mechanisms including clean air exchange within the lungs (exhalation) or metabolism of the substance
with eventual excretion. There are also an additional two days a week (the weekend for instance),
where it is assumed the exposure to the toxin is absent, thus providing additional clearance time prior to
re-exposure.
It becomes quickly apparent that the limited exposure times and extended clearance times
assumed within the OEL and REL do not readily translate to the uncontrolled use and potential abuse by
the public of e-cigs and/or vaporizers. In fact the exposure and clearance times could easily be reversed
with potentially more exposure to the toxic substance for 2/3 of the day and most likely higher
frequency exposures over the course of the weekend. Assuming an e-cig/vaporizer user is also exposed
during the course of their work day (the remaining 1/3 of the day), there is minimal clearance time for
the body to rid itself of the toxic substance.
Another concept of the time weighted average OEL and REL is the fact that they are based on an
average of exposure levels, meaning that while in some instances the exposure may be greater than 5
ppb for a given amount of time, over the course of an 8 hour day there will be periods of time that the
exposure will be less than 5 ppb or even no exposure will occur. While it could be argued that an ecig/vaporizer user would conceivably only be exposed to diacetyl for brief, periodic moments within a
day, given the uncontrolled exposure characteristics of the device and the severity of the risks
associated with diacetyl and 2,3-pentanedione inhalation it is impossible to make that argument with
any degree of conviction.
Due to the uncontrolled exposure capability inherent with using e-cig/vaporizer devices, any
attempts at producing a flavor solution containing DA or PD is not recommended. It should be noted
that cumulative exposure to inhalation toxicants is the most important metric to consider for respiratory
function and as such a consumer may potentially use these products over a significant portion of their
lifetime, greatly increasing their risks for developing airway obstructive diseases/syndromes including
bronchiolitis obliterans due to repeated, chronic exposure to DA or PD.
Confounding Cigarette Smoke Data and Additional Considerations Pertinent to e-cigs/vaporizers
A 2014 review article (Pierce et. Al.) describes the comparatively high yields of DA and PD from
conventional cigarettes versus what workers in the food and flavoring industry are exposed to in an
attempt to moderate the risks and exposure levels put forth by NIOSH concerning exposure to these
3

 Executive Summary

Toxicology of Diacetyl and 2,3-pentanedione

August 3rd 2014

compounds due to the of lack of bronchiolitis obliterans diagnoses in cigarette smokers and the paucity
of epidemiology data associated with PD exposure. [7]
The review article states that the authors are unaware of any cases of bronchiolitis obliterans
being attributed to cigarette smoking, even though DA and PD in cigarette smoke has been shown to be
much higher in concentration than any reported exposures in workplace studies. As noted above,
bronchiolitis obliterans is a rare disease that has the potential to be misdiagnosed by physicians as
asthma, bronchitis, emphysema, COPD and other airway obstruction. Cigarette smoking is strongly
associated with many respiratory diseases including the aforementioned, as well as airway obstruction.
As described in the previously referenced materials, workers exposed to diacetyl in the workplace also
exhibited airway obstruction at multiple times the expected rate based on comparison to national data.
Since the use of e-cig/vaporizer devices could be considered to better correlate with cigarette
smoking than with occupational exposures, and the Pierce article implies that the high concentrations of
DA and PD in cigarette smoke have not produced any documented cases of bronchiolitis obliterans, it is
reasonable to attain a better understanding of the inhalation dynamics for both cigarettes and ecigs/vaporizers. Fortunately, there is an abundance of data on the “puffing” patterns of humans with
many different types of cigarettes and electronic nicotine delivery systems. Kleinstreuer and Feng have
written a very thorough review article, which provides comparative studies concerning the deposition of
toxic aerosols in the lung for conventional cigarettes, less harmful cigarettes (LHC), potentially reduced
risk products (PREPs) and e-cigs. [8]
As described in the Kleinstreuer and Feng review, when it comes to considering how much of
any particular substance or even how many types of substances are inhaled from smoking devices, “less
mass and species being yielded is not equal to less health harm.” There are a number of factors to be
taken into consideration, specifically the problem of compensatory smoking (deeper/longer inhalation)
and an increased consumption of cigarettes (or in the case of e-cigs/vaporizers – increased number of
puffs). It has been shown that LHC smokers developed lung cancer further down into the lung than
conventional cigarette smokers. [9] Considering that on average, puff duration for conventional
cigarettes vs. e-cigs was found to be 2.4 s and 4.3 s respectively it becomes readily apparent that vapors
transferred from an e-cig device will be pulled deeper into the lung than smoke/vapor from
conventional cigarettes. [10] This is in fact supported through Computational Fluid-Particle Dynamics
(CF-PD) simulation models of airflow within the human respiratory system. [8]
From the same CF-PD modeling study of highly concentrated/dense smoke-aerosol particles
from conventional cigarettes contributes to enhanced deposition of the smoke components (including
DA and PD) in the upper airway region of the respiratory tract. [11] The data concerning the vapor
generated from an e-cig/vaporizer device is sparse, but one must assume based on the CF-PD modeling
of conventional cigarette smoke that the less concentrated/less dense aerosol will travel further into the
airway region and into the deeper lung where exposure to the bronchioles and alveoli to DA or PD is
significantly enhanced. This enhanced exposure may very well lead to increased risk of bronchiolitis
obliterans.

4

 Executive Summary

Toxicology of Diacetyl and 2,3-pentanedione

August 3rd 2014

The aerosol deposition dynamics within the lung, as modeled by CF-PD simulations, provide a
reasonable basis as to why according to Pierce et al. deeper lung diseases such as bronchiolitis obliterans
have not been reported in cigarette smokers even given the high concentrations of DA and PD in
cigarette smoke. The deposition of the aerosol particles in cigarette smoke occurs primarily in the upper
respiratory tract, whereas bronchiolitis obliterans is a deep lung disease.
Concomitantly, with longer puffing and less dense vapor from e-cigs, it is likely that the deeper
part of the lung (bronchioles and alveoli) will be exposed to the compounds within the vapor. If DA and
PD are components of the vapor then the risk of developing serious deep lung diseases such as
bronchiolitis obliterans becomes of critical concern.
Conclusion:
Upon critical evaluation of the available toxicology and epidemiology data described within this
monograph and more thoroughly within the associated references, along with the smoking behavior
data collected on both human subjects and CF-PD simulations describing compensatory smoking
behaviors and more deep lung exposures from e-cigs/vaporizers, it is concluded that diacetyl and 2,3pentanedione should not be considered as candidates for flavoring in the e-cig/vaporizer solution.
Furthermore, any consideration of reactive volatile organic carbonyl compounds for inclusion into the ecig/vaporizer solution should be thoroughly evaluated based upon the most recent toxicological and
epidemiological data.

R. Patrick Rainey, Ph.D., DABT
Technical Toxicology Consultant
Creative Process Solutions, LLC.
Raleigh, NC 27613

5

 Executive Summary

Toxicology of Diacetyl and 2,3-pentanedione

August 3rd 2014

References
1. International Union of Food, Agricultural, Hotel, Restaurant, Catering, Tobacco and Allied Workers’
Associations. Flavouring Ingredient Diacetyl Linked to Deadly LungDisease.
http://www.iuf.org/cgibin/dbman/db.cgi?db=default&ww=1&uid=default&ID=4675&view_records=1&en=1
2. Recommendation from the Scientific Committee on Occupational Exposure limits for Diacetyl.
SCOEL/SUM/149. June 2014. ec.europa.eu/social/BlobServlet?docId=6511&langId=en

3. http://www.cdc.gov/niosh/docket/archive/docket245.html
4. National Toxicology Program (2011). Short-term bioassay data for 2,3-butanedione (diacetyl).
5. The National Institute for Occupational Safety and Health (NIOSH). Occupational exposure to Diacetyl
and 2,3- pentanedione. Review #2 Follow up review of new content. Chapter 6: Quantitative Risk
assessment Based on Animal Data. Dec 2013.
http://www.cdc.gov/niosh/docket/review/docket245A/pdfs/245-A-Draft_Chapter6.pdf

6. Potera C. (2012). Still searching for better butter flavoring. Environ. Health Persp, 120, A457.
7. Pierce, J.S. Abelmann, A. Spicer, L. Adams, R. Finley, B. Diacetyl and 2,3-pentanedione exposures
associated with cigarette smoking: implications for risk assessment of food and flavoring workers. Crit
Rev Toxicol, 2014, p.1-16.
8. Kleinstreuer, C. Feng, Y. Lung Deposition Analysis of Inhaled Toxic Aerosols in Conventional and Less
Harmful Cigarette Smoke: A Review, Int. J. Environ. Res. Public Health, 2013, 10, 4454-4485.
9. Brooks, D.R.; Austin, J.H.; Heelan, R.T.; Ginsberg, M.S.; Shin, V.; Olson, S.H.; Muscat, J.E.;
Stellman, S.D. Influence of type of cigarette on peripheral versus central lung cancer. Cancer
Epidemiol. Biomark. Prev. 2005, 14, 576–581.
10. Alfi, M.; Talbot, P. Health-related effects reported by electronic cigarette users in online

forums.J. Med. Internet Res. 2013, 15, doi:10.2196/jmir.2324.
11. Robinson, R.J.; Yu, C. Deposition of cigarette smoke particles in the human respiratory tract.
Aerosol Sci. Technol. 2001, 34, 202–215.

6