Production of methane and ethylene from plastic in the environment
Production of methane and ethylene from plastic in the environment
Sarah-Jeanne Royer¤ 0 1 2
Sara FerroÂ n 0 1 2
Samuel T. Wilson 0 1 2
David M. Karl 0 1 2
0 Current address: International Pacific Research Center, School of Ocean and Earth Science and Technology, University of Hawaii at Manoa , Honolulu, Hawaii , United States of America
1 Daniel K. Inouye Center for Microbial Oceanography: Research and Education, School of Ocean and Earth Science & Technology, University of Hawaii at Manoa , Honolulu, Hawaii , United States of America
2 Editor: P. Pardha-Saradhi, University of Delhi , INDIA
Mass production of plastics started nearly 70 years ago and the production rate is expected to double over the next two decades. While serving many applications because of their durability, stability and low cost, plastics have deleterious effects on the environment. Plastic is known to release a variety of chemicals during degradation, which has a negative impact on biota. Here, we show that the most commonly used plastics produce two greenhouse gases, methane and ethylene, when exposed to ambient solar radiation. Polyethylene, which is the most produced and discarded synthetic polymer globally, is the most prolific emitter of both gases. We demonstrate that the production of trace gases from virgin lowdensity polyethylene increase with time, with rates at the end of a 212-day incubation of 5.8 nmol g-1 d-1 of methane, 14.5 nmol g-1 d-1 of ethylene, 3.9 nmol g-1 d-1 of ethane and 9.7 nmol g-1 d-1 of propylene. Environmentally aged plastics incubated in water for at least 152 days also produced hydrocarbon gases. In addition, low-density polyethylene emits these gases when incubated in air at rates ~2 times and ~76 times higher than when incubated in water for methane and ethylene, respectively. Our results show that plastics represent a heretofore unrecognized source of climate-relevant trace gases that are expected to increase as more plastic is produced and accumulated in the environment.
Data Availability Statement: All relevant data are
within the paper and its Supporting Information
Funding: This research was supported by National
Science Foundation (C-MORE, DBI-0424599 to
DMK and OCE-1260164 to Matthew J. Church and
DMK), the Simons Foundation (SCOPE Award ID
329108 to DMK), the Balzan Prize for
Oceanography (awarded to DMK) and the Gordon
and Betty Moore Foundation's Marine Microbiology
Initiative (grant #3794 to DMK). The funders had
no role in study design, data collection and
Over the past 50 years, polymer manufacturing has accelerated, from 2x106 metric tonnes
(Mt) per year in 1950 to 381x106 Mt per year in 2015, and is expected to double in the next 20
]. The total global production of plastics to date is estimated at 8300x106 Mt, with
polyethylene being the most common polymer [
], accounting for approximately 36% of all
plastic types . In the environment, plastics are vulnerable to weathering and degradation
processes, caused by environmental factors such as light, heat, moisture, chemical oxidation
and biological activity that are responsible for physical and chemical changes in the structure
of the polymer [
Polyethylene, like other plastics, is not inert and is known to release additives and other degradation products into the environment throughout its lifetime. For example, the additive
analysis, decision to publish, or preparation of the
bisphenol-A used in the manufacture of many plastic products [
] is leached as plastics age,
and hydrocarbon gases are produced during high-temperature decomposition (>202ÊC) [
These chemicals vary amongst different types of plastic and, once released, some can be toxic
and have adverse effects on the environment and human health [7±9]. Degradation processes
not only affect the chemical integrity of the plastic but also ultimately results in the
fragmentation of the polymer into smaller units increasing the surface area exposed to the elements.
Most plastic is synthesized from natural gases  and leaching is expected to occur during
the aging processes. However, to the best of our knowledge, no previous study has reported
hydrocarbon gas emissions from plastics under natural conditions. This study seeks to
investigate this phenomenon and its potential environmental consequences.
Our research investigated the production of hydrocarbon gases from polyethylene and
other plastics at ambient temperature, with an emphasis on methane (CH4), one of the most
potent atmospheric greenhouse gases [11±13] and ethylene (C2H4), which reacts with OH in
the atmosphere and increases carbon monoxide concentrations [
]. Given the substantial
rise in plastic production worldwide, understanding the extent of CH4 and C2H4 emissions
from plastic is essential. In addition, we report production rates of ethane (C2H6), the second
most abundant hydrocarbon in the atmosphere after CH4, known to enhance the level of
tropospheric ozone and carbon monoxide [
], and propylene (C3H6), also a hydrocarbon
pollutant in the atmosphere [
]. Since plastics come in different compositions and
morphologies, we conducted a series of experiments to evaluate gas production under a variety of
environmental conditions. We show that solar radiation initiates the production of these gases for
the polymers tested.
Materials and methods
Multiple experiments were conducted to assess the relative contribution of plastic in
hydrocarbon gas emissions. To assess if different types of plastic emit hydrocarbon gases under natural
conditions and the magnitude of these emissions, we tested some of the most common plastics
produced worldwide. After selecting the plastic category emitting the greatest amount of
hydrocarbon gases, we tested the effect of a change in density and morphology (i.e. pellets,
flakes, powder) on hydrocarbon gas emissions. Environmental factors, including solar
radiation, UVB (ultraviolet B; 280±320 nm) and solar radiation exposure history were varied as part
of this study. These tests were executed in parallel with two long-term experiments assessing
the effect of aging for 212 days using virgin pellets and for 152 days using marine plastic from
the ocean. Finally, to determine the impact of medium on plastic degradation and gas
emissions, we tested the production of gases from plastic incubated in either air or in water. All
experiments involving solar radiation were conducted on the roof of C-MORE Hale located in
Hawaii, at 21.3ÊN, 157.9ÊW.
Hydrocarbon gas production from plastics. Initial tests were conducted using different
substrates including polycarbonate (PC), acrylic (AC), polypropylene (PP), polyethylene
terephthalate (PET), polystyrene (PS), high-density polyethylene (HDPE) and low-density
polyethylene (LDPE). Each plastic type was incubated in seawater and exposed to ambient solar
radiation for several days, to measure the release of CH4 and C2H4 during incubation. Plastic
pieces measured 5.5 x 0.8 cm with a thickness of 0.3 cm and weighed approximately 1.5 g.
Each piece of plastic was incubated separately in 30 mL gas-tight cylindrical quartz vials (9.6
cm long by 2.0 cm diameter) filled with seawater filtered using a cartridge with a 0.2 μm pore
size. The seawater was collected at 25 m depth at Station ALOHA, a long-term monitoring site
located 100 km north of Oahu (Hawaii) in the North Pacific Subtropical Gyre (22.8ÊN,
2 / 13
158.0ÊW). The cruise was part of the Hawaii Ocean Time-series (HOT) program (http://
] and because it occurred in international waters, no special
permission was necessary to sample seawater. The field studies did not involve endangered or
protected species. The vials were incubated in an open, shallow (~15 cm depth) water bath
maintained at 26ÊC. For each plastic type, triplicate vials were exposed to ambient solar
radiation and another set of triplicates was incubated in the dark (by wrapping the vials with
aluminium foil), for a period of six and seven days, respectively. Triplicate quartz vials filled with
filtered seawater but without plastic were incubated in light and dark conditions as control
treatments. PS was tested in a separate experiment using the same vials, with pieces that were
triangular in shape (approximately 5.0 cm perimeter), a thickness of 0.3 cm and a weight of
approximately 0.2 g. A set of triplicate vials with no plastic served as control.
Effects of solar radiation exposure and ultraviolet radiation on hydrocarbon gas
production. Experiments were conducted to determine the effects of aging and irradiance
spectrum on hydrocarbon gas production. For these experiments, we used virgin LDPE pellets
(Marlex 1017, Chevron Phillips) with a diameter of approximately 5 mm and a density of
0.917 g cm-3. Replicate 30 mL quartz vials (n = 4) containing approximately 0.2 g of LDPE
pellets in MilliQ water were incubated on the roof of C-MORE Hale and the outer surfaces of the
vials were cleaned and shaken weekly. Hydrocarbon gas concentrations were measured
approximately every two weeks over a 212-day period. After each measurement, the inner and
outer surfaces of the vials were cleaned, and the vials filled with fresh MilliQ water, the
previously-exposed LDPE pellets added, and incubated again.
To determine the effect of UV radiation on hydrocarbon gas production, triplicate vials
containing 3.0 g of LDPE pellets immersed in MilliQ water were incubated for seven days
either in gas-tight 40 mL Pyrex1 bottles (11.8 cm long and 2.1 cm in diameter) or in 30 mL
quartz vials, after which the gas concentrations were measured. The dark treatment consisted
of virgin LDPE pellets with MilliQ water in vials wrapped in aluminium foil. The Pyrex1
bottles were scanned using a spectrophotometer and did not transmit UVB radiation (280±320
nm) and hence could help in deciphering the importance of shorter wavelengths for
To determine whether hydrocarbon gas production continues once the plastic is removed
from sunlight, triplicate vials containing 3.0 g of LDPE pellets were incubated in 40 mL serum
vials in the dark for 14 days at 25ÊC. The LDPE pellets had been exposed to light for 270 days
and 15 days with a control treatment of virgin pellets that had never been exposed to natural
Effects of polymer morphology and density. An additional test to determine the effect of
polymer morphology on hydrocarbon gas production used six different types of polyethylene.
Five of the samples were LDPE with three of the samples in pellet form while the other two
were either powder or flake. The remaining sample was linear low-density polyethylene
(LLDPE) with a density very close to one of the LDPE pellet treatments. For each treatment,
0.4 g of plastic was incubated in gas-tight quartz vials in MilliQ water and exposed to ambient
solar radiation for eight days on the roof of C-MORE Hale. For all experiments, the control
treatment (n = 3) consisted of vials with MilliQ water but without polyethylene.
Production of hydrocarbon gases from environmental plastic debris. In addition to
virgin plastic, marine plastic debris from the ocean was collected at Station ALOHA. Plastic
pieces were collected using a specialized manta trawl [
] skimming the surface of the ocean
for 30 min. The collected plastic (n = 4) was cleaned by soaking in 10% hydrochloric acid for 2
hours and incubated in MilliQ water on the roof of C-MORE Hale as separate pieces (with a
weight of approximately 0.2 g) for a period of 152 days. Gas measurements were made every
two weeks after which the inner and outer surfaces of the vials were cleaned, filled with MilliQ
3 / 13
water, the previously-exposed plastic added, and incubated again. Vials with MilliQ water and
without plastic were used as control treatment for each time point.
Emission of hydrocarbon gases by LDPE exposed to natural sunlight in air. To
determine if LDPE virgin pellets also emit hydrocarbons when exposed directly to air and if this
production differs under dark conditions, a pre-weighed amount of virgin LDPE pellets (3.0 g)
was placed in 30 mL quartz vials that were flushed with zero-CH4 air for 20 seconds and
immediately closed. The vials were either exposed to full sunlight or wrapped in aluminium foil. An
extra set of flushed vials (without LDPE pellets) for each treatment was also incubated and
used as control treatments. At each time point (1, 3, 9 and 17 days), three samples from each
treatment (control, light and dark) were removed and analyzed for CH4 and C2H4
concentrations in the headspace.
To identify the type of plastic collected in the ocean, spectra were obtained using a micro
Raman RXN system (Kaiser Optical Systems Inc.) as previously described . A 785 nm laser
excitation and a Leica automated microscope stage with a 50 μm slit width were used to
determine the micro-Raman spectra of the different plastic types. The environmental samples were
compared with known plastics and the spectra analysed using GRAMS/AI™ Spectroscopy
The dissolved gas concentrations were quantified using a purge and trap system  con
nected to a gas chromatograph (Agilent 7890A) equipped with a flame ionization detector. A
known volume of sample was transferred from the vials to a sparging chamber using positive
pressure supplied by ultra-high purity helium (Airgas) and sparged for 12 min (100 ml min-1).
The eluting gas stream was passed through a Nafion dryer and Drierite1 and Ascarite1
columns in order to remove water vapour and CO2 before being cryotrapped on a sample loop
containing Poropak Q 80/100 (Sigma-Aldrich) and immersed in liquid nitrogen. The trap was
subsequently heated for two min to approximately 90ÊC and the gas released was injected onto
a 30 m x 0.32 mm GS-CarbonPLOT analytical capillary column (J&W Scientific) with a
temperature program of 38ÊC for 1.0 min, 65ÊC for 1.3 min and 150ÊC for 2.2 min [
Calibration was performed by analysis of compressed gas standards containing 10 ± 2% ppm of CH4,
C2H4, C2H6 and C3H6 in nitrogen (Scott-Marrin). Set volumes of gas standard were analyzed
following the same procedure used for sample analysis [
]. The experimental precision of
these measurements, based on the coefficient of variation of ten replicate samples, was within
5% for all gases. The limit of detection, given the sample volume typically analyzed for these
experiments, was approximately 0.15 nM for all of the gases. The rate of CH4 and C2H4
production in the control treatments (i.e. vials incubated under the same conditions without
plastic) was approximately 5%.
Student's t-distribution (t-test) was used to determine differences between treatments with a level of significance of P < 0.05.
Preliminary experiments showed that all of the polymer types tested (PC, AC, PP, PET, PS,
HDPE and LDPE) produced measurable quantities of CH4 and C2H4 when exposed to
4 / 13
Relevant information regarding the polymer sources is also included. The errors represent the standard deviation of triplicate samples.
NS: final concentrations not significantly different from those in the control treatment (t-test, P>0.05).
: Polystyrene incubations lasted for 14 days and were conducted in MilliQ water.
ambient solar radiation (Table 1). The rate of production spanned more than two orders of
magnitude between the different types of plastic, from 10 to 4100 pmol g-1d-1 for CH4 and ~
20 to 5100 pmol g-1d-1 for C2H4. The highest production rates for both gases were measured
for LDPE. Most of the samples kept in the dark did not produce any gas, and those that did
had much lower production rates, reaching 120 pmol CH4 g-1 d-1 and 60 pmol C2H4 g-1 d-1 for
PS, and 50 pmol CH4 g-1 d-1 and 20 pmol C2H4 g-1 d-1 for PET and AC, respectively. Gas
production from all other plastic types incubated in the dark was either below the detection limit
of the instrument or not significantly different from the control treatment (t-test, P>0.11).
The spectra derived from Raman analysis on plastic debris pieces collected from the open
ocean were compared with different types of commercially sourced common plastics. The
spectra matched with LDPE and therefore the plastic collected from the open ocean was
hereafter referred to as ªaged LDPEº (Fig 1E). For both types of plastic, the cumulative amount of
hydrocarbon gases increased with time although the relative contribution of the gases
produced was variable (Fig 1A and 1B). The rates of hydrocarbon gas production are normalized
to weight and to convert to surface area, the pellet weight of approximately 35 mg had a surface
area ranging from ca. 70 mm2 to 315 mm2. For the virgin LDPE pellets, the cumulative gas
emissions during 212 days were 500, 1700, 340 and 1000 nmol g-1 of CH4, C2H4, C2H6 and
C3H6, respectively (Fig 1A). For aged plastics, the cumulative gas emissions during 152 days
were 700, 700, 180 and 160 nmol g-1 of CH4, C2H4, C2H6 and C3H6, respectively (Fig 1B). Gas
production rates increased over time for the virgin pellets (Fig 1C) but were relatively constant
for the aged plastic (Fig 1D). Except for CH4, the weight-normalized gas production rates were
lower for the aged plastic compared to virgin pellets. During this period of time (December
2016 to April 2017), the PAR intensity at noon, when the sun was at its zenith, varied from
1496 to 1787 μM m-2s-1.
As shown in previous experiments (Fig 1 and Table 1), LDPE releases large amounts of
CH4, C2H4, C2H6 and C3H6 into seawater and MilliQ water when exposed to sunlight. CH4
5 / 13
and C2H4 production rates from virgin LDPE pellets were larger (for similar incubation
periods) when the plastic was exposed to sunlight directly in air (Fig 2), rather than immersed in
water (Fig 1). In that case, production rates for both gases increased with time and ranged
Incubation time (days)
Incubation time (days)
1 ) 2000
Counts / Raman shift cm-1
Fig 1. Gas production from LDPE exposed to ambient solar radiation and identification of plastic collected in the
open ocean. The cumulative amount of hydrocarbon gases normalized to weight and production rates of CH4, C2H4,
C2H6 and C3H6 from (A) & (C) virgin LDPE pellets (0.2 g) and (B) & (D) LDPE plastic pieces (0.2 g) collected from the
open ocean. Measurements were taken at approximately two-week intervals during the incubation period. Error bars
represent the error propagation of the standard deviation of the mean for each time point. Panel (E) represents the
micro-Raman spectra for the aged LDPE collected in the open ocean (red) and the virgin LDPE pellets (blue).
Fig 2. Production rates of CH4 and C2H4 from virgin LDPE pellets exposed to sunlight in air. Mean production
rates of CH4 (A) and C2H4 (B) from LDPE virgin pellets incubated in air under ambient solar radiation for 1, 3, 9 and
17 days. The error bars represent the standard deviation of triplicate samples. The emission of CH4 in the dark
treatment was below the detection limit and less than 0.3 nmol g-1d-1 for C2H4. The black arrow indicates the
production rate of CH4 and C2H4 from virgin LDPE pellets incubated in MilliQ water for 14 days (Fig 1).
from 0.11 to 0.44 nmol CH4 g-1d-1 and 0.32 to 96 nmol C2H4 g-1d-1 after 17 days of incubation
(Fig 2). Those final rates are approximately 2 and 76 times larger than CH4 and C2H4
production from LDPE virgin pellets incubated in MilliQ water for 14 days (Figs 1 and 2). No
detectable emissions of any of the gases were observed in the control or dark treatments (data not
The emissions of hydrocarbon gases were measured using virgin LDPE pellets exposed for
seven days to the full solar spectrum (FULL; 280±700 nm) and to solar radiation without UVB
((-)UVB; 320±700 nm; Fig 3). Except for C3H6 (t-test, P = 0.636), the results showed significant
differences (t-test, P<0.002) between the two treatments, with consistently lower rates for the
(-)UVB treatment corresponding to 23%, 74%, and 68% of the FULL treatment for CH4, C2H4
and C2H6, respectively. All dark treatments showed emission rates less than 0.1 nmol g-1d-1.
LDPE pellets incubated in the dark after exposure to light produced measurable quantities of
hydrocarbon gases. The rates of CH4 (0.11 nmol g-1d-1) and C2H4 (0.52 nmol g-1d-1) production
were equivalent to 55% and 40%, respectively, of the rates when the pellets were incubated in
the light (0.20 nmol CH4 g-1d-1 and 1.3 nmol C2H4 g-1d-1). In addition, environmentally-aged
LDPE plastic collected from the open ocean and incubated for 14 days in the dark under
ambient outdoor temperature [18.5±32.5ÊC] also produced 0.37 ± 0.11, 0.14 ± 0.09, 0.21 ± 0.12 and
0.06 ± 0.03 nmol g-1d-1 of CH4, C2H4, C2H6, and C3H6, respectively. Pellets that had been
previously exposed to 270 days of natural solar radiation and incubated at 25ÊC were amongst the
highest production rates for all gases with 0.23 ± 0.02, 0.41 ± 0.05, 2.6 ± 0.22 and 0.70 ± 0.09
nmol g-1d-1 of CH4, C2H4, C2H6, and C3H6, respectively. These results indicate that once
initiated, the production of hydrocarbon gases continues in the dark.
Morphology has an important effect on hydrocarbon gas production. LDPE powder
exhibited weight-normalized CH4 and C2H4 production rates that were 488 and 132 times higher
than LDPE pellets, respectively (Table 2). All forms of LDPE, except LLDPE, produced
measurable quantities of hydrocarbon gases with CH4 and C2H4 production rates ranging from
Fig 3. Effect of spectral irradiance on hydrocarbon gas production from LDPE. Production rates of hydrocarbon
gases (CH4, C2H4, C2H6 and C3H6) from virgin LDPE pellets kept in the dark, exposed to full solar radiation (280±700
nm) or to solar radiation filtered for UVB (320±700 nm). Error bars represent the standard deviation of triplicate
0.08 to 55 nmol CH4 g-1d-1 and 0.16 to 21 nmol C2H4 g-1d-1. With respect to the LDPE pellets
of varying densities, there were significant differences in CH4 production between treatments
(t-test, P<0.03), with lowest rates for the lowest density treatment and highest rates for the
Mean production rates of hydrocarbon gases from different polyethylene products of different densities and morphologies exposed to ambient solar radiation for 14
days. Relevant information regarding the polymers is also included in the table: the source and product number, the polyethylene category, the morphology of the
product and the density at 25ÊC. The error represents the standard deviation of triplicate samples.
NS: final concentrations were not significantly different from those in the control treatment (t-test, P>0.05).
8 / 13
highest density treatment. No other consistent pattern for the other hydrocarbon gases was
observed with the different densities. Interestingly, the ratios of the different gases produced
varied across LDPE pellet densities. For example, the C2H4: CH4 ratio varied from 1.3 to 4.4
and the C2H4:C2H6 ratio varied from 1.9 to 8.6 indicating that degradation processes were
variable for the different morphologies under these experimental conditions.
The influence of plastic composition on hydrocarbon gas production
Our results show that different types of plastics, that are commonly used and dispersed in the
environment worldwide, produce CH4 and C2H4 under environmental conditions. We
hypothesize that the relative amounts of low-molecular-weight hydrocarbon gas molecules
that are released from plastic substrates depend on the molecular structure of the plastic
including the degree of branching, the addition of plasticizers, as well as the manufacturing
process. For example, among the plastic materials tested, LDPE produced the largest amounts
of CH4 and C2H4, probably due to its weaker structure and more exposed hydrocarbon
branches. In contrast, with a more compact structure, lower permeability and fewer accessible
active sites, degradation of HDPE resulted in lower emission.
Long-term hydrocarbon gas emission from plastics
The release of greenhouse gases from virgin and aged plastic over time indicates that polymers
continue to emit gases to the environment for an undetermined period. We attribute the
increased emission of hydrocarbon gases with time from the virgin pellets to
photo-degradation of the plastic, as well as the formation of a surface layer marked with fractures,
microcracks and pits [24±26]. With time, these defects increase the surface area available for further
photo-chemical degradation and therefore might contribute to an acceleration of the rate of
gas production. It is also known that smaller particles of secondary origin termed
] are eventually produced and may further accelerate gas production. The initial
shape of the polymer is also a potential factor contributing to the variability in hydrocarbon
production because items of the same mass but with different shapes have different
surface-tovolume ratios. Small fragments not only have a greater surface-to-volume ratio than larger
items, but they also tend to have longer edge lengths relative to their volume . This predicts
that in the environment, as plastic particles degrade and become smaller, they will also emit
more hydrocarbon gases per unit mass. The emission of gases from the aged plastic collected
from the ocean (age unknown at the start of the experiment) indicates that production may
continue throughout the entire lifetime of the plastic. We hypothesize that lower production
rates from environmentally aged plastic are probably due to the presence of plasticizers that
retard photo-degradation by counteracting the negative effect of UV radiation.
Driving factors for gas emissions
Manipulation of the radiation spectrum showed that UVB increases the production of
hydrocarbon gases relative to longer wavelengths, which is consistent with its key role in the
photooxidative degradation processes for polyethylene [
]. Indeed, solar photon flux, especially in
the UV portion of the spectrum, provides the activation energy necessary to initiate
degradation due to bond cleavage and depolymerisation, which affects the mechanical properties of
the material [
]. We hypothesize that this process yields hydrocarbon gas molecules
that are released into the environment. Although the emission of hydrocarbon gases is
enhanced by photo-degradation in the UVB spectrum, our results indicate that UVB radiation
9 / 13
is not essential for the initiation or the continuous production of hydrocarbon with time.
Once initiated by solar radiation, emission of the gases continues in the dark, at a rate that depends on previous radiation exposure.
Effects of morphology and density of polyethylene on gas emissions
While the production of hydrocarbon gases was prevalent in LDPE, the different morphologies
and densities within LDPE products resulted in a change in the relative concentrations of the
different gases. LDPE in powder form, with a much higher surface area than pellets with the
same density and weight, produced the highest amount of hydrocarbon gases. Along with
morphology and density, the average molecular weight that varies with the polymer chain
length and hence the number of branched molecules exposed may also be a controlling factor
in hydrocarbon emissions. Unfortunately, this information is proprietary and very rarely
available from suppliers. Also, even polyethylene, which is known to be the simplest polyolefin, can
form various conformational structures of the macromolecular chains, which in its solid state
assume various states of intermolecular order [
]. This heterogeneity within the same type of
plastic would affect the production rates of the different hydrocarbon gases and consequently
the ratios among them. Yet, due to the heterogeneity of the products even within the same
type of plastic, the nature of the reactions is unknown. The shortening of the long polymer
chain and its breakdown into smaller units can be caused by different mechanisms such as
random chain scission, end-chain scission, chain-stripping or cross-linking [
]. These reactions
would lead to the production of hydrocarbon gases including those presented in this study.
The global production of plastic is large [
] and the amount of plastic waste generated in 2010
from 192 countries was 275x106 Mt from which 4.8x106 to 12.7x106 Mt was estimated to enter
the ocean [
]. By 2025, the amount of plastic waste input to marine systems might increase by
an order of magnitude if waste management is not improved [
]. Because polyethylene is the
most common polymer, it is anticipated to be the most common form of plastic pollution in
surface ocean waters worldwide [33±35]. In addition, as microplastics with greater surface area
are produced hydrocarbon gas production rates will likely accelerate. The results from this
study indicate that hydrocarbon gas production may continue indefinitely throughout the
lifetime of plastics. The amount of plastic material in the environment exposed to full sunlight
exceeds the quantity of submerged plastic. Our experiments indicated that emissions of
hydrocarbon gases are even greater (up to 2 times higher for CH4 and up to 76 times higher for
C2H4) in air compared to in water. The difference in emission rate for plastic in water
compared to plastic exposed to air is partly due to temperature and heat build-up, [
] resulting in
the plastic material reaching a temperature higher than the surrounding medium.
Furthermore, plastic exposed to air is subjected to less biofouling than in aquatic environments
because of the lack of a fluid medium that facilitates microorganisms settling onto the plastic
substrate resulting in more surfaces directly exposed to solar radiation and more conducive to
deformation, degradation and hydrocarbon production. Therefore, in warm climates, higher
hydrocarbon production rates are expected for plastic exposed to air compared to plastic in
At this stage, very little is known about the contribution of plastic to CH4, C2H4 and other
hydrocarbon gas budgets. This study reports the first measurements of hydrocarbon
production by plastic under ambient conditions. Based on the rates measured in this study and the
amount of plastic produced worldwide CH4 production by plastics is likely to be an
insignificant component of the global CH4 budget. However, for the other hydrocarbon gases with
10 / 13
much lower global emissions to the atmosphere compared to CH4, the production from the
plastics might have more environmental and global relevance. The fate of these gases is also
not well constrained, but sinks are likely to be bacterial oxidation to carbon dioxide with a
portion of these gases emitted to the atmosphere. Given the ongoing rate at which plastic is being
produced, used and exposed on land and the future trend in mismanaged plastic waste ending
up in marine systems [
], the amount of plastic exposed to the environment will likely
increase with time and so too will the amount of CH4 and C2H4 emitted from polymers. In
addition, degradation of plastics in the environment leads to the formation of microplastics
with greater surface area, which may accelerate hydrocarbon gas production. Due to the
longevity of plastics and the large amounts of plastic persisting in the environment, questions
related to the role of plastic in the CH4 and C2H4 global budgets should be prioritized and
addressed by the scientific community.
S1 Dataset. Dataset used for Figs 1, 2 and 3.
We thank Eric Baer, Director of the Center for Layered Polymeric Systems (CLIPS) for advice
on polymer characteristics, Anupam Misra and Shiv Sharma for the use of the Raman
spectroscopy facilities at the University of Hawaii, Lance Fujieki for his help with meteorological
measurements and Blake Watkins for facilitating the collection of plastic at sea. We also thank
LabQMC for performer the Brunauer-Emmet-Teller analysis to determine the surface area of our samples.
Formal analysis: Sarah-Jeanne Royer.
Funding acquisition: David M. Karl.
Conceptualization: Sarah-Jeanne Royer, Sara FerroÂn, Samuel T. Wilson, David M. Karl.
Investigation: Sarah-Jeanne Royer, Sara FerroÂn, Samuel T. Wilson, David M. Karl.
Methodology: Sarah-Jeanne Royer, Sara FerroÂn, Samuel T. Wilson, David M. Karl.
Project administration: Sarah-Jeanne Royer, David M. Karl.
Resources: Sarah-Jeanne Royer, Sara FerroÂn, Samuel T. Wilson, David M. Karl.
Supervision: Samuel T. Wilson, David M. Karl.
Validation: Sarah-Jeanne Royer.
Visualization: Sarah-Jeanne Royer.
Writing ± original draft: Sarah-Jeanne Royer.
Writing ± review & editing: Sarah-Jeanne Royer, Sara FerroÂn, Samuel T. Wilson, David M.
11 / 13
12 / 13
1. Geyer R , Jambeck JR , Law KL . Production, use, and fate of all plastics ever made . Sci Adv . 2017 ; 3 : 25 ± 29 .
2. Andrady AL , Neal MA . Applications and societal benefits of plastics . Philos Trans R Soc Lond B Biol Sci . 2009 ; 364 : 1977 ± 1984 . https://doi.org/10.1098/rstb. 2008 .0304 PMID: 19528050
3. PlasticsEurope. Plastics±the Facts 2016 . 2016 ; https://doi.org/10.1016/j.marpolbul. 2013 . 01 .015
4. Gewert B , Plassmann MM , MacLeod M. Pathways for degradation of plastic polymers floating in the marine environment . Environ Sci Process Impacts . Royal Society of Chemistry; 2015 ; 17 : 1513 ± 21 . https://doi.org/10.1039/c5em00207a PMID: 26216708
5. Huang YQ , Wong CKC , Zheng JS , Bouwman H , Barra R , WahlstroÈm B , et al. Bisphenol A (BPA) in China: A review of sources, environmental levels, and potential human health impacts . Environ Int. Elsevier Ltd ; 2012 ; 42 : 91 ± 99 . https://doi.org/10.1016/j.envint. 2011 . 04 .010 PMID: 21596439
6. Beyler CL , Hirschler MM . Thermal decomposition of polymers . SPE Handbook of Fire Protection Engineering . 2001 . pp. 111 ± 131 . https://doi.org/10.1021/cm200949v
7. Teuten EL , Saquing JM , Knappe DRU , Barlaz MA , Jonsson S , BjoÈrn A , et al. Transport and release of chemicals from plastics to the environment and to wildlife . Philos Trans R Soc Lond B Biol Sci . 2009 ; 364 : 2027 ±45. https://doi.org/10.1098/rstb. 2008 .0284 PMID: 19528054
8. Thompson RC , Moore CJ , vom Saal FS , Swan SH . Plastics, the environment and human health: current consensus and future trends . Philos Trans R Soc Lond B Biol Sci . 2009 ; 364 : 2153 ± 2166 . https:// doi.org/10.1098/rstb. 2009 .0053 PMID: 19528062
9. Moore CJ . Synthetic polymers in the marine environment: A rapidly increasing, long-term threat . Environ Res . 2008 ; 108 : 131 ± 139 . https://doi.org/10.1016/j.envres. 2008 . 07 .025 PMID: 18949831
10. Vasile MD C. Handbook of polyolefins . 2nd ed. re. C. Vasile MD , editor. New York; 2000 .
11. Saunois M , Jackson RB , Bousquet P , Poulter B , Canadell JG . The growing role of methane in anthropogenic climate change . Environ Res Lett. IOP Publishing ; 2016 ; 11 : 1±5 . https://doi.org/10.1088/ 1748 - 9326/11/12/120207
12. Rigby M , Prinn RG , Fraser PJ , Simmonds PG , Langenfelds RL , Huang J , et al. Renewed growth of atmospheric methane . Geophys Res Lett . 2008 ; 35 : 2±7 . https://doi.org/10.1029/2008GL036037
13. Dlugokencky EJ , Bruhwiler L , White JWC , Emmons LK , Novelli PC , Montzka S.A. , et al. Observational constraints on recent increases in the atmospheric CH4 burden . Geophys Res Lett . 2009 ; 36 : 1±5 . https://doi.org/10.1029/2009GL039780
14. Aikin AC , Herman JR , Maier EJ , McQuillan CJ . Atmospheric chemistry of ethane and ethylene . J Geophys Res Ocean . 1982 ; 87 : 3105 ± 3118 .
15. Ghanta M , Fahey D , Subramaniam B . Environmental impacts of ethylene production from diverse feedstocks and energy sources . Appl Petrochemical Res . 2014 ; 4 : 167 ± 179 . https://doi.org/10.1007/ s13203-013-0029-7
16. Aydin M , Verhulst KR , Saltzman ES , Battle MO , Montzka S a , Blake DR , et al. Recent decreases in fossil-fuel emissions of ethane and methane derived from firn air . Nature . 2011 ; 476 : 198 ± 201 . https://doi. org/10.1038/nature10352 PMID: 21833087
17. Report of the Intergovernmental Panel on Climate Change. IPCC ( 2014 ) Climate change 2014: synthesis report. Contribution of working groups I, II and III to the fifth assessment . Cambridge; 2014 .
18. Karl DM , Lukas R. The Hawaii Ocean Time-series (HOT) program: Background, rationale and field implementation . Deep-Sea Res Part II Top Stud Oceanogr . 1996 ; 43 : 129 ± 156 . https://doi.org/10.1016/ 0967 - 0645 ( 96 ) 00005 - 7
19. Brown DM , Cheng L. New net for sampling the ocean surface . Mar Ecol Prog Ser . 1981 ; 5 : 225 ± 227 . https://doi.org/10.3354/meps005225
20. Misra AK , Sharma SK , Kamemoto L , Zinin P V. , Iu Q , Hu N , et al. Novel micro-cavity substrates for improving the Raman signal from submicrometer size materials . Appl Spectrosc . 2009 ; 63 : 373 ± 377 . https://doi.org/10.1366/000370209787598988 PMID: 19281655
21. Tilbrook BD , Karl DM . Methane sources, distributions and sinks from California coastal waters to the oligotrophic North Pacific gyre . Mar Chem . 1995 ; 49 : 51 ± 64 . https://doi.org/10.1016/ 0304 - 4203 ( 94 ) 00058 -L
22. Repeta DJ , FerroÂn S , Sosa OA , Johnson CG , Repeta LD , Acker M , et al. Marine methane paradox explained by bacterial degradation of dissolved organic matter . Nat Geosci . 2016 ; 9 : 884 ± 887 . https:// doi.org/10.1038/ngeo2837
23. Wilson ST , FerroÂn S , Karl DM . Interannual variability of methane and nitrous oxide in the North Pacific Subtropical Gyre . Geophys Res Lett . 2017 ; 9885 ± 9892 . https://doi.org/10.1002/2017GL074458
24. Tavares AC , Gulmine J V. , Lepienski CM , Akcelrud L . The effect of accelerated aging on the surface mechanical properties of polyethylene . Polym Degrad Stab . 2003 ; 81 : 367 ± 373 . https://doi.org/10. 1016/S0141- 3910 ( 03 ) 00108 - 3
25. Kupper L , Gulmine J V. , Janissek PR , Heise HM . Attenuated total reflection infrared spectroscopy for micro-domain analysis of polyethylene samples after accelerated ageing within weathering chambers . Vib Spectrosc . 2004 ; 34 : 63 ± 72 . https://doi.org/10.1016/j.vibspec. 2003 . 05 .002
26. Andrady AL . Microplastics in the marine environment . Mar Pollut Bull . 2011 ; 62 : 1596 ± 1605 . https://doi. org/10.1016/j.marpolbul. 2011 . 05 .030 PMID: 21742351
27. Browne MA , Galloway T , Thompson R . Microplastic-an emerging contaminant of potential concern? Integr Environ Assess Manag . 2007 ; 3 : 559 ± 561 . https://doi.org/10.1002/ieam.5630030412 PMID: 18046805
28. Arthur C , Baker J , Bamford H . Proceedings of the international research workshop on the occurrence, effects, and fate of microplastic marine debris . NOAA Tech Memo NOS-OR&R-30 Arthur C , Bak J , Bamford H , (Eds). 2009 ; 530 .
29. Ryan PG . Does size and buoyancy affect the long-distance transport of floating debris? Environ Res Lett . 2015 ; 10 : 1 ± 6 .
30. Raquez J-M , Bourgeois A , Heidi J , Degee P , Alexandre M , Dubois P. Effects of expandable graphite and modified ammonium polyphosphate on the flame-retardant and mechanical properties of wood flour-polypropylene composites . J Appl Polym Sci . 2011 ; 122 : 489 ± 496 . https://doi.org/10.1002/app
31. Amin MB , Hamid HS , Khan JH . Photo-oxidative degradation of polyethylene greenhouse film in a harsh environment . J Polym Eng . 1995 ; 14 : 253 ± 268 . https://doi.org/10.1515/POLYENG. 1995 . 14 .4. 253
32. Jambeck JR , Geyer R , Wilcox C , Siegler TR , Perryman M , Andrady A , et al. Plastic waste inputs from land into the ocean . Science . 2015 ; 347 : 768 ± 771 . https://doi.org/10.1126/science.1260352 PMID: 25678662
33. Frias JPGL , Otero V , Sobral P . Evidence of microplastics in samples of zooplankton from Portuguese coastal waters . Mar Environ Res . 2014 ; 95 : 89 ± 95 . https://doi.org/10.1016/j.marenvres. 2014 . 01 .001 PMID: 24461782
34. Hidalgo-Ruz V , Gutow L , Thompson RC , Thiel M. Microplastics in the marine environment: A review of the methods used for identification and quantification . Environ Sci Technol . 2012 ; 46 : 3060 ± 3075 . https://doi.org/10.1021/es2031505 PMID: 22321064
35. Reisser J , Shaw J , Wilcox C , Hardesty BD , Proietti M , Thums M , et al. Marine plastic pollution in waters around Australia: Characteristics, concentrations, and pathways . PLoS One . 2013 ; 8 : 1± 11 . https://doi. org/10.1371/journal.pone. 0080466 PMID: 24312224
36. Andrady AL . Weathering of polyethylene (LDPE) and enhanced photodegradable polyethylene in the marine environment . J Appl Polym Sci . 1990 ; 39 : 363 : 370 .