Thermal degradation of terpenes: Camphene, Δ3-carene, limonene, and α-terpinene

Article abstract of DOI:10.1021/es9810641

Emissions from wood dryers have been of some concern for a number of years, and recent policy changes by the Environmental Protection Agency have placed emphasis upon the gaseous emissions that lead to the formation of particulate matter as small as 2.5 μm diameter. In this qualitative study, camphene, Δ³-carene, limonene, and α-terpinene were thermally degraded in the presence of air to determine the number and kind of oxidative degradation products that might be expected under drying conditions used in processing wood products. Various chromatographic methods were used to isolate the products for proof of structure by NMR and/or GC-MS. The degradation products resulted from dehydrogenations, epoxidations, double bond cleavages, allylic oxidations, and rearrangements. A number of compounds not previously associated with the thermal degradation of these terpenes were identified. Emissions from wood dryers have been of some concern for a number of years, and recent policy changes by the Environmental Protection Agency have placed emphasis upon the gaseous emissions that lead to the formation of particulate matter as small as 2.5 μm diameter. In this qualitative study, camphene, Δ³-carene, limonene, and α-terpinene were thermally degraded in the presence of air to determine the number and kind of oxidative degradation products that might be expected under drying conditions used in processing wood products. Various chromatographic methods were used to isolate the products for proof of structure by NMR and/or GC-MS. The degradation products resulted from dehydrogenations, epoxidations, double bond cleavages, allylic oxidations, and rearrangements. A number of compounds not previously associated with the thermal degradation of these terpenes were identified.

Full text of DOI:10.1021/es9810641

Environ. Sci. Technol. 1999, 33, 4029-4033
The hydrocarbon emissions are typically divided into two
Thermal Degradation of Terpenes:
Camphene, ∆3-Carene, Limonene,
and r-Terpinene
G E R A L D W . M C G R A W , * , †
R I C H A R D W . H E M I N G W A Y, ‡
L E O N A R D L . I N G R A M , J R . , §
C A T H E R I N E S . C A N A D Y, † A N D
W I L L I A M B . M C G R A W †
categories: (a) the condensed hydrocarbons of higher
molecular weight that are responsible in part for the blue-
haze plume characteristic of dryer emissions and (b) the
lower molecular weight hydrocarbons (C2-C10), generally
referred to as volatile hydrocarbons. Both nongaseous
(condensed) and gaseous (volatile) hydrocarbons emitted
by wood dryers have been analyzed by a number of workers
(2-8). In general, the nongaseous fraction consists of a
mixture of resin acids and fatty acids and their esters as well
as some sesquiterpenoid compounds and undefined oxida-
tion products. The gaseous fraction is primarily made up of
monoterpenes present in the wood and some of their
oxidation products. Comparatively little is known about the
yields, structures, and biological properties of oxidation
products of monoterpenes.
Cronn et al. (3) studied the gaseous emissions from a
number of veneer dryers at mills in the northwest and
southern U.S. From a plywood veneer dryer in the southern
U.S. using a mixture of loblolly and shortleaf pines, it was
found that terpenes accounted for 98.9% of the total gaseous
hydrocarbon emissions. The authors also found that for
loblolly and shortleaf pines the total emissions were pre-
dominated by terpenes (86-93% and 56-62%, respectively).
For slash pine, the nongaseous fraction constituted the bulk
of emissions with only 42% gaseous. It was estimated that
the terpenic hydrocarbons were emitted at a rate of about
4 lb/ h, which would be the equivalent to the emissions of
125 acres of coniferous forest at noon (9).
Department of Chemistry, Louisiana College,
Pineville, Louisiana 71359, Southern Research Station,
USDA Forest Service, 2500 Shreveport Highway,
Pineville, Louisiana 71360, and Forest Products Lab,
Department of Forest Products, Mississippi State University,
Mississippi State, Mississippi 39762-9820
Emissions from wood dryers have been of some concern
for a number of years, and recent policy changes by the
Environmental Protection Agency have placed emphasis
upon the gaseous emissions that lead to the formation of
particulate matter as small as 2.5 µm diameter. In this
qualitative study, camphene, ∆3-carene, limonene, and
R-terpinene were thermally degraded in the presence of
air to determine the number and kind of oxidative degradation
products that might be expected under drying conditions
used in processing wood products. Various chromatographic
methods were used to isolate the products for proof of
structure by NMR and/or GC-MS. The degradation products
resulted from dehydrogenations, epoxidations, double
bond cleavages, allylic oxidations, and rearrangements. A
number of compounds not previously associated with
the thermal degradation of these terpenes were identified.
Some definitive work has also taken place regarding the
fate of these hydrocarbons in the atmosphere. Graedel (10)
reports that the oxidation of the terpenes in the lower
atmosphere is initiated by reaction with the ozone molecule
or hydroxyl radical. After measuring the rate constants for
these reactions, he was able to calculate the tropospheric
lifetimes for a number of terpenes after assuming constant
atmospheric concentrations for ozone and the hydroxyl
radical. The lifetimes ranged from under 1 min to slightly
less than 8 h, leading one to expect that most would react
near the point of emission.
Introduction
Most of the studies on the atmospheric oxidation of
terpenes have been carried out utilizing smog chambers,
although a few suspected oxidation products have been
detected in ambient air. Smog chamber experiments have
been performed using ozone, hydroxyl radical, and NOx.
Terpenes studied include R-pinene (5, 11), â-pinene (5, 12,
13), ∆3-carene (6, 12), limonene (5, 6, 14), and terpinolene
(6, 14) among others. It seems to be typical that, when the
terpenes react with ozone, hydroxyl radical, or NOx, the
products generated form aerosols; it is in these aerosol
particles that the products are identified. The degradation
products are usually oxygenated, and it has been estimated
that 80% of the terpenoid carbon is converted to particulate
matter within several hours (15).
The occurrence of terpenes in wood has been thoroughly
investigated (16) with considerable work directed toward
those in southern pines (17, 18). It has been shown that the
hydrocarbon effluent from drying kilns closely resembles
the hydrocarbon content of the wood and smog chamber
experiments on those hydrocarbons illustrate the fates
of those hydrocarbons in the atmosphere. However, although
reference is made to terpene oxidation products in the
effluents, little work has been done to identify those products.
This work takes a qualitative look at the nature of oxidation
products that might be generated in the drying of wood
products with the anticipation that knowing some of the
Monoterpenes and their derivatives are commonly found in
trace amounts in the troposphere. In addition to being
emitted by many species of vegetation (biogenic sources),
they are also introduced into the atmosphere in a variety of
activities associated with wood processing industries (an-
thropogenic sources). It has been estimated that the annual
global volatile organic compounds (VOC) flux is 1150 Tg C,
of which 11% is thought to be monoterpenes (1). Whereas
such compounds are normally recovered from wood pulping
processes and used in a variety of chemical markets, they are
not generally recovered in solid wood or wood composite
manufacture.
Emissions from wood dryers have been of some concern
for a number of years (2). Although these emissions may
include particulates, of more interest are the hydrocarbons
that make up the bulk of the effluent. The hydrocarbons are
of concern, not because of intrinsic toxicity or corrosiveness,
but because of their potential for the production of both
particulates and oxidants downwind from the source (3).
* Corresponding author phone: (318)487-7193; fax: (318)487-7191;
e-mail: mcgraw@lacollege.edu.
† Louisiana College.
‡ USDA Forest Service.
§
Mississippi State University.
9
10.1021/es9810641 CCC: $18.00
Published on Web 10/02/1999
1999 Am erican Chem ical Society
VOL. 33, NO. 22, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 4 0 2 9

products actually generated under such contitions and, in
the least, the nature of oxidation reactions occurring will
facilitate the identification of materials emitted from wood
dryers.
TABLE 1. Thermal Degradation Products of Camphene
Experimental Methods
Terpenes were used as obtained from Aldrich Chemical Co.:
(+)-camphene, 95% (91% by GC); ∆3-carene, 90% (89% by
GC); R-(+)-limonene, 97% (99% by GC); and R-terpinene,
85% (86% by GC). The thermal degradations were carried
out by placing 4.0 g of terpene (0.029 mol) into a 4-L screw-
cap bottle. Thus, the bottle contained 0.036 mol of O2 to give
a ratio of terpene to O2 of 1:1.24. The bottle was then sealed
and heated at 120 °C in a gas chromatograph oven for varying
lengths of time, depending on the reactivity of the terpene.
Upon cooling, the reaction vessel was rinsed with CH2Cl2
and acetone to remove oxidation products and any unreacted
terpene. No attempt was made to analyze gaseous products
that may have escaped upon opening the reaction vessel.
After removal of the solvent under reduced pressure, the
reaction mixture was subjected to flash chromatography (19)
over silica, eluting with increasingly polar mixtures of hexane/
ethyl acetate. The eluted fractions were monitored using TLC
on silica plates and visualized by their UV absorbance and/
or their color reaction with anisaldehyde spray (20). The
combined fractions usually required further purification by
column and/ or preparative thin-layer chromatography, also
using silica gel as the adsorbent. Because several separate
chromatographic procedures were typically used in order to
obtain materials for identification and each purification step
resulted in loss of sample, quantitative results are not
reported.
might be present. This makes it difficult to assign any
particular product or series of products to a specific oxidation
process but does more closely mimic wood-drying ovens or
kilns.
The degradation products of the terpenes are shown in
Tables 1 (camphene), 2 (∆3-carene), 3 (limonene), and 4 (R-
terpinene). The method of identification is also shown as
Unheated terpene samples were also subjected to flash
chromatography and compared to the heated samples using
GC-MS in order to be certain that the products isolated
were not artifacts resulting from the exposure of the terpenes
to silica.
NMR (compound was isolated and analyzed by 1H NMR, 13
C
NMR, and various 2D NMR techniques), GC-MS actual
(compound identified by GC-MS by comparing retention
time and mass spectrum to those of the actual compound),
and GC-MS library [compound was identified by comparing
mass spectrum to mass spectral terpene database (540
compounds) and/ or NIST database (10 000 compounds)].
The products identified in this study indicate that the
thermal degradation of the terpenes leads to four types of
oxidative reactions. First, six-membered rings containing one
or two double bonds readily undergo dehydrogenation to
form aromatic systems. This was observed with R-terpinene
(Figure 1) and limonene in the formation of p-cymene and
thymol, respectively. In addition, both camphene and ∆3-
carene yielded aromatic compounds among their degradation
products, resulting from rearrangement accompanied by
dehydrogenation.
A second common reaction is the oxidative cleavage of
carbon-carbon double bonds. This was observed with the
formation of camphenilone from camphene and the genera-
tion of two keto aldehydes by cleavage of the double bonds
in R-terpinene (Figure 2). Smog chamber studies on numer-
ous terpenes using ozone and/ or hydroxyl radical showed
the major degradation products to be those resulting from
oxidative cleavage of the olefinic bond (22, 25, 26), similar
to the results from this study.
Results and Discussion
The title monoterpenes were selected for their occurrence
in wood dryer emissions (3) and/ or their structural char-
acteristics (mono- vs bicyclic, conjugated vs nonconjugated,
etc.). After beginning the study, it was learned that similar
work with R- and â-pinene had already been carried out
(21), so work on these two common monoterpenes was
minimized. Punsuvon (21) heated R-pinene in the presence
of air at temperatures of 90-130 °C and reported that between
23 and 37% of the R-pinene had been rearranged or oxidized.
Products included â-pinene, R-pinene oxide, R-campholene
aldehyde, verbenol, pinocamphone, myrtenol, and ver-
benone. Similar air oxidation of â-pinene resulted in the loss
of 22% of this compound, and myrtenol was identified as a
major product. In our study, the amount of oxidation of the
terpenes roughly paralleled the atmospheric lifetimes of the
terpenes reported in the literature (1). The percent degrada-
tion of the terpenes (estimated by GC on the reaction
mixtures) and the length of heating time at 120 °C were as
follows: R-terpinene, 100% in 4 h; limonene, 50% in 24 h;
∆3-carene, 7% in 24 h and 36% in 72 h; and, camphene, 38%
in 72 h. Although none of the degradation products identified
were new compounds, many of them had not been reported
as thermal degradation products of the respective terpenes,
being generated instead in smog chamber studies by various
oxidizing agents or isolated as natural products themselves.
The products isolated from the thermal degradation of ∆3-
carene are examples of this (22-24).
The third reaction that was characteristic of the thermal
degradations was epoxide formation. This was observed with
∆3-carene, limonene (Figure 3), and R-terpinene. The ∆3-
carene and limonene products were the expected 1,2-epoxy
derivatives, but the products isolated and identified from
R-terpinene were 1,4- and 1,8-ethers (1,4-cineole and eu-
calyptol, respectively). Epoxides are readily generated in the
laboratory using peracids (27). Their formation under
atmospheric conditions has been attributed to the presence
of nitrate radicals and peroxy radicals (28).
The use of ambient air in this study means that any of the
oxidizing agents used in the smog chamber studies previously
referencedsozone, hydroxyl radical, and nitrate radicals
9
4 0 3 0 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 22, 1999

TABLE 2. Thermal Degradation Products of ∆3-Carene
TABLE 3. Thermal Degradation Products of Limonene
Allylic oxidation to form alcohols, ketones, and aldehydes
represents the fourth oxidation reaction observed under the
conditions investigated. The majority of the identified
degradation products from ∆3-carene (Figure 4) and limonene
resulted at least in part from such oxidation. Singlet oxygen
reacting with alkenes can lead to oxidation of the carbon
adjacent to the carbon-carbon double bond (allylic oxida-
tion) (28). Jay and Stieglitz have identified the allylic oxidation
product verbenone in smog chamber studies with R-pinene
and ozone (29).
In addition to the above oxidation reactions, some
degradation products resulted from carbon skeleton rear-
rangements. These products also incorporated at least one
of the previously described oxidation reactions, and it is
unclear whether the monoterpene or its oxidation product
underwent the rearrangement. In such manner camphene
yielded verbenone, ∆3-carene produced o-cymene and
4-hydroxy-2-methyl-2-cyclohexenone, and eucarvone was
generated from limonene (Figure 5).
It is not possible to accurately mimic the conditions of
drying of wood flakes for manufacture of oriented strand
board, drying wood veneers used in production of plywood,
or even drying of lumber in a simple bench-scale experiment.
However, knowledge of the reactions of these comounds at
9
VOL. 33, NO. 22, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 4 0 3 1

TABLE 4. Thermal Degradation Products of r-Terpinene
FIGURE 3. Limonene oxide from limonene.
FIGURE 4. trans-2-Hydroxy-3-caren-5-one from 3-carene.
FIGURE 5. Eucarvone from limonene.
wood flakes (at 90% moisture content dry basis) in a rotary
dryer with 40 000 cfm of air at an inlet temperature of between
400 and 550 °C. The temperature in the dryer would be
expected to drop to an outlet temperature of 100-130 °C
over a residence time of about 15 min. Evaporation of water
from the flake holds the temperature of the wood flake low
until the free water is removed (about 30% moisture content).
However, the wood temperature rises as it is dried over the
range of 30-5% moisture content in the later part of the
drying period.
When drying wood veneers for manurfacture of plywood,
a typical plant might dry 20 green ton of veneer (at 90%
moisture content), with air flowing over the veneer at
temperatures between 180 and 200 °C over a drying period
of 8-15 min. As in drying wood flakes, temperatures of the
wood veneers do not increase until the wood reaches the
fiber saturation point, and this zone gradually moves back
from the surface to the center of the veneer as it dries. Because
air is only vented periodically from the drying chamber, the
volatilized compounds are exposed to high temperatures but
also probably low oxygen levels until they are vented to the
atmosphere. Probably the most severe oxidation conditions
may be associated with veneer drying even though the initial
temperatures are much lower than for flake dryers because
of the comparatively long time periods the volatilized terpenes
are held in the vapor state at high temperatures.
FIGURE 1. p-Cymene from r-terpinene.
When drying lumber with “high temperature” schedules,
kiln temperatures above 100 °C can be employed over a period
of 24-30 h. Alternatively, low-temperature drying of nominal
2 in. framing lumber can involve a gradual increase in
temperature from about 70 to 90 °C over a total drying time
of 130 h. As in wood veneer drying, the lumber kiln is vented
intermittently to control the relative humidity throughout
the process. Because of the increased mass of porous material
involved, residence time for the terpenes within the drying
wood, probably with limited oxygen, is comparatively long.
FIGURE 2. 3,7-Dimethyl-6-oxo-2-octenal from r-terpinene.
120 °C (representing modest temperatures) together with
the availability of 1.24 mol of O2/ mol of terpene (representing
low levels of available oxygen) do provide a reasonable
reference point from which to consider what can happen in
all these wood-drying operations. For example, a typical
oriented strand board plant might dry 20 green ton/ h of
To interpret these results relative to actual wood-drying
operations, one must consider what proportion of volatile
terpenes remain within the porous wood structure and what
9
4 0 3 2 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 22, 1999

(6) Hakola, H.; Arey, J.; Aschmann, S. M.; Atkinson, R. J. Atmos.
Chem. 1994, 18, 75-102.
(7) Fraser, H. S.; Swann, E. P. Information Report VP-X-101;
Department of the Environment, Canadian Forestry Service,
Western Forest Products Laboratory: Vancouver, BC, 1972.
(8) Monroe, F. L.; Rasmussen, R. A.; Bamesburger, R. A.; Adams, D.
F. Report to Plywood Research Foundation; Washington State
University: Pullman, WA, 1972.
proportion has been volatilized and entered the dryer air
stream at various levels through the course of the drying
process. The rate at which the air stream is vented to the
atmosphere is also important. The conditions studied here
most closely represent the conditions seen by the terpenes
in lumber and veneer drying proceses.
The compounds selected for study provide models for
the structures of many other terpenes that may be subjected
to thermal oxidative rearrangement in processing wood or
other plant materials. Some generalizations may be useful
in providing clues to the type of products one might look for
in the vent gases from wood-drying operations, and such an
endeavor is the next logical step to take. Dehydrogenation
to form aromatic systems is a reaction shared by all of the
model systems studied. When a monocyclic six-membered
ring system with existing unsaturation is present, as in
R-terpinene and limonene, aromatization appears to be the
most common oxididation process. Oxidative cleavage of
double bonds to form the carbonyl compounds was observed
with three of these model compounds. Surely, where
aldehydes are formed, the corresponding carboxylic acids
would be expected and could probably be isolated as the
methyl esters by altering the workup procedure. All of the
model terpenes also showed some form of allylic oxidation,
and this suggests that such compounds should be expected
in the dryer effluents. Although not observed in our study,
the formation of epoxides, as seen with 3-carene and
R-terpinene, would probably lead to diols under wood-drying
conditions with the presence of acids and high moisture
found there.
(9) Zimmerman, P. R. Final Report to the Environmental Protection
Agency, Contract 68-02-2071, 1979.
(10) Graedel, T. E. Rev. Geophys. Space Phys. 1979, 17, 937-947.
(11) Schwartz, W. E.; Jones, P. W.; Riggle, C. J.; Miller, D. F. The
Organic Composition of Model Aerosols. Presented at Eastern
Analytical Symposium, New York City, November 1978.
(12) Gay, B. W., Jr.; Arnts, R. R. Proceedings of the International
Conference on Photochemical Oxidant Pollution and its Control;
EPA-600/ 3-77-001b; U.S. EPA: Washington, DC, 1977; pp 745-
751.
(13) Palen, E. J.; Allen, D. T.; Pandis, S. N.; Paulson, S. E.; Seinfeld,
J. H.; Flagan, R. C. Atmos. Environ. 1992, 26A, 1239-1251.
(14) Scheutzle, D.; Rasmussen, R. A. J. Air Pollut. Control Assoc. 1978,
28, 236-214.
(15) Duce, R. A. Pure Appl. Geophys. 1978, 116, 244-273.
(16) Dev, S. In Natural Products of Woody Plants, Vol. II; Rowe, J.
W., Ed.; Springer-Verlag: Berlin, 1989; pp 691-807.
(17) Zinkel, D. F., Russell, J., Eds. Naval Stores: Production, Chemistry,
Utilization; Pulp Chemicals Association: New York, 1988.
(18) Zinkel, D. F. In Natural Products of Woody Plants, Vol. II; Rowe,
J. W., Ed.; Springer-Verlag: Berlin, 1989; pp 953-978.
(19) Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43, 2923-
2925.
(20) Stahl, E., Ed. Thin-Layer Chromatography: A Laboratory
Handbook; Springer-Verlag: New York, 1969.
(21) Punsunvo, V. Ph.D. Dissertation, Mississippi State University,
1994.
(22) Hakola, H.; Arey, J.; Aschmann, S. M.; Atkinson, R. J. Atmos.
Chem. 1994, 18, 75-102.
(23) Bruns, W. D. P.; Carson, M. S.; Cocker, W.; Shannon, P. V. R. J.
Chem. Soc. C 1968, 3073-3079.
(24) Flynn, T.; Southwell, J. A. Phytochemistry 1987, 26, 1673-1686.
(25) Grosjean, D.; Williams, E. L., II; Grosjean, E.; Andino, J. M.;
Seinfeld, J. H. Environ. Sci. Technol. 1993, 27, 2754-2758.
(26) Arey, J.; Atkinson, R.; Aschmann, S. M. J. Geophys. Res. 1990, 95,
18539-18546.
Acknowledgments
This work was supported by USDA National Research
Initiative Competitive Grant Program, Grant 94-37500-1198.
Literature Cited
(1) Guenther, A.; Hewitt, C. N.; Erickson, D.; Fall, R.; Geron, C.;
Graedel, T.; Harley, P.; Klinger, L.; Lerdau, M.; McKay, W. A.;
Pierce, T.; Scholes, B.; Steinbrecher, R.; Tallamraju, R.; Taylor,
J.; Zimmerman, P. J. Geophys. Res. 1995, 100, 8873-8891.
(2) Adams, D. F.; Monroe, F. L.; Rasmussen, R. A.; Bamesburger,
W. L. Volatile Emissions form Wood; Fifth Washington State
University Symposium on Particle Board: Pullman, WA, June
1971.
(27) Fieser, L. F.; Fieser, M. Reagents for Organic Synthesis; John
Wiley and Sons: New York, 1967.
(28) Larson, R. A.; Weber, E. J. Reaction Mechanisms in Environmental
Organic Chemistry; Lewis Publishers: Ann Arbor, 1994.
(3) Cronn, D. R.; Truitt, S. G.; Campbell, M. J. Atmos. Environ. 1983,
17, 210-211.
(29) Jay, K.; Stieglitz, L. Air Pollut. Ecosyst. 1988, 542-547.
(4) Fehsenfeld, F.; Calvert, J.; Fall, J.; Goldan, P.; Guenther, A. B.;
Hewitt, C. N.; Lamb, B.; Liu, S.; Trainer, M.; Westberg., H.;
Zimmerman, P. Global Biogeochem. Cycles 1992, 6, 389-430.
(5) Grosjean, D.; Williams, E. L., II; Seinfeld, J. H. Environ. Sci.
Technol. 1992, 26, 1526-1533.
Received for review October 14, 1998. Revised manuscript
received July 6, 1999. Accepted August 25, 1999.
ES9810641
9
VOL. 33, NO. 22, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 4 0 3 3