Thermal behaviour of selected C10H16 monoterpenes

Article abstract of DOI:10.1002/ejoc.200600022

The presented work investigates the thermal behaviour of selected monoterpenes under various reactor temperatures and residence times (carrier gas, reactor inserts). In addition to the analysis of the liquid products by capillary GC and GC-MS, chemical derivatisation techniques (Diels-Alder reaction, hydrogenation) were used to identify the liquid-phase products. A thermal conversion of β-pinene (1), myrcene (2) and limonene (4) in a reaction network is presented and the experimental evidence for the formation of pyrolysis products by a biradical pathway is discussed. The reaction network was modified based on the identification of additional C₁₀H ₁₆ terpene isomers. Wiley-VCH Verlag GmbH & Co. KGaA, 2006.

Full text of DOI:10.1002/ejoc.200600022

FULL PAPER
DOI: 10.1002/ejoc.200600022
Thermal Behaviour of Selected C H Monoterpenes
1
0
16
[a]
[a]
[a]
[a]
Achim Stolle, Claudia Brauns, Matthias Nüchter, Bernd Ondruschka,*
[
b]
[c]
Werner Bonrath, and Matthias Findeisen
Keywords: Pyrolysis / Terpenes / Radical reactions / Rearrangements / Gas chromatography / Reaction mechanisms
The presented work investigates the thermal behaviour of
selected monoterpenes under various reactor temperatures
and residence times (carrier gas, reactor inserts). In addition
to the analysis of the liquid products by capillary GC and
GC-MS, chemical derivatisation techniques (Diels–Alder re-
action, hydrogenation) were used to identify the liquid-phase
products. A thermal conversion of β-pinene (1), myrcene (2)
and limonene (4) in a reaction network is presented and the
experimental evidence for the formation of pyrolysis prod-
ucts by a biradical pathway is discussed. The reaction net-
work was modified based on the identification of additional
C H16 terpene isomers.
10
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2006)
Introduction
It is well known that β-pinene (1) is thermally converted
[
1]
into myrcene (2). This reaction exemplifies the chemistry
of C H terpene isomers and is also interesting from a
1
0
16
mechanistic and thermodynamic viewpoint. Myrcene is an
important raw material for the synthesis of terpene alcohols
and aldehydes (e.g. citronellol and citral), which are relevant
[
2–4]
intermediates in the synthesis of vitamins A and E.
Pre-
liminary experiments revealed that in addition to myrcene
[
5–7]
(
2) many other C H monoterpenes are formed.
The Scheme 1. Biradical reaction intermediates in the thermal transfor-
1
0
16
detailed reaction mechanism of this thermal rearrangement mation of β-pinene (1).
to C H monoterpenes is not known. However, it seems
1
0
16
that differences in the experimental setup and residence
times are responsible for the contradicting yields and com-
positions of reaction products reported in the literature.
It is known that the pyrolysis of β-pinene (1) at tempera-
tures above 400 °C leads to a multitude of products that
arise from different parallel and consecutive reaction path-
ways. In addition to the open-chain main product 2, mono-
cyclic products like Ψ-limonene [p-mentha-1(7),8-diene
[
5–9]
Despite the fact that better analytical and technical meth-
ods have become available, very little work on the pyrolysis
of 1 has been published in recent times.[
10–12]
Most research-
[13]
[1,14]
(3)] and limonene (4) are also formed, see Scheme 1.
ers do agree that the formation of the reaction products
On the basis of our previous work experiences in the field
during the pyrolysis of 1 occurs via biradical intermediates
[15–18]
of hydrocarbon pyrolysis,
modern chromatographic
[
8]
(Scheme 1).
techniques were combined with chemical methods to gain
a closer insight into the reaction mechanism of such C H
16
10
terpene isomers. Herein we report on our results with the
envisioned and implemented experimental setups for the
thermal conversion of monoterpene hydrocarbons. These
findings can serve as a basis for future investigations deal-
ing with kinetic research and the trapping of transient inter-
mediates.
[
a] Institute for Technical Chemistry and Environmental Chemis-
try, University Jena,
Lessingstr. 12, 07743 Jena, Germany
Fax: +49-3641-948402
E-mail: bernd.ondruschka@uni-jena.de
b] DSM Nutritional Products, R&D Chemical Process Technol-
ogy,
[
[
P. O. Box 3255, 4002 Basel, Switzerland
c] Institute for Analytical Chemistry, University Leipzig,
Linnéstrasse 3, 04103 Leipzig, Germany
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
Results and Discussion
In order to study the influence of various process param-
eters (carrier-gas flow, temperature) on the pyrolysis of 1,
Eur. J. Org. Chem. 2006, 3317–3325
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3317

A. Stolle, C. Brauns, M. Nüchter, B. Ondruschka, W. Bonrath, M. Findeisen
FULL PAPER
a number of experiments were conducted (see Table 1). A creased by realising a high carrier gas flow. The throughput
comparison of the yields and selectivities of the collected was restricted to the flow rate of the carrier gas.
liquid products indicates that the process parameters and
other features connected with the experimental setup exhi-
bit a tremendous influence.[
19]
Table 1. Process parameters of selected pyrolysis experiments (T =
550 °C; substrate: β-pinene).
Experiment
1
2
3
HCP
empty[
0.15
120
11.8
2.2
HCP
packed
0.15
30
2.7
3.2
32%
93
Reactor
DGP[c]
a]
[b]
Flow rate (N
2
) [L/min]
1
30
0.6
Heated reactor volume [mL]
Average residence time[ [s]
Substrate flow rate [mmol/min]
Substrate flow rate/total flow rate
Yield liquid products [%]
d]
0.018
360 ppm
25%
96
[e]
–
β-Pinene in liquid product [%]
Ͻ5
Ͻ6
Ͻ0.01
[
a] High-concentration pyrolysis, empty quartz tube, see the follow-
ing text. [b] High-concentration pyrolysis, packed quartz tube, see
Exp. Sect. [c] Dilution-gas pyrolysis, see the following text. [d] For
the calculation of the average residence time, see Supporting Infor-
mation. [e] Not measurable under our conditions.
The observed differences are due to variations in carrier
gas and average residence times of the experimental setup
employed. In the case of high-concentration pyrolysis
(HCP), high substrate throughput was realised by fast and
direct evaporation of the starting material. The substrate
was passed through a heated evaporator (see Figure 1). It Figure 2. Schematic experimental setup for dilution-gas pyrolysis
(
DGP); 1 – rotameter for carrier gas, 2 – septum for dosing sub-
was shown in the experiments that high yields of liquid
products and a nearly quantitative conversion of starting
material are achieved by using a low carrier gas flow
strate, 3 – quartz ladle, 4 – electrical heating, 5 – quartz reactor,
– variable insert, 7 – exhaust for pyrolysis and carrier gas, 8 –
coolant (liquid N ), 9 – Dewar vessel, 10 – cold finger, 11 – liquid
product collector.
6
2
(Table 1).
Despite the fact that there were only minor differences in
the experimental setup (Figures 1 and 2), the differences in
Both pyrolysis techniques were comparable. A high
the average residence times for the two sets of pyrolysis ap- throughput was realised by using HCP (packed quartz reac-
paratus were significant (Table 1). In the case of dilution- tor), so that within short experimental times enough liquid
gas pyrolysis (DGP), the average residence time can be de- products for further investigations were generated. On the
Figure 1. Schematic experimental setup for high-concentration pyrolysis (HCP); 1 – rotameter for carrier gas, 2 – septum for dosing
substrate, 3 – substrate reservoir, 4 – quartz reactor, 5 – condenser 1 (air cooling), 6 – condenser 2 (cold trap), 7 – Dewar vessel, 8 –
coolant (liquid N
2
), 9 – jacketed coil condenser, 10 – exhaust for pyrolysis and carrier gas.
3318
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 3317–3325

Thermal Behaviour of Selected C10H16 Monoterpenes
FULL PAPER
other hand, because of the small amounts of starting mate- product 2 decreased dramatically with increasing tempera-
rial required for the DGP experiments, this method was ture, while it remained constant for the by-product limo-
suitable for screening the chemistry of such hydrocarbons nene (Figure 4).
at elevated temperatures.
Because of the slow and continuous evaporation of the
substrate and the low substrate concentration in the total
gas flow, reproducible experimental conditions in the pyrol-
ysis reactor were realised in the case of DGP. Thus, poly-
merisation reactions were also minimised. The residence
time of the starting material was determined from the car-
rier gas velocity and the surface/volume ratio. The selectiv-
ity of myrcene (2) increases with increasing surface/volume
ratio (Figure 3).
Figure 4. Selectivity for myrcene (2) and limonene (4) during the
pyrolysis of β-pinene (1) as a function of the pyrolysis temperature
and experimental setup, (carrier gas: N ), HCP – high-concentra-
2
tion pyrolysis (packed quartz tube), DGP – dilution-gas pyrolysis.
The conversion to known and unknown β-pinene pyroly-
sis products, besides 2, changes with rising temperature. In
order to obtain larger amounts of liquid pyrolysis products
for identification by chemical derivatisation, 1 was con-
verted at 550 °C using the HCP setup (packed quartz tube).
Comparison with the results obtained from the DGP setup
shows differences in the selectivity (Figure 4). The differ-
ences in product distribution resulted from the different car-
rier gas/substrate rates realised from both experimental set-
ups (Table 1). On summarising these observations, it was
concluded that myrcene (2) is preferentially formed if either
a low residence time combined with a low flow rate and
Figure 3. Selectivity of selected liquid products of β-pinene (1) py-
rolysis depending on the surface/volume ratio (dilution-gas pyroly-
2
sis, T = 500 °C, carrier gas: N , flow rate: 1 L/min).
In further investigations using the reactor setup of DGP, temperature (450 °C), or longer residence times and higher
compound 1 did not react below 350 °C under the condi- temperatures (550 °C) are realised.
tions investigated. Apparently, this temperature was too low
The identification of the unknown products was rather
to induce ring conversions. At 450 °C, no other side-prod- difficult and required complex analytical approaches. For
uct was formed, besides Ψ-limonene (3) and limonene (4), this purpose, a test mixture was prepared from readily avail-
in noteworthy concentrations. Increasing the temperature to able monoterpene hydrocarbons and analysed with GC-
650 °C or even higher leads to the formation of aromatics FID and GC-MS. Comparative gas chromatograms are de-
and coke. Overall, the yield and selectivity of the target picted in Figure 5.
Figure 5. Comparison of gas chromatograms (GC-FID) of a test mixture containing monoterpene hydrocarbons and the liquid pyrolysis
products of β-pinene; 1 – β-pinene, 2 – myrcene, 3 – Ψ-limonene, 4 – limonene, 5 – 5-ethylidene-1-methylcyloheptene (5-EMCHp), 6 –
+
+
iridane-1(6),8-diene, X1 – unknown (M : m/z = 136, C10
H16), X2 – unknown (M : m/z = 134, C10H14), a – α-pinene, b – camphene, c –
trans-isolimonene, d – α-phellandrene, e – α-terpinene, f – p-cymene, g – trans-ocimene, h – γ-terpinene, i – terpinolene, k – alloocimene.
Eur. J. Org. Chem. 2006, 3317–3325
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3319

A. Stolle, C. Brauns, M. Nüchter, B. Ondruschka, W. Bonrath, M. Findeisen
FULL PAPER
Figure 5 shows that besides the target compound 2 and thus a 3:1 mixture of cis/trans-p-menthane (4-methyl-1-iso-
limonene (4), many other products were formed during the propylcyclohexane) was obtained. These findings are in
pyrolysis of 1. Most of the products were identified as good agreement with results derived from the hetero-
C H isomers by GC-MS. Mass spectrometry also re- geneous hydrogenation of the orange oil with PtO as the
1
0
16
2
[
20]
vealed one compound with a molar mass that is 2 units catalyst.
In a side reaction, p-cymene was formed. The
smaller than that of the starting material (X2, Figure 5). oxidation of 4 to p-cymene under reducing reaction condi-
The structure has not yet been identified. However, initial tions can be explained by Pd-catalysed disproportion reac-
[
21]
data does not indicate aromatic structures like that found tions as reported in the literature.
in p-cymene (4-methyl-1-isopropylbenzene). Combined re- In contrast to 4, compound 2 is a triply unsaturated
sults of GC-FID experiments and spectroscopic data (GC- open-chain molecule containing one conjugated double-
MS) only provide evidence for the structures of C H iso- bond system. Similar to 4, hydrogenation of 2 takes place
1
0
16
mers. To extend these results, chemical derivatisation reac- stepwise producing several C H and C H hydro-
10
18
10 20
tions were conducted.
carbons as intermediates (Figure 7). At the end of the hy-
In the first step, a Diels–Alder (DA) reaction was carried drogen addition 2,6-dimethyloctane was formed.
out in order to gain a closer look into the diversity of
known and unknown hydrocarbon structures. Hydro-
carbons containing a 1,3-diene system, such as compound
2
or α-phellandrene, undergo a DA reaction, thus making
it an adequate instrument to characterise the mixture of
pyrolysis products. Depending on the solvent and the dien-
ophile, quantitative conversion was observed over a very
short time range using 2 and α-phellandrene as the sub-
strate. Similar reactions of liquid pyrolysis products re-
vealed only one DA-active compound – myrcene (2). As
a consequence of this result, the formation of monocyclic
hydrocarbons with a 1,3-diene system (e.g. α-phellandrene,
α-terpinene) by thermal rearrangement of 1 can be excluded
Figure 7. Liquid-phase hydrogenation of myrcene (2) in ethyl ace-
(
see Supporting Information).
tate at room temperature and atmospheric pressure as a function
The combination of GC-MS analyses and the heteroge- of reaction time (hydrogenation catalyst: Pd on a porous glass sup-
port).
neously catalysed hydrogenation of unsaturated hydro-
carbons in ethyl acetate as the solvent is an established and
powerful method for structural investigations. In order to
obtain an overview of the possible hydrogenation products,
pure compounds (1, 2 and 4) were hydrogenated in addition
to the liquid pyrolysis products received from HCP of hy-
drocarbon 1.
When β-pinene (1) was used as the starting material, the
hydrogenation catalyst isomerised the double bond of 1 in
the first step, thus leading to α-pinene. A mixture of cis-
and trans-pinane (2,6,6-trimethylbicyclo[3.1.1]heptane) was
obtained after hydrogenation of α-pinene (Scheme 2). The
Liquid-phase hydrogenation of compound 4 revealed
that hydrogen addition takes place in two steps (Figure 6).
In the first step, C H hydrocarbons were formed by hy-
1
0
18
drogenation of the less hindered exocyclic double bond.
Subsequently, the remaining double bond was reduced, and ethyl acetate.
Scheme 2. Liquid-phase hydrogenation products of β-pinene (1) in
Figure 6. Liquid-phase hydrogenation of limonene (4) in ethyl acetate at room temperature and atmospheric pressure as a function of
reaction time (hydrogenation catalyst: Pd on a porous glass support).
3320
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 3317–3325

Thermal Behaviour of Selected C10H16 Monoterpenes
FULL PAPER
isomerisation of either pinene isomer into the other is a carbons (e.g. α-phellandrene, α-terpinene, terpinolene) were
well-known and technically important reaction.[
22]
formed during the thermal isomerisation of compound 1.
[
5–9]
Our findings are in good agreement with research con-
As described in the literature,
the thermal rearrange-
ducted on the hydrogenation of α-pinene.[ The hydrogen- ment of compound 1 leads to products 2–4. Apparently, the
ation of α-pinene is an important step in the technical syn- conversion proceeds via biradical intermediates as shown in
thesis of linalool. In situ reduction with hydrogen (Li/ Scheme 1. A concerted mechanism (by cycloreversion of the
C H NH ) gives a mixture of cis- (56–57%) and trans-pi- cyclobutane ring in β-pinene) can be excluded, because of
23]
2
5
2
[
24]
nane (43–44%).
The heterogeneous hydrogenation leads the product distribution and DA reactivity: in the case of
to mixtures of pinane isomers with the cis selectivity rang- such a concerted reaction, two DA-active compounds
[
19]
ing from 49 up to 99% depending on the catalyst (Pt/C, Pd/ should have been observed. However, the liquid pyrolysis
[
24]
C or Ni), the H pressure and the reaction temperature.
mixture yielded only one product with the 1,3-diene sys-
2
Liquid-phase hydrogenation in ethyl acetate at ambient tem – hydrocarbon 2. Therefore, the initial reaction and
pressure and temperature with Pd on a porous glass sup- thermal rearrangement of 1 must proceed via biradical in-
port gave mixtures with a 75% cis selectivity.
termediates.
After the hydrogenation of the liquid pyrolysis product
It is also possible that the pyrolysis of hydrocarbon 1
derived from HCP of 1, all hydrogenated products de- progresses via lower molecular intermediates, e.g. isoprene
scribed above were identified in the reaction mixture, either (Scheme 4). This can be elucidated by pyrolysing 3,3,10,10-
2
by mass spectra (GC-MS) or by a comparison of the reten- [ H ]-β-pinene (β-pinene-D ). Thermal conversion of β-pin-
4
4
tion times of the pure compounds. Comparing the GC-MS ene-D would lead to a mixture of un-, tetra- and octadeu-
4
data of the crude and hydrogenated pyrolysis mixture al- terated products (statistical distribution: D /D /D = 1:2:1)
0
4
8
lowed for the identification of two other C H hydro- if the reaction involved lower molecular fragments. It was
1
0
16
carbons: 5-ethylidene-1-methylcycloheptene (5, 5-EMCHp) shown that pyrolysis products with various amounts of deu-
and iridane-1(6),8-diene (6) (Scheme 3).
terium (DGP of 1 and β-pinene-D in a 1:1 ratio) were sep-
arated in GC analyses.
4
[
25]
Scheme 3. Further thermal isomerisation products from β-pinene
(
1
1) pyrolysis [5-ethylidene-1-methylcycloheptene (5) and iridane-
Scheme 4. Hypothetical reaction pathway via lower molecular in-
termediates.
(6),8-diene (6)].
In summary, the combined experimental studies (DA, hy-
drogenation reactions), comparison with the literature val-
The formation of un- and octadeuterated pyrolysis prod-
ues, and the results of the GC-FID and GC-MS investi- ucts was excluded, because the mass spectrum of the pyroly-
gations allowed for the identification of the majority of the sis products of β-pinene-D did not contain signals with
4
rearrangement products obtained from β-pinene pyrolysis. m/z = 136 and 144 (Figure 8). A reaction pathway involving
By a comparison with the reference compounds it can be lower molecular reaction intermediates was thus ruled out.
ruled out that camphene and further menthadiene hydro- Therefore, the thermal gas-phase isomerisation of 1 pro-
Figure 8. Selected ion chromatograms (GC-MSD) obtained from the dilution gas pyrolysis (DGP) of β-pinene-D
gas: N , flow rate: 1 L/min), 2 – myrcene-4,4,10,10-D , 3 – Ψ-limonene-2,2,7,7-D , 4 – limonene-6,6,7,7-D , 5 – 5-ethylidene-1-methylcy-
loheptene-4,4,6,6-D , 6 – iridane-1(6),8-diene-5,5,6,6-D , X1 – unknown (M : m/z = 140, C10 ).
4
, (T = 500 °C, carrier
2
4
4
4
+
4
4
12 4
H D
Eur. J. Org. Chem. 2006, 3317–3325
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3321

A. Stolle, C. Brauns, M. Nüchter, B. Ondruschka, W. Bonrath, M. Findeisen
FULL PAPER
ceeds by a monomolecular mechanism under the conditions
investigated.
On the basis of these results, the formation of thermal
isomerisation products of 1 must proceed via biradical in-
termediates. The formation of 3 and 4 from biradicals 1a
and 1b, respectively (Scheme 1), can be explained by a [1,5]-
Scheme 6. Formation of 5-EMCHp by an ene reaction from myr-
cene.
(3) and [1,7]-H shift (4). Target product 2 was formed either
from biradical 1a or from 1b by homolytic bond cleavage
followed by (concerted) intramolecular radical recombina-
tion. Compounds 2–4 were directly formed from the sub-
strate 1, and are thus referred to as primary pyrolysis prod-
ucts (ppp).
Cope rearrangement of iridane-1(6),8-diene (6) via radical
intermediates that are stabilised by allylic rearrangement
could also lead to 5-EMCHp.
Furthermore, during the investigation of the thermal
conversion of 1, it was observed that the ppp undergo con-
secutive reactions. The pyrolysis products 5 and 6 were
formed from target product 2. The occurrence of 6 in con-
nection with β-pinene pyrolysis was not found in the litera-
ture. It is known that pinene analogues (e.g. nopinone, pin-
ane, pinan-2-ol) undergo cyclisation reactions involving
open-chain intermediates (analogues of 2) and lead to re-
lated substances as shown in Scheme 5.[
10,26–29]
From a
combination of mass spectra and hydrogenation experi-
ments it was concluded that the thermal conversion product
6
was formed from primary pyrolysis product 2 by an ene
reaction.
Scheme 7. Cope rearrangement (A) and formation of 5-EMCHp
by a nonclassical Cope rearrangement (B) of iridane-1(6),8-diene,
see ref.[
26]
Thermal isomerisation of iridane-1(6),8-diene to 5-
EMCHp can be explained by the nonclassical pathway of
the Cope rearrangement (Scheme 7). The unusual draw-
ing of the starting molecule [iridane-1(6),8-diene] and the
Scheme 5. Cyclisation of pinene analogues of the related cyclopen-
tane derivatives by an ene reaction.
[
26]
The formation of compound 5 (5-EMCHp), which was target 5-EMCHp in Scheme 7 illustrate the similarity be-
identified in the liquid pyrolysis-product mixture of sub- tween the classical and non-classical Cope rearrangements.
strate 1, can be explained in different ways. According to Probably both reaction pathways participate in the forma-
the literature, 5-EMCHp is built from compound 2 by an tion of 5-ethylidene-1-methylcycloheptene (5).
ene reaction as depicted in Scheme 6.[
30]
The composition profile of the identified liquid pyrolysis
Another possible reaction pathway leading to 5-EMCHp products derived from 1 indicates a strong dependence con-
could be based on the Cope rearrangement (Scheme 7, A). cerning the reaction temperature for the DGP setup (Fig-
In addition to the classical reaction pathway, a nonclassical ure 9). On the one hand, increasing the temperature de-
Figure 9. Selectivity of liquid pyrolysis products of β-pinene (1) as a function of the reaction temperature, (dilution-gas pyrolysis, carrier
gas: N
2
, flow rate: 1 L/min).
3322
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 3317–3325

Thermal Behaviour of Selected C10H16 Monoterpenes
FULL PAPER
Figure 10. Content of myrcene (2) and Ψ-limonene (3) in the pyrolysis product and selectivities of compounds 5 and 6 as a function of
the reaction temperature (dilution-gas pyrolysis, starting material: 2, carrier gas: N , flow rate: 1 L/min).
2
creased the amount of collected liquid pyrolysis products. haviour of terpene hydrocarbons. It was shown that chemi-
On the other hand, the selectivity of the target product 2 cal derivatisation techniques were as effective as modern
decreased and approached the amount of the other prod- gas-chromatographic experiments. Together they provide
ucts.
The results obtained from the pyrolysis of myrcene (2) position of hydrocarbon mixtures.
and high-temperature pyrolysis of 1 are in agreement. Hy- In agreement with previous results,
the essential tools needed for the identification of the com-
[
5,8,10,27,30]
the main
drocarbon 2 was not noticeably transformed until the reac- products [myrcene (2), Ψ-limonene (3) and limonene (4)]
tion temperature reached 470 °C as shown in Figure 10, and from β-pinene pyrolysis were identified. The existence of 5-
the concentration decreased with increasing temperature. ethylidene-1-methylcycloheptene (5) was also confirmed.[
The selectivity, corrected for the conversion of 2, for the Within this investigation, a new substance in the thermal
formation of 5 and 6 were in the same range as in the pyrol- network of C H hydrocarbons – iridane-1(6),8-diene
30]
10
16
ysis of 1.
(6) – was detected. The internal crosslink of products start-
The occurrence of Ψ-limonene (3) was also observed in ing from β-pinene (1) is depicted in Scheme 8. It was shown
the reaction of 1 at 450 °C. In contrast to the pyrolysis of that the formation of primary pyrolysis products (2–4) does
1, compound 3 is not a thermal isomerisation product of 2, not proceed by the concerted mechanism or via lower mo-
but an inseparable impurity found in the starting material. lecular weight fragments (isoprene). Rather, it appears to
For the definite identification of 3, the reaction mixture progress via joint biradical reaction intermediates. Second-
of the myrcene pyrolysis was heated at 100 °C for 24 h. Un- ary pyrolysis products (5, 6) were formed in consecutive re-
der these reaction conditions, 2 dimerised by a DA reaction. actions.
Compound 3 was separated from the myrcene dimer by Ku-
gelrohr distillation and was characterised by a combination
of analytical methods based on GC, GC-MS and HPLC-
NMR. Here it should be noted that the findings are in good
agreement with results obtained from the pyrolysis of hy-
drocarbon 1 in a microreactor at 510 °C under normal pres-
[
12]
sure.
The spectroscopic data are comparable with the
[31]
data from the literature.
No conversion of limonene (4) was detected at 450 °C
Figures 3, 4 and 9). Hydrocarbon 4, like 2, seems to be
(
thermodynamically stable under the experimental condi-
tions investigated. Therefore, additional pyrolysis experi-
ments were carried out at temperatures of 550 and 600 °C.
Unconverted 4 was still found in the pyrolysis products at
Scheme 8. Product network of β-pinene pyrolysis (legend: see Intro-
duction and Scheme 3).
550 °C. This observation additionally confirms the thermo-
dynamic stability of 4. A temperature increase to 600 °C
yielded monoaromatic products such as toluene or xylene.
Considering the product distribution of β-pinene pyroly-
sis, a concerted reaction mechanism can be ruled out. The
reaction proceeds through biradical intermediates. Ad-
ditional experiments with other/additional carrier gases
Summary
Two efficient and complementary experimental setups (H , D , toluene) or trapping methods (very low tempera-
2
2
were established for the investigation of the gas-phase be- tures, spin traps) may yield further information on the bi-
Eur. J. Org. Chem. 2006, 3317–3325
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3323

A. Stolle, C. Brauns, M. Nüchter, B. Ondruschka, W. Bonrath, M. Findeisen
FULL PAPER
radical reaction mechanism. On the basis of the pyrolysis red-heated quartz reactor with an approximate length of 40 cm.
The pyrolysis temperature was controlled with an Ni/CrNi thermo-
findings, the chemical derivatisation and the gas chromato-
couple placed in the middle of the reactor. The temperature gradi-
graphic results presented within this paper, kinetic measure-
ent (see Supporting Information) reveals that the pyrolysis oc-
ments are planned in order to gain further insight into the
curred in a 20 cm long section of the whole reactor. In order to
gas-phase behaviour of β-pinene.
increase the reactor surface, the quartz tube was packed with small
pieces of quartz glass calcinated in an air stream at 900 °C. The
condenser system is shown in Figure 1. The second condenser was
designed as a cold trap (liquid nitrogen) to allow for complete con-
Experimental Section
densation of the pyrolysis products. A small amount of the product
collected (50 µL) was diluted in 1 mL of DCM and analysed by
GC-FID and GC-MS after addition of 3 µL of hexadecane as an
internal standard. After every pyrolysis experiment with the HCP
setup, the reactor was cleaned in an air stream at 900 °C.
General Remarks: β-Pinene (1, ca. 99%), myrcene (2, ca. 80%) and
limonene (4, ca. 97%) were purchased from Fluka. The purity was
determined by capillary gas chromatography. Dichloromethane
(DCM) was used for the dilution of liquid pyrolysis products. All
other chemicals were readily available from commercial sources and
were used without further purification. Analyses were carried out
with a 6890 Series GC and 5890 Series II/5972 Series MSD GC
from Agilent Technologies. Hexadecane (3 µL) was used in all ex-
periments as an internal standard. Products were identified by com-
paring either the retention time and/or mass spectra of pure refer-
Diels–Alder (DA) Reaction: The DA reactions were performed in
two-necked flasks (100 mL) with reflux condensers and septums for
sampling. Maleic anhydride (after purification by recrystallisation
from chloroform) and tetracyanoethylene were used as dienophiles.
Ethyl acetate (AcOEt), benzene and dichloromethane were used as
solvents without further purification. The starting material (myr-
cene or α-phellandrene; 1.50 g, 11 mmol; 2 g in the case of the py-
rolysis mixture) was dissolved in the solvent (AcOEt, benzene or
DCM; 15 mL) together with hexadecane (200 µL) as an internal
standard. After adding the dienophile (12 mmol) and further sol-
vent (15 mL), a sample (500 µL) was taken from the reaction mix-
ture. While refluxing for 2–4 h, samples were taken every 30 min
and were analysed by GC-FID.
ence compounds. GC-FID: HP 5, 30 m×0.32 mm×0.25 µm, H
2
–
5
psi, program: 35 °C (hold 1 min), 4 K/min up to 80 °C, 4.5 K/min
up to 90 °C, 35 K/min up to 280 °C (hold 3 min), injector tempera-
ture: 250 °C, detector temperature: 280 °C. GC-MSD: HP 5,
3
5
0 m×0.32 mm×0.25 µm, He – 7 psi, program: 55 °C (hold 1 min),
K/min up to 150 °C, 20 K/min up to 280 °C (hold 5 min), injector
temperature: 280 °C, EI (70 eV). NMR spectra were recorded with
a Bruker Avance 200 MHz system (measurement frequency was
2
Hydrogenation: Liquid-phase hydrogenation was carried out in
ethyl acetate as solvent. The purity of the hydrogen used was higher
than 99.998%. The catalyst consisted of a porous glass support
coated with palladium.
00 MHz for 1H NMR and 50 MHz for C NMR) in a 5-mm
13
tube at room temperature. Measurements were carried out using
trichloromethane-D (CDCl ) as solvent. IR spectra were measured
3
with a Perkin–Elmer FT-IR spectrum 100 series device equipped
with a universal ATR sampling accessory.
(
a) Preparation of Catalyst: Silica was the main component (Ͼ95%)
®
of the porous glass support TRISOPOR (VitraBio GmbH, Stein-
ach, Germany). In a round-bottom flask (100 mL), Pd(OAc)
20.2 mg, 0.09 mmol) was dissolved in dichloromethane (50 mL).
Pyrolysis: The investigated C10H16 hydrocarbons were pyrolysed
over the temperature range 350–750 °C using two different experi-
mental setups explained below and presented in Figures 1 and 2.
2
(
The porous glass support (1 g) was added and the solvent was com-
pletely removed in vacuo. The coated glass was calcinated at 300 °C
for 3 h (heating to 300 °C in 20 min) in the muffle furnace “mls
1200 pyro” (MLS GmbH, Leutkirch, Germany).
(a) Dilution-Gas Pyrolysis (DGP): Pyrolyses were carried out with
an electrically heated quartz tube with a length of 50 cm and with
a pyrolysis zone of approximately 20 cm (Figure 2).[ In all experi-
ments, oxygen-free nitrogen with a purity of 99.999% was used as
the carrier gas. Residence times were varied by either selecting dif-
ferent carrier gas flow rates (0.5, 1 and 1.2 L/min) or by variation
of the surface/volume ratios of the reactor (see Supporting Infor-
mation). The starting material was introduced onto a quartz ladle
at the top part of the pyrolysis apparatus using a glass syringe (50
or 100 µL). The starting material was carried along with the nitro-
gen stream into the reactor. Vaporisation of the starting material
was supported by heating the ladle to 50 °C with a hot blast. Pyrol-
ysis products were collected in a cold trap (liquid nitrogen) and
were dissolved in 1 mL of DCM. The liquid products obtained
were analysed by GC-FID and GC-MS using 3 µL of hexadecane
as an internal standard.
15]
(
(
b) Experimental Setup: In a Schlenk flask, the starting material
β-pinene, myrcene, limonene: 2.04 g, 15 mmol; 3 g in the case of
the pyrolysis mixture) was dissolved together with hexadecane
200 µL) as an internal standard in ethyl acetate (60 mL). After
(
adding the catalyst (0.4 g), a sample (500 µL) was taken. The flask
was connected to the hydrogen reservoir by a plug valve and the
reaction vessel was flushed with hydrogen (3×400 mL). Hydrogen-
ation was carried out at room temperature and atmospheric pres-
sure in a shaking apparatus “HS 501 digital” (IKA Labortechnik,
Staufen, Germany). Every 30 min (after 180 min every 60 min) a
sample (500 µL) was taken. The hydrogenation process was ob-
served by consumption of hydrogen and by analysing the samples
with GC-FID and GC-MS (Figures 6 and 7). After the reaction
was finished, the catalyst was filtered off, washed with ethyl acetate
(2×5 mL) and dried at 105 °C before reusing it.
(b) High-Concentration Pyrolysis (HCP): The pyrolysis setup con-
sists of three different structural elements: an evaporator, the pyrol-
ysis reactor and a condenser. The evaporator is schematically de-
picted in Figure 1. In order to secure the complete and fast vapori-
sation of the starting material, the lower part of the evaporator was
heated by an air fan up to 250 °C. Gaseous starting material was
carried along with the nitrogen stream (0.15 L/min) into the reac-
tor. The central piece of this experimental setup was the infrared
furnace “IRF 10” (Behr Labortechnik GmbH, Düsseldorf, Ger-
many), which was integrated into the setup consisting of an infra-
Synthesis
6,6-Dimethylbicyclo[3.1.1]heptan-2-one (Nopinone): Potassium per-
manganate (26.4 g, 0.17 mol) was added to a mixture of β-pinene
(6 g, 44 mmol), aluminium oxide (60 g, 0.59 mol) and water (5 g,
0.28 mol). The mixture was transferred into a beaker (agate) with
seven agate balls, inserted into a planetary ball mill “Pulverisette
5” (Fritsch GmbH, Idar-Oberstein, Germany) and milled at
3324
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 3317–3325

Thermal Behaviour of Selected C10
H16 Monoterpenes
FULL PAPER
300 rpm for 20 min. Subsequently, the reaction mixture was ex-
tracted with DCM (400 mL). After removal of the solvent, the
crude product was purified by column chromatography (silica gel
[1] Kirk-Othmer, Encyclopedia of Chemical Technology (Chief
Eds.: J. I. Kroschwitz, H. Howe-Grant), 4th ed., John Wiley &
Sons, New York, 1997, vol. 23, pp. 833–882.
40, 0.063–0.2 mm, Merck) using a mixture of ethyl acetate/n-hex-
[
2] R. P. Sood, B. Singh, V. Sigh, J. Essent. Oil Res. 1992, 4, 425–
ane (1:3) as solvent. Yield: 3.4 g per beaker (56% based on β-pinene
2
0
426.
after purification). n
D
= 1.480 [reference (1R)-(+)-nopinone:
[
[
3] B. M. Lawrence, Prog. Essent. Oils 1997, 3, 49–56.
4] C. Merceir, P. Chabardes, Pure Appl. Chem. 1994, 66, 1509–
1
1
1
2
2
2
C
.479]. H NMR (CDCl ): δ = 0.85 (s, 3 H), 1.33 (s, 3 H), 1.59 (d,
3
H), 1.96–2.06 (m, 2 H), 2.24–2.25 (m, 1 H), 2.35–2.38 (dd, 1 H),
): δ = 21.3, 21.9, 25.3,
1518.
.51–2.58 (m 3 H) ppm. 13C NMR (CDCl
3
[
5] R. Rienäcker, Brennstoffchemie 1964, 7, 20–23.
5.7, 32.7, 40.2, 41.0, 57.7, 214.8 ppm. IR (ATR): ν˜ = 2949, 2928,
837 (νaliph. C–H), 1706 (νC=O), 1459 (δC–H) cm . MS (EI 70 eV,
9
H14O): m/z (%) = 139 (0.9) [M + 1], 138 (8.6) [M ], 123 (16.5),
[6] L. A. Goldblatt, S. Palkin, J. Am. Chem. Soc. 1941, 63, 3517–
–1
3522.
+
+
[7] R. L. Burwell, J. Am. Chem. Soc. 1951, 73, 4461–4462.
[
8] A. A. Frost, R. G. Pearson, Kinetik und Mechanismus homo-
gener chemischer Reaktionen, Verlag Chemie, Weinheim, 1964,
pp. 349–354.
109 (26.3), 95 (40.8), 83 (100), 81 (37.3), 67 (23.1), 55 (41.4).
3
,3,10,10-Tetradeuterio-6,6-dimethyl-2-methylenebicyclo[3.1.1]hep-
tane (β-Pinene-D
condenser, gas inlet and septum, a mixture of methyltriphenylphos-
phonium bromide and sodium amide (13 g, 31.5 mmol; Fluka) was
4
): In a three-necked flask (100 mL) with reflux
[9] J. E. Hawkins, J. W. Vogh, J. Phys. Chem. 1953, 57, 902–905.
[10] L. Lemée, M. Raiter, J. G. Duboudin, B. Delmond, Synth.
Commun. 1995, 25, 1313–1318.
[
[
[
11] C. Coeur, V. Jakob, I. Denis, P. Foster, J. Chromatogr. A 1997,
86, 185–187.
12] J. Y. Lou, H. Z. Wang, S. J. Peng, Linchan Huaxue Yu Gongye
000, 20, 47–50; Chem. Abstr. 2001, 134, 30604.
13] To avoid long and confusing names for the pyrolysis products
in the following text, the terpene nomenclature is used.
dissolved in dried DMSO-D
flushing nitrogen through the mixture, nopinone (2.5 g, 18 mmol),
dissolved in DMSO-D was slowly added through a glass syringe.
After complete addition, the reaction mixture was stirred at 50 °C
for 4.5 h. The colour of the reaction mixture changed from pale
yellow to auburn and sodium bromide precipitated. The reaction
was monitored by taking samples and analysing them by GC-MS.
The reaction was quenched with water (300 mL) and the mixture
extracted with n-pentane (5×20 mL). The combined organic ex-
tracts were washed with water and dried with anhydrous sodium
6
(15 mL; Aldrich). While stirring and
7
6
2
[14] H. Kwart, K. King, Chem. Rev. 1968, 68, 415–447.
[15] G. Zimmermann, M. Nüchter, M. Remmler, M. Fineisen, H.
Hopf, L. Ernst, C. Mlynek, Chem. Ber. 1994, 127, 1747–1753.
[
16] B. Ondruschka, Zum Cyclisierungs- und Aromatisierungsver-
halten von mehrfach ungesättigten Kohlenwasserstoffen in der
Gasphase, PhD Thesis (Habilitation), Leipzig, 1989.
17] S. Hagen, U. Nüchter, M. Nüchter, G. Zimmermann, Tetrahe-
dron Lett. 1994, 35, 7013–7014.
3
sulfate. The mixture was then washed with an NaHSO solution
and the solvent removed in vacuo. The crude product was purified
by column chromatography (silica gel 40) using n-pentane as sol-
[
[
18] G. Zimmermann, B. Ondruschka, M. Nüchter, F. D. Kopinke,
M. Remmler, J. Prakt. Chem. 1994, 336, 415–420.
2
0
vent. Yield: 1.74 g (69% based to nopinone). n
D
= 1.480 [reference
): δ = 0.73 (s, 3 H), 1.24 (s,
H), 1.40 (d, 2 H), 1.83 (s, 2 H), 1.97 (m, 1 H), 2.46 (m, 1 H) ppm.
IR (ATR): ν˜ = 2945, 2920, 2869 (νaliph. C–H), 2299, 2200, 2101 (νaliph.
1
(–)-β-pinene: 1.479]. H NMR (CDCl
3
[19] A. Stolle, Beitrag zum Pyrolyseverhalten von ß-Pinen – Tren-
nung von Pyrolysegemischen durch chemische Derivatisierung,
Diploma Thesis, Jena, 2005.
[20] E. E. Stashenko, J. Chromatogr. A 1996, 752, 217–222.
[21] H. E. Eschinazi, E. D. Bergmann, J. Am. Chem. Soc. 1950, 72,
5651–5654.
3
–
1
C–D/C=D), 1607 (νC=C), 1457 (δC–H), 720 (δCH2), 684 (δCD2) cm .
+
MS (EI 70 eV, C10
H
12
D
4
): m/z (%) = 141 (0.9) [M + 1], 140 (9.2)
+
[
M ], 125 (12.5), 97 (100), 96 (48.8), 95 (22), 81 (16.6), 79 (17.5),
[
[
22] K. A. D. Swift, Top. Catal. 2004, 27, 143–155.
69 (51.2).
23] V. A. Semikolenov, I. I. Ilyna, I. L. Simakova, Appl. Catal., A
2
001, 211, 91–107.
24] W. Cocker, P. V. R. Shannon, P. A. Staniland, J. Chem. Soc. C
966, 1, 41–47.
Supporting Information (see footnote on the first page of this arti-
cle): Mass spectra, properties of the reactor inserts used for the
DGP setup, temperature gradient in the pyrolysis reactor of the
HCP setup and calculation procedure for the average residence
times.
[
[
[
1
25] C. Brauns, Pyrolyseverhalten isotopmarkierter β-Pinenderivate
bei 500 °C, PhD Thesis, Jena, 2005.
26] H. Strickler, G. Ohloff, E. Kováts, Helv. Chim. Acta 1967, 50,
7
59–798.
[
[
27] C. F. Mayer, J. K. Crandall, J. Org. Chem. 1970, 35, 2688–2690.
28] W. D. Huntsman, V. C. Solomon, D. Eros, J. Am. Chem. Soc.
1958, 80, 5455–5458.
Acknowledgments
[
[
29] J. Nowicki, Molecules 2000, 5, 1033–1050.
30] J. de P. Teresa, I. S. Bellido, M. R. Alberdi Albistegui, A.
San Feliciano, M. G. Benito, Anal. Quim. 1978, 74, 305–310.
31] D. J. Brecknell, R. M. Carman, R. A. Edwards, K. A. Hans-
ford, T. Karoli, W. T. Robinson, Aust. J. Chem. 1997, 50, 689–
We gratefully acknowledge Dr. Annegret Stark (Jena) for the sup-
port given. We thank Sebastian Losse (Jena) for performing the
hydrogenation experiments in his laboratory. For supporting us
with technical help (GC, etc.) we gratefully acknowledge the techni-
cal staff of the Institute for Technical Chemistry and Environmen-
tal Chemistry, especially Antje Tied and Gisela Gottschalt.
[
700.
Received: January 13, 2006
Published Online: May 10, 2006
Eur. J. Org. Chem. 2006, 3317–3325
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3325