Novel route to a fruitful mixture of terpene fragrances in particular phellandrene starting from natural feedstock geraniol using weak acidic boron based catalyst

Article abstract of DOI:10.1016/j.catcom.2012.05.021

Myrcene, ocimene and in particular phellandrene were selectively generated as products by dehydration of the natural feedstock geraniol over a weak acidic boron pentasil zeolite catalyst in a gas phase reaction. Additionally linalool was formed by rearrangement reaction. The total selectivity of these 4 terpenes is up to 99%.

Full text of DOI:10.1016/j.catcom.2012.05.021

Catalysis Communications 26 (2012) 214–217
Contents lists available at SciVerse ScienceDirect
Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom
Short Communication
Novel route to a fruitful mixture of terpene fragrances in particular phellandrene
starting from natural feedstock geraniol using weak acidic boron based catalyst
⁎
Matthias Eisenacher, Stefan Beschnitt, Wolfgang Hölderich
RWTH Aachen University, Worringerweg 1, D-52074 Aachen, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Myrcene, ocimene and in particular phellandrene were selectively generated as products by dehydration of
the natural feedstock geraniol over a weak acidic boron pentasil zeolite catalyst in a gas phase reaction. Additionally
linalool was formed by rearrangement reaction. The total selectivity of these 4 terpenes is up to 99%.
© 2012 Elsevier B.V. All rights reserved.
Received 31 March 2012
Received in revised form 23 May 2012
Accepted 28 May 2012
Available online 1 June 2012
Keywords:
Boron pentasil zeolite
Gas phase reaction
Heterogeneous catalysis
Natural feedstock
Terpenes
Phellandrene
1. Introduction
been investigated [11]. The conversion of pinenes to p-cymene has
also been investigated [12]. In addition, the treatment of linalool
Monoterpenes, commonly used as fragrances and flavors [1], are of
growing interest as industrial building blocks for fine chemicals [2].
Generally, monoterpenes such as citral, myrcene (2), phellandrene
(7) and pinenes are available from natural sources by distillation
and extraction [3]. Other monoterpenes can be acquired via hydro-
genation and pyrolysis processes. For instance, geraniol (1) is
obtained by hydrogenation of citral [4], whereas pyrolysis of
β-pinene leads to myrcene (2) [5]. In addition, myrcene can also be
acquired by the dimerization of isoprene [6]. Linalool (8) can be
obtained by reduction of dehydrolinalool by alkalimetals leading to
the formation of waste due to the formation of metal salts [7].
Ocimene can be derived from geranyl-diphenyl-phosphate [8], or
via a five step reaction starting from 2-methyl-3-vinyloxy-propen
[9]. The preparation of phellandrene starts from highly priced sabinene
and uses highly toxic compounds such as triphenylpyrylium tet-
rafluoroborate and 1,4-dicyanobenzene as auxiliaries in a photo-
chemical reaction [10]. These processes cause a lot of side products
and waste streams. Additionally, they are partially based on fossil
feedstock.
(8), geraniol (1), nerol, their acetates and their glucosides with
weakly acidic Broensted acids such as citric acid in the liquid has
been examined [13,14].
Acidic treatment of geraniol (1) in the gas phase over strong acidic
catalysts mainly led to the formation of acyclic products like myrcene
(2), and ocimene (3) (Scheme 1) [15], whereas monocyclic products
are basically formed if—instead of geraniol—nerol is used as the
starting material. The monocyclic terpenes limonene (4), terpinolene
(5) and α-terpinene (6) were expected as products. Due to this work a
ratio of 9:1 acyclic to monocyclic terpenes is expected if acidic catalysts
like zirconia phosphates, organo-substituted zirconia phosphates or
acidic alumina are used for the gas phase reaction of geraniol (1)
[15].
Several monoterpenes are easily accessible by separation from
natural feedstock. Thus, the current research is more focused on the
conversion of such compounds than on their synthesis. Consequently,
the investigation of catalytic transformation of easily available mono-
terpenes to less accessible and therefore more valuable derivates is of
high economical interest. Such a compound is certainly phellandrene.
In this protocol the investigation of various reaction parameters
for the conversion of geraniol (1) to different monoterpenes is
reported.
Transformation of monoterpenes over hydrogenation catalysts such
as CuO or Raney-Nickel [2] have been studied extensively. Moreover,
the liquid phase hydrogenation, isomerization and dehydrogenation
of limonene and derivatives with supported palladium catalysts has
2. Discussion and results
Looking for an appropriate acidic catalyst to convert geraniol (1), a
⁎
Corresponding author. Fax: +49 241 80 22291.
E-mail address: hoelderich@rwth-aachen.de (W. Hölderich).
strong acidic H-ZSM5 (SiO2/Al2O3 ratio=26) and a weak acidic boron
1566-7367/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.catcom.2012.05.021

M. Eisenacher et al. / Catalysis Communications 26 (2012) 214–217
215
Scheme 1. Acidic treatment of geraniol (1) [15].
pentasil zeolite (SiO2/B2O3 ratio=33) were used in initial experiments
under the same reaction conditions (Table 1). The X-ray diffraction
analyses suggest that both catalysts consist of well-crystallized zeolites.
The BET surface area was 338.0 m2/g for the boron zeolite and
370.6 m2/g for H-ZSM5, the micropore surface was 247.3 m2/g for the
boron zeolite and 218.2 m2/g for H-ZSM5.
In contrast to previous protocols [13–15], limonene (4) terpinolene
(5) and terpinene (6) were not found in the reaction mixtures. Instead
of these compounds phellandrene (7) was surprisingly found as the
only monocyclic terpene.
The use of H-ZSM5 led to huge amounts of CO and CO2 which were
identified by gas chromatography. Linalool (8) was just found in
traces. In contrast in the presence of a boron pentasil zeolite a fruitful
mixture of monoterpenes was found: phellandrene (7), myrcene (2),
ocimene (3) and linalool (8) were identified in the reaction mixture
(Scheme 2).
In order to understand the different product spectra obtained
from the weak acidic boron pentasil zeolite and the strong acid
H-ZSM5 catalyst, the differences between weak acidic boron pentasil
zeolite and strong acidic H-ZSM5 zeolite have been investigated by
ammonia-TPD and Fourier transformation infrared spectroscopy
(FT-IR). FT-IR measurements have been carried out by pyridine ad-
sorption experiments in order to study the surface acidity (Figs. 1
and 2) [16].
In order to optimize the selectivity for the formation of the different
monoterpenes 2, 3, 7, and 8 the influence of the reaction parameters
was investigated by means of a Design of Experiments. A response
surface design was chosen as suitable tool [17]. Reaction parameters
were varied in the ranges described in Table 2.
The results (Table 3) clearly illustrate that a total selectivity of up
to 99% to the four terpenes. The selectivity towards linalool (8) can be
increased to 25% if absolute pressure is diminished to 50 mbar.
To the best of our knowledge, formation of phellandrene (7) via
acidic treatment of geraniol (1) has not been previously reported. Selec-
tivity up to 27% towards phellandrene could be achieved.
The reaction mixture obtained in the present experiments usually
contained about 27% phellandrene (7) which results in a acyclic to
monocyclic ratio of 4:1. Previous work mentioned a 9:1 ratio for the
acidic gas phase treatment of geraniol [15].
The following general tendencies could be ascertained during our
experiments:
Increase of the temperature was reflected in the formation of de-
hydration products phellandrene (7), myrcene (2) and ocimene (3).
Low pressures enhanced the selectivity to the rearrangement product
linalool (8). Shorter residence times, due to increased N2-flow, lead to
higher selectivity regarding the dehydration products and linalool
(8). Longer retention time causes the formation of CO and CO2 in
higher quantities. Higher catalyst loadings lead to an increased for-
mation of myrcene (2). However, the other products were not formed
in higher amounts. These facts indicate that dehydration of geraniol
(1) to myrcene (2) might be kinetically controlled.
Ammonia-TPD showed that H-ZSM5 contains weak acidic sites
(desorption of NH3 at 230 °C) as well as strong acidic sites (desorption
of NH3 at about 500 °C) In contrast the boron pentasil zeolite has only
weak acidic sites (desorption of NH3 at about 210 °C). Therefore,
H-ZSM5 appears to be much more acidic than boron pentasil zeolite.
Pyridine adsorption IR measurements (Fig. 2) showed that boron
pentasil zeolite provides more Lewis acid sites than H-ZSM5 zeolite
(wave numbers 1485 cm−1 and 1628 cm−1). The amount of Broensted
acidic sites is comparable for both catalysts.
Raising the temperature up to 250 °C or higher results in more
than 50% conversion of geraniol (1). However thereby, the undesired
byproducts CO, CO2 and H2O were observed as the main products .
3. Mechanistic considerations
The weak acidic catalyst having a high amount of Lewis acidic sites
and low number of Broensted acid centers led to a more selective re-
action but a lower conversion was achieved. The strong acidic catalyst
H-ZSM5 behaved strictly opposite.
Therefore, subsequent optimization experiments have been carried
out using boron pentasil zeolite to achieve highest selectivity in forma-
tion of one of the monoterpenes 2, 3, 7, and 8.
A tentative mechanistic explanation towards formation of
phellandrenes from geraniol should involve acid-catalyzed cycliza-
tion of myrcene to form a-terpinyl carbocation which might under-
go a long range hydride shift from the allylic position of the
endocyclic double bond, followed by deprotonation. Such unusual
carbocation rearrangement could be imposed by the constrained
environment of the pentasil zeolite pores that marginally
Table 1
Initial experiments using boron pentasil zeolite and H-ZSM5 as catalyst.
Cat.
Reaction parameters
Conversiona
[%]
Selectivity [%]b
T
N2-flow
Geraniol
[g/h]
M Cat
[g]
Pressure
[mbar]
Linalool
Myrcene
Phellandrene
(mixture of α and β)
Ocimene (mixture
of cis and trans)
Unidentified
CO,
CO2
[°C]
[L/min]
B-Zeolite
H-ZSM5
220
220
10
10
30
30
3.00
3.00
80
80
22.4
74.2
15.2
0.3
36.7
27.7
22.8
1.6
11.4
0.9
1.4
11.3
12.5
58.2
a
Geraniol conversion (%)=(moles of reacted geraniol)/(moles of geraniol fed in).
Product selectivity (mol%)=(moles of carbon in a defined product)/(moles of carbon in reacted geraniol) the amount of dehydration water was neglected.
b

216
M. Eisenacher et al. / Catalysis Communications 26 (2012) 214–217
0,6
H-ZSM5
boron
pentasil
zeolite
Lewis
acidic
0,5
Lewis
acidic
Lewis +
Broensted
acidic
0,4
0,3
0,2
0,1
0,0
-0,1
Broensted
acidic
Broensted
acidic
-0,2
1500
1550
1600
1650
1700
Wave number [cm-1]
Fig. 2. FT–infrared spectroscopic study of surface acidity by pyridine adsorption.
Instrument). X-ray diffraction analyses were carried out on a Siemens
D 500. As computer software for the Design of Experiments Design
Expert 5.04 from State Ease Company was used. Surface analyses
were carried out on a Micrometrics ASAP 2000.
Scheme 2. Reactions network during gas phase reaction of geraniol.
accommodate compounds like phellandrene and its isomers. The
formation of phellandrene may also be caused by the zeolites
framework. It has been reported that exocyclic elimination from
carbocation is preferred over endocyclic elimination in presence of
zeolite Y [18].
The boron zeolite was prepared according to example 1 [19]:
1000 g of 50% strength hexamethylenediamine solution, 80 g of Aerosil
and 15 g of boric acid are mixed. The mixture is stirred until homoge-
neous and then heated for 5 days in a steel autoclave at 150 °C. The
resulting product is filtered off, washed and calcined at 550 °C for 6 h.
The H-ZSM5 was prepared by mixing 1000 g of 10% strength
tetrapropylammonium hydroxide solution, 80 g of Aerosil and 35 g
of tetraethyl orthosilicate. The mixture is stirred until homogeneous
and then heated for 5 days in a steel autoclave at 150 °C. The resulting
product is filtered off, washed and calcined at 550 °C for 6 h.
The dehydration of geraniol was conducted in a 100 cm long con-
tinuous plug flow fixed bed reactor in form of a coil, made out of a
stainless steel (Type 316 TI) tube with a 0.6 cm internal diameter.
The solid catalyst with 0.1–0.15 mm particle size was filled into the
reactor, which was placed afterwards in a temperature-controlled
oven. At the adjusted reaction temperature geraniol was pumped
into the reactor. The reaction products were collected in a double
jacket flask, which was kept at 20 °C. A sample for analysis was
taken after 2 h from the collected product mixture and analyzed via
GC.
4. Experimental
All chemicals used in this work were purchased from Sigma-
Aldrich (St. Louis, United States) and used as received without any
further purification. Substances in the reaction mixtures were identi-
fied by gas chromatography–mass spectroscopy (GC–MS) and by
spiking gaschromatography (GC) analyses with reference substances.
GC analyses of liquid samples were carried out on a Fisons GC 8000
gas chromatograph equipped with a flame ionization detector and a
Chromosorb W-AW column (i.d. y 2 mm, l=2.5 m), which was set
to on-column injection. GC analyses of gaseous samples were mea-
sured under the use of a thermal conductivity detector. Ammonia-
TPD measurements of the catalysts have been carried out on a
Netzsch STA 409 C. Pyridine adsorption IR measurements have been
carried out on a Nicolet Proteget 460. Bulk elemental analyses of
both catalysts were carried out by means of inductive couple plasma
atomic emission spectroscopy on a Spectroflame D (Spectro Analytic
5. Summary and outlook
Using a weak acidic boron pentasil zeolite with a high amount of
Lewis acid sites and a low number of Broensted acidic centers, the
gas phase treatment of geraniol (1) was studied. Surprisingly,
phellandrene (7) was found as product which was not expected due
to considerations from the literature [13–15]. Influences of nitrogen
flow, geraniol feed, temperature, amount of catalyst and pressure
were evaluated. The results demonstrated that the selectivity towards
linalool (8), myrcene (2), ocimene (3) or phellandrene (7) can be
1000
H-ZSM5
boron pentasil
zeolite
800
600
400
200
0
Table 2
Parameters for design of experiments.
Parameter
Range
N2-flow [L/h]
Geraniol feed [g/h]
0.9–15.0
8.5–30.0
2.0–5.0
200
400
600
800
M
Catalyst [g]
Temperature [°C]
Pressure [mbar]
Temperature [°C]
35–100
190–270
Fig. 1. Ammonia-TPD comparison of H-ZSM5 and boron pentasil zeolite.

M. Eisenacher et al. / Catalysis Communications 26 (2012) 214–217
217
Table 3
Selected results of the Design of Experiments.
Reaction parameters
Conversiona
[%]
Selectivity [%]b
T
N2-flow
Geraniol
[g/h]
M Cat.
[g]
Pressure
[mbar]
Linalool
Myrcene
Phellandrene
(mixture of α and β)
Ocimene (mixture
of cis and trans)
Others
CO,
CO2
[°C]
[L/min]
194
210
210
230
10
15
15
10
27.50
38.75
38.75
47.99
3.50
2.75
4.25
3.50
57.5
79.1
50.7
57.7
9.4
13.4
13.4
25.0
19.7
16.8
25.1
14.0
40.0
41.5
35.5
43.5
26.0
26.7
26.1
24.9
13.4
13.2
12.4
12.5
0.1
0.5
0.5
1.5
0.8
1.3
0.4
3.7
Bold data shows the highest selectivity to the terpenes.
a
Geraniol conversion (%)=(moles of reacted geraniol)/(moles of geraniol fed in).
Product selectivity (mol%)=(moles of carbon in a defined product)/(moles of carbon in reacted geraniol) the amount of dehydration water was neglected.
b
[6] A. Murata, S. Tsuchiya, A. Konno, C. Funabashi, J. Tanaka, K. Takabe, S. Hamamatsu,
US Pat., 4003958 (August 26, 1974) assigned to Nissan Chemical Industries Ltd.
[7] J.D. Surmatis, US Pat., 2848502 (November 27, 1956) assigned to Hoffmann – La
Roche Inc.
[8] J.A. Miller, H.C.S. Wood, Angewandte Chemie 76 (1964) 301.
[9] O.P. Vig, Indian Journal of Chemistry 7 (1969) 1111–1113.
[10] M.J. Climent, M.A. Miranda, H.D. Roth, European Journal of Organic Chemistry
8 (2000) 1563–1567.
controlled by variation of the reaction parameters and a total selectivity
of up to 99% to these four terpenes could be achieved.
Thus, a novel flexible process for the production of high value
added chemicals was reported starting from the natural feedstock
geraniol (1).
The current work focuses on catalyst development, explanation of the
reaction mechanism and detailed study of various process parameters,
reactor and process design.
[11] R.J. Grau, P.D. Zgolicz, C. Gutierrez, H.A. Taher, Journal of Molecular Catalysis A
148 (1999) 203–214.
[12] a) D. Buhl, D.M. Roberge, W.F. Hölderich, Applied Catalysis
A 188 (1999)
287–299;
b) D.M. Roberge, D. Buhl, J.P.M. Niederer, W.F. Hölderich, Applied Catalysis A 215
(2001) 111–124.
Acknowledgments
[13] T. Ohta, Y. Morimitsu, Y. Sameshima, T. Samuta, T. Ohba, Journal of Fermentation
and Bioengineering 72 (1991) 347–351.
[14] R.L. Baxter, W.A. Laurie, D. McHale, Tetrahedron 34 (1978) 2195–2199.
[15] a) M. Conceicao Cruz Costa, R.A.W. Johnstone, D. Whittaker, Journal of Molecular
Catalysis A 129 (1998) 79–89;
The authors are grateful to Dr. W. Bonrath and Dr. R. Karge, both
DSM Nutritional Products, Kaiseraugst, Switzerland, for the steady
interest in such kind of work.
b) C.A. Bunton, O. Cori, D. Hachey, J.P. Leresche, Journal of Organic Chemistry 44
(1979) 3238–3244;
c) K.L. Stevens, L. Jurd, G. Manners, Tetrahedron 28 (1972) 1939–1944.
[16] T. Barzetti, E. Selli, D. Moscotti, L. Forni, Journal of the Chemical Society, Faraday
Transactions 92 (1996) 1401–1407.
[17] G.E.P. Box, K.B. Wilson, Journal of the Royal Statistical Society 13 (1951) 1–45.
[18] C.J.A. Mota, J. Sommer, M. Hachourny, R. Jost, Journal of Catalysis 172 (1997)
194–202.
[19] L. Marosi, J. Stabenow, M. Schwarzmann, US Pat., 4 401 637 (Feb. 17, 1983)
assigned to BASF.
References
[1] a) (for further information see:)E. Breitmaier, Terpenes: Flavors, Fragrances,
Pharmaca, Pheromones, 1st ed. Wiley-VCH, Weinheim-New York, 2006;
b) L.M. Surhone, M.T. Timpledon, S.F. Marseken, Terpene, 1st ed. Betascript Publish-
ing, Saarbrücken, 2010;
c) S. Koul, A Comprehensive Treatise on Terpenes: Perspective, 1st ed. Lap Lambert
Academic Publishing, Saarbrücken, 2011.
[2] K.A.D. Swift, Topics in Catalysis 27 (2004) 143–155.
[3] B.B. Jones, B.C. Clark Jr., G.A. Jacobucci, Journal of Chromatography. A 178 (1979)
575–578.
[4] K. Steiner, US Pat., 3 953 524 (March 11, 1974) assigned to Hoffmann–La Roche
Inc.
[5] a) L.A. Goldblatt, S. Palkin, Journal of the American Chemical Society 63 (1941)
3517–3522;
b) A. Stolle, W. Bonrath, B. Ondruschka, Journal of Analytical and Applied Pyrol-
ysis 83 (2008) 26–36.