Zinc catalyst recycling in the preparation of (all-rac)-α-tocopherol from trimethylhydroquinone and isophytol

Article abstract of DOI:10.2478/s11696-011-0027-0

Recycling of ZnCl₂ as a catalyst in the cyclocondensation of trimethylhydroquinone and isophytol to all-rac-a-tocopherol was studied. ZnCl2 was recycled at over 98 % efficiency in required purity by repeated extraction of the reaction streams with water, followed by switching of solvent to butyl acetate and then back to water. Recycled ZnCl2 had no negative effect on the yield of all-rac-a- tocopherol (maintaining over 90 %) at nearly total conversion of isophytol and only a 3 % molar excess of trimethylhydroquinone.

Full text of DOI:10.2478/s11696-011-0027-0

Chemical Papers 65 (4) 427–436 (2011)
DOI: 10.2478/s11696-011-0027-0
ORIGINAL PAPER
Zinc catalyst recycling in the preparation of (all-
)-α-tocopherol
from trimethylhydroquinone and isophytol
b
c
aJuraj Go¨m¨ory, Milan Králik, Alexander Kaszonyi
aZentiva, a.s., Nitrianska 100, 920 27 Hlohovec, Slovakia
bVUCHT j.s.c., Research Institute of Chemical Technology, Nobelova 34, 836 03 Bratislava, Slovakia
cDepartment of Organic Technology, Slovak University of Technology, Radlinského 9, 812 37 Bratislava, Slovakia
Received 21 June 2010; Revised 24 January 2011; Accepted 24 January 2011
Recycling of ZnCl2 as a catalyst in the cyclocondensation of trimethylhydroquinone and isophytol
to all-rac-α-tocopherol was studied. ZnCl2 was recycled at over 98 % efficiency in required purity
by repeated extraction of the reaction streams with water, followed by switching of solvent to butyl
acetate and then back to water. Recycled ZnCl2 had no negative effect on the yield of all-rac-α-
tocopherol (maintaining over 90 %) at nearly total conversion of isophytol and only a 3 % molar
excess of trimethylhydroquinone.
c
ꢀ 2011 Institute of Chemistry, Slovak Academy of Sciences
Keywords: all-rac-α-tocopherol, trimethylhydroquinone, isophytol, ZnCl2, catalyst recycling,
mechanism
Introduction
by the combination of boric and carboxylic acids (Bon-
rath & Netscher, 2005; Bonrath et al. 2007a, 2007b;
Baldenius et al., 1996).
all-rac-α-Tocopherol (synthetic Vitamin E) repre-
sents the most important industrial fat-soluble antiox-
idant. The bulk of its production (74 %) is used in the
manufacturing of animal feed (Bonrath & Netscher,
2005; Bonrath et al., 2007a, 2007b; Baldenius et al.,
1996). The production of animal feed represents thus
the most valuable market for all-rac-α-tocopherol, one
amounting annually to more than 30000 tons (Bonrath
et al., 2007a). Other applications include production
of cosmetics (3 %), pharmaceuticals, and food. The
main part (98 %) of produced tocopherol is used in
the acetate form (Bonrath et al., 2007b; Baldenius et
al., 1996).
The last step in the industrial synthesis of all-
rac-α-tocopherol involves the reaction of trimethylhy-
droquinone (TMHQ) and isophytol (IPh), phytol, or
phytyl halide (Bonrath et al., 2007a; Baldenius et al.,
1996) (Fig. 1). Traditionally, this reaction is catalyzed
by a combination of a Lewis acid, e.g. zinc(II) dichlo-
ride, and a Br¨onsted acid, such as hydrochloric acid, or
More recent papers (Bonrath et al., 2007b) men-
tion the use of triflates (tris-trifluoromethanesulfonates)
of rare earth metals (MIII(OTf)3), e.g. Gd(OTf)3, as
suitable catalysts for the reaction of TMHQ and IPh.
The reaction can be carried out in several solvent
systems, e.g. two-phase systems based on ethylene or
propylene carbonate, and on hexane, heptane or oc-
tane, respectively. The reaction temperature is kept
in the range of 80–150◦C. Compared with ZnCl2 as
the catalyst, the same conversion of raw materials and
selectivity of tocopherol formation can be achieved
with lower catalyst amounts. Similar activity is ex-
hibited by microencapsulated catalysts (MC) (Schager
& Bonrath, 2000), e.g MC–imides (CF3SO2)2NH and
triflates MC–Sc(OTf)3, with maximal yields of toco-
pherol of 87.5 % and 86.8 %, respectively. However,
when using these catalysts, a problem of active phase
leaching occurs; and to make it worse economically,
the MC-catalysts’ activity cannot be restored.
*Corresponding author, e-mail: alexander.kaszonyi@stuba.sk

428
J. G¨om¨ory et al./Chemical Papers 65 (4) 427–436 (2011)
Nanocomposite catalysts Nafion/SiO2 (Nafion
were used as alternative media, including supercritical
liquids (Wang et al., 2000; Bonrath & Netscher, 2005;
Bonrath et al., 2009). The use of polar and aprotic
phases, particularly a biphasic solvent system contain-
ing, e.g., ethylene or propylene carbonate and heptane
for polar and non-polar components, respectively, gave
good results even with “traditional” catalysts such as
ZnCl2/HCl. From an industrial viewpoint, these meth-
ods suffer from two major disadvantages: i) potential
contamination of waste water with zinc ions, ii) cor-
rosion problems increasing with the increasing con-
centration of mineral acid, an environmental problem
that has to be solved, e.g. by effective recycling of cat-
alysts and solvents.
Laboratory procedure described in this paper cap-
italizes wide industrial experience provided by a fairly
simple recycling procedure of zinc dichloride tested in
our laboratories. Due to the fact that the production
was shut down recently, it is now possible to publish
these results with relatively broad details.
mass content between 5 % and 20 %) or Nafion NR50
in toluene as a solvent under nitrogen atmosphere pro-
duce tocopherol from TMHQ and IPh in yields of
over 90 % (Wang & Xu, 2004). However, attempts
to reuse the above nanocomposite catalysts have not
succeeded due to the significant decrease in activity.
A more general use of composite Nafion/silica cata-
lyst for the synthesis of pure chemicals, including toco-
pherol, has been described in the recent paper of Hinze
et al. (2009). The authors compared the dependence
of tocopherol yields obtained using several composite
catalysts differing in the Nafion content, whereby the
highest tocopherol yield (88 %) was achieved using
40 % of Nafion. However, reused catalysts displayed
significantly lower activity. Nafion, an ion-exchanging
resin, similarly to the earlier mentioned catalysts, is
prepared from a perfluorinated organic compound.
Kemnitz et al. (2009), in their European patent, de-
scribe the reaction of IPh and TMHQ catalyzed by a
heterogeneous fluoride catalyst of the general formula
MgF2−x(OH)x or AlF3−x(OH)x in a two-phase sol-
vent system (propylene carbonate and heptane). The
reaction was carried out at a solvent reflux in a closed
system. A quantitative yield of tocopherol, relative to
IPh, was achieved at the molar ratio of TMHQ : IPh
exceeding 2 : 1, and the catalyst mass content of about
7 %. However, the examples in this patent do not men-
tion any reuse of the catalysts. The same authors (Co-
man et al., 2008) also described the preparation of a
nanocomposite catalyst based on aluminum fluoride
made from aluminum isopropylate by “sol-gel” fluori-
nation with hydrofluoric acid. The reaction conditions,
reactants and results are similar to those claimed in
the above mentioned patent. Once again, the reuse
of catalyst was not mentioned. Xing et al. (2009) de-
scribed a continuous process for the synthesis of all-
rac-α-tocopherol catalyzed by an ionic liquid based on
the tetrafluoroborate anion. The propylene carbonate
solution of TMHQ and the solution of IPh in supercrit-
ical CO2 were continuously fed into a reactor. During
the reaction, the product was continuously separated
from the reaction mixture by extraction with super-
critical CO2. Decompression of a supercritical mixture
affords a catalyst free product. The adjusted tempera-
ture/pressure regimen affects the continuous reaction
as well as the product separation. all-rac-α-Tocopherol
was prepared with a 90.4 % yield at 100◦C, 20 MPa,
and the residence time of 12.6 min using the molar
ratio of IPh : TMHQ : propylene carbonate : CO2 of
1 : 1.5 : 35 : 3358.
Experimental
Isophytol, IPh (3,7,11,15-tetramethylhexadec-1-
en-3-ol) (Hoffman La Roche, Swiss), 95 %; 2,3,5-
trimethylhydroquinone, TMHQ (Hoffman La Roche,
Swiss), 98.5 %; water (Zentiva, Slovakia), conduc-
tivity 2 µS; ZnCl2 (Mikrochem, Slovakia), 99 %;
butyl acetate (BuAc) (Zentiva, Slovakia), 99 %; NaOH
(Mikrochem, Slovakia), 98 %; HCl (Mikrochem, Slo-
vakia), 35 % aqueous solution, were used in this study.
all-rac-α-Tocopherol (TPh) was determined by gas
chromatography on a Hewlett-Packard (USA) appara-
tus equipped with an autosampler and an HP-5 (Ul-
tra 2) column, 50 m × 0.32 mm, using phenyl- and
methylpolysiloxane, 5 % and 95 %, respectively. An-
alytical conditions were set as follows: column tem-
perature to 245◦C, temperature delay to 2 min, tem-
perature gradient to 10◦C min−1 up to 300◦C, injec-
tor temperature to 260◦C, FID temperature to 300◦C,
carrier gas: He, flow rate of 1.4 mL min−1, injection:
0.2 µL, sample concentration: 2 mg mL−1, solvent:
hexane. TPh was identified by its retention time us-
ing an external standard all-rac-α-tocopherol, Fluka,
Swiss), analysis duration: 30 min.
Concentration of ZnCl2 in aqueous solution was
determined by argentometry and complexometry.
Preparation of all-rac-α-tocopherol
The apparatus consisted of a 1.5 L glass sulfona-
tion flask (Kavalier Sázava, Czech Republic) with a
nitrogen inlet tube, glass stirring rod fitted with a
teflon horseshoe stirrer (frequency > 600 min−1), re-
flux condenser with a water collector, thermometer
and a dropping funnel. The reactor was heated in an
oil bath (Heidolph, USA). During the work up of the
reaction mixture, phases were separated in a 2 L or 1 L
From the so far published data, it appears that
catalysts containing fluorine in organic or inorganic
form have been intensely studied although their use is
hampered by their high price and problematic reuse.
For the selection of recyclable catalysts, economically
and environmentally acceptable, as well as the one of a
suitable reaction medium, multiphase solvent systems

J. G¨om¨ory et al./Chemical Papers 65 (4) 427–436 (2011)
429
separatory funnel. The stirrer was powered by a vari-
HO
able speed motor (Heidolph, model RZ 20-51, USA).
The reaction mixture was concentrated in a variable
speed laboratory rotary evaporator (Heidolph, model
VV 2000, USA).
+
OH
OH
IPh
− H2O
TMHQ
The 1st cycle: The described glass sulfonation flask
was charged with BuAc, TMHQ, and pure ZnCl2 in
amounts given in Tables 1–4. The stirred reaction mix-
ture was heated up to 110◦C in order to achieve com-
plete dissolution of TMHQ and ZnCl2. Subsequently,
IPh was added at the rate controlled by the capacity
of the reflux condenser. Water from the reaction was
trapped in the water collector of the azeotropic exten-
der. After the whole amount of IPh had been added,
the reaction mixture was left to reflux for another hour
with a concomitant rise of temperature to 129◦C. Af-
ter cooling the reaction mixture to about 30◦C, ZnCl2
was removed by a triple extraction with 200 mL of
water. The extracts were designated as A(1-1), A(1-
2) and A(1-3). The last extract contained less than
0.15 % of the used ZnCl2. The remaining organic so-
lution was then extracted four times (4 × 160 mL)
with aqueous NaOH (8 %), in order to remove the un-
reacted hydroquinone together with the last remnants
of ZnCl2. pH of the raffinate was then changed by its
triple extraction with 200 mL of 0.1 % HCl solution.
These extracts no longer contained zinc. The raffinate
was concentrated under reduced pressure to remove
BuAc. The raw product, consisting largely of crude
tocopherol, was acylated by acetic anhydride in the
next step. ZnCl2 was regenerated from extract A(1-
1) in several steps. In the first step the main part of
water (160 mL) was distilled off. Next, the distillation
residue was diluted with 100 mL of BuAc and the re-
maining water was removed by azeotropic distillation
yielding a BuAc solution of catalyst B1. The BuAc so-
lution of the catalyst was filtered and then treated by
triple extraction with 50 mL of water resulting once
again in an aqueous solution of the catalyst, the ex-
tracts being designated as C(1-1), C(1-2), and C(1-
3), respectively. After the determination of the ZnCl2
concentration, an aliquot of the first extract, C(1-1),
was prepared with a calculated amount of ZnCl2 and
reused in the next cycle (Fig. 2). The amounts of other
reaction components were proportional to the amount
of the reused catalyst.
catalyst
HO
OH
catalyst
HO
O
TPh
Fig. 1. Formation of all-rac-α-tocopherol (TPh) from trimethyl-
hydroquinone (TMHQ) and isophytol (IPh).
TPh was worked up in the same way as in cycle 1.
The A(2-1) extract was concentrated and worked up
as the A(1-1) extract from the preceding cycle giv-
ing once again a BuAc solution of ZnCl2. Next, three
extractions followed, the first one with extract C(1-
2) from the preceding cycle giving extract C(2-1), the
second extraction with extract C(1-3) giving extract
C(2-2) and the third one with pure water giving ex-
tract C(2-3) (Fig. 2).
The 3rd and possible further cycles: BuAc and cat-
alyst extract from the preceding cycle, C(2-1), were
like in the preceding cycle charged into the appara-
tus. Once again, TMHQ and IPh were added. As be-
fore, tocopherol and three extracts, C(3-1), C(3-2),
and C(3-3), were obtained, out of which the most
concentrated one, C(3-1), was used in the next cycle
(Fig. 1).
Results and discussion
The reaction is known to proceed in two formal
steps. The first step involves the alkylation of TMHQ
with IPh followed by the second step, formation of
cyclic ether. Thereby, some catalysts allow high yield
of the intermediate product (Coman et al., 2008).
Side products of the reaction originate in the re-
actions of both starting materials and intermediates.
Since the hydroquinone structure is easily oxidized to
a quinone, the reactions are, as a rule, carried out in
inert atmosphere, e.g. under nitrogen, or in the pres-
ence of a reducing agent.
The 2nd cycle: BuAc and the calculated amount
of catalysts extract C(1-1) were charged into the ap-
paratus. The mixture was heated to reflux and water
was removed by azeotropic distillation. TMHQ was
then added and after its dissolution, IPh was added
in the same way as in cycle 1. Unlike in the first cy-
cle, the first extraction of catalyst from the reaction
mixture was carried out with the second extract of
the preceding cycle, A(1-2), giving extract A(2-1). The
second one was done using the A(1-3) extract result-
ing in A(2-2). Finally, the third extraction was done
with 200 mL of pure water. Raw product containing
In the presence of acid catalysts, IPh dehydra-
tion results in the formation of a mixture of var-
ious isomeric dienes, which in turn can be dimer-

430
J. G¨om¨ory et al./Chemical Papers 65 (4) 427–436 (2011)
Cycle 1
reaction mixture
extract A(1-1)
1) 200 mL water
2) 200 mL water
extract A(1-2) for the 1st extraction in the 2nd cycle
extractions
3) 200 mL water
extract A(1-3) for the 2nd extraction in the 2nd cycle
raw product in butyl acetate
distilled water
extract A(1-1)
butyl acetate
solvent exchange
extraction of B1
- B1
solution of ZnCl2 in butyl acetate
1) 50 mL water
2) 50 mL water
extract C(1-1) containing ZnCl2 charged into 2nd cycle as catalyst
extract C(1-2) for the 1st extraction of B2 in the 2nd cycle
extract C(1-3) for the 2nd extraction of B2 in the 2nd cycle
3) 50 mL water
butyl acetate
Cycle 2
reaction mixture in the 2nd
extract A(2-1)
1) A(1-2)
2) A(1-3)
extract A(2-2) for the 1st extraction in the 3rd cycle
extractions
3) 200 mL water
extract A(2-3) for the 2nd extraction in the 3rd cycle
raw product in butyl acetate
distilled water
extract A(2-1)
butyl acetate
solvent exchange
extraction of B2
solution of ZnCl2 in butyl acetate
- B2
1) C(1-1)
2) C(1-2)
extract C(2-1) containing ZnCl2 charged into 3rd cycle as catalyst
extract C(2-2) for the 1st extraction of B3 in the 3rd cycle
extract C(3-3) for the 2nd extraction of B3 in the 3rd cycle
3) 50 mL water
butyl acetate
Cycle 3
reaction mixture in the 3rd cycle
extract A(3-1)
1) A(2-2)
2) A(2-3)
extract A(3-2) for the 1st extraction in the 4thcycle
extractions
3) 200 mL water
extract A(3-3) for the 2nd extraction in the 4th cycle
raw product in butyl acetate
distilled water
extract A(3-1)
butyl acetate
solvent exchange
extraction of B3
- B3
solution of ZnCl2 in butyl acetate
1) C(2-1)
2) C(2-2)
extract C(3-1) containing ZnCl2 charged into 4th cycle as catalyst
extract C(3-2) for the 1st extraction of B4 in the 4th cycle
extract C(3-3) for the 2nd extraction of B4 in the 4th cycle
3) 50 mL water
butyl acetate
Fig. 2. Block scheme of ZnCl2 extractions during its recycling from the reaction products of all-rac-α-tocopherol preparation.

J. G¨om¨ory et al./Chemical Papers 65 (4) 427–436 (2011)
431
CH3
CH2
H
OH
OH
OH
− H2O
H
HO
catalyst
+ TMHQ
− H2O
− H
− H
HO
Fig. 4. Dehydration of IPh and formation of dienes.
O
I
Fig. 3. Formation of dihydrobenzofuranol (I) derivatives.
HO
+
OH
ized (one of them is mentioned in the British
pharmacopoeia (The British Pharmacopoeia Com-
mission, 2010) as (all-RS,all-E)-2,6,10,14,19,23,27,31-
octamethyldotriaconta-12,14,18-triene as an impurity
in TPh). In order to suppress these processes, alkyla-
tion can be carried out in the presence of water. Un-
fortunately, water also decelerates the second step of
the reaction – dehydration. The favorable effect of wa-
ter was observed only by Coman et al. (2008) in the
process catalyzed by fluoride catalysts, where water
is responsible for changing the acidity of active sites
of the catalysts rather than for affecting the reaction
equilibrium.
II
H
HO
OH
III
HO
OH
IV
Acidic catalyst (a proton) can activate IPh at its
double bond. In such a case, the equilibrium is shifted
towards the more stable secondary cation. Thus, in
the presence of protons, IPh, due to kinetic factors,
reacts preferentially at its secondary carbon (cations),
giving rise to secondary products such as dihydroben-
zofuranols (I ) (Fig. 3). However, the target product,
TPh (see Fig. 1) arises from the rather less favored
primary cation.
H
–
HO
O
V
Fig. 5. Formationof TPh from TMHQ and dehydrated IPh.
In the British pharmacopoeia (The British Phar-
macopoeia Commission, 2010), this compounds are
mentioned as all-rac-cis- and all-rac-trans-2,3,4,6,7-
pentamethyl-2-(4,8,12-trimethyltridecyl)-2,3-dihydro-
benazofuran-5-ol.
Alternatively, TMHQ reacting with the activated
and dehydrated IPh (II) can form an etheric C—O—
C bond in the first reaction step. Cyclization then
occurs by formation of a C—C bond in the second
reaction step, only. However, the appropriate ether
to be used as the intermediate compound of an al-
ternative mechanism in the reaction mixture could
not be found in our reaction mixture, or in papers
dealing with the reaction of TMHQ and IPh. As de-
protonized forms of species III and IV (which is the
same compound) of the suggested mechanism were
found in the reaction mixture (Coman et al., 2008;
The British Pharmacopoeia Commision, 2010), the
suggested reaction steps in Fig. 5 represent the more
probable mechanism of this reaction in the presence
The primary cation can be generated also by de-
hydration of IPh (Fig. 4).
A cation with positive charge located at the termi-
nal carbon (primary cation) is able to alkylate TMHQ
more easily than a tertiary cation, which is sterically
hindered by alkyl groups. In this case, kinetics favours
alkylation of TMHQ at its sole free aromatic posi-
tion by the primary carbon of the dehydrated and
activated molecule of IPh. Further steps of the tenta-
tive mechanism of TPh formation can be the following
(Fig. 5).

432
J. G¨om¨ory et al./Chemical Papers 65 (4) 427–436 (2011)
CH3
of ZnCl2 as the catalyst and at conditions applied in
this work.
Under appropriate reaction conditions, TPh re-
mains stable since it is formed in an oxygen-free atmo-
sphere. Moreover, its ether bond is fairly stable against
hydrolysis.
H
C
δ+
CH3
HO
H2C
H
O
H3C
H
O
Zn
δ−
CH3 Cl
Cl
The procedure of tocopherol preparation by the
reaction of IPh with TMHQ described in this work is
based on industrial practice utilizing reactors of a few
cubic meters and it takes into account all conditions
required to reach maximal selectivity of the reaction
running according to the suggested mechanism. It is
worthwhile to add that mixing of the reaction mixture
was rather intensive; therefore, mass-transport phe-
nomena did not exhibit a significant hindrance to the
course of the process. Generation of quinonoid com-
pounds was suppressed by reaction conditions without
the access of oxygen, or alternatively by an addition
of a reducing agent, e.g. sodium hyposulfite. On the
other hand, some side reactions of IPh were suppressed
using ZnCl2 without the addition of HCl as a source
of acidic protons. Moreover, IPh was added dropwise
so that its concentration relative to that of TMHQ re-
mained very low. IPh was removed by its fast reaction
with TMHQ, which in turn produced water limiting
its dehydration consequently.
In the context of the coordination chemistry, ZnCl2
is known to supplement its coordination sphere by at
least two additional ligands in solutions (Cotton et al.,
1995). The Zn—Cl bond is more of covalent than of
ionic nature. In our reaction mixture, the role of lig-
ands was fulfilled by either of the compounds present,
namely water, IPh, TMHQ, or BuAc. All these lig-
ands can be bound to zinc by a lone electron pair
of their oxygen atoms. This type of ligand—metal
bonds belongs to the preferred bonds in the coordi-
nation chemistry of Zn (Cotton et al., 1995). At the
reaction outset, in aqueous solution of ZnCl2, the co-
ordination sphere of zinc is filled by water molecules
and OH groups. After the water removal by azeotropic
distillation, TMHQ, as well as BuAc (solvent), enters
the coordination sphere of zinc. Although these lig-
ands account for the solubility of ZnCl2 in the reac-
tion mixture, their bonding is not as tight as that of
water. Consequently, the formed complex is strongly
hygroscopic and can enter further reactions with the
OH groups of the compounds present in the reaction
mixture.
Fig. 6. Proposed active complex of ZnCl2, TMHQ, and IPh.
tivated cation II inside the coordination sphere of
Zn in close vicinity of TMHQ (a π–allylic complex
is formed). After this dissociation, the molecule of
TMHQ, now formally with negatively charged phe-
nolate group, in the coordination sphere of Zn is acti-
vated for nucleophilic alkylation at its single free po-
sition. During the alkylation with cation II , a proton
can migrate from the six membered ring of TMHQ to
the alkyl group of IPh, generating a tertiary cation on
the 3rd carbon of the original IPh (cation IV ). In the
subsequent cyclization reaction with the negatively
charged phenolate group of alkylated TMHQ, cation
IV can create an ether bond. Then, the molecule of
TPh leaves the coordination sphere of ZnCl2. In other
words, it is assumed that all necessary reaction steps
take place in the coordination sphere of the Zn com-
plex. By the migration of the first proton in this oth-
erwise neutral complex, both IPh and TMHQ are ac-
tivated for alkylation of TMHQ to its free position
by the 1st carbon of IPh and consequently for cy-
clization between O of TMHQ and the 3rd carbon of
IPh. Water is removed from the coordination sphere
of zinc by azeotropic distillation, making ZnCl2 ready
for the next catalytic cycle, i.e. coordination with an-
other pair of TMHQ and IPh. Azeotropic distillation
continually removes water (and BuAc) and inciden-
tally also helps to maintain the reaction temperature.
The amount of water collected in the azeotropic trap
indicates the degree of conversion of IPh. As soon as
the rate of water elimination slows down with a con-
comitant rise of temperature, the condensation reac-
tion is complete. Owing to the fact that water and
OH− groups are more tightly bound to ZnCl2 than
IPh, TPh, TMHQ, or its phenolate, zinc can be al-
most completely removed from the reaction mixture
by the extraction with water and a dilute NaOH solu-
tion. However, dilute NaOH is mainly used to remove
unreacted TMHQ by converting it into a water soluble
phenolate.
Just as water in the zinc coordination sphere,
TMHQ can also release a proton, contributing thus
to the acidity of the reaction mixture. When IPh is
added, its tertiary OH group can displace the more
weakly bound BuAc from the coordination sphere of
zinc subsequently creating a zinc complex involving
both IPh and TMHQ (Fig. 6).
It is assumed that in this complex, the acidic pro-
ton of TMHQ migrates to the OH group of IPh and
catalyses its dehydration and the formation of ac-
Owing to the presence of butyl acetate and acidic
protons, TPh prepared by the reaction of IPh and
TMHQ contains also up to 2 % of tocopherol acetate
formed by acid-catalyzed transesterification of BuAc
by the free OH group of TPh. This reaction produces
butanol as a side product, which can be converted
back to BuAc in the next step of vitamin E prepara-
tion; where TPh is acetylated with acetanhydride to
all-rac-α-tocopheryl acetate (commercial form of vita-
min E acetate).

J. G¨om¨ory et al./Chemical Papers 65 (4) 427–436 (2011)
433
After the reaction mixture has been worked up,
the catalyst can be recycled. After exchanging water
from the first aqueous extract of the reaction mixture,
ZnCl2was extracted again from the less polar BuAc
layer to the aqueous layer creating thus another con-
centrated aqueous solution of the catalyst with suffi-
cient purity to be reused in the synthesis of TPh. As
can be seen from the details of the procedure (Fig. 2),
the second aqueous extract of the reaction mixture as
well as the third extract of the BuAc solution of ZnCl2
were used in the next cycle: the second extract for the
first extraction, and the third one for the second ex-
traction of the corresponding mixture. In both cases,
the third extraction was done with distilled water.
Such procedure resulted in more concentrated ZnCl2
solutions, and the energy demand of this work up is
lower than that of dilute ZnCl2 solutions.
Acylated derivatives of dihydrobenzofuranol (< 1 %)
and quinones (oxidation products of the correspond-
ing hydroquinones and TPh) dissolved in all-rac-
α-tocopheryl acetate represent the most significant
impurities. Phytadiene together with other volatile
components are removed from the final all-rac-α-
tocopheryl acetate by a short path distillation which
in the end also separates all-rac-α-tocopheryl acetate
from higher boiling components. Thus, a high quality
product is obtained.
Phytadienes and their dimers, e.g. (all-RS,all-E)-
2,6,10,14,19,23,27,31-octamethyl-dotriaconta-12,14,
18-triene, are formed also during the distillation of the
reaction mixture as a consequence of heat delivered to
the liquid mixture. Since under such conditions, ZnCl2
does not escape, its remnants left over by imperfect
extraction of the reaction products remain in the dis-
tillation residue which is less than 6 % of the raw TPh.
However, after standard sextuple extraction of the re-
action mixture (first three times with aqueous solution
and then three times with NaOH solution), the zinc
content of heavy residues of only a few ppm was ob-
tained, an amount that can be safely neglected in the
total zinc balance. Thus, ZnCl2 as catalyst has no dele-
terious effect on the final product purification process.
The remaining zinc can be completely separated from
heavy organic residues only after its mineralization.
However, this procedure is a little expensive. Taking
into account that zinc, at the ppm level, is a biogenic
element, this procedure has no economic significance.
The amount of ZnCl2 in the respective extracts of
the reaction mixture and BuAc solutions are given in
Tables 1–4, along with the amounts of starting ma-
terials and the TPh yield achieved with catalyst re-
cycled under various reaction conditions. All experi-
ments described here were conducted under conditions
normally imposed by the regulation of industrial pro-
duction. To compare the effectiveness of the recycled
catalyst with that of the initial catalyst, the molar ra-
tio of reagents and catalysts were kept constant, and
they represent a 3 % molar excess of TMHQ over IPh.
Table 1 displays the results of the first series of ex-
periments carried out under standard conditions ex-
cept for using catalyst recycled from the first experi-
ment in the second and third ones. Table 2 shows the
results of experiments carried out under nitrogen at-
mosphere, under the same conditions as those above,
except for the absence of oxygen. In the third series
(Table 3), only 50 % of the recycled catalyst was used
and fresh catalyst was added. In the fourth series of
experiments, sodium hyposulfite was added to the re-
action mixture as the reducing agent in three parallel
experiments with fresh catalyst used also for checking
the reproducibility of the measured data.
The very first concentrated BuAc solution of ZnCl2
(B1, B2, or B3) is rather dark in color owing to the
presence of quinonoid polymeric particles partly solu-
ble both in BuAc and in water.
On the other hand, the first concentrated aqueous
extract from the concentrated BuAc solution of ZnCl2
is nearly colorless and thus suitable to be used in fur-
ther batch of the tocopherol production. Hence, to this
solution, BuAc was added and after the azeotropic wa-
ter removal, TMHQ was also added. Such a mixture
was charged into the subsequent reaction cycle.
Apart from high selectivity and low price, ZnCl2 as
catalyst has the advantage in allowing the use of stan-
dard technological equipments, i.e. an enameled reac-
tor equipped with a glass condenser and an azeotropic
water collector, interconnected by glass or enameled
tubing with other equipments. The system does not
contain any emulgating component nor is any such
compound added; hence, layer separation after ex-
traction and distillation is fast and easy. During the
reaction, the system remains perfectly homogeneous.
Thus, stirring is not a problem compared to systems
using heterogeneous catalysts (G¨omo¨ry et al., 2010).
Recycling of the ZnCl2 catalyst, while preserving its
selectivity and reaction yield, represents a viable al-
ternative to working up ZnCl2 for other purposes (as
a source of biogenic zinc in fertilizers, or zinc sul-
fate for metal smelters, zinc white (ZnO), etc.) as it
is routinely done in industry. Polar inorganic impu-
rities are separated from ZnCl2 during dissolution in
BuAc, whereas non-polar organic impurities are re-
moved during its extraction to water. On the other
hand, ZnCl2 does not escape during azeotropic distil-
lation; it remains stable and can be practically quan-
titatively recycled from its aqueous solutions by the
above method.
After the removal of unreacted TMHQ and ZnCl2,
raw product (BuAc solution of raw TPh) is acety-
lated to tocopheryl acetate, as already mentioned.
During the process, by-products are acetylated as
well and must be removed from the final product.
Data in Tables 1–4 should be supplemented by
the fact that blank experiment (without any catalyst)
failed to produce tocopherol. An insignificant amount
of tocopherol was also formed in experiments with the

434
J. G¨om¨ory et al./Chemical Papers 65 (4) 427–436 (2011)
Table 1. Cyclocondensation of TMHQ with IPh under reflux of BuAc; ZnCl2 as catalyst recycled in the form of an aqueous
solution; molar ratio of reactants and catalysts TMHQ : IPh : ZnCl2 = 1.03 : 1 : 0.2
Charged amounts
Reaction products
ZnCl2, extracts Ab ZnCl2, extracts Cc
Cycle, n
TMHQ
IPh
mol
ZnCl2
BuAc
mL
TPha
%
Yield of TPh
%
1-nd
2-ne
3-nf
1-nd
2-ne
3-nf
mmol
1
2
3
1.24
1.11
1.07
1.2
1.07
1.03
0.25
0.23
0.22
400
360
340
91
90.7
93
247
222
216
3.3
7.3
5.6
0.3
1.4
0.5
233
220
213
13.2
11
5.8
0.7
0.3
0.4
91.6
92.5
92.2
a) Concentration of TPh in raw TPh after all extractions and evaporation of BuAc; b) amount of ZnCl2 in the aqueous extracts of
the raw product; c) amount of ZnCl2 in the aqueous extracts of the BuAc solution of ZnCl2 in the given cycle; d), e), and f ) first,
second, and third extracts in the given cycle of the raw product or the BuAc solution of ZnCl2, respectively.
Table 2. Cyclocondensation of TMHQ with IPh under nitrogen and reflux of BuAc; ZnCl2 as catalyst recycled in the form of an
aqueous solution; molar ratio of reactants and catalysts TMHQ : IPh : ZnCl2 = 1.03 : 1 : 0.2
Charged amounts
Reaction products
ZnCl2, extracts Ab ZnCl2, extracts Cc
Cycle, n
TMHQ
IPh
mol
ZnCl2
BuAc
mL
TPha
%
Yield of TPh
%
1-nd
2-ne
3-nf
1-nd
2-ne
3-nf
mmol
1
2
3
1.24
1.055
1.0
1.2
1.02
0.97
0.25
0.21
0.2
400
340
325
91.5
90.0
93.4
246
212
203
3
5.9
4.8
0.3
1
0.4
229
209
200
9
6.6
5.1
0.4
0.4
0.3
93.3
90.2
91.8
a–f ) Same as for Table 1.
ZnCl2 catalyst, but without solvent whose role was to
achieve a homogenous reaction mixture and a better
contact of reactants with the catalyst. Comparative
experiments have shown that the addition of HCl to
the reaction mixture does not increase the yield of
TPh although it significantly increases the corrosivity
of the reaction mixture and all extracts. Moreover, it
contaminates water vapor released during the reaction
and during the concentration of the catalyst solution.
These experiments were done to compare the results of
using the combined catalyst ZnCl2/HCl known from
literature with those of using ZnCl2 only.
The obtained results indicate that in each exper-
iment, the content of tocopherol in raw product, ob-
tained after the extraction of the final reaction mix-
ture by aqueous solutions and evaporation of BuAc,
exceeded 90 %. Also, the yield of TPh in each experi-
ment exceeded 90 %. All these data are very important
for the overall economy of this process.
Volatile impurities (5–6 %, mostly solvent residues,
phytadienes, and dihydrobenzofuranols) are more
volatile than TPh, as mentioned previously. The least
volatile ones (up to 1 %, mainly phytadiene dimers)
cannot be removed by distillation, which underlines
the importance of using an appropriate catalytic sys-
tem. Other components are oligomeric forms of phy-
tadienes and quinonoid structures. Their amounts are
between 0.5–2 %, and they are generated even during
the distillation following the acetylation.
While in the first experiment, 91 % of ZnCl2 was
recycled (the rest remained in the aqueous extracts),
in the third and further cycles, more than 98 % of cat-
alyst was recycled. The combined amounts of ZnCl2
in the aqueous extracts already exceed ZnCl2 charged
into the second and further cycles as more diluted ex-
tracts from the preceding experiment were used to ex-
tract the final reaction mixture in the next cycle. Less
than 2 % of the charged amount of ZnCl2 was ex-
tracted with NaOH solutions since a part of ZnCl2
was lost in the samples taken for the measurements
of its concentration in different extracts and in the
reaction mixture.
From Tables 1 and 2 follows that recycled ZnCl2
allows achieving higher purity of raw TPh, which is
enhanced from 91 % to 93 % after the second recy-
cling of the catalyst, nearly at the same level of TPh
yields. However, these yield variations are within the
reproducibility of the experiments. Surprisingly, nei-
ther quality nor yield was improved by the experi-
ments in which a part of ZnCl2 was replaced by fresh
catalyst (Table 3). Essentially the same result was ob-
tained in the presence of the reducing agent. The ef-
fect of the use of sodium hyposulfite or nitrogen at-
mosphere is shown by lighter color of crude tocopherol

J. G¨om¨ory et al./Chemical Papers 65 (4) 427–436 (2011)
435
Table 3. Cyclocondensation of TMHQ with IPh under reflux of BuAc; catalyst combined from fresh ZnCl2 and the recycled one
in the form of an aqueous solution; molar ratio of reactants and catalysts TMHQ : IPh : ZnCl2 = 1.03 : 1 : 0.2
Charged amounts
Reaction products
ZnCl2, extracts Ab ZnCl2, extracts Cc
Cycle, n
TMHQ
IPh
ZnCl2
fresh/recycled
BuAc
mL
TPha
%
Yield of TPh
%
1-nd
2-ne
3-nf
1-nd
2-ne
3-nf
mol
mmol
1
2
3
1.24
1.24
1.24
1.2
1.2
1.2
0.25 / 0
0.126 / 0.125
0.126 / 0.125
400
400
400
92.4
90.7
93.2
244
246
245
3.7
3.1
3.4
0.2
0.3
0.2
230
228
232
11
9.7
10.5
0.5
0.2
0.4
92.0
90.7
92.5
a–f ) Same as for Table 1.
Table 4. Cyclocondensation reaction of TMHQ with IPh under reflux of BuAc in the reductive environment of sodium hyposulfite;
molar ratio of reactants, catalysts and sodium hyposulfite: TMHQ : IPh : ZnCl2 : Na2S2O4 = 1.03 : 1 : 0.2 : 0.02
Charged amounts
IPh ZnCl2
mol
Reaction products
TPha
Yield of TPh
TMHQ
BuAc
mL
Sodium hyposulfite
mol
%
1.24
1.24
1.24
1.2
1.2
1.2
0.25
0.25
0.25
400
400
400
0.02
0.02
0.02
92.6
90.1
91.3
93.5
91.8
91.5
a) Content of TPh in raw TPh.
containing less quinonoid compounds. The difference
was, however, not significant and these results show
that in laboratory conditions, the solvent vapors re-
leased during the reflux represent a sufficient barrier
against TPh oxidation.
by a solvent exchange to BuAc allows for the ZnCl2
catalyst to be recycled with over 98 % efficiency with
the catalyst loss of less than 2 %. The recycled cat-
alyst allowed the yield of all-rac-α-tocopherol of over
90 %, while achieving virtually total conversion of IPh
at only a 3 % molar excess of TMHQ. The reaction
carried out under nitrogen atmosphere or in the pres-
ence of a reducing agent provided results comparable
to those obtained using refluxed solvent.
In another comparative experiment, the influence
of water on the quality and yield of tocopherol was
assessed. Since water and BuAc are immiscible, only
aqueous solution of ZnCl2, without BuAc, was added
to the reaction mixture. The reaction mixture was
placed on a rotary evaporator and heated to 120◦C, at
which temperature TMHQ and IPh form a homoge-
neous liquid. Water was partially lost by evaporation
and the solution of ZnCl2 got more concentrated. In
order to facilitate water removal, vacuum was used (6
kPa). When the reaction had run for 1 hour, the ob-
tained mixture was dissolved in BuAc and worked up
under standard conditions as described above. As can
be expected, the yield of tocopherol decreased below
60 %. In addition, the tocopherol content decreased
to about 70 % in the raw products. Most importantly,
the yield of the secondary products amounted to more
than three fold compared to the standard experiment.
More detailed analysis of the influence of the wa-
ter content and vacuum will be published separately
(Go¨mo¨ry et al., 2011).
Acknowledgements. Financial support from the Slovak
Grant Agency VEGA 1/0768/08 is gratefully acknowledged.
References
Baldenius, K.-U., Bussche-Hu¨nnefeld, L.v.d., Hilgemann, E.,
Hoppe, P., & Stu¨rmer, R. (1996). Ullmann’s encyclopedia
of industrial chemistry (pp. 478–488). Weinheim, Germany:
Wiley–VCH.
Bonrath, W., Aquino, F., Haas, A., Hoppmann, S., Netscher, T.,
Pace, F., & Pauling, H. (2009). Methanetrisulfonic acid: A
highly efficient strongly acidic catalyst for Wagner-Meerwein
rearrangement, Friedel-Crafts alkylation and acylation reac-
tions. Examples from vitamin E synthesis. Sustainability, 1,
161–168. DOI: 10.3390/su1020161.
Bonrath, W., Dittel, C., Giraudi, L., Netscher, T., & Pabst,
T. (2007a). Rare earth triflate catalysts in the synthesis of
Vitamin E and its derivatives. Catalysis Today, 121, 65–70.
DOI: 10.1016/j.cattod.2006.11.022.
Bonrath, W., Eggersdorfer, M., & Netscher, T. (2007b). Catal-
ysis in the industrial preparation of vitamins and nutraceu-
ticals. Catalysis Today, 121, 45–57. DOI: 10.1016/j.cattod.
2006.11.021.
Conclusions
Applying multiple extractions with water followed

436
J. G¨om¨ory et al./Chemical Papers 65 (4) 427–436 (2011)
Bonrath, W., & Netscher, T. (2005). Catalytic processes in vi-
tamins synthesis and production. Applied Catalysis A: Gen-
eral, 280, 55–73. DOI: 10.1016/j.apcata.2004.08.028.
Coman, S. M., Wuttke, S., Vimont, A., Daturi, M., & Kemnitz,
E. (2008). Catalytic performance of nanoscopic, aluminium
trifluoride-based catalysts in the synthesis of (all-rac)-α-
tocopherol. Advanced Synthesis & Catalysis, 350, 2517–2524.
DOI: 10.1002/adsc.200800411.
Kemnitz, E., Coman, S. M., Ru¨diger, S., & Wuttke, S. (2009).
European patent application No. EP 2050743 A1. European
patent office.
Schager, F.,
& Bonrath, W. (2000). Synthesis of D,L-α-
tocopherol using ‘microencapsulated’ catalysts. Applied
Catalysis A: General, 202, 117–120. DOI: 10.1016/S0926-
860X(00)00456-7.
The British Pharmacopoeia Commission (2010). The British
pharmacopoeia (pp. 1–4). London, UK: The Stationary Of-
fice.
Cotton, F. A., Wilkinson, G., & Gaus, P. L. (1995). Basic in-
organic chemistry. New York, NY, USA: Wiley.
Go¨m¨ory, J., Králik, M.,
&
Kaszonyi, A. (2011). Towards
Wang, H., & Xu, B.-Q. (2004). Catalytic performance of
the synthesis of (all-rac)-α-tocopherol from trimethylhydro-
quinone and isophytol at reduced pressure. Acta Chimica
Slovaca, 4(1), 3–14.
Nafion/SiO2 nanocomposites for the synthesis of α-tocophe-
rol. Applied Catalysis A: General, 275, 247–255. DOI:
10.1016/j.apcata.2004.07.038.
Go¨m¨ory, J., Králik, M., Kaszonyi, A., & Mravec, D. (2010).
Study of (all-rac)-α-tocopherol synthesis from trimethylhy-
droquinone and isophytol at the presence of solid catalysts.
Acta Chimica Slovaca, 3(2), 110–121.
Wang, S., Bonrath, W., Pauling, H., & Kienzle, F. (2000).
The synthesis of D,L-α-tocopherol in supercritical media.
The Journal of Supercritical Fluids, 17, 135–143. DOI:
10.1016/S0896-8446(99)00050-9.
Hinze, R., Laufer, M. C., H¨olderich, W. F., Bonrath, W., &
Netscher, T. (2009). The use of Nafion/silica composite cat-
alysts for synthesis of fine chemicals. Catalysis Today, 140,
105–111. DOI: 10.1016/j.cattod.2008.07.008.
Xing, H., Wang, T., & Dai, Y. (2009). Continuous synthesis
of D,L-α-tocopherol catalyzed by sulfonic acid-functionalized
ionic liquid in supercritical carbon dioxide. The Journal
of Supercritical Fluids, 49, 52–58. DOI: 10.1016/j.supflu.
2008.12.003.