Toxaphene chemistry: Separation and characterisation of selected enantiomers of the Polychloropinene mixtures

Article abstract of DOI:10.1016/j.chemosphere.2006.05.082

The primary goal for the study presented here was the preparation and characterisation of enantiomeric pure chlorobornane standards (Toxaphene). In this context, we partially modeled the procedure for Polychloropinene production in the former USSR. The initial reaction was ionic addition of hydrogen chloride to (1S)-α-pinene resulting predominantly in (1S)-2-endo-chlorobornane. Further photochlorination gave mixtures of chlorinated terpenes with different average content of Cl per molecule. The resulting mixtures were separated on a silica-gel column and a number of known hepta to decachlorobornanes were identified in fractions with the help of NMR and GC (using electron capture and mass spectrometric detection) - but in very unusual ratios as compared to the technical Toxaphene mixture formerly produced by Hercules (USA). Also several previously unknown congeners were isolated or detected. Three of the isolated congeners were obtained in crystalline state and X-ray crystallography showed their enantiomeric purity.

Full text of DOI:10.1016/j.chemosphere.2006.05.082

Chemosphere 67 (2007) 1695–1700
www.elsevier.com/locate/chemosphere
Toxaphene chemistry: Separation and characterisation of selected
enantiomers of the Polychloropinene mixtures
a
a
b
c
Alexey Trukhin , Fedor Kruchkov , Lars Kr. Hansen , Roland Kallenborn ,
a
a,
*
Anastasia Kiprianova , Vladimir Nikiforov
a
Department of Chemistry, St. Petersburg State University, Universitetsky pr. 26, 198504 St. Petersburg, Russia
b
Department of Chemistry, University of Tromsø, 9037 Tromsø, Norway
Norwegian Institute for Air Research (NILU), 9296 Tromsø, Norway
c
Accepted 26 May 2006
Available online 10 January 2007
Abstract
The primary goal for the study presented here was the preparation and characterisation of enantiomeric pure chlorobornane stan-
dards (Toxapheneꢀ). In this context, we partially modeled the procedure for Polychloropinene production in the former USSR. The ini-
tial reaction was ionic addition of hydrogen chloride to (1S)-a-pinene resulting predominantly in (1S)-2-endo-chlorobornane. Further
photochlorination gave mixtures of chlorinated terpenes with different average content of Cl per molecule. The resulting mixtures were
separated on a silica-gel column and a number of known hepta to decachlorobornanes were identified in fractions with the help of NMR
and GC (using electron capture and mass spectrometric detection) – but in very unusual ratios as compared to the technical Toxapheneꢀ
mixture formerly produced by Hercules (USA). Also several previously unknown congeners were isolated or detected. Three of the iso-
lated congeners were obtained in crystalline state and X-ray crystallography showed their enantiomeric purity.
ꢁ 2006 Elsevier Ltd. All rights reserved.
Keywords: Polychloroterpenes; Pinene photochlorination; Enantiomeric composition; X-ray analysis; NMR characterisation
1. Introduction
Natural camphene occurs in a number of essential oils in
(+)-form as well as (ꢀ)-form (Windholz et al., 1976). Pro-
duction of camphene starts from natural a-pinene, which is
also chiral. All types of camphene were used for Toxa-
pheneꢀ synthesis by different manufacturers. Some congen-
ers in selected samples of technical Toxaphene were shown
to have small deviations from racemic distribution (Buser
Toxaphene is an insecticidal mixture (average formula
C10H10Cl8), produced as technical product by Hercules
Agrochemical Inc. (Passaic, NJ, USA) through the con-
trolled photochlorination of camphene (Saleh, 1991),
Fig. 1.
In the USSR, a similar insecticide – Polychloropinene
(average formula C10H10.5Cl7.5) was produced by chlorina-
tion of a-pinene with AIBN (azobisisobutyronitrile) as ini-
tiator with preliminary hydrochlorination in the cyclic
reactor; hydrogen chloride evolved through the chlorina-
tion is returned to the reactor for hydrochlorination,
Fig. 1 (Nikiforov, 2002).
and Muller, 1994). Nevertheless, there is no information
¨
about any enantiomerically pure congeners of Toxaphene,
although deviations from the racemic distribution were
reported for technical Melipaxꢀ mixtures, a similar prod-
ucts as Toxapheneꢀ produced in the former German Dem-
ocratic Republic (Vetter et al., 1997).
The amount of Polychloropinene produced in the USSR
is estimated to be 160000 tons, which is a significant
contribution to the total global production. There is no
information on composition of Soviet Polychloropinene
and unfortunately, there is no traceable sample of
*
Corresponding author. Tel.: +7 812 4284079; fax: +7 812 5912630.
E-mail address: vovanikiforov@yahoo.co.uk (V. Nikiforov).
0045-6535/$ - see front matter ꢁ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.chemosphere.2006.05.082

1696
A. Trukhin et al. / Chemosphere 67 (2007) 1695–1700
Cl2, hν
Clx
Clx
+
+
+
Clx
Clx
-HCl
Camphene
Cl2, AIBN, t
HCl (gas)
Clx
Cl
α-Pinene
Fig. 1. Preparation of Toxaphene by chlorination of camphene and preparation of Soviet Polychloropinene by hydrochlorination and chlorination of
a-pinene.
Polychloropinene available. In the present work, we report
on the preparation of model Polychloropinene mix-
tures with different content of chlorine in an attempt to find
the difference between Toxaphene and Soviet Polychloro-
pinene.
evaporated. The solid was then recrystallised from a mini-
mum amount of n-hexane to give 20 g of crude (1S)-2-
endo-chlorobornane ([a]D ꢀ26ꢂ to ꢀ32ꢂ, Et2O, c = 2).
2.1.2. Preparation of ‘‘7–8’’ mixture
At the same time it was an attempt to prepare pure
enantiomers of important Toxaphene congeners from opti-
cally active pinene, to establish their configuration and thus
to provide more opportunities for enantiomer-specific envi-
ronmental analysis of Toxaphene residues.
Crude (1S)-2-endo-chlorobornane (8.5 g) was added to
1000 ml of 0.4 M solution of Cl2 in CCl4. The solution
was irradiated with a UV-lamp until the yellow color disap-
peared. GC analysis confirmed that the product contains
mainly hepta and octachloroterpenes. After evaporation
of the solution, 20 g of colorless oil was obtained.
2. Materials and methods
2.1.3. Preparation of ‘‘9–10’’ mixture
Seven grams of crude (1S)-2-endo-chlorobornane was
dissolved in 350 ml of carbon tetrachloride. Resulting
solution was saturated with chlorine for 10 min and then
irradiated with UV-lamp until until the yellow color disap-
peared. This saturation–irradiation cycle was repeated four
times. After evaporation of the solution 16 g of colorless oil
was obtained.
(1S)-a-pinene (Aldrich, St. Louis, MO, USA) used for
this study has an enantiomeric excess of 81+%. Gaseous
hydrogen chloride was prepared from sodium chloride
and concentrated sulfuric acid with subsequent scrubbing
with concentrated sulfuric acid. Chlorine gas in cylinder
was obtained from a commercial source. Photochlorination
experiments were carried out with a 400 W medium-pres-
sure mercury arc lamp. GC equipment used for monitoring:
Fisons MEGA with ECD (Fisons, Milan, IT), chiral capil-
lary column based on b-cyclodextrin (BGB 172 capillary
column, 15 m length, 0.25 mm i.d., 0.25 mm film thickness,
BGB Analytics, Boeckten, Switzerland). Helium (5.0 qual-
ity) was used as carrier gas (Hydro gas, Porsgrunn, Nor-
way). An aliquot of 2 ll was injected on-column in the
enantioselective GC column. The ECD detector was held
at 300 ꢂC. The following temperature programme was
used for separation: initial temperature: 60 ꢂC (isothermal
2 min) ꢀ temperature rate 3 ꢂC/min to 180 ꢂC · Tempera-
ture rate 5 ꢂC/min to 280 ꢂC (isothermal 5 min) · 15 ꢂC/
min to 300 ꢂC (final temperature).
2.1.4. Separation of the polychloroterpenes mixture
Separation of the prepared mixtures (16 g of ‘‘9–10’’
mixture or 20 g of ‘‘7–8’’ mixture) was carried out on a
‘‘home-made’’ silica column: i.d. 4.7 cm; height – 100 cm;
eluent – n-hexane. Each fraction was examined by 1H
NMR. Selected fractions were then purified by additional
silica column or through multiple crystallisations.
2.1.5. Dechlorination of polychloroterpenes with copper
The combined fractions from separation of mixture ‘‘9–
10’’ where Parlar no. 50 (7, Fig. 3) was enriched, were
evaporated to dryness to yield ca. 400 mg of semi-solid res-
idue. Fifty milliliters of glacial acetic acid and 1 g of Cu
powder were added to it and the mixture was refluxed for
10 h. The mixture was allowed to cool, diluted with
200 ml of water, the solid collected and extracted with
3 · 10 ml of boiling n-hexane. The combined extracts were
separated on a 60 cm long, 2 cm i.d. silica column with n-
hexane as eluent. In addition to Parlar no. 50, four previ-
ously unknown polychlorobornenes (3–6, Fig. 3) were
isolated.
2.1. Syntheses
2.1.1. Hydrochlorination of a-pinene
Dry hydrogen chloride was bubbled through the solution
of 28 g (0.2 mole) (1S)-a-pinene ([a]D ꢀ47ꢂ, neat) in 50 ml of
n-hexane for 6 h at ambient temperature. The resulting
solution was washed with 10% aqueous sodium hydrocar-
bonate, dried with anhydrous sodium sulfate and solvent

A. Trukhin et al. / Chemosphere 67 (2007) 1695–1700
1697
2.2. X-ray crystallography
mixture of (1S)-2-endo-chlorobornane and 2-endo-chloro-
fenchane in ca. 10:1 ratio (Fig. 2). The crude product was
found to be optically active ([a]D ꢀ26ꢂ to ꢀ32ꢂ, Et2O,
c = 2); we concluded that enantiomeric purity is retained,
though we had no data for specific rotation of chlorofench-
ane. Formation of two different intermediate ion pairs with
non-classical carbocations was postulated to explain the
composition of the product mixture. Rearrangement
includes 1,2-shift of either methylene or isopropylidene
bridge with formation of less strained norbornane carbon
sceleton.
Data collection: Enraf-Nonius CAD-4 diffractometer,
CAD-4-PC Software (Enraf-Nonius, 1992). Cell refine-
ment: CELDIM in CAD-4-PC Software. Data reduction:
XCAD4 (McArdle and Higgins, 1995). Program used to
solve and refine structure: SHELXS97 (Sheldrick, 1997).
Molecular graphics: ORTEX (McArdle, 1995).
2.3. NMR spectroscopy
The crude 2-endo-chlorobornane (bornyl chloride) was
then used for chlorination without further purification.
Chlorination was conducted in light-induced conditions
in contrast to industrial synthetic procedure (thermal initi-
ation with AIBN). Two different mixtures were prepared.
‘‘7–8’’ contained mainly hepta and octachlorinated com-
pounds, and ‘‘9–10’’ was obtained by exhaustive chlorina-
tion and contained nona and decachloroterpenes. Both
mixtures of chlorinated terpenes were separated on silica-
gel column with n-hexane as eluent according to a previ-
ously developed procedure (Nikiforov et al., 1995).
All NMR experiments were performed on dilute CDCl3
1
solutions at 30 ꢂC. For the new compound 11 H and 13C
DEPT spectra were recorded in the temperature range
ꢀ30 to +24 ꢂC, at intervals of ca. 10 ꢂC. In addition to
measurements at ꢀ30 ꢂC, NOESY-EXSY and EXSY spec-
tra for the mentioned compound were recorded. The NMR
spectrometer used for serial analysis and the chemical shift
assignments was a Varian Mercury 400 MHz equipped
with a z-gradient accessory, low temperature measurements
were carried out with a Bruker DPX-300 (Bruker, Karl-
sruhe, Germany).
The chemical shift assignments for new compounds are
based on two-dimensional 1H–1H COSY and 1H–13C
COSY.
3.2. Composition of Polychloropinene-like mixtures
On GC/ECD (gas chromatography with electron cap-
ture detection) all mixtures had peak patterns different
from those on chromatograms of mixtures of chlorinated
camphene with similar content of chlorine.
3. Results and discussion
3.1. Preparation of Polychloropinene-like mixtures
A number of known hexa to decachlorobornanes were
identified in fractions with help of NMR and GC
(Fig. 3). By the time of writing, 11 polychlorobornanes
The initial reaction was ionic addition of hydrogen
chloride to (1S)-a-pinene ([a]D ꢀ47ꢂ, neat) resulting in a
HCl (gas), hexane
H
+
H
Cl
Cl
(1
S
)-2-endo
-
(1R)-2-endo-
chlorofenchane
(1
S
)-(-)-
α-Pinene
chlorobornane
91-93 %
7-9 %
Cl
H
Me
Me
Me
Me
Cl
Me
Me
Me
Cl
H
H
H
H
H
H
H
H
Me
Me
Me
Cl
Me
Me
Me
Me
Me
H
H
H
H
H
H
H
H
Cl
Me
H
Me
Me
Cl
Cl
Cl
Fig. 2. Hydrochlorination of optically active a-pinene and its schematic mechanism.

1698
A. Trukhin et al. / Chemosphere 67 (2007) 1695–1700
Cl2HC
CH2Cl
Cl2HC
Cl
CH2Cl
Cl
ClH2C
CH2Cl
ClH2C
Cl
CH2Cl
Cl
ClH2C
Cl
CHCl2
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
CH2Cl
CHCl2
CH2Cl
CH2Cl
CH2Cl
Cl
Cl
4
1
2
3
5
Cl2HC
Cl
CH2Cl
ClH2C
Cl
CH2Cl
Cl2HC
Cl
CHCl2
ClH2C
Cl
CHCl2
ClH2C
Cl
CH2Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
CHCl2
Cl
CHCl2
CHCl2
CHCl2
CHCl2
Cl
Cl
Cl
Cl
6
9
10
7
8
ClH2C
CHCl2
Cl2HC
Cl
CH2Cl
Cl2HC
Cl
CH2Cl
Cl
ClH2C
Cl
CHCl2
Cl
Cl
Cl
Cl2HC
Cl
CH2Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
CHCl2
11
CHCl2
Cl
Cl
CHCl2
13
CHCl2
CHCl2
15
Cl
Cl
Cl
12
14
Fig. 3. Structures of polychloroterpenes obtained from a-pinene.
Table 1
and 4 polychlorobornenes (after treatment of fractions
with Cu in boiling acetic acid) were isolated in sufficiently
pure state to establish or confirm their structures by NMR.
Compounds 1, 7–10, 12–15 are known chlorination
products of camphene. Compounds 2–6 and 11 were previ-
ously unknown.
Composition of perchlorinated bornyl chloride (Mixture ‘‘9–10’’) and
perchlorinated Toxaphene
Relative content of selected
polychlorobornanesa
7
9
12
13
14
15
Mixture ‘‘9–10’’
Perchlorinated Toxaphene 2.3
2.8
2.0
4.8
2.0
2.4
3.7
0.4
1
1
2.0
2.0
From ‘‘7–8’’ mixture, two compounds were isolated: the
well-known Toxicant B (1; Parlar 32) and previously
unknown 2-endo,3-exo,5-exo,6-exo,8,9,10-heptachlorobor-
nane (2). Ring substitution pattern of the latter: 2-endo,
3-exo,5-exo,6-exo – has never been observed in polychloro-
bornanes prepared from camphene. Theoretically, this
heptachlorobornane can be obtained by chlorination of
camphene, but its presence in environmental samples
would likely point to Polychloropinene as a source of pol-
lution. None of the typical Toxapheneꢀ congeners (except
Toxicant B) were major components in fractions. Several
other new compounds were identified, their structure eluci-
dation is in progress.
Mixture ‘‘9–10’’ contained the same main components
as perchlorinated camphene, 2-exo,10-dichlorobornane,
2-exo,10,10-trichlorobornane or Toxaphene – two nona-
chlorobornanes (7, 9) and four decachlorobornanes – 12–
15. 13 is a major component of the mixture ‘‘9–10’’.
This is completely different from what we see in perchlori-
nated camphene or Toxapheneꢀ, where 13 is the smallest
of the mentioned six congeners (Table 1). Also the ratio
of Parlar 50 (7) and Parlar 62 (9) is opposite for the
two mixtures (‘‘9–10’’ and perchlorinated Toxapheneꢀ,
respectively).
a
Determined by NMR spectra.
eluting compound was the well-known decachlorobornane
12, the last eluting was the previously unknown 2-endo,
3-exo,5,5,8,8,9,10,10-nonachlorobornane (11). Its struc-
ture was established by NMR and then confirmed by
X-ray crystallography (Hansen et al., 2004). The com-
pound has no chlorine atoms in positions 2-exo or
6-exo. The ring substitution pattern ‘‘2-endo,3-exo,5,5-’’
has never been observed in polychlorobornanes prepared
from camphene. Thus, it is confirmed that Polychloro-
pinene-specific polychlorobornanes exist and can serve
as indicator compounds for source elucidation in the
environment.
3.3. Structural differences and similarities of Toxaphene
and Polychloropinene
• All polychlorobornanes isolated from chlorination
products of pinene contain at least one Cl in 2-endo or
6-endo positions.
Two compounds, the first and last major eluting compo-
nents from the silica column, were purified to >99% by suc-
cessive crystallisations from enriched fractions. The first
• Environmentally relevant Toxapheneꢀ congeners
2-exo,3-endo,6-exo,8,9,10-hexachlorobornane (Hx-Sed),
2-exo,3-endo,5-exo,9,9,10,10-heptachlorobornane (B7-

A. Trukhin et al. / Chemosphere 67 (2007) 1695–1700
1699
1000), 2-exo,3-endo,5-exo,8,9,9,10,10-octachloroborn-
ane (Parlar no. 41), 2-exo,5,5,8,9,9,10,10-octachloro-
bornane (Parlar no. 44) have not been detected among
chlorination products of pinene.
• No chlorobornanes with unsubstituted bridgehead
methyl group have yet been detected among chlorina-
tion products of pinene.
• Ratios of common polychlorobornanes must have
been different in Toxapheneꢀ and Polychloropinene,
because the ratios of major nona and decachloroborn-
anes are different in their respective perchlorination
products.
3.4. Characterisation of selected polychloroterpenes
The main method for identification of known polychlo-
roterpenes was NMR. NMR data for new compounds with
provisional assignments is listed in Table 2.
1
A number of H-signals in NMR spectrum of 11 are
broad. NMR investigation of 11 at different temperatures
revealed the molecule has slow internal rotation of the
C10-dichloromethyl moiety. This effect causes broadening
1
of H NMR signals of protons in its proximity (Koivisto
et al., 2001). At ꢀ40 ꢂC, all signals are sharp and the struc-
ture corresponds to the most stable conformation found in
the crystalline state.
3.5. Enantioselectivity during preparation of
polychlorobornanes from optically active a-pinene
One of the goals of our project was development of a
general enantioselective synthetic method for polychloro-
terpenes. We had the technical means to use three methods
for observation of the enantiomeric purity of isolated
compounds:
• measurement of optical rotation;
• GC analysis on chiral phases;
• direct resolution of structure by X-ray crystallography.
Fig. 4. Chiral-GC/ECD profiles of 7 (Parlar 50): (a) standard solution,
Promochem GmbH and (b) part of a chromatogram of ‘‘9–10’’ mixture.

1700
A. Trukhin et al. / Chemosphere 67 (2007) 1695–1700
(1R,2S,3S,4R,7S)-2-endo,3- (1S,2S,3S,4S,5R,6S)-2-endo,3-
(1S,4S,5S,6S)-2,2,5-endo,6-
exo,5,5,8,8,9,10,10-
nonachlorobornane (11)
exo,5-exo,6-exo,8,9,10-
heptachlorobornane (2)
exo,8,9,10-heptachloro-
bornane (1, Parlar no. 32)
Fig. 5. ORTEP plots for X-ray investigated polychlorobornanes.
active bornyl chloride ! enantiomers of polychloro-
bornanes.
With regard to differences and similarities in structures
Considering minuscule amounts of isolated pure com-
pounds (milligrams or tens of milligrams) and low solubil-
ity in organic solvents (1–5%), measurement of optical
rotation required a sensitive instrument, which was avail-
able to us only at the very beginning of the project. There-
fore we managed to measure optical rotation for just one
compound – isolated from mixture ‘‘9–10’’ congener 12:
[a]D 10ꢂ 2 (CCl4, c = 1 mg cmꢀ3).
A number of isolated compounds were tested on chiral
GC/ECD using t-butyldimethylsilylated b-cyclodextrin as
chiral selector (BGB 172 capillary column, 15 m length,
0.25 mm i.d., 0.25 mm film thickness, BGB Analytics,
Boeckten, Switzerland). In all cases we observed single
peaks. Minor peaks, identified as enantiomers or isomeric
impurities were usually <10% compared to the main enan-
tiomer signal in the chromatogram. At least, in one case we
achieved confirmation of a single enantiomer (compound 7,
Parlar 50). For this compound we had a racemic standard,
which gave well-resolved enantiomers on chiral GC/ECD
chromatogram. The comparison of compound 7, isolated
from mixture ‘‘9–10’’ and from standard solution con-
firmed the enantiomeric purity (Fig. 4).
and ratios of polychloroterpenes related to Toxapheneꢀ
and to Polychloropinene, our findings suggest that the ear-
lier Soviet Polychloropinene, unlike Toxaphene contained
non-racemic polychloroterpenes. Therefore observed devi-
ations from racemic composition in the environment may
in some samples be due to partial input of a non-racemic
source, and not only to enantioselective degradation or
metabolism.
Acknowledgements
This work was generously supported by Norwegian Re-
search Council, project number 152806/730. We gratefully
thank Arnfinn Kvarsnes from the University of Tromsø,
Norway and Stanislav Selivanov from St. Petersburg State
University, Russia for their invaluable help with NMR
experiments.
References
Three of the isolated congeners yielded crystals of suffi-
cient quality for direct X-ray determination of structure –
1, 2 from mixture ‘‘7–8’’ and 11 from mixture ‘‘9–10’’.
All three congeners were found to be single enantiomers
with preservation of the configuration of starting (1R)-
endo-2-chlorobornane (Fig. 5).
Based upon the above information, we concluded that
the three independent methods used in our study confirm
the preservation of the enantiomeric purity during hydro-
chlorination and subsequent chlorination of optically
active a-pinene. No evidence of racemization was found
during our experiments.
Buser, H.R., Muller, M.D., 1994. J. Agric. Food Chem. 42, 393–400.
¨
Enraf-Nonius, 1992. CAD-4-PC Software. Version 1.1 Enraf-Nonius,
Delft, The Netherlands.
Hansen, L.K., Kallenborn, R., Kiprianova, A., Nikiforov, V., Trukhin,
A., 2004. Acta Cryst. E60, o1089–o1091.
Koivisto, J.J., Kolehmainen, E.T., Nikiforov, V.A., Nissinen, M.J.,
Tuppurainen, K.A., Pera¨kyla¨, M., Miltsov, S.A., Karavan, V.S.,
2001. ARKIVOC 3, 95–113.
McArdle, P., 1995. J. Appl. Cryst. 28, 65.
McArdle, P., Higgins, T., 1995. XCAD. National University of Ireland,
Galway, Ireland.
Nikiforov, V.A., 2002. Organohalogen Compd. 59, 315–318.
Nikiforov, V.A., Tribulovich, V.G., Karavan, V.S., 1995. Organohalogen
Compd. 26, 379–382.
This implies that the strategy of enantioselective prepa-
ration of lower chlorinated terpene and its further free-rad-
ical chlorination, followed by column chromatography was
the right strategy for the preparation of enantiomerically
pure chlorobornane materials. The general pathway in
the synthesis of enantiomeric pure chloroterpenes was iden-
tified as follows: optically active a-pinene ! optically
Saleh, M.A., 1991. Rev. Environ. Contam. Toxicol. 118, 1–85.
Sheldrick, G.M., 1997. SHELXL97 and SHELXS97. University of
Go¨ttingen, Germany.
Vetter, W., Klobes, U., Luckas, B., Hottinger, G., 1997. Chromatographia
45, 255–262.
Windholz, M., Budavari, S., Stroumtsos, L.Y., Fertig, M.N., 1976. The
Merck Index: An Encyclopedia of Chemicals and Drugs, ninth ed.
Merck, Rahway, NJ, p. 1734.