Nucleophilic reactions of charge delocalised carotenoid mono- and dications

Article abstract of DOI:10.1039/b406913g

In the present study insight was gained on the larger complexity of cationic mixtures of diaryl (φ,φcarotene, isorenieratene) and aliphatic (Φ,Φ-carotene, lycopene) carotenes, prepared by reaction with BF ₃-etherate, compared with β,β-carotene. Chemical reactions of the mono- and dications prepared in situ from the allylic carotenols β,β-caroten-4-ol (isocryptoxanthin) and β,β-carotene-4,4'- diol (isozeaxanthin), and from isorenieratene and lycopene were investigated using selected O, N and S nucleophiles; water, methanol, azide and thioacetate. In total 22, including 18 new, neutral carotenoid products were isolated and identified by VIS, MS and NMR (in part) spectroscopy. Their structures were compatible with the structures of the cationic intermediates. The formal addition of hydride to the various dications, required to rationalise minor reaction products, is discussed in terms of more likely hydrogen radical or proton transfer in cationic reactions. Extensive E/Z isomerisation was observed for all quenching products. The potential use of carotenoid cations for the synthesis of 4,(4')-substituted β,β-carotenes and 7-oxabicycloheptane derivatives is discussed.

Full text of DOI:10.1039/b406913g

View Article Online / Journal Homepage / Table of Contents for this issue
A R T I C L E
Nucleophilic reactions of charge delocalised carotenoid mono- and
dications†
Geir Kildahl-Andersen, Liv Bruås, Bjart Frode Lutnaes and Synnøve Liaaen-Jensen*
Department of Chemistry, Norwegian University of Science and Technology (NTNU), NO-7491,
Trondheim, Norway. E-mail: slje@chem.ntnu.no; Fax: +47 73594256; Tel: +47 3594097
Received 10th May 2004, Accepted 2nd July 2004
First published as an Advance Article on the web 10th August 2004
In the present study insight was gained on the larger complexity of cationic mixtures of diaryl (φ,φ-carotene,
isorenieratene) and aliphatic (,-carotene, lycopene) carotenes, prepared by reaction with BF3-etherate, compared with
,-carotene. Chemical reactions of the mono- and dications prepared in situ from the allylic carotenols ,-caroten-4-ol
(isocryptoxanthin) and ,-carotene-4,4′-diol (isozeaxanthin), and from isorenieratene and lycopene were investigated
using selected O, N and S nucleophiles; water, methanol, azide and thioacetate. In total 22, including 18 new, neutral
carotenoid products were isolated and identified by VIS, MS and NMR (in part) spectroscopy. Their structures were
compatible with the structures of the cationic intermediates. The formal addition of hydride to the various dications,
required to rationalise minor reaction products, is discussed in terms of more likely hydrogen radical or proton transfer in
cationic reactions. Extensive E/Z isomerisation was observed for all quenching products. The potential use of carotenoid
cations for the synthesis of 4,(4′)-substituted ,-carotenes and 7-oxabicyclo[2,2,1]heptane derivatives is discussed.
Nucleophilic reactions of the dication 6
Introduction
Previously we have prepared from ,-carotene (1) the ,-carotene
dication (2) by treatment with BF3-etherates in CHCl3 solution, and
elucidated the structure of this charge delocalised cation by NIR and
detailed NMR spectroscopy at −15 °C.1,2
Subsequently we have obtained from ,-caroten-4-ol (3,
isocryptoxanthin) the monocation 4 (as four stereoisomers) by
reaction with CF3COOH in CH2Cl2 solution.3,4 The corresponding
bisallylic diol ,-carotene-4,4′-diol (5, isozeaxanthin) was con-
verted to the dication 6 by similar treatment, and to the symmetrical
dicarbocation 7 (three stereoisomers) by reaction with the stronger
Brønsted acid CF3SO3H.4
,-Carotene-4,4′-diol (5) or its diacetate 12 was used as substrate
for generating the dication 6, which was subsequently reacted with
water as nucleophile. In both cases E/Z isomerised isozeaxanthin
(5) was the major product. Methanolysis of the dication 6 provided
mainly the E/Z isomerised dimethyl ether 13, characterised by VIS,
NMR and mass spectrometry.
When sodium azide was employed as the nucleophile, NMR and
MS analyses identified the monoazide 14, and the diazide 15, as
major products, Scheme 3. Attempts to convert the produced azides
6
to amines by hydrogenation with Lindlar catalyst5 or with SnCl2
failed.
The structure elucidation of the monocation 4 and the dications
6 and 7 was conducted in the same manner by detailed NMR
spectroscopy, including COSY, HSQC, HMBC and 2D ROESY
techniques. The structures of the delocalised carotenoid cations
prepared are given in Scheme 1. The distribution of the charge is
illustrated by the diameters of the filled circles.
An alternative approach to gain structural information on the
delocalised carotenoid cations is to study their reactions with suit-
able nucleophiles, including detailed product analysis.
The reactions of the ,-carotene dication 2 (20  electrons) with
H2O as nucleophile (in acetone) provided isocryptoxanthin (3),
isocarotene (8), mutatochrome (9) and ,-carotene (1).2
Treatment of the dication 2 with CH3ONa/CH3OH furnished
the methyl ethers 10 and 11 and isocarotene (8). The reactions,
summarised in Scheme 2, were rationalised and consistent with the
dication structure determined by NMR spectroscopy.2
Reactions of the dication 6 with thioacetic acid as nucleophile
provided three main products, which were tentatively identified as
the monothioester 16, the dithioester 17,7 and the hydroxythioester
18, after partial characterisation by HPLC, VIS, MS and by H
NMR (18 only).
Reduction of the dithioester 17 to the thiol with LiAlH4 failed
according to earlier experience.7 Attempted conversion of 16, 17
and 18 to the free thiols using KOH resulted in the formation of
isocarotene (8), judged by HPLC, VIS and MS data.
Other carotenoid thioacetates8 and azides9 have been prepared
previously by different approaches.
1
Nucleophilic reactions of the symmetrical dication 7 (22 
electrons)
The symmetrical dicarbocation 7, obtained from ,-carotene-4,4′-
diol (5) with CF3SO3H, was reacted with H2O in acetone. Product
analysis including HPLC, TLC and spectroscopy (VIS, MS, and
in part 1H NMR) resulted in plausible identification of three major
E/Z isomerised products as the carotenols 19, 20 and 4′-hydroxy-
,-caroten-4-one (21), Scheme 4 (for definition of P1, P2, P3, and
P4, see Schemes 2 and 3). Product 19 with nonaene chromophore
had molecular ion m/z 550, and MS showed M − 92, but no M −
H2O fragment ions. This product, characterised by NMR, was
resistant to acetylation. Product 20 with decaene chromophore had
the same molecular ion as 19 and also no fragment ion for loss of
water upon MS. Product 21 had max 462 nm, showed strong M −
H2O ion upon electron impact MS, and gave isozeaxanthin (5) upon
NaBH4 reduction.
In the present work we report on the reactions of the charge
delocalised monocation 4, and dications 6 and 7 with selected O,
N and S nucleophiles.
Also included are nucleophilic reactions performed with cations
prepared in situ from carotenes with two aryl or two aliphatic end
groups.
Results
Nucleophilic reaction of the monocation 4
The monocation 4, resulting from ,-caroten-4-ol (3) treated with
CF3COOH, was converted to E/Z isomerised 3 when reacted with
H2O as nucleophile in acetone.
Treatment of the dicarbocation 7 with methanol as a nucleophile
provided several products analysed by HPLC, TLC, and VIS and
mass spectroscopy. HPLC revealed strong E/Z isomerisation of each
† No. 6 in the series Charged Carotenoid Species. No. 5 = Ref. 4.
2 4 9 6
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 4 9 6 – 2 5 0 6
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4

View Article Online
Scheme 1
Scheme 2
Scheme 3
product with up to eight peaks (isomers). Plausible structures were
assigned based on relative polarity, VIS spectra (max and spectral
fine-structure, %III/II10) in comparison with known chromophores,
including 3,4-didehydro-,-carotene, and MS data (molecular
weight, number of double bond equivalents and fragmentation
patterns). Products 8 and 22–25 are given in Scheme 4 according to
polarity by TLC, assuming that 25a had undergone dehydration to
25b during work-up or MS analysis.
Isorenieratene (26) provided with BF3-etherate in CHCl3 or
CH2Cl2 a product with max 910 nm. EPR studies revealed the
presence of free radicals in the reaction mixture, Fig. 1. A line
width of ca. 13 G was compatible with a delocalised carotenoid
cation radical.11 The product mixture was less stable than for the
,-carotene dication (2) and NMR analysis at −15 °C revealed
a complex mixture where the expected dication 28 could not be
identified. A minor monocation 29, occasionally formed, was
structurally identified, Scheme 5.4
Hydrolysis of the cationic intermediate with H2O in acetone
provided the mono-ol 30, identified by VIS and MS data, and a
diol 31, identified by polarity, VIS and MS data. The by HPLC
least polar products were tentatively identified as diastereomers of
the furanoxide 32, judged by VIS data, M = 544 (C40H48O) by MS,
and reaction with CF3COOH to a blue, tentative oxonium ion 33.
Reversion to max 412 nm upon addition of KOH to 33 is compatible
with the formation of the hemiketal (10-ol).12
The formation of the products obtained by nucleophilic reactions
from the dications 6 and 7 is rationalised in the Discussion.
Preparation, characterisation and nucleophilic reactions of the
isorenieratene (26) and lycopene (27) cations
As a supplement to the dicyclic cationic substrates prepared
from ,-carotene (1), its 4-ol (3) and 4,4′-diol (5) derivatives,
it was considered of interest to examine nucleophilic reactions
of cations prepared in situ from carotenes with two aryl end
groups, isorenieratene (26), and with two aliphatic end groups,
lycopene (27).
Lycopene (27), treated with BF3-etherate, provided a product
with max 940 nm in CHCl3 and max 920 nm in CH2Cl2, which
according to NIR data appeared to be relatively stable.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 4 9 6 – 2 5 0 6
2 4 9 7

View Article Online
length of the polyene system/product stability was a decisive factor
in directing the nucleophilic attack. This was the case even when
the highest partial positive charge was located in the middle of the
polyene system.
The dication 6 with a protonated hydroxy group has a good leav-
ing group for SN2 reactions. Substitution at C-4′ is evident from
all products obtained from 6 with various nucleophiles, Scheme 3.
The hydroxy substituent in product 18 is considered as a remnant
of the protonated hydroxy group in 6. As to the fate of the second
delocalised charge, addition of the nucleophile to the C-4 position
was generally the case. For the minor products 14 and 16, formally
a hydride has been added to C-4, see below.
The products (5, 13–18) obtained with water, methanol, azide
or thioacetic acid as nucleophiles are compatible with the structure
assigned to the dication 6, and demonstrate the potential use of
this dication in partial synthesis for obtaining C-4,4′ disubstituted
derivatives of ,-carotene (1). No in-chain substitution appeared
to be favoured for this delocalised cation (6).
The symmetrical dication 7 was reacted with water (aqueous
acetone) or methanol as nucleophiles, Scheme 4. With methanol,
disubstitution at C-4,4′ (25a), monosubstitution at C-4 (22, 24) and
elimination of one proton (23, 24, 25b) were noted. The formation
of products 19 and 23 is rationalised in Scheme 7. The hydroxy
substituent of 24 and 25a might be the result of incomplete dication
formation, cf. 6, of ,-carotene-diol (5) with CF3SO3H. The
didehydro product isocarotene (8) formally requires one hydride
transfer.
The reactions of the dication 7 with water resulted partly in
C-4,4′ disubstitution (21, considered obtained by ready oxidation
of product 5), and gave partly in-chain hydroxylation (19, 20), as
rationalised for product 19 in Scheme 7.
In conclusion, the products obtained by reaction of the symmetri-
cal dication 7 with selected nucleophiles (water and methanol) were
compatible with the structure assigned to this dication.
Scheme 4
The large extent of E/Z isomerisation for the products isolated by
reaction of 6 and 7 with nucleophiles was evident when employing
an HPLC system suitable for the separation of geometrical isomers
of carotenoids, and is consistent with product formation from
delocalised carotenoid cations.1–3 As an illustration is included the
HPLC profile of the E/Z isomerised reaction mixture containing
mainly isocarotene (8), obtained from ,-caroten-4-ol (3) by treat-
ment with dilute HCl via the monocation 4, cf. ref. 3, Fig. 3.
The strong E/Z isomerisation observed for all the carotenoid
products is characteristic for the products formed in these reactions,
as might be expected from the low rotational barrier of the cationic
intermediates.13 However, the structures of the carotenoid cations
determined by NMR have shown the all-E configuration to be
preferred also in the charged state.1,2,4
Fig. 1 EPR spectrum of isorenieratene (26) treated with BF3-dee.
Except for the double bond connecting the end group to the
polyene chain in retro carotenoids, only the E-configuration was
observed in these cations. Similar observations have previously
been made for ,-diphenyl polyenyl anions.14,15 This preference
for the all-E configuration has been explained by a higher charge
delocalisation, which is due to a better overlap of the -orbitals.
There seem to be two possible explanations for this apparent
anomaly. Either the E/Z isomerisation occurs only upon the
quenching of the cations, or, more likely, the Z isomers are present
also in the charged state, but without being observed in the NMR
experiments. If the latter explanation is correct, the lifetime of the Z
isomers in the charged state must be too short to be detected within
the NMR time frame.
MS showed prominent ions at M + 4, M + 2, M, M − 2 and
M − 4. EPR studies showed a line width of ca. 13 G, compatible
with the presence of delocalised free radicals.11 The amount of
radicals was reduced over time, Fig 2.
NMR studies of the cation prepared from lycopene (27)
were unsuccessful.4 When the in situ prepared cationic mixture
of lycopene (27, M = 536) was reacted with H2O in acetone a
large number of products (>70 peaks) were apparent by HPLC.
Two major fractions isolated by TLC had MS compatible with
M = 536 + H2O and M = 536 + 2 O, both fractions containing
products with nonaene and decaene chromophores. Dication 34
and product 35, Scheme 6, are referred to later.
Discussion
Nucleophilic reactions of ,-carotene dication (2) related
dications
Nucleophilic reactions of monocation 4 and dications 6 and 7
As to carotenoid cations prepared by reaction with BF3-etherates
from hydroxylated ,-carotenes we have previously examined
the reactions of (2R)-,-caroten-2-ol (36) and (2R,2′R)-,-
carotene-2,2′-diol (37), resulting in the formation of the 7-
oxabicyclo[2.2.1]heptane derivatives 38 and 39, assumed to occur
by internal attack by the hydroxy groups of a C-5(5′)-cation,16
When the monocation 4, obtained from ,-caroten-4-ol (3)
treated with CF3COOH,3,4 was hydrolysed with H2O in acetone,
E/Z isomerized 3 was obtained. The result is consistent with the
structure assigned to the delocalised monocation 4 and leading to
carotenol formation with maximum length of the polyene chain.
Also in the following, it was usually experienced that maximum
2 4 9 8
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 4 9 6 – 2 5 0 6

View Article Online
Scheme 5
of the assumed furanoxide 32 can be depicted by analogy with the
formation of the furanoxide mutatochrome (9, Scheme 2) from the
,-carotene dication (2).2
The reaction products obtained from the isorenieratene (26)
cation mixture are consequently compatible with the presence of
the dication 28 and the monocation 29. Moreover, the reaction
products encountered, as well as the monocation 29, support neg-
ligible delocalisation of the charge of 28 and 29 to the aromatic
end groups, which would cause a reduction of the Hückel (4n + 2)
aromaticity.
As for isorenieratene (26), the reaction of lycopene (27) with
BF3-etherate proceeds via free radicals. Athough absorption was
observed in the NIR region no resolution in the NMR spectrum
could be obtained. The opportunity for intermolecular cyclisations
of cationic intermediates such as the hypothetical dication 34 is
obvious. Quenching with H2O in acetone revealed a very com-
plex reaction mixture. Major coloured reaction products exhibited
mainly decaene and nonaene chromophores. The apparent addition
of H2O to lycopene (27) might involve a formal addition of hydride
to the dication 34, and addition of H2O to the resulting monocation,
providing a mono-ol with decaene chromophore (M = 554)
The alternative addition of 2 oxygens (M = 568) points to
epoxide type products. Previously the formation of the 5,6-diol 35
requiring M = 572, Scheme 6, identified by chemical correlations,
has been claimed from similar experiments.17 This diol 35, would
be a predicted product of the dication 34.
Fig. 2 EPR spectrum of lycopene (27) treated with BF3-dee.
Electron impact mass spectra recorded of freshly prepared
cationic mixtures of isorenieratene (26) and lycopene (27) showed
no evidence of BF3 addition products, cf. refs. 2,16,17. For lycopene
(27), the MS might support hydrogen transfer in the ionised state,
see below.
Scheme 6
Scheme 8. The relevant dications 40 and 41 obtained from 36 and
37 can now be formulated by analogy with the dication 2, prepared
from ,-carotene (1), Scheme 8. The relatively high positive
charges at C-5(5′) serve to explain the course and good yield of
these nucleophilic reactions.
Hydrogen transfer reactions of carotenoids
Hydride as leaving group has been documented during the reac-
tion of carotenoid 5,8-furanoxides with very strong acids, forming
delocalised oxonium ions,12 cf. 32 and 33, Scheme 6. Also the
formation of the monocation 42 from echinenone (43) reacted with
CF3COOH requires H− as a leaving group,4,18 Scheme 9. In all these
cases, H2 must be produced.
The rationalisation of products 8 (from 7), 14, 16, 22, and 29
encountered here formally requires the addition of a hydride to a
dication, and has represented a puzzle. The combined evidence
demonstrates that
Nucleophilic reactions of cations obtained from isorenieratene
(26) and lycopene (27)
The predicted dication 28 of isorenieratene (26), Scheme 5, could
not be identified in the complex reaction mixture by NMR. The
identified monocation 29 might be a product of the dication 28
formed by a formal hydride attack, vide infra. Reaction with H2O
in acetone provided the mono-ol 30, which may be rationalised by
reaction of the monocation 29 with H2O, whereas the diol 31 is an
expected product from the hypothetic dication 28. The formation
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 4 9 6 – 2 5 0 6
2 4 9 9

View Article Online
Scheme 7
Scheme 9
i) Although higher proportions of products with apparent
Fig. 3 HPLC (system 1) of the reaction mixture of isocryptoxanthin
(3) treated with 0.03 N HCl in CHCl3. For peak identification, see the
experimental part.
hydride-addition have been observed in particular solvent batches,
no selective transfer of deuterium from either CDCl3 or CD3OD has
been observed.
ii) Formal hydride transfer is observed using both BF3-etherates
and CF3COOH as reagent for generating carotenoid cations. Free
radicals are involved in the former, but not in the latter reaction. In
strong CF3COOH, H− cannot survive.
iii) Hydrogen transfer occurs as a minor reaction (products usu-
ally in the order of 10% of the total) to a positively charged carbon
atom at the end of the delocalised -electron system, resulting in
cations, which subsequently react with added nucleophiles. The
monoazide 14 and monothioacetate 16 were obtained from the
dication 6, Scheme 3. Moreover, the methyl ether 22 was obtained
from the symmetrical dication 7.
The above evidence suggests that the formal hydride transfer
may occur by reactions of the various dications, Scheme 10.
However, referring to A, no experimental support has so far been
obtained for the existence of a trication of this type, or derivatives
thereof.
Scheme 10
Scheme 8
2 5 0 0
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 4 9 6 – 2 5 0 6

View Article Online
The 7,8-dihydro-8-methoxy derivative 11 previously obtained
from the ,-carotene dication (2) was then rationalised as a
hydrogen radical attack of an intermediary radical cation,2 see B,
Scheme 10. For carotenoid cations generated by treatment with
BF3-etherate, radical reactions involving transfer of hydrogen radi-
cals remain an open alternative. For isorenieratene (26) treated with
BF3-etherate, the monocation 29 was isolated,4 as well as product
30 when the reaction was quenched with water, Scheme 5. In this
case the monocation with a hydrogen added was detected by NMR.
The intensity of the signals from the monocation 29 was constant
during the experiments (ca. 10 h), indicating that the H-transfer
was taking place during the formation of the cations only, and that
the neutral carotenoid was the H-donating species, compatible with
routes B or C.
However, products 14, 16 and 22 were formed in radical-free
media (CF3COOH). Here reaction C involving reaction between
dication and neutral carotenoid with H+ transfer is a plausible alter-
native to reaction A.
In summary, it appears more likely that hydrogen radical trans-
fer reactions (reaction B), rather than hydride transfer reactions
(reaction A) occur in free radical media, whereas hydride transfer
(reaction A) or proton transfer (reaction C) are required in radical-
free media. In strongly acidic media (CF3COOH or CF3SO3H)
proton transfer (C) is more likely.
with a direct inlet to the ion source, 70 eV, ion source temperature
250 °C. NMR spectra were obtained in CDCl3 on a Bruker Avance
DPX 400 instrument using a 5 mm QNP probe, or a Bruker Avance
DRX 500 instrument using a 5 mm TXI probe. Chemical shifts are
cited relative to TMS with calibration against residual CHCl3 at
7.27 ppm for 1H.
General methods for the analysis of the cations by NIR and NMR
are given elsewhere.4
EPR spectra were recorded on a Bruker ESP 300E instrument,
rectangular cavity, flat cell.
HPLC was carried out on a Hewlett Packard Series 1050
instrument equipped with a diode array detector (DAD). Detection
wavelengths were set at 420, 450 and 480 nm. VIS spectra
of the carotenoid components were recorded on-line during
chromatography. Spectral fine structure is reported as %III/II.10
Three different HPLC systems were used:
System 1,19 WatersYMC Carotenoid C30 column, 250 × 4.6 mm.
Mobile phase 0 min: methanol:tert-butyl methyl ether:water
(81:15:4 v/v/v, 1.0 ml min−1), 60 min: methanol:tert-butyl methyl
ether:water (31:65:4 v/v/v, 1.0 ml min−1), 70 min: methanol:tert-
butyl methyl ether:water (16:80:4 v/v/v, 1.0 ml min−1).
System 2, Interchrom Uptisphere 5 ODB column, 250 × 4.6 mm.
Mobile phase 0 min: methanol:acetone (100:0 v/v 1.0 ml min−1),
100 min: methanol:acetone (0:100 v/v, 1.0 ml min−1).
System 3, Interchrom Uptisphere 5 ODB column, 250 × 4.6 mm.
Mobile phase 0 min: methanol:acetone (90:10 v/v, 1.0 ml min−1),
90 min: methanol:acetone (0:100 v/v, 1.0 ml min−1).
Preparative TLC was carried out on self-made TLC plates
(silica:calcium carbonate 2:1).
Conclusions
The reactions of the cations 4, 6 and 7, prepared in situ from allylic
,-carotenols with various O, N and S nucleophiles (water in
acetone, methanol, sodium azide and thioacetate) were investi-
gated. Product analysis included TLC, HPLC-DAD, VIS, MS and
NMR spectroscopy, and chemical derivatisation reactions. In total,
22 neutral carotenoid products were identified, including 18 new
carotenoids.
The potential use of the dications 6 and 7 for the synthesis of
4,(4′)-substituted ,-carotenes, and of the 2-monohydroxy and
2,2′-dihydroxy analogues of the dication 2 for the preparation of
7-oxabicyclo[2.2.1]heptane derivatives is pointed out.
Preparation of substrates (3, 5, 12, 26). ,-Caroten-4-ol
(isocryptoxanthin, 3) and ,-carotene-4,4′-diol (isozeaxanthin, 5)
were prepared by NaBH4-reduction from synthetic ,-caroten-4-
one (echinenone, 43) and ,-carotene-4,4′-dione (canthaxanthin).4
,-Carotene-4,4′-diol (5) was converted to its diacetate (12) by
standard procedure. φ,φ-Carotene (isorenieratene, 26) was prepared
by total synthesis.4
Insight was gained into the larger complexity of cationic mixtures
of diaryl- and aliphatic carotenes, prepared by reaction with BF3-
etherate, compared with ,-carotene (1).
Treatment of ,-caroten-4-ol (3) with acids
VIS/NIR and NMR experiments.
0.03 M HCl at room temperature. ,-Caroten-4-ol (3, ca. 20 g)
was dissolved in 0.03 M HCl in CHCl3 (3 ml). The VIS/NIR spec-
trum was recorded at 5 min intervals. An immediate bathochromic
shift to max 479 nm was observed, as well as a weak absorption in
NIR around 1010 nm. A slight decrease in both intensity and fine
structure (%III/II) of the main peak occurred during 20 min.
The structures of all products with nucleophiles were rationalised
from those of the delocalised cations. The formal hydride addition to
carotenoid dications, required to rationalise the formation of some
products (8, 14, 16, 22, 30), has been discussed in terms of cationic
reations involving hydrogen radical or proton transfer reactions.
All carotenoid products obtained from the delocalised cations
were strongly E/Z isomerised, according to HPLC analysis,
compatible with cationic intermediates.
0.03 M HCl at −20 °C. The experiment was performed as
described above with the cuvette placed in a cuvette holder cooled
by cold methanol (−20 °C) from a cryostat. The change of absorp-
tion maximum progressed much slower, giving approximately equal
intensity of max 464 nm and max 487 nm after 20 min. Initially, the
peak around 1010 nm was more pronounced than at room tempera-
ture, with 20% of the intensity of the main peak. However, the NIR
absorption decayed rapidly to ca. 25% of the initial absorption dur-
ing 20 min.
Experimental
Materials
Commercially available CF3COOH from Acros and Merck and
CF3COOD from Acros were used. CF3SO3H and CF3SO3D were
supplied from Aldrich. Ethanol stabiliser present in CHCl3 was
removed on an alumina column prior to use. Pyridine was distilled
from BaO. BF3-diethyl etherate (BF3-dee) was obtained fromAcros.
Carotenoids used were synthetic samples obtained from Hoffmann-
La Roche, except isorenieratene (26), which was prepared by total
synthesis.4
CF3COOH at −20 °C. ,-Caroten-4-ol (3, ca. 5 g) was
dissolved in CHCl3 (3 ml) containing CF3COOH (0.013 M) at
−20 °C. The VIS/NIR spectrum was recorded with cooling as
described above. An absorption maximum of max 977 nm was
measured, with 16% loss of intensity during 30 min. Exchange of
CHCl3 for CH2Cl2 gave max 1028 nm, with <10% decay observed
during 3 h.
General methods
Spectroscopy and chromatography. Reactions and mani-
pulations were carried out as far as possible in darkness and
under nitrogen atmosphere. Visible light (VIS) and near infra-
red (NIR) spectra were recorded on a Varian Cary 50 UV-VIS
Spectrophotometer (190–1100 nm) or a Varian Cary 5 UV-VIS-
NIR Spectrophotometer (170–3500 nm). EI mass spectra were
recorded on a Finnigan MAT 95XL ThermoQuest spectrometer
For NMR analysis at −10 °C of the monocations 4, see refs. 3,4.
Elimination of water with 0.03 M HCl.
At room temperature. ,-Caroten-4-ol (3, ca. 20 g) was
dissolved in 0.03 M HCl in CHCl3 (3 ml). After 20 min, the reac-
tion mixture was washed with water and saturated NaCl solution
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 4 9 6 – 2 5 0 6
2 5 0 1

View Article Online
and analysed by HPLC, Fig. 3. HPLC (System 1) RT = 28.1 (5%,
mixture turned blue. Cooled methanol (10 ml) was added after
10 min stirring, giving an orange solution. The pigments were
transferred to hexane and washed with water and saturated NaCl


max/nm 475; %III/II 0; unidentified carotenoid), RT = 40.2 (12%,
max/nm 465; %III/II 0; 3,4-didehydro-,-carotene), RT = 50.7
(8%, max/nm 439sh 461 491; %III/II 28; 4′,5′-didehydro-4,5′-
retro-,-carotene, 8), RT = 52.1 (6%, max/nm 440 465 495; %III/II
39; Neo D 8 (For neo-nomenclature see ref. 20)), RT = 55.0 (7%,
solution. Pigment recovery 56% (E1%,
mixture was characterised without further purification.
2500). The reaction
1 cm
4,4-Dimethoxy-,-carotene (13). max(hexane)/nm 424sh 449
475, %III/II 28, %AB/AII 13; HPLC (System 1) RT = 18.7 (12%,

max/nm 442 465 495; %III/II 36; 8), RT = 55.7 (5%, max/nm 443sh
469 498; %III/II 48; Neo B 8), RT = 60.2 (16%, max/nm 444sh 467
497; %III/II 38; Neo A 8), RT = 65.1 (12%, max/nm 446 472 503;
%III/II 51; all-trans 8); HPLC (System 3) RT = 15.9 (8%, max/nm
425sh 451 479; %III/II 40; 3), RT = 16.8 (16%, max/nm 423sh 449
479; %III/II 30; 3), RT = 29.4 (7%, max/nm 439 459 491; %III/II
60; unidentified carotenoid), RT = 30.7 (15%, max/nm 461; %III/II
0; 3,4-didehydro-,-carotene), RT = 33.6 (22%, max/nm 441 465
495; %III/II 28; 8), RT = 34.2 (21%, max/nm 444 471 499; %III/II
43; 8).




max/nm 420sh 445 471; %III/II 18; 13-cis 13), RT = 21.8 (55%,
max/nm 424sh 451 478 %III/II 37; all-trans 13), RT = 25.8 (4%,
max/nm 420sh 446 473; %III/II 44; 9-cis 13), RT = 28.4 (14%,
max/nm 426sh 452 479; %III/II 33; unidentified carotenoid);
H(400 MHz, CDCl3) 1.02 (6H, 16/17/16′/17′-H), 1.04 (6H,
16/17/16′/17′-H), 1.26 (lipid), 1.37 (2/2′-H), 1.68 (2H, 2/2′-H),
1.77 (3/3′-H), 1.81 (6H, 18/18′-H), 1.82 (3/3′-H), 1.97 (19/19′-H),
1.98 (20/20′-H), 3.40 (6H, -OCH3), 3.52 (2H, 4/4′-H), 6.10–6.23
(6H, 7/8/10/7′/8′/10′-H), 6.27 (2H, 14/14′-H), 6.38 (2H, 12/12′-H,
J11,12 = J11′,12′ = 14.8 Hz), 6.62–6.68 (4H, 11/15/11′/15′-H); m/z (EI)
597 (47%, M + 1), 596 (M+, 100), 566 (32, M − HCOH), 564 (30,
M − CH3OH), 534 (14), 532 (16, M − 2 CH3OH), 504 (13, M − 92),
474 (5, M − 92 − HCOH), 472 (5, M − 92 − CH3OH), 442 (3, M −
92 − HCOH − CH3OH), 440 (2, M − 92 − 2 CH3OH).
At −20 °C. ,-Caroten-4-ol (3, ca. 20 g) was dissolved in
0.03 M HCl in CHCl3 (2.5 ml) at −20 °C. After 5 min an aliquot
was withdrawn and washed with water and saturated NaCl solution.
The product mixture was analysed by HPLC. HPLC (System 1)
RT = 24.2 (6%, max/nm 429sh 450 479; %III/II 50; all-trans 3),
RT = 26.9 (7%, max/nm 470; %III/II 0; unidentified carotenoid),
RT = 43.5 (6%, max/nm 431sh 459 487; %III/II 40; 8), RT = 48.0
(5%, max/nm 433sh 459 489; %III/II 50; Neo E 8), RT = 49.7 (8%,
Reaction of dication 6 (from 5) with NaN3. ,-Carotene-
4,4′-diol (5, 1.3. mg) was dissolved in CH2Cl2 (5 ml) and cooled
to −10 °C. CF3COOH (40 l) was added, whereupon the reaction
mixture turned blue. After 10 min stirring, a cooled solution of
NaN3 (2.0 g) in water (5 ml) was added, giving a two-phase system.
After vigorous stirring, the organic layer turned brown/orange. The
pigments were transferred to hexane and washed with water and
saturated NaCl solution to give an orange solution, max(hexane)/nm

max/nm 438 460 489; %III/II 40; 8), RT = 51.1 (6%, max/nm 438
465 495; %III/II 45; 8), RT = 53.4 (6%, max/nm 440 461 491; %III/
II 50; 8), RT = 54.3 (7%, max/nm 440sh 465 493; %III/II 40; Neo
B 8), RT = 55.0 (5%, max/nm 442 471 499; %III/II 50; 8), RT = 59.8
(15%, max/nm 445sh 467 495; %III/II 45; NeoA8), RT = 64.8 (11%,

max/nm 449 471 503; %III/II 60; all-trans 8).
10
426sh 449 476, %III/II 26, %AB/AII 14; HPLC (System 1) RT
16–33 min. Preparative TLC developed with 10% acetone in hexane
and eluted with acetone gave two main products:
Isomerisation of ,-caroten-4-ol (3). To ,-caroten-4-ol (3,
0.35 mg), dissolved in CH2Cl2 (5 ml) at −10 °C, CF3COOH (25 l)
was added, whereupon the reaction mixture turned blue. A cooled
solution of 25% H2O in acetone (10 ml) was added, giving a yellow
reaction mixture. The pigments were transferred to hexane and
the organic phase washed with water and saturated NaCl solution,
pigment recovery 89% (E1%, 1 cm 2500), max(hexane)/nm 425sh 448
475; %III/II 18. HPLC (System 1) RT = 17.4 (9%, max/nm 422sh
444 470; %III/II 11; 13-cis 3), RT = 18.3 (8%, max/nm 420sh 444
470; %III/II 13; 3), RT = 21.0 (63%, max/nm 428sh 451 478; %III/II
33; all-trans 3), RT = 23.7 (9%, max/nm 422sh 447 474; %III/II 36;
9-cis 3).
4-Azido-,-carotene (14, 38% yield). RF 0.85–0.96;

max(acetone)/nm 426sh 451 476, %III/II 6; HPLC (System 1)
RT = 25.8 (9%, max/nm 420sh 445 470; %III/II 17; 13-cis 14),
RT = 26.5 (9%, max/nm 420sh 444 472; 14), RT = 30.5 (54%,


max/nm 426sh 452 480; %III/II 34; all-trans 14), RT = 32.8 (11%,
max/nm 420sh 448 473; 14), RT = 34.4 (4%, max/nm 420sh 448
472; 14); H(400 MHz, CDCl3) 1.04–1.07 (12H, 16/17/16′/17′-
H), 1.26 (lipid), 1.47 (2-H), 1.49 (2′-H), 1.63 (3′-H), 1.67 (2-H),
1.73 (3H, 18′-H), 1.84 (3H, 18-H), 1.89 (3-H), 1.95 (3-H), 1.98
(12H, 19/20/19′/20′-H), 2.03 (4′-H), 3.72 (1H, 4-H), 6.13–6.21
(6H, 7/8/10/7′/8′/10′-H), 6.27 (2H, 14/14′-H), 6.36 (1H, 12′-H,
J11′,12′ = 14.9 Hz), 6.39 (1H, 12-H, J11,12 = 14.6 Hz), 6.63–6.70 (4H,
11/15/11′/15′-H); m/z (EI) 578 (10%, M + 1), 577 (M+, 21), 549
(31, M − N2), 536 (32), 535 (51, M − N3), 534 (100, M − HN3), 442
(21, M − 92 − HN3).
Treatment of ,-carotene-4,4′-diol (5) with CF3COOH
VIS/NIR and NMR experiments at −20 °C. ,-Carotene-
4,4′-diol (5, 16 g) was dissolved in CH2Cl2 containing CF3COOH
(0.013 M) at −20 °C. The VIS/NIR spectrum was recorded with
cooling as described above. max 980 nm was measured, with <20%
loss of intensity observed during 3 h.
6
Attempted reduction with SnCl2 failed to give the desired 4-
amino-,-carotene.
NMR analysis at −20 °C of dication 6, see ref. 4.
4,4′-Diazido-,-carotene (15, 53% yield). RF 0.70–0.85;
max(acetone)/nm 426sh 451 478, %III/II 20; HPLC (System 1)
Reaction of dication 6 (from 5) with water. ,-Carotene-
4,4′-diol (5, 0.57 mg) was dissolved in CH2Cl2 (5 ml), and cooled
to −10 °C. CF3COOH (25 l) was added, whereupon the reaction
mixture turned blue. A cooled solution of 25% H2O in acetone
(8 ml) was added after 10 min stirring, giving an orange solution.
The pigments were transferred to hexane and washed with water
and saturated NaCl solution. Pigment recovery 100% (E1%, 1 cm

RT = 22.5 (16%, max/nm 420sh 443 469; %III/II 17; 13-cis 15),
RT = 25.4 (60%, max/nm 424sh 451 478; %III/II 41; all-trans 15),
RT = 29.4 (11%, max/nm 422sh 446 473; %III/II 41; 9-cis 15);
H(400 MHz, CDCl3) 1.04 (6H, 16/17/16′/17′-H), 1.06 (6H, 16/17/
16′/17′-H), 1.26 (lipid), 1.48 (2/2′-H), 1.67 (2H, 2/2′-H), 1.84 (6H,
18/18′-H), 1.88 (3/3′-H), 1.96 (3/3′-H), 1.98 (19/20/19′/20′-H),
3.72 (2H, 4/4′-H), 6.09–6.21 (6H, 7/8/10/7′/8′/10′-H), 6.28 (2H,
14/14′-H), 6.39 (2H, 12/12′-H, J11,12 = J11′,12′ = 15.0 Hz), 6.62–6.69
(4H, 11/15/11′/15′-H); m/z (EI) 619 (8%, M + 1), 618 (M+, 17),
590 (6, M − N2), 576 (18, M − N3), 575 (22, M − HN3), 548 (31,
M − N2 − N3), 547 (32, M − N2 − HN3), 535 (28), 534 (67), 533
(35), 532 (61, M − 2 HN3), 517 (6), 484 (4, M − 92 − N3), 456 (14,
M − 92 − N3 − N2), 442 (14, M − 92 − 2 N3), 440 (14, M − 92 − 2
HN3), 43 (100).
10
2500), max(hexane)/nm 424sh 448 474, %III/II 22, %DB/DII 15;
HPLC (System 1) RT = 10.5 (5%, max/nm 419sh 443 470; %III/II
33; 5), RT = 11.8 (55%, max/nm 426sh 450 477; %III/II 27; all-trans
5), RT = 17.0 (6%, max/nm 422sh 447 474; %III/II 61; 9-cis 5),
RT = 21.7 (15%, max/nm 426sh 452 480; %III/II 51; unidentified
carotenoid).
Reaction of dication 6 (from 5) with methanol. ,-Carotene-
4,4′-diol (5, 1.0. mg) was dissolved in CH2Cl2 (6 ml), and cooled
to −10 °C. CF3COOH (40 l) was added, whereupon the reaction
Attempted hydrogenation with Lindlar catalyst5 did not give the
desired 4,4′-diamino-,-carotene.
2 5 0 2
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 4 9 6 – 2 5 0 6

View Article Online
Reaction of dication 6 (from 5) with CH3COSH. ,-Carotene-
4,4′-diol (5, 1.7 mg) was dissolved in CH2Cl2 (6 ml), and cooled
to −10 °C. CF3COOH (40 l) was added. After 10 min stirring,
a cooled solution of 25% CH3COSH in acetone (1.25 ml) was
added. The pigments were transferred to hexane and washed
with 5% NaHCO3, 0.01 M NaOH and saturated NaCl solution.
cooled to −10 °C. CF3COOH (25 l) was added, whereupon the
reaction mixture turned blue. A cooled solution of 25% H2O in
acetone (10 ml) was added after 10 min stirring, giving an orange
solution. The pigments were transferred to hexane and washed
with water and saturated NaCl solution. Pigment recovery 95%
(E1%, 1 cm 2500), max(hexane)/nm 424sh 448 475, %III/II 23; HPLC
(System 1) RT = 10.5 (5%, max/nm 424sh 445 474; %III/II 70; 5),
RT = 11.8 (62%, max/nm 426sh 450 477; %III/II 30; all-trans 5),
RT = 16.1 (5%, max/nm 426sh 452 480; %III/II 97; unidentified
carotenoid), RT = 17.0 (5%, max/nm 424sh 447 476; %III/II 83;
9-cis 5), RT = 21.6 (6%, max/nm 483; mixture).

max(hexane)/nm 410 441; HPLC (System 1) RT 15–42 min. Pre-
parative TLC developed with 10% acetone in hexane and eluted
with acetone gave five main products, of which all had a distinct
smell of the parent thioacid:
Unknown A (12% yield). RF 0.88–0.96; max(acetone)/nm 430sh
Treatment of ,-carotene-4,4′-diol (5) with CF3SO3H
453 478sh.
VIS/NIR and NMR experiments at −15 °C. ,-Carotene-4,4′-
diol (5, ca. 10 g) was dissolved in CH2Cl2 containing CF3SO3H
(0.011 M) at −15 °C. The VIS/NIR spectrum was recorded as
described above. max 1020 nm was measured, with 18% loss of
intensity observed during 1 h. However, the total intensity decay
after 21 h storage at −20 °C was not greater than 22%.
Unknown B (6% yield). RF 0.77–0.88; max(acetone)/nm 429sh
453 479sh.
4-Acetylthio-,-carotene (16, 17% yield). RF 0.69–0.77;
max(acetone)/nm 429sh 453 477, %III/II 2; HPLC (System 1)

RT = 27.2 (10%, max/nm 420sh 446 472; 13-cis 16), RT = 28.8
(4%), RT = 32.2 (54%, max/nm 428sh 452 480; %III/II 24, all-
trans 16), RT = 35.0 (14%, max/nm 422sh 448 475; 16); m/z (EI)
611 (12%, M + 1), 610 (M+, 27), 536 (79), 535 (60), 534 (100,
M − CH3COSH), 533 (10), 532 (17), 456 (17), 442 (20, M − 92 −
CH3COSH).
For NMR analysis at −20 °C see ref. 4.
Reaction of dication 7 (from 5) with water. ,-Carotene-
4,4′-diol (5, 1.3 mg) was dissolved in CH2Cl2 (6 ml), and cooled
to −15 °C. CF3SO3H (5 l), dissolved in CH2Cl2 (2 ml), was added,
whereupon the reaction mixture turned blue. An aliquot was with-
drawn and examined by VIS/NIR, showing max 1010–1020 nm.
However, the colour changed to pale yellow upon dilution prior
to VIS/NIR. The reaction mixture was stirred for 10 min, with no
colour changes observed. A cooled solution of 25% H2O in acetone
(8 ml) was added, and the pigments were transferred to hexane and
washed with water and saturated NaCl solution. A green colour
appeared during evaporation. N-Ethyldiisopropylamine (4 drops)
was added and the yellow colour reappeared. Pigment recovery
62% (E1%, 1 cm 2500), max(hexane)/nm 447; HPLC (System 2) RT
9–41 min. Products were isolated by preparative TLC developed
with 20% acetone in hexane and eluted with acetone to give:
4′-Acetylthio-,-caroten-4-ol (18, 19% yield). RF 0.60–0.69;

max(acetone)/nm 429sh 455 478, %III/II 1; HPLC (System 1)
RT = 31.0 (5%, max/nm 422sh 446 470sh; 18), RT = 31.3 (7%,
max/nm 424sh 447 474sh; 18), RT = 35.3 (68%, max/nm 428sh 454
479, %III/II 21; all-trans 18), RT = 38.6 (5%, max/nm 424sh 449
476; %III/II 24; 18), RT = 39.2 (5%, max/nm 424sh 449 476; %III/II
22; 18); m/z (EI) 627 (1.5%, M + 1), 626 (M+, 3.1), 610 (12), 608
(11, M − H2O), 536 (36), 535 (36, M − H2O − CH3COS), 534 (68,
M − H2O − CH3COSH), 533 (19), 532 (30), 518 (8), 456 (6), 442
(13, M − 92 − H2O − CH3COSH), 91 (100).

4,4′-Diacetylthio-,-carotene (17, 17% yield). RF 0.47–0.60;
max(acetone)/nm 428sh 454 479, %III/II 1; HPLC (System 1)
4′,5′-Didehydro-4,5′-retro-,-carotene (8, 32%). RF 0.88–0.96;
max(acetone)/nm 447 470 497sh.


RT = 25.2 (13%, max/nm 420sh 446 471; %III/II 8; 13-cis 17),
RT = 28.9 (65%, max/nm 426sh 453 480; %III/II 23; all-trans
17), RT = 33.0 (9%, max/nm 424sh 448 472; %III/II 28; 9-cis 17);
H(400 MHz,CDCl3)1.03(16/17/16′/17′-H),1.05(16/17/16′/17′-H),
1.50 (2/2′-H), 1.65 (2/2′-H), 1.75 (3/3′-H), 1.96 (6H, 19/19′-H),
1.98 (6H, 20/20′-H), 2.13 (3/3′-H), 4.14 (2H, 4/4′-H), 6.11–6.19
(6H, 7/8/10/7′/8′/10′-H), 6.26 (2H, 14/14′-H), 6.37 (2H, 12/12′-H,
J11,12 = J11′,12′ = 14.9 Hz), 6.61–6.68 (4H, 11/15/11′/15′-H), NMR
spectrum shows low purity; m/z (EI) 685 (6%, M + 1), 684 (M+,
12), 610 (16), 609 (18, M − CH3COS), 608 (13, M − CH3COSH),
536 (19), 535 (31), 534 (58), 533 (20), 532 (23, M − CH3COSH),
442 (20), 43 (100). Attempted reductions of 17 with LiAlH4 and
LiEt3BH failed to give the desired ,-carotene-4,4′-dithiol.
Hydrolysis with 10% KOH in methanol of a crude reaction mix-
ture dissolved in diethyl ether gave, after aqueous work-up, mainly
isomerised 4′,5′-didehydro-4,5′-retro-,-carotene (8), max(diethyl
ether)/nm 445sh 467 494, %III/II 4; HPLC (System 1) RT = 44.3
(5%, max/nm 440 462 492; %III/II 41; 8), RT = 46.4 (7%, max/nm
440 462 492; %III/II 42; 8), RT = 48.0 (10%, max/nm 440 466 495;
%III/II 38; Neo D 8), RT = 51.2 (5%, max/nm 441 466 495; %III/II
43; 8), RT = 51.9 (6%, max/nm 445 470 500; %III/II 53; Neo B
8), RT = 56.6 (20%, max/nm 446 470 499; %III/II 43; Neo A 8),
RT = 61.8 (24%, max/nm 447 473 504; %III/II 48; all-trans 8); m/z
(EI) 536 (17%, M + 2), 535 (45, M + 1), 534 (M+, 100), 442 (16,
M − 92), 428 (17, M − 106).
3,4-Didehydro-4′,9-retro-,-caroten-9-ol (19, 19%). RF
0.78–0.84; max(acetone)/nm 421 443 470, %III/II 30; HPLC
(System 2) RT = 25.1 (57%, max/nm 419 441 467; %III/II 34; 19),
RT = 25.6 (27%, max/nm 418 442 471; %III/II 53; 19), RT = 41.9
(16%, max/nm 453 479; %III/II 16; unidentified carotenoid); H
(500 MHz, CDCl3) 1.32 (16′/17′-H), 1.54 (2′-H), 1.66 (19-H), 1.89
(18-H), 1.92 (18′-H), 1.98–2.02 (20/19′/20′-H), 2.14 (3′-H), 4.67
(8-H), 5.80 (4/4′-H), 5.96 (7-H), 6.40 (7′-H), 6.80 (8′-H) (Impure);
m/z (EI) 551 (45, M + 1), 550 (M+, 100), 458 (19, M − 92), 444 (17,
M − 106).
An aliquot of product (19) was dissolved in pyridine (1.0 ml) and
acetic anhydride was added (0.1 ml). Work-up after 18 h and subse-
quent analysis by TLC revealed no formation of new products.
3,4-Didehydro-4′,7-retro-,-caroten-7-ol (20, 11%). RF
0.63–0.69; max(acetone)/nm 434sh 456 484sh; HPLC (System 2)
RT = 23.5 (44%, max/nm 430sh 452 483; %III/II 28; 20), RT = 24.2
(53%, max/nm 435 458 485; %III/II 45; 20); m/z (EI) 551 (1.8%,
M + 1), 550 (M+, 5.6), 548 (2.1), 534 (0.3), 458 (0.3, M − 92),
41 (100).
Mixture of unknowns (21%). RF 0.34–0.49; max(acetone)/
nm 447.
4′-Hydroxy-,-caroten-4-one (21, 18%). RF 0.26–0.31;


max(acetone)/nm 455, %III/II 0; HPLC (System 2) RT = 9.7 (64%,
max/nm 462; %III/II 0; 21), RT = 11.2 (32%, max/nm 449, %III/II 0;
Treatment of 4,4′-diacetoxy-,-carotene (12) with CF3COOH
For NMR analysis at −20 °C of dication 6, see ref. 4.
unidentified carotenoid); m/z (EI) 564 (1.9%, M − 2), 549 (8), 548
(20, M − 18), 456 (1.9), 41 (100).
An aliquot of product 21 was reacted with NaBH4,21 resulting in a
hypsochromic shift and increased fine structure. HPLC (System 2)
Reaction of dication 6 (from 12) with water. 4,4′-Diacetoxy-
,-carotene (12, 0.4. mg) was dissolved in CH2Cl2 (5 ml), and
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 4 9 6 – 2 5 0 6
2 5 0 3

View Article Online
Treatment of φ,φ-carotene (26) with Lewis acid
UV/VIS, EIMS and 1H (500 MHz) and 13C (125 MHz) NMR data
for φ,φ-carotene (26) were recently published elsewhere.4
RT = 8.0 (47%, max/nm 424sh 450 475; %III/II 27; ,-carotene-
4,4′-diol, 5), RT = 10.4 (24%, max/nm 419sh 445 469; %III/II 32;
unidentified carotenoid). 5 was identified by co-chromatography
(HPLC) with an authentic sample.
VIS/NIR experiments. φ,φ-Carotene (26, 0.5 mg) was dis-
solved in the solvent (0.5 ml) and BF3-dee (0.25 ml) was added at
room temperature. The mixture was diluted and the UV/VIS/NIR
spectrum recorded. max/nm (CHCl3) 910, 750 (weak) (tꢀ/min 10);
Reaction of dication 7 (from 5) with methanol. ,-Carotene-
4,4′-diol (5, 1.1 mg) was dissolved in CH2Cl2 (10 ml), and cooled
to −15 °C. CF3SO3H (5 l), dissolved in CH2Cl2 (2 ml), was added,
giving a blue reaction mixture. After 10 min stirring, methanol
(1 ml) was added, whereupon the reaction mixture turned yellow.
The pigments were transferred to hexane and washed with 5%
NaHCO3, water and saturated NaCl solution. A slightly brownish
hue was removed upon addition of N-ethyldiisopropylamine
(4 drops). Pigment recovery 71% (E1%, 1 cm 2500), max(hexane)/nm
445; HPLC (System 2) RT 17–42 min. Products were isolated by
preparative TLC developed with 10% acetone in hexane and eluted
with acetone to give:

max/nm (CH2Cl2) 930, 750 (strong) (tꢀ/min 10); max/nm (benzene)
730. For VIS/NIR experiments at −20 °C, see ref. 4.
EPR analysis. φ,φ-Carotene (26) was dissolved in CHCl3. The
same volume of BF3-dee was added, and the solution analysed,
Fig. 2. After 1 h, the signal was reduced by 35%; line width 14 G.
MS analysis. φ,φ-Carotene (26, 1 mg) was dissolved in CHCl3
(0.4 ml) and BF3-dee (0.4 ml) added. An aliquot was evaporated
with N2. Adding a drop of xylene facilitated the evaporation. The
MS was recorded 15 min after the preparation of the reaction mix-
ture. m/z (EI) 528 (M, 7%), 526 (M − 2, 9), 511 (2), 448 (2), 263 (8),
187 (11), 173 (70), 133 (31), 69 (23), 55 (27), 28 (100).
4′,5′-Didehydro-4,5′-retro-,-carotene
(8,
11%).
RF
0.85–0.91; max(acetone)/nm 447 469 496sh; HPLC (System 2)
RT = 36.7 (9%, max/nm 439 461 489; %III/II 36), RT = 40.5 (7%,

max/nm 437sh 457 485; %III/II 19), RT = 41.4 (32%, max/nm 443
465 494; %III/II 24; 8), RT = 42.1 (29%, max/nm 446 469 499;
%III/II 37; 8), RT = 48.2 (7%, max/nm 445sh 467 499; %III/II 21),
RT = 48.9 (7%, max/nm 449sh 473 500; %III/II 20); m/z (EI) 535
(44%, M + 1), 534 (M+, 97), 532 (23), 442 (19, M − 92), 428 (20,
M − 106), 57 (100).
Reaction of φ,φ-carotene (26) cations with H2O at room
temperature. φ,φ-Carotene (26, 1.3 mg) was dissolved in CHCl3
(2 ml) and BF3-dee (0.54 ml) added. The mixture was flushed with
N2, shaken for 2 min and poured into 20% H2O in acetone (30 ml).
The carotenoids were transferred to hexane. HPLC (System 1)
revealed the presence of E/Z isomerised φ,φ-carotene (26) and
two more polar products, max/nm (HPLC solvent) 385sh, 405, 430
(10% of total) and max/nm (HPLC solvent) 408sh, 425, 455 (20% of
total). MS was recorded of the reaction mixture, m/z (EI) 544 (9%),
528 (100), 513 (46), 395 (25), 355 (12), 315 (11), 275 (15), 211 (16),
173 (42), 133 (92). Test for elimination of allylic OH with 0.03 N
HCl in CHCl321 was negative.
4-Methoxy-,-carotene (22, 8%). RF 0.71–0.81; max(acetone)/nm
450 473sh; HPLC (System 2) RT = 26.8 (10%, max/nm 445; %III/II
0; unidentified carotenoid), RT = 27.4 (17%, max/nm 461; %III/II
0; unidentified carotenoid), RT = 28.1 (5%, max/nm 457; %III/II 0;
unidentified carotenoid), RT = 31.0 (43%, max/nm 427sh 450 476;
%III/II 11; 22), RT = 31.6 (6%, max/nm 423sh 446 471; %III/II 16;
9-cis 22), RT = 39.3 (11%, max/nm 450 475sh; unidentified carot-
enoid); m/z (EI) 567 (44%, M + 1), 566 (M+, 100), 565 (33), 564
(64), 535 (25), 534 (49, M − CH3OH), 532 (25), 474 (17, M − 92),
442 (10, M − CH3OH − 92).
Reaction of φ,φ-carotene (26) cations with H2O at low
temperature. In two parallel experiments, φ,φ-carotene (26,
1.0 mg) was dissolved in CDCl3 (0.5 ml) and CH2Cl2 (0.5 ml)
at −35 °C. To each solution was BF3-dee (0.5 ml) added, giving
a colour change to blue via green. After ca. 30 min, the reaction
mixtures were poured into 20% H2O in acetone (10 ml), provid-
ing yellow solutions. The pigments were transferred to hexane
and washed with water. Pigment recovery 35% (CDCl3) and 31%
(CD2Cl2), E1%, 1 cm = 2200, max/nm (hexane) 426. Both mixtures
showed similar chromatographic behaviour by HPLC (System 3),
and were combined prior to preparative TLC. Preparative TLC
plates were developed with 10% acetone in hexane, and the prod-
ucts were eluted with acetone to give:
Co-chromatography (TLC) with authentic 222 gave no
separation.
4′,5′-Didehydro-3-methoxy-4,5′-retro-,-carotene (23, 21%).
RF 0.56–0.65; max(acetone)/nm 448sh 470 500, %II/II 17; HPLC
(System 2) RT = 31.4 (65%, max/nm 443 467 496; %III/II 29; 23),
RT = 32.1 (33%, max/nm 446 471 502; %III/II 43; all-trans 23); m/z
(EI) 565 (41%, M + 1), 564 (M+, 91), 534 (33), 533 (29), 532 (63,
M − CH3OH), 518 (18), 472 (13, M − 92), 458 (13, M − 106), 119
(100).
3′,4′-Didehydro-,-caroten-4-ol (24) and ,-caroten-4-ol (3)
(24%). RF 0.12–0.21; max(acetone)/nm 427sh 451 474sh; HPLC
(System 2) RT = 17.3 (5%, max/nm 448, %III/II 0; unidentified
carotenoid), RT = 18.1 (36%, max/nm 461; %III/II 0; all-trans
24), RT = 19.0 (11%, max/nm 455; %III/II 0; 24), RT = 20.8 (38%,
φ,φ-Carotene (26, 7%). RF = 0.87–0.93; max/nm (acetone)
428sh, 452, 478; %III/II 3.
Product 1 (32 isomers, 4%). RF = 0.70–0.73, 0.77–0.83; max/nm
(acetone) 386sh, 406, 428; %III/II 17; RT (System 2) = 19.4 (65%,
max/nm 383sh, 405, 427; %III/II 37), RT (System 2) = 23.9 (26%,
max/nm 409, 429, 457; %III/II 51); m/z (EI) 545 (13%, M + 1), 544
(M+, 30), 542 (17), 411 (9), 410 (23), 147 (83), 133 (100).
An aliquot of this TLC fraction was tested for allylic hydroxyl
by treatment with 0.03 M HCl in CHCl3,21 giving no observable
change of chromophore in UV/VIS. Likewise, an aliquot was dis-
solved in CF3COOH, giving max/nm (CF3COOH) 687. Subsequent
neutralisation with 10% KOH in methanol recovered the carotenoid
spectrum, max/nm 412.

max/nm 424sh 451 473; %III/II 10; 3); m/z (EI) 553 (7%, M + 1, 3),
552 (M+, 3, 18, M + 2, 24), 551 (13, M + 1, 24), 550 (M+, 24, 26),
536 (25), 535 (34), 534 (79, M − H2O, 3), 533 (30), 532 (65, M −
H2O, 24), 442 (13, M − 92 − H2O, 3), 440 (10, M − 92 − H2O, 24),
91 (100).


Co-chromatography (HPLC, System 2) with authentic 3 gave no
separation at RT = 20.8.
4′,5′-Didehydro-4,5′-retro-,-caroten-3-ol
4-methoxy-,-caroten-4′-ol (25a) (16%). RF 0.06–0.12;
max(acetone)/nm 447sh 469 495sh; HPLC (System 2) RT = 13.2
(13%, max/nm 425sh 449 475; %III/II 22; 25a), RT = 20.7 (46%,
max/nm 446 469 498; %III/II 30; 23b); MS of mixture containing
(23b)
and

Product 2 (32 isomer, 5%). RF = 0.47–0.53; max/nm (acetone)
384sh, 406, 425; %III/II 2; RT (System 2) = 11.8 (23%, max/nm
457), RT (System 2) = 17.6 (43%, max/nm 384sh, 405, 428; %III/II
30), RT (System 2) = 19.3 (13%, max/nm 405sh, 425, 445sh); m/z
(EI) 545 (4%), 544 (12), 542 (24), 410 (9), 133 (100).

23b and 25a/b, m/z (EI) 565 (6%), 564 (11, M; 25b), 551 (6), 550
(6, M; 23b), 549 (6), 548 (7), 534 (9), 533 (13), 532 (24), 523 (6),
440 (4), 368 (7), 313 (8), 72 (100).
2 5 0 4
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 4 9 6 – 2 5 0 6

View Article Online
An aliquot of this TLC fraction was tested for allylic hydroxyl
by treatment with 0.03 M HCl in CHCl3,21 giving no observable
change of chromophore in UV/VIS. Likewise, an aliquot was dis-
solved in CF3COOH, giving max/nm (CF3COOH) 605sh, 651, 722.
Subsequent neutralisation with 10% KOH in methanol failed to give
recovery of carotenoids.
with CH3ONa as described previously.2 The compound of interest,
8-methoxy-7,8-dihydro-,-carotene was isolated with preparative
HPLC. Full characterisation is presented elsewhere.2
8-Methoxy-7,8-dihydro-,-carotene (6%). RT (System 3) =
30.9 min; m/z (EI) 569 (5.2%, M + 1), 568 (M+, 7.7), 538 (10),
536 (17, M − CH3OH), 434 (15), 431 (44), 119 (100). M + 1/
M+ = 68%.
Product 3 (unidentified, 4%). RF = 0.35–0.43; max/nm (acetone)
407; RT (System 2) = 11.8 (64%, max/nm 459), RT (System 2) = 12.8
(9%, max/nm 441); m/z (EI) 546 (1.5%), 544 (0.3), 530 (14), 413
(11), 412 (33), 410 (15), 133 (100).
Treatment of ,-carotene (lycopene, 27) with Lewis acid
VIS/NIR experiments. ,-Carotene (27, 1.0 mg) was dis-
solved in the solvent (1.0 ml) and BF3-dee (0.5 ml) added at room
temperature. The mixture was diluted and the UV/VIS/NIR spec-
trum recorded at room temperature. max/nm (CHCl3) 940 (tꢀ/h 1.5),
7,8-Dihydro-φ,φ-caroten-8-ol (30, 67%). RF = 0.25–0.32;
max/nm (acetone) 408sh, 429, 455; %III/II 37; RT (System 2) = 16.0

(85%, max/nm 405sh, 428, 453; %III/II 50), RT (System 2) = 17.2
(12%, max/nm 403sh, 423, 447; %III/II 59); H(400 MHz, CDCl3,
Me4Si) 1.96–2.00 (19/20/20′-H), 2.09 (19′-H), 2.24 (16′-H),
2.27–2.29 (16/17/17′/18″-H), 2.36 (18-H), 2.89 (dd, 2JH,H = 13.7 Hz,
3JH,H = 3.4 Hz, 7-H), 3.03 (dd, 2JH,H = 14.1 Hz, 3JH,H = 9.6 Hz, 7-H),
4.32 (8-H), 6.23 (10′-H), 6.25 (10-H), 6.29 (14/14′-H), 6.35 (12-H),
6.41 (12′-H), 6.54 (11-H), 6.61 (15/15′-H), 6.69 (11′-H), 6.97–7.00
(3/4/3′/4′-H); m/z (EI) 547 (1.2%, M + 1), 546 (M+, 3.5), 530 (10),
528 (16, M − H2O), 133 (100).

min/nm (CHCl3) 1413; max/nm (CH2Cl2) 920 (tꢀ/h 7.5); max/nm
(benzene) 850 (tꢀ/h 1).
EPR analysis. ,-Carotene (27) was dissolved in CHCl3, the
same volume of BF3-dee was added, and the solution analysed,
Fig. 3, line width 13 G. After 45 min the signal intensity was
reduced by 30%.
MS analysis. ,-Carotene (27, 1.0 mg) was dissolved in CHCl3
(1 ml), and BF3-dee (1 ml) added. An aliquot was evaporated with
N2 and the MS was recorded 15 min after the preparation of the
reaction mixture. m/z (EI) 540 (M + 4, 28%), 538 (M + 2, 72), 536
(M, 91), 534 (M − 2, 52), 532 (M − 4, 23), 521 (13), 493 (7), 347
(10), 267 (19), 233 (17), 145 (35), 105 (33), 91 (39), 28 (100).
7,8-Dihydro-φ,φ-carotene-7,8-diol (31, 13%). RF = 0.03–0.10;


max/nm (acetone) 405, 428, 455; HPLC (System 2) RT = 10.9 (84%,
max/nm 406sh, 428, 454; %III/II 52), RT = 11.9 (13%, max/nm 403,
423, 448; %III/II 47); m/z (EI) 562 (M+, 1.9%), 544 (10, M − H2O),
412 (24), 410 (20), 133 (100).
NMR analysis. The cationic mixture was prepared in CDCl3 as
for the UV/VIS/NIR experiment in a glove box under argon. No
meaningful NMR spectrum was achieved at −20 °C.4
Reproducibility of the low temperature experiments.
Attempts to reproduce the results from the above experiments
in larger scale failed, giving φ,φ-carotene (26) as the main
product, with only minor amounts converted to other products.
An experiment performed at room temperature with commercial
CHCl3 (with stabiliser present) as solvent, gave 30 and 26 as the
main products in a 2:1 ratio. Performed with φ,φ-carotene (26)
from another source, different reactions were observed in duplicate
experiments. The reaction mixtures had max/nm (hexane) 379, and
evaporation of solvent caused a colour change to green. The orange
colour was restored upon the addition of N-ethyldiisopropylamine
(4 drops). Preparative TLC developed with 10% acetone in hexane
and subsequent elution with acetone revealed the presence of
numerous compounds with shorter chromophore (max/nm (acetone)
360–385), covering a wide range of polarities (RF = 0.07–0.80). Due
to the low amounts of carotenoids present, characterisation was only
attempted for two of these products.
Reaction of ,-carotene (27) cations with H2O. ,-Carotene
(27, 3.4 mg) was dissolved in CHCl3 (3 ml), BF3-dee (1 ml) was
added, and the reaction mixture shaken for 2 min before addition of
20% H2O in acetone (40 ml), whereupon the reaction mixture turned
yellow. The products were transferred to hexane and analysed by
UV/VIS, HPLC (System 1) and MS. max/nm (HPLC solvent) 410,
434, 454 (pigment recovery 38%). MS of product mixture: m/z (EI)
568 (1%), 552 (5), 536 (100), 534 (43), 444 (12), 430 (6), 347 (6),
281 (6), 223 (14), 197 (23), 171 (24), 159 (3), 157 (34), 145 (47),
119 (42). HPLC (System 1) showed 7 major peaks with VIS spectra
corresponding to heptaene chromophore (1 peak, 13%), octaene (2
peaks, 20%), nonaene (2 peaks, 13%), decaene (2 peaks, 6%) and
undecaene (2 peaks, 2%). Preparative TLC on silica (5% acetone in
heptane) showed 3 zones.
Product A (31 isomer, 19%). RF = 0.15–0.22; max/nm (acetone)
380; RT (System 2) = 9.4 (52%, max/nm 381, 395sh); m/z (EI) 563
(1.0%, M + 1), 562 (M+, 2.2), 560 (2), 544 (8, M − H2O), 526 (6,
M − 2 H2O), 133 (100).
Fraction A. RF = 0.88–0.94; max/nm (acetone) 436, 456 (74%
of total); HPLC (System 1) comprised 9 peaks with VIS data
compatible with mainly decaene and undecaene chromophores.
Fraction B. RF = 0.73–0.81; max/nm (acetone) 406, 430, 456
(8%); HPLC (System 1) showed 6 peaks with VIS data compatible
with nonaene and decaene chromophores; m/z (EI) 568 (M1, 6%),
554 (M2, 32%), 552 (M1 − 16, M2 − 2, 13), 536 (M2 − 18, 19), 462
(M1 − 106, M2 − 92, 4), 410 (25), 368 (lipid, 11), 267 (11), 239 (15),
197 (23), 185 (26), 173 (28), 157 (37), 145 (50), 95 (51), 81 (53),
69 (68), 43 (100).
Product B (31 isomer, 24%). RF = 0.07–0.14; max/nm (acetone)
380; RT (System 2) = 7.6 (53%, max/nm 382, 396sh), RT = 8.2 (42%,

544 (10, M − H2O), 526 (13, M − 2 H2O), 347 (24, M − C15H19O),
133 (100).
A reaction was also performed in CHCl3 with the ordinary stabi-
liser (ethanol) replaced with CD3OD. The reaction mixture was
worked up as described above. The compound of interest, 7,8-
dihydro-φ,φ-caroten-8-ol (30), was isolated by preparative TLC.
max/nm 381, 396sh); m/z (EI) 563 (2.6%, M + 1), 562 (M+, 5.6),
Fraction C. RF = 0.47–0.56; max/nm (acetone) 450, 464, 500
(17%); HPLC (System 1) exhibited 5 peaks with decaene and
tentatively monocyclic undecaene chromophores; m/z (EI) 568 (M1,
8%), 554 (M2, 58), 552 (M1 − 16, M2 − 2, 20), 537 (33), 536 (M2 −
18, 71), 534 (19), 462 (M1 − 106, M2 − 92, 12), 444 (M2 − 18 − 106,
15), 410 (19), 333 (9), 281 (12), 239 (18), 223 (23), 209 (35), 197
(43), 171 (49), 159 (62), 157 (67), 145 (93), 119 (99), 69 (100).
Fraction C was tested for allylic hydroxyl by treatment with
0.03 N HCl.21 HPLC analysis revealed no products with prolonged
chromophore.
7,8-Dihydro-φ,φ-caroten-8-ol (30, 11%). RF = 0.29–0.39;

max/nm (acetone) 406sh, 428, 453; %III/II 11; m/z (EI) 547 (7.4%,
M + 1), 546 (M+, 14.6), 529 (13), 528 (26, M − H2O), 436 (6), 413
(10), 133 (100). M + 1/M+ = 0.51.
A similar study with CD3OD added in 10-fold excess to the
ethanol stabiliser already present, was undertaken with ,-carotene
(1). The dication formed by the addition of BF3-dee, was quenched
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 4 9 6 – 2 5 0 6
2 5 0 5

View Article Online
9 H. R. Sliwka and S. Liaaen-Jensen, Tetrahedron Asymmetry, 1993, 4,
2377–2382.
10 G. Britton, in Carotenoids Vol 1B: Spectroscopy, ed. G. Britton,
S. Liaaen-Jensen and H. Pfander, Birkhäuser, Basel, 1995, pp. 13–62.
11 J. L. Grant, V. J. Kramer, R. Ding and L. D. Kispert, J. Am. Chem. Soc.,
1988, 110, 2151–2157.
Acknowledgements
We thank T.A. Vestad and E. Malinen, University of Oslo, for
assistance in recording EPR spectra, and Hoffmann-La Roche,
Basel, for a research grant to SLJ and for synthetic carotenoids
(lycopene, echinenone and canthaxanthin).
12 J. A. Haugan and S. Liaaen-Jensen, Acta Chem. Scand., 1994, 48,
68–75.
13 G. Gao, C. C. Wei, A. S. Jeevarajan and L. D. Kispert, J. Phys. Chem.,
1996, 100, 5362–5366.
References
14 L. M. Tolbert and M. E. Ogle, J. Am. Chem. Soc., 1989, 111,
5958–5959.
1 B. F. Lutnaes, L. Bruas, J. Krane and S. Liaaen-Jensen, Tetrahedron
Lett., 2002, 43, 5149–5152.
15 L. M. Tolbert and M. E. Ogle, J. Am. Chem. Soc., 1990, 112,
9519–9527.
16 K. Aareskjold, H. Kayser and S. Liaaen-Jensen, Tetrahedron Lett., 1981,
22, 4541–4544.
2 B. F. Lutnaes, L. Bruas, G. Kildahl-Andersen, J. Krane and S. Liaaen-
Jensen, Org. Biomol. Chem., 2003, 1, 4064–4072.
3 G. Kildahl-Andersen, B. F. Lutnaes, J. Krane and S. Liaaen-Jensen,
Org. Lett., 2003, 5, 2675–2678.
17 W. V. Bush and L. Zechmeister, J. Am. Chem. Soc., 1958, 80,
2991–2999.
4 B. F. Lutnaes, G. Kildahl-Andersen, J. Krane and S. Liaaen-Jensen,
J. Am. Chem. Soc., 2004, 126, 8981–8990.
18 G. Kildahl-Andersen, B. F. Lutnaes and S. Liaaen-Jensen, Org. Biomol.
Chem., 2004, 2, 489–498.
5 E. J. Corey, K. C. Nicolaou, R. D. Balanson and Y. Machida, Synthesis,
1975, 9, 590–591.
19 L. C. Sander, K. E. Sharpless, N. E. Craft and S. A. Wise, Anal. Chem.,
1994, 66, 1667–1674.
6 S. N. Maiti, M. P. Singh and R. G. Micetich, Tetrahedron Lett., 1986,
27, 1423–1424.
20 L. Zechmeister, Cis–trans Isomeric Carotenoids Vitamins A and
Arylpolyenes, Springer Verlag, Wien, 1962.
7 H. R. Sliwka and S. Liaaen-Jensen, Acta Chem. Scand., 1990, 44,
61–65.
21 G. Britton, S. Liaaen-Jensen and H. Pfander, Carotenoids Vol 1A:
Isolation and analysis, Birkhäuser, Basel, 1995.
8 H. R. Sliwka and S. Liaaen-Jensen, Tetrahedron Asymmetry, 1993, 4,
361–368.
2 5 0 6
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 4 9 6 – 2 5 0 6