Longer polyenyl cations in relation to soliton theory

Article abstract of DOI:10.1039/b706212e

The carotene-like polyenes decapreno-β-carotene (C₅₀), C₅₄-β-carotene (C₅₄, first synthesis) and dodecapreno-β-carotene (C₆₀) with 15, 17 and 19 conjugated double bonds, respectively, were synthesized by double Wittig reactions. Introduction of a leaving group in allylic position failed, and cations were obtained by hydride elimination effected by i) triphenylcarbenium tetrafluoroborate-d₁₅, prepared by a new method, or ii) treatment with trifluoroacetic acid-d. Deuterated reagents were employed for product analysis by ¹H NMR. Parallel experiments were performed with β,β-carotene (C₄₀). NIR spectra at room temperature and at -15 °C were employed for characterisation and stability studies of the cationic products. In CH₂Cl₂ λ_max in the 900-1350 nm region was recorded. NMR data for the cationic product of β,β-carotene obtained by the two new preparation methods were consistent with the two monocations previously characterised. The cationic products of the longer polyenes provided downfield-shifted, broadened signals, compatible with C₅₀-monocation, mixed C₅₄-mono- and dication and C₆₀-dication. Combined NIR and NMR data suggest that the extent of charge delocalisation is limited by the maximum soliton width for cations obtained from linear polyenes with more than ca. 20 sp ²-hybridized carbon atoms. The Royal Society of Chemistry.

Full text of DOI:10.1039/b706212e

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Longer polyenyl cations in relation to soliton theory†
Geir Kildahl-Andersen, Thorleif Anthonsen and Synnøve Liaaen-Jensen*
Received 24th April 2007, Accepted 11th July 2007
First published as an Advance Article on the web 24th July 2007
DOI: 10.1039/b706212e
The carotene-like polyenes decapreno-b-carotene (C50), C54-b-carotene (C54, first synthesis) and
dodecapreno-b-carotene (C60) with 15, 17 and 19 conjugated double bonds, respectively, were
synthesized by double Wittig reactions. Introduction of a leaving group in allylic position failed, and
cations were obtained by hydride elimination effected by i) triphenylcarbenium tetrafluoroborate-d15,
prepared by a new method, or ii) treatment with trifluoroacetic acid-d. Deuterated reagents were
employed for product analysis by 1H NMR. Parallel experiments were performed with b,b-carotene
(C40). NIR spectra at room temperature and at −15 ◦C were employed for characterisation and stability
studies of the cationic products. In CH2Cl2 kmax in the 900–1350 nm region was recorded. NMR data for
the cationic product of b,b-carotene obtained by the two new preparation methods were consistent with
the two monocations previously characterised. The cationic products of the longer polyenes provided
downfield-shifted, broadened signals, compatible with C50-monocation, mixed C54-mono- and dication
and C60-dication. Combined NIR and NMR data suggest that the extent of charge delocalisation is
limited by the maximum soliton width for cations obtained from linear polyenes with more than ca. 20
sp2-hybridized carbon atoms.
are superimposed on the characteristics of the solitons contained
Introduction
in the polyacetylene. For accurate description of the physical
properties of the soliton itself, its availability in a pure compound is
desired. Experimental information about bond lengths and charge
distributions is of particular interest.
The theoretical fundament of charge distribution in very long
polyenes, such as doped polyacetylene, was provided by Heeger
and co-workers, leading to the Su–Schrieffer–Heeger (SSH)
model.1,2 According to this model, charges are carried in holes
or defects in the polyene structure, called solitons. They consist
of regions of intermediate bond lengths, in contrast to the strict
alternating nature of single and double bonds surrounding them.
The charge inside a soliton can mathematically be described as a
charge density wave, in the case of a delocalised cation oscillating
between finite positive values and zero on alternating carbon
atoms, with the highest values in the center of the soliton. In
this respect, the model is in agreement with the more familiar
concept of resonance structures for smaller delocalised charged
species. However, whereas the drawing of resonance structures
imposes no limitations as to the extent of delocalisation in the
molecule, the SSH-model implies a limiting value for how far a
polyenyl cation or anion can delocalise, described by the soliton
half-width, l. This remains one of the more controversial aspects
of the soliton theory.3 Physically, the parameter l describes the
width of the region containing half of the charge of the soliton, in
terms of the number of bonds.
Carotenoids, such as b,b-carotene (1, Scheme 1), are a con-
venient source of long polyenes. Charged carotenoids, including
radical cations, monocations and dications, have been studied
experimentally using the techniques mentioned above.8 Moreover,
it is possible through the study of smaller molecules to prepare
spinless charge-delocalised polyenes in a controlled manner, and
thereby study the charged polyene by 1H and 13C NMR. The 13
C
chemical shift is of particular importance in this respect, because
Experimental studies of soliton properties have been carried
out mostly with polyacetylene, both doped and undoped, em-
ploying techniques such as resonance Raman spectroscopy, cyclic
voltammetry, NIR spectroscopy, EPR and ENDOR.2,4–7 Using
doped polyacetylene, the properties of the bulk of the polyene
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 73594099
† Electronic supplementary information (ESI) available: 1H NMR (olefinic
region) of cations from b,b-carotene (1), decapreno-b-carotene (4) and
dodecapreno-b-carotene (6). See DOI: 10.1039/b706212e
Scheme 1
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2803–2811 | 2803
©

View Article Online
of the empirical correlation of the shift with atomic charge, as
first shown by Spiesecke and Schneider for aromatic compounds,9
and later extended to other systems with sp2-hybridised carbon
atoms.10 NMR studies of a,x-diphenylpolyene anions of various
length have been reported, employing these empirical correlations
to give a description of the charge distribution.3,11,12
We have previously reported on the formation of delocalised
carotenoid monocations formed by acid-induced dehydration of
the allylic alcohol b,b-caroten-4-ol (2, isocryptoxanthin), creating
the longest delocalised carbocation 3a,b fully characterised by
NMR to date (Scheme 1).13,14 By comparison of the 13C chemical
shifts of the resulting monocations 3a,b with model compounds,
we were able to provide an estimate for the charge distribution, and
thereby also the extent of delocalisation, as described by the soliton
half-width l, yielding l = 7.8. In the present work we aimed at
the preparation of even longer delocalised polyenyl carbocations,
and by NIR and NMR studies to gain insight on the extent of
delocalisation in these systems.
Scheme 3
diethyl etherate, followed by quenching with water,19,20 gave the
desired mono-ol product 2, albeit in a poor yield in experiments
with b,b-carotene (1). When applied to the longer C54 polyene 5,
it became apparent that the resulting allylic alcohol was rather
labile, and another strategy had to be considered.
We have previously noted the ability of carotenoids to eliminate
hydride in allylic positions under acidic conditions.13,21 This was
first observed by the formation of blue oxonium ions from
a 5,8-furanoid derivative neochrome (13) (Scheme 4, i),21 and
later during NMR studies of b,b-caroten-4-one (14, echinenone,
Scheme 1) treated with dilute trifluoroacetic acid-d (Scheme 4,
ii).13 It is known that hydride abstractions can be achieved by
treatment with triphenylcarbenium ions.22 Triphenylcarbenium
tetrafluoroborate-d15 (15) was therefore prepared as an NMR-
compatible hydride abstraction reagent. The initial triple Grig-
nard reaction from bromobenzene-d5 (16) and diethyl carbonate
offered a convenient pathway to triphenylmethanol-d15 (17), and
represented an alternative to a published method.23 Treatment of
17 with tetrafluoroboric acid in propionic anhydride provided the
carbenium ion 15 (Scheme 5).22
Results
Preparation of substrates
The elongated, carotene-like compounds referred to as decapreno-
b-carotene (4, C50), C54-b-carotene (5) and dodecapreno-b-
carotene (6, C60) were prepared as substrates. Decapreno-b-
carotene (4) has previously been synthesized via a 15,15ꢀ-didehydro
derivative,15 and dodecapreno-b-carotene (6) by a C15 + C30 + C15
approach.16 In the present work the C50- (4), C54- (5) and C60- (6)
polyenes were prepared by an analogous scheme by Wittig cou-
pling reaction between b-ionylidenethyltriphenylphosphonium
bromide (7), synthesised from b-ionone (8) via vinyl-b-ionol
(9), and the appropriate diapocarotenals 10–12, using NaH for
generation of the ylide17 (Scheme 2). This approach yielded the
desired target molecules in satisfactory quantities.
Various strategies were employed for the introduction of a
leaving group in allylic position of the parent polyene (Scheme 3).
Addition of NBS in the presence of excess acetic acid to b,b-
carotene (1) gave the diacetate, as reported by others,18 whereas
attempts to obtain the monoacetate failed. Reaction with BF3–
Preparation of polyenyl cations
The polyenyl cations were prepared from the parent polyenes
1, 4, 5 and 6 using two main strategies, i) either by employing
triphenylcarbenium tetrafluoroborate, or ii) by treatment with
trifluoroacetic acid. Both reagents were used in dilute solution in
Scheme 2
2804 | Org. Biomol. Chem., 2007, 5, 2803–2811
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Fig. 1 Initial absorptions at room temperature of cationic products
from b,b-carotene (1), decapreno-b-carotene (4), C54-b-carotene (5) and
dodecapreno-b-carotene (6) in dichloromethane containing 4 mg mL−1
Ph3CBF4.
Scheme 4
Table 1 Absoption maxima and relative stability of cationic products
formed in 4–6 mg mL−1 solution of triphenylcarbenium tetrafluoroborate
in CH2Cl2
Room temperature
−15 ◦
C
Starting polyene
kmax/nm, 0 h kmax/nm, 1 h Rel. e, 1 h kmax/nm a
b,b-Carotene (1)
1011
966
1031
1140
1241
0.88
0.40
0.57
0.60
1033
1130
1215
1290
C50-b-Carotene (4) 1123
C54-b-Carotene (5) 1180
C60-b-Carotene (6) 1229b
a Constant for 1 h. b After 5 min kmax 1260 nm was measured.
ii) With trifluoroacetic acid. VIS/NIR measurements of the
absorption spectra for the polyenyl cations from the polyenes 1 and
4–6 were performed in an analogous manner to that described
above. In addition to varying the temperature from −15 ◦C to
room temperature, two different acid concentrations were used.
Data for the lower concentration, 0.013 M, are given in Table 2,
and for the higher concentration, 1.3 M, in Table 3.
The influence of the acid concentration is obvious; a higher
concentration of acid gives in general a hypsochromic shift
of the kmax of the resulting polyenyl cation. Compared to the
trityl experiments above, the stability at room temperature is
considerably higher when trifluoroacetic acid is used, although
higher acid concentrations reduce the stability.
Scheme 5
chlorinated solvents. For NMR experiments, deuterated analogs
were employed.
VIS/NIR spectra
i) With triphenylcarbenium tetrafluoroborate. The VIS/NIR
absorption spectra were recorded for b,b-carotene (1) and
the polyenes 4–6 both at room temperature and at −15 ◦C
in a dichloromethane solution containing triphenylcarbenium
tetrafluoroborate. Large bathochromic shifts, characteristic for
the formation of positively charged polyenes, were observed at
both temperatures, illustrated by the initial room temperature
absorptions depicted in Fig. 1.
At room temperature, the stability of the cationic products from
the polyenes 4–6 was considerably lower than for the products
from b,b-carotene (1). The position of the absorption maxima
also drifted during the experiments. When the temperature was
lowered to −15 ◦C, constant absorptions were observed. Data for
stability and kmax are given in Table 1.
Table 2 Absoption maxima and relative stability of cationic products
formed in 0.013 M solution of trifluoroacetic acid in CH2Cl2
Room temperature
−15 ◦
C
Starting polyene
kmax/nm, 0 h kmax/nm, 1 h Rel. e, 1 h kmax/nm a
b,b-Carotene (1)
1022
1010
1199
1263
1307
2.00
0.99
0.96
0.99
980
1153
1238
1300
C50-b-Carotene (4) 1207
C54-b-Carotene (5) 1277
C60-b-Carotene (6) 1330
a All absorptions unstable, with kmax drifting significantly. Initial values are
given.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2803–2811 | 2805
©

View Article Online
Table 3 Absoption maxima and relative stability of cationic products
Table 4 Chemical shifts of C40-monocations 3a, prepared from b,b-
carotene (1) in CDCl3 at −15 ◦C by treatment with (a) deuterated trityl
reagent 15 and (b) CF3COOD
formed in 1.3 M solution of trifluoroacetic acid in CH2Cl2
Room temperature
−15 ◦
C
dH/ppm
Carotenoid numberinga (a)
dC/ppm
DdC/ppm
(a) − (b)
Starting polyene
kmax/nm, 0 h kmax/nm, 1 h Rel. e, 1 h kmax/nm a
(b)
(a)
(b)
b,b-Carotene (1)
997
967b
1076
1154
1222
0.49
0.80
0.88
0.88
853, 922
1123
1216
1308
C50-b-Carotene (4) 1099
C54-b-Carotene (5) 1152
C60-b-Carotene (6) 1236
1
2
3
4
5
35.9
39.7
23.2
36.0 −0.1
1.58
2.24
6.18
1.59
2.25
6.21
39.7
23.2
0
0
137.3 137.8 −0.5
a Unstable absorptions with significant drift. b Another peak at kmax 798 nm
apparent after 1 h.
135.5 135.1 0.4
6
7
8
159.9 160.1 −0.2
121.4 121.7 −0.3
6.55
7.63
6.58
7.57
147.5 147.3
0.2
9
138.5 138.7 −0.2
NMR spectra
10
11
12
13
14
15
16
17
18
19
20
1ꢀ
7.51
6.92
7.71
7.29
6.97
7.29
160.6 158.9 1.7
126.5 126.6 −0.1
Several polyenes were treated with the deuterated trityl reagent
15, namely b,b-carotene (1), decapreno-b-carotene (4), echinenone
(14, Scheme 1) and neochrome (13, Scheme 4) for NMR char-
acterisation. All polyene/trityl cation mixtures yielded blue or
green solutions, compatible with the formation of delocalised
cations.8 However, except for b,b-carotene (1), no detailed inter-
162.3 159.5
140.5 140.4
164.6 161.4
132.0 131.8
2.8
0.1
3.2
0.2
0
0
0.1
0
7.95
6.98
1.38
1.38
1.97
2.06
2.12
7.41
6.97
1.36
1.36
1.97
2.08
2.14
29.1
29.1
21.5
11.8
12.4
34.1
39.7
18.9
33.6
29.1
29.1
21.4
11.8
1
pretation of the H spectra was possible. For b,b-carotene (1),
12.5 −0.1
the expected monocationic product 3 was observed, as part of a
mixture with other unidentified polyene compounds. The chemical
shifts deviated significantly from those earlier reported for 3a,b
(Scheme 1),14 which may be ascribed to different counterions, i.e.
tetrafluoroborate versus trifluoroacetate.
b,b-Carotene (1) treated with trifluoroacetic acid-d in CDCl3 at
−15 ◦C also gave the monocation 3, with only minor differences
in 1H and 13C chemical shifts from values reported earlier in
CD2Cl2.14 NMR chemical shift data of the main isomer 3a are
given for the two new preparation methods in Table 4, as well
as a chemical shift comparison of their respective 13C chemical
shifts.
The NMR spectra of C50-b-carotene (4), C54-b-carotene (5)
and C60-b-carotene (6) treated with trifluoroacetic acid-d yielded
downfield-shifted, but severely broadened signals. No accurate
assignments of the olefinic parts could be made for either cationic
product. However, the mean chemical shifts of olefinic protons
were calculated from the broadened signals using Simpson’s
formula,24 giving average chemical shifts of 6.88 ppm and 7.11 ppm
for the cations formed from 4 and 6, respectively. For the cations
formed from 5, an initial product with an average chemical
shift of 6.80 ppm is formed, which later shifts further downfield
to 7.13 ppm. These numerical estimates of the average olefinic
chemical shift compared to the average chemical shift of the
neutral polyenes gives a total change of chemical shift of 10.0 ppm
and 18.5 ppm for cations prepared from 4 and 6, respectively, and
10.3 ppm and 18.2 ppm for the two different forms from 5. This
was calculated by multiplying the difference in average shift by
20 (for 4), 24 (for 5) and 26 (for 6), which equals the number of
sp2-hybridised protons for the three polyenes.
34.1
39.5
18.6
0
0.2
0.3
2ꢀ
1.48
1.62
2.10
1.49
1.62
2.13
3ꢀ
4ꢀ
34.0 −0.4
5ꢀ
135.7 138.7 −3.0
137.7 138.0 −0.3
135.1 137.4 −2.3
6ꢀ
7ꢀ
6.71
6.38
6.88
6.47
8ꢀ
136.7 136.6
0.1
9ꢀ
150.9 153.5 −2.6
132.3 133.1 −0.8
140.6 142.6 −2.0
10ꢀ
11ꢀ
12ꢀ
13ꢀ
14ꢀ
15ꢀ
16ꢀ
17ꢀ
18ꢀ
19ꢀ
20ꢀ
6.45
7.55
6.67
6.52
7.67
6.71
137.2 137.3 −0.1
b
—
163.1
—
6.78
8.08
1.08
1.08
1.81
2.19
2.38
6.79
7.76
1.10
1.10
1.84
2.22
2.35
135.9 136.3 −0.4
156.0 153.9 2.1
28.8
28.8
22.1
13.5
14.4
28.9 −0.1
28.9 −0.1
22.1
13.5
14.0
0
0
0.4
a See Scheme 1. b Not observed due to missing long-range correlation
from 20ꢀ-CH3 caused by broadening in the 1H dimension. A dC value
of 163.1 ppm was used for 13ꢀ-C in subsequent calculations.
and isocarotene (18, Scheme 1) were formed, demonstrating an
alternative pathway for the introduction of new functional groups
in an allylic postition to the polyene chain (Scheme 6). Both
products can mechanistically be explained by initial removal of a
hydride by the trityl reagent, and subsequent nucleophilic addition
or deprotonation. Exchange of water for aqueous potassium
acetate yielded a mixture of the mono-ol 2 and its corresponding
acetate, in addition to small amounts of isocarotene (18).
Quenching experiments
Addition of water as a nucleophile to b,b-carotene (1) treated
with triphenylcarbenium tetrafluoroborate yielded products com-
patible with formation via cation intermediates. Judging by
UV/VIS spectra and HPLC-DAD analysis, b,b-caroten-4-ol (2)
Scheme 6
2806 | Org. Biomol. Chem., 2007, 5, 2803–2811
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
1
in account, whereas the Spiesecke–Schneider relationship for H
Discussion
estimates 10.6 ppm per charge.9 The value of 10.0 ppm for the
cationic product of the C50-polyene (4) and 10.3 ppm for the initial
product of the C54-polyene (5) indicates formation of monocations.
The higher values found for the second product of 5 and from
the C60-polyene (6), 18.2 ppm and 18.5 ppm, respectively, point
towards dication formation. Additional information on the charge
distribution of the longer polyenyl cations is not provided by
this crude numerical treatment. However, according to previous
experience it would be expected that the charge was located at
the centre of the polyene for monocations, whereas electrostatic
repulsion would force the two solitons in a dication towards the
ends of the polyene chain.13,14,20
The data for the monocation 3a prepared from b,b-carotene
(1), as given in Table 4, show only small differences from earlier
published results.14 Most notably, by employing the deuterated
trityl cation 15, some characteristics of the soliton shape are
altered. The centre has shifted slightly, cf. the values of DdC in
Table 4 for C-10,12,14,15ꢀ compared to C-5ꢀ,7ꢀ,9ꢀ,11ꢀ. The soliton
half-width, determined empirically as given previously,13 is also
10% smaller (calculation not included here). This shows that in
addition to the averaging effects discussed above, the picture is
Hydride abstraction
Due to the steric requirements of the triphenylcarbenium ion,
the competition between hydride abstraction and addition to
the polyene chain should favour the former reaction, which
is in agreement with the NMR data and the quenching trials
for b,b-carotene (1). However, the preference for elimination of
hydride over protonation of b,b-carotene (1) treated with dilute
trifluoroacetic acid (0.01–0.2 M) was surprising. In concentrated
trifluoroacetic acid-d, only complicated mixtures were previously
observed in the NMR spectra.25 In dilute solutions a mixture of
C-7 and C-5 protonated products was expected, as suggested by
previous AM1 studies.26
NMR data
From earlier experience, we expected the polyenyl cations from
the extended polyenes 4–6 to yield readily interpretable NMR
spectra, in which downfield shifts could be assigned to specific
positions in the chain.13,20 It turned out that for the longer
symmetrical polyenes 4–6, treatment with trifluoroacetic acid-d
yielded broad 1H NMR spectra with few distinct spectral features
in the olefinic region. The broadening could be caused by various
reasons, including aggregation, precipitation, free radicals and
dynamic aspects. Care was taken to ensure non-viscous solutions
at lowered temperatures, and the samples were visually inspected
for precipitates. Although aggregation can not be ruled out, the
broadening most likely arises from other phenomena. It has been
noted during work with long polyene anions that the longest
polyene anions gave broadened spectra, and EPR analysis revealed
the formation of radicals, although no radical initiators were
present.11 Radical formation has also been observed by treatment
of concentrated carotenoid solutions with trifluoroacetic acid.26
Radical species cannot be definitely ruled out in our experiments.
If the longer charge-delocalised cations contain free solitons,
the charge density wave may travel along the polyene chain with
only little cost of energy,2,6,7 a dynamic process which hypo-
thetically may give rise to signal broadening in the NMR spectra.
Consequently, NMR may not be a suitable method for the study
of free solitons, as the continuous rapid translational motion of
the delocalised charge would yield averaged signals only, and
would not provide information about the charge distribution of
the soliton itself. This phenomenon seems to have been overlooked
in the past, where the preparation of longer, charged polyenes
was proposed as a solution to the soliton half-width problem,27
and followed up here. The broadened appearance of the polyenyl
cation from the longer C50-, C54- and C60-polyenes (4, 5 and 6)
may suggest signal averaging, whereas the sharper signals of the
C40 monocation 3 is compatible with restricted movement of the
soliton.
further complicated by the influence of different counterions, here
CF3COO− and BF4
.
−
NIR data
The free-electron model, in which electrons are allowed to move
freely along the polyene chain at a constant potential energy, states
a linear correlation between absorption maxima and polyene chain
length.28 Work on shorter polyenes revealed a good linear fit to
data for polyenyl monocations in the range of 3–13 sp2-carbons
involved in the delocalisation, with the absorption maxima varying
from 300 nm to 700 nm (Fig. 2).29 In the upper end of the
spectrum, measurement of the so-called midgap absorption in
doped polyacetylenes has been determined to be 0.65–0.75 eV, or
1650–1900 nm when converted to wavelength units.4,5 The NIR
absorption of charged polyenes is therefore expected to converge
1
Although broadened, the H NMR data of the polyenes 4–6
treated with trifluoroacetic acid-d in CDCl3 and CD2Cl2 show
considerable downfield shifts in the olefinic region, compared to
the parent polyenes. The calculated values for total downfield shift
correspond nicely with values found for other polyenyl cations of
carotenoid origin.13,14 For instance, in the monocation 3 prepared
from b,b-caroten-4-ol (2), a value of 9.7 ppm per charge is obtained
when only the protons originally in the polyene chain are taken
Fig. 2 Absorption maxima kmax at room temperature as function of the
number of conjugated sp2 carbon atoms in polyenyl cations. ᭹: Data from
monocations in ref. 29, measured in 80–96% sulfuric acid. ᭺: Present data,
polyenes 1, 4–6 dissolved in 0.013 M trifluoroacetic acid in CH2Cl2.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2803–2811 | 2807
©

View Article Online
in this range. Absorption maxima for radical cations of the C50-
and C60-polyenes 4 and 6 are reported at 1250 nm and 1480 nm,
respectively.30 For shorter polyenes, similar values are expected for
their spinless counterparts.26 However, lower values are measured
in our experiments.
For polyenes with longer polyene chains, C50-b-carotene (4),
C54-b-carotene (5) and C60-b-carotene (6), NMR spectra after
trifluoroacetic acid-d treatment revealed broadened peaks. Charge
distributions of these longer cations were therefore not deter-
mined. The number of delocalised charges in the polyenes could
be estimated by numerical methods. The two longest polyenes, 5
(C54) and 6 (C60), were susceptible to dication formation.
NIR data recorded for the cationic products of the longer
polyenes showed deviations from the linearity expected from the
free-electron model for delocalised elctrons.28,29 The lower than
expected bathochromic shifts for the linear polyenes with 20 or
more sp2-hybridized carbon atoms indicate the transition from
charge delocalisation limited by chain length to delocalisation
limited by the maximum soliton width.
The ease with which dications are formed, as demonstrated by
NMR experiments for 5 (C54) and 6 (C60), add uncertainty to
the interpretation of the NIR results. Formation of dications is
expected to shift the absorption to lower wavelengths relative to
monocations.31 With the exception of the monocation 3 from b,b-
carotene (1), the relatively low values in Table 1 point towards
dication formation upon treatment with triphenylcarbenium
tetrafluoroborate. When trifluoroacetic acid is employed (Tables 2
and 3), the effect of the concentration is compatible with dication
formation at the highest acid concentration. Shortening of the
polyene chain by in-chain protonations can also be envisaged.
However, the values in Tables 1 and 3 are in fair agreement,
disfavouring in-chain protonation.
Experimental
Materials
The assumption is made that the absorptions measured in dilute
trifluoroacetic acid in Table 2 correspond to monocation forma-
tion. Compared to the linear fit for the shorter monocations,29 all
our measured values are markedly lower, as shown in Fig. 2, for
the room temperature absorptions. This result indicates deviations
from the free-electron model, and may instead be taken as evidence
for the emergence of free solitons in these longer polyenes.
Synthetic samples of b,b-carotene (1), echinenone (14), 2,7-
dimethyl-2,4,6-octatriene-1,8-dial (10), 6,6ꢀ-diapo-w,w-carotene-
dial (11) and 4,4ꢀ-diapo-w,w-carotenedial (12) were obtained
from Hoffmann–La Roche. Neoxanthin, extracted from Spinacea
oleracea, was converted to neochrome (13) according to literature
procedures.33 b-Ionone (8, pract. 90%) was obained from Fluka.
Vinyl chloride was supplied by Matheson Gas Products. All
other reagents were available in analytical grade from commer-
cial chemical suppliers and used without further purification.
Dichloromethane, diethyl ether and tetrahydrofuran were dried
and distilled before use. The ethanol stabiliser present in chloro-
form was removed prior to quenching trials by running the solvent
through an alumina column.
Delocalisation limits
The extent of delocalisation can be described through the soliton
half-width l, and the physical implications of l on bond lengths
and distribution of charge are thoroughly discussed elsewhere.27
In the absence of bond length and charge distribution data,
a more qualitative approach has been employed by several
workers,16,32 using the absolute number of double bonds in which
the charge resides. Due to the shape of the charge distribution
as a hyperbolical cosine function,1 these numbers are not easily
correlated. The latter approach will now be considered here.
We interpret the deviation of linearity in Fig. 2 as an indication
of at which stage the primary assumption of the free-electron
model, namely delocalisation over the entire p-electron system,
is no longer valid. This occurs at approximately 20 sp2 carbon
atoms, which corresponds to 10 conjugated double bonds. It is
interesting to note that the electronic properties of long dicationic
polyenes examined by cyclic voltammetry has shown that a
limiting oxidation potential is reached for conjugation lengths
of 25–30 double bonds,16 implying that two positively charged
solitons in the same polyenyl dication can reach a state of minimal
energy with 12.5–15 double bonds per charge. These results are
in fair agreement with our findings. For polyenes, a charge carrier
length of more than 20 double bonds, which has been suggested,32
is not supported by our data.
Methods
All polyenes were stored in a freezer (−20 ◦C) under nitrogen
atmosphere. Reactions and manipulations were carried out as
far as possible in darkness and under nitrogen atmosphere.
Ultraviolet (UV) and visible light (VIS) spectra were recorded
on a Varian Cary 50 UV-VIS spectrophotometer or a Shimadzu
UV160-A spectrophotometer. Visible light (VIS) and near infrared
(NIR) spectra of charged compounds, with cooling as described
previously for the characterisation of charged species, was per-
formed on a Varian Cary 5 UV–VIS–NIR spectrophotometer.20
For quantitative estimates, E1%, 1 cm = 2500 at kmax was employed
throughout. EI mass spectra were recorded on a Finnigan MAT
95XL ThermoQuest spectrometer. NMR spectra of intermediates
were obtained on a Bruker Avance DPX 400 MHz instrument
using a 5 mm 13C/1H dual probe, whereas data for the polyenes was
recorded on a Bruker Avance 600 MHz instrument, using a 5 mm
cryoprobe (TCI). Spectra of charged species at temperatures below
◦
0 C were obtained on a Bruker Avance DRX 500 MHz instru-
ment, using a 5 mm inverse probe (TXI), equipped with a nitrogen
dewar for cooling. Multidimensional pulse sequences were used
for complete assignment, i.e. COSY, ROESY, HMSC,34 HSQC,
HSQC-TOCSY. Chemical shifts are cited relative to TMS with
calibration against residual CHDCl2 at 5.32 ppm and 53.8 ppm
for 1H and 13C respectively for samples with dichloromethane-d2,
or against residual chloroform at 7.27 ppm for 1H in chloroform-d
Conclusions
Elimination of hydride ion from b,b-carotene (1) was demon-
strated upon treatment with both trifluoroacetic acid and triph-
enylcarbenium tetrafluoroborate. Characterisation of the resulting
carotenyl monocation 3 by NMR was supported by product
analyses after quenching of the cation with nucleophiles.
2808 | Org. Biomol. Chem., 2007, 5, 2803–2811
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
and 77.0 ppm for 13C. HPLC was carried out on a Hewlett
Packard Series 1050 instrument equipped with a diode array
detector (DAD). Detection wavelength was set according to kmax of
the crude products recorded by VIS-spectroscopy. For analytical
separations, a 250 × 4.6 mm Interchrom Uptisphere 5 ODB
column was employed. Mobile phase: 0 min: methanol–acetone
(90 : 10 v/v, 1.0 mL min−1), 60 min: methanol–acetone (30 : 70
v/v, 1.0 mL min−1). A similar column with 10 mm inner diameter
was used for preparative separations.
(2ꢀ/2ꢀꢀ-CH3), 28.8 (6ꢀ/6ꢀꢀ-CH3), 32.9 (3ꢀ/3ꢀꢀ-CH2), 34.1 (6ꢀ/6ꢀꢀ-
C), 39.4 (5ꢀ/5ꢀꢀ-CH2), 124.8 (5/22-CH), 125.0 (9/18-CH), 126.4
(1/26-CH), 129.2 (2ꢀ/2ꢀꢀ-C), 130.0 (13/14-CH), 130.7 (4/23-CH),
132.3 (8/19-CH), 132.8 (12/15-CH), 135.8 (3/24-C), 136.3 (7/20-
C), 136.5 (11/16-C), 137.2 (6/21-CH), 137.6 (2/25-CH), 137.7
(10/17-CH/1ꢀ/1ꢀꢀ-C); m/z (EI) 670 (21%, M + 2), 669 (24, M +
1), 668 (M+, 46), 666 (11, M − H2), 563 (10), 562 (22, M −
106), 510 (9), 91 (100); cf. properties reported in early work,15
and compatible with reported VIS, 1H NMR and MS data.16
3,9,13,18,22,28-Hexamethyl-1,30-bis(2,6,6-trimethylcyclohex-
1-enyl)triaconta-1,3,5,7,9,11,13,15,17,19,21,23,25,27,29-pentade-
caene (C54-b-carotene, 5). Yield 66%, spectrophotometrically
determined. kmax/nm (hexane) 487, 513, 548; %III/II 29; RT = 50.6
(99%, kmax/nm 493sh, 521 553); dH(600 MHz, CDCl3) 1.04 (6ꢀ/6ꢀꢀ-
CH3), 1.48 (5ꢀ/5ꢀꢀ-CH2), 1.63 (4ꢀ/4ꢀꢀ-CH2), 1.73 (2ꢀ/2ꢀꢀ-CH3), 1.96
(13/18-CH3), 1.98 (9/22-CH3), 1.99 (3/28-CH3), 2.03 (3ꢀ/3ꢀꢀ-CH2),
6.15 (2/29-CH), 6.16 (10/21-CH), 6.18 (1/30-CH), 6.21 (4/27-
CH), 6.28 (14/17-CH), 6.36 (8/12/19/23-CH), 6.40 (6/25-CH),
6.47 (7/24-CH), 6.64 (15/16-CH), 6.65 (5/26-CH), 6.66 (11/20-
CH); dC(150 MHz, CDCl3) 12.5 (13/18-CH3), 12.6 (3/9/22/28-
CH3), 19.2 (4ꢀ/4ꢀꢀ-CH2), 21.7 (2ꢀ/2ꢀꢀ-CH3), 28.9 (6ꢀ/6ꢀꢀ-CH3), 33.0
(3ꢀ/3ꢀꢀ-CH2), 34.1 (6ꢀ/6ꢀꢀ-C), 39.5 (5ꢀ/5ꢀꢀ-CH2), 124.9 (11/20-CH),
126.7 (1/30-CH), 129.2 (7/24-CH/2ꢀ/2ꢀꢀ-C), 129.7 (15/16-CH),
130.0 (5/26-CH), 130.7 (10/21-CH), 131.9 (4/27-CH), 132.8
(14/17-CH), 133.9 (6/25-CH), 134.5 (9/22-C), 135.9 (3/28-C),
136.3 (13/18-C), 137.2 (8/12/19/23-CH), 137.6 (2/29-CH), 137.7
(1ꢀ/1ꢀꢀ-C); m/z (EI) 722 (19%, M + 2), 721 (29, M + 1), 720 (M+,
51), 628 (16, M − 92), 91 (100).
Vinyl-b-ionol (9)
Alcohol 9 was prepared in a Grignard reaction,35 using gaseous
vinyl chloride bubbled through a tetrahydrofuran suspension of
magnesium (6.5 g, 0.27 mol, activated with iodine and ethyl
bromide) for preparation of the Grignard reagent. Addition of
b-ionone (8, 30.1 g, 0.156 mol) gave after work-up 33.1 g crude
product, 96% yield. dH(400 MHz, CDCl3) 0.99 (s, 6H), 1.43 (s,
3H), 1.45 (m, 2H), 1.61 (m, 2H), 1.67 (d, 3H, J = 0.9 Hz), 1.98
(m, 2H), 5.09 (dd, 1H, J = 1.2 Hz, J = 10.6 Hz), 5.28 (dd, 1H,
J = 1.2 Hz, J = 17.3 Hz), 5.55 (d, 1H, 16.2 Hz), 6.02 (dd, 1H, J =
10.6 Hz, J = 17.3 Hz), 6.09 (m, 1H); m/z (EI) 221 (5.5%, M + 1),
220 (M+, 12.6), 205 (43), 202 (23), 193 (32), 189 (31), 187 (35), 177
(100), 175 (34), 161 (39), 159 (32), 149 (32), 147 (41), 145 (29), 137
(67), 121 (75).
b-Ionylidenethyltriphenylphosphonium bromide (7)
Vinyl-b-ionol (9, 10.0 g, 45.5 mmol) and triphenylphosponium
hydrobromide (15.5 g, 45.2 mmol) were mixed in tetrahydrofuran
solution.36 The product salt failed to crystallize, and was used as
a crude oil in subsequent reactions.
3,7,11,15,20,24,28,32-Octamethyl-1,34-bis(2,6,6-trimethylcyclo-
hex-1-enyl)tetratriaconta-1,3,5,7,9,11,13,15,17,19,21,23,25,27,29,
31,33-heptadecaene (dodecapreno-b-carotene, 6). Yield 67%,
spectrophotometrically determined. kmax/nm (hexane) 503sh,
530, 564; %III/II 14; RT = 56.9 (42%, kmax/nm 504sh, 529,
561sh), RT = 57.9 (58%, kmax/nm 509sh, 536, 571); dH(600 MHz,
CDCl3) 1.04 (6ꢀ/6ꢀꢀ-CH3), 1.48 (5ꢀ/5ꢀꢀ-CH2), 1.63 (4ꢀ/4ꢀꢀ-CH2),
1.73 (2ꢀ/2ꢀꢀ-CH3), 1.99 (3/32-CH3), 2.00 (7/11/24/28-CH3),
2.01 (15/20-CH3), 2.03 (3ꢀ/3ꢀꢀ-CH2), 6.16 (1/2/33/34-CH), 6.17
(4/31-CH), 6.25 (8/27-CH), 6.27 (12/23-CH), 6.30 (16/19-CH),
6.39 (6/29-CH), 6.41 (10/14/21/25-CH), 6.65 (5/9/26/30-CH),
6.66 (17/18-CH), 6.67 (13/22-CH); dC(150 MHz, CDCl3) 12.6
(3/32-CH3), 12.7 (7/11/15/20/24/28-CH3), 19.1 (4ꢀ/4ꢀꢀ-CH2),
21.7 (2ꢀ/2ꢀꢀ-CH3), 28.8 (6ꢀ/6ꢀꢀ-CH3), 32.9 (3ꢀ/3ꢀꢀ-CH2), 34.1 (6ꢀ/6ꢀꢀ-
C), 39.4 (5ꢀ/5ꢀꢀ-CH2), 124.6 (5/30-CH), 124.7 (9/26-CH), 125.0
(13/22-CH), 126.3 (1/34-CH), 129.2 (2ꢀ/2ꢀꢀ-C), 130.1 (17/18-
CH), 130.7 (4/31-CH), 132.4 (8/27-CH), 132.9 (12/23-CH),
133.0 (16/19-CH), 135.8 (3/32-C), 136.5 (7/11/24/28-C), 136.7
(15/20-C), 137.1 (6/29-CH), 137.6 (2/33-CH), 137.7 (1ꢀ/1ꢀꢀ-C),
137.8 (10/25-CH), 137.9 (14/21-CH); m/z (EI) 802 (10%, M +
2), 801 (22, M + 1), 800 (M+, 35), 695 (12), 694 (21, M-106), 386
(16), 119 (100); compatible with reported VIS, 1H NMR and MS
data.16
General procedure for preparation of polyenes17
b-Ionylidenethyltriphenylphosphonium bromide (7, ca. 800 mg),
sodium hydride (ca. 300 mg) and the appropriate diapocarotene-
dial (10–12, 10–20 mg) was stirred in dichloromethane overnight.
Excess sodium hydride was decomposed by dropwise addition
of acetic acid. The products were extracted with hexane, and
washed with water and saturated sodium chloride solution. Initial
purification was carried out by column chromatography on a
silica column, with 1–2% acetone in hexane as eluent. Final
purification was achieved by preparative HPLC (mobile phase:
various methanol–acetone gradients, 2.5 mL min−1). Multiple
injections of each compounds from dichloromethane solution
were required.
3,7,11,16,20,24-Hexamethyl-1,26-bis(2,6,6-trimethylcyclohex-
1-enyl)hexacosa-1,3,5,7,9,11,13,15,17,19,21,23,25-tridecaene (de-
capreno-b-carotene, 4). Yield 77%, spectrophotometrically deter-
mined. kmax/nm (acetone) 477sh, 502, 534; %III/II 15; RT = 45.6
(7%, kmax/nm 473sh, 497, 531), RT = 46.8 (84%, kmax/nm 477sh,
501, 535), RT = 48.0 (8%, kmax/nm 479sh, 501, 525); dH(600 MHz,
CDCl3) 1.04 (6ꢀ/6ꢀꢀ-CH3), 1.48 (5ꢀ/5ꢀꢀ-CH2), 1.63 (4ꢀ/4ꢀꢀ-CH2), 1.73
(2ꢀ/2ꢀꢀ-CH3), 1.98 (3/24-CH3), 1.99 (11/16-CH3), 2.00 (7/20-CH3),
2.03 (3ꢀ/3ꢀꢀ-CH2), 6.16 (2/4/23/25-CH), 6.18 (1/26-CH), 6.24
(8/19-CH), 6.29 (12/15-CH), 6.38 (6/21-CH), 6.39 (10/17-CH),
6.65 (13/14-CH), 6.66 (5/9/18/22-CH); dC(150 MHz, CDCl3)
12.6 (3/7/20/24-CH3), 12.7 (11/16-CH3), 19.1 (4ꢀ/4ꢀꢀ-CH2), 21.6
Allylic functionalization of polyenes
a) With NBS. 18 To b,b-carotene (1, 5.9 mg, 11 lmol) in 5 mL
CHCl3, cooled to −20 ◦C in a CO2–ethylene glycol bath, a mixture
of N-bromosuccinimide (1.8 mg, 10 lmol) and acetic acid (50 lL,
0.87 mmol) in 4 mL CHCl3 was added, giving a brown solution.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2803–2811 | 2809
©

View Article Online
After 1 min, methylmorpholine (40 lL, 0.36 mmol) was added,
restoring the orange colour. The organic phase was washed with
water and saturated NaCl (aq.). After purification on a neutral
alumina column, developed with 10% acetone in hexane, HPLC
analysis revealed two main peaks: RT = 13.4 (43%, kmax/nm 427sh,
450, 477; %III/II 29), RT = 14.7 (31%, kmax/nm 423sh, 445,
473; %III/II 33). An authentic sample of the monoacetate of
b,b-caroten-4-ol gave: RT = 22.8 (kmax/nm 425sh, 451, 481; %III/II
54). Alteration of the b,b-carotene/acetic acid stoichiometry from
∼1 : 100 to 1 : 4 gave destruction of the carotenoid, as evidenced
by HPLC.
J = 7.6 Hz), 8.23 (t, 1H, para-CH, J = 7.4 Hz); dC(150 MHz,
CDCl3) 130.4 (meta-CH), 140.1 (1-CH), 142.8 (ortho-CH), 143.2
(para-CH), 211.3 (Ph3C).
VIS/NIR experiments
a) With triphenylcarbenium tetrafluoroborate. To 3 mL of a
4–6 mg mL−1 dichloromethane solution of triphenylcarbenium
tetrafluoroborate (from a commercial source) in a cuvette, were
added polyenes via a syringe from dichloromethane stock solu-
tions. The VIS/NIR spectra (500–1600 nm) were recorded repeat-
edly at maximum 10 min intervals for at least 1 h. Experiments were
performed both at room temperature and at −15 ◦C, although
the latter experiments suffered from condensation problems.
Polyenes employed: b,b-carotene (1), decapreno-b-carotene (4),
C54-b-carotene (5) and dodecapreno-b-carotene (6).
b) With BF3–diethyl etherate. The b,b-carotene dication, pre-
pared from addition of BF3–diethyl etherate to b,b-carotene (1),
was reacted with water in the same manner as reported before,20
except for modification of the temperature to −20 ◦C. Pigment
recovery was 23%. HPLC analysis showed only one main region
containing peaks, and only traces of other compounds: RT = 16.4
(72%, kmax/nm 427sh, 450, 476; %III/II 16), RT = 17.0 (12%,
kmax/nm 422sh, 446, 471; %III/II 25), RT = 17.6 (5%, kmax/nm
421sh, 445, 473; %III/II 45). An authentic sample of b,b-caroten-
4-ol (2) gave: RT = 17.0 (kmax/nm 427sh, 451, 479; %III/II 32).
Employing the same methodology with C54-b-carotene (5, 5.7
mg), a pigment recovery of 50% was obtained. The HPLC analysis
showed peaks appearing in two main regions, RT = 29.0–35.0
(57%, functionalised polyenes) and RT = 47.8–53.2 (38%, unpolar
polyenes). Main peaks: RT = 30.1 (22%, kmax/nm 491, 517, 553;
%III/II 30), RT = 48.8 (16%, kmax/nm 493, 518, 554; %III/II
25). The more polar fraction was purified by preparative HPLC
and partially characterised by NMR; dH(500 MHz, CD2Cl2)
1.47 (5ꢀꢀ-CH2), 1.62 (4ꢀꢀ-CH2), 1.66 (4ꢀ-CHaHb), 1.71 (2ꢀꢀ-CH3),
1.89 (4ꢀ-CHaHb), 2.02 (3ꢀꢀ-CH2), 3.97 (3ꢀ-CHOH). Decomposition
products arising from elimination of the allylic hydroxy group
emerged during acquisition in one of two samples.
b) With trifluoroacetic acid. VIS/NIR spectra were recorded
in the same manner as described above for the four polyenes 1, 4,
5 and 6 in dichloromethane solutions of trifluoroacetic acid, using
concentrations of 0.013 M and 1.3 M. The measurements were
performed both at room temperature and at −15 ◦C.
NMR of polyenes reacted with triphenylcarbenium
tetrafluoroborate-d15 (16)
a) Low temperature preparation. The polyene (1–2 mg) was
dissolved in 0.25 mL CDCl3, and cooled to −78 ◦C under N2.
Trityl cation 16 (2–4 mg), dissolved in 0.25 CDCl3, was added,
and the mixture was allowed to heat slowly. NMR spectra were
recorded at 5 ◦C. Polyenes employed: b,b-carotene (1), decapreno-
b-carotene (4) and neochrome (13).
b) Ordinary preparation. To 1 mg trityl cation dissolved in
0.5 mL CDCl3, solid polyene (1–2 mg) was added. NMR spectra
of the mixture were recorded at 5 ◦C. Polyenes employed: b,b-
carotene (1), echinenone (14).
Triphenylmethanol-d15 (17)
Bromobenzene-d5 (16, 5.17 g, 31.9 mmol) was mixed with
10 mL diethyl ether and added dropwise to magnesium (0.78 g,
32.1 mmol) containing a few iodine crystals. The reaction started
immediately and the mixture was refluxed for 1 h. Diethyl
carbonate (1.22 g, 10.3 mmol) in 5 mL diethyl ether,37 was added
dropwise, and spontaneous boiling started. After addition of the
carbonate and reflux for another hour, the mixture was poured
over 20 g ice and 25 mL 1 M H2SO4. The organic phase was
washed with additional 1 M H2SO4 and saturated sodium chloride
solution and then dried over anhydrous MgSO4. Hexane (20 mL)
was added and the ether was boiled off under an aspirator until
crystalization was initiated. This afforded 1.48 g white crystals,
c) Low temperature NMR. The polyene (1–2 mg) was dissolved
◦
in 0.4 mL CDCl3, and kept at −15 C while 2–4 mg of the trityl
reagent, dissolved in 0.1 mL CDCl3, was added. NMR spectra
were obtained at −15 ◦C. Polyenes employed: b,b-carotene (1),
decapreno-b-carotene (4) and neochrome (13).
All polyene/trityl cation mixtures yielded blue or green solu-
tions. In the case of neochrome (13), precipitates were formed
1
1
upon addition of the trityl reagent 15. Initially, H and H,1H
COSY were recorded, with additional 2D experiments performed
on samples showing promising proton spectra.
◦
52% overall yield; mp 158–161 C; m/z (EI) 276 (4.6%, M + 1),
NMR of polyenes reacted with trifluoroacetic acid-d
275 (20, M+), 194 (16), 193 (100), 192 (22), 174 (16), 165 (21), 164
(19), 110 (82), 82 (39).
To a solution of polyene (1–2 mg) in CDCl3 at −15 ◦C (for
C54-b-carotene (5) and dodecapreno-b-carotene (6) in CD2Cl2),
10 lL of trifluoroacetic acid-d was added. NMR spectra were
recorded at this temperature or slightly higher, as indicated below.
Polyenes employed: b,b-carotene (1), decapreno-b-carotene (4),
C54-b-carotene (5, spectra recorded at −10 ◦C) and dodecapreno-
b-carotene (6, spectra recorded at −5 ◦C).
Triphenylcarbenium tetrafluoroborate-d15 (15)
From 0.78 g triphenylmethanol-d15 (17), 0.62 g of a red/orange
crystalline product was obtained.22 NMR of the corresponding
undeuterated product is given below. Traces of propanoic acid and
some unreacted triphenylmethanol were observed. dH(600 MHz,
CDCl3) 7.71 (d, 2H, ortho-CH, J = 7.4 Hz), 7.88 (t, 2H, meta-CH,
The average olefinic chemical shift for the cations from from 4,
5 and 6 was estimated using Simpson’s formula24 on 1D 1H spectra
2810 | Org. Biomol. Chem., 2007, 5, 2803–2811
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
for evaluation of the integrals in the quotient
9 H. Spiesecke and W. G. Schneider, Tetrahedron Lett., 1961, 3, 468–472.
10 D. H. O’Brien, A. J. Hart and C. R. Russell, J. Am. Chem. Soc., 1975,
97, 4410–4412.
ꢀ
(intensity × dH) ddH
ꢀ
.
11 L. M. Tolbert and M. E. Ogle, J. Am. Chem. Soc., 1990, 112, 9519–
(intensity) ddH
9527.
12 L. M. Tolbert, Acc. Chem. Res., 1992, 25, 561–568.
13 B. F. Lutnaes, G. Kildahl-Andersen, J. Krane and S. Liaaen-Jensen,
J. Am. Chem. Soc., 2004, 126, 8981–8990.
Quenching trials with b,b-carotene (1)
14 G. Kildahl-Andersen, B. F. Lutnaes, J. Krane and S. Liaaen-Jensen,
Org. Lett., 2003, 5, 2675–2678.
To a stirred solution of triphenylcarbenium tetrafluoroborate
(72 mg) in 5 mL CHCl3, b,b-carotene (1, 5 mg), dissolved in
5 mL CHCl3, was added. The resulting green solution was mixed
with water in a separation funnel and washed with 2% aqueous
sodium bicarbonate solution until the orange colour was restored.
The organic phase was washed with saturated NaCl (aq.), and
VIS/NIR analysis showed a pigment recovery of 16%. HPLC
analysis revealed three main peaks: RT = 16.8 (55%, kmax/nm
427sh, 450, 477; %III/II 25; 2), RT = 34.4 (15%, kmax/nm 442,
465, 491; %III/II 29; 18), RT = 35.2 (17%, kmax/nm 449, 472, 505;
%III/II 40; 18).
When the resulting green solution instead was shaken with
20 mL 5% potassium acetate, the following results were obtained
by HPLC analysis: RT = 16.7 (26%, kmax/nm 427sh, 450, 476;
%III/II 21; 2), RT = 22.7 (15%, kmax/nm 427sh, 451, 478; %III/II
21; monoacetate of 2), RT = 25.2 (3%, kmax/nm 457), RT = 34.2
(5%, kmax/nm 463, 491; 18), RT = 35.0 (6%, kmax/nm 449, 473, 499;
18).
15 P. Karrer and C. H. Eugster, Helv. Chim. Acta, 1951, 34, 28–33; O. Isler,
M. Montavon and R. Ru¨egg, Liebigs Ann. Chem., 1957, 603, 129–144;
J. D. Surmatis and A. Ofner, J. Org. Chem., 1961, 26, 1171–1173.
16 G. Broszeit, F. Diepenbrock, O. Gra¨f, D. Hecht, J. Heinze, H.-D.
Martin, B. Mayer, K. Schaper, A. Smie and H.-H. Strehblow, Liebigs
Ann., 1997, 11, 2205–2213.
17 H. Kjøsen and S. Liaaen-Jensen, Acta Chem. Scand., 1973, 27, 2495–
2502.
18 R. Entschel and P. Karrer, Helv. Chim. Acta, 1958, 41, 402–413.
19 F. J. Petracek and L. Zechmeister, J. Am. Chem. Soc., 1956, 78, 3188–
3191.
20 B. F. Lutnaes, L. Brua˚s, G. Kildahl-Andersen, J. Krane and S. Liaaen-
Jensen, Org. Biomol. Chem., 2003, 1, 4064–4072.
21 J. A. Haugan and S. Liaaen-Jensen, Acta Chem. Scand., 1994, 48, 152–
159.
22 H. J. Dauben, Jr., L. R. Honnen and K. M. Harmon, J. Org. Chem.,
1960, 25, 1442–1445.
23 S. J. Kitchin, M. Xu, H. Serrano-Gonza´lez, L. J. Coates, S. Z. Ahmed,
C. Glidewell and K. D. M. Harris, J. Solid State Chem., 2006, 179,
1335–1338.
24 C. H. Edwards and D. E. Penney, Calculus, 5th edn, Prentice Hall,
Upper Saddle River, NJ, 1998, pp. 348–351.
25 B. F. Lutnaes, Masters thesis, NTNU, Trondheim, 1999.
26 V. V. Konovalov and L. D. Kispert, J. Chem. Soc., Perkin Trans. 2, 1999,
901–909.
Acknowledgements
G. K.-A. was supported by a KOSK PhD fellowship from The
Research Council of Norway.
27 J. R. Reimers, J. S. Craw and N. S. Hush, J. Phys. Chem., 1993, 97,
2778–2787.
28 H. H. Jaffe´ and M. Orchin, Theory and Applications of Ultraviolet
Spectroscopy, John Wiley & Sons, New York, 1962, pp. 220–241.
29 T. S. Sorensen, J. Am. Chem. Soc., 1965, 87, 5075–5084; T. S. Sorensen,
in Carbonium Ions, Vol 2: Methods of formation and major types, ed.
G. A. Olah and P. von R. Schleyer, Wiley-Interscience, New York, 1970,
pp. 807–835.
References
1 W. P. Su, J. R. Schrieffer and A. J. Heeger, Phys. Rev. Lett., 1979, 42,
1698–1701.
2 A. J. Heeger, S. Kivelson, J. R. Schrieffer and W.-P. Su, Rev. Mod. Phys.,
30 J. Lafferty, A. C. Roach, R. S. Sinclair, T. G. Truscott and E. J. Land,
J. Chem. Soc., Faraday Trans. 1, 1977, 73, 416–429.
31 G. Kildahl-Andersen, T. Anthonsen and S. Liaaen-Jensen, Tetrahedron
Lett., 2006, 47, 4693–4696.
32 M. Rumi, A. Kiehl and G. Zerbi, Chem. Phys. Lett., 1994,
231, 70–74.
33 C. H. Eugster, in Carotenoids, vol. 1A: Isolation and analysis, ed.
1988, 60, 781–850.
3 L. M. Tolbert and M. E. Ogle, J. Am. Chem. Soc., 1989, 111, 5958–5959.
4 A. Feldblum, J. H. Kaufman, S. Etemad, A. J. Heeger, T.-C. Chung and
A. G. MacDiarmid, Phys. Rev. B: Condens. Matter, 1982, 26, 815–826.
5 N. Suzuki, M. Ozaki, S. Etemad, A. J. Heeger and A. G. MacDiarmid,
Phys. Rev. Lett., 1980, 45, 1209–1213.
6 A. J. Heeger and J. R. Schrieffer, Solid State Commun., 1983, 48, 207–
210.
¨
G. Britton, H. Pfander and S. Liaaen-Jensen, Birkhauser, Basel, 1995,
pp. 71–80.
7 B. R. Weinberger, E. Ehrenfreund, A. Pron, A. J. Heeger and A. G.
MacDiarmid, J. Chem. Phys., 1980, 72, 4749–4755.
8 S. Liaaen-Jensen and B. F. Lutnaes, in Studies in Natural Products
Chemistry, Vol. 30: Bioactive Natural Products (Part K), ed. Atta-ur-
Rahman, Elsevier, Amsterdam, 2005, pp. 515–557.
34 R. Burger, C. Schorn and P. Bigler, J. Magn. Reson., 2001, 148, 88–94.
35 Y. Ishikawa, Bull. Chem. Soc. Jpn., 1964, 37, 207–209.
36 W. Sarnecki and H. Pommer, Ger. Pat., 1959, 1 060 386.
37 J. Kobayashi, K. Goto, T. Kawashima, M. W. Schmidt and S. Nagase,
J. Am. Chem. Soc., 2002, 124, 3703–3712.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2803–2811 | 2811
©