The longest polyene

Article abstract of DOI:10.1021/ol302577d

A microwave assisted Wittig reaction allowed the synthesis, in good yields, of the longest polyene so far recorded with 27 conjugated double bonds. The synthesis of this stable, well-soluble polyene represents a noteworthy step in the direction of ultimat

Full text of DOI:10.1021/ol302577d

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
The Longest Polyene
†
Muhammad Zeeshan, Hans-Richard Sliwka,* Vassilia Partali, and Ana Martınez*
,†
†
,‡
´
Department of Chemistry, Norwegian University of Science and Technology,
Trondheim, Norway, and Instituto de Investigaciones en Materiales, Universidad
ꢀ
Nacional Autonoma de Mexico, Circuito Exterior s/n, Coyoacan 04510, Mexico
ꢀ
ꢀ
hrs@nvg.ntnu.no; martina@iim.unam.mx
Received September 19, 2012
ABSTRACT
A microwave assisted Wittig reaction allowed the synthesis, in good yields, of the longest polyene so far recorded with 27 conjugated double
bonds. The synthesis of this stable, well-soluble polyene represents a noteworthy step in the direction of ultimate λmax
.
Since discovering conjugation in 1899 chemists have
struggled with varying success to extend the length of
1,3-butadiene, the elementary conjugated chain of four
carbons with two adjacent double bonds, C4:2.1 All chem-
ists have appended a double bond to butadiene at least
one time during their studies and have added the legendary
FieserꢀWoodward increment of 30 nm when calculating
the λmax of a sketched compound. However, in the labora-
tory, adding CdC units to butadiene leads to an impasse at
C8:4 (terminal ene) or C12:5 (terminal methyl).2,3 Capping
the chain ends with phenyl or tert-butyl groups induces
stability and tolerates chain elongation to C28:8 and C42:13,
respectively.4,5
Annulenes, another class of polyenes, could not be in-
creased beyond 15 double bonds ([30]annulene C30:15).6
Five benzene rings can be linearly fused to stable pentacene
C22:11; hexacene C26:13 decomposes.7 Cyclic annulation of
benzene (C6:3) rings forms durable polycyclic aromatic
hydrocarbons (PAH) reaching via fullerenes and nanotubes
an unlimited amount of CdC bonds in graphite.8 However,
the inertness of PAHs, the limiting double bonds in plain
polyenes, annulenes, and acenes, complicates investigating
their optical, photochemical, and, most topical, semiconduc-
tor properties and precludes answering the pertinent ques-
tion: How many double bonds are required to reach the
ultimate λmax? This question is constantly addressed since
the concept of conjugation became prevailing and has
gathered different answers from extrapolations and molec-
ular calculations.9ꢀ11 Missing is the experimental verifica-
tion of λult, which is hampered by the lack of stable polyenes.
The longest naturally occurring polyene offers 14 double
bonds.12 The authentication of arduously HPLC-separated
polyenes with 100, 240, or even 1190 CdC bonds (claimed
“infinitely long polyacetylenes”) failed: these polyenes with
λmax around 550 nm express no more than 20 conjugated
CdC bonds.13,14 A laborious synthesis with a molybdenum
compound gave a complex product mixture from which
a C167:23 ester in a distorted zigzag-conjugated cis/trans
arrangement was isolated by repeated HPLC15 (Scheme 1S,
Supporting Information).
† Norwegian University of Science and Technology.
‡
ꢀ
ꢀ
Universidad Nacional Autonoma de Mexico.
€
(8) Rieger, R.; Mullen, K. J. Phys. Org. Chem. 2010, 23, 315.
(1) Thiele, J. Liebigs Ann. Chem. 1899, 306, 87.
(2) Bohlmann, F. Chem. Ber. 1952, 85, 386.
(9) Kuhn, W. Helv. Chim. Acta 1948, 31, 1780.
(10) Kohler, B. E.; Samuel, I. D. W. J. Chem. Phys. 1995, 103, 6248.
(11) Schmalz, T. G.; Griffin, L. L. J. Chem. Phys. 2009, 131, 22430.
(12) Kushwaha, S. C.; Gochnauer, M. B.; Kushner, D. J.; Kates, M.
Can. J. Microbiol. 1974, 20, 241.
(13) Kohler, B. E.; Woehl, J. C. J. Chem. Phys. 1995, 103, 6253.
(14) Ledoux, I.; Samuel, I. D. W.; Zyss, J.; Yaliraki, S. N.; Schattenmann,
F. J.; Schrock, R. R.; Silbey, R. J. Chem. Phys. 1999, 245, 1.
(15) Scriban, C.; Amagai, B. S.; Stemmler, E. A.; Christensen, R. L.;
Schrock, R. R. J. Am. Chem. Soc. 2009, 131, 13441.
(3) Spangler, C. W.; Little, D. A. J. Chem. Soc., Perkin Trans. 1 1982,
2379.
(4) Kuhn, R.; Winterstein, A. Helv. Chim. Acta 1928, 11, 87.
(5) Klein, D.; Kilickiran, P.; Mlynek, C.; Hopf, H.; Dix, I.; Jones,
P. G. Chem.;Eur. J. 2010, 16, 10507.
(6) Spitler, E. L.; Johnson, C. A.; Haley, M. M. Chem. Rev. 2006, 106,
5344.
(7) Mondal, R.; Adhikari, R. M.; Shah, B. K.; D. C. Neckers, D. C.
Org. Lett. 2007, 9, 2505.
r
10.1021/ol302577d
XXXX American Chemical Society

Since plain polyenes, annulenes, and acenes are not
expandable, elongation synthesis has to rely on stabilized
chains. Stable polyenes appear in nature as methyl
branched, and many are, in addition, terminated with
cyclohexenyl groups in the carotenoid series. All references
to previous attempts demonstrate that chain lengthening
of carotenoids is, so far, the only compelling recourse
to predefined long polyenes. The Wittig reaction provides
an efficient synthesis to carotenoids.16 Even so, the pre-
paration of long-chain carotenoids under classical Wittig
conditions always attends substantial decomposition;
the stability of polyenes decreases with increasing length.
A C70:23 carotenoid decomposed at low temperatures
under argon.17 Not surprisingly, the dictum expressed
in 1951 by Karrer and Eugster still reigns over carot-
enoid chemistry: ‘The synthesis of dodecapreno-β,β-
carotene (C60:19) appears to represent the limit of our
method.’18
Scheme 1. Microwave Assisted Wittig Synthesis of C80Zea:27
We have now taken a great stride toward λult. We found
that the reaction time can substantially be reduced
and yields increased when the Wittig reaction is performed
with microwave irradiation. Shorter reaction times pre-
vent the formation of byproducts and result in in stable
compounds. The microwave Wittig synthesis facilitated
the addition of ylide C15ZeaP with C50-dialdehyde
(C50diAld) to blue C80-zeaxanthin in 63% yield (Scheme 1).
C50diAld was obtained by several sequences: two pro-
tected C5P aldehydes and C20diAld gave C30diAld, which
reacted with two other C5P units to C40diAld and con-
secutively to C50diAld (Scheme 2).
Although the short wavelength part of the vis-spectrum
of C80Zea:27 deviates from the habitual C40Zea:11 shape
(Figure 1) it conforms to the fine structure seen with
C60:19-carotene.18,19 Nevertheless, since the polyene chain
may be sensitive to spatial variations the occurrence of
cis-(Z)-isomers cannot be excluded; after all, more than one
million isomers are theoretically expected.20 On the other
hand, the short high-temperature microwave reaction fa-
vors the formation of the all trans-(E)-isomer;21 the longer
the polyene, the more the trans-isomer is stabilized.4,22 The
absence of spectral changes after the classical cis f trans
23
isomerization experiment with light and I2 supports the
predominance of the all-trans-isomer (Figure 1S, Support-
ing Information).
Molecular modeling with time-dependent density func-
tional theory (TDDFT) using the M06-2X functional24
in combination with the 6-311g basic set outlined the
geometry of C80Zea and then generated the absorption
spectra in CH2Cl2 with the polarizable continuum model.
The molecule appeared with an in-plane bended structure
typical for carotenoids (Figure 2).
(16) Ernst, H. Pure Appl. Chem. 2002, 74, 1369.
(17) Broszeit, G.; Diepenbrock, F.; Graf, O.; Hecht, D.; Heinze, J.;
Martin, H. D.; Mayer, B.; Schaper, K.; Smie, A.; Strehblow, H. H.
Liebigs Ann. Recl. 1997, 2205.
(18) Karrer, P; Eugster, C. H. Helv. Chim. Acta 1951, 34, 1805.
(19) P. O. Andersson, P.; Gillbro, T. J. Chem. Phys. 1995, 103, 2509.
(20) Zechmeister, L. Cis-trans Isomeric carotenoids, vitamins A, and
arylpolyenes; Springer: Wien, 1962; p 11.
(21) Koyama, Y.; Fujii, R. Cis-trans carotenoids in photosynthesis:
Configurations, excited-state properties and physiological functions. In
The photochemistry of carotenoids; Frank, H. A., Young, A. J., Britton, G.,
Codgell, R. J., Eds.; Kluwer: Dordrecht, 1999; p 172.
(22) Christensen, R. L.; Barney, E. A.; Broene, R. D.; Galinato,
M. G. I.; Frank, H. A. Arch. Biochem. Biophys. 2004, 430, 30.
(23) Kon, M.; Britton, G. Food Sci. Technol. Res. 1999, 5, 347.
(24) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215.
The distinct alignment with positive and negative am-
plitudes hinders C80Zea to extend to twice the length of
the barely curved C40Zea25 (C80Zea lcalc = 63 A, C40Zea
˚
˚
calc = 40 A). The prolonged conjugation is expressed by
l
an adjustment of the bond lengths (calculated range in
˚
˚
C80Zea: CdC 1.3697ꢀ1.3887 A, C;C 1.4347ꢀ1.4898 A,
0
˚
C80Zea mid CdC (31,31 ) = 1.3836 A, measured in
€
(25) Linden, A.; Burgi, B.; Eugster, C. H. Helv. Chim. Acta 2004, 87,
1254.
B
Org. Lett., Vol. XX, No. XX, XXXX

Scheme 2. Synthesis of C50-Dialdehyde
Figure 2. Calculated structure of trans-C80Zea:27.
Figure 3. Calculted vis-spectrum of trans-C80Zea:27, λmax
672 nm (CH2Cl2).
=
spectrum of the 31,310-cis-isomer is quite different in shape
(Figure S2, Supporting Information), thus supporting the
experimentally indicated absence of cis-isomers.
Figure 1. Vis-spectrum of trans-C80Zea:27, λmax = 576 nm
(CH2Cl2).
The microwave Wittig reaction has surpassed the solu-
tion-based method heading to C80Zea:27, a noteworthy
step in the direction of ultimate λmax. This long, well-
defined polyene, when modified at the peripheral rings, is
moreovera likelycandidateformolecularwirestudies.27ꢀ29
Despite recently refined molecular modeling methods,
the errors in spectra calculation of polyenes are still notice-
able. The insolubility in current solvents of dialdehydes
with more than C50 prevents going beyond C80Zea:27
with Wittig reactions based on prefabricated synthons
(Schemes 1, 2). The challenge is now to develop a synthesis
for carotenoids even longer than C80Zea:27; maximal
elongated, soluble, and stable polyenes will be attained
with shorter dialdehydes CndiAld (n < 50) and longer
Wittig salts CnZeaP n > 15.
˚
˚
C40Zea: CdC 1.3216ꢀ1.3666 A, C;C 1.4255ꢀ1.4737 A,
25
0
˚
mid CdC (15,15 ) = 1.3416 A).
Long-chain polyenes point not only to synthetic com-
plications but also to obstacles for theoretical calculations.
The λmax of the computed spectrum (672 nm, CH2Cl2)
deviates considerably from experimental λmax = 576 nm
(CH2Cl2); the calculation overestimated the absorption by
3.7 nm for each conjugated CdC bond (the ring double
bonds are not fully conjugated with the chain bonds due to
sterical deviation). Spectral prediction of polyenes, includ-
ing C40Zea, is susceptible to inaccuracies,26 although
TDDFT schemes operate with lower average errors than
ZINDO/S and M06-2X performs better than other func-
tionals.24,26 The calculated spectrum of trans-C80Zea
shows an absorption shoulder in the lower wavelength
range (Figure 3) in agreement with the recorded data. The
Acknowledgment. We thank H. Ernst, BASF SE, for
a generous gift of starting compounds, S. V. Gonzales for
HRMS, and DGTIC for computer time at KanBalam.
(26) Martins, J. B. L.; Duraes, J. A.; Sales, M. J. A.; Vilela, A. F. A.;
Silva, G. M. E.; Gargano, R. Int. J. Quantum Chem. 2009, 109, 739.
(27) Sliwka, H. R.; Liaaen-Jensen, S. Tetrahedron: Asymmetry 1993,
4, 361.
(28) Ion, A.; Partali, V.; Sliwka, H. R.; Banica, F. G. Electrochem.
Commun. 2002, 4, 674.
(29) Leatherman, G.; Durantini, E. N.; Gust, D.; Moore, T. A.;
Moore, A. L.; Stone, S.; Zhou, Z.; Rez, P.; Liu, Y. Z.; Lindsay, S. M.
J. Phys. Chem. B 1999, 103, 4006.
Supporting Information Available. Schemes, figures,
experimental details, and characterization. This material
is available free of charge via the Internet at http://pubs.
acs.org.
The authors declare no competing financial interest.
Org. Lett., Vol. XX, No. XX, XXXX
C