Exciton spectroscopy of red polydiacetylene chains in single crystals

Article abstract of DOI:10.1016/j.cplett.2007.07.031

The absorption and emission properties of excitons are investigated in a pure 'red' phase of luminescent polydiacetylene wires. The values of the binding energy (E_b = 0.556 eV) and Bohr radius (r ～ 14 ?) make the 'red' exciton similar to the one of the extensively studied 'blue' phase, so the conformational modifications which lead to a reordering of excited states and luminescent emission do not strongly affect the exciton structure. The opening of a thermally activated non radiative channel leads to an efficient quenching of the intense luminescence observed at low temperature.

Full text of DOI:10.1016/j.cplett.2007.07.031

Chemical Physics Letters 444 (2007) 309–313
www.elsevier.com/locate/cplett
Exciton spectroscopy of red polydiacetylene chains in single crystals
a,
a
b
c
c
c
T. Barisien *, L. Legrand , G. Weiser , J. Deschamps , M. Balog , B. Boury ,
c
a
S.G. Dutremez , M. Schott
a
Institut des NanoSciences de Paris, UMR7588 CNRS, Universite´s Pierre et Marie Curie (Paris 6) et Denis Diderot (Paris 7), Campus Boucicaut,
140 rue de Lourmel, 75015 Paris, France
b
Faculty of Physics and Centre of Material Science, Philipps-Universita¨t, 35037 Marburg, Germany
ˆ
Institut Charles Gerhardt Montpellier, UMR 5253 CNRS-UM2-ENSCM-UM1, Universite´ Montpellier II, Bat. 17, CC 1700,
c
`
Place Eugene Bataillon, 34095 Montpellier Cedex 5, France
Received 5 June 2007
Available online 19 July 2007
Abstract
The absorption and emission properties of excitons are investigated in a pure ‘red’ phase of luminescent polydiacetylene wires. The
˚
values of the binding energy (Eb = 0.556 eV) and Bohr radius (r ꢀ 14 A) make the ‘red’ exciton similar to the one of the extensively stud-
ied ‘blue’ phase, so the conformational modifications which lead to a reordering of excited states and luminescent emission do not
strongly affect the exciton structure. The opening of a thermally activated non radiative channel leads to an efficient quenching of the
intense luminescence observed at low temperature.
ꢀ 2007 Elsevier B.V. All rights reserved.
1. Introduction
The differences in electronic properties of each phase are
necessarily associated to modifications of the ground state
structure. As the current data from X-ray diffraction expe-
riments do not discriminate between red and blue geome-
tries, the structural differences in the ground state must
be quite small. The only results showing clear variations
are from solid state 13C NMR spectroscopy [6,7]. Determi-
nation of the red chain conformation is thus a still unre-
solved issue which is related to the general question of
the relationship between electronic configuration and struc-
tural parameters in conjugated polymers.
Given this situation, we have searched for a purely red
system of PDA chains with the further requirement of
being able to achieve sufficiently low polymer contents in
a good crystallographic environment. The spectroscopic
results presented in this Letter indicate that the polydiacet-
ylene chain obtained from the polymerization of 1,6-
bis(diphenylamino)-2,4-hexadiyne (THD) leads to the
formation of a purely red phase, promising for compara-
tive studies between the two conformational isomers of
PDA. Up to now the structural and electronic properties
of poly-THD have been addressed in a limited number of
papers. The room temperature crystal structure has been
Polydiacetylene (PDA) chains exist in the solid state as
either of two electronic structures commonly called the
blue and red phases. Single isolated chains of red-phase
poly-3BCMU have remarkable electronic properties. A
chain behaves like a quasi-perfect one-dimensional system
[1] with the spatial coherence of the red exciton extending
over macroscopic distances [2]. However, in these crystals,
red chains coexist with a majority population of blue
chains. The red chains can be discriminated from the
weakly emitting blue ones only because of their strong
luminescence. The coexistence of two populations of chains
in the same crystal is a limiting factor since the absorption
properties of the red polydiacetylenic structure cannot be
selectively probed. For instance, photo-induced absorption
experiments which have allowed the investigation of the
relaxation dynamics of the blue phase exciton are impossi-
ble [3–5].
*
Corresponding author. Fax: +33 1 43 54 28 78.
E-mail address: Thierry.Barisien@insp.jussieu.fr (T. Barisien).
0009-2614/$ - see front matter ꢀ 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.cplett.2007.07.031

310
T. Barisien et al. / Chemical Physics Letters 444 (2007) 309–313
1H NMR, 13C NMR and IR data fully agree with the
expected molecular structure and previous literature
reports [14]. Monomer crystals were grown from a 4:1
methylene chloride–diethyl ether mixture (c ꢀ 15 g/l), at
0 ꢁC, in the dark. The crystals are needle-shaped with a
width of a fraction of millimetre, the chains being oriented
along the main dimension.
Polarized reflectance spectra in the energy range
between 1.5 and 5.8 eV were obtained from flat areas of
about 0.5 mm · 1 mm in size using a windowless vacuum
spectrometer and calcite polarizers. For electroreflectance
measurements, Cr/Mg contacts spaced by a 0.5 mm dis-
tance were evaporated to apply along the polymer back-
bone an external sinusoidal voltage at 1 kHz frequency.
The field modulated light was collected by lock-in at twice
the field frequency and evaluated as described in previous
studies [15]. Absorption spectra were derived by Kra-
mers–Kronig transformation for comparison with absorp-
tion for light polarized perpendicular to the chain. The
latter spectra were measured in the double beam mode of
a Cary 5000 double monochromator using freshly prepared
samples with low polymer content and selecting 0.1 mm
Fig. 1. Reflectance and electroreflectance spectra with light polarized
parallel to the polymer chain. The ER spectrum obtained for 30 kV/cm is
reduced in size by a factor 4. The dotted line shows the derivative of the
reflectivity.
˚
wide homogeneous areas. The spectral resolution was 4 A
˚
for reflectance and transmission measurements and 5 A
for electroreflectance spectra.
Raman spectra were measured using a Jobin Yvon
U1000 double monochromator and the excitation was the
647.1 nm line of a mixed Ar+/Kr+ laser. For both Raman
and absorption measurements the samples were mounted
on the cold finger of a He gas exchange cryostat that allows
studies down to 10 K.
The emission was studied in a confocal-like configura-
tion. Detailed description of the set-up can be found else-
where [16].
determined [8]. Room temperature reflectance and Raman
scattering spectra have been measured by Morrow et al. [9].
The Kramers–Kronig transformation of the reflectance
yields a zero phonon absorption band peaked at 568 nm
and characteristic of a red PDA phase. The fluorescence
spectra of thin films made of microcrystallites dispersed
in a PMMA matrix have also been measured [10]. So far,
all of the studies have dealt with fully or almost fully poly-
merized crystals.
In this work, the exciton spectroscopy of highly ordered
poly-THD chains dispersed in their monomer crystal is
investigated at low temperature. As previous reports indi-
cated a strong thermal reactivity of crystalline THD, par-
ticular attention has to be paid to avoid thermal reaction.
In our work, the polymer content in weight, xp, was kept
around a few percent except for reflectance and electrore-
flectance experiments where xp was ꢀ65%, as inferred by
the absolute value of the reflectance and the sum rule of
the oscillator strength (see Fig. 1).
3. Results
3.1. Reflectance and electroreflectance
The reflectivity of the samples in the visible region is
very anisotropic. For light polarized perpendicular to the
polymer chain, the reflectivity shows no structure and
remains below 5% up to 3.8 eV where the absorption by
the phenyl side-groups sets in. For parallel polarization
of light, the room temperature spectrum in Fig. 1 shows
a broad exciton peak of about 30% reflectance followed
by a vibrational band as reported earlier [9]. Low temper-
ature reflectance spectra show some narrowing of the main
peak, and a manifold of vibronic exciton peaks appears as
is typically observed in high-quality polydiacetylene single
crystals [15,17].
Much more structure is resolved in low temperature
electroreflectance (ER) spectra (Fig. 1). Two spectra scaled
quadratically to the fields of 15 and 30 kV/cm (full and
dashed lines, respectively) are almost identical. This proves
that all features below ꢀ2.6 eV increase quadratically with
field strength and are reproduced in their spectral position
2. Experimental
1,6-Bis(diphenylamino)-2,4-hexadiyne (THD) was pre-
pared in two steps starting from commercial diphenyl-
amine. In the first step, lithium diphenylamide was
allowed to react with stoichiometric amounts of propargyl
bromide, in THF, at ꢁ78ꢁC. N,N-Diphenylpropargylamine
was obtained as a white crystalline solid after purification
by vacuum distillation and column chromatography
(yield = 50%). THD was obtained in 65% yield by oxida-
tive dimerization of N,N-diphenylpropargylamine [11–13].

T. Barisien et al. / Chemical Physics Letters 444 (2007) 309–313
311
and shape by the first derivative of the reflectivity, dis-
played on top by the dotted line. The electroreflectance
spectrum thus reveals a rigid redshift of the exciton and
its vibronic satellites by a quadratic Stark effect. The shift
in a field of 15 kV/cm is about 8 leV, similar to that
observed in other polydiacetylene crystals [15,18]. The
derivative spectrum does not reproduce the large signal
above 2.7 eV which increases sublinearly at high fields. This
feature arises from the Franz–Keldysh (FK) effect of the
exciton continuum [15,19]. The large signal at small fields
indicates the excellent quality of the investigated area since
it can be observed only if the scattering rate is small enough
to allow a free electron hole pair to gain significant energy
in an external field [15,20].
Kramers–Kronig transformation yields, with an accu-
racy better than 5 meV, the exciton transition energy,
2.171 eV and the free carrier gap, 2.727 eV, indicating an
exciton binding energy Eb ꢀ 556 meV (see Table 1). This
is about 10% larger than that obtained in bulk blue PDA
crystals with exciton below 2 eV, but similar to that of
red bulk crystals like TCDU [21]. The derivative lineshape
of the ER spectrum is compatible with coupling of the exci-
ton by the electric field F to its continuum of ionized states
with the mean square displacement r of electron and hole
representing the transition dipole [15]
could not be deduced. Its amplitude is nevertheless of the
same order than in bulk DCHD crystals [15], indicating
highly ordered THD chains along which coherent motion
occurs.
3.2. Absorption and luminescence
Thermal reactivity of the THD crystal is so high that the
polymer content at the end of the sample growth precludes
absorption in parallel polarization to be measured. We
thus compare in Fig. 2 the experimental absorption for a
polarization of the field perpendicular to the chain axis,
a^, to the calculated absorption obtained by Kramers–
Kronig transformation of the reflectance shown in Fig. 1.
a^ reaches only 100 cmꢁ1 for a value of xp estimated to
be 2–3%, which confirms the strong anisotropy by more
than 2 orders of magnitude of p–p* absorptions. The larger
linewidth of the exciton peak at high energy in parallel
polarization is due to the rapid decrease of the refractive
index n above the exciton resonance and is absent in spec-
tra of the dielectric constant, Im(e) = nca/x. The arrow
indicates the energy gap which, due to strong Coulomb
coupling of electron and hole, does not have sufficient
strength to be resolved in absorption under the excitonic
progression.
The luminescence spectrum is well resolved and is nearly
a mirror image of the absorption. The inhomogeneous
broadening of the zero phonon line is ꢀ40 meV at 10 K,
comparable to the absorption linewidth. It is affected by
re-absorption and cannot be considered as an origin for
the emission. This origin is recovered by adding the energy
of ground states normal modes measured by Raman spec-
troscopy to the corresponding vibronic lines in the emis-
sion. Using the C@C and C„C stretching modes values
2
~
~
jer ꢂ F j
DE ¼
:
ð1Þ
Eexc ꢁ Eg
Inserting the experimental values yields an exciton radius
of 14 A, slightly larger than that derived for the blue chains
˚
of DCHD [15]. The exciton ground state thus extends over
5 diacetylene repeat units and must be delocalized on the
chain. ‘Blue’ and ‘red’ excitons are intrinsically similar
states, with internal structure quasi-independent of the
electronic structure of the chain. Note that presently, due
to interfering contributions of band edges corresponding
to vibronic exciton replica to the FK signal, a reliable
determination of the coherence length of continuum states
Table 1
Transition energies of the exciton and its satellites, taken from spectra of
Im(e) derived from the electroreflectance at low temperature
Assignment
E
DE
(meV)
Raman
(298 K)
Raman
(15 K)
(eV)
Exciton E
2.171
2.175
–
–
5 (0)
–
–
–
a
b
c
4.7
6.7
5.0
6.0
19.1
–
–
19.2
d
S
2.226
2.260
2.276
2.327
2.356
2.439
2.727
54 (7)
89 (2)
105 (5)
156 (4)
185 (2)
267 (7)
556
54.8
86.9
106.1
153.2, 149.9
184.1
261.7
55.2
87.5, 87.0, 86.2
106.3
153.3, 149.8
183.8
261.5
D*
D
T
Fig. 2. Absorption spectra with polarization parallel (Kramer–Kronig
transform) and perpendicular (experimental) to the polymer chain, derived
from the reflectance and transmission spectra, respectively. The lumines-
cence at 10 K (dots) is for a 488.0 nm continuous wave excitation and a
sub-microwatt pump fluence.
Bandgap
DE refers to the energy difference to the exciton and yields the vibrational
energies in good agreement with Raman modes measured at room tem-
perature [9] and low temperature (this work).

312
T. Barisien et al. / Chemical Physics Letters 444 (2007) 309–313
Fig. 3. Typical low temperature Raman spectrum of a THD crystal
having low polymer content (xp 6 3%). The excitation power is 10 mW at
647.1 nm. The frequency of the double bond and triple bond stretching
modes are, respectively, 1481 2 cmꢁ1 and 2107 2 cmꢁ1. For a detailed
discussion of the attribution of the other lines, see [9].
Fig. 4. Luminescence spectra at 11.3 K (solid line), 16.0 K (dots) and
44.0 K (dashes) excited at 488.0 nm for a polarization of excitation parallel
to the chains. Inset: Arrhenius plot of the luminescence integrated over the
566–800 nm range (squares) and the associated linear fit.
However, the poly-THD exciton energy is lower by
ꢀ150 meV than that of red poly-3BCMU, and its Raman
frequencies are also intermediate between those of red
and blue poly-3BCMU.
(Fig. 3) places the origin at 2.153 eV, that is ꢀ18 meV
below the exciton energy deduced from electroreflectance.
This Stokes shift cannot be associated to diffusion of exci-
tons to low energy sites since ER data imply a highly reg-
ular chain similar to poly-3BCMU red chains. The
presence of such sites would indeed suggest a disordered
state of the chains, which is inconsistent with the observa-
tion of the FK effect in the field modulated signals. It thus
corresponds to a real shift between the emission and
absorption origins. No such Stokes shift is observed in
the red phase of highly diluted poly-3BCMU chains [22]
and its origin in poly-THD is under study.
The emission of crystals with low polymer content is
surprisingly strong. Given that the geometry of the collect-
ing optics does not differ from one experiment to the other,
it is possible to compare the emission intensities of poly-
THD and poly-3BCMU isolated chains, the yield of which
is known [23]. A rough estimate of the quantum yield, gF,
leads to gF(THD) >0.6 at 10 K. An extremely fast drop in
the luminescence intensity with increasing temperature is
also observed; this phenomenon follows an Arrhenius-type
law with a thermal activation energy Eact ꢀ 4.3 meV
(Fig. 4). This corresponds to the opening of a non radiative
channel. The origin of this process is not known yet but it
seems to be a generic property of the red structure. Isolated
poly-3BCMU chains indeed show a similar behaviour
above 50 K, with a much higher activation energy,
Eact ꢀ 38 meV [24].
The blue chain is planar and the red one cannot be so
[25]. A model in which successive repeats units are alterna-
tively tilted out of the average plane by +h and ꢁh angles is
reasonable, and furthermore agrees well with NMR results
[6,7]. The unit cell of the red chain thus contains two non-
translationaly equivalent monomer units. In poly-3BCMU,
side-groups are linked by hydrogen bonds which certainly
influence the value of h. In poly-THD there are not such
bonds and successive repeat units differ in the packing
(C–Hꢂ ꢂ ꢂp interactions) of phenyl groups. This may lead
to a different h value.
Simple considerations show that increasing h decreases
conjugation, hence increases the energy of the excited
states. The value of h in poly-THD would then be smaller
than in poly-3BCMU. Different excited states shift differ-
ently, and the occurrence of fluorescence may correspond
to a much faster shift of the ‘dark’ Ag states which are pres-
ent in the optical gap of blue chains [3], so that in red ones
they lay above the optical gap.
The activated decay of fluorescence yields may thus be
associated with an activated transfer of the excitonic pop-
ulation to this Ag state. However, the quenching mecha-
nism can also imply a virtual state positioned slightly
above the gap and corresponding to the dissociation
threshold of a singlet exciton in two triplets, as observed
in blue chains [26]. In the latter case, delayed luminescence
should result associated to the long lifetime of the triplets.
Time resolved luminescence experiments will give an
insight into the nature of the process.
4. Discussion and conclusions
The present results can be compared to those obtained
on red and blue poly-3BCMU chains. At the lowest tem-
perature, poly-THD is more fluorescent than these red
chains, and much more than any other PDA red phase.
Poly-THD chains thus form quantum wires, the exci-
tonic properties of which, on the whole, compare well with

T. Barisien et al. / Chemical Physics Letters 444 (2007) 309–313
313
[7] H. Tanaka, M.A. Gomez, A.E. Tonelli, M. Thakur, Macromolecules
22 (1989) 1208.
[8] V. Enkelmann, G. Schleier, Acta. Crystallogr. B 36 (1980) 1954.
[9] M.E. Morrow, K.M. White, C.J. Eckhardt, D.J. Sandman, Chem.
Phys. Lett. 140 (1987) 263.
[10] N.V. Agrinskaya, E.G. Guk, I.A. Kudryavtsev, O.G. Lyublinskaya,
Phys. Solid State 37 (1995) 526.
[11] A.S. Hay, J. Org. Chem. 27 (1962) 3320.
[12] L. Brandsma, Preparative Acetylenic Chemistry, second edn., Else-
vier, Amsterdam, 1988.
those of isolated poly-3BCMU red chains. The observation
of an intense Franz-Keldysh effect in the electroreflectance
spectra is the signature of a mixed monomer–polymer crys-
talline phase of ‘red’ nature possessing a high degree of
order. The latter appears as a promising model system of
1D crystals, in particular for the investigation, through
time-resolved non-linear optical spectroscopy, of how the
electronic structure is affected by the conformations in
the frame of strongly interacting electrons.
[13] B.N. Ghose, Synth. React. Inorg. Met. Org. Chem. 24 (1994) 29.
[14] A. Sharifi, M. Mirzaei, M.R. Naimi-Jamal, J. Chem. Res. (S) (2002)
628.
Acknowledgements
[15] G. Weiser, Phys. Rev. B 45 (1992) 14076.
´
[16] T. Guillet, J. Berrehar, R. Grousson, J. Kovensky, C. Lapersonne-
We wish to thank Dr. J.-F. Mohrange for his assistance
in Raman Scattering experiments. G. Weiser thanks the
INSP for its hospitality. This work was funded in part by
the Agence Nationale de la Recherche through Contract
ANR-06-NANO-013-02.
Meyer, M. Schott, V. Voliotis, Phys. Rev. Lett. 87 (2001) 087401.
[17] A. Horvath, G. Weiser, C. Lapersonne-Meyer, M. Schott, S.
Spagnoli, Phys. Rev. B 53 (1996) 13507.
[18] Y. Tokura, T. Koda, A. Itsubo, M. Miyabashi, K. Okuhara, A.
Ueda, J. Chem. Phys. 99 (1986) 85.
[19] G. Weiser, A. Horvath, in: N.S. Sariciftci (Ed.), Primary Photoex-
citations in Conjugated Polymers, World Scientific, Singapore, 1997,
Chapter 12.
References
[20] A. Jaeger, G. Weiser, Phys. Rev. B 58 (1998) 10674.
[21] Y. Tokura, Y. Oowaki, T. Koda, Chem. Phys. 88 (1984)
437.
´
[1] F. Dubin, J. Berrehar, R. Grousson, T. Guillet, C. Lapersonne-
Meyer, M. Schott, V. Voliotis, Phys. Rev. B 66 (2002) 113202.
[2] F. Dubin, R. Melet, T. Barisien, R. Grousson, L. Legrand, M. Schott,
V. Voliotis, Nature Phys. 2 (2006) 32.
´
´
[22] R. Lecuiller, J. Berrehar, C. Lapersonne-Meyer, M. Schott, Phys.
Rev. Lett. 80 (1998) 4068.
´
`
´
[23] R. Lecuiller, J. Berrehar, C. Lapersonne-Meyer, M. Schott, J.D.
[3] B. Kraabel, M. Joffre, C. Lapersonne-Meyer, M. Schott, Phys. Rev. B
58 (1998) 15 777.
Ganiere, Chem. Phys. Lett. 314 (1999) 255.
´
´
`
[24] R. Lecuiller, J. Berrehar, J.D. Ganiere, C. Lapersonne-Meyer,
P. Lavallard, M. Schott, Phys. Rev. B 66 (2002) 125205.
[25] M. Schott, J. Phys. Chem. B 110 (2006) 15864.
[26] B. Kraabel, D. Hulin, C. Aslangul, C. Lapersonne-Meyer, M. Schott,
Chem. Phys. 227 (1998) 83.
[4] J.-Y. Bigot, T.-A. Pham, T. Barisien, Chem. Phys. Lett. 259 (1996)
469.
[5] Y. Yuasa, M. Ikuta, T. Kobayashi, Phys. Rev. B 72 (2005) 134302.
[6] A.L. Cholli, D.J. Sandman, W. Maas, Macromolecules 32 (1999)
4444.