Spectral dependence of nonlinear optical properties of symmetrical octatetraynes with p-substituted phenyl end-groups

Article abstract of DOI:10.1039/c5cp01661d

Organic compounds containing conjugated carbon chains have been extensively investigated due to their interesting properties including nonlinear optical response. Polyynes are a group of compounds where the conjugation can be modified by appropriate selec

Full text of DOI:10.1039/c5cp01661d

PCCP
View Article Online
View Journal | View Issue
PAPER
Spectral dependence of nonlinear optical
properties of symmetrical octatetraynes with
p-substituted phenyl end-groups†
Cite this: Phys.Chem.Chem.Phys.,
2015, 17, 13680
Agata Arendt,a Radosław Kołkowski,bc Marek Samocb and Sławomir Szafert*a
Organic compounds containing conjugated carbon chains have been extensively investigated due to
their interesting properties including nonlinear optical response. Polyynes are a group of compounds
where the conjugation can be modified by appropriate selection of end-groups, enabling ‘‘control’’ and
improvement of nonlinear effects. In this work, we investigate three newly synthesized aryl end-capped
octatetraynes, exhibiting strong nonlinear absorption and nonlinear refraction properties, which can be
attributed to the presence of aryl end-groups with electron-withdrawing functional groups suitable for
further extending the conjugation. Nonlinear optical measurements were performed using a f-scan
(modified Z-scan) technique with a femtosecond laser system at various wavelengths in the visible and
near-infrared range (540–1600 nm), revealing strong negative third-order nonlinear refraction (3NR) and
two-photon absorption (2PA) below 600 nm and a noticeable three-photon absorption (3PA) at longer
wavelengths. The results of spectroscopic characterization and the crystallographic data for the
investigated compounds are also presented.
Received 22nd March 2015,
Accepted 15th April 2015
DOI: 10.1039/c5cp01661d
www.rsc.org/pccp
stable graphene-like structures.8 Stabilization of polyynes often
involves the use of bulky end-groups, preventing the conjugated
Introduction
Carbon-rich compounds have been extensively explored during chains from approaching each other. The synthesis of polyynes
the last few decades. One class of such species is polyynes – with a variety of end-groups has been widely discussed in the
compounds with a linear carbon chain consisting of alternating literature. The first phenyl end-capped oligoyne was obtained
triple and single bonds between carbon atoms,1,2 which can by Glaser in 1869.9 To date, the syntheses of polyynes with
serve as a model for carbyne (–CRC–)n, a linear sp-hybridized aryl,10–12 bulky ‘super trityl’,13 organometallic,14,15 and silyl
allotrope of carbon that still remains a hypothetical material.3,4 end-groups16,17 have been reported.
Oligoynes and polyynes are foreseen to be of interest for
Third-order nonlinear optical (NLO) properties of polyynes
numerous applications, for instance, as components of wires have been extensively studied by means of theoretical model-
or switches in optoelectronic devices5,6 or as components of ling, addressing in particular the dependence of cubic hyper-
co-crystals with conducting polymers.7
polarizability g on the length of acetylenic chains. It has been
One of the main issues associated with such conjugated found that increasing the number of repeated units in polyynes
organic systems is their thermodynamic instability resulting does not improve the NLO properties as much as in a similar
from the tendency of the chains to cross-link, forming more series of polyenes.18,19 Other computational studies concerned
the effect of terminal groups, showing that phenyl end-capped
a Faculty of Chemistry, University of Wroclaw, Joliot-Curie 14, Wroclaw 50-383,
Poland. E-mail: slawomir.szafert@chem.uni.wroc.pl; Fax: +48 71 328 2348;
Tel: +48 71 375 7122
b Advanced Materials Engineering and Modelling Group, Faculty of Chemistry,
Wroclaw University of Technology, Wybrzeze Wyspianskiego 27, 50-370 Wroclaw,
Poland. E-mail: Marek.Samoc@pwr.edu.pl; Fax: +48 71 320 3364;
polyynes exhibit superior g values compared to other end-
groups.20 These findings suggested that moderately long aryl
end-capped polyynes may be an optimal design concerning a
compromise between the NLO properties, molecular weight of a
molecule and its chemical stability.
Tel: +48 71 320 4466
Experimental confirmation of the theoretical predictions
has been challenging due to the stability issues associated with
polyynes. To the best of our knowledge, only several size-selective
experimental NLO studies on organic polyynes with various end-
groups and chain lengths were successfully conducted. Tykwinski
and co-workers studied the nonlinear properties of triisopropylsilyl
c
´
´
Laboratoire de Photonique Quantique et Moleculaire, Ecole Normale Superieure de
´
Cachan, 61 Avenue du President Wilson, 94230 Cachan, France
† Electronic supplementary information (ESI) available: Preparation, NMR spec-
tra, X-ray details and higher order NLO effects. CCDC 1042757 and 1042758. For
ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/
c5cp01661d
13680 | Phys. Chem. Chem. Phys., 2015, 17, 13680--13688
This journal is ©the Owner Societies 2015

View Article Online
Paper
PCCP
(TIPS)16,17 and phenyl21 end-capped polyynes using a differential Experimental section
optical Kerr effect technique (DOKE), confirming the higher NLO
Detailed synthetic procedures and characterization data for the
investigated compounds can be found in the ESI.† UV-vis
spectra were recorded on a Cary 300 Bio spectrophotometer.
The 1H and 13C NMR spectra were obtained using a Bruker
Avance 500 MHz spectrometer. HRMS spectra were recorded
using a spectrometer with a TOF mass analyzer and an ESI ion
source. X-Ray diffraction data were collected on a Xcalibur
diffractometer with a Ruby CCD camera (o scan technique)
equipped with an Oxford Cryosystem–Cryostream cooler at
100 K. The space groups were determined from systematic
absences and subsequent least-squares refinement. Lorentz
and polarization corrections were applied. The structures were
solved by direct methods and refined by full-matrix, least-
squares on F2 using the SHELXTL package.26 Crystallographic
data for the structures of E1–C8–E1 and E2–C8–E2 were deposited
with cambridge crystallographic data centre as supplementary
publication no. CCDC 1042757 and 1042758, respectively.
Nonlinear optical properties were studied using the newly
developed f-scan technique,25 which is very similar to the well-
known Z-scan,27 except that the position of the sample (z) is
fixed, whereas the focal length (f) of an electrically focus-
tunable lens (Optotune EL-10–30 series) is changed. The light
source used in f-scan measurements was a femtosecond laser
system, tunable over the visible and near-infrared spectral range,
consisting of an optical parametric amplifier (OPA, Quantronix
Palitra) pumped by a Quantronix Integra Ti:Sapphire regenera-
tive amplifier (output wavelength 800 nm, pulse duration 130 fs,
repetition rate 1 kHz). Femtosecond pulsed light source plays a
crucial role in the determination of NLO parameters, removing
any possible contribution of parasitic effects occurring at different
time scales, such as molecular reorientation which in the case of
anisotropically polarizable molecules may become a dominant
process responsible for nonlinear refraction measured with pico-
or nanosecond pulses.28 The use of the f-scan technique signifi-
cantly shortened the time of a single measurement compared to
the Z-scan, minimizing the exposure of our potentially unstable
compounds to the high-power pulsed laser radiation.23 Moreover,
the total duration of the whole series of measurements was
reduced, lowering the risk of long-term degradation of com-
pounds. A detailed description of the f-scan technique, including
the scheme of the experimental setup and theoretical analysis,
can be found in a recent publication.25
performance of the phenyl end-capped molecules. An unprofit-
ability of the polyynic chain extension has been found in the
experimental studies of organometallic ruthenium-terminated
polyynes22 and chromophore-terminated polyynes and polyenes,19
performed by degenerate four-wave mixing (DFWM), however,
in these studies, the NLO activity was attributed more to the
end-groups rather than polyynic chains. Very recently, NLO
properties of a group of phenyl-end-capped polyynes were
investigated using a nanosecond Z-scan at the wavelength of
532 nm,23 pointing out some additional phenomena such as
molecular reorientation which can contribute to the unexpectedly
large g values observed when the measurements were performed
using pulsed light sources of longer than picosecond pulse
durations.
While in most cases the research has been directed towards
the investigation of the chain length dependence of g, spectrally
resolved studies have been rarely addressed. Such studies are
crucial for the correct interpretation of the results, providing
better insights into molecular energy levels participating in
multiphoton transitions. Some results obtained only at a single
wavelength, such as the astonishingly rapid increase of the
g values with increasing acetylenic chain length in silyl end-
capped polyynes16,17 could be better explained in terms of a red
shift of the resonantly enhanced multiphoton transition bands,
which has been observed in the wavelength dependence of the
cubic nonlinearity of platinum-terminated polyynes, measured
by Samoc et al. using the Z-scan technique.24 Moreover, inves-
tigation of the full spectral dependence may reveal higher-order
NLO effects, which in the case of polyynes have not been
investigated yet.
In this work, we report on the synthesis of three new
octatetraynes with electron withdrawing ester-substituted phenyl
end-groups and present the results of spectrally-resolved investi-
gation of NLO properties of these compounds, performed using
the femtosecond f-scan technique.25 The choice of the com-
pounds seemed highly rational since no symmetrical systems
with electron withdrawing substituents have yet been selected
for NLO measurements. Ester-substituted polyynes appeared
reasonably stable and considering the enhancement of con-
jugation in such systems it was expected to give interesting
results. Indeed, the measurements have revealed negative non-
linear refraction in the visible range and multiphoton absorp-
tion in the visible and near-infrared range. The observed values
of g are in the order of 10ꢀ32 esu, outperforming polyynes
of similar size investigated under comparable experimental
conditions.24 These large values of g can be attributed to the
Results and discussion
Synthesis
presence of the electron-withdrawing groups causing further The synthetic route for the preparation of the octatetraynes
extension of the conjugation. Additionally, we quantitatively E1–C8–E1, E2–C8–E2, and E3–C8–E3 is outlined in Scheme 1.
describe the three-photon absorption (3PA) in the studied In the first step, the Sonogashira coupling between bromo-
polyynes, which is, to the best of our knowledge, performed or iodo-substituted aryl esters was carried out yielding silyl-
for the first time for these kinds of organic compounds. We protected phenylacetylenes. In the next step, 1-bromoacetylenes
have also characterized the compounds using 1H, 13C NMR, (E1–E3–C2Br) were obtained with the use of NBS (N-bromo-
and UV-vis spectroscopy and HRMS. Solid-state structures for succinimide) and a AgNO3/KF reaction system. The carbon chain
two of the studied compounds were also obtained.
elongation was attained via Cadiot–Chodkiewicz cross-coupling
This journal is ©the Owner Societies 2015
Phys. Chem. Chem. Phys., 2015, 17, 13680--13688 | 13681

View Article Online
PCCP
Paper
Scheme 1 Syntheses of octatetraynes.
using TMSC2H in the presence of a Pd(PPh3)2Cl2/CuI catalyst
and NH(iPr)2 as a base.29 The cross-coupling reactions were
carried out in THF for 2.5–4 hours. The key step was the oxidative
dimerization of trimethylsilyl-protected butadiynes (E1–E3–C4TMS)
that led to the aryl end-capped octatetraynes. The products
E1–C8–E1 and E2–C8–E2 were obtained via Hay–Glaser homo-
coupling30 with in situ deprotection using a CuCl/TMEDA
(N,N,N0,N0-tetramethylethylenediamine) catalytic system in the
presence of TBAF as the deprotection agent, TMSCl (chlorotri-
methylsilane) as the Fꢀ ion scavenger, and O2 as an oxidant.31
The reactions were carried out in dry acetone for 20–22 hours. If
the reaction time was too short, only products of deprotection
of E1–E3–C4TMS were observed. Our attempts to obtain the
compound E3–C8–E3 via Hay–Glaser coupling failed and there-
fore Eglinton homocoupling of deprotected butadiyne was per-
formed yielding the target octatetrayne. Workup on flash column
chromatography gave pure products with acceptable yields as
presented in Scheme 1.
Fig. 1 Molecular structure of E1–C8–E1.
Crystal structures of octatetraynes
Fig. 2 Molecular structure of E2–C8–E2.
The crystal structures of E1–C8–E1 and E2–C8–E2 were deter-
mined and crystallographic data are presented in the ESI.† The
monocrystals of E1–C8–E1 were obtained by slow evaporation
of its acetone solution while E2–C8–E2 was crystallized from
1.363(2) Å for the single bonds C2–C3 and C4–C4A, and they are
similar to those in E2–C8–E2 and other tetraynes.2,31,32 The
polyyne carbon chains are only a little distorted from linearity
with low contraction coefficients as described in Table 1.
%
DCM (dichloromethane). E1–C8–E1 crystallized in the P1 space
group, whereas E2–C8–E2 in the C2/c space group. In both cases
only half of the molecule resides in the asymmetric unit and the
whole molecule is generated by symmetry operations.
Packing motifs and reactivity implications
As presented in Fig. 1 and 2, the molecules of E1–C8–E1 and
%
E2–C8–E2 are almost ideally planar (the angles between the The two structures crystallize in triclinic (P1, E1–C8–E1) and
phenyl rings are 01 for both compounds due to the symmetry), monoclinic (C2/c, E2–C8–E2) systems. As a consequence,
with carbon atoms from the chain being no further than molecules of E1–C8–E1 form one set of parallel chains, while
0.160 and 0.112 Å out of the planes of phenyl substituents for E2–C8–E2 are packed to form two non-parallel sets of parallel
E1–C8–E1 and E2–C8–E2, respectively. Such geometry creates chains forming an angle of 13.91. Fig. 3 and 4 show the packing
very favourable conditions for a strong conjugation across the diagrams for E1–C8–E1 and E2–C8–E2.
entire molecule. In the structure of E1–C8–E1 the bond lengths
The closest chain–chain separation distance was then ana-
within the carbon chain are 1.2060(17) Å and 1.2108(17) Å for lyzed. As defined earlier,2 we approximate the closest chain–
the triple bonds C1RC2 and C3RC4, and 1.3631(17) Å and chain distance as the closest carbon–carbon distance from two
13682 | Phys. Chem. Chem. Phys., 2015, 17, 13680--13688
This journal is ©the Owner Societies 2015

View Article Online
Paper
PCCP
Table 1 Interatomic distances and angles for E1–C8–E1 and E2–C8–E2a
Table 2 Packing parameters for E1–C8–E1 and E2–C8–E2
E1–C8–E1
E2–C8–E2
Compound
E1–C8–E1
E2–C8–E2
Interatomic distances [Å]
C11–C1
C1–C2
Chain–chain contact (Å, parallel)
f (1)2
3.854
67.8
1.456
0.12
—
3.863
83.0
0.471
0.04
13.9
3.742
1.4340(16)
1.2060(17)
1.3631(17)
1.2108(17)
1.363(2)
1.4537(14)
1.2084(14)
1.431(2)
1.201(2)
1.362(2)
1.207(2)
1.360(3)
1.3585(18)
1.1940(19)
Offset distance (Å)2
C2–C3
Fractional offset
Angle between nonparallel chains (1)
Chain–chain contact (Å, nonparallel)
C3–C4
C4–C4Ai,ii
O1–C17
—
O2–C17
The 1,n-topochemical polymerization is one of the most
important transformations that can take place in the case of
organic polyynes.34 The geometric requirements for such a
process for tetraynes (1,8-polymerization) are fulfilled when
the nearest parallel chains are separated by ca. 3.5 Å and the
angle f is close to 211. Searching for such values, it can be
easily noticed (see Table 2) that there is no good candidate
for 1,8-polymerization among the investigated compounds and
both structures present more ladder type architectures (f =
67.81 and 83.01; for ideal ladder type f = 901) with potential for
a 1,2 process.
C11–C11A dist
C11–C11A sum
% Contraction
C1–C1A dist
C1–C1A sum
% Contraction
11.757
11.791
0.29
8.913
8.923
0.11
11.746
11.762
0.14
8.895
8.900
0.06
Interatomic angles [1]
C2–C1–C11
C1–C2–C3
175.08(12)
177.07(12)
176.48(13)
179.49(17)
176.48(13)
112.73(10)
124.04(11)
—
177.11(17)
177.71(17)
177.20(17)
179.5(2)
123.44(15)
119.95(12)
—
C4–C3–C2
C3–C4–C4Ai,ii
O2–C17–O1
C17–O1–C14
O2–C17–C14
O2–C17–C18
Av. angle
126.40(15)
177.88
UV-vis spectra
177.03
The UV absorption spectra of the octatetraynes are presented
in Fig. 5 and summarized in Table 3. Samples were prepared
by dissolving the compounds in DCM with concentrations
presented in Table 4. The UV-vis spectra were recorded in the
225–800 nm range (at intervals of 1 nm), however, the absorp-
tion bands are observed only up to 425 nm.
a
Symmetry codes for E1–C8–E1 (i) ꢀx + 2, ꢀy + 1, ꢀz; E2–C8–E2 (ii) ꢀx + 1,
ꢀy + 1, ꢀz.
The most intense absorption peaks (lmax) are observed
within the 275–335 nm wavelength region. Narrow absorption
peaks observed near 350, 370 and 400 nm are very similar for all
the compounds due to the analogous length of the polyyne chain.
Fig. 3 Packing diagram for E1–C8–E1 with the shortest chain–chain
distance of 3.854 Å for C2B–C2D.
Fig. 4 Packing diagram for E2–C8–E2 with the shortest chain–chain
distances for parallel and non-parallel neighbors. The distances [Å] are
C2B–C3AD, 3.780; C3B–C3AD, 3.742; and C1AD–C1AB, 3.863.
neighboring carbon chains. Accordingly, the nearest chains
with a parallel orientation for E1–C8–E1 are only 3.854 Å apart
and this separation is slightly larger than a sum of the van der
Waals radii (3.56 Å) and is analogous to that found for
TMS(CRC)4TMS (3.853 Å).33
Fig. 5 UV-vis extinction spectra of octatetraynes.
Table 3 Selected spectral data for octatetraynes
Despite different space groups, the shortest distance for
E2–C8–E2 is almost identical (3.863 Å) to that of E1–C8–E1,
which is the consequence of a similar size of the end-group.
Additionally, for E2–C8–E2, the chain–chain distance between
non-parallel chains is even shorter than that between parallel
chains and reaches 3.742 Å.
Region 1
lmax [nm], e [Mꢀ1 cmꢀ1
Region 2
lmax [nm], e [Mꢀ1 cmꢀ1
]
]
E1–C8–E1
E2–C8–E2
E3–C8–E3
306 (120 000)
290 (150 000)
293 (110 000)
407 (30 000)
401 (27 000)
402 (21 000)
This journal is ©the Owner Societies 2015
Phys. Chem. Chem. Phys., 2015, 17, 13680--13688 | 13683

View Article Online
PCCP
Paper
Table 4 Sample concentration for UV-vis spectra and nonlinear optical
measurements
UV-vis spectra
Nonlinear optical measurements
E1–C8–E1
E2–C8–E2
E3–C8–E3
5.1 ꢁ 10ꢀ6
4.1 ꢁ 10ꢀ6
4.6 ꢁ 10ꢀ6
M
M
M
4.4 ꢁ 10ꢀ2
4.8 ꢁ 10ꢀ2
1.0 ꢁ 10ꢀ2
M
M
M
Such spectra are characteristic of aryl-end-capped polyynes.10,12
The highest molar absorptivity was observed for E2–C8–E2
(e = 150 000 Mꢀ1 cmꢀ1 for l = 290 nm) and the values of the
molar absorptivities are comparable to the octatetraynes with
other end-groups.21
Nonlinear optical measurements
The f-scan measurements were performed at different wave-
lengths in the spectral range of 540–1600 nm. All the com-
pounds were dissolved in DCM forming transparent solutions
of relatively high concentrations, as specified in Table 4. The
solution of E3–C8–E3 could be obtained only with a slightly
lower concentration due to some stability and solubility problems.
Each investigated solution was placed in a 1 mm-thick glass
cuvette, and the transmitted intensity of a focused laser beam
passing through the cuvette was recorded simultaneously in
closed aperture and open aperture configurations as a function
of focal length f (which was then recalculated for z, that is, the
position of the sample with respect to the focus).
In addition to the cuvettes containing solutions of octa-
tetraynes in DCM, which will be referred to as samples, each
measurement also included three references: (1) an empty
space (required for a closed aperture f-scan), (2) a 4.66 mm-
thick silica glass plate and (3) a cuvette containing pure solvent
(DCM). The first reference is used to ‘‘correct’’ the closed
aperture f-scan curves, so that the line-shape resembles the
usual Z-scan data.25 The second and third references are needed
to calibrate the light intensity (making use of the known nonlinear
refractive index of fused silica) and to calculate the nonlinear
optical parameters referring to the pure solute35 (pure octa- Fig. 6 Dispersion of cubic hyperpolarizability g, real (filled circles) and
imaginary parts (empty circles), of the investigated octatetraynes, obtained
tetrayne). In addition, each open aperture f-scan curve of the
sample is divided by the open aperture curve of the solvent
(third reference), in order to remove any possible contribution
using the f-scan. For wavelengths above 620 nm the Im(g) values also
account for the higher order nonlinear effects.
of the nonlinear absorption of the solvent, as well as to get rid
of the artefacts which may be present in both samples and exhibit strong third-order NLO properties; 2PA and negative
solvent measurements.
3NR are observed in the 540–580 nm spectral range, which
The experiment at a given wavelength consisted of a sequence roughly correspond to the doubled wavelength of the peaks in
of three consecutive measurements, and the series of measure- the main one-photon UV absorption band. It needs to be
ments at different wavelengths in the 540–1600 nm range was mentioned here that due to the centrosymmetric character of
repeated after several days under slightly different experimental the compounds, exact correspondence between the one-photon
conditions (different laser power and beam diameter). The error and two-photon spectra is not expected, the one-photon transi-
of the determined NLO parameters is estimated based on the tions being allowed to states of ungerade character and the two-
statistics, depending on the repeatability of the results (when the photon transitions to gerade states. However, because of the
results were poorly repeatable, the error was larger).
presence of a large number of possible electronic excited states
The values of the real and imaginary parts of cubic hyper- as well as the possible deviations from centrosymmetry, the
polarizability g, corresponding to third-order nonlinear refrac- rough coincidence of the appropriate wavelength ranges is not
tion (3NR) and two-photon absorption (2PA), respectively, are unexpected. A more detailed understanding of the 2PA mecha-
plotted in Fig. 6. All the investigated compounds are found to nisms would require identification of the electronic transitions
13684 | Phys. Chem. Chem. Phys., 2015, 17, 13680--13688
This journal is ©the Owner Societies 2015

View Article Online
Paper
PCCP
E1–C8–E1, ꢀ4 ꢁ 10ꢀ33 esu for E2–C8–E2, and ꢀ1.4 ꢁ 10ꢀ32 esu
for E3–C8–E3. At the same time, the magnitude of Im(g) varies
in the range from 1 to 5 ꢁ 10ꢀ34 esu, which is one order of
magnitude lower than Re(g). Based on these results one may
expect that organic compounds similar to the investigated ones
could be promising for efficient optical switching, which requires
the merit factor |Re(g)/Im(g)| to be greater than 4p.37 The highest
values of this factor are found at 580 nm – |Re(g)/Im(g)| 4 20 in
E1–C8–E1 and E2–C8–E2, and |Re(g)/Im(g)| 4 50 in E3–C8–E3.
Molecular engineering of such compounds, aiming at optimiza-
tion of the NLO performance in the useful (telecom) spectral
range is a promising challenge for future research work.
Although a direct comparison between the values of the real
part of the second hyperpolarizability g obtained using different
techniques such as Z-scan, DOKE and DFWM may be erro-
neous, we find it appropriate to set together and discuss the
values obtained in our work and the values determined by other
research groups for similar compounds. It is done in Table 5,
taking into account polyynes of the same number of triple
bonds (4) as well as the longest polyynes studied in a given work
(with an exception of ref. 23, where polyynes up to 4 triple
bonds were studied). Our results fall in between the relatively
small magnitudes of g reported in ref. 16, 17 and 21, and
the significantly larger values reported in ref. 22 and 23. As
already mentioned, low g values of silyl and phenyl end-capped
polyynes measured at 800 nm are probably a result of an off-
resonant character of the NLO phenomena at this wavelength.
On the other hand, results presented in ref. 22 and 23 were
obtained using nanosecond light sources, which make the NLO
measurements susceptible to contributions of parasitic pheno-
Fig. 7 Representative f-scan traces of the investigated octatetraynes at
560 nm: experimental data (empty circles) fitted using theoretical curves
(solid lines). Left: closed aperture curves involving negative 3NR accom-
panied by 2PA. Right: open aperture curves exhibiting optical limiting
behaviour due to 2PA. The asymmetric shape of the curves (compared
to traditional Z-scan) is a peculiar feature of the f-scan and is fully
accounted for by the theoretical model.
involved in this process, which could be possible by quantum mena such as molecular reorientation. In view of the above
chemical calculations supported by appropriate analytical incompatibility issues, our results can be directly compared
tools.36
only with the work presented in ref. 24. Indeed, the values of g
The values of the relevant NLO parameters were extracted by appear comparable, and the polyynes investigated in our work
fitting the experimental data (silica, solvent and samples) using seem to slightly outperform compounds of similar polyynic
appropriate theoretical curves based on the f-scan modification chain lengths terminated by organometallic moieties.
of the theory proposed by Sheik-Bahae et al.27 Representative
In the wavelength range of 600–680 nm the results were more
closed aperture and open aperture f-scan traces at 560 nm for ambiguous, large discrepancies being found between different
all three investigated samples are shown in Fig. 7, together with series of measurements, leading to relatively large errors of the
the theoretical fits.
determined third-order NLO parameters. Analysis of the results
As for organic compounds of relatively low molecular weight
(394.42 for E1–C8–E1, 366.37 for E2–C8–E2 and 490.50 for
E3–C8–E3), the magnitudes of the obtained NLO parameters
can be considered as very high. The effect of negative 3NR in all
studied octatetraynes in the spectral range of 540–580 nm was
so strong that it surpassed the positive 3NR of the solvent and
the cuvette glass, causing inversion of the closed aperture curve
(exchange of the peak and valley positions along the z axis).
For example, the nonlinear phase shift DF0 measured in the
sample E1–C8–E1 (cuvette containing DCM solution of this
compound) was equal to ꢀ5.24 rad, whereas the values of
DF0 equal to 0.88 rad and 1.34 rad were obtained under exactly
the same experimental conditions as the silica plate and the
cuvette with pure DCM, respectively. Taking into account the
concentration of the solutions, the values of Re(g) at the spectral
maximum (around 560–580 nm) exceed ꢀ5 ꢁ 10ꢀ33 esu for
Table 5 Comparison of Re(g) values of different polyynes reported in the
literature and the present work
Triple End-
bonds group |g| ꢁ 1035 esu l [nm] Solvent
Technique Ref.
4
10
4
8
4
8
2
4
4
Pr3Si 1.25 ꢂ 0.21
800
THF
DOKE
DOKE
16 and 17
64.6 ꢂ 2.7
Ph
Ru2
Ph
Pt
4.9 ꢂ 1.8
800
THF
21
58.8 ꢂ 3.6
25 000 ꢂ 4000 532
18 000 ꢂ 3000
Toluene DFWM
Methanol Z-scan
22
17 000
532
590
23
130 000
320 ꢂ 80
DCM
DCM
Z-scan
f-scan
24
12
4
4
7500 ꢂ 1200 520
‘E1’
‘E2’
‘E3’
520 ꢂ 70
400 ꢂ 30
1400 ꢂ 300
580
This work
4
This journal is ©the Owner Societies 2015
Phys. Chem. Chem. Phys., 2015, 17, 13680--13688 | 13685

View Article Online
PCCP
Paper
led to the conclusion that part of the discussed spectral range is
probably a region of spectral overlap and an interplay between
the NLO phenomena of different order. In particular, both two-
and three-photon (3PA) processes may contribute to the resultant
nonlinear absorption behaviour of the samples. This makes the
values of effective cubic hyperpolarizability (g) dependent on
the laser intensity, in a manner that is difficult to predict, as
the relative contribution of different nonlinear absorption
processes cannot be easily assessed.
In addition, the possible instability of the samples cannot be
excluded – one should be concerned that chemical changes,
such as cross-linking of the polyynic chains, could have occurred
during the measurements, affecting to some extent the NLO
properties.23 However, we presume that such effects are rather
weak, as the NLO parameters in the 540–580 nm range were
rather well reproducible during different series of measure-
ments. The occurrence of 2PA and 3PA is summarized in Fig. 8,
which shows the dispersion of multiphoton absorption cross-
sections s(2) and s(3) (filled and empty squares, respectively)
against the one-photon absorption spectra plotted as a function
of doubled and tripled wavelengths (grey solid and dashed
lines, respectively). 3PA becomes dominant in the open aper-
ture f-scan curves at 640 nm, which is shown in Fig. 9(a) in the
case of E2–C8–E2. At wavelengths above 640 nm 2PA is no
longer observed, whereas 3PA is clearly seen up to 825 nm,
occurring in the spectral region which corresponds to the tripled
wavelengths of the UV-vis absorption region 225–275 nm. An
example of the open aperture curve featuring 3PA, measured in
E1–C8–E1 solution at 800 nm is presented in Fig. 9(b).
The values of 3PA cross-sections were calculated directly
from the 3PA absorption coefficient of the solution (a3),
obtained by fitting open aperture data using 3PA theoretical
curves based on the f-scan modification of the 3PA formula
given by Gu et al.38 It is also possible to calculate s(3) from the
(2)
relation: s(3) = Ese(2ff)/I0, where s is the value of 2PA cross-
eff
section obtained from fitting the open aperture data using 2PA
theoretical curves, E = hc/l is the photon energy and I0 is the Fig. 8 Dispersion of 2PA and 3PA absorption cross-sections s(2) and s(3)
(black and white squares, respectively), superimposed with UV-vis spectra
laser intensity obtained from a closed aperture measurement of
plotted as a function of doubled and tripled wavelengths (grey solid and
the silica reference. Such calculations give nearly the same
dashed lines). The values of s(2) may be affected by the contribution of 3PA
values of s(3). Each open aperture experimental curve at the
above 580 nm. At longer wavelengths, the values of s(2) and s(3) may be
wavelengths of 640 nm and above was fitted using both 2PA
affected by the contribution of 4PA (above 825 nm), and additionally by
and 3PA, assuming contribution of only one of these pheno- 5PA (above 1000 nm).
mena during each fitting. Both s(3) and effective s(2) values were
calculated in the spectral region dominated by 3PA, so that the
relative strength of both optical limiting phenomena can be obtained in our experiments above 640 nm are fitted separately
directly compared.
using 2PA and 3PA theories, and possible contributions of higher-
Fig. 8 reveals the spectral regions, where two different NLO order absorption processes are accounted for by effective Im(g) in
absorption processes are simultaneously present, resulting in a Fig. 6 (calculated from 2PA fits), and by effective cross sections s(2)
large error due to variations of the multiphoton cross-sections and s(3) in Fig. 8. These values are exact in the regions where only
obtained at a given wavelength under different laser intensities. one particular kind of nonlinear absorption is clearly dominant –
Such effects are visible not only in 2PA values in the range of s(2) at 540–580 nm and s(3) at 640–825 nm. Outside these regions
600–680 nm, but also in the range of 850–950 nm, where they should be used only for indicative description of the optical
another process – four-photon absorption (4PA) – is probably limiting strength under assumption of a pure contribution of the
involved together with 3PA. Because the analysis of higher- given nonlinear absorption process. Continuation of the discussion
order NLO phenomena is associated with very large uncertainties, on higher-order phenomena, also including the observation of
we refrain from going beyond 3PA. All of the open aperture curves higher-order nonlinear refraction, can be found in the ESI.†
13686 | Phys. Chem. Chem. Phys., 2015, 17, 13680--13688
This journal is ©the Owner Societies 2015

View Article Online
Paper
PCCP
Notes and references
1 (a) F. Cataldo, Polyynes: Synthesis Properties and Applications,
CRC Press Taylor & Francis Group, Boca Raton, 2006;
(b) R. J. Lagow, J. J. Kampa, H.-C. Wei, S. L. Battle, J. W.
Genge, D. A. Laude, C. J. Harper, R. Bau, R. C. Stevens,
J. F. Haw and E. Munson, Science, 1995, 267, 362–367.
2 (a) S. Szafert and J. A. Gladysz, Chem. Rev., 2003, 103,
4175–4205; (b) S. Szafert and J. A. Gladysz, Chem. Rev.,
2006, 106, PR1–PR33.
Fig. 9 Examples of the open aperture f-scan curves revealing the 3PA
phenomenon.
3 F. Cataldo, Polym. Int., 1997, 44, 191–200.
4 F. Diederich, P. J. Stang and R. R. Tykwinski, Acetylene
Chemistry; Chemistry, Biology and Materials Science, Wiley-
VCH, Weinheim, 2005.
5 (a) K. West, C. Wang, A. S. Batsanov and M. R. Bryce, Org.
Biomol. Chem., 2008, 6, 1934–1937; (b) A. Ricci, M. Chiarini,
C. Lo Sterzo, R. Pizzoferrato and S. Paoloni, J. Photochem.
Photobiol., A, 2014, 298, 1–8.
6 J. Cornil, D. Beljonne, J.-P. Calbert and J.-L. Bredas, Adv.
Mater., 2001, 13, 1053–1067.
7 A. Sun, J. W. Lauher and N. S. Goroff, Science, 2006, 312,
1030–1034.
8 G. Eres, C. M. Rouleau, M. Yoon, A. A. Puretzky, J. J.
Jackson and D. B. Geohegan, J. Phys. Chem. C, 2009, 113,
15484–15491.
9 C. Glaser, Ber. Dtsch. Chem. Ges., 1869, 2, 422–424.
10 S. M. E. Simpkins, M. D. Weller and L. R. Cox, Chem. Commun.,
2007, 4035–4037.
11 M. Gulcur, P. Moreno-Garcia, X. Zhao, M. Baghernejad,
A. S. Batsanov, W. Hong, M. R. Bryce and T. Wandlowski,
Chem. – Eur. J., 2014, 20, 4653–4660.
12 T. Gibtner, F. Hampel, J.-P. Gisselbrecht and A. Hirsch,
Chem. – Eur. J., 2002, 8, 408–432.
Conclusions
The investigated compounds exhibit very strong negative third-
order nonlinear refraction in the 540–580 nm spectral range,
with peak values of Re(g) in the order of ꢀ5 and ꢀ4 ꢁ 10ꢀ33 esu
in compounds E1–C8–E1 and E2–C8–E2, respectively, and ꢀ1.4 ꢁ
10ꢀ32 esu in compound E3–C8–E3. In all compounds the non-
linear refraction is accompanied by two-photon absorption
with the maximum cross-section s(2) varying from B100 GM
in E2–C8–E2 up to B200 GM in E1–C8–E1 and E3–C8–E3,
corresponding to Im(g) between 2.5 and 4.5 ꢁ 10ꢀ34 esu. These
results are comparable with the data for the compound PtC8Pt
in ref. 24 where peak values of Re(g) = ꢀ3.2 ꢁ 10ꢀ33 and Im(g) =
7 ꢁ 10ꢀ34 esu were found in the vicinity of 580 nm, not very
different from the values in the present report. The high values
of the |Re(g)/Im(g)| merit factor obtained for the studied
compounds confirm the potential suitability of the polyyne
structural motif for optical switching. At longer wavelengths
the samples exhibit complex nonlinear absorption behaviour,
involving various multiphoton phenomena, with three-photon
absorption being a dominant process in the range of 640–
825 nm, showing the maximum cross-section s(3) at 640 nm
in the order of 5 ꢁ 10ꢀ79 cm6 s2 in E2–C8–E2, and up to 1 ꢁ
10ꢀ78 cm6 s2 in E1–C8–E1 and E3–C8–E3.
The present results suggest that polyynes remain a very
interesting group of organic compounds from the point of
view of their NLO properties, especially that, apparently, even
small molecules of such type – of molecular mass not exceeding
500 g molꢀ1 – can exhibit quite large values of the NLO
parameters. Further investigation of these compounds, including
spectrally resolved NLO measurements and theoretical analysis of
the electronic transitions involved in the NLO mechanisms, such
as those presented in ref. 36, may allow for the rational design of
polyyne molecules suitable for photonic applications in the useful
spectral regions.
13 W. A. Chalifoux and R. R. Tykwinski, Nat. Chem., 2010, 2,
967–971.
14 (a) W. Mohr, J. Stahl, F. Hampel and J. A. Gladysz, Chem. –
Eur. J., 2003, 9, 3324–3340; (b) N. J. Long and C. K. Williams,
Angew. Chem., Int. Ed., 2003, 42, 2586–2617; (c) T. Bartik,
B. Bartik, M. Brady, R. Dembinski and J. A. Gladysz, Angew.
Chem., Int. Ed. Engl., 1996, 35, 414–417; (d) R. Dembinski,
T. Bartik, B. Bartik, M. Jaeger and J. A. Gladysz, J. Am. Chem.
Soc., 2000, 122, 810–822.
15 Z. Cao, B. Xi, D. S. Jodoin, L. Zhang, S. P. Cummings, Y. Gao,
S. F. Tyler, P. E. Fanwick, R. J. Crutchley and T. Ren, J. Am.
Chem. Soc., 2014, 136, 12174–12183.
16 S. Eisler, A. D. Slepkov, E. Elliott, T. Luu, R. McDonald,
F. A. Hegmann and R. R. Tykwinski, J. Am. Chem. Soc., 2005,
127, 2666–2676.
17 A. D. Slepkov, F. A. Hegmann, S. Eisler, E. Elliott and R. R.
Tykwinski, J. Chem. Phys., 2004, 120, 6807–6810.
18 J.-W. Song, M. A. Watson, H. Sekino and K. Hirao, J. Chem.
Phys., 2008, 129, 024117.
19 B. B. Frank, P. R. Laporta, B. Breiten, M. C. Kuzyk, P. D.
Jarowski, W. B. Schweizer, P. Seiler, I. Biaggio, C. Boudon,
J.-P. Gisselbrecht and F. Diederich, Eur. J. Org. Chem., 2011,
4307–4317.
Acknowledgements
RK and MS would like to acknowledge the financial support
from the Foundation for Polish Science (WELCOME grant
‘Organometallics in NanoPhotonics’) and the National Science
Centre Poland (NCN grant UMO-2013/10/A/ST4/00114). SS
would like to thank the National Science Centre (Grant number
UMO-2013/08/M/ST5/00942) for financial support.
This journal is ©the Owner Societies 2015
Phys. Chem. Chem. Phys., 2015, 17, 13680--13688 | 13687

View Article Online
PCCP
Paper
20 J.-W. Song, M. A. Watson, H. Sekino and K. Hirao, Int.
J. Quantum Chem., 2009, 109, 2012–2022.
21 T. Luu, E. Elliott, A. D. Slepkov, S. Eisler, R. McDonald, F. A.
Hegmann and R. R. Tykwinski, Org. Lett., 2005, 7, 51–54.
22 G.-L. Xu, C.-Y. Wang, Y.-H. Ni, T. G. Goodson III and T. Ren,
Organometallics, 2005, 24, 3247–3254.
Ed. Engl., 1992, 31, 1101–1123); (d) R. R. Tykwinski and
Y. Zhao, Synlett, 2002, 1939–1953; (e) Y. Xu, M. D. Smith,
M. F. Geer, P. J. Pellechia, J. C. Brown, A. C. Wibowo and
L. S. Shimizu, J. Am. Chem. Soc., 2010, 132, 5334–5335;
( f ) J. Xiao, M. Yang, J. W. Lauher and F. W. Fowler, Angew.
Chem., 2000, 112, 2216–2219 (Angew. Chem., Int. Ed., 2000,
39, 2132–2135); (g) Y. Takeoka, K. Asai, M. Rikukawa and
K. Sanui, Chem. Commun., 2001, 2592–2593; (h) V. Enkelmann,
Chem. Mater., 1994, 6, 1337–1340; (i) V. Enkelmann, Adv.
Polym. Sci., 1984, 63, 91–136; ( j) J. L. Foley, L. Li,
D. J. Sandman, M. J. Vela, B. M. Foxman, R. Albro and
C. J. Eckhardt, J. Am. Chem. Soc., 1999, 121, 7262–7263 and
23 E. Fazio, L. D’Urso, G. Consiglio, A. Giuffrida, G. Compagnini,
`
O. Puglisi, S. Patane, F. Neri and G. Forte, J. Phys. Chem. C,
2014, 118, 28812–28819.
24 M. Samoc, G. T. Dalton, J. A. Gladysz, Q. Zheng. Y. Velkov,
H. Ågren, P. Norman and M. G. Humphrey, Inorg. Chem.,
2008, 47, 9946–9957.
´
references cited therein; (k) F. Carre, N. Devylder,
25 R. Kolkowski and M. Samoc, J. Opt., 2014, 16, 125202.
26 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr.,
2008, A64, 112–122.
27 M. Sheik-Bahae, A. A. Said, T.-H. Wei, D. J. Hagan and
E. W. Van Stryland, IEEE J. Quantum Electron., 1990, 26,
760–769.
28 J. G. Breitzer, D. D. Dlott, L. K. Iwaki, S. M. Kirkpatrick and
T. B. Rauchfuss, J. Phys. Chem. A, 1999, 103, 6930–6937.
29 W. Chodkiewicz, Ann. Chim., 1957, 2, 819–869.
30 A. S. Hay, J. Org. Chem., 1962, 27, 3320–3321.
31 N. Gulia, K. Osowska, B. Pigulski, T. Lis, Z. Galewski and
S. Szafert, Eur. J. Org. Chem., 2012, 4819–4830.
32 I. Deperasinska, A. Szemik-Hojniak, K. Osowska, M. F.
´
S. G. Dutremez, C. Guerin, B. J. L. Henner, A. Jolivet,
V. Tomberli and F. Dahan, Organometallics, 2003, 22,
2014–2033; (l) H. Irngartinger and M. Skipinski, Tetrahedron,
2000, 56, 6781–6794; (m) J. R. Neabo, K. I. S. Tohoundjona
and J.-F. Morin, Org. Lett., 2011, 13, 1358–1361; (n) S. R.
Diegelmann, N. Hartman, N. Markovic and J. D. Tovar,
J. Am. Chem. Soc., 2012, 134, 2028–2031; (o) Y. Nemoto
and M. Sano, J. Phys. Chem. B, 2011, 115, 12744–12750;
(p) J. Nagasawa, M. Yoshida and N. Tamaoki, Eur. J. Org.
Chem., 2011, 2247–2255.
35 M. Samoc, A. Samoc, B. Luther-Davies, Z. Bao, L. Yu,
B. Hsieh and U. Scherf, J. Opt. Soc. Am. B, 1998, 15, 817–825.
´
Rode, A. Szczepanik, Ł. Wisniewski, T. Lis and S. Szafert, 36 (a) M. Sun, J. Chen and H. Xu, J. Chem. Phys., 2008,
J. Photochem. Photobiol., A, 2011, 217, 299–307.
33 B. F. Coles, P. B. Hitchcock and D. R. M. Walton, J. Chem.
Soc., Dalton Trans., 1975, 442–445.
128, 064106; (b) J. Liu, J. Xia, P. Song, Y. Ding, Y. Cui,
X. Liu, Y. Dai and F. Ma, ChemPhysChem, 2014, 15,
2626–2633.
34 (a) G. Wegner, Z. Naturforsch., B: J. Chem. Sci., 1969, 24, 37 B. Luther-Davies and M. Samoc, Curr. Opin. Solid State
824–832; (b) U. H. F. Bunz, Y. Rubin and Y. Tobe, Chem. Soc. Mater. Sci., 1997, 2, 213–219.
Rev., 1999, 28, 107–119; (c) F. Diederich and Y. Rubin, 38 B. Gu, J. Wang, J. Chen, Y.-X. Fan, J. Ding and H.-T. Wang,
Angew. Chem., 1992, 104, 1123–1146 (Angew. Chem., Int. Opt. Express, 2005, 13, 9230–9234.
13688 | Phys. Chem. Chem. Phys., 2015, 17, 13680--13688
This journal is ©the Owner Societies 2015