Content uploaded by Anna De Girolamo Del Mauro
Author content
All content in this area was uploaded by Anna De Girolamo Del Mauro
Content may be subject to copyright.
Electroluminescence and fluorescence
emission of poly(n-vinylcarbazole) and
poly(n-vinylcarbazole)-Irppy3-based
organic light-emitting devices prepared
with different solvents
Annalisa Bruno
Anna De Girolamo Del Mauro
Giuseppe Nenna
Maria Grazia Maglione
Saif A. Haque
Carla Minarini
Electroluminescence and fluorescence emission of
poly(n-vinylcarbazole) and poly(n-vinylcarbazole)-
Irppy3-based organic light-emitting devices
prepared with different solvents
Annalisa Bruno,a,b Anna De Girolamo Del Mauro,aGiuseppe Nenna,a
Maria Grazia Maglione,aSaif A. Haque,band Carla Minarinia
aItalian National Agency for New Technologies, Energy and Sustainable Economic (ENEA),
Portici Research Center, p.le E. Fermi 1, 80055 Portici, Italy
annalisa.bruno@enea.it
bImperial College London, Chemistry Department, South Kensington Campus,
London SW7 2AZ, United Kingdom
Abstract. We present a study of spectroscopic proprieties of poly(n-vinylcarbazole) (PVK) and
PVK doped with iridium complexes tris[2-phenylpyridinato-C2,N]iridium(III) (IrðppyÞ3) films
prepared by spin-coating from toluene and chlorobenzene solutions. A different molecular
organization of the polymer on the substrate during the spin-coating process can be produced
using solvents with different boiling temperatures. The modified molecular rearrangement
affects the emission properties of the PVK material and the consequent energy transfer to
the doping molecules. Both static and dynamic fluorescence emissions properties have been
studied, for pure PVK and PVK doped with different weight percentage of IrðppyÞ3.
Different organic light-emitting devices, using a simple architecture, have been prepared
with both solvents to test the change in electroluminescence spectral shape and in electrical
characteristic, and the final efficiencies of the devices have been evaluated. ©2013 Society of
Photo-Optical Instrumentation Engineers (SPIE). [DOI: 10.1117/1.JPE.3.033599]
Keywords: organic light-emitting devices; spectroscopy; electroluminescence.
Paper 12047P received Jul. 25, 2012; revised manuscript received Nov. 28, 2012; accepted for
publication Nov. 29, 2012; published online Jan. 15, 2013.
1 Introduction
In the last few years there has been a fast-growing interest in developing efficient organic
light-emitting devices (OLEDs).1Despite OLEDs having reached the commercialization
level, there is still a need for materials development with regard to efficiency, color purity,
and stability. Phosphorescent dyes dopant into charge transporting hosts as emissive layer
have attracted intensive attention due to the ability to produce highly efficient emissions
compared with conventional fluorescent OLEDs.2,3Through radiative recombination of both
singlet and triplet excitons, the internal quantum efficiency of the phosphorescent OLEDs
can reach 100% (Refs. 4and 5).
Small molecules and polymers are currently the preferred candidates. Polymers are generally
of lower purity than small molecules but can access full colour and larger display sizes at much
lower costs using solution-based deposition techniques.
Previous works have exhibited high brightness devices using a blue emitting polymer such as
poly(n-vinylcarbazole) (PVK) as a host material to which an organometallic complex, like
Iridium complexes tris[2-phenylpyridinato-C2,N]iridium(III) (IrðppyÞ3), is added as a dopant.6
The heavy metal atom at the center of these complexes exhibits strong spin-orbit coupling, facili-
tating intersystem crossing between singlet and triplet states. By using these phosphorescent
materials, both singlet and triplet excitons will be able to decay radiatively, hence improving
0091-3286/2013/$25.00 © 2013 SPIE
Journal of Photonics for Energy 033599-1 Vol. 3, 2013
the internal quantum efficiency compared to a standard OLED where only the singlet states will
contribute to emission of light.
In this work, pure PVK and PVK doped with different concentrations of IrðppyÞ3solutions
were prepared using two organic solvents (toluene and chlorobenzene), and the corresponding
thin films were obtained by spin-coating on glass and quartz substrates. This system has been
previously widely investigated being promising combination for efficient green emission.7–9
Previously these two solvents have shown to be the best in dissolving PVK polymer.7The effect
of using solvents with different boiling temperature on different organization of the polymer on
the substrate is discussed. The modified molecular rearrangement affects the emission properties
of the PVK material and the consequent energy transfer to the doping molecules has been
analyzed.
Two green phosphorescent OLED devices, with a simple structure, were fabricated and their
performances characterized.
2 Experimental
2.1 Materials
IrðppyÞ3is a green emitter widely used in phosphorescent polymer OLEDs, while PVK is a
polymer host with large band gap and a triplet state energy above that of the phosphorescent
dye to guarantee the confinement of the triplet excited state on the guest.
PVK (average Mn: 25,000 to 50,000), IrðppyÞ3, and the organic solvents (toluene and
chlorobenzene) were purchased from Aldrich Co. and used as receivers. The chemical structures
and the energy levels (HOMO and LUMO) of PVK and IrðppyÞ3are reported in Fig. 1(a) and
1(b), respectively.
The thin films of PVK matrix were prepared from a 10 mg∕ml PVK solution in toluene, and
one at the same concentration in chlorobenzene.
Various amount of IrðppyÞ3(1%, 3%, 5%, 8% in weight) were dispersed in the two solutions
of PVK, and films with thicknesses of about 80 nm were obtained by spin-coating in air on
quartz, and glass substrates varying the spinning conditions (from 1500 to 800 rpm). The
substrates have been precleaned by sonication in acetone and subsequently isopropanol. All
the preparation process was done in clean room, in air, at room temperature.
OLEDs device structures were fabricated on commercially 130-nm-thick indium tin
oxide (ITO) (sheet resistance <20 Ω∕sq), with the device structure: ITO/poly(3,4 ethylenedioxy-
thiophene):poly(styrenesulfonate) (PEDOT:PSS) ð50 nmÞ∕PVK∕IrðppyÞ3ð80 nmÞ∕calcium
ð50 nmÞ∕aluminum (Al) (150 nm) cathode
2.2 Setups
The absorption and emission spectra were performed using, respectively, a single-beam
spectrophotometer, Lambda 900 Perkin Elmer, and a spectrofluorimeter, Fluorolog FL-3-22
Fig. 1 (a) Chemical structures and (b) HOMO LUMO levels of PVK and IrðppyÞ3.
Bruno et al.: Electroluminescence and fluorescence emission of poly(n-vinylcarbazole). . .
Journal of Photonics for Energy 033599-2 Vol. 3, 2013
Horiba Jobin Yvon, both commercially available instruments. Roughness was measured by
means of a contact Alpha-TENCOR profilometer.
Time-resolved fluorescence measurements were carried out using a diode laser-based
time-correlated single-photon-counting (TCSPC) spectrometer from IBH (UK). A subnanosec-
ond LED (1-MHz repetition rate) with time width of 500 ps was used as excitation source.
A MCP-PMT detector was used for collecting the fluorescence decay measurements. For life-
time measurements, the fluorescence decays were recorded at the magic angle (54.7 deg) with
respect to the vertically polarized excitation light. All the characterizations were carried out in
ambient air.
The electrical characteristics of the devices were characterized through I through V measure-
ments by using a Keithley 2400 power-supply source meter keeping a constant increment
steps and delay time of 1 s between each measurement point. An integrating sphere and a
photodiode (Newport 810UV) connected to a Keithley 6517A electrometer were employed
for the electroluminescence (EL) analysis, while the electroluminescence spectra were evaluated
through a Ocean-Optics USB4000 spectrometer.
3 Results
3.1 Pure PVK
PVK films prepared with toluene (Tol) and chlorobenzene (CB) have been spectroscopic char-
acterized through ultraviolet (UV)-visible absorption and emission measurements: The spectra
are reported in Fig. 2(a) and 2(b), respectively.
The absorption spectra are similar for the Tol and CB films studied, showing a peak absorp-
tion in the UV and slowly decreasing around 400 nm and a being negligible in the visible. The
emission spectra (excitation wavelength 280 nm) show a broadening and a red shift effect for the
CB film. These effects could suggest already a possible different arrangement of the polymer on
the substrate7–10 respect to the Tol film.
The roughness of the two films has been studied using a profilometer with a vertical
resolution of 1 nm. The roughness has been measured of 30 2nmfor the Tol film and 10
2nmfor the CB one, indicating that the PVK can arrange differently on the substrate when it is
spun with the different solvents.
In order to better investigate the spectral change of the two films, we have studied the time
behavior of the fluorescence signal for different emission wavelengths inside the emission band
for the Tol and CB films using TCSPC. Figure 3reports the time decays of the fluorescence
collected for various emission wavelengths for the two films. Mainly two aspects can be noted.
The first is that the lifetimes of the fluorescence signals became systematically longer going from
the UV and progressively to the visible range for both the films. The second is that, for CB film,
Fig. 2 (a) Absorption spectrum of Tol and CB films of pure PVK; (b) steady-state emission
fluorescence of Tol and CB films, with excitation wavelength at 280 nm, and detection in the
360- to 650-nm range.
Bruno et al.: Electroluminescence and fluorescence emission of poly(n-vinylcarbazole).. .
Journal of Photonics for Energy 033599-3 Vol. 3, 2013
the emission lifetime is always slightly longer in respect to the one of the emission of the Tol film
at the same wavelength, and this effect gets more evident at longer wavelengths. This can suggest
that the smoother morphology assumed from the polymer on the CB films could be more effi-
cient in transfer energy to a possible dopant.10 A longer lifetime could correspond to a larger
interaction radius for the exciton generated into the material.11
A slower decay time and a broader emission spectrum for the PVK CB films, together with a
more flat substrate, could already suggest the possibility that the CB prepared film could be more
efficient in transferring energy to the dopants.
3.2 Adding Irðppy Þ3to PVK Matrix
The good overlap between the PVK emission and IrðppyÞ3absorption spectra is reported in
Fig. 4. This is already an indication of a good energy transfer between the matrix and the dopant.
According to Forster energy-transfer theory, it is possible to calculate the energy transfer critical
Fig. 3 Fluorescence time decays for detection various wavelengths: (a) 375, (b) 400, (c) 425,
(d) 450, and (e) 475 nm for Tol (black squares) and CB (green circles) films.
Fig. 4 PVK emission (blue circles) and IrðppyÞ3(violet circles) absorption spectrum excited at
282 nm.
Bruno et al.: Electroluminescence and fluorescence emission of poly(n-vinylcarbazole). . .
Journal of Photonics for Energy 033599-4 Vol. 3, 2013
radius R0, which is the distance between a host and a guest at which the efficiency of energy
transfer is 50%, from the relative overlap of the PVK emission and the IrðppyÞ3absorption
spectra.12 Indeed the critical radius was about 4.2 nm for our IrðppyÞ3doped PVK system
and Forster/singlet energy transfer from PVK to IrðppyÞ3is expected to be around 5%.
The emission spectra of the Tol and CB PVK films with different concentrations of IrðppyÞ3
are reported in Fig. 5(a) and 5(b), respectively. It is interesting to note that for both solvents and
for all the investigated concentrations the spectra are double peaked around 500 and 570 nm. In
the Tol films, the spectral variation as function of concentration seems to not follow a clear trend.
For the CB films, indeed it is clear that increasing the dopant concentration in the matrix the
emission spectra is progressively more red shifted and also the relative intensity of the peak at
570 nm increase respect to the one at 500 nm.
The roughness of the two films has been studied also in this case, showing a similar trend
with respect to the film prepared with pure PVK. In this case an average roughness of 20 2nm
for the Tol film and 52nmfor the CB one has been measured.
In order to clarify the differences between the two emission spectra, the dynamic of
fluorescence signal has been temporally analyzed for a fixed-emission wavelength of
550 nm. In this work we want to study the effects of the energy transfer from the matrix to
the dopant, monitoring the changes in the decays of the fluorescence emission of the dopants.
Previous works have studied the quenching of the fluorescence signal of the PVK as function of
the dopant concentrations, monitoring the PVK emission peak (371 nm).13 At this wavelength,
the photoluminescence decay time of IrðppyÞ3doped PVK thin films is systematically shorter
than that of pure PVK thin films (for all the dopants concentrations analyzed), indicating that
Forster energy transfer in IrðppyÞ3doped PVK thin films is taking place.
In Fig. 6(a) and 6(b), the fluorescence decays are shown for different concentrations of IrðppyÞ3
in the PVK matrix for both Tol and CB samples, respectively, where the excitation wavelength was
360 nm and detection 550 nm. The decay times get slower, increasing the IrðppyÞ3concentrations
for both set of samples. For the Tol PVK∕IrðpyyÞ3film the decays show a “saturation”effect for
concentration higher than 5% without a real trend with concentration [see also Fig. 6(a)]. The
saturation effect showed for the toluene PVK∶IrðpyyÞ3film fluorescence lifetime [Fig. 6(a)]is
well correlated also with the steady photoluminescence saturation [Fig. 5(a)]. In both cases,
3% concentration is the maximum in photoluminescence broadening and the maximum in the
lifetime, and there is not a real trend versus the concentration.
Moreover the CB films show a monotonic increase of fluorescence lifetime with increasing
doping concentrations. A similar monotonic increase has been observed for the photolumines-
cence peak at 570 nm [Fig. 5(b)]. These different trends for the two solvents can be explained
also thinking to a better dopant dispersion using the CB solvent.
In particular the dynamics are generally always much slower for CB films similarly to what
has been observed also for the pure PVK-CB films (see Fig. 3). In this case, we are dealing with
the emission of phosphorescent dye excited through an energy-transfer process from the PVK
Fig. 5 Emission spectra of PVK films with concentrations of IrðppyÞ3varying from 1% to 8% for
(a) tol dispersed film and (b) CB dispersed films.
Bruno et al.: Electroluminescence and fluorescence emission of poly(n-vinylcarbazole).. .
Journal of Photonics for Energy 033599-5 Vol. 3, 2013
matrix, where the emission state population can be modified with the polymer arrangement. In
particular the different arrangement in the polymer chain in the presence of CB solvent could
affect the singlet-singlet transfer mechanism, because with less clustering on the surface, less
energy is lost in no-radiative processes. The subsequent intersystem crossing populate more
efficiently the triplet IrðppyÞ3emission state varying the time decay [Fig. 6(b)] and the emission
spectra [Fig. 5(b)].
3.3 PVK ∕Irðppy Þ38% Devices
OLED device structures using Tol and CB to spin the active layer were fabricated to investigate
the correlation between the emission proprieties in devices and to test the change in the transfer
mechanism when the active layer is prepared with the two solvents. We chose the 8% IrðppyÞ3
concentration that corresponds with the maximum of emission for the CB films and within the
saturation region for the Tol films.
OLED devices have been prepared according to the following structure: ITO ð130 nmÞ∕
PEDOT∶PSS ð50 nmÞ∕PVK∕IrðppyÞ3ð80 nmÞ∕calcium ð50 nmÞ∕aluminum (Al) (150 nm)
cathode. The schematic representation is reported in Fig. 7. On the glass ITO-coated substrate,
a PEDOT:PSS layer was spooned acting as an hole injection layer (HIL) and on top of which a
single layer of the active material (PVK∕IrðppyÞ3blends) has been deposited using solvents that
do not damage the HIL film. Finally, the calcium/aluminum cathode was deposited by thermal
evaporations under a high vacuum condition.
The electroluminescence and the electrical behavior have been characterized for both the Tol
and CB devices. Figure 8(a) shows that normalized electroluminescence spectra from the CB
device is broader and more red-shifted with respect to the Tol-prepared OLED, in agreement
with what suggested from the photoluminescence measurements [Fig. 5(b)].
Fig. 6 Fluorescence decays at 550 nm for PVK∕IrðppyÞ3films with concentrations varying from
1% to 8%, together with the response signal (prompt) at the scattering wavelength: (a) tol
dispersed film and (b) CB dispersed films.
Fig. 7 Scheme of the OLEDs device structure
Bruno et al.: Electroluminescence and fluorescence emission of poly(n-vinylcarbazole). . .
Journal of Photonics for Energy 033599-6 Vol. 3, 2013
Figure 8(b) reports the device current as function of the applied voltage to the device; it is
clear that in the Tol devices there is more current flowing. but the charges do not recombine
efficiently in to produce electroluminescence because of leakages inside the active layer
due to a less homogeneous structure. Indeed in Fig. 8(c), it is clear that the luminescence
from the CB device is one order of magnitude higher than the one prepared in Tol. Both
of these devices are not particularly luminescent (70 and 3Cd∕m2, respectively), and this
is due to the simple device structure used. Substantial improvements in the device performances
could be obtained using a more complex device structure and through codoping of hole-
blocking and hole-transport molecules as reported in Ref. 14. In this case, the luminescence
absolute values are interesting just as direct comparison of the effect of the morphology induced
from the different solvents on the final performances. In Fig. 8(d), the luminescence is also
reported as function of the current density in the device, to make more clear the better efficiency
of the CB devices respect to the toluene one.
From all the Fig. 8(a)–8(d), it is clear also that the CB-prepared device is sensibly more
efficient, around one order of magnitude, with respect to the one prepared in Tol. This effect
indicates that a better (smoother) arrangement of the polymer chain in presence of CB solvent
can guarantee more efficient emission states and prevent not only current leakages but also
nonradiative decay mechanisms confirming the assumptions made on the photoluminescence
spectra and the lifetime regarding Figs. 5and 6and from the morphology measurements.
4 Conclusions
In this work, we have studied the effects of using two different solvents such as toluene and
chlorobenzene in the preparation process of the active layer of OLED devices on static and
dynamic spectroscopic proprieties for pure PVK and PVK∕IrðpyyÞ3blends. The CB films
present always a smoother morphology with respect to the Tol films. It has been shown that
there are spectral changes and different dynamics for the films prepared with the two solvents.
Fig. 8 Toluene (black squares) and chlorobenzene (green circles) device electro-optical charac-
terization: (a) electroluminescence spectra of the device; (b) device current versus applied
voltage; (c) luminescence versus applied voltage; (d) luminescence versus device current density
for Tol and CB devices.
Bruno et al.: Electroluminescence and fluorescence emission of poly(n-vinylcarbazole).. .
Journal of Photonics for Energy 033599-7 Vol. 3, 2013
In particular, CB films show a broader spectra and a slower dynamic, with respect to the Tol
films. This effect has been observed both when the PVK matrix has been studied alone and on all
the blends compositions investigated.
This suggests that the arrangement of the PVK polymer in CB films is more favorable to
guarantee an efficient energy transfer from the matrix to the dopant.
Moreover, OLED devices have been realized using PVK∕IrðppyÞ3as active layer with
concentration of 8 wt.% of dopant using both solvents. The devices prepared using CB as
dissolving solvent for active layer preparation show a higher efficiency. Our study suggests
a clear correlation between the arrangement of the polymer chain using different solvents
and the energy-transfer mechanism.
References
1. E. L. Williams et al., “Excimer-based white phosphorescent organic light-emitting diodes
with nearly 100% internal quantum efficiency,”Adv. Mater. 19(2), 197–202 (2007),
http://dx.doi.org/10.1002/(ISSN)1521-4095.
2. M. A. Baldo et al., “Highly efficient phosphorescent emission from organic electrolumi-
nescent devices,”Nature 395(6698), 151–154 (1998), http://dx.doi.org/10.1038/25954.
3. M. A. Baldo et al., “Very high-efficiency green organic light-emitting devices based on
electrophosphorescence,”Appl. Phys. Lett. 75(1), 4–6 (1999), http://dx.doi.org/10.1063/
1.124258.
4. C. Adachi et al., “Nearly 100% internal phosphorescence efficiency in an organic light-
emitting device,”J. Appl. Phys. 90(10), 5048–5051 (2001), http://dx.doi.org/10.1063/1
.1409582.
5. M. A. Baldo, M. E. Thompson, and S. R. Forrest, “High-efficiency fluorescent organic
light-emitting devices using a phosphorescent sensitizer,”Nature 403(6771), 750–753
(2000), http://dx.doi.org/10.1038/35001541.
6. N. J. Lundin et al., “Synthesis and characterization of a multicomponent rhenium(I) com-
plex for application as an OLED dopant,”Angew. Chem. 45(16), 2582–2584 (2006), http://
dx.doi.org/10.1002/(ISSN)1521-3773.
7. K. M. Vaeth and C. W. Tang, “Light-emitting diodes based on phosphorescent guest/
polymeric host systems,”J. Appl. Phys. 92(7), 3447–3454 (2002), http://dx.doi.org/10
.1063/1.1501748.
8. C. L. Lee, K. B. Lee, and J. J. Kim, “Polymer phosphorescent light-emitting devices doped
with tris(2-phenylpyridine) iridium as a triplet emitter,”Appl. Phys. Lett. 77(15), 2280–2283
(2000), http://dx.doi.org/10.1063/1.1315629.
9. F. C. Chen et al., “Energy transfer and triplet exciton confinement in polymeric electrophos-
phorescent devices,”J. Pol. Sci. Part B: Polym. Phys. 41(21), 2681–2690 (2003), http://dx
.doi.org/10.1002/(ISSN)1099-0488.
10. M. K. Vaeth and J. DiCillo, “High-efficiency doped polymeric organic light-emitting
diodes,”J. Polym. Sci. Part B: Polym. Phys. 41(21), 2715–2725 (2003), http://dx.doi
.org/10.1002/(ISSN)1099-0488.
11. D. Gupta et al., “Polymer light-emitting diode using a new electrophosphorescent
cyclometalated iridium complex,”Mater. Manufact. Process. 21(3), 285–289 (2006),
http://dx.doi.org/10.1080/10426910500464693.
12. T. Tanaka, Experimental Methods in Polymer Science: Modern Methods in Polymer
Research and Technology (Polymers, Interfaces and Biomaterials), pp. 100–120,
Academic Press, London (1999)
13. P. Y. Liu et al., “Energy transfer in phosphorescent dye doped polymer thin films,”J. Kor.
Phys. Soc. 46, S66–S69 (2005), http://dx.doi.org/10.3938/jkps.46.66
14. X. Yang et al., “Highly efficient single-layer polymer electrophosphorescent devices,”
Adv. Mater. 16(2), 161–166 (2004), http://dx.doi.org/10.1002/(ISSN)1521-4095.
Biographies and photographs of the authors are not available.
Bruno et al.: Electroluminescence and fluorescence emission of poly(n-vinylcarbazole). . .
Journal of Photonics for Energy 033599-8 Vol. 3, 2013