M7 Electroluminescence of Polymers

Universität Potsdam Institute of Physics and Astronomy Advanced Physics Lab Course March 2020

M7 ELECTROLUMINESCENCE OF POLYMERS

I. INTRODUCTION

The recombination of holes and electrons in a luminescent material can produce light. This emitted light is referred to as electroluminescence (EL) and was discovered in organic single crystals by Pope, Magnante, and Kallmann in 1963.[1] EL from conjugated polymers was first reported by Burroughes et al.[2] The polymer used was poly(p-phenylenevinylene) (PPV). Since then, a variety of other polymers has been investigated. Organic EL devices have applications in a wide field ranging from multi-color displays and optical information processing to lighting. Polymers have the advantage over inorganic and monomolecular materials in the ease with which thin, structurally robust and large area films can be perpared from solutions. Using printing techniques, patterned structures can be produced easily. Even flexible displays can be produced because of the good mechanical properties of polymers. In this lab course, basic optical and electrical properties of conjugated polymers will be investigated. Advanced Lab Course: Electroluminescence of Polymers 2

EXPERIMENTAL TASKS

 Measure the absorption spectra of your polymers (thin films spin coated onto glass substrates).  Characterize the setup used for luminescence measurements. Identify possible sources of error and collect data necessary for their correction.  Measure the photoluminescence emission spectra for the polymer films, using suitable excitation wavelengths.  Measure the photoluminescence excitation spectra for the polymer films, using suitable detection wavelengths.  Measure the current through the OLEDs and the spectral radiant intensity of electroluminescence as a function of applied voltage (the current-radiance-voltage characteristics).  Measure the electroluminescence emission spectrum at a suitable voltage or current.  Measure the radiance of an OLED as a function of viewing angle.  Determine the radiometric and luminous efficiencies of the OLEDs.  Summarize your results.

CONTENTS

I. INTRODUCTION ...... 1 EXPERIMENTAL TASKS ...... 2 CONTENTS ...... 2 II. FUNDAMENTALS ...... 3 II.1 CONJUGATED POLYMERS ...... 3 II.2 LIGHT EMISSION BY CONJUGATED POLYMERS ...... 3 II.3 LIGHT-EMITTING DIODES ...... 6 II.4 POLYMERS FOR LIGHT-EMITTING DIODES ...... 12 III. EXPERIMENT EXECUTION AND ANALYSIS...... 14 III.1 GENERAL REMARKS ...... 14 III.2 ABSORPTION SPECTRA ...... 14 III.3 LUMINESCENCE SPECTRA ...... 15 III.4 ELECTROLUMINESCENCE ...... 16 III.5 SUMMARY OF EXPERIMENT ...... 18 IV. INSTRUMENTATION AND POSSIBLE SOURCES OF ERROR ...... 19 IV.1 CORRECTION OF SPECTRA ...... 19 IV.2 INNER FILTER EFFECTS ...... 20 V. REFERENCES AND FURTHER READING ...... 21 VI. RADIOMETRIC AND PHOTOMETRIC QUANTITIES ...... 22

Advanced Lab Course: Electroluminescence of Polymers 3

II. FUNDAMENTALS

II.1 CONJUGATED POLYMERS Organic electronics can be divided into two major classes of materials. One of them is the class of small organic molecules. Electroluminescence was first observed in crystals of the small molecule anthracene in 1961.[1] However, these materials are usually not soluble and have to be deposited by thermal evaporation. Hence the production of small molecular electronics needs expensive technology comparable to inorganic electronics. The second class of materials used are polymers. While most polymers are electrical insulators, for electronics applications, materials with semiconducting properties are needed. Conductivity needs the presence of free, mobile charges. In polymers, these can be provided by a conjugated 휋 electron system. This conjugated system is often depicted as an alternating chain of single and double bonds between the carbon atoms. However, quantum chemical calculations show that this picture is not correct, the electrons forming the double bonds are delocalized over the main chain. In a thin solid film, many chains are in close contact with each other (allowing orbital overlap between the chains), so that excess charges can travel over macroscopic distances in an electrical field. Table II.1 shows a selection of polymers which have been studied with respect to their semiconducting properties. While the polymers shown in Table II.1 possess a conjugated backbone, that is not generally necessary. Other polymers have been used which are built from an saturated backbone bearing conjugated side groups. The conjugated system can be easily disturbed by either conformational defects (e.g., kinks) or chemical impurities in the material. In addition, for entropic reasons it is very unlikely for a polymer chain to be completely extended. At the resulting geometric distortions, the conjugated system is interrupted. As a result, a polymer chain is not one single conjugated system but a chain of conjugated segments of different length. The segments are often called “chromophores” because they determine the optical and electronic properties of the materials. A polymer chain can be regarded as a chain of chromophores of different lengths and thus different properties. The length distribution can be regarded as being Gaussian.

II.2 LIGHT EMISSION BY CONJUGATED POLYMERS Absorption of energy by atoms, molecules or condensed matter will result in the generation of excited states, i.e. it increases the potential energy of the electrons in the substance rather than its heat. If the excited state decays under emission of visible light, that emission is called luminescence. Luminescence is observed from many inorganic and organic substances, the luminophores, and can be induced by various physical processes. Two possible excitation mechanisms relevant for this lab course are the absorption of photons, which leads to photoluminescence, or the recombination of injected charges, which is called electroluminescence. Luminescence spectra are determined by the properties of the material. In contrast, the thermal generation of light by heat radiation is called incandescence. It is mainly determined by the material’s temperature. The excited state is called an exciton, a coulombically bound electron-hole pair. In inorganic (crystalline) semiconductors, the binding energy of an exciton is on the order of or even below the thermal energy at room temperature (about 0.025 eV), such that the exciton will dissociate into free charges. In organic materials, the exciton is much stronger bound (about 0.5 eV) and will not dissociate easily. The generated excitons on polymer chromophores thus resemble the excited states of small molecules. Thus, absorption and (photo-)luminescence can be described in a molecular picture. To understand the optical properties of a thin polymer film, one needs to keep in mind that such films comprise a distribution of chromophores in close contact.

Advanced Lab Course: Electroluminescence of Polymers 4

Table II.1: Polymers which have been extensively studied for EL [3] PPV poly(p-phenylene vinylene) n

O MEH-PPV poly[2-methoxy-5-(2’-ethyl-hexyloxy)-1,4-phenylene vinylene]

n O OR1

PPE n poly[2,5-dialkoxy-1,4-phenylene ethynylene] R O 2 N PPy n poly(p-pyridine)

n PF poly(9,9-dialkylfluorene) R R R P3AT

n poly(3-alkylthiophene) S

PPP n poly(para-phenylene) C10H21

C H R 6 13 LPPP “ladder-type poly(para-phenylene)” n R=H, LPPP R H13C6 R=CH3 (methyl), Me-LPPP

H C 21 10 II.2.1 Photoluminescence In the following chapter, the numbers in brackets correspond to the numbers in Figure II.1. At room temperature, most chromophores are in the vibrational and electronic ground state S0,0: The occupation of higher vibrational states follows the Boltzmann distribution. But since typical vibrational energies for polymers are in the range of 100 meV, population of higher vibrational states in thermal equilibrium is very unlikely. In general, excitation (1) by absorbed energy is comprised of transitions S0,0 − S푛,휈′ into an electronically excited state 푛, accompanied by an additional vibrational excitation 휈′. The generated state decays at very short times by non- radiative transitions (2) to vibrational states of the first excited singlet state S1,휈′′ (internal conversion) and vibrational relaxation (3) to the lowest vibrational level S1,0. The radiative transition (4) S − S from the lowest electronically excited state to the electronic ground state 1,0 0,휈 is denoted as fluorescence. In general, the transition will not only decay to the ground state S0,0, but also to higher vibronic states S0,휈 of the singlet ground state. This gives rise to the vibronic progression observed in the fluorescence spectra of most conjugated systems. Similarly, electronic excitation usually involves transitions into more than one vibrational state. The radiative transition competes with nonradiative deactivation pathways. The first is isoenergetic internal conversion (not shown) from S1,0 to a high vibronic level of the electronic ground state followed by vibrational relaxation to S0,0. The second path is inter-system crossing, a non-radiative transition (5) from S1,0 to triplet states T푛,휈′′′with similar energy, followed by Advanced Lab Course: Electroluminescence of Polymers 5 vibrational relaxation (6) and possibly internal conversion steps to the lowest excited triplet state T1,0. Inter-system crossing involves a change of spin states from S = 0 for the singlet state to S = 1 for the triplet. There are three different ways for de-excitation of the triplet state, the first of which is repeated inter-system crossing (7) to a level S0,휈 and subsequent distribution of the vibrational energy (8). The second – radiative – path is phosphorescence, the transition (9) to S0,휈. However, since this process requires the simultaneous change of spin and energy (orbital) of the electron, this process shows only a low probability in polymers. Another possibility for deactivation of the triplet state is the transition (10 + 11) from T1,0 to the S1,0 state by means of thermal energy, which is only possible if the energy difference between the two states is not more than some 푘퐵푇. This gives rise to thermally activated delayed fluorescence (TADF), a process employed to increase quantum yields of organic or polymer light-emitting diodes. In case of high excitation densities, a bimolecular interaction of two molecules in triplet state T1 may also yield an molecule in the S1 state with the other one in S0 (triplet-triplet-annihilation). For most conjugated systems, the strongest absorption band is associated with a transition from the S0,0 singlet ground state to multiple vibronic levels of the lowest excited singlet state S1,휈′′. Oscillator strengths and positions of the different transitions depend on the shape and position of the molecular potentials in the ground and excited states (Figure II.2).

Figure II.1: Typical energy level diagram of organic molecules. The description of the processes of excitation and de- excitation is given in the text.

Figure II.2: Simplified potential energy curves showing how the mirror-image relationship between the absorption and emission spectrum of an organic molecule is caused. Adapted from [4] In the case of conjugated polymers, it is important to keep in mind that these substances can be regarded as a chain of chromophores with Gaussian length distribution. Due to disorder in the film, the conjugation length will vary throughout the film. To a first approximation, the dependence of electronic transition energies on the chromophore length can be described with the quantum mechanical “particle in a box” model: The longer the conjugation, the lower the Advanced Lab Course: Electroluminescence of Polymers 6 energies of absorption and fluorescence. Absorption of a photon in a (dense) polymer film will occur whenever the energy of a photon matches an allowed transition on any chromophore the photon passes in the film. Thus, when the incident light penetrates the polymer, it will sample a large variety of chromophore lengths leading to a broad, unstructured absorption band. However, prior to fluorescence the generated exciton will migrate to states of lower energy not only by means of internal conversion and vibronic relaxation, but also by energy transfer to adjacent molecular segments (this is sometimes called “spectral diffusion”, see Figure II.3). As a result, the fluorescence spectrum will be more structured and considerably red-shifted with respect to absorption. For polymers with a stiff backbone, both absorption and fluorescence will show clear vibronic progressions because most chromophores are of the same length and thus the energetic distribution of states is narrow. In this case, the red-shift of fluorescence with respect to absorption is smaller.

Figure II.3: Illustration of spectral diffusion: Absorption of a photon occurs into a distribution of possible states around a central transition energy 〈S1 ←S00 − 0〉. Prior to emission, the exciton migrates to sites of lower energy (a). Only for very low excitation energies below νloc, this migration is supressed (“localization threshold”, b). [5] II.2.2 Electroluminescence Emission of light can also be driven by the recombination of injected charge carriers. In this case, the excited state is formed when two oppositely charged carriers meet on one chromophore (e.g., segment of a polymer chain). In a molecular picture, the formed exciton represents an excited electronic state, e.g. S1, of the chromophore. Its decay follows the same principles as in photoluminescence. Based on spin statistics, injection of “uncorrelated” charge carriers produces singlet and triplet excitons in the ratio 1:3, as there is only one combination for an electron and a hole to form an exciton with total spin zero, but three combinations with non-vanishing spin components. The singlet excitons decay promptly, yielding what is referred to as prompt EL, whereas the triplet excitons either decay directly (electrophosphorescence), make use of thermal energy to produce TADF or they fuse to form singlet excitons (triplet-triplet annihilation, see above). The latter two processes contribute to delayed EL. However, in most conjugated polymers, triplet excitons decay mainly non-radiatively.

II.3 LIGHT-EMITTING DIODES A simple organic light-emitting diode (OLED) device geometry is shown in Figure II.4. It consists of one emission layer sandwiched between a hole and an electron injecting contact, denoted anode and cathode, respectively. Carriers of opposite sign are injected separately at opposing contacts when a sufficiently high voltage is applied. The most simple polymer-based OLED consists of a single layer of semiconducting fluorescent polymer sandwiched between two electrodes, a semitransparent, high work function anode like indium tin oxide (ITO) or gold and an opaque, low work function cathode made of calcium, aluminum or other materials. The thickness of the organic layer is typically in the order of 100 nm and, for experimental convenience, the active device area is in the order of a few square millimeters. More elaborated multilayer structures (Figure II.5) include additional charge-transporting layers. The function of e.g. the hole-transporting layer is to facilitate hole injection from the anode Advanced Lab Course: Electroluminescence of Polymers 7 to the emission layer and to prevent electrons from leaving the emission layer and reaching the anode without recombining with a hole. The electron-transporting layer functions accordingly.

Figure II.4: Schematic drawing of a single-layer electroluminescent diode. An applied electric field leads to injection of holes and electrons into the light-emitting film from the two electrode contacts. Formation of an electron-hole pair within the material may then result in the emission of a photon.

Figure II.5: Layout of a multilayer OLED comprising hole and electron- transporting layers.

II.3.1 Theory of Electroluminescence Electroluminescence in organic layers involves a series of steps: 1. injection of charge carriers from a metal contact into the organic layer 2. transport of charge carriers within the film 3. recombination of holes and electrons into excitons 4. migration and radiative decay of the exciton 5. emission of the generated photon to the outside The efficiency of these processes is closely related to the device geometry and to particular properties of the applied polymers such as the positions of energy levels of the charge transport materials or the solid state photoluminescence efficiency of the emitting polymer. II.3.1.1 Charge Injection One of the fundamental processes occurring in OLEDs is the injection of excess charges from the metal contacts into the electroluminescent polymer film. This charge injection can be qualitatively understood by considering the electronic energy structure of the thin polymer film with respect to the work functions (Fermi levels) of the injecting contacts (Figure II.7). As an example, the relevant electronic energy levels of MEH-PPV are shown in Figure II.6 along with the work function of various metals used as contacts in OLEDs. The ionization potential (퐼푃) of PPV, i.e. the energy required to remove an electron from the energetically highest occupied molecular orbital (HOMO) to vacuum, is roughly 5.2 eV. The electron affinity (퐸푎), i.e. the energy gained when adding an electron to the lowest unoccupied molecular orbital (LUMO) from vacuum, is roughly 2.5 eV. The energy gap, 퐼푃 − 퐸푎, is about 2.7 eV. To inject electrons, the contact must be able to donate electrons into the lowest unoccupied state 2.5 eV below vacuum. Similarly, to inject holes (remove electrons), the according contact must be able to accept electrons from an energy 5.2 eV below vacuum (more accurately, inject holes into that level). Advanced Lab Course: Electroluminescence of Polymers 8

Figure II.6: Electron energy level diagram of PPV and work functions of selected contact metals used in OLEDs. [3]

E (a) (b) (c) (d) Vacuum level

IP EA φC LUMO φ A ΔEe

HOMO ΔEh Anode Polymer Cathode x Figure II.7: Schematic energy levels of an OLED. (a) depicts the situation without contact between the components: The anode material needs to have a Fermi level that allows the extraction of electrons from (injection of holes into) the polymer HOMO, the cathode material needs a Fermi level that allows the injection of electrons into the polymer LUMO. After contact and at zero external bias (b), the Fermi levels of the electrodes will align, giving rise to an internal electric field (built-in field) in the polymer layer indicated by the tilted HOMO and LUMO levels. The application of an external voltage will first reduce the band bending back to the “flat-band-condition” (c) and further to an inverse bending (d). Situation (d) is the one needed for charge injection and transport in OLEDs, while the region between (b) and (c) is used for photovoltaic devices. Note that the work functions φA, φB, electron affinity EA, ionization potential IP, and hence the energy barriers ΔEh, ΔEe do not change upon contact or application of an external field. The (dotted) color arrows indicate the (possible) direction of motion for holes (red) and electrons (blue).

Electron injection is limited by the barrier Δ퐸푒 between the Fermi level of the cathode and the position of the LUMO. The rate of hole injection is similarly restricted by the barrier Δ퐸ℎ between the work function of the anode material and the HOMO of the polymer. Since many π conjugated polymers have ionization energies of about 5 eV, high work function materials such as gold, copper and indium-tin-oxide (ITO) are needed for hole injection. LUMO positions are typically located between 3 eV and 1.5 eV. Effective electron injection thus requires low work function metals such as calcium. However, these metals are very unstable in air and require device encapsulation. For a balanced charge injection, similar barriers Δ퐸푒 ≅ Δ퐸ℎ are needed. For most polymeric systems, balanced charge injection is difficult to realize in single layer devices. Ideally, there are no injection barriers. As the work function of the two contacts is different, the current will not only depend on the absolute field but also on its direction. Higher currents will flow if the contact with the higher work function is biased as the anode and the contact with the smaller work function as the cathode. This situation is denoted forward bias. Inverting the bias (reverse bias) will lead to smaller currents. In addition, a certain forward bias is required to overcome the so-called built-in field due to the different electrode Fermi energies. The operation conditions for forward bias are described in Figure II.7. Note that the minimum turn-on voltage is determined by the work function difference of the electrodes, additional injection barriers will cause an additional increase in operating voltage. An example for the resulting diode-like current-voltage characteristics are shown in Figure II.8. Advanced Lab Course: Electroluminescence of Polymers 9

Figure II.8: (left) Typical current-voltage and radiance-voltage characteristics of an MEH-PPV-based OLED with an Au anode and a Ca cathode. [3] (right) Band diagram for electron and hole injection into a polymer layer.

If the injection barriers Δ퐸푒 and/or Δ퐸ℎ become too high, they may limit the device performance. Two different models to describe injection-limited currents are thermionic injection (1) and field emission (Fowler-Nordheim tunneling) (2), 4휋푒푚 푘2 Δ퐸 푒푈 eff 퐵 2 푒,ℎ (1) 푗 = 3 푇 exp { } [exp { } − 1] ℎ 푘퐵푇 푘퐵푇 푒2푘 푚 8π√푚 2 퐵 eff eff (2) 푗 = 퐹 2 √ exp {− }. ℎ Δ퐸푒,ℎ 2 3푒퐹 In the equations, F is the electric field, all other symbols have their usual meaning. In principle, the barrier heights for electron and hole injection can be determined on the basis of these equations. Note, however, that one needs to separate hole and electron currents to determine the barriers. II.3.1.2 Charge Carrier Motion and Recombination After charges have been injected into the polymer, they need to be transported across the layer. The motion of charge carriers is generally described by the carrier mobility µ, which is defined by the ratio of the drift velocity 푣 and the electric field F, 푣 휇 = . (3) 퐹 In typical conjugated polymers, µ is in the range (10−7 … 10−1) cm2/Vs. In general, the mobility of holes and electrons is different (with the electron mobility typically being smaller). As discussed in sections II.1 and II.2.1, the polymers in thin films can be regarded as “chains of chromophores” where each chromophore has electronic properties that can be described in a molecular picture. Excess charges injected from the contacts will localize on single chromophores. Charge transport then requires the motion of the charges between these localized states. This process is called “hopping” transport in contrast to “band-like” transport that occurs in well ordered (crystalline) systems. This hopping transport mechanism is the main reason for low charge mobilities in disordered polymeric semiconductors. In the case of small injection barriers, the low carrier mobility of organic semiconductors determines the charge transport through the layer. After injection, charges will only slowly move away from the contacts. The accumulated charges lead to the formation of a space charge, i.e. charge carrier density and electric field are not constant over the film thickness. In the most simple case with no trapping sites in the semiconducting layer, the current density in the space- charge limited regime becomes 9 푈2 푗 = 휀 휀 휇 . (4) 8 0 푟 푑3 Advanced Lab Course: Electroluminescence of Polymers 10

In this case, the current depends quadratically on the applied voltage. For the meaningful extraction of a mobility value one needs to make sure that only one type of charges is present in the film. In addition, this simple behavior is usually not found in polymers due to the Gaussian distribution of transport states. Here, the low-energy states act as shallow traps for carrier transport. The best experiment to distinguish between injection limited current and transport limited current is the measurement of the thickness dependence: Injection limited currents do not depend on the film thickness while transport limited currents do. The recombination of an electron and a hole under the condition of low mobility is described by the Langevin recombination mechanism, which involves the drift of the charges in their mutual electric field. Due to the low carrier mobility, the exciton formation is much slower than the radiative decay of the exciton and controlled by the mobility and densities of the charge carriers. II.3.1.3 Migration and Radiative Decay of the Excitons A critical parameter in determining the operating efficiency of OLEDs is the luminescence quantum efficiency 휂(PL) of singlet excitons in the polymer, i.e. the probability that a singlet exciton will decay radiatively. This probability is limited by the intrinsic (intramolecular) quantum efficiency for radiative decay on an isolated molecule (as determined by the PL efficiency in dilute solution). However, in the solid state, several mechanisms can further reduce the quantum efficiency. For example, electronic coupling to neighboring molecules might alter the electronic states and thus the efficiency for radiative exciton decay. Furthermore, during its lifetime (typically 100 ps…1 ns) the excitons in polymers can diffuse to non-radiative sites (so- called quenching sites) or they might be deactivated at the metal electrodes via energy transfer or dissociation. Therefore, the photoluminescence quantum efficiency 휂(PL) in the solid state is generally smaller than in dilute solution. For “good” polymers, values range between 40% and 60%. However, with proper device and material design, internal quantum efficiencies approaching 100% have been reported. II.3.1.4 Emission of the Generated Photon to the Outside Not every photon generated inside the emission layer will escape the device and become visible to an external observer. Photons might be absorbed either by the emissive material itself (reabsorption), by additional charge transport layers or by the electrodes. Further, multiple reflection of photons at both electrodes may occur (see section II.3.2.1). Finally, the question is how a light-emitting diode appears to an external viewer. It is known from many surfaces that they have the same apparent brightness, independent of the viewing angle. If such a “Lambertian” surface is viewed under an angle 휃, the effective visible surface is given by (cf. Figure II.9)

푑퐴eff = 푑퐴⃑ ∙ 푟̂ = 푑퐴 ∙ cos 휃. (5)

Therefore, for an extended source the luminous intensity 퐼푣,휃 (that is the radiant energy emitted per unit time and unit solid angle by a source along a given direction) follows with (5)

퐼푣,휃 = 퐼푣,0 ∙ cos 휃. (6) where 퐼푣,0 is the luminous intensity normal to the surface. Due to multiple scattering and reflection events in the active layer, OLEDs follow to a good approximation Lambert’s emission law. Advanced Lab Course: Electroluminescence of Polymers 11

Figure II.9: Projection of an emitting surface element of an extended source II.3.2 Efficiency Considerations II.3.2.1 Quantum efficiency The efficiency of organic light-emitting devices is described by several parameters. First, a material with a high photoluminescence quantum yield is needed. The PL quantum yield is defined as the number of photons emitted per photons absorbed. It is determined by the radiative and non-radiative decay rates kr and knr, respectively.

The internal quantum efficiency of electroluminescence 휂𝑖푛푡(EL) is defined as the number of photons generated in the material per injected charge carrier,

휂𝑖푛푡(EL) = 휙푟 × 휙푒 × 휂(PL). (7) The individual factors denote the fraction of injected charges that recombine to form an exciton 휙푟, the fraction of emissive excitons 휙푒 and the quantum efficiency of photoluminescence 휂(PL), respectively. The fraction of recombining charges can be driven close to unity, but in general 휙푟 < 1 1. For fluorescent materials, 휙 = due to the fact the recombination of “uncorrelated” charge 푒 4 carriers yields singlet and triplet excitons in a ratio 1:3. For phosphorescent emitters, this factor can be 1 if efficient inter-system crossing converts all formed singlet excitons to triplets. The photoluminescence quantum efficiency enters the equation because once formed, the excitons created by charge recombination decay in the same manner as excitons formed by absorption of a photon.

The external quantum efficiency 휂푒푥푡(EL) < 휂𝑖푛푡(EL) is defined as the number of „detectable“ photons per injected charge carrier. The refractive index 푛 of the emissive polymer (typically 1.7…2.0) is larger than that of the supporting glass substrate (1.5). Thus, photons generated in the polymer layer, propagating at an oblique angle with respect to the surface normal, are reflected at the polymer-glass interface and may be waveguided in the device. For an isotropic material, where the transition dipoles are randomly oriented, 휂 (EL) 휂 (EL) = 𝑖푛푡 (8) 푒푥푡 2푛2 accounting for waveguiding and scattering losses inside the device.[6] As a result, even for ideal recombination conditions 휙푟 = 1 and 휂(PL) = 1, for a fluorescent material with refractive index 푛 = 1.7 the maximum external efficiency is about 5 %. II.3.2.2 Power efficiency

Relevant for a device application is the external power efficiency 휂푃(EL), defined as the emitted light power divided by the electrical driving power. Using the definition of the external quantum efficiency, 휂푃(EL) can be rewritten as ℎ휈 휂 (EL) = 휂 (EL) (9) 푃 푒푈 푒푥푡 where ℎ휈 is the energy of the emitted photons, e is the electric charge of an electron and U is the applied voltage. For polychromatic emission, the energy distribution of photons needs to be taken into account. A high power efficiency requires a high external quantum efficiency 휂푒푥푡(EL) and a low operating voltage U. For an external efficiency of 5%, an operating voltage of 3 V and an emission wavelength of 540 nm (ℎ휈 = 2.3 eV), 휂푃(EL) ≅ 4%. Advanced Lab Course: Electroluminescence of Polymers 12

Experimentally, the power efficiency can be determined by measuring the radiant flux Φ퐸 of the device and dividing it by the electrical driving power. The emission of an LED is usually measured as spectral radiant flux 휙퐸,휆 (see section VI) so that the power efficiency is calculated by

Φ퐸 ∫ 휙퐸,휆푑휆 휂푃(EL) = = . (10) 푃퐸 퐼 × 푈 II.3.2.3 Luminous efficiency The human eye, our own light detector, is characterized by non-uniform sensitivity to the various spectral components of light. An experimental function, the photoptic luminosity function 푉휆(휆) shown in Figure II.10 is used to represent the physiological effect of light throughout the visible part of the spectrum. At the maximum of the sensitivity curve (휆푚푎푥 = 555 nm), a radiant flux Φ퐸 = 1 W represents a luminous flux Φ푉 = 680 lm. The wavelength-dependent factor 푉휆(휆) relates physical units to physiological ones, i.e. it convert watts to lumens throughout the spectrum. The spectral luminous flux is related to the spectral radiant flux by

휙푉,휆 = 푉휆 × 휙퐸,휆. (11) The luminous efficiency is calculated by

Φ푉 휂푃(EL) = (12) 푃퐸 and has the unit lm/W. For different spectral regions, the same power efficiency will lead to different luminous efficiencies. White light incandescent lamps have luminous efficiencies of less than 20 lm/W while fluorescent lamps (and LEDs) can reach up to 100 lm/W.

700

600

500

400

) [lm/W] 300



(



V 200

Luminosity function Luminosity 100  =555 nm max 0 400 450 500 550 600 650 700 750 800 Figure II.10: Photoptic luminosity function V() depicting the sensitivity of the human eye Wavelength [nm] with the maximum sensitivity of 680 lm/W at 555 nm.[7]

II.4 POLYMERS FOR LIGHT-EMITTING DIODES II.4.1 MEH-PPV One well-known polymer is poly[2-methoxy,5-(2-ethylhexyloxy)-1,4-phenylene vinylene] (MEH-PPV), a soluble derivative of PPV. MEH-PPV has a fluorescence quantum efficiency in the solid state of approx. 10 %. The optical absorption and electroluminescence emission spectra of MEH-PPV are shown in Figure II.11. The absorption and emission spectra are typical for PPV- based polymers. The maximum absorption coefficient is about 2.5 × 105 cm−1. The absorption has a broad peak, which is roughly 0.5 eV wide with vibronic features evident at about 2.2 eV and 2.4 eV. The energy gap of MEH-PPV is about 2.4 eV. The electroluminescence emission spectrum is much more narrow, peaks at about 2 eV and contains several vibronic peaks. The width of the spectra as well as the large red shift in energy between the optical absorption and emission maxima is due to the disorder in the film, as described in section II.2.1. The photoluminescence Advanced Lab Course: Electroluminescence of Polymers 13 spectrum (not shown) is identical to the electroluminescence spectrum indicating that the excited states produced optically and electrically are identical.

Figure II.11: Electroluminescence and optical absorption spectrum of the soluble polymer MEH-PPV.[3] II.4.2 LPPP Another well-known polymer is ladder-type poly(para-phenylene) (LPPP). The emission color of this polymer is bluish-green. In the solid state, LPPP has a fluorescence quantum efficiency of approx. 40 %. It contains linking groups that preserve the structure of the polymer chain (see Table II.1) and thus has considerably sharper spectral features compared to MEH-PPV. However, this material tends to form aggregates or chemical defects. These sites can function as charge carrier traps and emission sites. Therefore, the EL spectrum of this polymer might differ from the solid state PL spectrum: In PL, the excited state is rather immobile during its short lifetime. In contrast to that, in EL the injected charges have to travel through the film bulk before they recombine. Therefore, the probability of being captured in traps is much larger than in PL leading to a higher contribution of defect emission in EL compared to PL. II.4.3 Phosphorescent OLEDs An elegant way to increase the efficiency of OLEDs is to harvest triplet excitons. To do so, heavy metal atoms with a strong spin-orbit coupling are required in the emitting species. Since the energy of the T1 state is generally lower than the energy of the S1 state, singlet excitons will transfer their energy to triplet states as well. Hence, in equation (7), 휙푒 = 1 and the internal quantum efficiency can reach 100%. An easy way of constructing phosphorescent OLEDs is to blend phosphorescent dyes into a polymer matrix. II.4.4 Multi-Component OLEDs In light-emitting diodes, the emitted light originates from a defined electronic transition. It is, therefore, monochromatic (although the emission spectrum can be relatively broad in molecular substances due to vibronic progression). To achieve a spectrally broad emission or even white light emission, chromophores with different transition energies and thus emission colors covering the entire visible spectrum have to be combined. Also, it is feasible to combine different moieties with specialized properties such as efficient emission or good charge transport into the emissive layer e.g. by co-polymerization or simple blending. In such multi-component systems, interactions between the different components have to be taken into account. Steady-state spectra are influenced by two major processes: energy transfer and charge trapping. The energy of an excited state can be transferred to neighboring chromophores by different processes such as simple transfer (reabsorption), Förster and Dexter transfer. In Förster transfer, the transition dipole moments of two chromophores couple. This is a long-range interaction (several nanometers). In contrast, Dexter transfer is an electron-exchange interaction that requires wave function overlap and is thus short-ranged. All processes have in common that the energy of the emitted light is lowered. Advanced Lab Course: Electroluminescence of Polymers 14

The different emission wavelengths originate from the different energy levels (“band gaps”) of the materials. On an energy scale, the transport levels for electrons and holes will be closer together for a red-emitting material than they are for a blue emitter. In a blend of two such materials, the energetic position for at least one type of carrier will be much more favorable on the red emitter compared to the blue-emitting material. The situation is depicted in Figure II.12. If charges that are initially placed on a blue chromophore start to move through a blend film due to an applied field, they will eventually come close to a red chromophore and jump to its energy level. The return to the blue material now requires additional energy. If there are only few red chromophores, the charge can not move further and will essentially be trapped on the red dye. If a charge of opposite sign comes close, it will be coulombically attracted to the first charge, jump to the molecule and form an exciton. This process occurs in addition to energy transfer from blue to red excitons such that in electroluminescence, low-energy emission is more pronounced than in photoluminescence for the same system.

Figure II.12: Illustration of charge trapping in a multi- component system. After coming close to a low-energy site, the electron will jump on this chromophore (1). A release into the transport level requires additional energy and is thus unlikely (2). However, a hole may get transferred onto the chromophore due to coulombic attraction (3) followed by the formation of an exciton that then decays radiatively (4).

III. EXPERIMENT EXECUTION AND ANALYSIS

III.1 GENERAL REMARKS You will investigate two different materials. A sheet with “technical data” for your specific samples will be provided at the setup. Make sure to note experimental conditions such as wavelength settings or diode driving conditions. They need to be included in the report, all data presented must contain information on how it was collected. Typically, the absolute intensities of the different spectra can not be compared due to differences in excitation and measurement geometries. Relative differences within the spectral distributions are the main focus of this lab course. In the report, you are supposed to explain your data. Do not just present a uncommented collection of results. Store all raw measurement data in your folder on the lab course server (drive T:). In preparation for the experiment, familiarize yourself with the experimental techniques used and why they are applied in the order suggested by the experimental tasks.

III.2 ABSORPTION SPECTRA Measure the absorption spectra of the polymers (thin films spin coated onto glass substrates). Light absorption follows Lambert-Beer’s law,

퐼(푥) = 퐼0 exp{−훼(휆)푥} (13) with the light intensity 퐼(푥) at position 푥, 퐼0 the incident light intensity (푥 = 0) and the spectral absorption coefficient 훼(휆). Absorption spectra will be measured using a standard absorption spectrometer (Thermo Scientific Evolution 220). This spectrometer consists of two light paths (sample and reference path). By measuring a “baseline” curve without samples, differences in the two light paths are recorded for automatic correction of the measured spectra. When choosing “Extinction” as measurement mode, the optical density (decadic spectral extinction 퐸(휆)) Advanced Lab Course: Electroluminescence of Polymers 15

퐼0 퐸(휆) = log { } (14) 10 퐼(푑) with the sample thickness 푑 is then calculated from the recorded intensities of the sample and reference path and stored in a data file. To account for contributions of the glass substrates the films are deposited on, the absorption spectrum of an empty substrate should be recorded as well. A more detailed manual can be found at the instrument. From the spectra, suitable wavelengths for excitation of photoluminescence need to be determined.

III.3 LUMINESCENCE SPECTRA A schematic of the setup used for measuring excitation and emission spectra is shown in Figure III.1. For measurement of steady state photoluminescence excitation and emission spectra, a xenon maximum pressure lamp is used. The excitation and emission part of the spectrometer consist of two similar (reflective) grating monochromators controlled by step motors. The sample chamber contains a holder for polymer films on glass substrates. The sample holder is designed to hold the polymer film in the rotational axis of the holder in the intersection of excitation and emission path. When this sample holder is removed, a cuvette holder for liquid samples can be installed. A photomultiplier (Hamamatsu R928 in “digital mode”) detects the emitted light.

steady state Xe lamp

excitation monochromator emission monochromator photon polymer counter film Figure III.1: Schematic setup used for PL and EL measurements sample chamber (top view).

The monochromator connected to the Xe lamp is used to select which wavelength is exciting the sample (hence it is called the “excitation monochromator”) while the monochromator in front of the detector determines which wavelength will be transmitted from the sample chamber to the detector (“emission monochromator”). All spectra will be recorded in a “step-by-step” mode. This is necessary as a simplistic approach to shine broad-band white light on the sample and measure the light originating from the sample would make it almost impossible to identify the origin (and wavelength) of the detected light. Based on the instrument design, two main measurement modes are possible. First, the excitation monochromator can be set to a fixed wavelength while the emission monochromator scans a given wavelength range (one wavelength after the other). The resulting spectrum is called an “emission spectrum” as it shows the light that is emitted by the sample upon excitation with the (fixed) wavelength set by the excitation monochromator. The other mode is the reversed case: The emission monochromator is fixed to a certain wavelength while the excitation monochromator scans a given range. The resulting “excitation spectrum” contains information on how efficiently emission at the fixed emission wavelength is excited by different incoming wavelengths. Of course, a reasonable emission spectrum can only be recorded using a suitable excitation wavelength. Similarly, for the measurement of an excitation spectrum, the emission monochromator needs to be set to a wavelength at which the sample actually emits. When light encounters a surface, part of the intensity will always be reflected and, for real surfaces, scattered. This results in scattered excitation light being recorded by the detector. While Advanced Lab Course: Electroluminescence of Polymers 16 this light can have a quite high signal intensity, it is not relevant to the measured spectra. Thus, it can be excluded from the measurement range. However, keep in mind that scattered light will show up. Also, the grating monochromators diffract light not only in one diffraction order. III.3.1 Correction for instrument response Characterize the setup used for luminescence measurements. Identify possible sources of error and collect data necessary for correction. The setup does not have a constant sensitivity over the spectral range under investigation. This so-called instrument response has to be separated from the measurement data in order to get correct results. Section IV.1 lists possible instrument-related sources of error and different methods for the correction of excitation and emission spectra. Due to the construction of the setup, no absolute “intensity” measurements are possible. Moreover, they are not needed for most of this lab course. However, it is necessary to remove any wavelength-dependent factors that influence the measured spectra in order to be able to compare the (corrected) spectra with others. Determine which part of the setup will have a wavelength- dependent influence on which type of measured spectra. Design experiments that allow to correct each wavelength dependence of the setup. Collect the necessary data. The correction functions are to be applied to the according spectra before further discussion. In the report, describe carefully how you derived and applied your correction functions. Specify in what units the “brightness” of your samples will be displayed after correction. Use Table VI.1 for reference. Although listed here as the first item, the correction data can be collected at any point during the course of the experiment – assuming that the setup does not change its characteristics during the experiment. III.3.2 Photoluminescence Emission Spectra Measure the photoluminescence emission spectra for the polymer films in an appropriate measurement geometry, using suitable excitation wavelength(s). Keep in mind that for all spectra measured, a background correction needs to be performed. Considering the energetics of luminescence, you can reduce the wavelength range that needs to be covered for detection. From a more technical point of view, consider different origins of the detected light. How should the sample be mounted in order to suppress unwanted signals? Optional: Compare the front face measurement when emission is recorded at the excitation side of the film with a measurement in in-line geometry where emission is recorded at the side opposite to the excitation side of the film for one film (see section IV.2). III.3.3 Photoluminescence Excitation Spectra Measure the photoluminescence excitation spectra for the polymer films in an appropriate measurement geometry, using suitable detection (emission) wavelength(s). Keep in mind that for all spectra measured, a background correction needs to be performed. For the report, discuss the three types of spectra you collected for each of your samples. How are the different spectra related theoretically? Do your measured spectra show these relations? Also, compare the spectra of the different samples. Remember that absolute statements (“sample A is brighter than sample B”) can not be drawn from the data you collected, but similarities and differences in the spectral distribution should be discussed. To compare different spectra, it is very helpful to plot them together in a common diagram. As absolute intensities can not be compared anyway, normalization of spectra can also be helpful.

III.4 ELECTROLUMINESCENCE Figure III.2 shows the general structure of the OLEDs. Device fabrication starts using a glass substrate with pre-patterned ITO coverage. Onto these substrates, the active material is deposited by spin coating from solution. The solution viscosity and spin speed determines the thickness of Advanced Lab Course: Electroluminescence of Polymers 17 the dry film. Top electrodes are deposited by thermal evaporation in high vacuum. The full fabrication requires a few hours and is not part of the lab course. Figure III.2 shows one example device structure. The active area of an OLED is defined by the overlapping areas of both electrodes which form a sandwich bottom electrode – active polymer – top electrode. Hence, each substrate contains 6 independent “pixels” each having an area of 0.16 cm2 (4 mm × 4 mm). The active device region is encapsulated by means of a microscope cover glued onto the device. The electrodes are designed to be contacted outside the encapsulated region. The exact composition and geometry of the provided samples will be given at the setup.

Figure III.2: Example OLED device structure in top and side view. ITO-covered areas are shown in light blue, the metal cathode in black. The side view includes the polymer film (yellow, thicknesses not to scale). The dashed line indicates the encapsulation.

III.4.1 Current-Radiance-Voltage Characteristics Measure the current I through the OLED as a function of applied voltage U (I = f(U)). At the same time, measure the spectral radiant intensity of electroluminescence iE, as a function of voltage at a constant emission wavelength. The setup uses a voltage source and a separate Amperemeter to measure the current through the device. Plan the proper circuit you need to measure the current-voltage characteristics, i.e. how to connect voltage source, Amperemeter and sample. Sketch a wiring scheme if necessary. As the detector is mounted at the emission monochromator, you need to decide which wavelength may be suitable to measure the radiant intensity. It is helpful to recount possible processes that result in luminescence. The chosen wavelength needs to be set in the program for spectral measurements (“FELIX”). The optical excitation part of the setup will not be used in electroluminescence measurements, so the excitation wavelength is not relevant. The excitation light can (and should) be blocked with the appropriate shutter. For the report, make sure that your plots are showing all the interesting information: Both current and radiance will change over several orders of magnitude. Thus, a logarithmic representation may be suitable.

Plot iE, as a function of the current I through the OLED and discuss your findings. Discuss whether you need any correction or even calibration (to absolute units) of your radiance data. What device parameters determine the current-voltage characteristics (in particular, turn-on voltage) of an OLED? III.4.2 Electroluminescence spectra Measure the electroluminescence emission spectrum at a suitable voltage or current. The suitability of a voltage has mostly technical limits: There needs to be some light emission at all, and this emission must not saturate the detector system. Both limits relate back to the wavelength choice for the measurement of current-radiance-voltage characteristics. Compare the electroluminescence spectrum to the photoluminescence spectrum. What do you conclude concerning the mechanism of both luminescence processes? III.4.3 Lambert’s law Measure the radiance of an OLED as a function of viewing angle. For this measurement, the so-called “time-based” method of FELIX is be used. Here, the emission intensity is recorded while the emission monochromator (as well as the excitation monochromator) is at a fixed wavelength. The temporal resolution is about 0.1 seconds. It has been proven practical to alter the viewing angle between -70° and +70° in steps of 10° while keeping each angle for about 10 seconds. The resulting plot will be a multi-step graph whose steps can easily be assigned to the different angles. For the report, it is sufficient to use an approximated Advanced Lab Course: Electroluminescence of Polymers 18 mean value for each step (angle). Analyze the data according to Lambert’s law. To do so, plot your data as a function of cos 휃. Is your source a so-called Lambertian emitter? III.4.4 Efficiency Determine the radiometric and luminous efficiencies of the OLEDs. In order to calculate the efficiencies according to eqns. (10) and (12), the measured and corrected spectra need to be calibrated to absolute units. To do this, a light source of known absolute radiance is needed. For this reason, a calibrated inorganic LED which can be mounted at the position of the organic samples will be provided. The driving current and corresponding radiant flux will be given. The eye sensitivity curve 푉휆(휆) is available as data file at the work space. Depending on the correction function used, the values 푠휆,푐표푟푟 of corrected emission spectra are proportional but not equal to either the photon flux or the spectral radiant intensity. For an energetic calibration, the proportionality factor 푓푐푎푙 has to be determined. For spectra that are corrected proportional to photon flux, the procedure is as follows: The photon number, received 휋푑2 in a room angle Ω determined by the equipment (Figure III.3, Ω = with 푑 = 35 mm, 푙 = det det 4푙2 36 mm) is measured with a spectral bandwidth Δ휆 (about 3 nm for the setup used) at the wavelength 휆 in a defined time interval Δ푡. Thus, the spectral radiant intensity follows to

d퐼퐸 푠휆,푐표푟푟 ℎ푐 푖퐸,휆 = | = 푓푐푎푙 ∙ ∙ , (15) d휆 휆 Δ휆 ∙ Ωdet 휆 whereas the total spectral radiant flux of a Lambert radiator is dΦ 휙 = 퐸| = ∫ 푖 dΩ′ 퐸,휆 d휆 퐸,휆 휆 (16) 푠휆,푐표푟푟 ℎ푐 = 푓푐푎푙 ∙ ∙ ∙ 휋. Δ휆 ∙ Ωdet 휆

Using the total radiant flux Φ퐸 = ∫ 휙퐸d휆 of a calibrated light source measured under the same conditions as the light source under investigation, the wavelength-independent calibration factor 푓푐푎푙 can be determined. Once 푓푐푎푙 is known, the radiant flux and luminous flux of the OLEDs can be determined for efficiency calculations. Make sure to describe your derivations conclusively and note all necessary steps and quantities in your report. A formal algebraic derivation of the final equation before plugging in numbers may help to avoid lengthy numeric calculations. sample detector Wdet d

Figure III.3: Illustration of the room angle Ωdet for the measurement configuration. The geometric values are l l = 36 mm, d = 35 mm.

III.5 SUMMARY OF EXPERIMENT As a conclusion to the report, summarize your findings. Do the results meet your expectations? Are there differences to the theoretical predictions, how can they be explained? Do all samples behave in the same way or are there notable differences? Also, discuss whether the instrumentation you used was suited for the task and identify possible sources of error. Advanced Lab Course: Electroluminescence of Polymers 19

IV. INSTRUMENTATION AND POSSIBLE SOURCES OF ERROR

IV.1 CORRECTION OF SPECTRA Because of the spectral characteristics of optical components, the observed signal in excitation and emission measurement is distorted for several reasons:  The diffraction efficiency and the polarizing effect of the monochromators are functions of wavelength.  The light intensity from the excitation source is a function of wavelength. The intensity of the excitation light can be monitored via a beam splitter, and corrected by division. (In this case, the spectral sensitivity of the monitor has to be considered as well.)  The optical density of the sample may exceed the linear range, which is about 0.1 absorbance units, depending upon sample geometry and the slit width of monochromators.  The emission spectrum is further distorted by the wavelength dependent efficiency of the photo detector and the emission monochromator. The development of methods to correct excitation and emission spectra (photo- and electroluminescence) for wavelength dependent effects has been the subject of numerous investigations. Overall, none of these methods are completely satisfactory. Prior to correcting spectra, the researcher should determine if such corrections are necessary. Frequently, one only needs to compare spectra with other spectra collected on the same instrument. However, corrected spectra are needed for calculations of quantum yields (efficiencies) and overlap integrals. IV.1.1 Correction of excitation spectra The wavelength dependent intensity of the exciting light can be converted to a signal proportional to the number of incident photons by the use of a so-called quantum counter. Here we use a concentrated solution of Rhodamine B in ethylene glycol which absorbs all incident light and provides a fluorescence signal of constant wavelength, which is proportional to the photon flux of the exciting light. In a certain wavelength region (up to 550 nm), the fluorescence intensity is independent of excitation wavelength. In the setup, the emission intensity of Rhodamine is recorded by use of a triangular cuvette instead of the sample, like an excitation spectrum. The excitation correction function 푐푓푡푒푥(휆) is the inverse of the measured spectrum. Accordingly, all recorded excitation spectra have to be divided by the “Rhodamine excitation spectrum” to obtain the corrected excitation spectra. The fluorescence intensity of Rhodamine solution is independent of excitation wavelength, but the quantum efficiency is not known. Hence, this correction method does not give a calibration of the excitation path in terms of an absolute photon density (or similar). It rather yields a measure for the relative change in photon flux for the different excitation wavelengths. IV.1.2 Correction of emission spectra The correction of emission spectra requires knowledge of the wavelength dependent efficiency of the detection system, which consists of all components the emission light has to pass and the detector. The wavelength dependent correction factor is generally obtained by the measurement of light with a known spectral distribution. The sensitivity of the detection system 푆(휆) is calculated as the ratio of the measured signal 퐼(휆) and the known “true” spectral intensity distribution. The correction function for emission spectra 푐푓푡푒푚(휆) is the inverse of the sensitivity 푆(휆). The standard spectrum can be  the emission spectrum of a standard substance,  the spectral distribution of the excitation light (if known) measured using a wavelength independent scatterer, e.g. MgO or  the wavelength dependent output from a calibrated light source, e.g. a tungsten filament lamp of known color temperature. Advanced Lab Course: Electroluminescence of Polymers 20

The last method makes use of the fact that the (spatial) energy density in a frequency interval between 휈 and 휈 + 푑휈 for a so-called black body can be calculated from Planck’s formula [8] 8휋ℎ휈3 1 휚 (휈, 푇)푑휈 = 푑휈. 휈 3 ℎ휈 (17) 푐 exp { } − 1 푘퐵푇

A detector of spectral width 푑휈 will measure the spectral irradiance 푖휈(휈, 푇)푑휈 with 휚 ⋅ 푐 휚 ⋅ 푐 2ℎ휈3 1 푖 (휈, 푇) = 푣 = 푣 = . 휈 2 ℎ휈 (18) Ω 4휋 푐 exp { } − 1 푘퐵푇 A tungsten filament lamp is not an ideal black body. However, at proper operating conditions, it can be approximated as one using not the real filament temperature but its so-called color temperature for which equations (17) and (18) hold when including the emissivity 휀휈 < 1. Keep in mind that the setup uses gratings for wavelength selection. They divide the spectrum into wavelength intervals of constant length rather than constant frequency intervals. Hence, the intervals 푑휈 need to be transformed into wavelength intervals 푑휆. Also, the equations can be adjusted to give different correction functions such as photon flux or radiant flux. Not all light emitted by the tungsten lamp can and will be collected by the detection system: At its maximum wavelength, the lamp emits some 1021 photons per second. The total emission will thus be “spatially filtered” by (rather large) pinholes that do not change the spectral distribution of the emitted light. As a result, this correction method gives only a relative correction of recorded spectra but no energetic calibration. (This is why the unknown emissivity 휀휈 < 1 does not bother us.) IV.1.3 Background Even without any measurement signal, the detector still produces a count rate. To a first approximation, this wavelength-independent contribution stems from the thermal dark count rate of the detector. Stray light from the excitation source and laboratory lighting may also be included. Therefore, for all spectra a proper background measurement has to be performed and subtracted from the signal before further correction. The background is obtained best with the setup adjusted for an actual measurement, but with the excitation source turned off or blocked.

IV.2 INNER FILTER EFFECTS The apparent emission intensity and spectral distribution can depend on the optical density of the sample and the precise geometry of sample illumination. The path length of excitation light through the sample to the point observed by the detection channel and the path length of the emission light through the sample determine the influence of the absorption behavior on excitation and emission spectra. If there is a strong overlap of absorption and emission spectra one often observes a reduced emission intensity at blue side of the emission spectrum (called reabsorption, if the absorbing and emitting species are the same). In general, the influence of the absorption on the emission spectrum is called Post Filter Effect. In excitation spectroscopy, a high optical density at the excitation wavelength can reduce the excitation intensity in the observed volume element. As a result one might measure a smaller emission intensity compared to a sample with a smaller optical density (Pre Filter Effect). The correction of these effects is a complicated problem, since one needs the exact description of the geometrical light paths in the sample and the spectral characteristics of the sample.

Advanced Lab Course: Electroluminescence of Polymers 21

V. REFERENCES AND FURTHER READING

[1] M. Pope, H. P. Kallmann, P. Magnante, “Electroluminescence in Organic Crystals”, The Journal of Chemical Physics 38, 2042 (1963), DOI: 10.1063/1.1733929 [2] R. H. Friend, R. W. Gymer, A. B. Holmes, J. H. Burroughes, R. N. Marks, C. Taliani, D. D. C. Bradley, D. A. Dos Santos, J. L. Bredas, M. Logdlund, W. R. Salaneck, “Electroluminescence in conjugated polymers”, Nature 397, 121–128 (1999), DOI: 10.1038/16393 [3] G. Hadziioannou, P. F. (eds. . van Hutten, Semiconducting Polymers : Chemistry, Physics and Engineering, Wiley-VCH, Weinheim (2000) [4] M. Pope, C. E. Swenberg, Electronic Processes in Organic Crystals and Polymers, Oxford University Press, New York (1999) [5] H. Bässler, B. Schweitzer, “Site-Selective Fluorescence Spectroscopy of Conjugated Polymers and Oligomers”, Accounts of Chemical Research 32, 173–182 (1999), DOI: 10.1021/ar960228k [6] N. C. Greenham, R. H. Friend, D. D. C. Bradley, “Angular Dependence of the Emission from a Conjugated Polymer Light-Emitting Diode: Implications for efficiency calculations”, Advanced Materials 6, 491–494 (1994), DOI: 10.1002/adma.19940060612 [7] A. Stockman, A. Rider, B. Henning, R. Baabbad, D. Mylonas, P. Andrikopoulos, “Colour & Vision Research laboratory and database,” http://www.cvrl.org/ [8] M. Planck, “Ueber das Gesetz der Energieverteilung im Normalspectrum”, Annalen der Physik 4, 553–563 (1901), DOI: 10.1002/andp.19013090310 [9] A. D. Ryer, Light Measurement Handbook, International Light, Newburyport (1997)

 E.E. Krieszis, D.P. Chrissoulidis & A.G. Papagiannakis, Electromagnetics and Optics (World Scientific Publishing Co., Singapore, 1992).  P.W. Atkins, Physical Chemistry (Oxford University Press, Oxford, 1992).

References in German language:  M. Schwoerer, H.C. Wolf, Organische Molekulare Festkörper (Wiley-VCH, 2005)  P.W. Atkins, Physikalische Chemie (Wiley-VCH, 2001)

A selection of these references can be downloaded from the lab course web site http://www.uni-potsdam.de/u/physik/fprakti/Start.html (Link “Literatur”, authorization required). It is also available on the measurement computer or can be accessed via Windows Networking \\PPF-Server\PPF\M7 - Additional Information\References and Further Reading.zip. The other references are available in the university’s library. Advanced Lab Course: Electroluminescence of Polymers 22

VI. RADIOMETRIC AND PHOTOMETRIC QUANTITIES

Table VI.1: Derivation of radiometric and photometric quantities. Photometric quantities represent the visible part of the optical radiation field. Photometry provides physiological insight to the optical radiation field detected by the human eye. On the other hand, radiometry refers to the energy content of the optical radiation field. For consolidation, use e.g. [9]. radiometric (physical) photometric (physiological) Consider a light source S (Figure VI.1) with an area A, which emits an electromagnetic field into the space above the source. The energy emitted within this radiation field is the The luminous energy Qv is the energy of the visible radiant energy QE which is measured in Joules (J). radiation field. It is measured in lumen×second (lms). The radiant power (Strahlungsleistung) or radiant The luminous flux (Lichtsstrom) Φ푉 is defined as flux (Strahlungsfluss) 푑푄푉 Φ푉 = (A1b) 푑푄 푑푡 Φ = 퐸 (A1a) 퐸 푑푡 and is measured in lumens (lm). is the radiant energy emitted per time and is measured in Watts (W). The radiant flux represents an instantaneous radiometric quantity. The radiant intensity (Strahlungsstärke) The luminous intensity (Lichtstärke) 푑Φ 푑Φ 퐼 = 퐸 (A2a) 퐼 = 푉 (A2b) 퐸 푑Ω 푉 푑Ω is defined as radiant flux per unit solid angle emitted is defined as the luminous flux through the solid by a source along a given direction. It is measured in angle dW. It is measured in candela (cd) (basic unit W lm watts per steradian ( ). of SI), which is also referred to as . sr sr With regard to Figure VI.1, the solid angle 푑Ω subtended by the surface element 푑푠⃑⃑⃑⃑⃑ ∥푟⃑ is given by 푑푠 푑Ω = (A3) 푟2 The radiance (Strahldichte) LE of an extended source The luminance (Leuchtdichte) Lv expresses the is defined as emitted radiant intensity per unit luminous intensity per unit emitting area, emitting area dA (Figure II.9), 푑퐼푉 1 푑퐼푉 퐿푉 = = . (A4b) 푑퐼 1 푑퐼 푑퐴⃑⃑⃑⃑⃑⃑∙푟̂ cos 휃 푑퐴 퐿 = 퐸 = 퐸. (A4a) 퐸 푑퐴⃑⃑⃑⃑⃑⃑∙푟̂ cos 휃 푑퐴 Again, only the apparent emitting area is considered. In this definition, only the apparent emitting area lm cd Lv is measured in 2 or 2. that is the source area projected on a plane m ⋅sr m

perpendicular to the observation direction 푟̂ is taken W into account. The radiance is measured in . m2⋅sr

1979 SI definition: The candela is the luminous intensity, in a given direction, of a source that emits monochromatic radiation of frequency 540 ⋅ 1012 Hz and that has a radiant intensity in that 1 W direction of . 683 sr

Figure VI.1: Sketch to define the radiometric quantities of light emitted from the light source S