[page 40↓]

3.  The Electrochemiluminescence Experiments

3.1. Solution phase ECL: DPA

The model compound in the electrogenerated chemiluminescence experiments is 9,10-diphenylanthracene, abbreviated as DPA. Due to the consistency and the simplicity in the nature of the electrochemiluminescence process in DPA, it has been used as the compound for the standardization of the experimental conditions [56] It is chemically denoted as follows:

9,10 - diphenylanthracene (DPA).

3.1.1. Experimental details

The Pt substrate was polished with 0.7 μm diamond paste and the surface was checked for uniformity with a Nikon PFX 104 optical microscope. Electrochemical experiments were carried out in a three-electrode configuration with Pt disc as the working electrode (WE) and Pt wires as quasi-reference (QRE) and counter electrodes. Anhydrous CH3CN (Aldrich No. 27,100-4) was used as the solvent. DPA corresponding to the concentration 1 mM, tetrabutylammonium perchlorate (TBAClO4) or tetraethylammonium perchlorate (TEAClO4)corresponding to 0.1 M in concentration were taken in the evacuated mixing chamber. The mixture was continuously heated while evacuating the container. (The supporting electrolyte and DPA were also dried in the oven for 12 hours to remove traces of water present in them prior to transferring them to the mixing vessel). Acetonitrile was pumped in to make the solution of the DPA and the supporting electrolyte. The solution was bubbled with argon (Ar) to chase out the traces of oxygen and was heated frequently to accelerate the [page 41↓]solubility of DPA in acetonitrile. It was then transferred to the electrochemical cell in-situ. Electrochemical measurements were accomplished with a homemade potentiostat. Potential pulses were applied by a Agilent 33250 A function generator via the potentiostat. ECL emission intensities were measured with an RCA photo multiplier tube (type number 7326), which was polarized at -700 V by a Keithley 246 high voltage supply. The Pt working electrode was cycled in the potential range from -2.1 V to +1.6 V vs. Pt (QRE) at a scan rate of 100 mV s-1. The range of potential used in the sequential potential step experiment to generate the ECL was determined from the oxidation and reduction peaks in the cyclic voltammetry.

3.1.2. Cyclic voltammetry

The cyclic voltammogram has the same features for both TEAClO4 and TBAClO4. The potentials for the oxidation and reduction of the polymer in the ECL experiment were determined from the E°' value from the cyclic voltammogram.

Fig. 3.1.2a: Cyclic voltammogram (CV) of Pt in 1mM DPA + 0.1M TBAClO4 in acetonitrile at a scan rate of 100 mV/s.

3.1.3. Current transients

Current transients are obtained from the potential step experiments. Under the conditions of potential step experiments, initially the charging of the [page 42↓]electrode/solution interface takes place for a duration depending on the time constant τ (see section 2.1.2.ii and 2.1.2.iii) followed by the electron transfer reaction. The parameters determining the time constant, the resistance of the electrolytic solution (Ru) and the double layer capacitance Cdl were evaluated first by a separate experiment where the potential step was made to a lower magnitude, thereby avoiding the electron transfer reaction at the electrode. In the case of DPA, the potential steps were made from 0 V to 0.5 V or - 1 V of duration 50 ms. From the decay characteristics of the corresponding current transients, the uncompensated solution resistance Ru and the double layer capacitance were calculated. For the cathodic transient they were 71 Ω, 1.99 μF and for the anodic transient 148 Ω, 0.9 μF respectively. The values of the resistance and capacitance were averaged out to be 109.5 Ω and 1.5 μF, respectively. The area of the electrode is 0.07065 cm2.

Then the rate parameters of diffusion were calculated by making potential steps of higher magnitudes past the standard potentials for oxidation/reduction. These lead to the production of radical cations or radical anions depending on whether it is a negative or a positive potential step. The current flow after capacitive charging of the electrode in this case is controlled by the diffusion of the radical ions which is described by the Cottrell equation (Eq. (2.1.2b)) The diffusion coefficient was evaluated to be 5x10-6 cm2 s-1 from the slope of the diffusion current (i) vs. the inverse square root of time (t-1/2).

Fig. 3.1.3a: Cottrell plot for 1 mM DPA in 0.1 M TBAClO4-AN when the potential step was made from 0 V to -1.95 V of duration 50 ms.


[page 43↓]

3.1.4.  The ECL experiment

As explained in the introduction, the ECL was produced in the third step of the triple potential step experiment. The typical form of the triple potential step is shown below:

Fig. 3.1.4a: Scheme of the sequential anodic and cathodic potential steps in ECL experiments.

In this case we expect cathodic ECL. The potential step can also be reversed to produce anodic ECL. The ECL experiments were done with 1 mM DPA in 0.1 M TBAClO4 in acetonitrile. The emission intensity increased when the magnitude of the potential step was raised due to the increased rate of formation of the reactant.

Fig. 3.1.4b: The potential step from Ef to Er was made from -1.9 V (Ef) to a positive (Er) value given in the inset. tf, tr = 0.05 s and tw = 0.2 s.


[page 44↓]

The mechanism of ECL generation should be the same, irrespective of the potential step sequence. Contrary to the expectation, the intensity and the kinetics of the anodic and cathodic ECL are found to be different. Several researchers have tried to reason out the difference. In DPA, it has been reported in the literature that the intensity of the anodic ECL is more pronounced than the cathodic ECL. However, the occurrence of cathodic ECL is more consistent than the anodic ECL. The various reasons proposed for the occurrence of non-symmetrical ECL in DPA are as follows:

(1) The difference in the life times of the radical ions. It was analyzed with electron spin resonance (ESR) spectroscopic studies. The DPA radical cations and anions were turned out to be almost of equal stability. The half-lives of the radical cations are 0.5 min and radical anions are 1min [57].

(2) Different possibilities are there for the formation of the excited states. There are two major routes ¯ the singlet route (S route) and the triplet route (T route). In the singlet route the annihilation between radical cation and anion results directly in the production of the singlet-excited state, whereas in the T route first a triplet-excited state is formed. Two formed triplets undergo mutual annihilation resulting finally in the singlet-excited state. Due to the paramagnetic spin of the triplet-excited state, its resultant magnetic field can interact with an external magnetic field. Some experiments were done in this direction. Since there were no appreciable effects, it was concluded that it forms the excited state predominantly by the singlet route [58 and references therein]. Also the efficiency of the DPA ECL agrees with that expected for an energy sufficient system [51 and references therein]. The S route is reported to be predominant under the conditions of cathodic ECL than that of anodic ECL in DMF solvent [59]. The major criteria set for the S route behavior are (i) the congruence of transients generated by the two possible reactant generation sequences and (ii) an ability to linearize the transients with the inverse square root of time [60, 61]. Although discrepancies were reported [62, 63], the absence of any effect on the DPA ECL by a magnetic field suggests that DPA follows the S route for the generation of the excited state.

(3) Ionic association of the radical ions with the counter ions from the supporting electrolyte could lead to slow diffusion of one of the ions. This could also result in the non-sysmmetry of the ECL process. To analyze this feature, the ECL experiments were performed with different supporting electrolytes such as TBABF4, TBACF3SO3 and TBAP. It was found that the maxima of the intensity [page 45↓]as well as the time taken for its decay to zero for anodic ECL were roughly twice than for the cathodic ECL, in the former two supporting electrolytes. In TBAP, the cathodic ECL was nearly invisible in comparison to the anodic ECL. In addition, the anodic ECL showed a delay for the onset of the luminescence and decayed fast to zero. The results with the perchlorate were interpreted by assuming the ion pair formation between the radical cations and the perchlorate, with the rate constant on the order of 0.1 ms. However, there are several unsolved questions: How would a stable ionic association exist in acetonitrile which does not promote ion pair formation. Also pre-treatment procedures of the electrode have made changes in the ECL transients, which could not be explained based on the concept of ion association. It is also peculiar that perchlorate alone shows such an ion-pairing effect [64].

(4) The other possibility could be the adsorption of some species on the electrode other than the precursor molecule DPA, changing the kinetics of the electrochemiluminescence reaction. In fact, this had already been speculated and some analyses were made in this direction. The electrode material was changed to gold and glassy carbon. However, the responses were not reliable for analyzing this phenomenon. There were some electrochemical pre-treatments followed earlier, as, e.g., by changing the direction of the potential cycling prior the ECL reaction and polarizing the electrode at 1.3 V. The former could not be taken as a favorable pre-treatment procedure as the results were complicated [64]. In the latter process the potential of the Quasi Reference Electrode shifts due to the application of such a high voltage to the working electrode [65].

In earlier experiments, either a continuous rectangular AC voltage [66] or continuous potential pulses of unequal widths, i.e., tf is not equal to tr, were applied. Unequal pulse widths were preferred to verify the linearity between the current transient with an inverse square root of tr (or tr). This would result in the unequal concentration distribution of radical cations and anions, which would complicate the comparison of the ECL intensities. Hence, in this present study, the property of non-symmetrical ECL has been analyzed by varying the potential step width, magnitude, sequence and duration between the potential steps. The electrolyte was chosen to be TBAP¯acetonitrile, since this was reported to be the controversial system in the literature. Also the durations tr and tf were chosen to be the same, to provide similar conditions for the generation and distribution of radical cations and anions.


[page 46↓]

Under the conditions of normal potential steps (E+ = E+/0 + 250 mV, E- = E-0/-¯250 mV), of duration 5¯10 ms when applied continuously, only cathodic ECL of low intensity was observed (Fig. 3.1.4c). ECL could not be observed on reversing the potential step. When the waiting period (tw) between the successive double potential steps was increased, the intensity of the cathodic ECL also increased (Fig. 3.1.4d).

Fig. 3.1.4c: The potential step was made from 1.55 V (Ef) to - 2.05 V (Er). tf, tr = 0.01 s and tw = 0.02 s. No ECL was observed when reversing the potential step.

Fig. 3.1.4d: The potential step was made from 1.55 V (Ef) to -2.05 V (Er). tf, tr = 0.01 s and tw = 0.28 s.


[page 47↓]

On increasing the potential step width to 50 ms (tf and tr), the ECL intensity increased and anodic ECL was also observed. Unlike the cathodic ECL, the anodic ECL was independent of the waiting period (tw), which is shown in Figs. 3.1.4e and 3.1.4f.

Fig. 3.1.4e: The potential step was made from - 2.1 V (Ef) to 1.6 V (Er). tf, tr = 0.05 s and tw = 0.20 s.

Fig. 3.1.4f: The potential step was made from - 2.1 V (Ef) to 1.6 V (Er). tf, tr = 0.05 s and tw = 1.80 s.


[page 48↓]

Further on, at tf and tr = 50 ms, cathodic and anodic ECLs occurred with equal intensities under the appropriate magnitudes of the oxidation and reduction potentials. These results are contradicting the ones reported earlier which stated that no cathodic ECL was observed with TBAP. And also in the case of ECL experiments with another supporting electrolyte it was reported that the anodic ECL and cathodic ECL were of different intensities, the anodic being more intense than the cathodic.

Fig. 3.1.4g: Anodic ECL when the potential step was made from - 2.05 V (Ef) to 1.6 V (Er). tf, tr = 0.05 s and tw = 0.24 s.

Thus, the observation of equal intensities of anodic and cathodic ECLs in the TBAP-acetonitrile at a specific time scale in our experiments is a new result to be considered.


[page 49↓]

Fig. 3.1.4h: Cathodic ECL under the similar condition as in Fig. 3.1.4g; potential step was from 1.6 V (Ef) to - 2.05 V (Er). tf, tr = 0.05 s and tw = 0.24 s.

Equal intensities of anodic and cathodic ECLs could be observed under the same conditions as above with only changing the cation of the supporting electrolyte, to TEAClO4.

Fig. 3.1.4i: Anodic ECL for potential step from - 2.05 V (Ef) to 1.5 V (Er). tf, tr = 0.05 s and tw = 0.24 s.


[page 50↓]

Fig. 3.1.4j: Cathodic ECL under the similar condition as in Fig. 3.1.4g; potential step was from 1.5 V (Ef) to - 2.05 V (Er). tf, tr = 0.05 s and tw = 0.24 s.

These results show that polarizing the electrode at 0 V during the period tw = 240ms causes an increase in the ECL intensity. This could be due to the stripping off some electro inactive species from the electrode surface. The adsorption of this species should have occurred during the negative potential step, since only cathodic ECL shows variation with the increase in tw.

In an earlier publication, it was found that the addition of cyclohexanone in small concentration [0.05 M] makes the DPA anodic and cathodic ECLs to be almost similar in the cases of TBACF3SO3 or TBABF4 as the supporting electrolytes [65]. The hydrogen absorbed in platinum electrodes can be reduced to hydride in non-aqueous media when a sufficiently negative potential is applied. The hydride is very reactive and reacts with solvents to produce various products near the electrode surface [67]. At low temperatures ideal electrochemical behavior was observed in non-aqueous solvents, since the diffusion of hydrogen in Pt is sharply reduced as a consequence of its relatively high activation energy [68]. Cyclohexanone can react fast with hydride to produce its alkoxide, thus scavenging the hydride away. Thus, its addition improved the intensity and the decay of the cathodic ECL, where hydride could easily react with the DPA radical ions. This is in accordance with the present observation and thus, the effect of polarization of the Pt electrode at 0 V was interpreted as [page 51↓]a method of removing adsorbed hydrogen and thereby its interference with ECL.

3.1.5. The ECL Emission spectrum

The ECL was caused by the annihilation reaction (4.1.1c) between the radical cations and the radical anions of DPA [69 and 70]. Further evidence for the cause of electron transfer between the radical cation and anion for the generation of the excited state was provided by the effect of ionic strengths of the solvents [71] and the temperature effect [72] on the ECL. The free energy of the pair of radical ions compared with that of the ground state DPA pair was estimated to be 3.06 eV on the basis of the cyclic voltammogram in acetonitrile. The free energy of 1R* + R was calculated from the 0-0 transition energy in the fluorescence spectrum as 3.0 eV. The solvation energy was estimated as 0.4 eV based on the values for similar compounds in acetonitrile [73]. Thus, the formation of the singlet-excited state is in the 'normal region' (-ΔG° < λ) of electron transfer reactions.

Fig. 3.1.5a: ECL emission spectrum of DPA.

The experimental verification to find out the nature of the excited state is the ECL emission spectrum. The spectra of anodic and cathodic ECLs were taken by manipulating the waiting period (tw) to produce one kind of ECL at a time. The conditions of the potential steps were corresponding to that for the observation of equal intensities for anodic and cathodic ECLs. The spectra of [page 52↓]both were found to be identical with the emission maximum at 427 nm (≈ 2.91 eV) confirming that the emission was from the singlet excited state (Fig. 3.1.5a). The singlet excited state can also be generated by a triplet-triplet annihilation reaction, which can be analyzed by the effect of the magnetic field as discussed in the previous section.

3.2. Polymer phase ECL: MEH-PPV

The electrogenerated chemiluminescence (ECL) from a conducting polymer (CP) poly(2ethylhexyloxy 5 methoxy 1,4 phenylenevinylene), abbreviated MEH-PPV, was studied. The ECL experiment was carried out with the system configuration as electrode/polymer/solution. The effect of the supporting electrolyte on the ECL was examined.

poly(2-ethylhexyloxy-5-methoxy-1,4-phenylenevinylene) (MEH-PPV).

3.2.1. Experimental details

The polymer MEH-PPV was coated on the Pt electrode (with a surface area of approximately 0.09 cm2) by evaporating a solution of the polymer in CHCl3 in the dark. The Pt substrate was polished with 0.7 μm diamond paste, and the surface was checked for uniformity with a Nikon PFX 104 optical microscope before and after the deposition of the polymer. Electrochemical experiments were carried out in a three-electrode configuration with the polymer coated Pt as the working electrode (WE) and Pt wires as quasi-reference (QRE) and counter electrodes. Anhydrous CH3CN (Aldrich 27,100¯4) was used as the solvent and 0.1 M tetraethylammonium hexafluorophosphate (TEAPF6), tetraethylammonium tetrafluoroborate (TEABF4) or tetrabutylammonium tetrafluoroborate (TBABF4) were used as the supporting electrolytes. The [page 53↓]preparation of the electrolytic solution was done in dried Ar atmosphere following the procedure utilized by Orlik et al. [53], which was described in detail in section 2.2. Electrochemical measurements were accomplished with a homemade potentiostat. Potential pulses were applied by an Agilent 33250 A function generator via the potentiostat. ECL emission intensities were measured with an RCA photo multiplier tube (type number 7326), which was polarized at ¯1200 V by a Keithley 246 high voltage supply. The polymer coated Pt working electrode was cycled in the potential range from ¯1.9 V to +1.5 V vs. Pt (QRE) at 100 mV s¯1. The range of potential used in the sequential potential step experiment to generate the ECL was determined from the oxidation and reduction peaks in the cyclic voltammetry.

3.2.2. Cyclic voltammogram

The cyclic voltammograms were reproducible, when TEABF4 or TEAPF6 were used as the supporting electrolytes.

The ratio of the current peaks in the anodic and cathodic cycles (0.64 and 0.74, respectively) shows that the kinetics of the oxidation and reduction of the MEH-PPV polymer is not strictly reversible but nevertheless quasi-reversible. Thus, [page 54↓]similar to the case of the solution phase ECL process with DPA (chapter 3.1), the first requirement, the co-generation of positive and negative polarons, for the production of ECL is fulfilled with this polymer. Also the CV was devoid of any intervening peaks.

However, when the supporting electrolyte was changed from TEABF4 to TBABF4, which has a comparatively bigger tetrabutylammonium cation (TBA+), the polymer could not be reduced. The reason is the resistance offered by the polymer to the movement of bulkier ions through it. When the polymer is becoming reduced, it needs an equivalent amount of positive charged ions from the supporting electrolyte to counter balance the negative charge. If the positive ions cannot move inside the polymer, it cannot accept electrons from the electrode. This concept is discussed in detail in chapter 4. The cyclic voltammogram looks as follows:

Fig. 3.2.2.b: Cyclic voltammogram (CV) of Pt/MEH-PPV in 0.1 M TBABF4 in acetonitrile at a scan rate of 100 mV/s.

The MEH-PPV film used in Fig. 3.2.2a and Fig. 3.2.2b were of approximately the same thickness (1 μm). This shows how the size of the counterions, and hence the diffusivity of them, plays a major role in the electrochemistry of conducting polymers. Thus, the mechanism of charge transport in the polymer is different in the electrochemical environment than in the case of the solid-state electroluminescence process where such effects were not reported [74].


[page 55↓]

Some other conducting polymers were examined for ECL reaction. One is again a derivative of PPV, poly[bis(2,5-dimethyl-octylsilyl)-1,4-phenylenevinylene] (BDMOS PPV). The cyclic voltammetry showed irreversible oxidation of the polymer and no reduction at all. Thus, no ECL could be obtained from this CP, since feasible oxidation and reduction of the polymer are the foremost criteria for the generation of ECL.

Another polymer that was tried was poly(3-hexylthiophene) (PAT6). Upon oxidation the polymer was not intact on the Pt electrode. This caused instability in the cyclic voltammogram during oxidation, and no reduction could be observed. Even roughening of the electrode surface did not improve the adherence of the polymer with the electrode. These factors are the important ones in the selection of conducting polymers for ECL reaction.

3.2.3. Current transients

As with the case of DPA, the potential step was made to generate one kind of polarons. The current was linear with t¯1/2. The Cottrell equation (Eq. (2.1.2ib)) was used to evaluate the diffusion coefficient as 1x10¯13 cm2 s¯1 from the slope of the diffusion current (i) vs. the inverse square root of time (t¯1/2) plot.

Fig. 3.2.3a: Cottrell plot for Pt/MEH-PPV in 0.1 M TEABF4¯AN, when the potential step was made from 0 V to -1.9 V of duration 1 s.


[page 56↓]

3.2.4.  The ECL experiment

As with the case of DPA, the magnitudes of positive and negative potential steps were chosen from the formal potentials calculated from the cyclic voltammogram. Using a square wave alternating between the potential for the oxidation and reduction of the polymer (or vice versa), positive and negative polarons (similar to the cationic and anionic radicals in DPA) were generated. These species combine to form the luminescent moiety. The anodic ECL was obtained, when the potential was switched from negative to positive value, which is commonly referred to as ¯ to +; similarly, cathodic ECL is denoted as + to ¯. With the MEH-PPV polymer both, ¯ to + as well as + to ¯ were observed when TEABF4 was used as the supporting electrolyte in the acetonitrile solvent. These are shown in the following figures:

Fig. 3.2.4a: Anodic ECL when the potential was switched between ¯1.75 V and 1.45 V of duration 4 s each for MEH-PPV film on Pt in 0.1M TEABF4¯AN.

The ECL transient takes longer time for decaying (1 s) in the CP than the solution phase ECL with DPA (20 ms). This is due to the low transport coefficient in the conducting polymers (x10¯13 cm2 s¯1) compared to the diffusion coefficient in the solution phase (x10¯6 cm2 s¯1). This slower kinetics is one [page 57↓]important difference in the transport characteristics of the conducing polymers in comparison to that in the solution phase.

Fig. 3.2.4b: Cathodic ECL when the potential was switched between 1.45 V and ¯1.75 V of duration 4 s each for MEH-PPV film on Pt in 0.1 M TEABF4¯AN.

The anodic and cathodic ECLs shown in Fig. 3.2.4a and Fig. 3.2.4b have comparable intensities in the ECL experiments with MEH-PPV. The observation of symmetrical ECL is an important result of the present experiments with MEH-PPV, because in earlier experiments done with other polymers, such an observation had been difficult. This has to do with the relative stability of the positive and negative polarons [3]. In earlier studies, continuous pulsing between positive and negative potentials was done, without considering the stability aspect. As in the studies with DPA discussed in section 3.1.4, the widths of the potential steps and the waiting period were chosen to provide favorable conditions for the generation of positive and negative polarons. Also magnitudes of the potential steps were carefully chosen to prevent over oxidation/reduction of the polymer. Thus, it was possible to obtain symmetrical ECL .in our experiments. However, the kinetics was found to be different between the anodic and cathodic ECLs as seen from differences in the onset and decay of the ECL transients. This needs to be analyzed.


[page 58↓]

The experiments were done by varying the anion; using 0.1M TEAPF6 in acetonitrile. The ECL transients obtained are shown below:

Fig. 3.2.4c: Anodic ECL, when the potential was switched between ¯1.85 V and 1.35 V of duration 1 s each for MEH-PPV film on Pt in 0.1 M TEAPF6¯AN.

Fig 3.2.4d: Cathodic ECL, when the potential was switched between 1.35 V and ¯1.85 V of duration 1 s each for MEH-PPV film on Pt in 0.1 M TEAPF6¯AN.

Again the anodic (Fig. 3.2.4c) and the cathodic (Fig. 3.2.4d) ECLs are found to have comparable intensities but differ in the kinetics. The anodic ECL has more delay (≈120 ms) than the cathodic ECL (≈ 50 ms) for the onset of ECL. The [page 59↓]intensities of anodic and cathodic ECL with TEAPF6 are lower than those obtained with TEABF4. This could be due to the difficulty in purifying the former. Comparing Fig. 3.2.4a and Fig. 3.2.4c, we find that the relatively larger anion PF6 ¯ causes more delay in the onset of ECL than the smaller one, BF4 ¯. Further, one can find that even the cathodic ECL was suffering a larger delay in the presence of TEAPF6. This is because during the negative potential step at which cathodic ECL was observed, the charge compensation can be achieved by the movement of PF6 ¯ away from the polymer or the transport of the TEA+ towards it. Thus, an increase in the size of the anion affects the cathodic ECL also, due to the coupling between the counter ion movements themselves.

Fig. 3.2.4e: Comparison of anodic ECL and current decay for Pt/MEH-PPV in 0.1M TEABF4¯AN.

When the anodic current and ECL transients were compared, the decay of current seems almost instantaneous compared to the decay of the ECL in the conducting polymer (MEH-PPV in the present case). This is due to the slow transport of charges in the CP and the slow reaction kinetics for the annihilation reaction between the polarons - both are the parameters describing the ECL transient. This can be compared with the decay of the current and luminescence intensity in the solid-state electroluminescence (EL) process of the CP, in which both of them almost follow the same path [50]. This is because, in the ECL process, the transport is dominated by the slower moving counterion (Dap≤x10¯11 cm2 s¯1) from the supporting electrolyte in contrary to the usual electron (or hole) drift mode of charge transport in the CPs in the EL [page 60↓]experimental conditions (Dap≤ 10¯6 cm2 s¯1) Thus the ECL process in the CP differs from the solid-state EL process.


[page 61↓]

3.3.  Polymer phase ECL: DB-PPV

The electrogenerated chemiluminescence (ECL) from a conducting polymer (CP) 2,3- dibutoxy-1,4-polyphenylenevinylene (DB-PPV) was studied. The ECL experiment was carried with the system configuration as electrode/polymer/solution.

poly(2,3dibutoxy 1,4-phenylenevinylene) DB-PPV.

3.3.1. Experimental details

The polymer DB-PPV was coated on the Pt electrode (with surface area of approx. 0.07 cm2) by evaporating a solution of the polymer in CHCl3 in the dark. The Pt substrate was polished with 0.7 μm diamond paste, and the surface was checked for uniformity with a Nikon PFX 104 optical microscope before and after the deposition of the polymer. Electrochemical experiments were carried out in a three-electrode configuration with the polymer coated Pt as the working electrode and Pt wires as quasi-reference (QRE) and counter electrodes. Anhydrous CH3CN (Aldrich 27,100-4) was used as the solvent and 0.1 M tetraethylammonium perchlorate (TEAClO4), tetraethylammonium hexafluorophosphate (TEAPF6), tetraethylammonium tetrafluoroborate (TEABF4), or tetrabutylammonium tetrafluoroborate (TBABF4) were used as the supporting electrolytes. ECL emission intensities were measured with an RCA photo multiplier tube (type number 7326), which was polarized at ¯1000 V by a Keithley 246 high voltage supply.


[page 62↓]

3.3.2.  Cyclic voltammogram

The cyclic voltammograms were reproducible, when TEABF4 or TBABF4 were used as the supporting electrolytes. With TEAClO4 and TEAPF6 the oxidation and reduction were both irreversible, oxidation being the most affected reaction (Fig. 3.3.2a). The reason could be the difficulty in drying them.

Fig. 3.3.2a: Cyclic voltammogram (CV) of Pt/DB-PPV in 0.1M TEAClO4 in acetonitrile; scan rate: 100 mV/s.

Acetonitrile was found to be the best solvent for this purpose. The choice of the solvent was limited by the potential range and the purity of the solvent. As an example, when benzonitrile was used oxidation of the polymer could not be observed but there were multiple peaks. The presence of multiple electrochemical reactions is destructive for the production of ECL. Further, the high boiling point of benzonitrile (191 °C), as compared to acetonitrile (82 °C), caused difficulty in evacuating the vacuum lines between successive experiments.

The cyclic voltammetry did not show a marked difference between TBABF4 and TEABF4, when DB-PPV films of equal thickness were used (Figs. 3.3.2b and 3.3.2c).


[page 63↓]

Fig. 3.3.2b: Cyclic voltammogram (CV) of Pt/DB-PPV in 0.1 M TBABF4 in acetonitrile; scan rate = 100 mV/s.

Fig. 3.3.2c.: Cyclic voltammogram (CV) of Pt/DB-PPV in 0.1 M TEABF4 in acetonitrile; scan rate = 100 mV/s.

However, the anodic sweep was found to be less reversible than the cathodic one, which by itself, was quasi-reversible (Figs. 3.3.2d and 3.3.2e)

. [page 64↓]

Fig. 3.3.2e: Anodic potential sweep of Pt/DB-PPV in 0.1 M TBABF4 in acetonitrile; scan rate = 100 mV/s.

The cross-linked films did not show much difference from the uncross-linked polymer. But indeed, they show effect by the bulkier nature of the supporting electrolyte.[75]. When TBA+ was used instead of TEA+, there was no reduction. This is due to the restriction for the transport of bulkier cation (which is required for charge compensation in cathodic sweep) by the cross-linked polymer film.


[page 65↓]

Fig. 3.3.2f: Potential sweep of Pt/DB-PPV (thickness = 1μm, energy of cross-linking = 400 mJ/cm2) in 0.1 M TBABF4 in acetonitrile; scan rate = 100 mV/s.

3.3.3. Current transients

In the ECL experiments, the radical anions and cations are formed upon application of successive negative and positive potential steps. It was therefore considered essential to analyze the anodic and cathodic current transients [page 66↓]following the potential steps. The linearity between current and the 1/(square root of time) (i vs. t¯1/2), is a criterion for a transport controlled electrochemical process. Such plots are referred to as Cottrell plots which are shown in Figs. 3.3.3a, 3.3.3b and 3.3.3c. Note that at the shorter times (t-1/2 > 10 s-1/2), deviations from the linear behavior were observed that result from the charging current (section 2.1.2 iii). From the slope of these plots the apparent diffusion coefficient was estimated to be 2.637 x 10¯11 cm2 s¯1 using Eq. (2.1.2b) of chapter 2. The concentration term was estimated to be 6.098 x 10¯4 mol cm¯3 by assuming the dry density for DB-PPV as 0.6 g/cc and 4 DB-PPV units as one electroactive site. The area of the electrode is 0.07065 cm2.

Fig. 3.3.3b: Cottrell plot for the cathodic current on the negative potential step of the square wave between 0 and ¯1.7 V, with step width of 150 ms each, for Pt/DB-PPV in 0.1 M TBABF4 in acetonitrile.

The characterization of the electrode/polymer/solution also involves the evaluation of the double layer parameters: the double layer capacitance (Cdl) at the electrode/polymer interface and the uncompensated resistance (Ru) of the bulk consisting of the polymer and the electrolyte. Estimations of these double layer parameters were done by stepping the potential to a lower magnitude, where there will be no electrochemical reaction. The current in this region corresponds to the charging of the double layer at the electrode/DB-PPV (in the solution of the supporting electrolyte in acetonitrile). The magnitude of the current is related to the double layer capacitance and solution resistance as per Eq. (2.1.2c) in chapter 2 from the plot of ln (idl) vs. time.In a separate experiment [page 67↓]of stepping the potential to -1V, charging current was recorded against time. The resistance of the polymer/electrolyte system was calculated from the intercept of the plot of ln(idl) vs. time, as 1200 Ω and the double capacitance as 4 x 10¯6 F from the slope. The high value of resistance compared to that in the case of solution, e.g., DPA dissolved in the electrolyte, is due to the presence of the non-conducting polymer chains under this condition.

3.3.4. The ECL experiment

ECL was generated in the triple potential step chronoamperometry experiment. The oxidized and reduced forms of the conducting polymer were generated in succession using a square wave alternating between the potential for the oxidation and reduction of the polymer (or vice versa). These species combine to form the luminescent moiety.

Fig. 3.3.4a: Cathodic ECL when the potential was switched between 1.3 V and -1.85 V of duration 150 ms each for DB-PPV film of thickness ≈ 0.6 μm on Pt in 0.1 M TBABF4 - AN.


[page 68↓]

Fig 3.3.4b: Anodic ECL, when the potential was switched between -1.9 V and 1.3 V of duration 150 ms each for DB-PPV film of thickness ≈ 0.6 μm on Pt in 0.1M TBABF4-AN.

Fig. 3.3.4c: Anodic ECL when the potential was switched between - 1.9 V and 1.2V of duration 150 ms each for DB-PPV film of thickness ≈ 0.6 μm on Pt in 0.1M TEABF4-AN.

In polymers like MDOPPV and PAT6, it has been reported that of the anodic and cathodic ECLs, one is favored over the other [3]. In MDOPPV the cathodic ECL is more predominant than the anodic ECL, which is reversed in the case of PAT6. The behavior was explained based on the results from the [page 69↓]chronoabsorptiometry experiments. For PAT6 the anionic form was relatively stable, subsequently when the polymer was oxidized, anodic ECL could be observed. On the other hand, in MDOPPV, the reduced form was found to be unstable, which was the cause for less occurrence of anodic ECL. They have used different pulse widths for cathodic and anodic steps.

Fig. 3.3.4d: Scheme of the sequential anodic and cathodic potential steps in ECL experiments.

In the present case, equal pulse widths were used for anodic and cathodic potential steps with a waiting period tw between adjacent potential steps at which the potential was kept at '0' V. The potential step order was also reversed in the same way. This is to ensure identical experimental conditions for the generation of positive and negative polarons. It has been found experimentally that the ECL could be found when the potential step widths starting from 150 ms to 1 s (Figs. 3.3.4a., 3.3.4b. and 3.3.4c). Reproducible results were obtained at potential widths of 100 -150 ms. No preference was observed for cathodic or anodic ECL. However, we could find anodic and cathodic ECLs of almost equal intensity ( symmetrical ECL ), when the potential step width was of duration 150 ms for each and the waiting time between successive potential steps being 200 ms, (Figs. 3.3.4a and 3.3.4b). Thus, the absence of a preference for anodic or cathodic ECL in our experiments has to do with the time allowed for the existence of the polymer in the polaronic state. At 1 s only anodic ECL could be observed (Fig. 3.3.4c). This can be explained by lesser stability of positive polaron in comparison to the negative polaron, as seen from the CV. Thus, at longer potential width, (1 s in this case) the anion produced becomes decayed, so that there is nothing left to react with the negative polaron produced in the subsequent negative potential step. Thus, cathodic ECL could not be obtained in the + to - potential step experiment at longer step widths.


[page 70↓]

The onset of ECL had a delay to the maximum of 10 ms in the case of DB-PPV. Thus, the double layer parameters of the Pt/DB-PPV in the supporting electrolyte solution were calculated. The resistance was found to be 1200 Ω for a polymer film of thickness 2 μm and the capacitance was found to be 4 μF. The values differ from the ones we would expect from an electrode/organic solution interface (eg., Pt/DPA-SE), being larger by a factor of 10. DB-PPV was insoluble in acetonitrile in the neutral state. The higher resistance is obviously due to the presence of the polymer network on the electrode surface.

3.3.5. Studies concerning the stability of DB-PPV

Dissolution of the polymer was observed in the positive potential step. In fact, such dissolution was reported previously for other polymers [1]. This was a hindrance for the observation of ECL over a longer period. Fig. 3.3.5a shows decreasing intensity of the ECL transient with successive pulses. The dissolution has led to the change in the polymer film morphology before and after the ECL experiment.

Fig. 3.3.5a: Decrease in the anodic ECL intensity with successive potential steps.


[page 71↓]

Fig. 3.3.5b: Film morphology before the ECL experiment as seen with a microscope.

Fig. 3.3.5c: Film morphology after the ECL experiment as seen with a microscope.

To investigate whether there was complete dissolution of the polymer, which had resulted in the loss of its electroactivity, SEM (Scanning Electron Microscopy) image of the polymer coated electrode was taken after the ECL experiment.


[page 72↓]

Fig. 3.3.5d: SEM image of the DB-PPV coated electrode surface.

Fig. 3.3.5e: SEM image of the DB-PPV coated electrode surface at a higher magnification.


[page 73↓]

Fig. 3.3.5f: EDS analysis on the darker spots as seen in the SEM image.

The SEM pictures at two different magnifications (200 μm and 500 nm) showed the difference in the surface contour. Examination of the darker portion by EDS (Energy-dispersive X-ray Spectrometry) indicated carbonaceous material on the Pt electrode, which should be the polymer. Thus, there is polymer present on the electrode after the ECL experiment, which could not be felt by bare eyes, not even clearly under microscopic investigation. A scratch was made to analyze the interface between the polymer and the Pt electrode surface to examine the origin of dissolution, whether it is nearer to the electrode surface than the polymer bulk. But the SEM pictures did not give any insight into that aspect.

In order to prevent the dissolution, cross-linking of the polymer is one of the common procedures [76]. The DB-PPV was cross-linked using synchrotron radiation of energy in the range 100 - 500 mJ/cm2. The cross-linking, of course, improved the polymer stability. The degree of cross-linking was checked by coating a series of electrodes with the DB-PPV polymer of approximate thickness (0.63 μm), dipping them all into CHCl3 for 10 min. and measuring the charging current in 0.1 M TEAP-acetonitrile solution. This is directly proportional to the resistance of the polymer film, which is the parameter that can be influenced by cross-linking. However, generating ECL in cross-linked film was found to be difficult due to the slower counterion mobility inside the cross-linked polymer during the process of electrochemical reaction in the CP.


[page 74↓]

Fig. 3.3.5.g: Energy of cross-linking (mJ/cm2) used to cross-link the DB-PPV films against the magnitude of cathodic charging current.

The experiment was also done at low temperature of 5 °C with the hope that the deactivation kinetics of DB-PPV can be slowed down. Even this did not lead to observable ECL with cross-linked films. Hence, spectroscopic investigation of the polymer characteristics was undertaken. Ex-situRaman spectroscopy data shows that the polymer was intact upon spin coating or cross-linking by synchrotron radiation of energies specified above.


[page 75↓]

Fig.3.3.5i: Comparison of Raman spectra between the powder form and the cross-linked film of the polymer DB-PPV before the ECL experiment.

Fig. 3.3.5j: Comparison of the Raman spectra of the polymer DB-PPV before and after the ECL experiment.

But after the ECL experiment, there was a change in the spectrum when a thinner polymer film was used. The important feature is the change in the [page 76↓]magnitude and position of the doublets in the neutral polymer. The peak at 1599 cm¯1 corresponds to the aromatic C¯C bond, and that at 1634 cm¯1 to the vinylic C¯C bond. The intensity ratio between the two is 3:1, respectively. This shows that the aromatic system is actually a phenylic one. But after the experiment both the peaks were shifted to higher wave numbers 1613 and 1647cm¯1, respectively. This shift to higher energy and the equal intensities of aromatic and vinylic C¯C bonds has been attributed to quinonic structure [77]. So the polymer was disturbed during the double potential step experiment. It has been found that the positive polaronic and bi-polaronic levels co-exist in the PPV polymers [78]. The quinonic structure can be attributed to an irreversible derivative of the bi-polaronic level. The CV in the positive potential range is almost irreversible (Fig. 3.3.2e), confirming a chemical deactivation of the oxidized polymer. Further analysis suffers from the inability to characterize the product left on the electrode after the ECL experiment by spectroscopic techniques. The reason is the lesser quantity of the polymer on the electrode. Use of more polymer is limited by the slow ion transport.

Fig. 3.3.5k: DB-PPV in phenylic form.

Fig. 3.3.5l: DB-PPV in quinonic form.

It has been reported that EL of DB-PPV has more PL quantum yield (40 %) than that of MEH-PPV (15-20 %) [79]. This is also true in the case of ECL with these [page 77↓]polymers. The ECL signals from MEH-PPV could be detected when the PMT was biased -1200V , whereas it could be observed at the PMT bias voltage of .1000V. However, as it is seen from the results, the ECL process has complications due to the stability of the polymer. Such complications were not reported in the solid-state electroluminescence (EL) or photoluminescence (PL) experiments [80]. So the electrode/polymer/solution configuration and the transport properties associated with the ECL experiments could be more sensitive to the polymer characteristics. In DB-PPV, the two butoxy substituents are in the ortho positions to each other. The presence of two ortho substituents has two effects: 1. The steric repulsion between the two butoxy groups ¯ makes the substituents to stay away from each other, which might be the cause for the porous appearance of the film. This is because poor orientation of the substituents was found to affect the film forming property of these polymers [81]. 2. The adjacent alkoxy groups repel from each other. As a result, the phenyl rings are tilted at an angle of 30° with respect to each other, as found from theoretical quantum chemical calculations for the ortho substituted PPVs. This strain in the aromatic system can result in the less conversion of the neutral form to the polaronic form as well as to their low stability. This explains why we could not go to higher potential step widths. The longer the duration of the potential step, the lesser will be the polarons remaining intact in the polymer. Since positive polarons are found to be less stable, we could not observe cathodic ECL, when the potential step widths were of magnitude 1 s. When the potential was switched between positive to negative steps of duration 1s, the positive polarons become deactivated in the first potential step itself, thereby not yielding any ECL in the subsequent negative potential step. The lower limit of 150 ms for the potential step width is a consequence of the solid-state transport process, i.e., the slower ionic transport within the polymer.

3.3.6. Energetics of the ECL in DB-PPV

In order to understand the energy of the ECL process, the ECL emission spectra were recorded. The emission spectra from anodic and cathodic ECLs occur at the same wavelength. This shows that both anodic and cathodic ECLs are generated by the same process, the recombination of positive and negative polarons.


[page 78↓]

Fig. 3.3.6a: Cathodic ECL spectrum.

Fig. 3.3.6b: Anodic ECL spectrum.


[page 79↓]

Fig. 3.3.6c: Absorption spectrum of DB-PPV in CHCl3.

The wavelength of ECL emission was approximately 485 nm. This energy corresponds to the energy of the singlet state. This was compared with the emission spectrum from the solid state Electroluminescence (EL) process, where there will not be any liquid phase in contact with the polymer. The value of emission maximum reported for this polymer was 519 nm [82,83]. The polymer was also dissolved in CHCl3 and absorbance and fluorescence studies were made. The spectra are shown below:

Fig. 3.3.6d: Flourescence spectrum of DB-PPV in CHCl3.

The emission maximum seems to be different in the solid-state EL process, in [page 80↓]the ECL process where the solid polymer was kept in contact with electrolytic solution, and in the solution phase fluorescence process. The difference between the first and the others can be speculated, since solvent plays a major role in stabilizing the ground and excited states of an aromatic compound, which will in turn affect the energy of the emission. The difference between the emission maxima between ECL and fluorescence needs some consideration. The emission maximum occurs at approximately 500 nm in the solution phase. ECL occurs at a shorter wavelength in acetonitrile, 485 nm. One reason could be that in a good solvent such as CHCl3 of low dielectric strength, the polymer can exist with extended conjugation, which shifts absorption and emission to longer wavelengths [84]. More detailed explanation can be given from the knowledge about the nature of the ground and excited states. The electronic state having lower dielectric strength will be stabilized by CHCl3.


© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.
DiML DTD Version 3.0Zertifizierter Dokumentenserver
der Humboldt-Universität zu Berlin
HTML generated:
23.02.2004