Motivation

There are two parts in this dissertation. First, four small molecule donor material with D-A-A structure configuration were employed in organic solar cells.

Compared to our previous works, end acceptor of these materials were modified from dicyanoethylene into single cryno group⁸⁵. The HOMO levels were higher lying for triggering the charge transfer and separation state. Besides, the thermal properties of these materials were being improved. These four materials were provided by Dr.

Hao-Chun Ting and Prof. Keng-Tsung Wong’s group, Department of Chemistry, National Taiwan University. To pursing a high efficiency, a series of optimization in OSCs were taken with bulk heterojunction structure.

Second, exciton dynamics of amorphous rubrene thin film with different thicknesses was studied under different temperature and excitation energy by TrPL, supported by Dr. Juen-Kai Wang, Center for Condensed Matter Sciences, NTU, and Institute of Atomic and Molecular Sciences, Academia Sinica. A rate equation was

16

used to explain the dynamics and the parameters (such singlet fission rate, triplet fusion rate, singlet-singlet annihilation rate…) was quantitatively extracted from our TrPL results.

17

2 Chapter 2 Experiments

2.1 Introduction

In this chapter, all the fabrication and measurement systems in this thesis will be introduced.

Thin-film structures for OSC and SF studies in this thesis we obtained by thermal evaporator under high vacuum. Suitable patterning and treatments were applied before the thin-film process. And the encapsulation process was performed directly after the thin-film evaporation.

For OSC thin-film and devices, J-V characteristics under dark and 1-sun illumination was performed for calculating device performances such as V_OC, J_SC, R_s, Rsh, FF, and PCE. Recombination mechanism in our OSCs was studied by varying the illumination intensity of solar simulator. External quantum efficiency and absorption spectra were measured and hence internal quantum efficiency can be obtained.

Besides, photoelectron spectroscopy and elliposometry measurements were performed for the energy levels and optical constants of organic thin-films, respectively.

In our SF study, transient trPL was used with various laser pulse energy under different sample temperatures (77K to room temperature).

2.2 Device fabrication and measurement systems for OSC

2.2.1 Substrate patterning

Glass substrates with 150-nm indium tin oxide (ITO) were used (Lumtec Corporation, resistance 15 Ω/square). ITO pattern was obtained through

18

photolithography and etching processes in the clean room.

First we cleaned the substrates by detergent (DI water with detergent 10:1), acetone and isopropyl alcohol (IPA) consecutively each for 10 min and dried it with nitrogen gun. Then spin the positive photoresist (PR, S1813) on the substrates with two different rotation speed and time (550 and 1250 rpm for 5 and 25 sec respectively). followed by 10 min hard-bake to solidify the film. UV exposure (for 13 sec with photo masks followed by developine process (MF 319) for 13 sec was used to obtain the PR pattern. After that, soft bake was applied by remove the redundant vapor. Then, aqua regia (mixed by HCl: HNO₃= 3:1) was used to etch the ITO without PR. Finally, acetone was used to remove the PR in the substrate. Then, patterned substrates were cut into 1.78 × 2.78 cm² for the following processes.

2.2.2 Device fabrication

All the devices were fabricated in the thermal evaporate system under a high vacuum (1×10^-6 torr) which included 7 thermal cells for small molecule organic materials and 3 boat for metal and C60. Evaporation rates for materials were monitored (Mextek 350, 400) by the sense of quartz crystal microbalance (QCM).

With proper shadow masks for both organic materials and metal, active region can be defined for 0.04 cm² which contained three pixels on a substrate. Samples were then transferred to a 99.999% (5N) N2 glove box with the concentration of O2 and H2O below 0.1 ppm. After spreading the UV glues at the peripherals of the cover glass and attaching it with the thin-film substrates, samples were illuminated by UV lamp for 12 minutes to solidify the UV glues. Such encapsulation provided reliable storage lifetime at least over two weeks.

19

Fig. 2-1 Device configurations.

2.2.3 Device performances measurement

In device performances measurement, OSCs were illuminated by a class-A solar simulator (Newport Model 94022a) under a standard condition (1-sun, 100 mW/cm², air mass (AM) 1.5G) as shown in Fig. 2-2. The solar simulator was calibrated regularly to prevent from the decay of the lamp (155W) with measuring distance of 2.5 inches. Hence, uniform illumination region with precise illuminance can be achieved. Manual shutter was used when measuring the dark current of the OSCs, which were connected to power meter (Keithley 2400). A computer was used to control the power meter via GIPB interface for obtaining the J-V characteristics and analyzing the device parameters such as VOC, JSC, RS, RSH, FF and PCE. The solar simulator system is supported by Prof. Tien-Lung Chiu in Yuan Ze University, Taoyuan Taiwan.

20

Fig. 2-2 Set up of device performances measurement.

2.2.4 Measurement of external quantum efficiency (EQE)

Setup of EQE measurement was shown in Fig. 2-3. A halogen light source (Newport Model 66901) supported by power supply (Newport Model 69907) gave a continuous white light source which passed the monochromator (Oriel Conerstone130 1/8m) with a precise grating working for 200-1200 nm. Then it passed through a chopper with the rotation speed 350 Hz, controlled by a chopper controller (Stanford Research system SR540) which connected with lock-in amplifier (SR-830) to eliminate the noise from environment. After that a filter wheel (Newport 74040) was constructed to cancel the 2nd diffraction from grating in monochromator. A silicon photodiode was employed to calibrate the input photon density (count the photon number). Then the photocurrent was measured by lock-in amplifier (count the numbers of electron) and sent the information to computer via GPIB interface and the software was procvided by Forter Technology Corporation. EQE spectra can obtained by the ratio between photon and electron numbers.

21

Fig. 2-3 Set up of EQE measurement.

2.2.5 Absorption spectrum measurement

To gain the value for internal quantum efficiency (IQE), we have to measure the absorption spectra for the devices. A commercial spectrometer U4100 by Hitachi was employed and gives a wide range for absorption measurement from 240 nm to 2600 nm. Due to the zero transmission of metal cathode in OSC devices, we can obtain the absorption spectra from reflection spectra. And IQE spectrum can be achieved by using the EQE spectrum dividing the value for absorption spectrum from the equation below. The spectrometer were supported by Man-Kit Leung's Lab in Chemistry Department of National Taiwan University.

EQE (%)

abs (%) = IQE (%)………(2)

2.2.6 Measurement of optical constants

We can realize the optical characteristic and packing situation for molecules under the measurement of ellipsometry. In this research, a variable incident angle ellipsometry SOPRA GESP5 (Gonio-Ellipso-Spectro-Photometer) which was support by Radiation Technology was employed, shown in Fig. 2-4. For ellipsometry system,

22

we detected the elliptically polarized reflect light from the dielectric organic thin film by input a plane polarized beam. With extracting the information for amplitude and phase of the polarization from neat film which were deposited on the n-type silicon substrate with reflection mode, we can simulate and scan by these data then get the values for anisotropic refractive index and extinction coefficient accurately in scanning mode, without destroyed the thin film morphology. This variable angle goniometer constructed by an analyzer and polarizer (Xe lamp include) arm, which can be mounted by stepper meter and high-resolution monochromator, photon counting detector is contained. During the scan, the detected signal were collected into a optical fiber which connected to a monochromator.

Fig. 2-4 Set up of variable incident angle ellipsometry

2.2.7 Sun variation system

A sun variation experiment is powerful to give the standpoint for exciton and carrier behavior because a power dependent can separate the recombination between exciton and charges. The system is built as shown in Fig. 2-5. A Wacom light source for radiative parallel luminous of high luminous (Wacom HX-504) with a tunable input current supported (Wacom XDS-501SG) had ability for varied the incident

23

power from 0.001 to over 1sun.

The value of sun can be accurately measured and calibrated by a silicon diode (Model 99150 V). By tuning the wanted sun value detected by reference cell, our solar cells were replaced on the holder. All the measurements were taken under the dark room to preventing the noise from the environment, especially under low intensity measurement. And to avoid the thermal formation during the measurement, each test will place over 1 min to cool down when the measurement is too frequently.

The system is also supported by Prof. Tian Lung Chiu in Yuan Ze University, Taoyuan Taiwan

Fig. 2-5 Set up of Sun variation system

2.3 Sample fabrication and measurement systems for exciton

dynamics

2.3.1 Samples fabricated for exciton dynamic investigation

All the samples were fabricated under the same evaporation system as mentioned in section 2.2.2. Rubrene and LiF thin film were deposited on the quartz and glass substrates under high vacuum, which the quartz purity over 99.99% and the

24

transmission also greater than 95% in visible region. The samples then transfer to the nitrogen glove box to seal it with UV glues.

2.3 2 Steady state and time-resolve photoluminence (TrPL) at different temperatures

In this research, exciton dynamic was studied by the measurement of steady state photoluminence and TrPL. The system is established and supported by Dr. Pin-Hao Sher from Prof. Juen-Kai Wang's group in Institute of Atomic and Molecular Sciences, Academia Sinica, Taiwan and the detail construction was shown in Fig. 2- 6.

A Nd:VAN (High Q IC-1064-15000) laser source with 1064 nm emission wavelength was employed in this research. Laser pulsewidth was 6.5 ps in full width at half maximum (FWHM) with its origin repetition rate 76 MHz. The pulse passed through a second harmonic generation (SHG) crystal which generated 532 nm pulses.

Then the radio frequency (RF) acoustic pulse selector changed the repetition rate by by producing a periodic wave grating to reflect the unwanted pulses. The repletion rate we chose was 170 kHz. A first 4X lens and aperture served the function of space filter to eliminate the frequency noise from the light source. A combination for half wave plate and broad band polarizer (SM1PM10) was used to vary the input intensity.

By using the power meter diode (818-UV) and optical power meter (1916C, Newport) before the microscope, an accurate intensity density (in terms of W) can be measured. It can monitor and focus the sample by microscope (Olympus BX61W1) in front of sample. After passing the long pass filter, the reflecting fluorescence signal coming from samples went through the same path and back to the detector. To detect the signal, one can obtain the steady state behavior by detecting the signals with Charge-coupled Device (CCD) (Andor DU920P BR-DD) or transient characteristic by Time-Correlated Single Photon Counting (TCSPC) system with reverse mode to delay

25

the time, all the fluorescence was collected by monochromator (Horiba Jobin Yvon MicroHR).

Fig. 2-6 Set up of steady state and time-resolve photoluminence (TrPL) In our measurement, samples were mounted on a stage with temperature control system from 77K to room temperature. Here, samples attached to the special cryo carrier (Cryo industries CFM 1738-X6M102, 5119) as illustrated in Fig. 2-7. To provide a better thermal conduction, silver glue was applied between copper holder in cryo carrier and samples. The base pressure of the measurement chamber is 1×10^-5 by turbo pump (Varian vacuum tec TPS-compact). Liquid nitrogen was used to cool down the system to 77 K. Under the temperature controller (Lakeshore 331), the heater can control the temperature in carrier.

26

Fig. 2-7 Temperature-controlled sample holder.

27

3 Chapter 3 Optimization of bulk heterojunction OSC for

D-A-A configuration molecule with single cyano group as

the electron donor material

3.1 Introduction

In this chapter, four novel p-type electron donor materials, DTCPB, DTCTB, DTCPBO and DTCTBO were investigated and fabricated into OSC. These compounds were synthesized and supported by Prof. Keng-Tsung Wong’s group, Department of Chemistry, National Taiwan University. Compared with the previous work⁸⁶, dicyanovinylene (DCV) groups were replaced by single cyano in our study to provide larger ΔELUMO between electron donor and acceptor materials, which improves exciton dissociation at donor/acceptor interface and reduce recombination.

With systematically engineering the OSC device structures based on these four electron donor materials and C60 and C70 as the electron acceptor materials, we found that the optimized ones were bulk heterojunction due to their superior FF values. In our optimized device structures, power conversion efficiencies of DTCPB, DTCTB, DTCPBO and DTCTBO based OSCs were 6.55%, 4.40%, 5.98%, 4.65%, respectively.

Incident illumination power was varied for studying the recombination characteristics of OSCs with different electron donor materials.

3.2 Photophysical properties of the four single cyano group electron

28

donor materials

Fig. 3-1 Molecular structures of (a) DTCTB, (b) DTCPB, (c) DTCPBO, (d) DTCTBO and (e) DTDCPB.

Fig. 3-1 shows the molecular structures of DTCPB, DTCTB, DTCPBO and DTCTBO, respectively, which were originally modified from the compound 2-{[7-(4-N, N-ditolylaminophenylen-1-yl)-2,1,3-benzothiadiazol-4-yl]methylene}

malononitrile (DTDCPB), as shown in Fig. 3-1 (e). All these organic materials were synthesized by Prof. Keng-Tsung Wong’s group. With connecting a single cyano moiety attached as the electro-withdrawing end-group, LUMO value can be easily raised up (1 eV higher than acceptor) while maintaining a HOMO value which provides a suitable D-A electro-withdrawing ability compared to an over-strong one

29

in DTDCPB to prevent the formation of dipoles. The entire chemical configuration for four compounds with single cyano groups were shown in Fig. 3-1 However, there is one disadvantage for these four materials that the wide bandgap characteristics (higher LUMO with the same HOMO, compared to DTDCPB) results in blue-shift in absorption spectra combined with the lower absorption ability.

The differences between these four compounds are illustrated below: from the electron donor endcap, the electron-rich and fortified quinoidal characters of thiophene functional groups of DTCTB and DTCTBO result in the redshift (50-70 nm) in absorption spectra, compared to phenelyne attached molecules, DTCPB and DTCPBO. Then, considering the central bridging electron-withdrawing unit, benzoxadiazole (BO) and benzothiadiazole (BT), absorption spectra of BO-based molecules (DTCPBO and DTCTBO) are redshifted compared to those of BT-based ones (for DTCPB and DTCTB).and the HOMO levels were also deep-lying. Fig. 3-2 shows the absorption spectra of these four molecules in solution (CH3Cl) and thin films (by thermal evaporation).

Fig. 3-2 Absorption spectra of the four molecules (a) in solution, and (b) thin films.

The absorption peak for DTCPB, DTCPBO, DTCTB and DTCTBO are 491, 518, 563 and 583 nm for solution and 511, 537, 583, 600 nm for thin film, respectively.

30

HOMO values of these thin film measured by photoelectron spectrometer are shown in Fig. 3-3. The phenylene-containing groups help to lower the HOMO value respect to thiophene by ~0.1 eV, this shows a similar result in BO and BT moiety. All the parameters and thermal properties for intrinsic parameters were reorganized in the Table. 3-1 and Table. 3-2.

Fig. 3-3 HOMO of DTCPB, DTCPBO, DTCTB and DTCTBO thin films measured by photoelectron spectrometer.

Table. 3-1 Basic characteristics of DTCPB, DTCPBO, DTCTB and DTCTBO in solution.

Table. 3-2 Basic characteristics of DTCPB, DTCPBO, DTCTB and DTCTBO

31

3.3 OSC optimization of four electron donor materials with single

cyano substituent molecules

In this section, we illustrate the optimization procedures of OSCs by using these four electron donor materials with C60 and C70 electron acceptors. We optimized DTCTB first, followed by DTCPB, DTCPBO, and finally DTCTBO in sequence. In our previous research, we found that the optimized OSC structure consisted of an active layer and an electron transport layer (planar-mixed heterojunction structure, PMHJ)⁸⁵. Besides, donor buffer layer between anode and active layer can effectively prevent exciton quenching and increase J_SC and PCE in OSCs. Hence, we started our device architecture from PMHJ. However, interestingly, in OSCs based on these four electron donor materials, device performances were even better when using simple bulk-heterojunction (BHJ) structure. Besides, insertion of donor buffer layer showed worse efficiency in these OSCs. Hence, the device structure was very simple with two variables, mixing ratio and active layer thicknesses.

In our devices, we used MoOx and bathocuproine (BCP) as the hole extraction layer and exciton blocking layer, respectively. The workfunctions of electrodes and energy levels of materials were shown in Fig. 3-4.

32

Fig. 3-4 Energy diagrams of four single cyano groups moieties molecules with C60 and C70 under simple bulk heterojunction

3.3.1 Comparison between PMHJ and BHJ of DTCTB devices

3.3.1.1 C

₆₀

based OSC

First of all, we fixed our thickness of active region to 60 nm by changing the ratio thickness between mixing layer and acceptor C₆₀. In such a design, the optical field distribution did not change a lot. The mixing ratio was fixed at 1:1.6 ^[85].

Table. 3-3 Device configurations from A-1 to A-4 of DTCPB:C60 OSC. The unit is nm.

Device MoO₃ DTCTB:C₆₀ C₆₀ BCP Al

A-1

20

1:1.6 (Mixing ratio), 20 40

7 100

A-2 1:1.6 (Mixing ratio), 30 30

A-3 1:1.6 (Mixing ratio), 40 20

A-4 1:1.6 (Mixing ratio), 50 10

33

J-V characteristic of devices from A-1 to A-5 under dark condition and 1-sun AM 1.5 G solar illumination was shown in Fig. 3-5 and the device performance was shown in Table. 3-4. As the thickness of mixing layer increased, J_SC increased, RS decreased, and PCE increased, which meant C60 layer should be as thin as possible.

(a) (b)

Fig. 3-5 J-V performances of devices from A-1 to A-4 under (a) dark condition and (b) 1-sun solar illuminations.

Table. 3-4 Performances of devices from A1 to A-4 under 1-sun solar

34

(a) (b)

(c)

Fig. 3-6 (a) EQE, (b) IQE and (c) absorption spectrum of devices from A-1 to A-4.

3.3.1.2 C

₇₀

based OSC

Here, we attempt to apply another acceptor, C₇₀ to enhance our PCE. In our previous work⁸⁵, although C70 can induce a higher JSC, the electrical property was poor compared to C₆₀ so we can introduce a transporting by higher mobility C₆₀ while mixture with C70. In this section, we varied the transporting layer as C60 or C70, or give a replacement with a mixing layer for preventing from the variation of optical field distribution, in other words, to compare the PMHJ and BHJ in C70 system.

35

Devices configurations and performances were shown in Table. 3-5 and Table. 3-6.

The J-V performances for devices of A-5 to A-7 under dark condition and illuminations were shown in Fig. 3-7

Table. 3-5 Device configurations from A-5 to A-7 of DTCPB:C₇₀ OSC. The unit is nm.

Fig. 3-7 J-V performances of devices from A-5 to A-7 under (a) dark condition and (b) 1-sun solar illuminations.

Table. 3-6 Performances of devices from A-5 to A-7 under 1-sun solar

36

A-7 0.75 9.63 47.02 3.41 0.44 15.90

Among these three devices, one can see that BHJ device (A-7) exhibited highest RSH, FF, and PCE, which meant the introduction of acceptor layer (C60 or C70) resulted in carrier accumulation and carrier recombination.

(a) (b)

(c)

Fig. 3-8 (a) EQE, (b) IQE and (c) absorption spectrum of devices from A-5 to A-7.

3.3.2 Optimization of different mixing ratio and thickness of active

layer for DTCTB

37

3.3.2.1 C

₆₀

based OSC

Then, we tuned the mixing ratio of electron donor and acceptor material, as shown in Table 3-7. J-V characteristic of devices to under dark condition and 1-sun AM 1.5 G solar illumination was shown in Fig. 3-9 and performances were shown in Table. 3-8. One can see that the optimized mixing ratio is DTCTB:C₆₀ = 1:2.2 (A-10).

Comparing devices A-10 and A-13, C60 neat acceptor layer should be replaced by mixing layer for improving PCE.

Table. 3-7 Devices configuration of devices from A-4, A-8 to A-13. The unit is nm.

Device MoO₃ DTCTB:C₆₀ C₆₀ BCP Al

A-4 1:1.6 (Mixing ratio), 50 10

100 A-8 1:1.6 (Mixing ratio) , 60 10

A-9 1:1.9 (Mixing ratio) , 60 10

A-10 1:2.2 (Mixing ratio) , 60 10 7

A-11 1:2.4 (Mixing ratio) , 60 10

A-12 1:2.5 (Mixing ratio) , 60 10

A-13 1:2.2 (Mixing ratio) , 70 0

(a) (b)

Fig. 3-9 J-V performances from of device A-4 to A-12 under (a) dark condition

38

and (b) 1-sun solar illuminations.

Table. 3-8 Performances of devices from A-4, A-8 to A-13 under 1-sun solar

39

(b) (c)

Fig. 3-10 (a) EQE, (b) IQE and (c) absorption spectrum of devices A-8 and A-13.

3.3.2.2 C

₇₀

based OSC

In this section, we optimize the DTCTB: C₇₀ by varying the mixing ratio and thickness. The optimized condition is DTCTB: C70 = 1: 2.6 with the thickness of 70 nm, as shown in Fig. 3-10 and 3-11 and Tables 3-9 and 3-10.

Table. 3-9 Device configurations for different ratio and thickness of active layer of A-7, A-14 to A-17 of DTCTB:C₇₀ based OSC. The unit is nm.

Device MoO₃ DTCTB:C₇₀ C₇₀ BCP Al

A-7

20

1:2.2 (Mixing ratio) , 70

0 7 100

A-14 1:2.6 (Mixing ratio) , 70 A-15 1:3.0 (Mixing ratio) , 70 A-16 1:2.6 (Mixing ratio), 60 A-17 1:2.6 (Mixing ratio), 80

40

(a) (b)

Fig. 3-11 J-V performance of devices from 3-20 to 3-22 under (a) dark condition and (b) 1-sun solar illuminations.

Table. 3-10 Devices performance of different mixing ratio and thickness of active layer for DTCPB/ C70 based OSC.

Name VOC

41

(a) (b)

(c) (d)

(e) (f)

Fig. 3-12 (a) EQE, (b) IQE and (c) absorption spectrum of devices A-7, A-14 to A-15 with different mixing ratio and (d) EQE, (e) IQE and (f) absorption spectrum of devices from A-14, A-16 to A-17 with different thickness of active layer.

42

3.3.2.3 Replacement of different blocking layer for DTCTB: C

₆₀

based OSC

Here, we employed several kinds of blocking layer in order to block the exciton from anode side and promote the hole transport ability. These blocking layer showed different HOMO values in Fig. 3-13. The devices configurations and detail performances were also present in Table. 3-11 and Table. 3-12, respectively. The J-V performances for devices from A-18 to A-25 under dark condition and illuminations were shown in Fig. 3-14. However, the insertion of blocking layer resulted in higher R_S, and little enhancement (if any) on J_SC, which resulted in worse PCE.

Table. 3-11 Device configurations for different blocking layers applied on champion device for DTCTB:C60 based OSC. The unit is nm.

Device MoO3 Blocking layer DTCTB:C60 C60 BCP Al

43

Fig. 3-13 Configurations of devices and energy diagrams for different blocking layer of DTCTB:C60 based OSC.

(a) (b)

44

(c) (d)

(e) (f)

Fig. 3-14 J-V performances from A-18 to A-25 under (a) mCP inserted dark condition and (b) 1-sun solar illuminations, (c) NPB inserted under dark condition and (d) 1-sun solar illuminations and (e) DTDTB inserted under dark condition and (f) 1-sun solar illuminations.

Table. 3-12 Devices performances of for different blocking layer of DTCTB:C60

Table. 3-12 Devices performances of for different blocking layer of DTCTB:C60

在文檔中 具腈基小分子有機太陽能電池之優化與紅螢烯薄膜激子動態特性之研究 (頁 35-0)