Heterojunction Area-Controlled Inorganic Nanocrystal Solar Cells

Subscriber access provided by University of Sunderland

Functional Inorganic Materials and Devices

Heterojunction Area-Controlled Inorganic Nanocrystal Solar Cells Fabricated Using Supra-Quantum Dots Juwon Park, Sungjae Hwang, Sanghwa Jeong, Sungjee Kim, Jiwon Bang, and Seungho Cho ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b14752 • Publication Date (Web): 09 Nov 2018 Downloaded from http://pubs.acs.org on November 9, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Heterojunction Area-Controlled Inorganic Nanocrystal Solar Cells Fabricated Using Supra-Quantum Dots Juwon Park,† Sungjae Hwang,† Sanghwa Jeong,† Sungjee Kim,† Jiwon Bang,‡,* and Seungho Cho$,* † Department

of Chemistry, Pohang University of Science and Technology (POSTECH), 77

Cheongam-ro, Namgu, Pohang 37673, Republic of Korea. ‡

Electronic Conversion Materials Division, Korea Institute of Ceramic Engineering and

Technology, Jinju 52852, Republic of Korea. $

School of Materials Science and Engineering, Ulsan National Institute of Science and

Technology (UNIST), Ulsan 44919, Republic of Korea.

KEYWORDS: inorganic nanocrystal, self-assembly, supra-quantum dot, photovoltaic device, heterojunction area, domain size

ABSTRACT A supra-quantum dot (SQD) is a three-dimensional structure formed by the attachment of quantum dots. The SQDs have sizes of tens of nanometer and they maintain the characteristics of the individual quantum dots fairly well. Moreover, their sizes and elemental compositions can be tuned precisely. Based on their unique features, in this work, SQDs are used as constituents of the interpenetrating photoactive layers of inorganic nanocrystal p–n heterojunction solar cells in

ACS Paragon Plus Environment

1

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 21

order to control the p-type and n-type domain sizes (i.e., p–n heterojunction areas) for optimizing the charge-carrier collection. SQD-containing p–n heterojunction solar cells exhibit improved charge transport and thereby, higher power conversion efficiency (PCE) (3.03%), owing to their intermediate p-type and n-type domain sizes, which are between those of a bilayer nanorod p–n heterojunction solar cell (PCE: 1.21%) and an interpenetrating nanorod p–n heterojunction solar cell (PCE: 2.40%). This work demonstrates that the self-assembly of nanoscale materials can be utilized for tailoring the spatial distributions of charge carriers, which is beneficial for obtaining an enhanced device performance.

Introduction Semiconducting nanocrystals (NCs) have attracted considerable attention as key materials in optoelectronic devices such as light-emitting diodes,1-2 field-effect transistors,3-4 photodetectors,5-6 and photovoltaic devices,7-11 because of their distinctive optical and electrical properties. Colloidal NCs offer several advantages over the conventional bulk materials, such as solution processability and tunability of quantum size effect, which allow large-scale and low-cost fabrication of photovoltaic devices with optimized optical absorption properties.12-14 The performances of the all-inorganic NC photovoltaic devices have been improved by engineering the NC active-layer architectures using Schottky junction,15-16 depleted17-20 (bulk21-22) p–n heterojunction, or quantum funnel.23 Great efforts have been made to achieve a certified power conversion efficiency (PCE) exceeding 10%, in the recent years.19-20,22 Owing to the electronic spatial asymmetries created by the junctions, p–n heterojunction NC solar cells show clear improvements in charge separation, when


2

Page 3 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


compared to those solar cells composed of single-material NCs.24-26 However, restricted junction areas of bilayer p–n heterojunction cells can impede efficient charge separation. In order to increase the junction areas, blends of p-type and n-type NCs have been used as photoactive layers. Rath et al. reported solar cells based on blends of n-type Bi2S3 and ptype PbS NCs, which exhibit more than three-fold improvement in the device performance, compared to the corresponding bilayer devices.21 Yang et al. recently reported high-efficiency bulk heterojunction quantum dot (QD) solar cells (PCE: 10.4%) fabricated by spin-casting n-type and p-type PbS QD mixed ink. The homogenous blended p- and n-type nanoscale domains show efficient charge separation at nanoscale interfaces and can overcome the short carrier diffusion lengths in the photoactive films.22 However,

excessively

large

heterojunction

areas

can

cause

the

following

disadvantages: (i) Too many charge-separation events can lead to charge-carrier accumulation, giving rise to carrier leakage and recombination. (ii) Small p-type and/or ntype domain sizes can inhibit depletion-layer (i.e., built-in potential) formation because of spatial confinement, reducing the open-circuit voltage (VOC).24,27-28 Therefore, controlling p-n heterojunction areas is important for achieving high device performance. However, finding proper methods to control the heterojunction areas remains a big challenge. In this work, we demonstrate p-n heterojunction area control in CdTe/CdSe NC solar cells, using self-assembled QDs. The solar cells have CdTe NCs as electron acceptors and CdSe NCs as electron donors, based on their cascaded band-edge alignments (i.e., the conduction and valence band edge positions of CdTe are higher in energy than the corresponding ones of CdSe),24,28 and p-n heterojunction is formed at the CdTe/CdSe interface.29 We evaluate the performances of photovoltaic devices with controlled


3


Page 4 of 21

interfacial areas between the p-type CdTe domains and n-type CdSe domains, fabricated using the three-dimensional (3D) supra-quantum dot (SQD) structures and onedimensional (1D) nanorod (NR) structures, as a proof-of-concept study. The SQD is a 3D superstructure formed by the attachment of hundreds of QDs.30 The sizes of the SQDs can be easily tuned from dozens to hundreds of nanometers by controlling the synthesis protocol, which also allows facile surface modification that can retain the colloidal stability in any solvent. In addition, SQDs have interconnected structures of QDs, without organic ligands between them, which is beneficial for efficient photogenerated carrier transport.30 Based on these unique features of SQDs, we incorporate SQDs into the interpenetrating blending layers of the all-inorganic solar cells to control the p-type and ntype domain sizes (i.e., p–n heterojunction areas) for efficient charge transfer and reduced recombination loss. The SQD and NR-containing p–n heterojunction solar cells show superior photovoltaic properties, when compared to simple bilayer p–n heterojunction solar cells with smaller heterojunction areas and p–n heterojunction NR solar cells with larger heterojunction areas.

Results and discussion We fabricated three types of all-inorganic solar cells whose photoexcited charge-carrier– generating layers (i.e., photoactive layers) were composed of p-type CdTe NCs and n-type CdSe NCs: i) bilayer p–n heterojunction cell (Structure I), ii) interpenetrating p–n heterojunction cell with SQDs and NRs (Structure II), and iii) interpenetrating NR p–n heterojunction cell (Structure III). Colloidal CdTe SQDs30 and NRs of CdTe and CdSe31-33 were synthesized using methods that were slightly different from those previously reported. Figures 1a-d show the TEM


4

Page 5 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


images and absorption spectra of the SQDs and NRs used as the photoactive layer materials. The CdTe SQDs had nearly monodispersed spherical 3D structures with an average diameter of 79 nm (Figure 1a). The CdTe NRs (56.2 nm × 3.3 nm, length × diameter, average values) and CdSe NRs (44.1 nm × 4.4 nm) are shown in Figures 1b and 1c, respectively. Both the CdTe and CdSe NRs exhibit sharp band edge excitonic absorption peaks due to the strong quantum confinement along two dimensions, whereas the CdTe SQDs show a rather broad and featureless band edge absorption profile, presumably, due to the weakly confined excitons in the internal SQD structures with connections of a few nanometer-sized particles.30 Figures 1e, 1f, and 1g show schematic illustrations of the three different types of photoactive layers on ITO glasses with top metal contacts (Structures I, II, and III, respectively). Prior to photoactive layer deposition, long-alkyl-chain ligands of the assynthesized NCs were replaced with pyridine to improve their electrical conductivity.12,24 Each structure had a CdTe NR layer spin-cast onto an ITO anode as a selective hole transport layer (HTL) for reducing shunting and reverse currents.34-35 A CdTe NR layer and a CdSe NR layer were sequentially spin-cast on the CdTe HTL for Structure I (p–n bilayer device, Figure 1e). For the interpenetrating p–n heterojunction devices, a mixed solution of CdTe SQDs and CdSe NRs and a mixed solution of CdTe NRs and CdSe NRs were spin-cast twice on the CdTe HTLs for Structure II (Figure 1f) and Structure III (Figure 1g), respectively. The nominal thickness of the photoactive layers in the devices was approximately 300 nm. In the case of the photoactive layer of Structure II, CdSe NRs filled the voids created by packing the spherical SQDs (Figure 2). In addition, CdSe NRs and CdTe NRs that possess identical surface ligands can be well-mixed with close packing via van der Waals and dipole interactions between the NRs.36 After spin-casting


5


Page 6 of 21

the NC layers, the films were sintered with Cd ion treatments under nitrogen atmosphere, which allowed the subsequent coating of NCs onto the predeposited NC layers without dissolution of the predeposited layer and facilitated the charge-carrier transfer owing to crystal grain growth.31 The X-ray diffraction (XRD) pattern of the NC film after sintering shows the sharper peaks of the NCs, indicating the higher crystallinity of the NC film after sintering, Figure S1, Supporting Information). The XRD data before and after the sintering process indicate no significant oxidation of the nanocrystals occurred by the sintering process. Pinholes or voids were not observed on the cross-sections of the photoactive films for Structure I, II, and III indicating well-packing of CdSe NRs and CdTe NRs or CdTe SQDs. The absorption spectra of the CdTe/CdSe photoactive layers after sintering were extended to the longer wavelength regions in comparison with those of the NC solutions (Figure 1d and Figure S2a in the Supporting Information). The photovoltaic properties of Structures I, II, and III were measured under standard AM 1.5G illumination (100 mW cm−2). Figure 3 shows the current density–voltage (J–V) curves and external quantum efficiency (EQE) spectra of the best performing devices of Structures I, II, and III. The statistics and averaged photovoltaic parameters are shown in Table 1. The structures have different p-n heterojunction areas (Structure III > Structure II > Structure I; i.e., CdTe and CdSe domain sizes: Structure III < Structure II < Structure I) as illustrated in Figures 1e, 1f, and 1g. Structures II and III with the interpenetrating CdTe and CdSe domains exhibited approximately three-fold higher short-circuit current densities (JSC) than Structure I with the CdTe–CdSe bilayer. JSC values depend on exciton generation rates (light absorption) and charge carrier extraction rates from the light absorber layers. The devices of Structures I, II, and III had similar thicknesses of the light


6

Page 7 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


absorbing layers as shown in Figure S2b in the Supporting Information, and had the similar optical densities (Figure S2a in the Supporting Information). Thus, difference in the JSC values between Structures I, II, and III can originate from the photo-generated carrier separation, transport, and extraction abilities of the structures. The increase in the EQE of the interpenetrating CdTe–CdSe cells (Structures II and III), compared to that of the bilayer heterojunction cell (Structure I), led to an enhanced JSC (Figures 3a and 3b). The EQE of Structure I was markedly reduced in the blue region of the spectrum, where photons are mostly absorbed on the front part of the cell. The EQEs for the interpenetrating cells (Structures II and III) particularly showed over five-fold enhancement at wavelengths below 450 nm. The significantly low EQE in the blue region for Structure I could be because the photogenerated excitons at the front part of the device were severely lost via the recombination processes before the photogenerated charge carriers reached the CdTe/CdSe interface (i.e., p-n heterojunction) in the bilayer of Structure I because of the limited exciton diffusion lengths. In contrast, the interpenetrating cells could form continuous percolation pathways for both electrons and holes through the photoactive layers of the devices, which allowed efficient carrier transport over all the photoactive layers. The open-circuit voltage (VOC) values of the best performance devices of Structures I, II, and III were 0.61, 0.57, and 0.44 V, respectively (Table 1). Too large heterojunction areas resulting from too small p-type and n-type domain sizes in bulk heterojunction devices (Structure III, in this study) could induce interface charge-carrier recombination and charge-carrier leakage, and could inhibit depletion-layer formation because of spatial confinement.22,27 The dark current densities of Structures I, II, and III were shown in


7


Page 8 of 21

Figure 3c. The reverse currents increased as decreasing the p-type and n-type domain sizes. Structure III showed the higher dark current leakage with a poor rectifying behavior due to insufficient formation of depletion layers of this device. Consequently, the VOC could decrease with increase in the p–n heterojunction areas, owing to which Structure III with the smallest domain sizes had VOC lower than that of Structures I and II. In the case of Structure II, where SQDs were used as constituents of the photoactive film, significant JSC enhancement was achieved with minor compromise on the VOC, in comparison with that of the bilayer p–n heterojunction NC solar cell (Structure I), leading to 2.5 times PCE enhancement. The film surface roughness of the photoactive layers for Structures I, II, and III was assessed by atomic force microscopy (AFM) because film morphology could influence the device performance (Figure S3, Supporting Information). The film morphology could be affected by the solubility of NCs in an active layer solution. In this study, we used only a good solvent, pyridine, for the ligand exchange of all the NCs, and hence, the possibility of a film morphology change by the solubility of the NCs in the solutions could be ruled out. Therefore, the film morphologies were likely to be affected only by the shapes and sizes of the NCs we used. It was expected that incorporation of SQDs with larger volumes, in the case of Structure II, would lead to a film morphology rougher than those of Structures I and III. As expected, Structure II had a higher film roughness value (Figure S3, Supporting Information). Nevertheless, Structure II showed an enhancement in PCE, which implied that the device performance of the SQD solar cell could be enhanced further after optimization of the film surface morphology of the cell.


8

Page 9 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Devices with a mixed configuration of Structure II and Structure III were also prepared as control devices (Structure M, Figure 4). The photoactive layer of Structure M consisted of three layers: a CdTe HTL, a layer of a blend of CdTe SQDs and CdSe NRs, and a layer of a blend of CdTe NRs and CdSe NRs, as illustrated in Figure 4a. The photovoltaic performance of the cell is shown in Figure 4b. Structure M showed intermediate photovoltaic characteristics, between those of Structures II and III: the JSC, VOC, and PCE values of Structure M were between those of Structure II and Structure III, which indicated that the results of this study were reliable.

Conclusion In summary, we designed and demonstrated an approach for controlling the p–n heterojunction areas by using SQDs as constituents of the photoactive layers. The SQD-containing p–n heterojunction solar cells (Structure II) exhibited enhanced photogenerated charge-carrier transport, and thereby, higher PCE values originating from the intermediate p-type and n-type domain sizes (i.e., p–n heterojunction areas), between those of the bilayer NR p–n heterojunction solar cell (Structure I) and the interpenetrating NR heterojunction solar cell (Structure III). The SQD-containing solar cells investigated in this study were not rigorously optimized in terms of the compositions, film morphologies and sizes of SQDs, constituent ratios of blends, surface modification of NCs, film thickness, or sintering temperatures.11,37-38 Better device configurations (e.g., by the introduction of electron transport layers and/or optimized hole transport layers such as PEDOT:PSS or NiO, and/or by control of device thickness) can further improve the device performances.39 Therefore, there is considerable room for improvement for the performance of an SQD-based heterojunction-type solar cell. The new approach proposed in


9


Page 10 of 21

this study, which uses 3D assembled NCs to tune the domain sizes of devices, can be applied in a wide range of technological areas such as memories, sensors, light-emitting devices, and energyharvesting systems.

Materials and Methods Materials. Tri-n-octylphosphine oxide (TOPO) (90% and 99%), cadmium acetate dihydrate (Cd(Ac)2·2H2O, 98%), stearic acid (SA, 95%), tri-n-octylphosphine (TOP, 90%), octadecylphosphonic acid (ODPA, 97%), octylphosphonic acid (OPA, 97%), tellurium shot (99.999%, 1-2 mm), cadmium chloride hemi-pentahydrate (CdCl2·2.5H2O, >98%), anhydrous chlorobenzene (99.8%), cadmium oxide (CdO, 99.5%), and pyridine (99+%) were purchased from Aldrich. Acetone and iso-propanol were purchased from J. T. Baker. Methanol was purchased from Mallinckrodt Chemicals. These materials were used as purchased. Synthesis of CdTe supra-quantum dots (SQDs). SQDs were prepared using the previously reported method.30 Cadmium acetate dihydrate (2 mmol), stearic acid (0.05 mmol), and TOPO (6 g) were degassed at 100–120 °C for two hours in a 25 mL three-neck round bottle flask, and heated to 270 °C under N2 gas flow. TOPTe solution (0.5 mmol of tellurium shots dissolved in 1 mL of TOP) was injected at 270 °C and maintained at that temperature for five minutes before the heating mantle was turned off. The reaction mixture was cooled to room temperature, and diluted with hexanes. To remove the excess organics, the SQD crude product was precipitated with the excess ethanol collected by centrifugation, and redispersed using a small amount of hexanes.


10

Page 11 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Synthesis of CdSe and CdTe nanorods (NRs). NRs were prepared using a method similar to those previously reported.31-33 For CdTe NRs, CdO (0.3 g, 2.34 mmol), ODPA (1.218 g, 3.645 mmol), OPA (0.4725 g, 2.43 mmol), and TOPO (99%, 3.975 g) were loaded in a 50 mL threeneck flask, dried under vacuum at 110 °C for three hours, and heated to 320 °C under N2 atmosphere. The colorless CdO solution was cooled to 120 °C and degassed for 30 min. After removing O2 and H2O, the solution was heated again to 320 °C under N2. At 320 °C, 0.72 mL of 1 M TOPTe and 3 mL TOP solution were injected rapidly and NCs were allowed to grow for four minutes. For CdSe NRs, CdO (0.4112 g, 3.2 mmol), TDPA (0.888 g, 3.2 mmol), HPA (0.528 g, 3.2 mmol), and TOPO (90%, 12.9168 g) were degassed and heated up to 320 °C. The solution was then cooled to 120 °C. After reheating to 320 °C, 0.8 mL 1 M TOPSe and 15 mL TOP solution were injected. After the injection, the solution temperature was dropped to 250 °C and maintained for 24 min and then 2.4 mL 1 M TOPSe was injected. After cooling the solution, iso-propanol was added to precipitate the NRs. Sample characterization. High-resolution transmission electron microscopy (HR-TEM) was performed using a JEOL JEM-2200FS operated at 200 kV. Scanning electron microscopy (SEM) was performed using JEOL JSM-7401F. X-ray diffraction (XRD) measurements were performed using a Rigaku Max-2500V, which was equipped with a Cu Kα X-ray source. Atomic force microscope observations were made using a Veeco Dimension 3100 AFM system with a Nanoscope V controller (Veeco Metrology Group). Device fabrication and characterization. Purified NCs were collected and dried. For surface ligand exchange to pyridine, 100 mg NCs were suspended in 15 mL pyridine and refluxed overnight under an N2 atmosphere. After the ligand exchange, the NCs were purified by centrifugation using hexane. Pyridine-capped NCs were dissolved in pyridine of concentration


11


Page 12 of 21

70 mg mL-1. Indium tin oxide (ITO)-patterned glass substrates were cleaned by sonication with detergent, acetone, and iso-propanol. After cleansing, the NC solution was spin-coated and thermally annealed on a hot plate at 150 °C for two minutes for every layer. For sintering, the spin-coated film was dipped into a CdCl2 saturated methanol solution and washed with isopropanol, after which the samples were sintered on a hot plate at 350 °C in a glovebox for one hour. Under high vacuum (

Heterojunction Area-Controlled Inorganic Nanocrystal Solar Cells

Recommend Documents