用户,您好!欢迎您进入大气污染防治技术进展平台!登录 | 注册  NSTL重点领域信息门户  帮助中心  RSS
 您当前的位置: 首页 > 资源详细信息
资源基本信息 
发布日期: Jun 1, 2020
资源类型: 527.09KB
资源性质: 科技论文
重要度:  
资源评价:

资源推荐:

主题相关资源
系统抽取主题
抽取主题     
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)

Cathode engineering with perylene-diimide interlayer enabling over 17% efficiency single-junction organic solar cells

Results .Materials synthesis and characterization . The device structure is provided in Fig.? 1a , and the photovoltaic materials (PM6 and Y6) for the active layer are shown in Supplementary Fig.? 1 . As shown in Fig.? 1b , PDINN was easily prepared in one step, by condensation of N , N -dimethyldipropylenetriamine with perylene-3,4,9,10-tetracarboxylic dianhydride in methanol. The chemical structure of PDINN was verified by mass spectrometry (Supplementary Fig.? 2 ) and NMR spectroscopy (Supplementary Fig.? 3 ). The reaction proceeded almost entirely in methanol with a yield of ca. 95%. Notably, 64.4?g large-scale syntheses of PDINN can be readily conducted in methanol under mild reaction conditions (70?°C) and short reaction time (8?h). And the condensation reaction is convenient for further scale-up if needed. With the model for cost calculation43, the cost of material ( Cg , cost-per-gram) for PDINN is 1.6?$?g ?1 (Supplementary Table? 1 ). The Cg value is less than one fifth of the well-known low-cost donor P3HT (10.0?$?g ?1 ), and less than one percent of the fullerene (240?$?g ?1 ), which is the starting material to synthesize effective fullerene derivative interlayers25. The result indicates that PDINN is really a low-cost material convenient for large-scale application.Fig. 1: Chemical structures and characterization of the PDI derivatives CIMs.a Device structure of the OSCs. b Synthetic route of PDINN along with the chemical structures of PDINN and PDINO. c Normalized UV–vis film absorption spectra of PDINN and PDINO films. d Cyclic voltammograms of PDINN and PDINO. e UPS spectra of the PDINN and PDINO-treated electrodes. f Electron spin resonance spectra of PDINN and PDINO in solid. Full size image To better understand the role of the aliphatic amine group, the thermal, physicochemical, and the photovoltaic properties of PDINN are compared with the widely used PDINO CIM. The thermal stability of PDINN and PDINO was investigated by putting the two samples on a hot plate under 150?°C for 1?h followed by dissolving the samples in methanol. It can be seen that PDINO becomes insoluble in methanol, whereas, clear PDINN solution can be still obtained for the PDINN sample after the thermal treatment (Supplementary Fig.? 4 ). The result indicates that PDINN with the aliphatic amine group is more stable than PDINO with N -oxide amine group. The good thermal stability of PDINN is further confirmed by their thermogravimetric analysis (TGA), which show the 5% weight loss temperature at 103.4?°C for PDINIO and 240.1?°C for PDINN (Supplementary Fig.? 4b ).Intermolecular interactions with active layer . PDINN shows a good solubility of 26.7?mg?mL ?1 in methanol without the assistance of any acid, while most organic CIMs (such as PDIN or PFN) need hydrochloric acid or acetic acid to make a clear alcohol solution for processing24 , 36. In addition, PDINN shows a better interfacial compatibility on a wide range of active layers of OSCs compared with PDINO. For example, when a dilute solution (0.5?mg?mL ?1 ) of PDINN or PDINO was spin-coated onto the PM6:Y6 active layer at high spin speed (5000?rpm), PDINN can form a uniform film, whereas PDINO shows a poor film formation, as shown in the photo-induced force microscopy (PiFM) images in Supplementary Fig.? 5 . This phenomenon is likely related to the hydrogen bond formed between the secondary amine in the aliphatic amine groups of PDINN with the active layer surface.To confirm this, we initially conducted the variable-temperature and titration 1 H NMR experiments, with PDINN and Y6 in solution (Supplementary Fig.? 6 ). The 1 H NMR signal at δ1.54 (assigned to hydrogen in the secondary amine) broadens and downfield shifts. To provide more evidence on the interactions between Y6 in the active layer and PDINN CIM in the OSC devices, we performed attenuated total reflectance (ATR) FT-IR measurement for the solid blend sample of Y6 and PDINN (w/w, 1:1), and the result is shown in Supplementary Fig.? 7a . In contrast to the ATR FT-IR spectrum of pure Y6 sample with the C–F mean peak at 1350?cm ?1 , the C–F peak of the Y6/PDINN blend sample moves to 1297?cm ?1 and is broadened. Moreover, we further studied the interaction between PDINN and solid Y6 by measuring 19 F NMR spectra of solid Y6 power in methanol-D 4 NMR tubes with and without PDINN. The 19 F NMR spectra of the samples are shown in Supplementary Fig.? 7b . It can be seen that there is a 19 F signal at ?152?ppm in the PDINN methanol-D 4 solution, which is from the F atom of Y6. However, for the control sample without PDINN in methanol-D 4 , there is no such signal for Y6 because Y6 is not soluble in methanol. Thus, the above results collectively provide strong evidence for the formation of the hydrogen bond between secondary amine group of PDINN and C–F group of Y6 in the solid state. The results also suggest that PDINN can provide intermolecular interaction with the organic semiconductor (especially those bearing fluorine groups4), such as Y6 during the deposition of PDINN from its methanol solution. The hydrogen-bonding feature of PDINN indicates that PDINN as CIM in OSCs can improve the interfacial compatibility, with the active layer blow via a spontaneous adsorption process driven by the hydrogen-bonding interactions, thus reduce the series resistance of the devices44.Physicochemical properties . Figure? 1c shows UV–vis absorption spectra of PDINN and PDINO film. PDINN displays a blue-shifted absorption relative to that of PDINO, with an absorption edge at 630?nm and a maximum absorption at 473?nm. To explain the phenomenon of the blue-shifted absorption, we measured the hydrogen-bonding property of PDINN in comparison with PDIN (ref.36) by measuring the ATR FT-IR spectra of PDINN and PDIN, as shown in Supplementary Fig.? 7c . The carbonyl vibration at 1692?cm ?1 for PDIN is shifted to 1687?cm ?1 for PDINN, indicating the hydrogen-bonding formation in PDINN. Therefore, the blue-shifted absorption of PDINN film could be related to the different aggregation of PDINN molecules in comparison with the PDINO molecules due to its hydrogen-bonding property45. Cyclic voltammetry measurements were performed to measure the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels ( EHOMO/LUMO ) of the materials. The onset oxidation/reduction potentials ( φox/red ) can be obtained from the cyclic voltammograms, as shown in Fig 1d . According to the equations of EHOMO/LUMO ?=?? e ( φox/red ??? φFc+/Fc ?+?4.8) (eV), the ELUMO value was calculated to be ?3.78?eV for PDINN and ?3.63?eV for PDINO. The LUMO energy-level values of the PDI derivatives, especially for PDINN, are close to those of the non-fullerene acceptors4 , 9, which can form a better energy-level alignment at the cathode interface and hence enhance electron collections. And the low-lying EHOMO values of ?6.02?eV for PDINN and ?6.21?eV for PDINO mean that holes from various donors will be blocked sufficiently at the cathodes, with the PDI derivative CIMs. Grazing-incidence wide-angle X-ray scattering (GIWAXS) was used to measure the crystallinity of the PDI derivatives deposited on silicon substrates. As shown in Supplementary Fig.? 8 , both of the two PDI interlayers show a semicrystalline structure, where their defined peaks are associated with their sufficient aggregation of the PDI core in the solid for efficient charge transport, when they are used as the CIMs. With a smaller terminal group in PDINO, although it shows a stronger aggregation behavior, its predominant edge-on orientation as revealed by the strong (100) peak in the out-of-plane direction may weaken its carrier extraction ability from the active layer.Ultraviolet photoelectron spectroscopy (UPS) was used to probe the WF values of different metal electrodes of Al, Ag, and Cu modified with the CIMs. The WF modification ability is frequently associated with the interfacial dipole provided by the CIMs (refs.19 , 26 , 32 , 46 , 47). From Fig.? 1e and Table? 1 , it can be seen that the two CIMs successfully reduced the WFs of the electrodes, the WF values of Al, Ag, and Cu were reduced to 3.83, 3.88, and 4.24?eV for the PDINO CIM, and 3.67, 3.72, and 4.08?eV for the PDINN CIM, respectively (see Table? 1 ). PDINN shows a stronger WF lowering ability over that of PDINO for each electrode, with ca. 0.16?eV lower WFs for the PDINN-modified electrodes than that of the PDINO-modified electrodes. It is well accepted that the dipole moment in the amine group-containing CIMs is originated from the electron transfer from the nitrogen lone pair of neutral amine group to the electrode20 , 26. Thereby, the lower WFs obtained for PDINN should be associated with its higher dipole moment from the aliphatic amine group, which is just as poly(ethyleneimine) (PEI) or its derivative does in effectively lowering the WFs of different electrodes20 , 48. These results encouraged us to explore PDINN as cathode interlayer with high WF metals (such as Ag and Cu) as cathode.Table 1 Comparison of the physicochemical properties of the CIMs of PDINO and PDINN.Full size table The doping effects of the CIMs were investigated by the electron spin resonance (ESR) spectroscopy (Fig.? 1f ). PDINN shows a strong self-doping behavior ( g value of 2.001) relative to PDINO ( g value at 2.002), which is enabled by the unpaired electrons in aliphatic amine group of PDINN and beneficial to improving the electrical conductivity of the materials46 , 49 , 50. The electrical conductivities of the CIM films were measured by depositing two parallel silver electrodes on the two sides of the films. The conductivities of PDINO and PDINN calculated from the I – V curves (Supplementary Fig.? 9 ) are 2.4?×?10 –6 ?S?cm ?1 and 5.0?×?10 –4 ?S?cm ?1 , respectively, which indicate that PDINN has a better electron-transport ability.Photovoltaic performance . To evaluate the photovoltaic performance of the PDINN CIM in OSCs, we fabricated conventional devices with a structure of indium tin oxide (ITO)/PEDOT: PSS (poly(3,4-ethylenedioxythiophene): poly(styrene-sulfonate)) /PM6:Y6/CIM/electrode. As for the active layer, PM6: Y6 blend represents the state-of-the-art system with initially reported efficiency of 15.7% by Zou, using the PDINO/Al cathode10. Recently, intensive device engineering with adding a third component in the PM6:Y6 system increased the efficiency of the ternary OSCs over 16.5% (refs.51 , 52 , 53). Without tedious component optimization, here, we demonstrate a simple yet effective cathode modification with PDINN to realize a high efficiency over 17% for binary PM6:Y6 OSC (see Table? 2 ). To better understand the high performance of PDINN interlayer, the device performance of PDINN was compared with that of PDINO, especially for the devices with air stable metal Ag and Cu as cathode.Table 2 Photovoltaic performance of the OSCs base on PM6:Y6 with Ag as cathode under the illumination of AM 1.5?G, 100?mW?cm ?2 .Full size table Current density–voltage ( J–V ) characteristics of the optimal OSCs without the CIM (treated with methanol), and those with the CIM are compared in Fig.? 2a , and the relevant photovoltaic parameters are listed in Table? 2 . Figure? 2b shows the incident photon-to-converted current efficiency (IPCE) curves, and the integrated currents from the IPCE spectra agree quite well with the Jsc values measured from the J–V curves. The PDINO/Ag device delivers a good PCE of 15.17%, along with a Voc of 0.821?V, a Jsc of 25.58?mA?cm ?2 , and an FF of 72.24%, which is much higher than that (13.84%) of the Ag-only device. The OSC devices with PDINN/Ag cathode demonstrated an even higher maximum PCE of 17.23%, the remarkable improvement in PCE of the PDINN/Ag-based device stems from the substantial increase in the key device parameters of Voc (from 0.821?V to 0.847?V) and FF (from 72.24 to 78.59%) in comparison with the PDINO/Ag-based device. The efficiency histograms of PDINN-based and PDINO-based devices are provided in Supplementary Fig.? 10 . The photovoltaic performance of the optimized PDINN/Ag device was certified by NREL, and a PCE of 16.77% together with a Voc of 0.843?V, an FF of 77.5%, and a JSC of 25.704?mA?cm ?2 was confirmed (Supplementary Fig.? 11 ).Fig. 2: Device performance of the OSCs based on PM6:Y6 with Ag as cathode.aJ – V curves of the best OSCs under the illumination of AM 1.5?G, 100?mW?cm ?2 . b IPCE spectra and integrated current density of the best devices. c Dependence of PCE and FF of the OSCs on the thickness of PDINN CIM. d Dark currents of the OSCs. Full size image Interestingly, the photovoltaic performance of the PDINN-based OSCs exhibit good tolerance to the PDINN thickness variation in fabricating the OSC devices, which is important for large area fabrication of the OSCs. While the commonly used PEI and PFN CIMs that only works well at critical thin thickness (<5?nm)17 , 26. For example, when the PDINN thickness rises to 27?nm, the PCE of the devices still maintains at a high value over 15.0% (Supplementary Fig.? 12 , Supplementary Table? 3 ). Impressively, the FF still maintains a high value of 74.09% when the PDINN thickness is further increased to 38?nm, indicating the excellent electron-transport and collection ability of the PDINN CIM. Supplementary Fig.? 12c, d shows the effect of CIM thickness on the photovoltaic performance of the PDINO CIM-based OSCs. The PCE decreased from 15.46% for the optimized device with PDINO thickness of 9?nm to 11.72% for the OSC with PDINO thickness of 39?nm. Notably, the PCE of the OSC with 38?nm thickness PDINN CIM is even comparable to the peak value (15.46%) for the PDINO-based devices at its optimum thickness (9?nm).Based on the PDINN thickness insensitivity of the PDINN-based OSCs, we also fabricated large area devices of 1?cm 2 to evaluate the prospect of the PDINN/Ag cathode in real application. The J – V curve and IPCE spectrum of the large area device are plotted in Supplementary Fig.? 13 . As a result, a PCE value of 15.82% is obtained, along with a Voc of 0.850?V, a Jsc of 25.42?mA?cm ?2 (calculated Jsc from IPCE spectrum is 24.95?mA?cm ?2 , with a mismatch of 1.85%), and an FF of 73.19%. The performance is a prominent one for the 1?cm 2 device54.The effect of CIMs on device performance . For the two CIMs, we further evaluate the charge carrier recombination behavior of the OSCs, and studied the dependence of Jsc on the light intensity ( P ), as well as the dependence of Voc on P (ref.55). The dependence of Jsc (and Voc ) on P is shown in Fig.? 3a . The relationship between Jsc and P can be expressed as Jsc∝ P α(ref.56). The value of α should be 1 if all the free carriers were extracted by the electrode55. In our cases, the α value of the PDINO-based device is 0.963, while for the PDINN-based device, its α value is increased to 0.982 that is very close to 1. This means there is effective carrier collection and well-suppressed bimolecular recombination in the OSCs with PDINN as CIM. Figure? 3a shows the dependence of Voc on light intensity P . The slope of Voc versus ln( P ) line should be kT / q for bimolecular recombination57, while the competition between bimolecular and Shockley–Read–Hall type, trap-assisted recombination makes the slope between kT / q and 2 kT / q . The OSCs with methanol treatment (without CIM) and with PDINO or PDINN CIM exhibit the slopes of 1.42, 1.24, and 1.14 kT / q , respectively. It is obvious that the slope of the device with PDINN CIM is closer to kT / q , which indicates the suppression of the trap-assisted recombination in the devices. The results collectively suggested that charge recombination was effectively suppressed in the PDINN-based OSCs.Fig. 3: The investigation of the charge carrier recombination behavior at the contact.a Light intensity dependence of Jsc and Voc of the OSCs based on PM6:Y6. Free charge dynamics of PM6:Y6/PDINN, PM6:Y6/PDINO, and PM6:Y6 films b at 633?nm probe and c at 770?nm probe, respectively. Full size image Furthermore, we performed transient absorption (TA) measurements in a timescale up to tens of microseconds to study the effect of the CIM on the dynamics of photo-excited charges decay in the OSCs. The experimental details are described in “Methods” section. For the static transport measurements10, the average diffusion distance at a delay time of 20?ns can be estimated to be on the order of 10?nm at room temperature with no voltage applied. The diffusion length may be much longer if the time-dependent local charge mobility is considered58 , 59. Thus for our samples, excited carrier can diffuse to the active layer surface, and the related probe can give some information on the carrier behavior that affects device performance. On the timescale of tens of ns and longer, the TA data are mainly induced by the charge-separated state of free polarons as confirmed by the photo-induced-absorption measurements (Supplementary Fig.? 14 ). In the PM6:Y6 blend, the charge-separated state is configured with electron polaron at the acceptor site and hole polaron at the donor site, which may be captured by the excited-state absorption features of the free polarons, and the ground-state bleach features of donors and acceptors (Supplementary Fig.? 14 ). With increasing pump fluence, the decay of free polarons becomes faster due to bimolecular recombination process (Supplementary Fig.? 15 ). In comparison with the dynamics in the sample with PDINO, the charge recombination in the sample with PDINN is slower at the early stage (Fig.? 3b, c ), implying that the CIM of PDINN suppresses some loss channel.For the different dynamics in the two CIM-treated samples, although we have no further evidence to tell how the PDI interlayers affect the charges that diffuse to the surface. The longer-lived photo-excited charges observed in PDINN-treated blend extend the charge lifetime, which is beneficial for charge collection, and thus probably responsible for the improved performance in the PDINN-treated OSCs. Typically, charge recombination is directly related with the key device parameter FF of the devices, the suppressed carrier recombination, and enhanced charge collection in the PDINN-based devices is well consistent, with its higher FF value (78.59%) in comparison with that (72.24%) for the PDINO-based device.The J–V characteristics of the OSCs at dark, (see Fig.? 2d ) show a slightly large rectification ratio thus a better diode quality for the device with PDINN CIM. To better understand the improved FF of the OSC with PDINN CIM, the electrochemical impedance spectroscopy (EIS) measurements were performed to examine the interface resistance of the devices60. Supplementary Fig.? 16 shows Nyquist plots of the OSCs with PDINN or PDINO CIM at dark. A bias voltage equal to Voc was applied to dissipate the total current. The data were fitted using the equivalent-circuit model. The series resistances ( Rseries ) of the PDINN-based device (0.62?Ω?cm 2 ) is significantly smaller than that of the PDINO-based device (280.80?Ω?cm 2 ). The lower series resistance should be beneficial for the higher FF of the PDINN-based devices.Device performance with other active layers . With the successful application of PDINN CIM in the PM6:Y6-based OSCs, we further studied the universality of the PDINN CIM for the OSCs with other active layers, including PTQ10: IT-4F (ref.61) and J11: m -ITTC (ref.62; Supplementary Figs.? 17 and 18 ). Using Ag as the top cathode, the direct comparison of the PCEs of the OSCs with different CIMs (PDIN/Ag in Supplementary Fig.? 19 ) and different photoactive layers are shown in Fig.? 4a , and the related photovoltaic parameters are collected in Supplementary Table? 4 . Clearly, the PDINN-based devices exhibit better PCEs in all the OSCs, with different active layers compared with that of the corresponding PDINO-based devices, indicating that PDINN is a universal and effective CIM for various OSCs. In consideration that all of the investigated active layers bearing halogen atom or carbonyl group, the formation of the hydrogen bond intermolecular interactions between the active layer and PDINN is also plausible. Our results also suggest that constructing intermolecular interactions between the active layer and the electrode interlayer is an effective way for further improving photovoltaic performance, and realizing application of the OSCs.Fig. 4: Photovoltaic performance of the OSCs base on PM6:Y6.a The comparison of the PCEs of the OSCs with different CIMs and different photoactive layers under the illumination of AM 1.5?G,100?mW?cm ?2 . bJ – V curves of the best OSCs based on PM6: Y6 blend with copper as cathode under the illumination of AM 1.5?G,100?mW?cm ?2 . c IPCE spectra and the integrated current density of the best devices with Cu as cathode. d Stability of the OSCs with different top cathode of PDINN/Ag or PDINN/Cu. The devices kept in nitrogen environment and in dark between the J – V measurements. Full size image Device stability and effect of cathodes on device performance . At the optimal CIM thickness, we also explored high WF Cu as cathode of the OSCs based on PM6: Y6 blend to take its advantage of high conductivity and low price. The J – V curves and IPCE spectra of the OSCs are shown in Fig.? 4 . Notably, with the PDINN CIM, Voc value (0.844?V) of the Cu-based devices is similar to that of the Ag-based devices (0.847?V). Impressively, a high PCE exceed 16% and a high FF of 78.8% are obtained for the PDINN/Cu-based device, indicating that Cu is also a good choice for cheap and stable top electrode in future commercialization of the OSCs, with PDINN as CIM. The lower IPCE value in the wavelength region from 300 to 600?nm is associated with the low reflectivity of the Cu-based device (Supplementary Fig.? 20 ), which explains the slightly lower Jsc value with Cu as cathode. In addition, the result suggested that a higher Jsc value thus higher PCE could be obtained by suitable optical management34in the Cu electrode-based OSCs, with PDINN as CIM.The stability of the interlayer itself and interlayer/electrode contact is important issue for real application of OSCs. With the good air and thermal stability of PDINN, we further investigated its device stability using Ag or Cu as top electrode. After ~50 days, the PCEs of the PDINN/Ag-based and PDINN/Cu-based devices can retain 87 and 88% of their initial PCE values in a glove box filled with nitrogen (Fig.? 4d ), while using PDINO as CIM, the PCEs remained 73 and 85% of their initial PCEs for the Ag-based and Cu-based devices, respectively. Supplementary Fig.? 21a displays the steady-state efficiency of the OSCs (in a glove box filled with nitrogen) under continuous air mass (AM) 1.5?G (100?mW?cm ?2 ) illumination. The primary results over 600?s show a better stability of the PDINN-based devices (98.21% of initial efficiency) than that of the PDINO-based devices (95.90% of initial efficiency). For the unencapsulated device in air (Supplementary Fig.? 21b, c ), the PDINN/Ag and PDINN/Cu devices maintained 93.5% and 86.3% of their initial PCE value, whereas the PDINO/Ag and PDINO/Cu devices only showed 35.2% and 70.0% of their initial efficiency after storing 150?h. The big difference between the stability of PDINN and PDINO-based devices is closely related to the CIMs used. The low thermal stability and moisture sensitive of PDINO can lead to an irreversible damage to the CIM–cathode contact, which renders the device stability poor. All the above results indicate that the PDINN-based OSCs show better stability than that of the PDINO-based devices.To better understand the effect of the top electrodes (Al, Ag, and Cu) and the CIMs on the device performance, corresponding device parameters are summarized in Supplementary Tables? 5 and 6 for comparison ( J – V curves and IPCE spectra of the Al-based devices are shown in Supplementary Fig.? 22 ). From Supplementary Table? 5 , it can be seen that for the PDINO-based devices with cathodes from Al, Ag to Cu, the Voc values, and FF values are gradually decreased due to a limited WF tunability for PDINO, they are 0.843?V and 76.81% for Al, 0.821?V and 73.52% for Ag, and 0.814?V and 72.61% for Cu, respectively. In contract, the PDINN-based devices demonstrated almost unchanged Voc (ca. 0.844?V) and FF (ca. 78.0 %) values for all the OSCs with Al, Ag, and Cu cathode (Supplementary Table? 6 ) benefited from a strong WF tunability of PDINN. In addition, the advantage of Ag electrode with strong reflectivity and low resistance should also account for its higher PCE values over that of the PDINO-based OSCs. The trends of variation in Voc and FF are plotted in the inset of Fig.? 4d for a clear comparison. The lower WF of the cathode increases the built-in electric field of the device, which is beneficial to increasing charge extraction efficiency and reducing recombination losses for higher FF of the PDINN-based OSCs. The results suggested that PDINN is an excellent CIM for future application of the OSCs with high WF stable metals as cathode. .