Efficient red perovskite quantum dot light-emitting diode fabricated by inkjet printing

Perovskite quantum dots (PeQDs) are considered potential display materials due to their high color purity, high photoluminescence quantum yield (PLQY), low cost and easy film casting. In this work, a novel electroluminescence (EL) device consisting of the interface layer of long alkyl-based oleylammonium bromide (OAmBr), which passivates the surface defects of PeQDs and adjusts the carrier transport properties, was designed. The PLQY of the OAmBr/PeQD bilayer was significantly improved. A high-performance EL device with the structure of indium tin oxide/poly(3,4-ethylenedioxythiophene) polystyrene sulfonate/poly(bis(4-phenyl)(2,4,6-trimethylphenyl)amine)/OAmBr/PeQDs/2,2′,2″-(1,3,5-benzinetriyl)-tris(1-phenyl-1H benzimidazole)/LiF/Al was constructed using a spin-coating method. A peak external quantum efficiency (EQE) of 16.5% at the emission wavelength of 646 nm was obtained. Furthermore, an efficient matrix EL device was fabricated using an inkjet printing method. A high-quality PeQD matrix film was obtained by introducing small amounts of polybutene into the PeQDs to improve the printing process. The EQE reached 9.6% for the matrix device with 120 pixels per inch and the same device structure as that of the spin-coating one.


Introduction
In recent years, the efficiency of perovskite light-emitting diodes (PeLEDs) has developed rapidly.Based on the Original content from this work may be used under the terms of the Creative Commons Attribution 4.0 licence.Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.spin-coating technique, the efficiency of red and green perovskite LEDs has exceeded 20%, [1][2][3][4] and that of blue perovskite LED has reached 12% [5,6].As known, inks for the spin-coating method have a choice of multiple solvents for the possible production of high-quality uniform PeLED films.However, for color display utilization, the inkjet printing technique is almost exclusively used for commercialization rather than the spin-coating method [7].To date, research on inkjetprinting PeLEDs has seldom been reported and their electroluminescence (EL) performance is still poor compared with that of the devices fabricated using the spin-coating method [8,9].

Future perspectives
Inkjet printing technology has great potential in the preparation of low-cost, large-size displays, and the excellent photoelectric properties of perovskite materials have also attracted the attention of many researchers.In addition to the stability problem of perovskite materials, for inkjet printing, the development of stable and efficient inks and the realization of high-resolution displays still require further effort by researchers.We developed stable red perovskite quantum dot inks and improved the performance of the device from two aspects: material design and device structure optimization, providing a solution for inkjet printing highefficiency perovskite light-emitting devices.
The main reasons for this are probably the ink formulation for inkjet printing, small-size matrix film quality control, EL device structure design, etc.
Inkjet printing perovskite films are usually divided into two material systems according to the ink solvent.One is the perovskite precursor system using polar solvents, such as dimethylformamide and dimethyl sulfoxide.The other is a pre-synthesized perovskite system, such as perovskite quantum dots (PeQDs) using non-polar solvents, such as octane and dodecane.For inkjet printing perovskite precursor, Sun's group reported a green PeLED device with an external quantum efficiency (EQE) of 9.0% with the device structure of indium tin oxide (ITO)/poly[ (9,9- A PVP layer was functioned to suppress the leakage current and prevent PVK layers from being dissolved by the perovskite precursor ink solvent.In the device fabrication, a preheated substrate was used to facilitate crystallization of the perovskite precursor, reduce the grain size, and produce an out-of-plane orientated crystal structure for efficient charge transport [9].The control of grain size and orientation complicates the fabrication of perovskite precursor films when using the inkjet printing technique.The quality of the printed perovskite precursor layers relied mostly on the precise control of nucleation and crystallization during solvent evaporation [10]. However, the pre-synthesized PeQDs with non-polar solvent do not require precise control of nucleation and crystallization during inkjet printing.For inkjet printing PeQDbased LEDs, our group first reported a green matrix PeQD-LED with EQE of 2.8% by using the mixing solvent of octane and dodecane PeQD ink with the device structure of ITO/poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS)/PVK/poly(4-butylphenyldiphenylamine) (poly-TPD)/PeQDs/TPBi/LiF/Al [8].By adjusting the solvent ratio of octane and dodecane, we obtained flat films without a coffee ring.Then, Li's group achieved an EQE of 3.03% from a pixel-sized green PeQD-LED using the inkjet printing method with the device structure of ITO/PE-DOT:PSS/TFB/PeQDs/TPBi/LiF/Al [11].Compared to the performance of PeLEDs based on the spin-coating method, the PeQD-LED efficiency based on the inkjet printing method is still low.However, to our knowledge, red PeQD-LED based on the inkjet printing technique has not yet been reported, prob-ably due to red PeQD instability and its greater sensitivity to water and oxygen.However, red emission is necessary for fullcolor display.
Here, we contribute some excellent results from red PeQD matrix EL devices prepared using the inkjet printing technique.PeQDs with Sr, Zn co-doped CsPb(I/Br) 3 were designed and synthesized, since the substitution of Pb 2+ with a slight amount of Sr 2+ and Zn 2+ enhances the stability of perovskite via reducing structural distortion [12][13][14][15].Furthermore, an efficient EL device structure was designed by introducing long-side-chain oleylammonium bromide (OAmBr) as the interfacial layer between poly(bis(4-phenyl)(2,4,6trimethylphenyl)amine) (PTAA) and PeQDs functioned to both efficiently passivate PeQD surface defects and balance carrier transport properties.The EL devices fabricated using both spin-coating and inkjet printing methods exhibited efficient red emission.The EQEs of the EL devices fabricated using spin-coating and inkjet printing methods are 16.5% and 9.6%, respectively.Here, for inkjet printing devices, a slight amount of polybutene (PB) polymer was introduced into the PeQD ink to improve the quality of the PeQD pixel film, which might negatively influence the EL performance.

Preparation of Cs-oleate
In order to synthesize Cs-oleate, 0.1628 g of Cs 2 CO 3 and 0.5 ml of OA were mixed with 8 ml of ODE in a 50 ml threeneck flask, and the mixed solution was degassed under vacuum at 120 • C for 30 min, then heated under N 2 to 140 • C until all the Cs 2 CO 3 reacted with the OA.

Synthesis of Sr, Zn-doped CsPb(I/Br) 3 NCs
PbBr 2 (0.0348 g), PbI 2 (0.0433 g), SrI 2 and ZnI 2 (0.188 mmol and mixed in different ratios), OA (0.5 ml) and OAm (0.5 ml) were loaded into a 50 ml three-neck flask and dried under vacuum for 30 min at 120 • C.Then, the temperature was raised to 150 • C and Cs-oleate solution (0.4 ml, prepared as described above) was quickly injected.After about 5 s, the reaction mixture was cooled in an ice-water bath.Then, 2 ml of the crude solution was added to 1 ml of acetonitrile and 4 ml of ethyl acetate.The precipitate was collected after centrifugation at 12 000 rpm for 3 min and dispersed in 1 ml of toluene.Then, 2 ml of ethyl acetate was added.The precipitate was collected after centrifugation at 12 000 rpm for 3 min and dispersed into octane.The element ratios of Sr, Zn and Pb were confirmed by using coupled plasma atomic emission spectrometry, as shown in table 1.

Device fabrication
The EL devices were constructed with the structure of ITO/PEDOT:PSS/PTAA/OAmBr/PeQDs/TPBi/LiF/Al.An ITO-coated glass substrate was cleaned in sequence, with tetrahydrofuran, isopropanol, detergent, deionized water and isopropanol in an ultrasonic bath, and then dried in a vacuum oven.The cleaned ITO-coated glass substrates were then treated with a plasma for 10 min and then covered with a 40 nm thick PEDOT:PSS film by spin-coating at 3000 rpm to form a hole injection layer.The coated ITO glass substrate was then baked at 150 • C for 15 min to dry thoroughly.PTAA (5 mg ml −1 in chlorobenzene) and OAmBr (10 mg ml −1 in ethanol) were sequentially spin-coated at 4000 rpm followed by baking at 120 • C for 15 min.The thickness of the PTAA was 10 nm, while that of the OAmBr was 25 nm.For the spincoating device, the PeQD layer was spin-coated at 1200 rpm and then annealed at 60 • C for 10 min.For the inkjet-printing device, the matrix PeQD layer was formed in air using a 30 µm diameter nozzle, dried in a vacuum of 1.3 kPa for 15 min, then annealed at 60 • C for 10 min.The pixel size was 190 µm in length and 50 µm in width.Both the PEDOT:PSS and PTAA films were prepared sequentially by the spin-coating method.
The electron injection and transport layer (TPBi/LiF), and the cathode Al were thermally deposited under a pressure of 2 × 10 −4 Pa onto the PeQD layer in sequence.Finally, the PeQD device was encapsulated with UV curing glue and a glass cover in a glove box and then characterized in air.

Characterization and device measurements
UV-visible absorption spectra were recorded with a UVvisible spectrophotometer (HP 8453E).Photoluminescence (PL) spectra were recorded with a PL spectrometer (Biaoqi Photoelectric).The time-resolved PL spectra were obtained using a time-correlated single-photon counting system (C11367-11, Hamamatsu Photonics).The x-ray diffraction (XRD) patterns of the PeQD film structure were measured on an x-ray diffractometer (Panalytical) equipped with a Cu-Ka radiation source (l = 1.5405Å).The transmission electron microscopy (TEM) images were collected from an instrument at 200 kV (TECNAI G2 20 LaB6).Atomic force microscopy (AFM) and scanning electron microscopy (SEM) images were acquired with an atomic force microscope (Bruker Multimode 8) and scanning electron microscope (Regulus 8100), respectively.X-ray photoelectron spectroscopy (XPS) measurement was conducted using an Escalab Xi + XPS (ThermoFisher) with a monochromatic Al Kα (1486.68 eV) x-ray source.
The physical properties of PeQD inks (i.e.surface tension, contact angle and viscosity) were measured at room temperature on an optical contact angle measuring instrument (One Attension Theta Lite) and a viscometer (Brookfield Rotational Viscometer), respectively.The current density (J)and voltage (V) characteristics were recorded on a source meter (Keithley 2400), while the luminance (L) was obtained using a luminance meter (Konica Minolta CS-200).The EL spectra were recorded on a photometer (PR-705).PL and EL images were recorded on a polarizing microscope (RANGBO LB-CX).

Results
The PL spectra of the PeQD solution with different ratios of Sr and Zn were blue-shifted when increasing the proportion of Sr element, which mainly resulted from lattice contraction, as shown in figure S1(a) (available online at stacks.iop.org/MF/1/015301/mmedia) [15].The photoluminescence quantum yield (PLQY) was 38%, 48%, 60% and 56%, respectively, corresponding to the ratios of Sr and Zn of 2:8, 4:6, 6:4 and 8:2.This indicates that the ratio of 6:4 is the best.The EL performance of the device with a 6:4 ratio of Sr and Zn was also the best one among different ratio devices, as shown in figures S1(b) and (c).The 6:4 PeQD film absorbance edge was at 640 nm and the PL peak at 648 nm, as shown in figure 1(a).The average size of the PeQDs was 12 nm (figure 1(b)), and the size distribution chart is displayed in the upper-right corner.The diffraction peaks of the as-prepared 6:4 PeQDs deposited on a glass substrate were located at 14.8 • , 29.5 • and 35.5 • , which were assigned to the (100), ( 200) and (210) planes, respectively, implying that the PeQDs have cubic crystal structure (figure 1(c)).Quantum dot (QD) films generally exhibit relatively low PLQY compared to that of the QD solution, which may be caused by ligand shedding during the film formation process [16].The ligand binding to the PeQD surface is highly dynamic, and ligands often desorb from the PeQD surface upon purification, aging and film formation.These processes lead to the formation of surface defects, and thus a reduction in PLQY [17].Various ligands with higher binding affinity were reported to overcome these difficulties [3,18,19].
Here, in order to passivate the PeQD surface defects, the EL device with the emissive layer blend of OAmBr and PeQDs was designed.The device structure was ITO/PE-DOT:PSS/PTAA/OAmBr:PeQDs/TPBi/LiF/Al.The ratio of OAmBr and PeQDs is 1 wt%.The device exhibited current efficiency of 6.7 cd A −1 , which is much higher than that of the device without OAmBr (2.1 cd A −1 ), as shown in figures 4(b) and (c).The turn-on voltage of the PeLED with OAmBr was 2.8 V, lower than that of the control device (3.1 V) because of the suppressed leakage current.However, the device efficiency is still not ideal.In order to improve the carrier balance and simultaneously passivate the surface defects of the PeQD film, an OAmBr interfacial layer was introduced between the PTAA and PeQDs layers.The experiments imply that the OAmBr interfacial layer could play a role in reducing the PeQD surface defects and blocking carrier transport.Experimental results suggest that Br ions in OAmBr could enter the PeQDs and passivate the Br vacancies of PeQDs, as discussed below.Consequently, the combined OAmBr/PeQD film showed stronger fluorescence and its PLQY was improved from 13% (pristine film) to 25% (OAmBr/PeQD film).The transient PL of the OAmBr/PeQD film increased to 4.9 from 1.7 ns of the pristine film (figure 1(d)).The AFM and SEM images showed that the pristine PeQD film had a slight aggregation, while the OAmBr/PeQD film had clear crystal grains without aggregation, which may be consistent with higher PLQY of the OAmBr/PeQD film, as shown in figures 2(a) and (b).The root-mean-square roughness (Rq) values were 3.88 and 5.21 nm for the pristine PeQD film and OAmBr/PeQD film, respectively.The increase in Rq comes from the greater roughness of the OAmBr film itself.
XPS showed that the content of Br and N of the OAmBr/PeQD film increased compared to the pristine film (figures 3(a)-(c)), while the content of I of the OAmBr/PeQD film decreased compared to the pristine film, as shown in table S1.In addition, the OAmBr/PeQD film showed a slight blue-shift of the PL spectrum and higher PLQY compared to the PeQDs pristine film, which indicated that OAmBr penetrated into the PeQD film and passivated the Br vacancies of PeQDs caused by the desorption of ligand, as shown in figure 3(d).
In order to gain more insight into the interaction between PeQDs and OAmBr, we estimated the trap density using space-charge-limited current measurement.We fabricated both hole-only devices (ITO/PEDOT:PSS/PTAA/(OAmBr)/PeQDs/4,4,4 ′′tris(carbazol-9yl)triphenylamine (TCTA)/MoO 3 /Al) and electron-only devices (ITO/ZnO/(OAmBr)/PeQDs/TPBi/LiF/ Al), respectively.The hole and electron trap densities from hole-only and electron-only devices were calculated, respectively, according to the equation N traps = 2ε 0 ε r V TFL /(qd 2 ), where ε 0 , ε r , q and d are the vacuum permittivity, relative dielectric constant of perovskite, elementary charge and distance between the electrodes, respectively [5].As shown in figures 3(e) and (f), the density of hole traps was reduced to 3.95 × 10 16 cm −3 after incorporation of the OAmBr layer, which was just half that of the controlled device (N traps = 7.49 × 10 16 cm −3 ).The density of the electron traps was also reduced from 4.93 × 10 17 to 2.59 × 10 17 cm −3 .Furthermore, the current density of the OAmBr-based device was lower than that of the controlled device, both in hole-only and electron-only devices.Among them, the current density of single-electron devices was reduced by about three orders of magnitude, and the current density of single-hole devices was reduced by an order of magnitude.The reduction of the  For inkjet printing, the morphologies of the films strongly affect the efficiency, stability and performance of the devices [20].To obtain a uniform film, various methods were proposed, such as cosolvents [20,21], mixing polymer in inkjet printing QDs [22].Due to the instability in moisture and oxygen [23], research on inkjet printing perovskite films processed in air has lagged behind.How to obtain a high-quality perovskite film using inkjet printing is still a challenge.We  developed a PeQD-polymer composite ink to improve the uniformity of the inkjet printing of PeQDs films, to enhance the EL performance of inkjet-printed PeQD-LED.The ink was composed of PeQDs and a small amount of PB dissolved in octane.PB is a kind of polymer with a high viscosity and boiling point, and good solubility in octane.It can bind to the PeQDs so that they do not diffuse to the pixel edge due to capillary flow, and a uniform film can be obtained; otherwise it is difficult to get a uniform film, as shown in figure 5(a).The viscosity, surface tension and contact angle of PTAA of the PeQD ink with and without PB are summarized in table 2. There was basically no change in viscosity and surface tension because of the small amount of PB, and the contact angle was slightly increased for the PB-based ink, as shown in figure S3, indicating that the small amount of PB had almost no influence on the ink's properties.The inkjet printing device was constructed with the structure of ITO/PEDOT:PSS/PTAA/OAmBr vapor/ D Li et al PeQDs:PB/TPBi/LiF/Al.It is noted that the OAmBr film on the pixel substrate prepared by spin-coating was uneven, so OAmBr vapor treatment method was used instead of spin coating OAmBr.The device with PeQDs:PB reached a maximum luminance of 131 cd m −2 , a peak current efficiency of 8.2 cd A −1 and EQE of 9.6%, as shown in figures 5(b) and (c).The pixel morphologies are shown in figures 5(d) and (e), which are uniform emission under a bias of 4 V.It is noted that under different biases, the morphologies of the PeQD film by inkjet printing were different due to the thickness difference in pixels, influenced by the spin coating underlayers on the matrix structure substrate, as shown in figure S4.The difference in the performance between spin-coated and inkjet-printed devices is partly a result of the difference in film morphology, fabrication method and environment.The coverage of the inkjet printed perovskite film was significantly lower than that of the spin coating one shown in figure S5.Consequently, improving film quality is still the focus through improving the device performance of inkjet printing.

Conclusion
In this work, we propose OAmBr as an interfacial layer, which fills the Br vacancies of PeQDs and improves the PLQY of the PeQD films.The device EL efficiency was increased from 3.6% to 16.5% due to the balanced carrier transport and reduced density of hole and electron traps.The PB in PeQD inks can bind PeQDs and lead to a uniform film fabricated by inkjet printing.The red matrix device exhibited an EQE of 9.6%, which to date is the best one in the literature, to the best of our knowledge.

Figure 1 .
Figure 1.(a) Absorption and PL spectra of PeQDs deposited on a quartz plate.(b) TEM image of PeQD.Inset: size distribution of PeQDs.(c) XRD pattern of PeQDs with and without OAmBr film deposited on a glass substrate.(d) PL decay curves of PeQD film with and without OAmBr.Inset: photo of PeQD film with (right) and without OAmBr (left) under UV light.

Figure 3 .
Figure 3. (a)-(c), XPS spectra of PeQD films with and without OAmBr.(a) Br 3d spectra.(b) I 3d spectra.(c) N 1s spectra.(d) Absorption and PL spectra of PeQDs with and without OAmBr.(e) J-V curve of the hole-only device with and without OAmBr layer.(f) J-V curve of the electron-only device with and without OAmBr layer.

Figure 5 .
Figure 5. (a) Drying process of one droplet of PeQDs with and without PB in air over time.(b) J-V-L characteristic, and (c) LE-J characteristic of inkjet printing PeQD-LED with and without PB.(d) PL image of inkjet-printed matrix PeQD-LED.(e) Light-on image of inkjet-printed matrix PeQD-LED with a bias of 4 V (scale bar: 200 µm).

Table 1 .
The elemental ratios of Sr, Zn and Pb in PeQDs.

Table 2 .
Viscosity and surface tension of PeQD inks.
a Measured at room temperature.