Electrically Switchable Amplified Spontaneous Emission from Lead Halide Perovskite Film.

Electrically tunable optical devices that allow for modulation and detection of the optical signals would be extremely beneficial for the next photonic and electronic technologies. Perovskite materials as an emerging excitonic one provide promising platforms because they offer excitons manipulated by an external electrical field and efficient coupling to light. However, so far, electrically modulated switches based on perovskite amplified spontaneous emission (ASE) still remain unexplored. Here, we prepared perovskite films on indium tin oxide substrates by a spin-coating method and characterized their ASE behaviors. Based on it, we designed and fabricated electrically switchable ASE devices of perovskite film based on a light-emitting diode device configuration. Under the externally applied current, this device exhibits good controllable optoelectronic switching behaviors. Furthermore, this photoelectric response can be modulated by the different current densities. Our strategy for electrically switchable perovskite ASE will promote integrated applications in optoelectronic devices and provide valuable experience for the development of electrically pumped perovskite lasers.

Introduction

The demand for increasing information technologies calls for an efficient optoelectronic conversion device to complete the shift of signal processing from the electric to the optical domain. Among these devices, the excitonic device has been paid great attention to because they can manipulate excitons by the use of external electric or magnetic fields and efficiently couple back to light.[1−3] Of particular interest are electrical tuning techniques because they open a route to on-chip integration of photonic devices with electronics, potentially enabling high-speed modulation. For instance, carrier concentration control produces metasurfaces that have demonstrated spectral tuning with switching speeds of up to 30 MHz.[4] In addition, the integrated systems of electrically tunable lasers would significantly enhance detection sensitivity by the regulation of the laser wavelength and intensity.[5] Therefore, investigation of electrically switchable excitonic devices’ optical and electrical properties is essential to pave the way for further application in optoelectronic devices.

Lead halide perovskites, an emerging class of semiconductor materials, have attracted more attention in recent years not only because of diversity in the molecular compositions and solvable fabrication but also because of a high optical absorption coefficient, fluorescence quantum yield and equilibrium carrier transport.[6−8] They have also been demonstrated as promising candidates for optoelectronic devices, such as solar cells, light-emitting diodes (LEDs), laser diodes, photodetectors, and photocatalysis.[9−16] In the past few years, many types of perovskite amplified spontaneous emission (ASE) and lasers have been experimentally reported, for example, perovskite film ASE, Fabry–Pérot lasers, whispering-gallery-mode lasers, distributed-feedback Bragg lasers, vertical-cavity surface-emitting lasers, and polariton lasers.[14,17−21] The corresponding laser-modulation applications are also extended to single-mode laser output[22] and all-optical switch.[23] Despite the continuous success, the existing control methods for the optical switch are mainly focused on optical technologies, including linear mode coupling and bistable and nonlinear mode interaction.[22,23] The dependence on the complicated device configuration and the inferior material properties has hindered the development of electrically switchable devices. Therefore, electrically controllable perovskite ASE or lasers are highly desired.

Here, we design and fabricate electrically switchable ASE devices of perovskite film based on an LED device configuration. The CH3NH3PbBr3 (MAPbBr3) perovskite films we fabricated exhibit excellent laser gain behaviors. Under the externally applied current, this device exhibits good controllable optoelectronic switching behaviors. Furthermore, this photoelectric response can be modulated by the different current densities. Our strategy for electrically switchable perovskite lasers will promote integrated applications in optoelectronic devices and provide valuable experience for the development of electrically pumped perovskite lasers.

Results and Discussion

The MAPbBr3 perovskite film was prepared with the facile spin-coating technique (see Material and Methods).[17] From the bright-field image of the obtained sample (Figure a), it is found that the large-area and uniform perovskite film has been fabricated. Figure b depicts the absorption (solid line) and emission (dashed line) spectra of this perovskite film. A broad band gap absorption posits at about 539 nm and a photoluminescence (PL) band is at 540 nm, which are in agreement with the reported experimental results.[24] X-ray diffraction (XRD) peaks (Figure c) show that the characteristic peaks are at 14.9 and 30°, which belong to the (001) and (002) faces of the cubic perovskite phase, respectively.[25] We also checked the thickness and the quality of the obtained perovskite film by atomic force microscopy (AFM) measurement. The average thickness and the root-mean-square (rms) surface roughness are determined to be about 120 and 4.3 nm, respectively (Figure d).

Figure 1

(a) Optical image, (b) absorption and emission spectra, (c) AFM image, and (d) XRD pattern of MAPbBr3 films.

To investigate the ASE behavior of the obtained MAPbBr3 perovskite film, a 400 nm femtosecond laser was focused into a rectangle strip on the surface of the perovskite film, and the PL spectra were recorded by collecting the light emitted from the edge of the sample. Figure a shows the PL spectra as a function of pump energy density (P). At low P = 2.6 μJ cm–2, a broad spontaneous emission band is observed at 539 nm with a full width at half-maximum (fwhm) of about 29 nm (black line). When P exceeds a threshold (Pth), a sharp peak develops on the red side of the spontaneous emission peak at around 555 nm. Meanwhile, the spectral narrowing is evident by the fact of the fwhm decrease to 7.4 nm. Figure b shows the log–log plot of integrated PL intensities versus P, in which a clear inflection reveals a threshold of Pth = 2.8 μJ cm–2. We fitted the intensity dependence to a power-law x with p = 0.94 ± 0.02 and 2.0 ± 0.10 below and above the threshold, respectively, indicating a typical ASE evolution process from a linear increase for spontaneous emission to a superlinear increase for ASE.[26,27] Notably, the intensity dependence value of p = 0.94 ± 0.02 below the threshold is close to 1, suggesting that the absence of an exciton–exciton annihilation process in the perovskite film and benefit for the establishment of the ASE process.

Figure 2

(a) Pump fluence dependence of the emission from MAPbBr3 films with the excitation laser (400 nm, 150 fs pulses). (b) Corresponding threshold curve and gain curve.

To investigate the optical gain of the perovskite film, we further measured the model gain coefficient (g), an important figure-of-merit that indicates the efficiency of light amplification of laser materials, by using the variable strip method.[26,28] The ASE threshold can be expressed by the relationship I = [(Ip × A)/g] × [exp(g × L)−1], where A is a constant related to the emission cross section, I is the emission intensity, and L is the length of excitation laser stripe.[17,29] The inset of Figure b shows the emission intensity as a function of the excitation laser stripe length. The fitting curve (red line) is in good agreement with the measured lasing threshold values (black circles), giving rise to a gain coefficient of g = 186 cm–1 comparable to the value reported for thin films.[26,30] For comparison, the structure and ASE behavior of perovskite films with different thicknesses have also been systematically studied. Perovskite films with thicknesses of 50 and 500 nm are prepared. With the increase of thickness, the grain size of the surface becomes larger. The roughnesses of the film are 3.7, 4.3, and 17.4 nm for 50, 120, and 500 nm-thick films, respectively (Figure S1). Figure S2 shows XRD patterns and absorption emission spectra of 50 and 500 nm-thick perovskite films. Strong (210) and (220) crystal plane peaks appear in the 500 nm-thick perovskite film indicating that the grain accumulation in the perovskite film is more disordered. In addition, the ASE tests show that the thresholds are 1.5 and 6.2 μJ cm–2 and the gain coefficients are 168 and 80 cm–1 for 50 and 500 nm-thick films, respectively (Figure S3). With the increase of thickness, the gain coefficient of perovskite films increases slightly. For the 500 nm-thick film, it is difficult to obtain a high-quality perovskite film, resulting in the decrease of the gain coefficient. Therefore, we choose the best perovskite film with a thickness of 120 nm to construct electrically modulated ASE devices.

It is highly desirable to integrate electrical modulation to a coherent light source in a single device because this can minimize the optoelectric device volume and increase functionality in integrated photonics and electronics. In order to detect the photoelectric response relationship between the perovskite ASE and injected current, we build an electrically modulated ASE switch device based on a typical LED device (see the Experimental Section).[31] The energy level diagram of the perovskite LED device is shown in Figure a. For the measurement of the performance of this device, we built microarea optoelectronic detection equipment for the perovskite LED, and its schematic diagram is shown in Figure b.[27,32] In short, the 400 nm femtosecond laser is focused into a 100 μm circular spot and irradiated on the surface of the perovskite film through a microscope and passes through the indium tin oxide (ITO) and PEDOT layers. Upon laser excitation, the emitted green ASE from the perovskite film is collected by the microscope and then is guided into the spectrometer.

Figure 3

(a) Band alignment of each function layer in the devices. (b) Schematic diagram of test equipment for the electrically modulated perovskite laser. Spectral intensity changes with the power on (c) and off (d).

We adjust the energy density of the pumped laser to 120 μJ cm–2, and the clear ASE from the perovskite film can be observed. Figure c depicts PL spectra from the perovskite LED device under an operation current of 40 mA cm–2. Within the current injection, the ASE intensity at 547 nm rapidly decreases from 25 (black line) to 15 (red line). Only spontaneous fluorescence emission at 540 nm remains, and its intensity remains unchanged after the injected current is applied for 3 s. Interestingly, when we remove the applied current, the ASE peak at 547 nm appears immediately within 1 s. Upon prolongation of the applied current time, the ASE intensities gradually increase and eventually recover to the intensity of 25 (Figure d). Thus, a reversible device, which functions like an electrically modulated optical switch, has been successfully realized through monitoring the relationship between the ASE signal intensity and current density injection. As this photoelectric response should be very fast (less than 1 s), regrettably, we cannot obtain the ultrafast response speed limited by our simple instrument.

We attempt to understand the mechanism of this electrical switch. As it is well known, the PL origin of three-dimensional perovskites is carrier recombination at the band edge between free electrons in the valence band and free holes in the conduction band,[33,34] and the laser gain is through an electron–hole plasma-stimulated emission mechanism.[26,35] After pumped by laser excitation, a large number of photogenerated excitons are produced and establish the process of population inversion. However, the stability of excitons is very susceptible to the influence of carrier concentration and temperature. For example, the excitons are easy to decompose on the condition of the high carrier concentration due to the shielding effect of free charge on the Coulomb field. With the increase of the electron or hole concentration, the Coulomb repulsion of electron–electron or hole–hole is significantly improved, which often reduces the Coulomb attraction and may lead to exciton dissociation. Therefore, in our case, electrons and holes are produced under the action of an electric field when the current is injected into the perovskite LED. At a low current density such as 10 mA cm–2 (the experimental results are not shown), the quantity of electron-generated excitons originated from the injected current is relatively small and their Coulomb interaction is not enough to bring about significant exciton dissociation. Thus, the obvious effect of the electric switch was not observed. Only when larger current densities are applied, such as 40, 50, and 110 mA cm–2, abundant electrons and holes are produced and their interaction results in the significant decrease of the effective quantity of photogenerated excitons, which gives rise to destruction of the population reversion process and makes perovskite ASE disappear. So, the clear electric switching effect of ASE has been observed experimentally. Therefore, the interactions between photogenerated excitons and electron-generated electrons and holes result in the regulation of the population reversion process and the establishment of the electric switch of perovskite ASE. Indeed, the mechanism of the shielding effect of free charge on the Coulomb field is also used to explain the efficiency roll-off caused by exciton quenching in the perovskite LED,[36,37] that is, the current efficiency will not increase or even decrease when the injected current is greater than 20 mA cm–2.

To check the performance of the obtained optoelectronic switch device, we measure ASE behaviors under different injection current densities, such as 20, 40, 50, and 110 mA cm–2. As shown in Figure a, it should be noted that the current ASE switch has no obvious effect when the current density is less than 20 mA cm–2. With the increase of current density, the response of ASE intensity to current density becomes faster and greater. For example, when the applied current density is 20 mA cm–2, the intensities of the ASE spectrum (black line) decrease from 27 to 24 instantaneously and are stable at 18 after about 10 s. When the applied current is removed, the spectral intensity immediately rises to 24 and gradually returns to 27 within 10 s. For a current density of 40 mA cm–2, the intensities of the ASE spectrum (red line) decrease from 27 to 14 instantaneously and are stable at 11 after about 10 s. The spectral intensity immediately rises to 21 and gradually returns to 27 as soon as the applied current is removed. For a higher current density of 110 mA cm–2, the intensity of the ASE spectrum (blue line) suddenly drops to 4 and remains stable. After removing the current, the spectral intensity immediately rises only to 5.7 and finally slowly recovers to about 18. The decrease in the ASE intensity might be because the perovskite film could be damaged under the high current injection. Remarkably, under a current density of 20 mA cm–2, this switch can be reversibly switched over at least 10 cycles without any fatigue (Figure b).

Figure 4

(a) Laser behavior of the perovskite at different current densities. (b) Cycle performance of the switch.

Conclusions

In summary, perovskite films with different thicknesses were prepared by a one-step solution method. We investigated their ASE gain behavior and selected a 120 nm-thick perovskite film with the largest optical gain coefficient of 186 cm–1 to construct an electrically modulated ASE device. Under different applied currents, the ASE behavior of the perovskite film shows different responses. When the current density is lower than 20 mA cm–1, the spectral intensity of ASE has no obvious change. When the current density is higher than 20 mA cm–1, the high concentration of carriers will induce exciton dissociation, resulting in the decrease of spectral intensity and the disappearance of the ASE peak. Moreover, the spectral response rates increase significantly as the applied current increases. Our strategy for electrically switchable perovskite ASE will promote integrated applications in optoelectronic devices and provide valuable experience for the development of electrically pumped perovskite lasers.

Materials and Methods

Lead bromide (99.99% metals basis) was purchased from Alfa Aesar. MABr, PEDOT:PSS, and TPBi were purchased from Xi’an Polymer Light Technology Corp. N,N-dimethylformamide (DMF) and chlorobenzene were purchased from Beijing Chemical Agent Ltd., China. All salts and solvents were used as received without any further purification.

MAPbBr3 precursor solution was prepared by dissolving MABr and PbBr2 with the molar ratio of 1:1 in DMF. Then, MAPbBr3 films were prepared in a nitrogen-filled glovebox by spin-coating the precursor solution (with the concentration of 0.6 M) onto ITO at a speed of 6000 rpm. During this process, a drop of chlorobenzene was added to obtain a better quality, followed by annealing on a hot plate at 100 °C for 30 min. Here, the addition of chlorobenzene antisolvent will help to obtain a smooth MAPbBr3 perovskite film.

In a typical progress, PEDOT:PSS solutions (Baytron P VP Al 4083, filtered through a 0.22 μm filter) were spin-coated onto the ITO substrates at 3000 rpm for 60 s and baked at 140 °C for 15 min. Then, the perovskite film was prepared by spin-coating MAPbBr3 precursor solution at 6000 rpm for 60 s. Finally, TPBi (60 nm) and LiF/Al electrodes (1 nm/100 nm) were deposited using a thermal evaporation system through a shadow mask under a high vacuum of 2 × 10–4 Pa. The device active area was estimated to be 10 mm2 as defined by the overlapping area of the ITO and Al electrodes.

The steady-state optical absorption was characterized on a Shimadzu UV-3600 spectrometer. The XRD patterns were measured by a D/Max 2400 X-ray diffractometer with Cu Kα radiation (λ = 1.54050 Å) operated in the 2θ range from 5 to 40°. The excitation source is a Xenon lamp equipped with a band-pass filter (330–380 nm for UV-light and 460–490 nm for blue light), and the samples were deposited onto an ITO substrate. The morphology of these perovskite films was characterized by AFM tests carried out on a Bruker MultiMode 8 atomic force microscope.

Room-temperature gain measurements were performed using standard VSL methods. The femtosecond excitation laser was focused by a cylindrical lens (with focal length f = 2 cm) into a rectangle strip (0.018 cm width and variable length) on the surface of MAPbBr3 thin films. The excitation stripe length was varied through an adjustable slit actuated by a micrometer which was placed at the focal line of the cylindrical lens. Moreover, the PL spectra were recorded by collecting the light emitted from the edge of the sample. For ASE behavior in the LED device, the second harmonic (400 nm, 150 fs, 1 kHz) of a regenerative amplifier (Spitfire, Spectra-Physics) seeded with a mode-locked Ti:sapphire laser (Tsunami, Spectra-Physics) was focused to a 100 μm diameter spot to excite the perovskite films. Then, PL spectra were collected under a reflection mode through the same 50 × 0.9NA objective that was mounted on a 3D movable stage. A 430 nm long-wave-pass dielectric filter was used to block any scattered excitation light. Finally, the collected PL was coupled to an optical fiber and detected using a liquid nitrogen-cooled CCD (SPEC-10-400B/LbN, Roper Scientific) attached to a polychromator (Spectropro-550i, Acton). The spectral resolution is 0.1 nm.