A laser film has to live between contacts
Metal halide perovskites are attractive laser materials: they are solution-processed, bright, colour-tunable and have a large optical gain. Optically pumped perovskite lasers already run continuously at room temperature, and perovskite devices already carry more than 1 kA cm−2. An electrically pumped laser needs both at once, so the perovskite must be sandwiched between transport layers that bring in electrons and holes.
In light-emitting diodes such interfaces often quench the light. Do they also raise the lasing threshold? To find out, the authors spin-coat either the hole-transport polymer PTAA or the electron-transport molecule Bphen on top of a MAPbBr3 film on glass, and pump the film through the glass with femtosecond pulses.
Schematic, not to scale, after Fig. 1 and the Experimental Section. All films are sealed with epoxy and a thin cover glass.
A narrow line grows out of the glow
At low pump fluence (14.0 μJ cm−2 at 343 nm) the film shows ordinary spontaneous photoluminescence: a broad band peaking at 538 nm, 24 nm wide (FWHM). At higher fluence an extra narrow peak appears near 551 nm and soon dominates the spectrum.
That narrow peak is amplified spontaneous emission: light emitted in the film is amplified by stimulated emission as it travels through it. It needs more excited states than ground states, the same population inversion a laser needs, so the ASE threshold is a stand-in for the lasing threshold.
Peak positions and width from the text; spectra schematic, after Fig. 1b, each normalised to its maximum.
Where the slope bends and the width jumps
Plot the integrated emission against the pump fluence on log–log axes. Below threshold it grows with a fixed slope; once gain sets in, the slope steepens. For most materials the spectrum narrows at the same point.
Here the ASE line sits so far from the spontaneous peak that the spectrum first widens when ASE appears and only narrows when ASE takes over. The threshold is taken where both the slope and the FWHM change clearly: 64.2 μJ cm−2 for the bare MAPbBr3 film.
Schematic, after Fig. 1c: only the threshold and the 24 nm starting width are from the paper.
PTAA helps, Bphen hurts
Neither layer absorbs in the emission range, and neither changes the perovskite underneath: SEM images, X-ray diffraction, absorption spectra and the emission peak position stay the same. The ASE does change. With PTAA on top, the threshold falls by 22.9%, from 64.2 to 49.5 μJ cm−2, and the ASE intensity rises by 18.8%.
With Bphen the opposite happens: the threshold rises by 14.6%, to 73.6 μJ cm−2, and the ASE intensity drops by 16.2%. A transport layer is clearly not a neutral passenger.
All values from the text (Fig. 1e,f; Bphen in Fig. S4).
Less glow, yet a lower threshold
A familiar explanation would be defect passivation: a coating heals surface traps, so the film glows longer and brighter. Below threshold (36.0 μJ cm−2, only spontaneous emission) PTAA does the opposite. The emission gets weaker and its lifetime shortens from 7.8 to 6.4 ns, because holes move from the perovskite into PTAA, which the band alignment allows.
Bphen makes the film look brighter but leaves the lifetime almost unchanged at 7.7 ns. Electrons are not extracted; the layer, with a refractive index between perovskite and air, only lets more light out of the film. So the sample with the lowest spontaneous efficiency has the lowest ASE threshold, which passivation cannot explain.
Lifetimes from the text (Fig. 2b, Fig. S5). The decays are drawn as single exponentials with these lifetimes (schematic).
Above threshold, PTAA speeds up stimulated emission
Could PTAA simply change how light leaves the film? PTAA and Bphen have similar refractive indices (about 1.7), so they should change the out-coupling in a similar way, yet their effects on ASE have opposite signs.
Far above threshold (705.4 μJ cm−2) the ASE from the PTAA-coated film is much stronger, and the emission lifetime at the ASE peak drops from 6.7 to 1.2 ns. Transient absorption agrees: the photobleach at 527 nm, which tracks the carrier density, decays much faster with PTAA. The carriers are used up quickly by stimulated emission, a sign of larger gain. With Bphen the decay is the same as in the bare film.
Values from the text (Fig. 2c,d; Fig. 3e). Decays drawn as single exponentials with these lifetimes (schematic).
Holes that leave help the rest cool down
A 343 nm photon carries far more energy than the band gap. The electron and hole it creates start out "hot", high above the band edges, and cool down by handing the excess energy to lattice vibrations. In transient absorption, cooling shows up as a red shift of the bleach peak with time.
Exponential fits give a cooling time of 0.49 ps for the bare film and the Bphen film, and 0.42 ps with PTAA. Hole cooling in MAPbBr3 is known to be slow: hot holes keep the lattice vibrations excited, a hot-phonon bottleneck. Extracting holes into PTAA eases that bottleneck, so the carriers cool faster.
Cooling times from the text (Fig. 3f, Fig. S6). Band diagram schematic; the energy levels themselves are in Fig. S1, not reproduced here.
Every loss raises the bar
The authors write the net gain as G = G0 − L. G0 grows with the number of excited carriers; L collects every process that competes with stimulated emission. Threshold is reached at G = 0. A larger loss L therefore needs more carriers, and so a higher pump fluence, before gain appears.
Hot holes add to L. Perovskites are ionic, so carriers made with high-energy photons form large polarons that distort the lattice and boost vibrations, which speeds up phonon-assisted Auger recombination, a loss that grows fast at high carrier density. The authors add a second effect: the film is slightly p-type, and excess holes disturb the coherent interaction between electron–hole pairs that lasing relies on. Removing holes with PTAA lowers both.
Schematic, after the discussion and Fig. 4a,c: the lines show the logic, not computed values.
Fewer hot carriers, lower threshold
If hot carriers are the problem, pumping with photons closer to the band gap should help even without PTAA. It does. For the bare film the threshold drops from 64.2 μJ cm−2 at 343 nm to 25.7 μJ cm−2 at 400 nm and 7.2 μJ cm−2 at 500 nm.
At 508 nm, where few hot holes are made at all, the bare and PTAA-coated films need about the same pump power (≈20 μW): with little Auger loss left, hole extraction has little to fix. The PTAA film still has the shortest emission lifetime at other excitation wavelengths, so its hole extraction does not depend on the pump colour.
Thresholds from the text (Fig. 4b; 508 nm in Fig. S8). Photon energies and the excess energy above the 538 nm emission peak (2.30 eV) are computed here from the wavelengths. Counted in photons, the threshold falls about 6-fold from 343 to 500 nm, not 9-fold, because a 500 nm photon carries less energy (also computed here).