Skip to main content
NCBI home page
Science Advances logoLink to Science Advances
. 2023 Sep 29;9(39):eadi9347. doi: 10.1126/sciadv.adi9347

Uncovering the mechanisms of efficient upconversion in two-dimensional perovskites with anti-Stokes shift up to 220 meV

Bo Wu 1,2,†,*, Aocheng Wang 2,3,†, Jing Fu 4,†, Yutong Zhang 2,3, Cheng Yang 1, Yiyang Gong 1,2, Chuanxiu Jiang 2,3, Mingzhu Long 1, Guofu Zhou 1, Shuai Yue 2,3, Wei Ma 4,*, Xinfeng Liu 2,3,*
PMCID: PMC10541006  PMID: 37774031

Abstract

Phonon-assisted photon upconversion holds great potential for numerous applications, e.g., optical refrigeration. However, traditional semiconductors face energy gain limitations due to thermal energy, typically achieving only ~25 milli–electron volts at room temperature. Here, we demonstrate that quasi–two-dimensional perovskites, with a soft hybrid organic-inorganic lattice, can efficiently upconvert photons with an anti-Stokes shift exceeding 200 milli–electron volts. By using microscopic transient absorption measurements and density functional theory calculations, we explicate that the giant energy gain stems from strong lattice fluctuation leading to a picosecond timescale transient band energy renormalization with a large energy variation of around ±180 milli–electron volts at room temperature. The motion of organic molecules drives the deformation of inorganic framework, providing energy and local states necessary for efficient upconversion within a time constant of around 1 ps. These results establish a deep understanding of perovskite-based photon upconversion and offer previously unknown insights into the development of various upconversion applications.

Upconversion photoluminescence in two-dimensional perovskite originates from the strong fluctuation of the soft hybrid lattice.

INTRODUCTION

Photon upconversion (UC) is a physical process where the emitted photon energy is greater than the absorbed photon energy, which differs from the down conversion (DC) process that follows Stokes’ law. UC can be achieved through various methods, including multiphoton nonlinear absorption, triplet-triplet annihilation in organic molecules, energy transfer in rare-earth metal doped materials, and thermal-assisted photon absorption (1–3). The latter, which generates anti-Stokes photoluminescence (PL), forms the basis for solid-state laser cooling (4), as well as numerous other applications, such as photovoltaics (1), fluorescence imaging (3), anticounterfeiting (5), and lasers (6). A few recent studies have shown that some low-dimensional materials such as nanotubes (7), nanobelts (4), quantum dots (8, 9), transition metal dichalcogenide monolayers (10, 11), and diamond color centers (12) can achieve large energy gains of around 100 meV, which greatly surpasses the thermal energy (~25 meV). However, the role of phonons in enabling such a large anti-Stokes shift in these materials is not yet fully understood. Some researchers propose that the UC is a multiphonon process occurring through intermediate levels such as subgap traps, surface states, and Urbach tail (7–9, 13, 14) and can be enhanced by doubly resonant Raman scattering (10, 11) and excitonic effects (15). Others attribute it to a single-step process involving the absorption of a large optical phonon with energy greater than 100 meV (7). Further investigation into the microscopic physics of photon-electron-phonon interactions behind the UC processes is crucial to enhance the efficiency and energy gain of thermal-assisted UC.

In this work, we show that a quasi–two-dimensional (2D) perovskite, namely, phenethylammonium lead iodine perovskite [(PEA)2PbI4, or PEPI], can facilitate efficient UC with an enormous anti-Stokes shift of up to 220 meV. As far as we are aware, this energy shift is the highest reported for phonon-assisted UC. By combining microscopic transient absorption spectroscopy with density function theory (DFT) calculations, we reveal that the exceptional energy gain originates from the strong fluctuation of the soft organic-inorganic hybrid lattice, which brings about a notable band energy variation of more than 0.3 eV at room temperature. The motion of the organic molecule initiates the lattice fluctuation, driving the harmonic and anharmonic motion of the inorganic framework. The strong lattice fluctuation produces abundant phonon-dressed electronic density of states below the bandgap enabling considerable subgap photon absorption. The electron-hole pairs generated below the bandgap then couple to the picosecond-fast lattice deformation and move to the equilibrium states where they radiatively recombine. Our findings clarify the UC mechanisms in perovskite materials with a soft lattice and have substantial implications for the design of previously unknown UC materials.

RESULTS

The UC PL properties

The chemical structure of PEPI is shown in Fig. 1A. PEPI crystalline platelets were grown using an antisolvent method developed by Shi et al. (16) and further refined by our group. The platelets have lateral dimensions of tens of micrometers (μm) and a thickness of a few hundred nanometers (nm), as shown in representative optical and scanning electron microscopic images in Fig. 1B, with a thickness of around 300 to 400 nm (fig. S1). As-grown PEPI platelets exhibit high phase purity and excellent crystalline quality, which are evident from energy-dispersive x-ray (EDX) spectroscopy analysis (Fig. 1B), x-ray diffraction pattern, and Raman spectrum (fig. S2).

Fig. 1. The DC and UC PL properties of PEPI micro platelets.

Fig. 1.

Open in a new tab

(A) Chemical structure of PEPI. (B) Scanning electron microscopy image and EDX mapping of the C, Pb, and I elements of a PEPI micro platelet. Scale bar, 5 μm. (C) Absorption, PLE, and excitation energy–dependent PL spectra of a PEPI micro platelet. a.u., arbitrary units. (D) Excitation power dependence of the PL intensity for a PEPI micro platelet. (E) Temperature dependence of the UC and DC PL intensity and bandwidth.

Figure 1C displays the PL spectra of a PEPI platelet that was excited using DC and UC processes at room temperature. The DC PL excited by a 400-nm laser is centered at 526 nm (2.357 eV) and has a full width at half maximum of around 74 meV. A prominent low-energy tail can be observed in PL as well as absorption spectra that arise from momentarily self-trapped exciton by the lattice vibration (17). The UC PL excited by 552-, 566-, and 580-nm lasers is centered at the same position. For excitation wavelength λexc < 580 nm, both DC and UC PL display a linear trend on the excitation density, indicating that the PL originates from exciton radiative recombination (Fig. 1D). The PL excited via two-photon absorption starts to prevail over that via one-photon absorption for λexc ~ 620 nm, which shows a quadratic excitation density dependence. Although Auger recombination could also contribute to UC PL by providing energy for a third charge carrier to excite to the band edge (18), we did not observe any cubic excitation density dependence of the UC PL and therefore can exclude the contribution from Auger recombination. The PL results are consistent with the PL excitation (PLE) spectrum, which shows considerable intensity below the emission peak up to approximately 580 nm (Fig. 1C). The anti-Stokes shift, defined as the energy difference between the excitation and emission peak photon energy, is around 220 meV for the PEPI platelets obtained using excitation wavelength λexc = 580 nm (2.138 eV). This energy gain is the highest reported for anti-Stokes PL to our knowledge (table S1) and approximately nine times that of the thermal energy (~25 meV). In the classic physical picture, the thermal energy available from the phonon bath seems too small to explain such a large anti-Stokes shift.

To unravel the potential mechanisms of the UC in perovskites, we conducted temperature-dependent studies on the UC PL. The results show that while the PL intensity of the DC process excited by a 405-nm laser decreases with temperature, the trend for UC PL excited by a 560-nm laser was the opposite, indicating that it is thermally activated and proportional to the phonon occupation number (Fig. 1E). The energy gain reduces to ~120 meV as the temperature decreases to 80 K (fig. S3), further confirming the thermal origin of UC PL. The electron-phonon coupling mechanisms can be pinpointed through the temperature-dependent PL linewidth measurements (Fig. 1E). The DC PL linewidth increases with temperature and can be well described by the Debye-Einstein approximation: ΔΓ = Γ0 + γacT + γ0/[exp(E0/kBT) − 1], where the first, second, and third terms at the right side represent the inhomogeneous broadening and the homogeneous broadening by acoustic and longitudinal optical (LO) phonon scattering, respectively. γac and γo are the electron-acoustic phonon and electron-LO phonon (Fröhlich) coupling strengths, respectively. Eo is the optical phonon energy. By neglecting the minor contribution by acoustic phonon scattering, we obtained a γo = 321 ± 93 meV and Eo = 49 ± 6 meV. These values are in consistency with previous reports (19). The large γo indicates that the quantum confinement endows the perovskite with strong electron-phonon coupling. The extracted optical phonon energy is similar to the energy spacing between absorption sidebands at low temperature (20). The latter was previously attributed to either the exciton-phonon renormalization energy (20) or the torsional motion of the NH3 group and the adjacent Pb-X cage (21). In either case, the motion of organic molecules may couple to that of electrons and alter the electronic properties (22). The broader linewidth of UC PL (30 to 40 meV) compared to DC PL may be caused by additional broadening from scatterings by defects and phonons during relaxation, as well as different electron-phonon coupling with different excitation energies (23).

The UC dynamics

We further carried out microscopic transient absorption (TA) measurements to disentangle how lattice vibration assists the photon UC process in the quasi-2D perovskite. Figure 2 shows the TA spectra of thin platelets (~30 nm thick) measured in transmission mode and fig. S4 is the TA spectra of thick platelets (~300 nm thick) measured in reflection mode. The TA spectra upon 400 nm above bandgap excitation displays a negative peak at the exciton resonance (ΔA < 0) and two positive peaks at either side (ΔA > 0) (Fig. 2A). These features are a combination of effects including the state filling of exciton band and exciton band broadening and shift by many-body interactions in the presence of other excitons and carriers (23, 24). Upon below gap excitation (565 nm), the TA spectra display an instantaneous giant response at time zero corresponding to the blueshift of exciton (Fig. 2B and fig. S5). This is due to the optical Stark effect (OSE) that emerges from the hybridization of equilibrium and Floquet states, i.e., the photon dressed electronic states (25, 26). After this instantaneous blue shift, the transient spectral feature for below gap excitation is similar to that for above gap excitation, except for that the former is coupled to prominent oscillatory component. Figure 2C displays the transient dynamics monitored at the exciton band (~522 nm) for DC and UC excitations. The rise time for the exciton band bleach following 400-nm DC excitation is around 0.35 ± 0.15 ps, which corresponds to the time required for excitations to scatter by optical phonons to the band edge. (27) For 565-nm UC excitation, after removing the strong nonlinear response by OSE, it is found that the rising of exciton band bleach is relatively slower with a time constant of 1.2 ± 0.2 ps. Similar rising time can be observed in the reflection mode (fig. S4). Further increase of the pump wavelength leads to more pronounced oscillation component that covers the rise of exciton band bleach. In addition, two-photon absorption may contribute for longer wavelength pumping due to a much larger pump fluence used in TA than PL. These make it difficult to extract the exact time constant of the rise. We estimate that the rise time does not change substantially upon longer wavelength pumping, i.e., being around 1 ps (fig. S6). The slow rise implies that the photon UC is not a one-step process via the absorption of a high-energy organic molecule vibration mode (>100 meV), which would occur at a timescale on the order of a phonon period (~10 fs).

Fig. 2. TA of PEPI micro platelets.

Fig. 2.

Open in a new tab

TA spectra of a PEPI micro platelet upon (A) DC (400 nm) and (B) UC (565 nm) excitations. Top: TA pseudocolor mapping. Bottom: TA spectra at selected delay time. (C) Exciton band bleach dynamics monitored at 522 nm for DC and UC excitations.

The slow rise of the exciton band bleach is accompanied by the damping of the oscillatory component. As shown in Fig. 3A, we extracted the oscillatory component from the TA data using analysis method elaborated in Materials and Methods. The oscillatory component can be attributed to coherent phonon oscillations (CPOs) that are generated by the intense pump optical field driving the coherent motion of the lattice via impulsive stimulated Raman scattering (23, 28–30). Through fast Fourier transformation (FFT), we identify two dominant oscillation frequencies: 42.5 and 47.5 cm−1 (Fig. 3B). These two modes correspond well with M3 and M4 phonon modes reported previously (31). M3 can be attributed to the Pb-I-Pb bending and Pb-I stretching modes. M4 is assigned to the scissoring of Pb-I-Pb angle (fig. S7) (23). It indicates the energy of excited species upon UC excitation is strongly modulated by the motion of a vibrational wave packet along the normal coordinate defined by M3 and M4 modes, more specifically, the change of Pb-I bonds in length or orientation (23). Furthermore, M1 mode (18.9 cm−1) related to the twisting of the PbI6 octahedra along two pseudo-cubic axes can also be faintly observed. However, the dominant phonon mode M2 (~35 cm−1) that observed for DC excitation shows diminished intensity for UC excitation (31). M2 is partially assigned to the Pb displacement, which couples strongly to the conduction band electrons (31). Therefore, we rationalize that UC excitation generate charge-neutral cold excitons whose energy is relatively immune to the Pb displacement, in contrast to free charge carrier above the bandgap.

Fig. 3. CPOs of PEPI micro platelets extracted from TA measurements.

Fig. 3.

Open in a new tab

(A) The pseudocolor mapping of CPOs near the exciton resonance. (B) FFT of the CPOs showing prominent contribution from M1, M3, and M4 modes but not M2 mode. (C) The damping of CPOs monitored at 524 nm and its fitting using a multioscillator model. (D) Schematics of the motions of organic molecules and associated lattice distortion.

The time constant of the CPO damping is around 1.6 ± 0.7 ps, which matches well with that of the exciton relaxation in the UC (Fig. 3C). The CPO damping is as a result of the rapid loss of vibration coherence through phonon-phonon scattering. This happens in halide perovskites due to the strong anharmonicity of the soft hybrid lattice (29, 32). Thermal fluctuation can strongly alter Pb─X bonds mainly by increasing the tilting of inorganic octahedra and therefore modify the electronic/excitonic energy that is derived from lead and halide orbitals (33). Organic molecules at A-site may provide additional structural disorder through its thermal motions, including the rotation within the sheet plane and the tilting with respect to the sheet normal (Fig. 3D) (22, 34, 35). These motions have characteristic time on the order of a few picoseconds, which drive the dynamic deformation of the inorganic framework at a similar rate via the noncovalent binding between the two (22). Therefore, the CPO damping time can be severely affected by the A-site organic molecule. For another 2D perovskite butyl ammonium lead iodide (BAPI), we observed that the cold excitons relax to the thermal equilibrium state with a time constant of 0.51 ± 0.15 ps upon 560-nm UC excitation (fig. S8). BAPI is less rigid than PEPI. The latter can form extensive network of pi-hydrogen bonds and be more space filling with the aromatic ring (36). Therefore, it shows more pronounced bandgap fluctuation (37) and faster CPO damping rate (29), which agrees well with the faster UC than PEPI. Given that the energy of excitons generated upon UC excitation is strongly modulated by Pb-I phonon modes, we infer that the cold excitons relax to thermal-equilibrium state by coupling to the lattice vibrations during its anharmonic motion in a few picoseconds.

DISCUSSION

To understand the intricate interplay between the electronic excitation and the organic-inorganic hybrid lattice in the UC, DFT calculations were further performed. Figure 4A illustrates a representative time evolution of the band edge states during the 5-ps ab initio nonadiabatic molecular dynamics (NAMD) simulations at 300 K. We observe that both the valence band maximum (VBM) and conduction band minimum (CBM) exhibit strong fluctuations which increase with temperature. Within the 5-ps trajectory, the standard variations of the band energies rise from 77 (132) meV at 100 K to 146 (171) meV at 300 K for the VBM (CBM) (Fig. 4B and fig. S9). This brings about an increase of bandgap variation from 152 to 179 meV. The giant band edge fluctuations correlate with the large variations in the Pb─I bonds in PEPI. We imitated the UC process by initially populating the phonon renormalized CBM (CBM0) with electrons and monitoring their relaxation dynamics. The phonon renormalized CBM represents the lowest energy levels that can be excited by subgap photon absorption. Figure 4C shows the time evolution of electron population at the band edges. At 100 K, there is negligible exchange in electron population between different energy levels within 2 ps. However, at 300 K, the electrons at CBM0 show a drastic drop in population, with a time constant of about 1.3 ps to decrease to 1/e of the initial population. The time correlates well with the exciton relaxation and lattice anharmonic motion time observed in above experiments (1.2 to 1.6 ps). Concomitantly, the higher energy levels just above the CBM (CBM1 + CBM2, ECBM1,2 − ECBM0 = 0.2 to 0.4 eV) are populated, suggesting that photoexcited electrons have jumped to adjacent higher excited states by coupling to lattice vibrations. These higher-energy levels may approach the real band edge states where electrons accumulate and form a quasi-equilibrium distribution at room temperature. The phonon modes participated in UC is analyzed by Fourier transform of the autocorrelation functions of energy difference fluctuations between the initial and final states during NAMD simulations. We can clearly observe M1 and M3 modes that are prominent in TA measurement (fig. S7). Another mode M5 (~68 cm−1) related to the Pb-I symmetric stretching shows the highest coupling strength that is, however, very weak in TA measurement (38). This discrepancy could be due to many factors. For example, our NAMD calculations are founded on electronic bands, whereas excitons emerge as the dominant excited species in actual scenarios. Moreover, the coupling modes are greatly influenced by the electronic landscape (23). We also find substantial contributions from high-frequency phonon modes to UC transitions at room temperature, which diminish at low temperatures (fig. S7). These modes may include high-frequency inorganic modes (e.g., octahedral tilting in-plane), or coupled vibronic modes between organic and inorganic parts (37, 38). This further support that the system has high lattice deformation and strong polaronic binding at elevated temperature driving the electronic UC process.

Fig. 4. MD simulations of band energy fluctuations and carrier relaxation in UC process.

Fig. 4.

Open in a new tab

(A) Band energy evolution in a representative MD simulation at 300 K. VBM0 and CBM0 are in cyan and red. CBM1 and CBM2 are in green and orange. (B) The standard variations of the band energy at CBM and VBM at different temperatures. (C) The simulated carrier relaxation in the UC process at 100 and 300 K.

We highlight that the photophysical mechanism of UC in perovskite materials differs from that in traditional semiconductors because of the presence of a soft lattice. In rigid lattice semiconductors, anti-Stokes emission occurs when the excitation of a higher vibrational state in the ground state absorbs one or more phonons in the UC process. This low probability of UC in these materials can be explained by the Boltzmann distribution. In hybrid perovskites, however, the lattice deformation resulting from strong electron-phonon coupling leads to notable band energy fluctuations. The motion of the organic cation at picosecond timescales causes the entire lattice to deform, resulting in rapid changes in the phonon-renormalized electron/exciton energy. This provides the low-energy excitons with sufficient energy to reach quasi-equilibrium states (free exciton) where they can recombine radiatively (schematic diagram shown Fig. 5). Hence, the efficient UC in quasi-2D perovskites is not directly attributed to the absorption of specific phonon modes in the perturbative theory. Instead, it arises from the alteration of electronic energy linked to the pronounced thermal-driven deformation of the entire lattice or, alternatively, the formation of dynamic polarons (17, 39, 40). The band energy fluctuation, reaching approximately ±180 meV at room temperature due to the strong nonperturbative interaction with lattice deformation, allows for a remarkable UC energy gain that cannot be attained in traditional semiconductors.

Fig. 5. Schematics of the energy diagram and dynamics of UC in perovskites.

Fig. 5.

Open in a new tab

Strong lattice fluctuation produces abundant phonon-dressed (polaronic) states below the bandgap for light absorption. Excitons are generated following subgap light absorption. These low-temperature excitons jump to the high-lying free exciton (FE) states by coupling to strong lattice fluctuation within ~1 ps, where they radiatively recombine and yield UC PL.

To sum up, we have investigated the effective conversion of photons in a quasi-2D perovskite material PEPI. Our results show that, even at room temperature, the anti-Stokes shift can reach up to 200 meV. We have connected this substantial energy gain in the UC process to the rapid lattice deformation, which occurs within ~1 ps, possibly assisted by the movement of organic cations. The MD simulations have revealed that the rapid lattice deformation can lead to strong fluctuations in band energy and provide adequate lattice energy for UC. Our findings can be extended to other perovskites. In the future, more research is needed to optimize the electron-phonon coupling strength and rate by adjusting the organic cations and hybrid soft lattice for even more efficient UC.

MATERIALS AND METHODS

PEPI platelet synthesis

The PEPI micro platelets were synthesized by solution antisolvent methods. PEA·HI (10 mg; 99.99%, Sigma-Aldrich), 14 mg of MAI3 (99.99%, Sigma-Aldrich), and 18.4 mg of PbI2 (99.99%, Sigma-Aldrich) were fully dissolved in chlorobenzene (CB) (Macklin, 1 ml) and DMF (Macklin, 1 ml) mixed solvent and stewing for 20 min, which is stocked for further use. For each synthesis, 200 μl of stock solution was added into 11.8 ml of CB, acetonitrile, and dichlorobenzene cosolvent to dilute 60 times. Diluted solution (20 μl) was dropped onto a precleaned sapphire substrate, which was then put into a small beaker (10 ml). The small beaker was further put into a large beaker (25 ml) that contains CB (3 ml) as antisolvent and sealed by tinfoil. The sealed beaker was stored in a refrigerator at a temperature of −10°C for 2 days, and the micro platelets would grow on the substrate.

TA measurements

Macroscopic TA measurements on thin films were performed using the HELIOS commercial system (Ultrafast Systems). Microscopic TA measurements on micro platelets were performed using a home-built microscopic system integrated into HELIOS. Fundamental 800-nm pluses (1 kHz, 100 fs) from a Coherent Astrella regenerative amplifier were used to pump an optical parametric amplifier (Coherent, OperA Solo) to obtain the frequency-tunable pump beam. The probe beam was obtained by focusing a small fraction of the fundamental 800-nm beam on to a sapphire plate. The time trace of oscillating components at the probe wavelength is obtained by subtracting the population dynamics (empirically approximated as ninth-order polynomials) from the transient spectrum. A Lomb-Scargle periodogram was coded using MATLAB to obtain the phonon modes with high-frequency resolution from the transient unevenly sampled data.

PL measurements

PL spectra were collected using a Princeton Instrument (SP2500i) spectrometer system. The laser beam was focused onto PEPI micro platelets using an objective lens (Leica, 100×, numerical aperture = 0.95). The emission signal was collected by the same lens and an optical fiber coupled into spectrometer with a liquid nitrogen cooled charge-coupled device detector.

DFT calculations

First-principles calculations were carried out to study the ground- and excited-state properties of perovskite based on DFT. Geometry optimization, electronic structure calculation, and ground-state molecular dynamics (MD) are performed with the VASP code using plane wave basis sets, Perdew–Burke–Ernzerhof exchange-correlation functional, and the projector augmented wave potentials. Excited electronic dynamics make use of the quantum classical decoherence-induced surface hopping technique, implemented within the time-dependent Kohn-Sham theory. The time-dependent ab initio NAMD simulations were performed using the Hefei-NAMD code. The optimized lattice vectors were a = 9.13 Å, b = 7.12 Å, and 18.11 Å for (PEA)2PbI4. The calculated structure yields a bandgap of 2.35 eV, which is in good agreement with previous studies. A 400-eV energy cutoff for plane waves and 4 × 6 × 2 Γ-centered Monkhorst-Pack k-points mesh were used. The convergence threshold for the energy of electronic self-consistent iteration was set to be 10−6 eV. Geometries were fully relaxed with forces on each atom less than 0.01 eV/Å. Then, the system was heated to 100 and 300 K by repeated velocity rescaling. After that, a 5-ps ab initio molecular dynamics trajectory was obtained using the microcanonical ensemble with a 1-fs time step. The 2-ps nonadiabatic Hamiltonians were iterated 500 times to simulate the dynamics on the nanosecond scale. The NAMD results were obtained by averaging more than 300 random initial configurations selected from the MD trajectory.

Acknowledgments

Funding: The research was supported by the following grants: The authors thank the support of the National Key Research and Development Program (2022YFA1603701), the National Natural Science Foundation of China (22073022, 62175068, 22173025, 11704207, and 62004072); the Strategic Priority Research Program of Chinese Academy of Sciences (XDB36000000); the Guangdong Basic and Applied Basic Research Foundation (2023B1515020024 and 2019B151502028), and the Guangdong Provincial Key Laboratory of Optical Information Materials and Technology (2017B030301007).

Author contributions: Conceptualization: X.L. and B.W. Methodology: X.L., B.W., W.M., M.L., S.Y., and C.J. Investigation: B.W., A.W., Y.Z., J.F., and Y.G. Visualization: C.Y. Supervision: X.L., B.W., and W.M. Writing—original draft: B.W. and X.L. Writing—review and editing: B.W., X.L., W.M., and G.Z.

Competing interests: The authors declare that they have no competing interests.

Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials.

Supplementary Materials

This PDF file includes:

Figs. S1 to S9

Table S1

Click here for additional data file. (869.6KB, pdf)

REFERENCES AND NOTES

  • 1.Richards B. S., Hudry D., Busko D., Turshatov A., Howard I. A., Photon upconversion for photovoltaics and photocatalysis: A critical review. Chem. Rev. 121, 9165–9195 (2021). [DOI] [PubMed] [Google Scholar]
  • 2.Auzel F., Upconversion and anti-stokes processes with f and d Ions in solids. Chem. Rev. 104, 139–174 (2004). [DOI] [PubMed] [Google Scholar]
  • 3.Zhou J., Liu Z., Li F., Upconversion nanophosphors for small-animal imaging. Chem. Soc. Rev. 41, 1323–1349 (2012). [DOI] [PubMed] [Google Scholar]
  • 4.Zhang J., Li D., Chen R., Xiong Q., Laser cooling of a semiconductor by 40 kelvin. Nature 493, 504–508 (2013). [DOI] [PubMed] [Google Scholar]
  • 5.Liu X., Wang Y., Li X., Yi Z., Deng R., Liang L., Xie X., Loong D. T. B., Song S., Fan D., All A. H., Zhang H., Huang L., Liu X., Binary temporal upconversion codes of Mn2+-activated nanoparticles for multilevel anti-counterfeiting. Nat. Commun. 8, 899 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Scheps R., Upconversion laser processes. Prog. Quantum Electron. 20, 271–358 (1996). [Google Scholar]
  • 7.Akizuki N., Aota S., Mouri S., Matsuda K., Miyauchi Y., Efficient near-infrared up-conversion photoluminescence in carbon nanotubes. Nat. Commun. 6, 8920 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Ye Z., Lin X., Wang N., Zhou J., Zhu M., Qin H., Peng X., Phonon-assisted up-conversion photoluminescence of quantum dots. Nat. Commun. 12, 4283 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Wang X., Yu W. W., Zhang J., Aldana J., Peng X., Xiao M., Photoluminescence upconversion in colloidal CdTe quantum dots. Phys. Rev. B. 68, 125318 (2003). [Google Scholar]
  • 10.Jadczak J., Bryja L., Kutrowska-Girzycka J., Kapuściński P., Bieniek M., Huang Y.-S., Hawrylak P., Room temperature multi-phonon upconversion photoluminescence in monolayer semiconductor WS2. Nat. Commun. 10, 107 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Jones A. M., Yu H., Schaibley J. R., Yan J., Mandrus D. G., Taniguchi T., Watanabe K., Dery H., Yao W., Xu X., Excitonic luminescence upconversion in a two-dimensional semiconductor. Nat. Phys. 12, 323–327 (2016). [Google Scholar]
  • 12.Tran T. T., Regan B., Ekimov E. A., Mu Z., Zhou Y., Gao W.-B., Narang P., Solntsev A. S., Toth M., Aharonovich I., Bradac C., Anti-Stokes excitation of solid-state quantum emitters for nanoscale thermometry. Sci. Adv. 5, eaav9180 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Poles E., Selmarten D. C., Mićić O. I., Nozik A. J., Anti-Stokes photoluminescence in colloidal semiconductor quantum dots. Appl. Phys. Lett. 75, 971–973 (1999). [Google Scholar]
  • 14.Khurgin J. B., Role of bandtail states in laser cooling of semiconductors. Phys. Rev. B 77, 235206 (2008). [Google Scholar]
  • 15.Rupper G., Kwong N. H., Binder R., Large excitonic enhancement of optical refrigeration in semiconductors. Phys. Rev. Lett. 97, 117401 (2006). [DOI] [PubMed] [Google Scholar]
  • 16.Shi E., Yuan B., Shiring S. B., Gao Y., Akriti Y. G., Su C., Lai M., Yang P., Kong J., Savoie B. M., Yu Y., Dou L., Two-dimensional halide perovskite lateral epitaxial heterostructures. Nature 580, 614–620 (2020). [DOI] [PubMed] [Google Scholar]
  • 17.Tao W., Zhang C., Zhou Q., Zhao Y., Zhu H., Momentarily trapped exciton polaron in two-dimensional lead halide perovskites. Nat. Commun. 12, 1400 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Driessen F., High-efficiency energy up-conversion at GaAs-GaInP2 interfaces. Appl. Phys. Lett. 67, 2813–2815 (1995). [Google Scholar]
  • 19.Ziegler J. D., Zipfel J., Meisinger B., Menahem M., Zhu X., Taniguchi T., Watanabe K., Yaffe O., Egger D. A., Chernikov A., Fast and anomalous exciton diffusion in two-dimensional hybrid perovskites. Nano Lett. 20, 6674–6681 (2020). [DOI] [PubMed] [Google Scholar]
  • 20.Neutzner S., Thouin F., Cortecchia D., Petrozza A., Silva C., Kandada A. R. S., Exciton-polaron spectral structures in two-dimensional hybrid lead-halide perovskites. Phys. Rev.Mater. 2, 064605 (2018). [Google Scholar]
  • 21.Urban J. M., Chehade G., Dyksik M., Menahem M., Surrente A., Trippé-Allard G., Maude D. K., Garrot D., Yaffe O., Deleporte E., Plochocka P., Baranowski M., Revealing excitonic phonon coupling in (PEA)2(MA)n−1 PbnI3n+1 2D layered perovskites. J. Phys. Chem. Lett. 11, 5830–5835 (2020). [DOI] [PubMed] [Google Scholar]
  • 22.Kang J., Wang L.-W., Dynamic disorder and potential fluctuation in two-dimensional perovskite. J. Phys. Chem. Lett. 8, 3875–3880 (2017). [DOI] [PubMed] [Google Scholar]
  • 23.Thouin F., Valverde-Chávez D. A., Quarti C., Cortecchia D., Bargigia I., Beljonne D., Petrozza A., Silva C., Srimath Kandada A. R., Phonon coherences reveal the polaronic character of excitons in two-dimensional lead halide perovskites. Nat. Mater. 18, 349–356 (2019). [DOI] [PubMed] [Google Scholar]
  • 24.Wu B., Ning W., Xu Q., Manjappa M., Feng M., Ye S., Fu J., Lie S., Yin T., Wang F., Strong self-trapping by deformation potential limits photovoltaic performance in bismuth double perovskite. Sci. Adv. 7, eabd3160 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Giovanni D., Chong W. K., Dewi H. A., Thirumal K., Neogi I., Ramesh R., Mhaisalkar S., Mathews N., Sum T. C., Tunable room-temperature spin-selective optical Stark effect in solution-processed layered halide perovskites. Sci. Adv. 2, e1600477 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Sie E. J., McIver J. W., Lee Y.-H., Fu L., Kong J., Gedik N., Valley-selective optical Stark effect in monolayer WS2. Nat. Mater. 14, 290–294 (2015). [DOI] [PubMed] [Google Scholar]
  • 27.Wu X., Trinh M. T., Zhu X.-Y., Excitonic many-body interactions in two-dimensional lead iodide perovskite quantum wells. J. Phys. Chem. C 119, 14714–14721 (2015). [Google Scholar]
  • 28.Batignani G., Fumero G., Srimath Kandada A. R., Cerullo G., Gandini M., Ferrante C., Petrozza A., Scopigno T., Probing femtosecond lattice displacement upon photo-carrier generation in lead halide perovskite. Nat. Commun. 9, 1971 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Quan L. N., Park Y., Guo P., Gao M., Jin J., Huang J., Copper J. K., Schwartzberg A., Schaller R., Limmer D. T., Vibrational relaxation dynamics in layered perovskite quantum wells. Proc. Natl. Acad. Sci. U.S.A. 118, e2104425118 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Debnath T., Sarker D., Huang H., Han Z.-K., Dey A., Polavarapu L., Levchenko S. V., Feldmann J., Coherent vibrational dynamics reveals lattice anharmonicity in organic–inorganic halide perovskite nanocrystals. Nat. Commun. 12, 2629 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Fu J., Li M., Solanki A., Xu Q., Lekina Y., Ramesh S., Shen Z. X., Sum T. C., Electronic states modulation by coherent optical phonons in 2D halide perovskites. Adv. Mater. 33, 2006233 (2021). [DOI] [PubMed] [Google Scholar]
  • 32.Lanigan-Atkins T., He X., Krogstad M., Pajerowski D., Abernathy D., Xu G. N., Xu Z., Chung D.-Y., Kanatzidis M., Rosenkranz S., Two-dimensional overdamped fluctuations of the soft perovskite lattice in CsPbBr3. Nat. Mater. 20, 977–983 (2021). [DOI] [PubMed] [Google Scholar]
  • 33.Li W., Vasenko A. S., Tang J., Prezhdo O. V., Anharmonicity extends carrier lifetimes in lead halide perovskites at elevated temperatures. J. Phys. Chem. Lett. 10, 6219–6226 (2019). [DOI] [PubMed] [Google Scholar]
  • 34.Chen T., Foley B. J., Ipek B., Tyagi M., Copley J. R., Brown C. M., Choi J. J., Lee S.-H., Rotational dynamics of organic cations in the CH3 NH3PbI3 perovskite. Phys. Chem. Chem. Phys. 17, 31278–31286 (2015). [DOI] [PubMed] [Google Scholar]
  • 35.Bakulin A. A., Selig O., Bakker H. J., Rezus Y. L., Müller C., Glaser T., Lovrincic R., Sun Z., Chen Z., Walsh A., Frost J. M., Jansen L. C., Real-time observation of organic cation reorientation in methylammonium lead iodide perovskites. J. Phys. Chem. Lett. 6, 3663–3669 (2015). [DOI] [PubMed] [Google Scholar]
  • 36.Seitz M., Magdaleno A. J., Alcázar-Cano N., Meléndez M., Lubbers T. J., Walraven S. W., Pakdel S., Prada E., Delgado-Buscalioni R., Prins F., Exciton diffusion in two-dimensional metal-halide perovskites. Nat. Commun. 11, 2035 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Zhang S.-F., Chen X.-K., Ren A.-M., Li H., Bredas J.-L., Impact of organic spacers on the carrier dynamics in 2D hybrid lead-halide perovskites. ACS Energy Lett. 4, 17–25 (2019). [Google Scholar]
  • 38.Spirito D., Asensio Y., Hueso L. E., Martín-García B., Raman spectroscopy in layered hybrid organic-inorganic metal halide perovskites. J. Phys. Mater. 5, 034004 (2022). [Google Scholar]
  • 39.Tao W., Zhou Q., Zhu H., Dynamic polaronic screening for anomalous exciton spin relaxation in two-dimensional lead halide perovskites. Sci. Adv. 6, eabb7132 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Tao W., Zhang Y., Zhu H., Dynamic exciton polaron in two-dimensional lead halide perovskites and implications for optoelectronic applications. Acc. Chem. Res. 55, 345–353 (2022). [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figs. S1 to S9

Table S1

Click here for additional data file. (869.6KB, pdf)

Articles from Science Advances are provided here courtesy of American Association for the Advancement of Science

RESOURCES