Charge-carrier transport in halide perovskites is often interpreted using quantities derived from an average crystal structure. Yet their atomic lattices are continuously fluctuating in a way that cannot be straightforwardly captured as a perturbation of the average crystal structures. This prompts many intriguing questions of how these emerging structural fluctuations modify the electronic states through which carriers move. I will present recent work in which we develop a new method that combines molecular dynamics with electronic-structure calculations and a Kubo treatment of the optical conductivity. Our approach makes it possible to follow the current response across a wide range of physical regimes in materials. At the same time, using machine-learnign techniques it retains first-principles accuracy while transgressing the limiting boundaries of traditional models of electronic transport and first-principles theories used therein. Using our method, I will show intriguing spectral signatures of halide perovskites and discuss what these features imply for the interpretation of mobility and the limits of conventional band-transport models.

Metal halide perovskites exhibit a remarkable combination of defect tolerance, facile processing, and outstanding optoelectronic performance. These same attributes, however, may also underlie many of the phenomena that limit device stability. While unusual behavior in perovskite devices is often attributed broadly to ion migration, a growing body of evidence points to a more complex picture involving coupled ionic, electronic, and chemical processes.

In this talk, I will discuss halide phase separation as a model system for understanding the reactive nature of metal halide perovskites. I will describe how photoelectrochemical processes and interfacial reactions can drive compositional evolution in mixed-halide materials, and why controlling these processes is essential for realizing stable wide-bandgap perovskites for perovskite-silicon tandem solar cells.

I will highlight recent studies of interfacial redox chemistry, halide transport, and degradation pathways, and discuss how concepts from electrochemistry can complement conventional semiconductor device physics in understanding these materials. Finally, I will explore an open question for the field: whether the exceptional defect tolerance of halide perovskites is fundamentally linked to their propensity for photochemical reactivity. If so, understanding this relationship may prove central to the design of the next generation of stable perovskite optoelectronic devices.

Metal halide perovskites are highly dynamic materials in which structural fluctuations, defects, and electronic excitations are strongly coupled across a broad range of length and time scales. These dynamic processes govern key properties relevant to photovoltaics, light emission, and radiation detection, yet many of the underlying atomistic mechanisms remain poorly understood.

In this talk, I will present our recent efforts to understand the dynamic behavior of halide perovskites using atomistic simulations spanning multiple scales. Starting from lattice vibrations and structural phase transitions, I will discuss how local distortions and dynamic disorder influence material properties and give rise to emerging phenomena such as chiral phonons and temperature-dependent chirality. I will then address the role of charged defects, polarons, and ion migration, highlighting how their interactions with the dynamic lattice affect transport, stability, and performance.

Many of these processes occur on timescales that are inaccessible to conventional simulations. By combining machine-learning interatomic potentials with enhanced-sampling and rare-event techniques, we can access slow activated processes such as low-temperature ion migration, crystal growth, and degradation pathways while retaining near first-principles accuracy.

Together, these studies provide a unified picture of perovskite dynamics, linking ultrafast lattice motions, defect physics, and long-timescale structural evolution. The insights gained help explain the unique properties of halide perovskites and provide a foundation for the rational design of next-generation functional materials.

References: 

1. Tyagi, V.; Pols, M.; Brocks, G.; Tao, S. Tracing Ion Migration in Halide Perovskites with Machine-Learned Force Fields. J. Phys. Chem. Lett. 2025, 16, 5153–5159.

2. Tyagi, V.; Brocks, G.; Tao, S. A Unified Microscopic Picture of Cation and Anion Migration in MAPbI₃. arXiv:2605.02685 (2026).

3. Tyagi, V.; Pols, M.; Brocks, G.; Tao, S. Halide Diffusion in Mixed-Halide Perovskites and Heterojunctions. Chem. Mater. 2026, 38, 4703–4711.

4. Pols, M.; Brocks, G.; Calero, S.; Tao, S. Temperature-Dependent Chirality in Halide Perovskites. J. Phys. Chem. Lett. 2024, 15, 8057–8064.

5. Pols, M.; Brocks, G.; Calero, S.; Tao, S. Chiral Phonons in 2D Halide Perovskites. Nano Lett. 2025, 25, 10003–10009.

6. Wilke, K.; Tyagi, V.; van Erp, T. S.; Tao, S. Bridging Timescales in CsPbBr₃: Low-Temperature Ion Migration from Machine-Learning Potentials and Path Sampling. Manuscript in preparation.

Lead halide perovskite nanocrystals (LHP NCs) - the latest generation of colloidal quantum dots (QDs) - possess dynamic, entropically stabilized soft lattices and electronically benign surfaces that, remarkably, do not compromise their textbook semiconductor optical quality. They are intrinsically bright emitters without the need for epitaxial wide-bandgap shells. In recent years, LHP NCs have emerged as the most intensively studied QD material, challenging the field's foundational paradigms in nearly every respect. They are the first QDs to exhibit excitonic coherence on timescales comparable to their radiative lifetimes. Their giant oscillator strength effect enables extremely fast emission (lifetimes as short as 60 ps) even in relatively large NCs, while maintaining single-photon emission. The excitonic fine structure of LHP QDs can be readily engineered through shape anisotropy. Periodic ensembles of LHP NCs have further demonstrated collective, accelerated radiative decay - superfluorescence - a phenomenon previously unseen in colloidal systems. This property unlocks future prospects of devising deterministic N-photon sources, through creation of N-QDs aggregate states. Our latest work focuses on expanding the operation of perovskite QDs to the red and near-infrared ranges, available with fully Tin-based and Lead-Tin-based compositions. The key enabler is to attain and maintain sufficiently low self-doping levels (originating from Sn vacancies and oxidation), in particular, through de-doping strategies and core-shell morphologies. The presentation will encompass our latest works along these research lines.

Recent proliferation of data-intensive applications has posed challenges for existing wireless connectivity [1]. Optical wireless communication (OWC) integrates data communication with ambient lighting, emerging for the next-generation high-speed free-space data communication [2]. OWC transmits data by modulating light intensity, with data transmission capacity dependent on the switching speed of the emitter. In addition to making faster emitters, the capacity can be enhanced by sending data simultaneously across multiple independent channels at different emission wavelengths.

Currently, light sources employed in OWC use conventional inorganic semiconductors that exhibit high carrier mobilities and fast switching speeds. However, such materials are expensive to produce and the emission wavelength is hardly tuneable. Perovskite nanocrystals (PNCs), hold great promise for advancing OWC: they can be processed from solution, they are bright, tuneable, and narrow emitters with short luminescence lifetimes, which meet the requirements for optical communication [3].

In this work, we study compositionally engineered PNCs as colour converters for multichannel OWC. Drop-cast CsPbBr₃, CsPbI₃, and FAPbI₃ nanocrystal films are used to generate spectrally separated emission channels from visible to near-infrared spectrum. We characterise important figures-of-merit using optical spectroscopy, including photoluminescence spectra (PL), PL quantum yield and PL lifetime, and correlate these properties with OWC performance. Using a prototype free-space link, we evaluate modulation bandwidth and data transmission performance over a 50 cm optical path. Under single-channel transmission of independently seeded pseudorandom bit sequences, the three colour-converted channels support gigabit-class pulse-amplitude-modulated data links, achieving data rates of 1.70, 1.60, and 1.76 Gbps for the CsPbBr₃, CsPbI₃, and FAPbI₃ channels, respectively. This corresponds to an aggregate data rate of 5.06 Gbps in single-channel measurements and 4.38 Gbps under wavelength-multiplexed operation. These results demonstrate the potential of compositionally tuneable perovskite nanocrystals as scalable, spectrally addressable colour converters for visible to near-infrared multichannel OWC.

The development of room-temperature quantum light sources remains a central challenge in quantum technologies. Metal halide perovskite quantum dots (QDs) are promising candidates due to their non-classical emission, including photon anti-bunching and bunching. However, photon bunching — a key signature of collective quantum states — has so far been demonstrated in perovskites only at cryogenic temperatures. In this presentation I will discuss collective blinking and photon bunching at room temperature from individual sub-wavelength CsPbBr3 QD superlattices.

Photoluminescence microscopy measurements showed that over 95% of superlattices exhibit collective two-level blinking, switching between a strongly emitting ON-state and a weak grey state, with ON-state intensity exceeding a single QD by more than two orders of magnitude — indicating synchronized emission from many QDs. Photon coincidence measurements revealed bunching in the ON-state with a degree up to 3.9 across 49 superlattices, in clear contrast to the anti-bunching of single QDs and the absence of bunching in QD ensembles.

Super-resolution imaging localized the ON-state emission to a 20–30 nm region, far smaller than the superlattice itself, pointing to long-range exciton migration toward a localized energy trap. Extended ON-state PL lifetimes (28–69 ns) further support multi-step Förster transfer prior to emission. Power spectral density analysis fits a stretched Lorentzian function, consistent with a well-defined photoactivatable quencher at the confined emission site driving the collective blinking.

Photon bunching is attributed to biexciton–exciton cascade emission at this localized site, where exciton funneling raises local exciton density enough to favor biexciton formation. This is supported by the decrease in bunching degree with increasing excitation power, a hallmark of cascade emission distinct from superfluorescence or superradiance. Fitting the central g2(τ) peak yields a fast ~1.3 ns component attributed to the biexciton lifetime, consistent with the known fast biexciton-to-exciton radiative rate in CsPbBr3. A red-shifted PL shoulder (binding energy 10–29 meV), absent in isolated QDs, further supports biexciton involvement.

These findings establish sub-wavelength CsPbBr3 QD superlattices as a practical platform for room-temperature collective optical phenomena, with promising implications for entangled photon sources and quantum optoelectronic devices.

Understanding and the interconversion between spin and charge current under chirality induced spin selectivity (CISS) is critical in leverage chiral semiconductors for developing room-temperature control over spin, charge and light. CISS phenomena arise from an interplay among structural chirality, electron spin orientation, and charge current. Steady-state observations such as magnetoresistance (MR) offer little insight into the timescales that govern the spin-charge interconversion. In contrast, inverse CISS involves the conversion of spin to a charge current. Terahertz (THz) emission spectroscopy (TES) offers a non-contact probe of the induced charge current that maps the charge current direction in three dimensions. We show how the THz emission can map the spin-to-charge conversion and study THz emission in FM/chiral perovskite heterostructures. We contrast with room temperature magnetoresistance in similar films. We show that both CISS and ICISS exhibit similar symmetry, i.e., for ICISS the current direction reverses either for changing the chirality or magnetization direction. These observations directly demonstrate the inherent coupling between spin and charge currents in chiral systems. These observations suggest constraints on the CISS mechanism. For a chiral/FM interface there is spin accumulation at the interface: injection of either charge (CISS) or spin (ICISS) drives the spin configuration out of equilibrium and movement of spin (CISS) or charge (ICISS) must compensate, thus driving the spin-to-charge interconversion.

Lead-free halide perovskites have emerged as promising alternatives to conventional Pb-based materials, offering new opportunities to develop environmentally benign semiconductors while uncovering fundamentally different structure-property relationships. This presentation will highlight recent advances toward understanding and developing lead-free perovskites across both bulk and nanocrystalline material platforms.

The first portion of the talk will focus on two-dimensional germanium halide perovskites, where temperature-dependent optical spectroscopy, single-crystal structural characterization, and theoretical modeling reveal how lattice rigidity governs exciton-phonon interactions. Compared to analogous Pb-based materials, the Ge perovskites exhibit significantly weaker coupling to longitudinal optical phonons, suppressed bound-exciton formation, and the narrowest reported emission linewidths for a Ge-based halide perovskite, demonstrating that careful structural design can fundamentally alter excited-state dynamics.

Building on these insights, the presentation will discuss ongoing efforts to extend lead-free perovskite chemistry into colloidal nanocrystal systems through the synthesis of double perovskite nanomaterials. Preliminary studies highlight both the synthetic challenges and emerging opportunities associated with preparing compositionally complex lead-free nanocrystals with controlled optical properties. Together, these examples illustrate how fundamental understanding of lattice dynamics and continued advances in synthetic chemistry are expanding the landscape of lead-free halide perovskites for future optoelectronic applications.

For conventional semiconductors the band gap values are usually governed by lattice expansion and electron-phonon interactions, leading to a gradual redshift upon heating. [1] In hybrid perovskites, however, such behavior can be non-monotonic due to stronger structural distortions and dynamic disorder of organic compounds. As a result the temperature dependence might be unconventional, resulting in phenomena’s such as increasing band gap values to a certain temperature as reported for 2-thiophenemethylammonium lead iodides.[2] Importantly, the evolution of the band gap is closely linked to the emission behavior, as temperature-induced modifications of the electronic structure affect exciton localization and recombination pathways. These changes not only alter the electronic structure but also affect carrier localization and radiative recombination, thereby influencing the emission properties. [3-4] Consequently, materials exhibiting unconventional temperature dependence of the band gap may display markedly different photoluminescence characteristics in comparison to classical semiconductors. In this presentation we present unusual evolution of the band hap in several halide perovskites combined with crystallography and optical studies to describe effects of such phenomena on the perovskite properties.

Lead halide perovskites have emerged as highly promising materials for solar energy conversion, with single-junction power conversion efficiencies exceeding 26%. Their soft crystal lattice combines high defect tolerance with mechanical flexibility, but also enables halide ion migration, particularly under illumination. This phenomenon, known as halide segregation, is generally considered detrimental due to its impact on charge-carrier mobility and open-circuit voltage. However, controlled halide redistribution may also offer opportunities for engineering local band structure. Such control over the local band gap is particularly relevant at device interfaces, where band alignment and interfacial recombination velocities govern charge extraction in perovskite solar cells. Harnessing these light-induced changes could enable dynamic tuning of interfacial band offsets, potentially reducing the need for static interlayers.

Here, we show that halide segregation at low iodide concentrations gives rise to dynamic emission peaks in MAPbBr₃ thin films doped with ~1% iodide. Under continuous illumination, photoluminescence (PL) spectra exhibit a stable emission peak around 540 nm, accompanied by multiple dynamic peaks spanning the range between pure bromide and pure iodide emission. These transient features emerge and disappear on sub-second timescales and can shift by more than 100 nm within seconds. They are observed across different cation compositions and doping levels. This behavior contrasts with halide segregation at higher iodide fractions, which is typically characterized by a gradual redshift toward a stable emission peak.

Temperature-dependent PL measurements (300 K to 3 K) reveal a slowing of these dynamics at lower temperatures, with dynamic behavior still visible down to 192 K. This temperature dependence supports a mechanism based on light-induced ion migration, leading to the fast formation and dispersion of iodide-rich regions with lower band gap. If this dynamic halide redistribution could be spatially controlled, these fluctuations could be localized at perovskite device interfaces. This would enable in-situ tuning of band offsets and recombination velocities under operating conditions, rather than relying on static interfacial layers.

These results demonstrate a previously underexplored regime of halide segregation in which low dopant concentrations lead to dynamically evolving emission peaks. This provides a pathway toward tuning interfacial band offsets and recombination velocities, which could enable improved charge extraction and reduced interfacial losses in perovskite solar cells.

The optoelectronic and excitonic properties of halide perovskites are intimately linked to their structural flexibility, from static disorder and structural heterogeneity to temperature-driven dynamic disorder. In this talk, I will discuss our work using many-body perturbation theory (GW+BSE) to unravel how these structural degrees of freedom govern excited-state phenomena across the halide perovskite family, from layered two-dimensional systems to three-dimensional bulk compounds. We show how static disorder, structural heterogeneity and exciton-phonon coupling affect the fine structure of excitons, and how dynamic disorder renormalizes band gaps and exciton binding energies with temperature, captured by combining molecular dynamics with GW+BSE calculations on thermally sampled configurations. Throughout, I connect our calculations directly to experimental observations and discuss the methodological challenges of capturing these effects from first principles. Finally, I will present preliminary results on a new methodological direction: excited-state forces, a key ingredient for accessing excited-state structural relaxation and dynamics directly from many-body perturbation theory. Together, these results illustrate that static and dynamic disorder play a decisive role in the excited-state landscape of halide perovskites, with direct consequences for their use in optoelectronic devices.

Electron-phonon coupling plays a central role in determining the optoelectronic response of two-dimensional layered perovskites, yet the specific lattice degrees of freedom that mediate this coupling remain difficult to identify. This challenge arises from the structural complexity of hybrid perovskites, where mechanically coupled organic and inorganic sublattices form large, anharmonic unit cells with many closely spaced vibrational modes. As a result, low-frequency Raman modes are often assigned to inorganic framework motion, while the role of the organic spacer layer is treated as secondary.

Here we combine vibrational spectroscopy, lattice engineering and first-principles analysis to resolve the hybrid lattice dynamics of layered perovskites. We show that the normal modes are intrinsically organic-inorganic in character and cannot be separated into purely molecular or framework vibrations. This hybridization persists even below 50 cm^-1, a spectral region conventionally associated with octahedral tilts and inorganic lattice distortions. Chemical substitution of the organic spacer selectively modifies these low-frequency modes, demonstrating that the organic sublattice provides an active handle for tuning the vibrational landscape.

We then examine the lattice response under above-bandgap excitation and identify spectroscopic signatures of polaronic distortion [1]. The excited-state spectra reveal that polaron formation is governed by hybrid vibrational coordinates involving both the inorganic framework and the organic spacer layer. In particular, the polaronic response is consistent with a vibronic mechanism in which a Franck-Condon progression of a Jahn-Teller-like framework distortion couples to a hybrid soft mode through Herzberg-Teller interactions. These results show that the organic sublattice is not merely a passive structural spacer but actively participates in stabilizing the photoinduced polaronic state.

By linking lattice hybridization, chemical control and excited-state vibrational response, this work establishes hybrid phonon engineering as a route to tune polaron formation in two-dimensional perovskites. More broadly, it highlights the need to resolve lattice dynamics under operating conditions in order to understand and design the photophysical properties of hybrid optoelectronic materials.

Halide perovskite nanomaterials provide a powerful platform for studying how synthesis conditions control nucleation, growth, compositional evolution, dopant incorporation, dimensionality, and light–matter interactions. However, these materials often form through fast, coupled, and highly condition-dependent pathways, making it difficult to resolve how precursor chemistry, reaction environment, and processing history collectively determine their optical properties. In this talk, I will discuss how self-driving laboratories can accelerate the experimental interrogation of these complex reaction spaces by integrating automated synthesis, in situ/online optical characterization, and machine-learning-guided experiment selection. I will highlight recent work from our group on flow- and batch-based autonomous platforms for metal halide perovskite quantum dots and low-dimensional perovskite nanostructures, including autonomous optimization of emission properties, data-rich mapping of dopant-mediated photoluminescence, accelerated exploration of anion-exchange and photoinduced transformation pathways, and synthesis–property modeling across high-dimensional experimental spaces. These studies show how autonomous experimentation can move beyond empirical materials optimization to uncover non-intuitive reaction conditions, identify key synthetic variables governing optical response, and generate mechanistic insight into how halide perovskite nanomaterials form, transform, and emit.

Wide band gap semiconductors absorb visible light but waste sub-band gap infrared photons. Likewise, near infrared band gap materials (like Si) use more of the solar spectrum, but waste higher-energy photons and produce low voltages in solar cells. An intriguing solution for harvesting IR photons, without compromising the voltage, is offered by intermediate-band solar cells. Here, an empty intermediate band is placed within the band gap of the semiconductor absorber to promote two-step photon absorption--from the valence to the intermediate band and from the intermediate to the conduction band--in addition to the band gap transition. Thus intermediate-band solar cells can produce high current density while maintaining a high voltage. However, current intermediate-band solar-cell designs suffer from a low intermediate-band density. I will present our work exploring a new design for intermediate-band solar cells, inspired by halide perovskites, that may offer new routes for increasing the intermediate band density and overall efficiency.

Two-dimensional hybrid perovskites are a highly intriguing class of materials, composed of alternating inorganic and organic molecular layers. Their reduced dimensionality combined with weak dielectric screening leads to the formation of tightly bound excitons that efficiently absorb and emit radiation. A central questions for excitons in perovskites from the perspectives of both fundamental physics and applications is their mobility. In addition, the flexibility of the material design allows for the integration of a variety of functional compounds including chiral molecules to enable polarization control of the optical response. Most importantly, the recently demonstrated presence of the exciton fine structure and the predictions of the associated non-equilibrium pose major questions regarding the energy transport in 2D perovskites mediated by excitonic carriers. In this talk I will focus on the transport of optically detected excitons in 2D perovskites via transient, ultrafast microscopy, featuring different regimes of propagation featuring free and localized states. I will demonstrate the strong impact of the exciton fine structure leading to extremely rapid propagation of hot excitons in 2D perovskites, discuss the impact of temperature and the associated transition to the equilibrated regime. This opens up interesting opportunities to design the exciton band structures with the possibilities to create scenarios enabling hot exciton extraction.

The instability of halide perovskite optoelectronic devices remains a bottleneck for long-term operation, largely due to persistent defect formation and associated non-radiative (NR) losses. However, unlike conventional semiconductors, MHPs possess an inherent self-healing (SH) ability that allows spontaneous defect repair without external intervention[1–5], which presents a pathway toward resilient, low-maintenance devices; however, its practical realization hinges on deciphering and regulating the microscopic origins of SH. The underlying microscopic mechanisms remain poorly understood, particularly the role of interfacial chemistry on trap dynamics and healing kinetics. Here, I will discuss SH and defect evolution in triple-cation mixed-halide (TCMH) perovskite films and their device-relevant charge-transport-layer heterostructures subjected to photo-induced damage. Using correlation clustering imaging (CLIM) [6], our recently developed local functional imaging tool, I will show how we map spatiotemporal photoluminescence heterogeneity to track defect dynamics in pristine and hetero-structure films (Figure 1). Our results demonstrate that the chemical nature of charge-transport layers modulates trap activity, healing kinetics, and halide redistribution, with hetero-structures exhibiting faster recovery than pristine films, a boon for device resilience. These findings provide new insights into the dynamic interactions among defects, interfaces, and ion migration and establish a framework for the rational design of durable, next-generation perovskite optoelectronic devices"

In light-emitting diodes (LEDs), the charge injection rate is critical to device performance, but it is generally limited by resistance–capacitance (RC) time constants to nanosecond or longer timescales. While optical excitation can produce carriers on femtosecond timescales, it inherently generates excitons, complicating the study of individual charge carriers. Here, we present an infrared (IR) optical control strategy that enables ultrafast electrical charge injection in operational perovskite LEDs (PeLEDs). Under electrical bias, space charge naturally accumulates at device interfaces, and an ultrafast IR pulse can modulate this interfacial charge distribution. This accelerates the injection of charges into the emissive perovskite layer on femtosecond timescales—far faster than intrinsic electronic response times—and thereby enhances electroluminescence.

To probe this process, we combine electrical injection and optical excitation in an operando electrical pump–IR push–visible probe spectroscopy setup. Using this technique, we observe the rapid emergence of band-edge carriers in the perovskite layer immediately following the IR pulse. Time-resolved electroluminescence measurements confirm that these IR-injected carriers directly enhance radiative recombination in the PeLED. Modified device architectures with selective charge-blocking layers show that the injected carriers originate predominantly from the electron-transport layer interface, demonstrating polarity-selective control of electrical injection.

We demonstrate that in halide perovskite devices, charge injection is governed not only by energy-level alignment but also by material-specific interfacial properties. Unlike conventional electrical injection, which relies on thermal occupation of interfacial states, an infrared (IR) optical pulse can transiently enhance carrier injection by manipulating the space-charge population accumulated near an injection barrier. Absorption of IR photons promotes carriers into higher-energy or more strongly coupled interfacial states, enabling transfer into the perovskite emissive layer while partially bypassing limitations imposed by barrier height, width, and electronic coupling. Our results reveal that charge injection is strongly influenced by material-level interfacial factors, including space-charge distribution, interfacial dipoles, and electronic coupling, providing new strategies for optical control of carrier transport and the design of faster, more efficient perovskite optoelectronic and quantum devices.

Transparent Conducting Oxides (TCOs) sit at nearly every critical interface in perovskite–silicon tandem solar cells, governing optical coupling, charge extraction, and recombination, while introducing resistive, absorptive, and reflective losses that can each cost more than 2% absolute efficiency.

Beyond performance, TCO choice and quality can directly affect long-term device stability: microstructure and interfacial chemistry determine resistance to moisture ingress, ion migration, and damp-heat degradation, factors that are too often considered separately from efficiency optimization. In this talk, I will argue that TCOs deserve treatment as an important design parameter, alongside the carrier extraction and perovskite absorber itself, in the pursuit of tandem cells that are both highly efficient and operationally durable. I will discuss the materials physics trade-offs involved, current material limitations including reliance on scarce indium-based oxides, and emerging strategies (new dopants, multilayer architectures, and computational screening) aimed at simultaneously improving optoelectronic performance and long-term stability for sustainable deployment.