Abstract
Spin polarized excitons induced by spin injection from magnetic ion to a single quantum dot, has been considered as a basic unit of quantum information transfer between spin and photon for spin-photonic applications. However, this state-of-the-art technology has only been found with limited coupling strength and weak excitonic emission. Here, we demonstrate a spin-polarized self-trapped exciton naturally formed in the zero-dimensional lattice of cesium copper iodide. Upon excitation, the conversion from Cu+ ion to spin-1/2 Cu2+ ion results in an in-situ self-trapped exciton, which facilitates a local Jahn-Teller distortion and guarantees the strong spin-exciton coupling and near-unity excitonic emission efficiency. Consequently, a giant Zeeman splitting of −53 meV and an effective excitonic g-factor of −93.5 are observed from magneto-photoluminescence. More importantly, this nano-scale coupling can also be driven by an external electric field, which generates electroluminescence with a circular polarization of 44.5% at 4.2 K and 8% at 300 K. The spin-optic properties of this copper compound will stimulate the fabrication of next-generation spin-photonic devices based on self-trapped excitons.
Similar content being viewed by others
Introduction
The integration of electrical, magnetic, and optical properties in a semiconductor nanostructure is pivotal for the advancement of spintronics1,2,3. For making a spin-photonic device, spin-polarized exciton was revealed in the diluted magnetic semiconductors (DMS) to facilitate the coupling between optoelectronic and spin properties4,5,6,7. The spatial overlapping of wavefunction between the magnetic ions and intrinsic exciton is crucial for constructing these coupling systems. This often necessitates the use of low-dimensional nanostructures such as Mn2+ or Cu2+ doped II−VI and IV−VI quantum dots (QDs), where the exciton was confined and the sp-d interaction was enhanced8,9,10. However, as shown in Fig. 1a, Wannier exciton exhibits a large Bohr radius with broad wavefunction (~5 nm)11,12,13, significantly exceeding the spatial extent of the electronic wave function of magnetic ions (~0.5 nm), resulting in limited coupling strength in DMS systems. Additionally, doping magnetic ions into semiconductors can lead to energy transfer from excitons to dopants14,15, drastically reducing excitonic luminescence.
a In conventional DMS QD, the overlapping of Wannier exciton wavefunction with that of magnetic dopant is small while for b STE (a unique Frenkel exciton), this overlapping is large. c The structural transition from [Cu2I5]3− (left) to [Cu2I5]2− (right) upon UV light excitation. The dumbbells represent the atomic orbital on the Cu and I sites; upon excitation, the I:5pz orbital and Cu:3dxy orbital overlap (blue dumbbells) and drive the J-T distortion. d Schematic diagram of energy levels for ground and excited states of Cs3Cu2I5 in tetrahedral crystal fields before and after UV excitation. e ESR spectra of Cs3Cu2I5 without (dark) and with UV-light illumination (Light, T = 300 K, 200 K, 120 K, 77 K), the gray curves show the simulation results considering a tetragonal J-T distortion.
Unlike the weakly bound Wannier excitons, Frenkel excitons show tightly-binding excitonic wavefunction in atomic scale16, enabling the effective coupling of magnetic ion and exciton. Self-trapped exciton (STE) is a unique Frenkel exciton with natural low-dimensional structurally confined holes17, as shown in Fig. 1b, it is typically formed in soft lattice materials with strong electron-phonon coupling which induces local lattice distortions, and highly confines the exciton within nano-meter range18,19. In the existence of a local magnetic ion, the atomic-scale spin-exciton coupling is expected to be greatly enhanced due to the low dimension of this exciton20. Furthermore, if this magnetic center itself is part of the STE, the nonradiative energy transfer could be avoided and the excitonic emission would be enhanced. This intrinsic low dimensional exciton structure by the confinement effect of STE enhances the interaction between spin and exciton without the nanostructure fabrication, thus broadening their applicability. Therefore, spin-polarized STE deserves further investigation and in-depth research. Recently, a copper-based halide compound Cs3Cu2I5 was found with an electronically zero-dimensional (0D) structure with typical STE emission21,22,23, and can be a promising candidate for spin-photonics applications.
In this work, a strong local spin-STE interaction is revealed in Cs3Cu2I5 from the observation of a giant Zeeman splitting and large circular polarization of luminescence in the presence of a magnetic field. As illustrated in Supplementary Fig. S1, it is found that the 0D structural component [Cu2I5]3− transform to [Cu2I5]2− upon excitation, charge conversion occurs, and the spinless Cu+ ions turn into spin-1/2 Cu2+ ions. The resultant structure is comparable to that of a DMS, but without energy transfer or surface effects. Leveraging the in-situ spin-STE interaction and the near-unity quantum yield of photoluminescence (PL), a prototype spin-photonic device has been fabricated, exhibiting the desired performance for circularly polarized light emission.
Results and discussion
In light of this, Cs3Cu2I5 thin films are prepared using single-source vacuum thermal evaporation. The homogeneous films prepared on quartz substrates exhibit orthorhombic structure at the growth temperature of 373 K (Supplementary Figs. S2–S4). PL peak at 445 nm with broad peak width (80 nm) originates from STE from the 0D [Cu2I5]2− units (Supplementary Figs. S5–S7), exhibiting a PLQY of 95% (Supplementary Fig. S8). Cs3Cu2I5 show large exciton activation energy and electron-phonon energy (Supplementary Fig. S9), with the PL lifetime of hundreds of nanoseconds (Supplementary Fig. S10). As shown in Fig. 1c, d, the charge conversion from [Cu+]−3d10 to [Cu2+]−3d 9 in tetrahedral coordination is stabilized by the Jahn-Teller (J-T) effect21,24,25, which splits the valence states of Cu2+ with an unpaired electron after excitation. This photo-assisted J-T effect instigates a conversion of the originally common line of the CuI3 triangle and CuI4 tetrahedron into two coplanar CuI4 units. The transformation not only reduces the optical bandgap, but also induces an overlap between the Cu:3dxy orbitals and the I:5pz orbitals, providing strong coupling between localized Cu2+ spin and the STE.
The above charge conversion and J-T effect can be evidenced by photon-excited electron spin resonance (ESR) spectroscopy (Supplementary Fig. S11). As shown in Fig. 1e, without illumination, there is no signal except a sharp transition at g = 2.003 (B = 350 mT), which is attributed to shallow halide vacancy defects formed during the sample growth process. In contrast, with the illumination of light at the wavelength of 280 nm, strong asymmetric ESR signals appear in the B range of 260−380 mT, revealing spin signals with strong magnetic anisotropy, which would be expected from a spin-1/2 electron at the Cu2+ site under J-T distortion26,27. At cooling down from 300 K to 77 K, the thermal broadening of the ESR signal reduced, revealing a clear structure corresponding to electron-nuclear hyperfine interaction between electron spin S = 1/2 and nuclear spin I = 3/2 (corresponding with the Cu nuclear spin of the two natural abundance isotopes)28. From the fitting using an effective spin Hamiltonian analysis, the deduced anisotropic electron g-factors are gx ≈ gy = 2.07(±0.01), gz = 2.36(±0.01). The anisotropic hyperfine parameters are Ax ≈ Ay = 50(±20) MHz, Az = 490(±10) MHz. The symmetry evaluation of the Cu2+ ESR signal suggests the symmetry of C3V or lower. A more detailed discussion can be found in Supplementary Note 1. The same parameters can be used to fit the ESR data at 300 K, by assuming the thermal broadening effect. We also note that under 450 nm illumination (below the absorption edge), no ESR signal was detected expect for the sharp resonance peaks attributed to intrinsic defects (Supplementary Fig. S12), which verifies the association of Cu2+ to the above band edge absorption and the generation of STEs. In the X-ray photoelectron spectroscope (XPS) pattern, the Cu-2p3/2 peak shows an energy shift towards the high energy side (Supplementary Figs. S13, S14), which indicates the electron loss upon UV light illumination. Additionally, the increased magnetic susceptibility under illumination also indicates the presence of magnetism in the excited state (Supplementary Fig. S15). These experimental results provide clear evidence of Cu+ to Cu2+ transition triggered by light illumination and the occurrence of tetrahedral-type J-T distortion in the local photo-excited [Cu2I5]2− structure components, which support previous theoretical conjectures21.
The existence of photoinduced localized Cu2+ in a locally J-T distortion lattice would result in the spin-exciton interaction, which then produces enhanced Zeeman energy splitting−energy splitting between the left-hand (σ+) and right-hand (σ−) circularly polarized PL peak energies denoted by ΔE = E+ − E−. This ΔE is the energy splitting of STE, which is different from the former Zeeman splitting of the photo-excited Cu2+ ions. As shown in Fig. 2a, the σ+- and σ−-polarized PL show opposite tendencies with increasing magnetic field in both intensity and peak energy, and both tendencies reverse when the field changes its polarities. The field-dependent E+, E−, and ΔE are shown in Fig. 2b, ΔE increases with B and saturates at |B | ≥ 15 T, with a saturation value of −53 meV. In the linear region at low fields (|B | ≤ 10 T), an effective exciton g-factor (geff) of −93.5 could be obtained. This giant Zeeman splitting for Cs3Cu2I5 is much larger than that of Pb-based perovskites and comparable with that of some DMS nanostructures8,29,30,31. Further, ΔE and geff for Cs3Cu2I5 are negative, which is similar to the giant Zeeman splitting of Mn- or Cu-doped CdSe quantum dots (Table 1), where sp-d exchange interaction occurs with energy level reversing and further splitting9,32. Therefore, we adopted the sp-d exchange model for DMS to interpret the result for Cs3Cu2I5, and the energy splitting is demonstrated in the inset of Fig. 2b. The giant Zeeman splitting in Cs3Cu2I5 should be due to the strong coupling between STE and localized Cu2+ ion through the 3dxy-5pz orbitals, where the Cu2+ itself is included for forming the band-edge excitonic states, rather than acting as a magnetic dopant. The ΔE can be fitted by the Brillouin function8,9:
The first term represents the linear Zeeman splitting due to the intrinsic Zeeman effect of excitons, whereas the second term accounts for the additional splitting arising from the s-pd exchange interaction. In this second term, fe and fh represent the exchange coupling strength between electrons and Cu²⁺, as well as holes and Cu²⁺, respectively, while N0α and N0β denote the s-d, p-d exchange interaction constants. The quantity \(\langle {S}_{Z}\rangle\) is characterized by the Brillouin function which describes the magnetization of paramagnetic Cu2+ moments, Further details are provided in Methods. An excellent fit using Eq. (1) was achieved with \({N}_{0}({f}_{e}\alpha -{f}_{h}\beta )\) = 0.5 eV. Additionally, the energy splitting decreases at rising temperatures according to the Boltzmann law, indicating the paramagnetic behavior of the excitonic spins33, as shown in Fig. 2c, d. At T = 30 K, ΔE = − 15 meV fitted by Eq. (1) with the sp-d exchange energy of 16 meV and geff = −8.27. At 60 K, ΔE = − 10 meV and geff = −4.47 (Supplementary Fig. S16).
a 3D pseudo-color plot of left- (σ+) and right-circularly (σ−) polarized PL spectra in magnetic fields from −16 T to 16 T at T = 4.2 K. b Peak energy of the σ+ and σ− polarized PL and energy splitting (ΔE) versus B. The inset illustrates the excitons energy levels, where the magneto-PL arises from the Zeeman splitting of Jz = ±1 states. The reversal of the +1 and −1 states, due to the highly localized Cu2+ ions, leads to the giant Zeeman splitting. c σ+-and σ−-polarized PL at different T. d ΔE versus T. e Comparison of exciton PLQY and effective exciton g-factor per spin in typical DMS materials (Co2+:CdSe QDs34, Mn2+:ZnSe-CdSe QDs8, Cu2+:ZnSe-CdSe QDs9, Mn2+:CdSe QDs35) and Cs3Cu2I5.
This observation of giant Zeeman splitting and geff in Cs3Cu2I5 is similar to that in DMS. However, while the excitonic emission is suppressed in most DMS8,9, in Cs3Cu2I5, no nonradiative exciton-dopant energy transfer occurs. The superior performance of Cs3Cu2I5 compared to typical DMS34,35, demonstrated by the near-unity PLQY with remarkable sp-d exchange interaction facilitate its application in spin-photonic devices (Fig. 2e). It is also found that the spin polarization of Cs3Cu2I5 can be optimized by growth temperature, through which the grain size and crystallinity can be modulated. This variation alters the electron-phonon coupling strength, which in turn influences the formation of self-trapped excitons (STEs), and ultimately affects the degree of spin polarization (Supplementary Fig. S17—21). Accordingly, the Cs3Cu2I5 thin film grown at 393 K is chosen for making a prototype of spin light-emitting diodes. A heterostructure of ITO/ZnO/Al2O3/Cs3Cu2I5/MCP/TAPC/HAT-CN/Al is fabricated36, utilizing the Cs3Cu2I5 thin films as the active region (Supplementary Fig. S22). The working principle of spin-polarized electroluminescence (EL) is illustrated in Fig. 3a, where holes and electrons are injected and the two join in the Cs3Cu2I5 layer to form STEs. According to the optical selection rules and conservation of angular momentum in the optical transitions37, the excitonic recombination would produce circularly polarized luminescence, which can be attributed to the imbalanced population of electrons on the two pairs of Zeeman energy levels.
a Schematic diagram of the transport and recombination of carriers with electrical excitation under the magnetic field. b Magnetic field-dependent σ+- and σ−- polarized EL spectra at T = 4.2 K with an applied bias voltage of 16 V. c Energy splitting of EL as a function of B. d The σ+- and σ−- polarized EL spectra at T = 4.2 K, 80 K, 180 K, and 300 K. ePEL versus B, with the solid curves fitted.
The magneto-EL shows similar behavior as that of magneto-PL. In Fig. 3b, the EL spectra show large circular polarization and Zeeman splitting at 4.2 K. The saturated ΔE is −20.4 meV at B ≥ 15 T with the geff value of −27.2 (Fig. 3c). These values are in a similar magnitude to that in magneto-PL, and can be also attributed to the J-T distortion and accompanied sp-d exchange interaction between the electric-field-driven Cu2+ ions and STE. The circular polarization of EL, PEL = [(I+ − I−) / (I+ + I−)], where I+ and I− represent the intensities of the σ+- and σ−-polarized photon emission, can be obtained from the EL spectra in Fig. 3d, and shown in Fig. 3e. At T = 4.2 K, the saturated PEL value reaches 44.5%, slightly larger than that of PPL (31%) (Supplementary Fig. S23). It is speculated that the reorganization of photon-generated excited states leads to spin dissipation21, while the electron-hole pair directly recombines in the electric-field-driven excited state and show less spin dissipation. It is known that the excitonic spin polarization would be easily diminished by thermal fluctuation at rising T for most semiconductor29,38,39,40, however, even at room temperature, as demonstrated in Fig. 3d, e, the significant spin polarization of EL can still be observed. The excitonic g values were estimated using Brillouin function \(P={{{{\boldsymbol{P}}}}}_{0}\,\tanh [{g}_{eff}{\mu }_{B}B/(2{k}_{B}T)]\), where P0 is the saturation polarization. geff values are 2.97, 3.11, and 3.06 at T = 80 K, 180 K, and 300 K, respectively, which are generally consistent with that of magneto-PL result shown in the Supplementary Fig. S23. Supplementary Fig. S24 shows the similar T-dependence of P and gex for both PL and EL at B = 15 T.
It is noteworthy that current-voltage curves of the heterostructure can be effectively modulated by the magnetic field, with an opposite tendency when B changes polarity at T = 180 K and 300 K. This indicates a significant magnetic field effect on the electric signals, driven by Zeeman effect-induced spin polarization in Cs3Cu2I5, which leads to spin-selective transport through variations in scattering probabilities and band alignment (Supplementary Fig. S25). The above magneto-EL results reveal a new type of room temperature spin-photonic device based on the strong coupling between charge conversion-induced STE and magnetic ion at the same lattice site. It is important to note that the effective spatial confinement of STE around the Cu²⁺ ion plays a crucial role in modulating the sp-d interactions. Given this, it is reasonable to raise the question of whether this confinement can be adjusted by the dimension of the local lattice. In particular, variations in the local lattice dimension could potentially alter the spatial overlap between the 3 d orbitals of Cu²⁺ and the 5p orbitals of the surrounding halide ions, thereby influencing the strength of the sp-d interactions and the overall spin polarization.
The extent of exciton confinement is comparatively studied in the 0D lattice of Cs3Cu₂I5 and the 1D lattice of CsCu2I3. As depicted in Fig. 4a, unlike the isolated Cu atoms in Cs3Cu2I5, they are arranged along the crystal c-axis in CsCu2I3 (ref. 41). This higher dimension arrangement is expected to weaken the confinement of STE, thereby reducing the sp-d exchange interaction and spin polarization. We successfully synthesized 1D CsCu2I3 using the antisolvent method and vacuum thermal evaporation (see Methods and Supplementary Figs. S26–S28). In Fig. 4b, c, we present the T-dependent PL spectra in the range of 80−300 K for the two thin films (see Supplementary Fig. S29 for PL spectra). The full width at half maximum (FWHM) of PL varies with T in Fig. 4d, and fits using\(FWHM=2.36\sqrt{S}{E}_{p}\sqrt{coth[{E}_{p}/(2{k}_{B}T)]}\) results in S = 95.5 and 23.4, for Cs3Cu2I5 and CsCu2I3, respectively, which indicates a more pronounced lattice distortion occurs in Cs3Cu2I5 (ref. 42). The PPL values of CsCu2I3 at different T are obtained by detecting circular polarized PL (Supplementary Fig. S30), and the exciton g-factors are derived from the linear fit of PPL versus B in Fig. 4e. A comparative analysis of S, gex, and PPL and for the distinct dimensional-confined samples is displayed in Fig. 4f, in which the 0D Cs3Cu2I5 show superior spin polarization over the 1D CsCu2I3 due to a more pronounced quantum confinement effect and J-T effect.
a The lattice arrangement of Cs3Cu2I5 shows a 0D confinement structure of the Cu-I structure components, and that of CsCu2I3 exhibits a 1D structure. J-T distortion induced STE PL spectra at T = 80-300 K for b Cs3Cu2I5 and c CsCu2I3, respectively. d The fit of T-dependent FWHM of Cs3Cu2I5 and CsCu2I3. ePPL versus B at different T for CsCu2I3. f Comparison of S, gex, and maximal PPL of Cs3Cu2I5 and CsCu2I3, respectively (PPL and gex are obtained at T = 80 K).
In summary, we report the first observation of giant Zeeman splitting in an undoped compound, Cs3Cu2I5, with high excitonic photo-emission efficiency. In this zero-dimensional copper lattice, the Jahn-Teller distortion and charge conversion generate inherent local spin angular momentum which is conserved and transferred through the strong sp-d exchange interaction from the optically or electrically Cu2+ ions, to the self-trapped excitons in nano-scale range. Taking advantage of the spin-polarized self-trapped excitons, magnetic field-induced circular polarization of excitonic luminescence is achieved, and a prototype of a spin-photonic device is fabricated. Importantly, this series of compounds with condensed spin-1/2 copper exhibits significantly stronger spin-carrier interactions, brighter luminescence and greater flexibility in spin-electronic manipulation compared with DMS quantum dots. This study opens a new path for seeking optimal nano-materials for spin-photonic applications using self-trapped excitons.
Methods
Preparation of Cs3Cu2I5 powder
Cs3Cu2I5 powder is synthesized by the anti-solvent method. 7.7943 g CsI (0.03 mol), 3.809 g CuI (0.02 mol), and 20 mL dimethyl sulfoxide (DMSO) are loaded into a centrifugal tube successively. The precursor in the centrifuge tube is shaken for 10 min for complete dissolution. The antisolvent dichloromethane (DCM, 60 ml) is added to the precursor solution. Then the precursor is centrifuged at 8170 g to obtain the Cs3Cu2I5 powder precipitation. Finally, the Cs3Cu2I5 powder is dried and ball-milled for thermal evaporation.
Preparation of Cs3Cu2I5 film
Cs3Cu2I5 film is prepared by single-source vacuum thermal evaporation (VTE). The Cs3Cu2I5 powder is placed into the quartz crucible and heated to achieve a desirable deposition rate (~1 Å/s) under a high vacuum (~10−4 Pa). At the same time, the substrate is heated to 150 °C for in-situ annealing. It is worth noting that the material in the crucible needs to be completely evaporated since Cs3Cu2I5 breaks down during evaporation (the thickness is determined by the mass of the material added to the crucible).
Preparation of CsCu2I3 film
CsCu2I3 powder is synthesized by the anti-solvent method. 2.598 g CsI (0.01 mol), 3.809 g CuI (0.02 mol), and 20 mL DMSO are sequentially added to a centrifugal tube and shaken for 10 min to ensure complete dissolution. Subsequently, 60 ml of DCM is introduced as the antisolvent, and the mixture is centrifuged at 8170 g to obtain the CsCu2I3 powder precipitation. The obtained powder is then dried and ball-milled for thermal evaporation. CsCu2I3 film is prepared by single-source VTE, following the same experimental parameters as those employed in Cs3Cu2I5 film preparation.
Device fabrication
The pattern indium tin oxides (ITOs) are ultrasonic cleaned in detergent, ethanol, acetone, and deionized water successively for 30 min and dried by flowing nitrogen. The ZnO: PEIE precursors are spin-coated at 4000 rpm/s for 45 s, and annealed at 120 °C for 15 min. A 5 nm Al2O3 layer is deposited on ZnO: PEIE by atomic layer deposition (ALD). The ITO/ZnO: PEIE/Al2O3 substrates are then transferred into the vacuum chamber for the deposition of 100 nm Cs3Cu2I5 emitting layer. After the substrates are cooled to room temperature, 10 nm MCP, 20 nm TAPC and 5 nm HAT-CN are sequentially evaporated on it at 0.2 Å/s rate. Finally, 60 nm Al is deposited as the anode. Noted that during all deposition processes, the substrates are rotated at 10 rpm/min under high vacuum (~10−4 Pa). 1,1-Bis[4-[N,N’-di(p-tolyl)amino]phenyl]cyclohexane (TAPC), 2,3,6,7,10,11-Hexacyano-1,4,5,8,9,12-hexaazatriphenylene (HAT-CN), and 1,3-Bis(N-carbazolyl)benzene (MCP) were purchased from Xi’an Polymer Light Technology Co., Ltd.
ESR and XPS measurements
ESR spectra were acquired in the temperature range of 77−300 K (A300-10/12, Bruker) at a microwave frequency of 9.651 GHz, with a microwave power of 20.11 mW. XPS measurements were performed using the Thermo Escalab 250Xi electron spectrometer with monochromatized Al Kα radiation (hν = 1486.6 eV). Both investigations employed in-situ testing methodologies, with the excitation light wavelength of 280 nm.
Magneto-PL/EL measurements
For PL measurements, multimode optical fibers are used to transmit excitation and emission beams. PL is excited by femtosecond lasers (840 nm, 76 MHz, 130 fs, Coherent) through a BBO crystal with triplet frequency generation to generate the 280 nm excitation and recorded by a spectrometer consisting of a monochromator (SP500i, Andor) and an electron multiplying charge-coupled device (EMCCD, Newton 970 P, Andor). The sample is positioned within a liquid helium cryostat, located at the center of a pulsed magnet capable of achieving a designed peak field strength of 60 T and a pulse duration of 400 ms. To ensure accurate measurements, the EMCCD is synchronized with the magnetic field pulses. For magneto-optical investigations involving circular polarization, left and right circular polarizers, comprising a quarter-wave plate and a thin film linear polarizer with axis angles of +45°/−45° respectively, are strategically positioned between the optical fiber and the sample detecting left and right circularly polarized PL. The difference between the EL of the device and the PL of the thin film is that the device’s luminescence does not require excitation light to be excited but driven by a DC voltage (K2400 Keithley).
Modified Brillouin function fit
Upon excitation, Cu²⁺ ions are generated, leading to strong sp-d exchange interactions with the 0D confined excitons. This interaction results in the observation of ΔE saturation at low temperatures and high magnetic fields. The average magnetization \(\left\langle {S}_{z}\right\rangle\) can be characterized by the Brillouin function SzsatBS[gCuμBJH/(kBT)], where the Szsat represents the effective saturation value of \(\left\langle {S}_{z}\right\rangle\) for each Cu²⁺ ion, gCu is the Cu2+ g-factor, μB is the Bohr magneton, and kB is the Boltzmann constant. The Brillouin function is given by BS(x) = [(2S + 1)/2S]coth[((2S + 1)/2S)x]−(1/2S)coth[(1/2S)x], where for Cu²⁺ ion, S = 1/2.
Data availability
All data supporting the findings of this study are provided in the article and the Supplementary Information.
References
Myers, R. C. et al. Zero-field optical manipulation of magnetic ions in semiconductors. Nat. Mater. 7, 203–208 (2008).
Almutlaq, J. et al. Engineering colloidal semiconductor nanocrystals for quantum information processing. Nat. Nanotechnol. 19, 1091–1100 (2024).
Atatüre, M., Englund, D., Vamivakas, N., Lee, S. & Wrachtru, J. Material platforms for spin-based photonic quantum technologies. Nat. Rev. Mater. 3, 38–51 (2018).
Léger, Y., Besombes, L., Fernández-Rossier, J., Maingault, L. & Mariette, H. Electrical control of a single Mn atom in a quantum dot. Phys. Rev. Lett. 97, 107401 (2006).
Dyakonov, M. I. Spin Physics in Semiconductors (Springer International Publishing, 2017).
MacDonald, A., Schiffer, P. & Samarth, N. Ferromagnetic semiconductors: moving beyond (Ga,Mn)As. Nat. Mater. 4, 195–202 (2005).
Pinchetti, V. et al. Excitonic pathway to photoinduced magnetism in colloidal nanocrystals with nonmagnetic dopants. Nat. Nanotechnol. 13, 145–151 (2018).
Bussian, D. A. et al. Tunable magnetic exchange interactions in manganese-doped inverted core-shell ZnSe-CdSe nanocrystals. Nat. Mater. 8, 35–40 (2009).
Pandey, A. et al. Long-lived photoinduced magnetization in copper-doped ZnSe-CdSe core-shell nanocrystals. Nat. Nanotechnol. 7, 792–797 (2012).
Erwin, S. C. et al. Doping semiconductor nanocrystals. Nature 436, 91–94 (2005).
Wannier, G. H. The structure of electronic excitation levels in insulating crystals. Phys. Rev. 52, 191–197 (1937).
Tilchin, J. et al. Hydrogen-like Wannier-Mott excitons in single crystal of methylammonium lead bromide perovskite. ACS Nano 10, 6363–6371 (2016).
Warlimont, H. & Martienssen, W. Springer handbook of materials data (Springer International Publishing, 2018).
Zhang, K. et al. Room-temperature magnetic field effect on excitonic photoluminescence in perovskite nanocrystals. Adv. Mater. 33, 2008225 (2021).
Poggio, M., Myers, R. C., Stern, N. P., Gossard, A. C. & Awschalom, D. D. Structural, electrical, and magneto-optical characterization of paramagnetic GaMnAs quantum wells. Phys. Rev. B 72, 235313 (2005).
Lidzey, D. G., Bradley, D. D. C., Armitage, A., Walker, S. & Skolnick, M. S. Photon-mediated hybridization of frenkel excitons in organic semiconductor microcavities. Science 288, 1620–1623 (2000).
Benin, B. M. et al. Highly emissive self-trapped excitons in fully inorganic zero-dimensional tin halides. Angew. Chem. Int. Ed. Engl. 57, 11329–11333 (2018).
Luo, J. et al. Efficient and stable emission of warm-white light from lead-free halide double perovskites. Nature 563, 541–545 (2018).
Chen, B., Chen, R. & Huang, B. Strong electro-phonon coupling induced self- trapped excitons in double halide perovskites. Adv. Energy Sustain. Res. 4, 2300018 (2023).
Iida, T., Nakaoka, Y., von der Wei, J. P. & Aegerter, M. A. Self-trapped exciton in caesium iodide: I. Theory. J. Phys. C Solid State Phys. 13, 983–992 (1980).
Jun, T. et al. Lead-free highly efficient blue-emitting Cs3Cu2I5 with 0D electronic structure. Adv. Mater. 30, 1804547 (2018).
Lian, L. et al. Photophysics in Cs3Cu2X5 (X = Cl, Br, or I): highly luminescent self-trapped excitons from local structure symmetrization. Chem. Mater. 32, 3462–3468 (2020).
Qu, K. et al. Zn (II)-doped cesium copper halide nanocrystals with high quantum yield and colloidal stability for high-resolution X‑ray imaging. Adv. Opt. Mater. 11, 2202883 (2023).
Kugel, K. I. & Khomskii, D. I. The Jahn-Teller effect and magnetism: transition metal compounds. Sov. Phys. Usp. 25, 231–256 (1982).
Jahn, H. A. & Teller, E. Stability of polyatomic molecules in degenerate electronic states I–Orbital degeneracy. Proc. R. Soc. Lond. A 161, 220–235 (1937).
Donoso, J. P., Magon, C. J., Lima, J. F. & Nascimento, O. R. Electron paramagnetic resonance study of copper–ethylenediamine complex ion intercalated in bentonite. J. Phys. Chem. C. 117, 24042–24055 (2013).
Schlapfer, C. W., Vlasova, N. N., Poznyak, S. K. & Kokorin, A. I. The structure and properties of TiO2-Cu(II)-EDTA ternary surface complexes. J. Colloid Interface Sci. 239, 200–208 (2001).
Han, Y. et al. Observation of the orbital quantum dynamics in the spin-1/2 hexagonal antiferromagnet Ba3CuSb2O9. Phys. Rev. B 92, 180410 (2015).
Zhang, C. et al. Magnetic field effects in hybrid perovskite devices. Nat. Phys. 11, 427–434 (2015).
Canneson, D. et al. Negatively charged and dark excitons in CsPbBr3 perovskite nanocrystals revealed by high magnetic fields. Nano Lett. 17, 6177–6183 (2017).
Neumann, T. et al. Manganese doping for enhanced magnetic brightening and circular polarization control of dark excitons in paramagnetic layered hybrid metal-halide perovskites. Nat. Commun. 12, 3489 (2021).
Yu, J. H. et al. Giant Zeeman splitting in nucleation-controlled doped CdSe:Mn2+ quantum nanoribbons. Nat. Mater. 9, 47–53 (2010).
Van Vleck, J. H. Quantum mechanics-the key to understanding magnetism. Rev. Mod. Phys. 50, 181–189 (1978).
Archer, P. I., Santangelo, S. A. & Gamelin, D. R. Direct observation of sp−d exchange interactions in colloidal Mn2+- and Co2+-doped CdSe quantum dots. Nano Lett. 7, 1037–1043 (2007).
Beaulac, R. et al. Spin-polarizable excitonic luminescence in colloidal Mn2+-doped CdSe quantum dots. Nano Lett. 8, 1197–1201 (2008).
Wang, L. et al. Exploration of nontoxic Cs3CeBr6 for violet light-emitting diodes. ACS Energy Lett. 6, 4245–4254 (2021).
Odenthal, P. et al. Spin-polarized exciton quantum beating in hybrid organic–inorganic perovskites. Nat. Phys. 13, 894–899 (2017).
Walsh, K. M., Pressler, K., Crane, M. J. & Gamelin, D. R. Ferromagnetism and spin-polarized luminescence in lead-free CsEuCl3 perovskite nanocrystals and thin films. ACS Nano 16, 2569–2576 (2022).
Kittel, C. Introduction to Solid State Physics Ch. 12 (John Wiley & Sons, Inc, 2005).
Fiederling, R. et al. Injection and detection of a spin polarized current in a LED. Nature 402, 787–790 (1999).
Du, M. Emission trend of multiple self-trapped excitons in luminescent 1D copper halides. ACS Energy Lett. 5, 464–469 (2020).
Hui, Y., Chen, S., Lin, R., Zheng, W. & Huang, F. Photophysics in Cs3Cu2I5 and CsCu2I3. Mater. Chem. Front. 5, 7088–7107 (2021).
Acknowledgements
Y.H. acknowledges the support from the National Key R&D Program of China (Grant No. 2022YFA1602702), and the Natural Science Foundation of China (Grant No. 11974126). J.L. thanks the support provided by the Natural Science Foundation of China (Grant No. 62322505). F.G. and W.C. acknowledge the support by Knut and Alice Wallenberg Foundation (Dnr. KAW 2019.0082) and the Swedish Government Strategic Research Area in Materials Science on Functional Materials at Linköping University (faculty grant SFO-Mat-LiU no. 2009-00971).
Funding
Open access funding provided by Linköping University.
Author information
Authors and Affiliations
Contributions
Y.H., J.L., and F.G. conceived and supervised the experiments. R.H., L.Y., F.Y., and G.L. performed the optical measurements and analysed the data. Y.P. analysed ESR data. L.Y., Q.H., J.L., and R.H. prepared the sample, characterized the crystal structure and morphology of the samples. J.H. provided theoretical calculation of the sample structure. Z.H. performed photomagnetic measurements. Y.H. and R.H. wrote the paper in consultation with F.G., J.L., L.Y., J.T., L.L., W.C., and Y.P. all authors commented on the manuscript.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks the anonymous, reviewers for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Huang, R., Yang, L., Yang, F. et al. Spin-polarized self-trapped excitons in low-dimensional cesium copper halide. Nat Commun 16, 7264 (2025). https://doi.org/10.1038/s41467-025-62704-y
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-025-62704-y