The role of SiO2 buffer layer in the molecular beam epitaxy growth of CsPbBr3 perovskite on Si(111) | Scientific Reports

Scientific Reports volume 14, Article number: 23618 (2024) Cite this article

711 Accesses

Metrics details

Here we present the growth of molecular beam epitaxy (MBE) CsPbBr3 perovskite films in the orthorhombic crystal structure, with unique structural and morphological properties. CsPbBr3 MBE perovskite films, with thickness ranging from a few nm to 300 nm, were grown in ultra-high vacuum on a Si(111)7 × 7 reconstructed surface, and after the formation of about 2 nm of SiO2, obtained exposing the clean reconstructed Si surface to molecular oxygen that serves to decouple the film from substrate. X-ray diffraction, and electron microscopies, such as scanning electron microscopy and high-angle annular dark-field scanning transmission electron microscopy measurements showed remarkable structural, as well as morphological features, indicating extremely high crystallinity over a large area and across the bulk of the perovskite film. Through the X-ray diffraction patterns we found very narrow (002) and (110) reflections of CsPbBr3 in pure orthorhombic phase, exhibiting a full width at half maximum of only 0.035°, value similar to a bulk Si single crystals, and a surface morphology composed of flat areas up to micrometres in lateral size. Our results shed new light both on preparation of high crystal quality perovskite films, and on the intrinsic properties of this striking fully-inorganic materials, which are exploitable for potential applications in electronic/optoelectronic devices and next generation photovoltaic solar cells.

The term perovskite describes a wide class of compounds having a crystal structure similar to the CaTiO₃1,2,3,4, where the unit cell generally consists of a Ti atom coordinated with six O atoms to form a TiO6 octahedron, and slightly rotated TiO6 octahedra are connected at a vertex to form a three-dimensional networked structure5,6.

Beside oxide perovskites, the halide perovskite gained the attention of the scientific community after 2009, when Kojima and collaborators published their seminal work on photovoltaic (PV) cells, based on hybrid halide organic–inorganic perovskite7. In fact, organometal halide perovskites CH3NH3PbBr3 and CH3NH3PbI3 (where CH3NH3+ and Pb2+ are the metallic cations, and I- and Br-, and Cl- the non-metal anions) were used as visible-light sensitizers for these two first prototypal PV devices, exhibiting a power conversion efficiency of 3.13 (%) and 3.81 (%), respectively7. This opened a path for an intense worldwide scientific push, leading towards the development of countless, and very varied optoelectronic and PV perovskite-based applications8,9. Furthermore, a large number of different elements and compounds can be combined to form a perovskite structure, and using this compositional flexibility, scientists aim to design perovskite crystals that exhibit different physical, optical, and electrical characteristics9.

Currently, all-inorganic caesium lead halide perovskites, CsPbX310,11,12,13, represent a very important class of compounds, particularly due to their increased stability and carriers conductivity compared to their hybrid organic–inorganic counterparts14. Therefore, the enhanced thermal and chemical stability observed in fully-inorganic CsPbBr3 perovskite layers with respect to hybrid perovskites, posed challenges in both synthesis and investigation of their physical and chemical properties. The greater stability is very relevant for potential technological applications, also owing to the significantly elevated decomposition temperature of CsPbBr3, approximately 577 °C15,16.

In CaTiO35, the rotation of TiO6 octahedra and the relative displacement of the cations reduces the crystal symmetry to the orthorhombic or tetragonal. Similarly, the ideal cubic Pm \(\overline{3 }\) m structure5,6 of the CsPbBr3 perovskite can be modified into two different structural phases, including, the tetragonal P4/mbm, and the orthorhombic Pbnm, in which the Pb atom is located in the centre of the octahedra formed by six Br atoms, whereas the Cs atoms are located in the three-dimensional framework cavities. At room temperature (RT) the CsPbBr3 perovskite is monoclinic distorted17. It crystallizes in the orthorhombic (Pnma) phase, adopting the distorted perovskite structure, with a = 8.2440(6) Å, b = 8.1982(8), and c = 11.7351(11) Å, as determined by the single-crystal diffraction18. The {PbBr6}4− octahedra are tilted with respect to the orthogonal geometry of the ideal cubic perovskite structure17. Two successive temperature phase transitions of orthorhombic CsPbBr3 occur at 88 °C and 130 °C, transforming the crystal structure to tetragonal (P4/mbm) and cubic (Pm-3 m), respectively18, as already observed by x-ray analysis and neutron diffraction, by Moller et al.17, and Hirotsu et al.19, as second and first order phase transitions, respectively.

A wide range of synthesis techniques have been applied to produce full-inorganic CsPbBr3 single crystal perovskites. For example, a vertical two-zone tube furnace using an electronic dynamic gradient method showed relevant structural and optical properties, exhibiting extremely high crystalline quality20. On the other hand, colloidal CsPbBr3 perovskite quantum dots with orthorhombic structure (Pnma) were successfully prepared21, and catalyst-free, solution-phase synthesis of CsPbX3 (X = Br, I) nanowires was reported22. Furthermore, monodisperse colloidal nanocubes (from 4 to 15 nm edge lengths) of fully inorganic caesium lead halide perovskites (CsPbX3, X = Cl, Br, and I or mixed halide systems Cl/Br and Br/I), showing nanocrystals with cubic shape and cubic perovskite crystal structure23, and also colloidal CsPbBr3 perovskite nanocrystals, with orthorhombic phase, exhibiting luminescence beyond traditional quantum dots24, were synthesised. Techniques such as physical and chemical vapor depositions for these materials have recently emerged, with the potential to enable a perovskite active layer in a device that can be more stable against exposure to air, mainly due to reduced moisture permeability and/or crystalline structure, as well as morphology modifications25,26. A comprehensive review on the state of the art and outlooks on halide perovskite nanocrystals has been recently reported by Dey et al.27, taking into account of the main outstanding physics/chemistry and device engineering achievements.

Vapour physical deposition methods embrace molecular beam epitaxy, thermal deposition, chemical vapour deposition, as well as plasma sputtering, and UHV-MBE considered the main technique for epitaxial deposition. Epitaxial and quasiepitaxial growth of halide perovskites represent the edges for succeeding new materials, with a better control over the film structures and quality, as well as low level of defects suitable for high end optoelectronics28.

Remarkably, epitaxial growth of cubic phase of CsPbBr3 has been demonstrated on a metal oxide perovskite SrTiO3(100) substrate, via pulsed laser deposition (PLD)29, and chemical vapor-phase deposition (CVD), at a reaction temperature of 450 °C30. In both cases29,30, films of high crystal quality were obtained, exhibiting a photoluminescence (PL) peak at 525 nm with a full width at half maximum (FWHM) of just 19 nm29, and a (100) X-ray diffraction (XRD) reflection wide of only 0.18° (FWHM)30. In 2023, fully inorganic CsPbBr3 films, exhibiting large columnar grains, with different sizes up to ~ 7 μm , and a preferential (001) crystal orientation, were grown at a base pressure of about 2.7 × 10−5 mbar, by thermal vapour deposition and subsequently annealing at high temperature of 375 °C31. In addition, molecular beam epitaxy (MBE) was used to synthetize both CsPbBr3 and CsSnBr3 on an Au(001) single crystal32. While these results are extremely valuable, in order to enable applications in real-world devices it is essential to expand the group of potential substrates, particularly including technologically-relevant materials such as silicon and other conventional device components.

In this work, we report for the first time, CsPbBr3 films grown in a wide thickness range (from 10 to 300 nm) on Si(111)7 \(\times\) 7 by ultra-high vacuum (UHV) molecular beam epitaxy (MBE). Thin perovskite film exhibited a × 2 reconstruction on Si(111)7 \(\times\) 7, whereas, for the thicker one the interaction with substrate was mediated by a very thin amorphous layer of SiO2, obtained by exposing the clean Si(111)7 \(\times\) 7 reconstructed surface to pure O2, at room temperature (RT). The presence of the thin SiO2 on the Si substrate permits the growth of highly ordered orthorhombic phase of thick CsPbBr3 (300 nm) film, exhibiting unique structural and morphological, properties, as probed by X-ray diffraction, scanning electron microscopy (SEM), as well as high-angle annular dark-field scanning transmission electron microscopy (HAADF STEM). The thin layer of SiO2 acted as a cushion layer to promote the MBE growth of very thick perovskite, decoupling the otherwise highly mismatched film from the silicon substrate.

The CsPbBr3 powder, to be used as precursor for MBE growth inside a UHV-MBE Knudsen cell, was first characterised. The X-ray diffraction pattern of such powder displayed orthorhombic symmetry (Pbnm (62) space group)15,18,20,33,34,35 (Supplementary Fig. S1). This is in agreement with the orthorhombic symmetry, for which Rietveld refinement measurements have been recently reported in the literature both through high-resolution (HR) synchrotron XRD at ALBA (Spain) facility35, and neutron powder diffraction SINQ spallation source, at PSI (Switzerland)35 on CsPbBr3 perovskite, and already applied to nanowires34, nanotwins36, nanodots21, as well as nanocrystal superlattices37. In particular, a clear splitting of the (110) and (220) peaks in the diffraction pattern was observed20. Moreover, in our measurements we do not find evidence of the presence of an additional PbBr2 phase originating from perovskite decomposition, typically found with XRD peaks at 14.37°, 18.58°, 21.63°, 22.04° and 23.69°, which correspond to the lattice planes (101), (002), (011), (102) and (111)38. This proves the high purity of the CsPbBr3 powder used for our UHV MBE film growths.

The samples were grown in a base pressure of 8.0 × 10−10 mbar in a Knudsen evaporation cell with a rate of ~ 0.1 Å/s, keeping the substrates at temperature of ~ 130 °C.

A preferential growth orientation of CsPbBr3 (Pnma space group, with a = 8.2440 Å, b = 8.1982, c = 11.7351 Å), on Si(111)7 × 7 (space group Fd3̅m, with aSi = b = c = 5.43 Å) is obtained, assuming a large lattice mismatch factor of 7% [f = (1 − a overlayer/a substrate)], for the planes (001)CsPbBr3 being parallel to (111)Si with the lattice constant aCsPbBr3 much higher than aSi. Despite such a large difference in lattice constants, one-unit cell of CsPbBr3, along the [1–10]Si direction, could match with two-unit cells of Si in a × 2, as shown on the RHEED pattern (Fig. 1). The RHEED pattern, collected with the direction of electron beam parallel to the [11–2]Si, of the Si(111)7 × 7 surface (Fig. 1a), and a deposited thin film of CsPbBr3 (Fig. 1b), show the streaks (− 1/7, 1/7) of the clean Si(111)7 × 7 surface before the growth, and (− 1/2,1/2) spots, induced by the film growth. The poor sharpness and spot-like appearance of the streaks (− 1/2,1/2) of the CsPbBr3 film indicate the coalescence of CsPbBr3 islands during the growth.

RHEED patterns collected on clean Si(111)7 × 7 (a), and after the deposition of ~ 100 Å of CsPbBr3, (b). The arrows indicate the bulk Si (blue) integer streaks order (00), (0–1), (01), and the fractional streaks order of the × 2 CsPbBr3 film (red), in reciprocal space; the primary energy (Ep) is 12 keV; the Ep incident beam was parallel to the [11–2] orientation of Si(111), while the fractional streaks order (0–1/7), and (01/7), stemming from the × 7 Si reconstruction are marked.

It is worth noting that the preferential growth mode of CsPbBr3 perovskite directly on reconstructed Si(111)7 × 7 is quite promising, but the film develops from the beginning more appropriately as island-like, Volmer-Weber mode, and/or as Stranski–Krastanov growth mode39,40, rather than as a needed continuous film.

To improve the crystal quality of the CsPbBr3 film, we designed a structure with a buffer layer between the film and the Si(111)7 × 7 surface. For this we chose SiO2, for which in-situ Auger Electron Spectroscopy spectra are shown (Supplementary Fig. S2). The formation of an amorphous SiO2 layer on the Si(111)7 × 7 surface is a simple and straightforward process. Furthermore, this is expected to be relevant for technological applications: in optoelectronic systems, active perovskite is frequently deposited on oxide layers12,13,29,30,31. The main idea behind this architecture is to decouple the surface atoms of the Si substrate from the perovskite film via a very thin amorphous SiO2 buffer layer, to minimise constraints on nucleation and favour epitaxial growth from the initial, seeding perovskite layer. This is facilitated by the intact arrival of the entire CsPbBr3 molecule on the substrate surface, as confirmed by mass spectrometry (Supplementary Fig. S3), enabling the CsPbBr3 molecules to self-organize based on the optimized temperature applied to the Si substrate and the molecular flux from the MBE cell.

Various CsPbBr3 films have been grown, by considering different substrate’s temperatures, from RT up to 250 °C, and with different Knudsen cell fluxes, from 0.1 up to 18 Å/s, to find optimal conditions for continuous growth of large-scale epitaxial films, trying to keep the same thickness of the films, as measured by a quartz microbalance.

The thickness of the films, as measured by a quartz microbalance, was kept constant. Using ex-situ analysis by XRD, EDS, SEM, and HAADF STEM, as well as differential phase contrast (DPC) STEM we find that the optimal conditions are Tsubstrate = 130 °C and cell flux of ~ 0.1 ÷ 0.2 Å/s .

The XRD pattern from a 300 nm-thick CsPbBr3 film grown on SiO2/Si(111)7 × 7 at 130 °C is shown (Fig. 2a). This pattern features a very intense peak at 15.195°, and two other peaks at 30.700°, and 28.435°, where the first two are associated with orthorhombic CsPbBr3 perovskite, and the latter is the (111) reflection from the substrate. The detailed structure of the perovskite peaks at 2ϑ of 15.2° and 30.7°, is reported (Fig. 2b and c).

XRD from CsPbBr3 films grown on thin SiO2/Si (111)7 × 7 at substrate temperature of 130 °C (a); (b) and (c) are the enlarged regions around 2ϑ of 15° and 30° of the profile (a). The least-square fitting procedure, by using the convolution of Gaussian and Lorentzian functions, employing four components for each peak are reported (pink, blue, green and yellow lines), identifying the main (002)/(004) and (110)/(220) components as due to the orthorhombic CsPbBr3 structure, and the others (green and yellow line) as their Cu Kα2 X-ray replica.

The XRD peaks for the CsPbBr3 film are not fully symmetric, due to the presence of multiple contributions. To get clearer physical insights and attribute each contribution of the X-ray diffraction peaks, signing, noteworthy for the first time, a limit of their FWHM, a least-square fitting procedure, using the convolution of Gaussian and Lorentzian functions, has been applied to isolate these components, findings four components for both peaks. In fact, we have to consider that each XRD peak contains two reflections, the (002) (110), (004) and (220) of orthorhombic CsPbBr3 phase. Additionally, the X-ray beam consists of two photon lines, Cu Kα1 and Cu Kα2, as the X-ray source is not monochromatic. The first two peaks at lower angles (pink and blue lines of Fig. 2b) are located at 15.187° ± 0.02°and 15.203° ± 0.02°, while the first two peaks at higher angles (pink and blue lines of Fig. 2c) are located at 30.688° ± 0.02°and 30.704° ± 0.02°. These components are attributed to (002) (110) and (004) (220) reflections of the orthorhombic CsPbBr3 perovskite phase (Pnma (62) space group), in good agreement with literature15,18,20,21,33,34,35,36,37. The other lines (green and yellow curves) with peaks at 15.225° ± 0.02°, and 15.240° ± 0.02°, (Fig. 2b) and 30.767° ± 0.02°, and 30.783° ± 0.02°, (Fig. 2c), are peak replica originating from the Cu Kα2 X-ray (λ = 1.54439 Å). According to the Bragg’s law, the spacing between the two lines, Cu Kα1, and Cu Kα2, split by a given energy interval, increases with the 2ϑ diffraction angle value, producing that the replica at higher 2ϑ angle (Fig. 2c) are better separated, from their main peaks, while are just a shoulder at lower angle (Fig. 2b). The intensity of each component have been fitted, under the constrain that the replica of the Cu Kα2 x-ray line have a natural half intensity with respect to the main Cu Kα1 X-ray line (λ = 1.54056 Å), and, furthermore, their angle positions imposed by the Cu Kα2 wavelength. It is worth noting that, for the four components of each peak similar parameters were used: (Gaussian width, σG, and Lorentzian width, ΓL) σ(002), (110)G = 0.020°;Γ(002), (110))L = 0.020°; σreplica (002), (110)G = 0.025°; Γreplica (002), (110)L = 0.025°); (σ(004), (220)G = 0.032°; Γ(004), (220)L = 0.032°; σreplica (004), (220)G = 0.032°; Γreplica(004), (220)L = 0.032°). The Δ (°) between the main (002) and (110); (004) and (220) components, as well as that of their replica is of 0.016°. On the other hand, the shift between the main component (002) and its replica is 0.038°, whereas that of (004), and its replica, is of 0.079°, as expected, and similar to that found for the (111) reflection peak of 0.070° from the bare Si(111) substrate (see Supplementary Fig. S4).

By applying the Bragg’s law (n\(\cdot \lambda\) = 2\(\cdot\) d \(\cdot\)sinϑ), where \(\lambda\)(Cu Kα1) = 1.54056 Å and n = 1 for the (110), and n = 2 for (002), we get the orthorhombic CsPbBr3 lattice constant values a CsPbBr3 = 8.235 Å, and c CsPbBr3 = 11.658 Å, in good agreement with previous literature15,18,20,21,33,34,35,36,37. This points to have the type (002) CsPbBr3 planes with preferentially aligning parallel to (111) Si planes.

Remarkably, the main peak, containing the (002) and (110) contributions (Fig. 2a) shows a FWHM of 0.06° in the raw profile, which drops down to 0.035° by separating the individual components via fitting. This width is the same order of magnitude as the (111) XRD peak of Si(111) single crystal substrate (0.025°, Supplementary Fig. S4). Moreover, the observed value is the lowest reported so far in the literature for orthorhombic CsPbBr3: for instance this is ~ 4.5 times narrower than that reported for a single crystal ingot (0.16°) produced in a two-zone furnace method, and sliced into wafers of 8 mm in diameter and 1.5 mm in thickness20. The high intensity of the (002) (110) peaks (Fig. 2) compared to the (111) peak from the Si substrate denotes a preferential film growth along the [001]c perovskite axis. By slightly tilting the sample while keeping the XRD beam and detector conditions unchanged, their intensity is reduced, showing, indeed, a more pronounced Si (111) peak of the substrate (Supplementary Fig. S5). The exceptional characteristics of the XRD data from our CsPbBr3 MBE films indicate a very low density of dislocations or other defects, which, in the literature, often contribute to widen the XRD peak, as well as an absence of residual stress, cushioned by the thin SiO2 buffer layer.

The high-angle annular dark-field (HAADF) STEM cross-sectional image of a 300 nm CsPbBr3 MBE film grown on SiO2/Si(111)7 × 7 substrate is reported (Supplementary Fig. S6), while the high-resolution HAADF STEM images of the same film are shown (Fig. 3).

HR HAADF STEM images of 300 nm CsPbBr3 MBE film grown on SiO2/Si(111)7 × 7: (a) HAADF image of the entire cross section; (b) atomic resolution HAADF (left) and DPC (right) STEM images of domains A and B tilted to the [120] and [110] zone axes of the CsPbBr3 film respectively; and (c) corresponding atomic model for the orthorhombic phase of CsPbBr3 (Cs: light blue, Pb: grey, Br: brown). The vertical dashed red line separates areas A and B.

The vertical dashed red line marks the step between two terraces of CsPbBr3 film. It separates two large domains indicated with A, and B, having a lateral size of at least 1 µm. The corresponding HAADF and differential phase contrast (DPC) STEM images aligned to the zone axis are also shown (Fig. 3b) for each domain. They are assigned to the [120] and [010] zone axes of the CsPbBr3. The film was slightly tilted to align onto the zone axis for areas A and B independently. Therefore, the [120], [110] CsPbBr3 and [110] Si planes are not perfectly parallel to each other. The misorientation is as large as a few degrees.

Remarkably, the domains grow sharing the same orientation of the c axis [001], in full agreement with the XRD-defined film texture. The corresponding atomic model of the orthorhombic CsPbBr3 from the given zone axes are also shown (Fig. 3c). Here, it is worth noting that the distance, dA [001] and dB [001], between the (002) planes (along the [001] direction), is equal to 11.65 Å for both domains, in good agreement with the c CsPbBr3 = 11.658 Å found from XRD measurements.

To highlight the overall elemental composition of the CsPbBr3 film, in addition to the HAADF image, the EDX measurements have been performed (Fig. 4). The HAADF cross-section (Fig. 4a) shows the Si substrate (black part, at the bottom), as well as the CsPbBr3 film grown on top (grey). The thin C/Pt protection layer, obtained during the focused ion beam (FIB) preparation of the cross section lamella, is, furthermore, observable at the topmost of the image. We confirm the film thickness estimated by quartz microbalance during the MBE growth, to be about 300 nm. In addition, in some areas on the sample surface, underlined, here, by the two blue circles, some cross-section of parallelogram-like structures, can be recognised, consistent with the surface SEM observations reported below.

HAADF image (a) and Pb (b), Br (c), and Cs (d) EDX elemental maps of 300 nm CsPbBr3 MBE film grown on SiO2/Si(111)7 × 7 substrate.

The corresponding Pb, Br, and Cs EDX elemental maps of this HAADF image is reported (Fig. 4b, c, d), clearly showing that all three elements are very homogeneously distributed throughout the entire film, exhibiting an average composition of 19 at. % Cs, 19 at. % Pb and 61 at. % Br (with a confidence interval of 1 at %), in good agreement with the stoichiometry of CsPbBr3 perovskite.

To detail better the composition of the entire film, the high-resolution HAADF STEM image of the film cross section tilted to the [110]CsPbBr3 zone axis of area B depicted in Fig. 3, is reported (Fig. 5), including the Si substrate at the bottom (Fig. 5a). Both HAADF image and EDX elemental maps of a similar region at the Si/perovskite interface are also shown (Fig. 5b). An amorphous layer of SiO2 about 2 nm thick, acting as a buffer layer between the Si substrate and perovskite film, can clearly be observed and identified. A highly ordered homoepitaxial CsPbBr3 film, almost free of defects and/or dislocations, has been obtained. The reasons for the high crystalline quality of this film growth may lie in a multifactorial process.

Cross-sectional HAADF-STEM image (a) and elemental maps (b) of the CsPbBr3 MBE film, including the Si substrate.

The presence of the SiO2 amorphous layer, in conjunction with optimized UHV-MBE growth conditions—specifically the substrate temperature, which primarily affects molecular kinetics, the Knudsen cell, which influences the velocity and flux of CsPbBr3 perovskite molecules and the fact that the stoichiometric CsPbBr3 molecules can arrive intact (Supplementary Information Fig. S3)—has likely facilitated the uniform onset of film growth. This results in a process we term the "slipping-and- sticking " mechanism on top of the oxide buffer layer. Consequently, this has promoted the homoepitaxial growth of the CsPbBr3 film to an impressive thickness of 300 nm. Notably, this amorphous thin oxide layer features sharp interfaces on both sides, interfacing with silicon on one side and perovskite on the other, which helps mitigate strain induced by the large lattice parameter mismatch between silicon and CsPbBr3 perovskite.

Finally, the surface morphology of the CsPbBr3 film has been investigated by surface SEM measurements. The reported images (Fig. 6) reveal a very compact film, composed by numerous particularly uniform, large and flat terraces. These extended flat terraces have lateral dimensions, ranging between several hundred nanometres and a few microns (Fig. 6a). They arrange in compact interlocking structures, with very narrow gaps. Small grains, like the one surrounded by a red dashed circle (Fig. 6b), are occasionally present but are comparatively rare. It is essential to highlight that this CsPbBr3 structure having obtained by UHV-MBE growth is quite exceptional. Up to now, conventional PLD, and CVD physical methods, have produced epitaxial films composed of much more irregular small grains29, and separated nanoplates30.

SEM images of the epitaxial CsPbBr3 film. In (a) markers indicate the size of terraces and other surface features. In (b) the red dashed circle and ellipse mark a small interstitial grain and a 1D structure respectively.

A second minority morphological phase is present: long and narrow, faceted 1D structures, very bright in the SEM contrast, can be seen on the surface (for example, in the dashed blue ellipse in Fig. 6b), akin to CsPbBr3 nanowires reported in Ref.41. They extend for several hundred nanometers in the longitudinal direction and are about 100 nm wide. The same features were visible in cross-section within red dotted circles (Fig. 4a). These nanowires, are only formed on the top surface of the film and do not affect the bulk properties of the material underneath. They represent a minority, and could be explained by the presence of a residual vacuum dominated by the perovskite, persisting for a few hours, even after the closing of the Knudsen shutter cell. They could be avoided placing, directly, a capping layer of other possible/useful materials, just after the film growth.

In summary, we reported the growth of thin (~ 10 nm) and thick (~ 300 nm) CsPbBr3 perovskite, by UHV-MBE, on Si(111)7 × 7 and SiO2/Si(111)7 × 7, after in-situ formation of a thin oxide layer, introduced as a buffer layer to decouple the Si substrate from the perovskite film. The growth of CsPbBr3 showed a RHEED pattern with × 2 surface reconstruction. Ordered thick films (up to 300 nm) of CsPbBr3 with orthorhombic Pbmn (62) structure were achieved on SiO2/Si(111)7 × 7, as attested by XRD, HAADF STEM, EDX, and SEM measurements, thanks to a slipping-and-sticking process induced by the thin amorphous SiO2 layer. The excellent crystal quality of the film was proven by the narrowest reported FWHM of the (002) and (110) XRD reflections (0.035°), comparable to that of Si(111) substrate (Supplementary Information Fig. S4), in addition, the electron microscopy studies also demonstrated the presence of micron-scale single crystal terraces. It is noteworthy that the results presented in this work are a significant advancement in the study and development of materials, crucial for optoelectronic and energy conversion applications. Specifically, halide perovskites such as CsPbBr3, synthesized from entirely inorganic precursor salts under ultra-high vacuum conditions, pave the way for the creation of new nanostructures with exceptional fundamental physical properties.

In-situ experiments were performed at the CNR-ISM laboratory of Tor Vergata, Rome (Italy). The Si(111) substrates were cleaned in a UHV MBE chamber (base pressure: 8.0 × 10−10 mbar) by applying several flashes, at about 1150 °C, until the Si(111)7 × 7 reflection high electron energy diffraction (RHEED) pattern of clean silicon was obtained. Molecular deposition of CsPbBr3 was carried out, keeping the substrate at 130 °C. The mass spectrometry was carried out by SRS RGA-300 AMU, during the CsPbBr3 films growth. AES spectra were collected at the normal incidence with a primary electron beam of EP = 3.0 keV. AES data were acquired in the first derivative mode with a PHY 255G double-pass cylindrical mirror analyser equipped with a coaxial electron gun, with a resolution of 0.5 eV. The Si(111)7 × 7 surface was exposed to 1.2 × 104 Langmuir (L) (1 L = 1 × 10−6 mbar · 1 s) of O2 at RT in order to oxidise the first few nm into amorphous SiO2. Thin (~ 10 nm) and thick (~ 300 nm) CsPbBr3 were evaporated on clean Si(111)7 × 7 surface, and after exposure to O2. The substrate was kept at ~ 130 °C during MBE growth with a rate of ~ 0.1 Å/s from a Knudsen cell (RIBER) loaded with stoichiometric CsPbBr3 powder, purchased from TCI CAS RN :15243-48-0, with low water content, without any further purification. (A thin film of native silicon(111) is also suitable for growth of CsPbBr3).

Ex-situ X-ray diffraction (XRD) measurements were performed at University of Roma 2, Engineering Dep., Rome (Italy) by means of a Rigaku SmartLab diffractometer, working in ((ϑ—2ϑ) Bragg–Brentano geometry, equipped with a Cu source (Kα1 = 1.54056 Å, Kα2 = 1.54439 Å) and a D/teX Ultra 250 silicon strips detector. XRD spectra (5°- 60°) were acquired in a single scan with a step of 0.05°, ~ 0.2 s/pt, and a horizontal slit of 5 mm. Scanning electron microscopy (SEM) measurements were performed by using a ZEISS SIGMA 300 FESEM.

High-resolution scanning transmission electron microscopy (STEM) characterisation was performed in a double corrected ThermoFisher Spectra 300 S/TEM operated at 300 kV. Images were acquired using a current of 50 pA. Segmented STEM detectors allow the acquisition of both High-Angle Annular Dark Field (HAADF) and Differential Phase Contrast (DPC) images. Compositional maps were acquired using Velox, with a probe current of ~ 150 pA and rapid raster scanning, collecting the Energy-Dispersive X-Ray (EDX) signal on a Dual-X system comprising two detectors on either side of the sample, for a total acquisition angle of 1.76 Sr. A cross-section of the films with a final thickness of 150 nm was prepared by Focused Ion Beam (FIB) milling using a Helios Nanolab FIB/SEM (Thermo Fisher Scientific). A thin layer of platinum (200 nm) was deposited over the target area of the film using the electron beam (2 nA, 5 kV) prior exposure to the ion-beam. A standard in-situ lift-out procedure was employed. Further thinning was performed at 30 kV/100 pA. Final polishing was at a reduced voltage of 2 kV. This procedure minimizes the amount of damaged material on the surface and reduces the levels of implanted gallium in the final thin foil.

All data generated or analysed during this study are included in this published article [and its supplementary information files].

Rose, G. Beschreibung einiger neuen Mineralien des Urals. Poggendorffs Ann. Phys. Chem. 124, 551–573 (1839).

Article Google Scholar

Katz, E. A. Perovskite: Name puzzle and german-russian odyssey of discovery. Helv. Chim. Acta 103, e200006–e200011 (2020).

Article Google Scholar

Megaw, H. Crystal structure of barium titanate. Nature 155, 484–485 (1945).

Article CAS Google Scholar

Rooksby, H. P. Compounds of the structure type of the calcium titanate. Nature 155, 484 (1945).

Article ADS CAS Google Scholar

Yashima, M. & Ali, M. Structural phase transition and octahedral tilting in the calcium titanate perovskite CaTiO3. Solid State Ionics 180, 120–126 (2009).

Article CAS Google Scholar

Shi, F., Fu, G., Xiao, E.-C. & Li, J. Lattice vibrational characteristics and dielectric properties of pure phase CaTiO3 ceramic. J. Mater Sci. Mater Electron. 31, 18070–18076 (2020).

Article Google Scholar

Kojima, A., Teshima, K., Shirai, Y. & Miyasaka, T. Organometal halide perovskites as visible-light sensitizers for photovoltaic cells. J. Am. Chem. Soc. 131, 6050–6051 (2009).

Article CAS PubMed Google Scholar

Lee, M. M., Teuscher, J., Miyasaka, T., Murakami, T. N. & Snaith, H. J. Efficient hybrid solar cells based on meso-superstructured organometal halide perovskites. Science 338, 643–647 (2012).

Article ADS CAS PubMed Google Scholar

Jena, A. K., Kulkarni, A. & Miyasaka, T. Halide perovskite photovoltaics: Background, status, and future prospects. Chem. Rev. 119, 3036–3103 (2019).

Article CAS PubMed Google Scholar

Iso, Y. & Isobe, T. Review: Synthesis, luminescent properties, and stabilities of cesium lead halide perovskite nanocrystals. ECS J. Solid State Sci. Technol. 7, R3040–R3045 (2018).

Article CAS Google Scholar

Pan, Y. et al. Progress in the preparation and application of CsPbX3 perovskites. Mater. Adv. 3, 4053–4068 (2022).

Article CAS Google Scholar

Liu, X., Li, J., Wang, X. & Yang, D. Inorganic lead-based halide perovskites: From fundamental properties to photovoltaic applications. Mater. Today 61, 191–217 (2022).

Article CAS Google Scholar

Wang, K. et al. Progress of inverted inorganic cesium lead halide perovskite solar cells. Cell Rep. Phys. Sci. https://doi.org/10.1016/j.xcrp.2023.101726 (2023).

Article PubMed PubMed Central Google Scholar

Guo, W. et al. Stability of hybrid organic-inorganic perovskite CH3NH3PbBr3 nanocrystals under Co-stresses of UV light illumination and temperature. Nanomaterials 9, 1158. https://doi.org/10.3390/nano9081158 (2019).

Article CAS PubMed PubMed Central Google Scholar

Zhang, C. et al. Thermal stability of CsPbBr3 perovskite as revealed by in situ transmission electron microscopy. APL Mater. 7, 071110–071116 (2019).

Article ADS Google Scholar

Kanak, A. et al. Melting and crystallization features of CsPbBr3 perovskite. Cryst. Growth Des. 22, 4115–4121 (2022).

Article CAS Google Scholar

Moller, C. K. Crystal structure and photoconductivity of caesium plumbohalides. Nature 182, 1436 (1958).

Article ADS CAS Google Scholar

Stoumpos, C. C. Crystal growth of the perovskite semiconductor CsPbBr3: A new material for high-energy radiation detection. Cryst. Growth Des. 13, 2722–2727 (2013).

Article CAS Google Scholar

Hirotsu, S., Harada, J., Iizumi, M. & Gesi, K. Structural phase transition in CsPbBr3. J. Phys. Soc. Jpn. 37, 1393–1398 (1974).

Article ADS CAS Google Scholar

Zhang, M. et al. Growth and characterization of all-inorganic lead halide perovskite semiconductor CsPbBr3 single crystals. Cryst. Eng. Comm. 19, 6797–6803 (2017).

Article CAS Google Scholar

Cottingham, P. & Brutchey, R. L. On the crystal structure of colloidally prepared CsPbBr3 quantum dots. Chem. Commun. 52, 5246–5249 (2016).

Article CAS Google Scholar

Zhang, D., Eaton, S. W., Yu, Y., Dou, L. & Yang, P. Solution-phase synthesis of cesium lead halide perovskite nanowires. J. Am. Chem. Soc. 137, 9230–9923 (2015).

Article CAS PubMed Google Scholar

Protesescu, L. et al. Nanocrystals of cesium lead halide perovskites (CsPbX3, X = Cl, Br, and I): Novel optoelectronic materials showing bright emission with wide color gamut. Nano Lett. 15, 3692–4369 (2015).

Article ADS CAS PubMed PubMed Central Google Scholar

Swarnkar, A. et al. Colloidal CsPbBr3 perovskite nanocrystals: Luminescence beyond traditional quantum dots. Angew. Chem. Int. Ed. 54, 15424–15428 (2015).

Article CAS Google Scholar

Tian, C. et al. Chemical vapor deposition method grown all-inorganic perovskite microcrystals for self-powered photodetectors. ACS Appl. Mater. Interfaces. 11, 15804–15812 (2019).

Article CAS PubMed Google Scholar

Bonomi, S. & Malavasi, L. Physical and chemical vapor deposition method supplied to all-inorganic metal halide perovskites. J. Vac. Sci. Tech. A. 38, 060803–060811 (2020).

Article CAS Google Scholar

Dey, A. et al. State of the art and prospects for halide perovskite nanocrystals. ACS Nano 15, 10775–10981 (2021).

Article CAS PubMed PubMed Central Google Scholar

Wang, L., King, I., Chen, P., Bates, M. & Lunt, R. R. Epitaxial and quasiepitaxial growth of halide perovskites: New routes to high end optoelectronics. APL Mater. 8, 100904–22 (2020).

Article ADS CAS Google Scholar

Yuan, B. et al. Epitaxial growth of quasi-intrinsic CsPbBr3 film on a SrTiO3 substrate by pulsed laser deposition. ACS Appl. Electron. Mater. 3, 5592–5600 (2021).

Article CAS Google Scholar

Chen, J. et al. Single-crystal thin films of cesium lead bromide perovskite epitaxially grown on metal oxide perovskite (SrTiO3). J. Am. Chem. Soc. 139, 13525–13532 (2017).

Article CAS PubMed Google Scholar

Zhang, Y. et al. Preparation of CsPbBr3 films with tens of micrometer-scale grains and preferential orientation for perovskite solar cells. ACS Appl. Energy Mater. 6, 12190–12197. https://doi.org/10.1021/acsaem.3c01133 (2023).

Article CAS Google Scholar

Rieger, J. et al. Epitaxial inorganic metal-halide perovskite films with controlled surface terminations. Phys. Rev. Mat. 7, 035403–035406 (2023).

CAS Google Scholar

Hooper, T. J. N., Fang, Y., Brown, A. A. M., Pu, S. H. & White, T. J. Structure and surface properties of size-tuneable CsPbBr3 nanocrystals. Nanoscale 13, 15770–15789 (2021).

Article CAS PubMed Google Scholar

Imran, M. et al. Colloidal synthesis of strongly fluorescent cspbbr3 nanowires with width tunable down to the quantum confinement regime. Chem. Mater. 28, 6450–6454 (2016).

Article CAS PubMed PubMed Central Google Scholar

López, C. A. et al. Crystal structure features of CsPbBr3 perovskite prepared by mechanochemical synthesis. ACS Omega 5, 5931–5938 (2020).

Article PubMed PubMed Central Google Scholar

Bertolotti, F. et al. Coherent nanotwins and dynamic disorder in cesium lead halide perovskite nanocrystals. ACSNano 11, 3819–3831 (2017).

CAS Google Scholar

Toso, S., Baranov, D., Giannini, C., Marras, S. & Manna, L. Wide-Angle X-ray diffraction evidence of structural coherence in CsPbBr3 nanocrystal superlattices. ACS Mater. Lett. 1, 272–276 (2019).

Article CAS PubMed PubMed Central Google Scholar

Ahmad, Z. & Mishra, A. Growth of PbBr2 microrods with unique structure and surface morphology. J. Mater. Sci. Mat. Electr. 31, 4672–4676 (2020).

Article CAS Google Scholar

De Padova, P. et al. Identification of the Si 2p Surface Core Level Shifts on the Sb/Si(001)-2 ✕ 1 Interface. Phys. Rev. Lett. 81, 2320–2323 (1998).

Larciprete. R., De Padova, P. et al. Ge/Si(001)c(4 ✕ 2) interface formation studied by high-resolution Ge 3d and Si 2p core-level spectroscopy. Phys. Rev B. 61, 16006–16014 (2000).

Eaton, S. W. et al. Lasing in robust cesium lead halide perovskite nanowires. PNAS 113, 1993–1998 (2016).

Article ADS CAS PubMed PubMed Central Google Scholar

Download references

This work was supported by the EU Horizon 2020 Framework Programme Call: H2020-LC-SC3-2020-RES-RIA CitySolar Project grant N: 101007084. A. D. C. acknowledges CANVAS project of Ministero della Transizione Ecologica – Ricerca di Sistema (CUP B53C22009710005), Cinzia Giannini, director of Institute of Crystallography of CNR, is recognized for the fruitful scientific discussion, mainly related to the XRD measurements. The authors are grateful to L. Imperatori, S. Priori, and A. Ippoliti for their invaluable technical support. Y. P. I. and G. D. thank Dr. Paola Parlanti and Dr. Mauro Gemmi for access to the FIB in CMI@SSSA.

National Research Council, Institute of Structure of Matter (CNR - ISM), Via Fosso del Cavaliere, 100, 00133, Roma, Italy

Paola De Padova, Carlo Ottaviani & Aldo Di Carlo

National Research Council, Institute of Atmospheric Sciences and Climate (CNR - ISAC), Via Fosso del Cavaliere, 100, 00133, Roma, Italy

Bruno Olivieri

Electron Spectroscopy and Nanoscopy, Italian Institute of Technology (IIT), Via Morego, 30, 16163, Genova, Italy

Yurii P. Ivanov & Giorgio Divitini

You can also search for this author in PubMed Google Scholar

P. D. P. and A. D. C. conceived the idea of this project. P. D. P., C. O. and B. O. grown CsPbBr3 films in the UHV MBE apparatus; cleaned the Si(111)7 × 7 substrates and performed the thin SiO2 layers on Si(111)7 × 7 substrates. P. D. P., C. O. and B. O. carried out the RHEED patterns, Auger electron spectroscopy and relative data analysis. P. D. P. performed the XRD, SEM and PL measurements and relative data analysis. Y. P. I. and G. D. carried out the STEM characterization and relative data analysis. P. D. P. wrote the original draft. C. O., B. O., I. Y., G. D. and A. D. C. helped to edit and review the manuscript. A. D. C. secured funding. All authors contributed to the discussion of results and the final manuscript preparation.

Correspondence to Paola De Padova, Giorgio Divitini or Aldo Di Carlo.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, 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 you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. 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-nc-nd/4.0/.

Reprints and permissions

De Padova, P., Ottaviani, C., Olivieri, B. et al. The role of SiO2 buffer layer in the molecular beam epitaxy growth of CsPbBr3 perovskite on Si(111). Sci Rep 14, 23618 (2024). https://doi.org/10.1038/s41598-024-67889-8

Download citation

Received: 04 April 2024

Accepted: 17 July 2024

Published: 09 October 2024

DOI: https://doi.org/10.1038/s41598-024-67889-8

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative