Ioannis
Konidakis
*a,
Konstantinos
Brintakis
a,
Athanasia
Kostopoulou
a,
Ioanna
Demeridou
ab,
Paraskevi
Kavatzikidou
a and
Emmanuel
Stratakis
*ab
aInstitute of Electronic Structure and Laser (IESL), Foundation for Research and Technology-Hellas (FORTH), 71110 Heraklion, Crete, Greece. E-mail: ikonid@iesl.forth.gr; stratak@iesl.forth.gr; Tel: +30-2810-391835 Tel: +30-2810-391274
bPhysics Department, University of Crete, 71003 Heraklion, Crete, Greece
First published on 1st June 2020
Owing to their exceptional optoelectronic properties, all-inorganic lead halide perovskites offer enormous potential for next generation photonic, light-emitting, and optoelectronic devices. However, their usage is significantly limited by their poor stability upon moisture exposure and lead toxicity issues. Moreover, many of the aforementioned applications rely on the development of confined perovskite patterns of various shapes and periodicities. Here we report a simple and low-temperature method enabling the controlled incorporation of photoluminescent all-inorganic metal halide PNCs into a silver phosphate glass (AgPO3) matrix which is transparent in most of the visible range. The developed fabrication protocol is based on a simple melting encapsulation process in which pre-synthesized perovskite crystals are inserted in the glass matrix, following the initial glass quenching. Using this novel approach, two types of composite perovskite glasses are prepared, one that hosts perovskite isles and the second in which a thin perovskite layer is embedded beneath the glass surface. Both types of composite glasses exhibit remarkable photoluminescence stability when compared to the ambient air-exposed perovskite crystals. More importantly, by means of a simple and fast cw-laser processing technique, we demonstrate the development of encapsulated dotted perovskite micropatterns within the composite perovskite glass. The ability of the proposed system to resolve stability and lead toxicity issues, coupled with the facile formation of highly luminescent perovskite patterns pave the way towards the broad exploitation of perovskite crystals in photonic applications.
While both encapsulation techniques are without doubt a remarkable advancement over the stability and toxicity issues, they give rise to various drawbacks. Namely, in the case of incorporating PNCs into macroscale polymeric matrices the induction of spectral shifts and reduction in the luminescence efficiency substantially affect their use in photonic device applications.22 On the other hand, the embedment of perovskite structures within transparent inorganic oxide glasses seems to address these issues, but it often requires costly high-temperature melting synthesis routes, often exceeding 800 °C, which are disadvantageous towards large scale production. More importantly, to date, the formation of composite perovskite glasses is based on mixing the perovskite precursors together with the chemical powders of the glass prior to melting, and then, heating all materials together at the desired melting and post-melting annealing temperatures.23–32 In this manner, the so-formed PNCs are distributed randomly throughout the entire glass network, i.e. without offering any possible way of controlling the position, growth, and morphology of the embedded structures,23–31 or requiring the employment of femtosecond laser irradiation for controlled crystal growth within the glass.32 Remarkably though, the importance of confined crystal growth for perovskite-based optoelectronic devices has been emphasized recently.34,35 Namely, the authors of these studies demonstrate the feasibility of laser-induced growth methods towards developing perovskite nanostructures with controlled size and location. Nevertheless, although stable upon formation, the reported laser-induced perovskite structures are located on the substrate surface, and thus, are vulnerable to the well-known negative effects of humidity exposure,16,18 also the lead toxicity concerns remain unresolved.
In this study, we demonstrate for the first time the development of composite perovskite glasses, where all-inorganic lead halide perovskite patterns are encapsulated within phosphate glass by using a post-glass quenching method. Based on this approach, pre-synthesized PNCs are incorporated into the glass matrix following the initial glass quenching. This novel approach introduces a significant advancement in enabling the confined spatial positioning of luminescent perovskite domains within a transparent medium in the visible, exclusively by means of glass melting procedures, i.e. without the need of additional means. Furthermore, the described post-glass quenching encapsulation process is performed at a relatively low temperature that does not exceed 160 °C, and thus, is cost-effective while diminishing structural degradation of the PNCs. In particular, all-inorganic lead halide PNCs are prepared following a previously described room temperature co-precipitation method,6 while silver metaphosphate glass (AgPO3) is employed as the hosting material. In recent years, phosphate-based glasses have been used frequently as the platform for many interesting photonic features such as second harmonic generation,36,37 lasing,38 and optical fiber applications.39,40 In this work, the selection of AgPO3 glass is encouraged mainly due to its soft nature, low glass transition temperature (Tg) of 192 °C, property of being transparent in most of the visible region, easy preparation, and non-toxic nature.37,41,42
The composite perovskite-glasses developed here exhibit impressive photoluminescence (PL) stability over the tested period of two months, when compared to the corresponding PL of the air-exposed PNCs. More importantly, using a simple and fast cw-laser processing technique, we demonstrate the formation of permanent perovskite micropatterns. The facile formation of highly luminescent perovskite patterns presented here enables the exploitation of the developed composite glasses in a broad range of photonic and optoelectronic applications, while simultaneously resolving the perovskite stability issues and environmental concerns of lead.
Two types of composite perovskite AgPO3 glasses are prepared. In the first type, isles, comprising clusters of PNCs embedded within a confined location inside the glass matrix (clustered perovskite samples, PvG-C), whereas in the second type, the incorporation of a thin perovskite layer beneath the glass surface is realized (perovskite-layer samples, PvG-L). As shown schematically in Fig. 1, the development of PvG-L from the as-prepared perovskite/toluene solution (50 mg ml−1) is based on a simple drop casting method. In particular, the AgPO3 glass substrate was positioned on a silicon wafer while a heating plate was employed to maintain a temperature of 70 °C. Several drops of the perovskite/toluene solution were added, while allowing 20 s intervals between each addition in order to ensure the gentle vaporization of the toluene solvent and the formation of the solid perovskite precipitate. Following this, the temperature was increased to 160 °C for 5 minutes, i.e. 32 °C below the glass transition temperature of the phosphate glass. At this temperature the AgPO3 glass gains viscosity and allows the smooth incorporation of the perovskite layer within the glass matrix. After the perovskite layer immersion was completed, the silicon wafer with the PvG-L sample was instantly removed from the heating plate and left to cool down to room temperature. Upon this post-annealing quenching the composite PvG-L glass solidifies again, while it can be readily retrieved from the silicon wafer.
Rather differently, the synthesis route of the composite PvG-C is based on encapsulating dry powder of PNCs within phosphate glass. In this case, the perovskite/toluene solution is dried under ambient conditions in order to yield PNCs in the form of powder. Then, around 50 mg of the extracted dry powder is positioned at the middle of the surface of the AgPO3 substrate and the sample is heated at 160 °C for 5 minutes (Fig. 1). This allows the total immersion of the perovskite powder within the glass, while the upper surface of the sample is splat-quenched and regenerates readily after the incorporation process is over. Immediately after this the composite sample is removed from the hot plate and allowed to cool down to room temperature.
For the room temperature photoluminescence measurements a previously described micro-photoluminescence (μ-PL) setup operating in the backscattering geometry was employed.44 The employed excitation source is a diode-pumped solid-state (DPSS) laser, Nd:YAG 473 nm (photon energy: 2.62 eV). The laser beam passes through a short pass (SP) filter (475 nm) in order to reduce noise at higher wavelengths, while the laser power is controlled via a neutral density (ND) filter. A 70:30 beam splitter is used to reflect and drive the beam to the 50× objective lens, which focuses the spot size down to ∼1 μm for sample excitation. The position of the PvG glasses was controlled with a XYZ mechanical translation stage. Following the excitation, the emitted PL signal passes through a Raman edge filter (473 nm) for the elimination of laser emission. An iHR-320 spectrometer (Horiba Scientific/Jobin Yvon Technology) was used to record the PL spectra.
In order to reveal information about the nature of the encapsulated PNCs we consider transmission electron microscopy (TEM) results and XRD analysis. Fig. 3a presents a typical TEM image of the as-prepared PNCs, whereas Fig. 3b depicts the corresponding profile after incorporation to form the PvG-L glass. Inspection of TEM images reveals that the initial shape of the nanohexagons (Fig. 3a) is partially retained following encapsulation into the glass matrix (Fig. 3b). Notably, TEM profiles of the embedded nanocrystals within PvG-L and PvG-C samples were found to be indistinguishable, suggesting that the encapsulation protocol had no effect on the nature of the incorporated particles. The obtained white large spots on the particles are attributed to the interaction of the HRTEM electron beam with the perovskite nanohexagons. In fact, soon after the HRTEM beam irradiation starts the nanohexagons are affected until they are totally damaged. Such hollow regions as a result of beam damage have also been reported by other groups for similar nanomaterials.45,46 Moreover, the corresponding XRD patterns in Fig. 3c show that the as-synthesized PNCs consist mainly of the rhombohedral Cs4PbBr6 phase (01-073-2478), while a secondary and strongly photoluminescent orthorhombic CsPbBr3 phase (ICSD 97851) is also present.47–49 The presence of both phases also within the composite PvG-L glass is confirmed by the corresponding XRD profile, as depicted in Fig. 3c. In particular, the XRD pattern of the PvG-L glass retains both characteristic peaks at 12.6° and 14.8° attributed to the (110) and (100) facets of Cs4PbBr6 and CsPbBr3 phases, respectively.47–49 However, due to the thermal annealing induced during the incorporation process, some additional XRD peaks are observed for the nanocrystals embedded into the PvG-L sample, possibly attributed to the inevitably newly created phases upon heating. This could also plausibly explain the darker spots that appeared within the embedded nanohexagons, as shown in the TEM image of Fig. 3b.
Raman spectroscopy was employed to probe any structural modifications of the phosphate glass network, upon perovskite encapsulation. Fig. 4a presents the room temperature Raman spectrum of the pristine AgPO3 glass, and the corresponding spectra of the two composite glasses PvG-L and PvG-C, respectively. It is well known that the metaphosphate network of the AgPO3 glass consists mainly of chains, which are formed by connected phosphate tetrahedral units with bridging and non-bridging (terminal) oxygen atoms.41,42,50,51 The two main Raman features shown in Fig. 4a are distinct signatures of such structural units. In particular, the key band at ca. 1144 cm−1 is attributed to the symmetric stretching vibration of terminal PO2− groups, vs(PO2−), whereas the broader band at ca. 675 cm−1 originates from the symmetric stretching movement of P–O–P bridges within the phosphate backbone, vs(P–O–P). The relative intensities of these two Raman bands provide direct evidence in the case wherein the population of terminal or bridging phosphate species changes at the expense of the other upon the incorporation of PNCs. As for instance, if oxygen was to be detached from the phosphate network, the population of P–O–P bridges would increase relative to that of terminal PO2− groups. The spectra of Fig. 4a are normalized at the strongest band at 1144 cm−1 in order to facilitate comparison. Notably, inspection of Fig. 4a reveals no significant change in the relative intensities of the Raman features of the PvG-L and PvG-C samples, when compared to the corresponding spectrum of pristine AgPO3 glass. This observation implies that in both types of composite glasses, the encapsulation of perovskite crystals has no effect on the connectivity of the phosphate glass network, and thus, no direct incorporation of additional oxygen from the phosphate network to the perovskite lattice occurs. This holds good despite the formation of newly created PNC phases during encapsulation heating. However, in the case of the PvG-L sample where the perovskite layer is widely spread near the surface, it appears to give rise to Raman activity below 250 cm−1, a region where Raman modes have been previously attributed to structural vibrations of lead-containing all-inorganic perovskite crystals.52,53
Although the incorporation of PNCs does not cause any noticeable structural modification to the phosphate network, it definitely alters the optical properties of the composite glasses. Fig. 4b shows the respective optical absorbance spectra. As expected from previous studies, the pristine binary AgPO3 glass is almost transparent, down to 380 nm.37 Rather differently, both PvG-L and PvG-C samples exhibit the characteristic absorption shoulder of the orthorhombic CsPbBr3 phase at around 510 nm (Fig. S3†).47–49 Notably, a more pronounced absorption is observed for the latter sample where the considerably thicker perovskite core is present. Both composite glasses also exhibit absorption profiles in the vicinity of 440 nm that are attributed to the formation of metallic silver nanoparticles due to the thermal annealing that occurred during encapsulation.37 It is also shown that the integration of perovskite crystals pushes the absorbance threshold of the composite glasses to slightly lower wavelengths of around 350 nm, due to the enhancement of light absorption within the glass matrix. Nevertheless, the characteristic low-wavelength absorption feature of the Cs4PbBr6 species at around 315 nm is not observed for any of the two composite glasses,47–49 as the binary AgPO3 phosphate glass is not transparent in that wavelength.
Remarkably, the inclusion of PNCs within the phosphate glass induces impressive photoluminescence properties. Fig. 5 presents typical photoluminescence (PL) spectra of the two composite glasses PvG-L and PvG-C. Both composite glasses exhibit the characteristic green PL feature of orthorhombic CsPbBr3 nanocrystals at around 525 nm.47–49 In addition, Video S1† illustrates the obtained transition of the green light photoluminescence upon moving the blue laser excitation source of the μ-PL setup inside and outside the confined perovskite isle of the PvG-C sample. It is emphasized that such a PL profile is totally absent from the spectrum of the pristine AgPO3 glass, as well as when a region outside the perovskite core of the PvG-C sample is excited (Video S1†). Fig. 5 also presents the corresponding PL spectra of an air-exposed perovskite film that is formed on the surface of AgPO3 glass by the same drop-casting method that was used for the synthesis of the composite PvG-L sample. The PL profile of the air-exposed perovskite dry powder is also included in the same figure. A slight blue shift of a few nm is noted for the PL signature of the PvG-L glass when compared to the reference film, suggesting a possible decrease in the size of the PNCs upon vitrification,47 possibly induced by the compression from the phosphate glass network. Impressively though, the perovskite layer encapsulation within the phosphate glass matrix turns out to be of excellent photoluminescence efficiency, as it causes negligible losses in the PL intensity. Rather differently, in the case of PvG-C glass the encapsulation of the perovskite dry powder induces a PL loss of around 40% (Fig. 5). Since the perovskite isle is entirely beneath the glass surface, the obtained luminescence losses might be attributed to scattering effects induced by the silver phosphate glass matrix that completely covers the confined luminescent isle.37
An even more remarkable feature of the developed PvG-L and PvG-C glasses lies in their impressive PL stability over time. Fig. 6a depicts the indicative PL spectra of the air-exposed reference film over a period of two months, whereas Fig. 6b presents the corresponding spectra of the PvG-L glass sample. The inset of Fig. 6b also shows a photograph of the green PL of the PvG-L sample captured at day one, while some scattering of the blue excitation laser is also visible. The actual PL intensities of the perovskite film and the PvG-L glass as a function of days of exposure to air are plotted in Fig. 6c, while maintaining all spectra acquisition conditions identical. The corresponding peak PL intensity, in terms of percentage with respect to the maximum intensity recorded on the first day, is plotted in Fig. 6d. Inspection of the latter two figures reveals that during the period of two months the PL intensity of the PvG-L sample exhibits only minimal losses of 8%, in contrast to the air-exposed film that exhibits significant losses of 96%. Even within the first month, the PvG-L glass retains around 97% of the its initial PL intensity value, whereas the air-exposed film exhibits losses of 55% within the same period. The impressive PL stability, together with the aforementioned efficiency of the encapsulation process, demonstrate the suitability of the proposed fabrication method towards the development of highly luminescent and stable encapsulated perovskite patterns within PvG-L samples.
Along similar lines, the PL stability of the PvG-C glass has been investigated. Fig. 7a shows the PL spectra of the air-exposed dry perovskite powder at various time intervals, whereas the corresponding indicative spectra of the PvG-C glass are shown in Fig. 7b. Fig. 7c and d present respectively the variation of PL intensity over time and the corresponding percentage with respect to the initial intensity recorded on the first day. Inspection of Fig. 7d reveals that the PL intensity of the air-exposed dry powder drops to 23% within two months, whereas the corresponding intensity of the PvG-C sample is maintained at 70% of the initial value within the same period. These findings imply, that despite being a less efficient inclusion method when compared to that for PvG-L, the core encapsulation still offers an important shield over the degradation of perovskite crystals upon exposure to ambient air.
As a final part of this study, we demonstrate the development of encapsulated luminescent perovskite micropatterns within the fabricated composite glasses by means of a simple cw laser patterning process. Fig. 8a presents a fluorescence microscopy image of the PvG-C glass within the confined location of the incorporated perovskite core, whereas Fig. 8b depicts an indicative photograph of the PvG-L glass before patterning. Notably, both perovskite glasses are highly luminescent under the UV lamp irradiation. However, among the two composite samples, the PvG-L glass is more suitable for laser inscription of patterns on the thin luminescent layer, instead of attempting to create patterns on the considerably thicker perovskite core.
Fig. 8c and d show fluorescence photographs of micro-dotted optical textures within PvG-L glasses after performing laser patterning with a cw, 455 nm laser. The adopted irradiation pattern is represented schematically in Fig. S4.† Namely, the laser beam initially moves horizontally, followed by vertical movement after completing dots of the first line. The distance between the dots is defined by the horizontal and vertical separation steps (Fig. S4†), while the diameter of the dots is readily controlled by the spot size of the employed laser. In this manner, the shape and size of the pattern can be varied in accordance to the irradiation pathway and beam profile. In particular, Fig. 8c presents a luminescent micro-dotted optical texture with a horizontal separation step of 400 μm and a vertical separation step of 500 μm, while Fig. 8d depicts a similar pattern with smaller separation steps of 300 μm and 400 μm, respectively. In both patterns, the dark dots are created upon laser irradiation with a power of 1.6 W for 100 ms, while the beam diameter is 250 μm. These irradiation conditions resulted in the confined degradation of the incorporated PNCs. Consequently, the green luminescence within the irradiated spots disappears. In contrast, the non-irradiated region between the dark dots remains highly luminescent. Fig. S5† presents the PL spectra of PvG-L glass before and after laser processing, where the latter PL spectrum is recorded for a non-irradiated region between the dark dots (Fig. 8c). The negligible differences between the two spectra shown in Fig. S5† strongly imply that the embedded PNCs outside the irradiation spots maintain their structure and photoluminescence properties. Finally, it is emphasized that based on the aforementioned procedures, the developed patterning process requires only widely available, compact, and low-cost cw laser sources, while being extremely fast and efficient.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/d0nr03254a |
This journal is © The Royal Society of Chemistry 2020 |