Encapsulation of CsPbBr3 Perovskite Nanocrystals in SiO2

(1)

Encapsulation of CsPbBr

3

Perovskite Nanocrystals in

SiO

₂

Report of bachelor project B´

eta-Gamma, major Physics (12EC)

Conducted between April 3 and June 27, 2017

Susan Rigter

10563199

Supervisors: Dr. Leyre G´omez Navascu´es & Prof. Dr. Tom Gregorkiewicz Second assessor: Prof. Dr. Wim Sinke

Institute of Physics, van der Waals-Zeeman Instituut

Faculteit der Natuurwetenschappen, Wiskunde en Informatica, UvA

June 27, 2017

Abstract

Cesium lead bromide perovskite nanocrystals (NCs) have lately gained a lot of atten-tion for their superior optical properties, which makes them attractive for optoelectronic applications. This is a report of an attempt at encapsulating CsPbBr3 perovskite NCs in

SiO2 nanospheres, to overcome their labile characteristics under humidity and high

tem-peratures. The perovskite NCs were not encapsulated in SiO2 nanospheres, but embedded

in larger structures. During the process, the perovskite NCs suffered; their emission and photoluminescence quantum yield were reduced drastically. Nonetheless, the final product is water stable for a period of at least two weeks, their emission is more stable up to temper-atures of 100°C than unencapsulated CsPbBr3perovskite NCs, and the photoluminescence

lifetimes is more stable over the whole measured range of temperature, up to 200 °C. This is a good prospect for further research into encapsulating cesium lead bromide perovskite NCs in SiO2nanospheres.

(2)

Dutch summary

Onlangs is er veel onderzoek gedaan naar een bepaald soort kristal: cesium lood bro-mide perovskiet nanokristallen; CsPbBr3. Het gaat om zeer kleine kristallen, in het

gebied van 4 tot 15 nanometer. Dit is een materiaal waarvan in de toekomst misschien zonnecellen, LED-lampen en inkjet inkt van kunnen worden gemaakt, omdat het goede opto-electronische eigenschappen heeft, en de emissie samen met soortgelijke kristallen het hele zichtbare spectrum dekt. Tot nu toe is er echter ´e´en groot probleem: het mate-riaal kan niet tegen water en tegen hoge temperaturen. Dat is wel nodig op het moment dat het gebruikt wordt voor de voorgestelde toepassingen. Daarom moet er onderzoek worden gedaan naar manieren om deze problemen op te lossen.

E´en mogelijkheid hiervoor is het omhullen van de nanokristallen in een ander materi-aal, dat wel tegen deze externe factoren kan. Eerder is dit al uitgevoerd met stearinezuur, wat de perovskieten beschermde tegen water, maar niet tegen hoge temperaturen. Een soortgelijk materiaal, CH3NH3PbBr3, is eerder omhuld met silica, en dit leverde een

prod-uct op dat waterstabieler was dan de originele kristallen. Daarom is nu geprobeerd om de kristallen te omhullen met bollen van silica van enkele nanometers groot, een materiaal wat beter tegen hoge temperaturen kan. Een groot deel van de perovskieten overleefde dit proces niet, en de perovskieten blijken zich niet in bollen, maar in grotere structuren sil-ica te bevinden. Toch zorgen deze silsil-icastructuren ervoor dat de perovskieten beter tegen water en hitte kunnen - ze blijven stabiel voor minimaal twee weken wanneer ze opgelost zijn in water, terwijl hetzelfde materiaal zonder silica dat slechts enkele minuten overleeft, en ze zijn stabieler dan de perovskieten zonder silica tot in ieder geval een temperatuur van 100 °C. Dit is een goed vooruitzicht voor verder onderzoek naar omhulling in silica van dit materiaal.

(3)

1 Introduction

Recently, perovskite nanocrystals (NCs) have received a lot of attention for the possibility of being the future of optoelectronic materials, because of their advantages (1, 2). A class of perovskite NCs that is especially interesting, are the all-inorganic cesium lead halide perovskite NCs. These perovskite NCs combine the advantages described above with quantum size effects of quantum dots (QDs). This makes this material have emissions which are tunable over the whole visible spectrum, even higher photoluminescence (PL) quantum yield, short PL lifetimes, as well as narrow emission bands (3–5).

Many suggestions have been made for the possibilities of CsPbBr3, such as LEDs,

photodetecting, solar cells and lasers (6–12). However, these perovskite NCs have some disadvantages: they don’t behave well when exposed to humidity or heat (13–15) making them unsuitable for these applications, unless a solution is found.

It is therefore important to find ways to protect the all-inorganic perovskite NCs against these external influences. Previously, this has been done by using solid-lipid structures (13). Although this did lead to water-stable perovskite NCs within these structures, this has a major disadvantage: the encapsulation material, mainly stearic acid, is only stable at room temperature. This is a report of an attempt at encapsulating CsPbBr3 in SiO2 nanospheres, to make them water-stable and temperature stable.

2 Theory

To be able to understand and recognize stability of encapsulated perovskite NCs, one must understand the properties of the unencapsulated perovskite NCs. In the introduction it was stated that they have several advantages: high quantum yields, narrow emission bands tunable over the whole visible spectrum, short photoluminescence lifetimes and quantum size effects. Below, these advantages will be elaborated. Also, one needs to understand what happens to the NCs when they are exposed to humidity and heat, and what could be done to protect them from these external factors. This, too, will be elaborated below.

2.1 Perovskite structure

Perovskites are cubic crystals, built up from two cations and three anions, such that they have the chemical formula ABX3. The schematic structure is visible in figure 1.

Perovskites are built up as follows: the A-cation is located at (0,0,0), the B-cation at (½,½,½) and the X-anions at (½,½,0). It is necessary for the A-cation to be larger than the B-cation, otherwise the structure of the perovskites is distorted (16).

These crystals can be both insulators, conductors and semiconductors, depending on whether the amount of electrons needed to fill the outer shell of the three anions is equal to the amount of valence electrons of the cations combined. If this is the case, the perovskite is an insulator. If the amount of electrons of the cations is less than that, it is a conductor,

(5)

Figure 1: Schematic representation of the orthorhombic CsPbBr3 structure (3).

and if the amount of valence electrons is more than that, it is a semiconductor (17). In the case of CsPbBr3, three electrons are needed to complete the 6p-orbitals of all three bromine

atoms, and there are 5 valence electrons available from cesium and lead. Therefore, it is a semiconductor, which makes it suitable for the proposed optoelectronic applications.

perovskite NCs that have lead as the B cation, exist as hybrid organic-inorganic and all-inorganic structures. In the hybrid organic-all-inorganic perovskite NCs, alkyl ammonium or formamidinium take the place of the A cation. This place is usually taken by cesium in the all-inorganic perovskite NCs, which have a higher temperature stability than the hybrid perovskite NCs. At high temperatures, the organic group will be lost by sublimation, which does not occur with cesium (12).

2.2 Emission of CsPbX

3

Cesium lead halide perovskite NCs have narrow photoluminescence (PL) peaks. How narrow a peak is, can be measured as full width at half maximum (FWHM), which is the complete width at half the heigth of the peak. The PL peaks are tunable by varying the size of the NCs, because when a perovskite NCsgrows larger, their quantum size effects decrease and the peak is redshifted (18). Because of this, the size distribution of mea-sured perovskite NCs can be derived from the FWHM. The bigger the FWHM, the more variation in size of the particles. By also using perovskite NCs with X –– (Cl/Br), (Br/I), the emission of CsPbX3 is tunable over the whole visible spectrum. The emissions of the

products of this tuning are visible in figure 2 (3, 13).

Although the fast anion exchange of these NCs (within 10-20 minutes) helps to create NCs with PL peaks between those of CsPbCl3, CsPbBr3 and CsPbI3, the same process

poses a difficulty when using them to create optoelectronic devices because they would need to maintain their original emission characteristics (13).

(6)

Figure 2: Emission of colloidal CsPbX3 NCs (X = Cl, Br, I and mixtures) (a) under a

UV-lamp and (b) their spectra (3).

2.3 Photoluminescence quantum yield

The photoluminescence quantum yield (PLQY) is defined as the ratio of emitted photons and absorbed photons, expressed in percentages:

P LQY = Ne Na

× 100% (1)

where Ne is the number of emitted photons, and Na the number of absorbed photons by

the sample, which can be acquired by integrating the emission and absorption expressed in photons (19). High quantum yields are an advantage for multiple applications, because this indicates that a substantial amount of excitation photon energy is used to excite electrons to a higher energy level, as opposed to being converted to kinetic energy or heat. The energy of these excited electrons can then for example be used as electrical energy, or be converted back into an emission photon. A high PLQY also shows there are few defects on the surface of the material, such as trap states (20).

The PLQY of CsPbX3 perovskite NCs is in the range of 50-90% (3). This indicates

that this material has high potentials, compared to silicon, the material of which most solar cells are made, which usually has quantum yields around 30% (21).

(7)

2.4 Photoluminescence lifetime

The photoluminescence lifetime expresses the time it takes for an exciton to recombine (19). The lifetime of a material can be calculated from the photoluminescence decay curve, defined as I(t) = A1e − t τ1 _{+ A}₂_e− t τ2 _{+ A}₃_e− t τ3_. ₍₂₎

Where I is intensity, A1,2,3 the relative strength of the components. This equation

shows how the intensity of the photoluminescence decays over time. From this, the average lifetime of the material can be obtained from

τav =

(A1 × τ12) + (A2× τ22) + (A3× τ32)

(A1 × τ1) + (A2× τ2) + (A3× τ3)

. (3)

The reason that this is a triexponential decay curve, is that the NCs are distributed in size, rather than all the same size (14). According to Protescu et al. (3), the lifetimes of CsPbX3 NCs are in the range of 1-29ns, with smaller lifetimes for NCs with bigger

bandgaps, thus for smaller perovskites. The decay curves of their samples are visualized in figure 3, where one sees a rapid PL decay of CsPbBr3.

Figure 3: PL decay curves of colloidal CsPbX3 NCs (X = Cl, Br, I and mixtures). The

lifetimes of these NCs vary between 1-29ns. The bigger the bandgap of the NC, the smaller the lifetime. The green curve represents CsPbBr3 (3).

Short lifetimes are advantageous for some optoelectronic devices and disadvantageous to others. For LEDs it is an advantage, because a short lifetime means a quick reaction

(8)

time (22). However, for photovoltaics, a short lifetime decreases the distance an electron travels before being recombined, thus reducing the chance of producing an electric current in a connected current loop (23).

2.5 Quantum size effects

Quantum size effects are the cause of differences between optical properties of quantum dots and bulk versions of the same material. As suggested above, it is possible to control the energy level of the band gap by controlling the size of the NC. In quantum dots, the electrons and holes are confined in three dimensions, which is the cause of an energy and motion quantization, which makes the QDs different from their bulk counterparts, because in bulk, electrons can have any energy above the band gap energy (24). Quantum confinement means that the possible movement of the exciton is equal to or smaller than its twice its bohr radius, d 6 2a∗B. When the possible movement is bigger than that, but

still comparable to twice the bohr radius, weak quantum confinement occurs. The bohr radius can be calculated by

a∗_B = (∞)

m∗ aB (4)

where (∞) is the dielectric constant of the material, m∗ the effective mass of the exciton and aBthe Bohr radius. When this situation occurs, the minimum kinetic energy increases

with an amount of δE ' ~ 2 2m∗_d2 x + ~ 2 2m∗_d2 y + ~ 2 2m∗_d2 z . (5)

Decreasing the size of the QD therefore increases the minimum kinetic energy, which causes a blueshift in the emitted photon when the exciton is recombined (17). This quantum confinement induces several advantageous properties.

The cesium lead halide perovskites, as synthesized by Protescu et al. (3), vary in size from 4 to 15nm in all three directions. The bohr radius of CsPbBr is ∼ 3.5nm, which shows that quantum confinement occurs within these NCs (25). This explains the higher emission wavelength of bigger inorganic perovskite NCs, because when the perovskite grows larger, the quantum confinement reduces, which means the minimum energy increase is reduced and the emission has a lower energy.

2.6 Exposure to heat

Some of the proposed applications, such as LEDs, require the perovskites to be stable up to temperatures of around 200 °C. This is not the case. When thin film samples of CsPbBr3 perovskite NCs are exposed to high temperatures, the ligands on the surface of

the perovskite NCs are removed (15). The removal of the ligands makes it possible for the perovskites to disintegrate and creates the possibility of existence of trap states on the perovskite surfaces. Trap states exclude electrons from the processes of energy exchange.

(9)

This is a disadvantage for the proposed applications for cesium lead halide perovskite NCs, because it affects the optoelectronic properties of the perovskites in multiple ways. Up to temperatures of around 30 °C, the PL intensity decreases slightly (14). When heated to 50°C, the PLQY and PL lifetime show a slight increase. At higher temperatures than this, rapid drops in intensity, PL lifetimes and PL quantum yield are observed. (14, 15). Another thing that is observed are red shifts of the emission of samples after heating them up to 150°C, indicating a decrease of quantum confinement. However, when heated to 200 °C, the emission is blueshifted. This is not yet fully understood (15).

2.7 Exposure to water

When exposed to polar solvents such as water, colloidal perovskite NCs lose their struc-tural integrity, due to their ionic structure. In figure 4 it is visible that the PL vanishes within 5 to 15 minutes after coming into contact with Milli-Q water, indicating that the perovskite NCs are disintegrated (13).

Figure 4: Normalized PL intensity as a function of time for unencapsulated CsPbBr3

per-ovskite NCs (red) and CsPbBr3 encapsulated in solid lipid nanoparticles (black), dispersed

in Milli-Q water (13).

2.8 Previous encapsulations

To overcome the limitations of different nanocrystals, several encapsulations have been explored. The most relevant of these are listed below.

CsPbX3 in solid lipid nanoparticles

Both CsPbBr3 and CsPb(Br0.2I0.8)3 have succesfully been encapsulated in solid lipid

(10)

in-tensity of CsPbBr3 perovskite NCs in SLNs is visible. The encapsulated perovskite NCs

are water stable for over two months and anion exchange was fully prevented. However, the stearic acid is stable up to temperatures of 70 °C, and this encapsulation does not overcome the difficulty of temperature instability.

Organic-inorganic perovskite NCs in SiO2

Huang et al. (26) demonstrated that organic-inorganic CH3NH3PbBr3 perovskite NCs can

be encapsulated in layers of SiO2. Usually, for encapsulations in silica, ’harsh’ reagents

like water or alcohols are required, but since that would disintegrate organic-inorganic perovskites, another method had to be found. The successful method Huang et al. used, did not cause a decrease in photoluminescence. Upon investigation, it was observed that some perovskite NCs were located on the surface of SiO2 nanospheres, but most of the

perovskite NCs were embedded inside them.

3 Methods

The previously mentioned encapsulations of CsPbX3perovskites in SLNs and CH3NH3PbBr3

perovskites in SiO2 have lead us to this attempt at encapsulating CsPbBr3 perovskites in

SiO2 nanospheres. Below, the methods of synthesis, encapsulation and characterization

are elucidated.

3.1 Synthesis and encapsulation

Chemicals

Cesium carbonate (Cs2CO3, 99.9%, Sigma-Aldrich), octadecene (ODE, 90%, Sigma-Aldrich),

oleic acid (OA, 90%, Sigma-Aldrich), oleylamine (OLA, 80-90%, Acros), lead(II) bromide (PbBr2, 98%, Sigma-Aldrich, toluene (ACS reagent > 99.5%, Sigma-Aldrich), tetramethyl

orthosilicate (TMOS, 98%, Sigma-Aldrich), methyl acetate (99.5%, Sigma-Aldrich). No further purifications were performed, except for the reported.

Perovskite nanocrystal synthesis

We synthesized the perovskite NCs using the method described by Protescu et al. (3). Cs-oleate was prepared by mixing 0.814 g Cs2CO3 and 40 mL ODE and 2.5 mL OA.

This mixture heated to 145 °C under a N2 atmosphere until the cesium carbonate was

completely dissolved. 0.7000g of PbBr2 and 30mL ODE were heated to dry at 120 °C

under N2 atmosphere, after which 5mL dried OA and 5mL dried OLA were added to

the reaction flask and temperature was raised to 160°C. After dissolution of PbBr2, the

warmed Cs-oleate solution was injected. Within seconds after this, the suspension was cooled down using an ice bath. The NCs were purified by centrifugation and redispersed

(11)

in toluene. The final product was then stored for further use. The concentration of the product is estimated to be 6.6mg perovskite NCs per mL.

Encapsulation of NCs in SiO2

We performed the encapsulation by means of the SiO2 generation method based on the

high hydrolysis rate of tetramethyl oscillate (TMOS) in analytical grade toluene, as de-scribed by Huang et al. to encapsulate hybrid organic-inorganic perovskite NCs (26). We dried multiple ratios of reactants. 1 mL of the original product was added to 6.6 mL toluene and varying amounts of TMOS and stirred for 24 hours. Afterwards, I added methyl acetate to the sample in a ratio of 1:1 to aid precipitation during the purification. The samples were then centrifuged and redispersed in toluene.

This is a report of measurements I have done on three different samples, varying only in amount of TMOS used during the synthesis. The difference between the samples, as well as the percentage that did not disintegrate during the encapsulation process, is located in table 1.

Table 1: The different samples, their TMOS-concentrations, percentages surviving the encapsulation process and percentages of perovskite NCs that appear to be encapsulated.

Sample Amount of original sample TMOS % survived % encapsulated 1 1000µL 10µL 34.3% 0.043%

2 1000µL 333µL 21.45% 0.042% 3 1000µL 3333µL 13.32% 0.142%

I calculated the percentages of perovskite NCs that survived the encapsulation process by comparing the emission of different samples in toluene to an unencapsulated sample of the same concentration. It is clear that the perovskite NCs suffer severely during the encapsulation. It is not surprising that the smaller the amount of TMOS added, the more perovskites survived the encapsulation process. In TMOS, there are trace amounts of water, which are likely to be the cause of the perovskites that disintegrated.

I calculated the percentages of encapsulated perovskite NCs by comparing the emission of the same concentrations of sample in toluene and in water. Since water causes the unencapsulated perovskite NCs to disintegrate, these won’t emit when dispersed in water, but they do emit when dispersed in toluene. Something that has to be noted is that some perovskites might be on the surface of the silica structures, similar to what was reported for the encapsulation of organic-inorganic hybrid perovskites in silica (26). Also, perovskite NCs might be inside silica structures with small defects, causing them not to be protected from water.

There are two other methods of encapsulation which we tried unsuccessfully. One of these was close to the synthesis of Huang et al. (26), with the TMOS replaced by 20 µL tetraethylorthosilicate (TEOS), and 4 µL was added to aid the reaction. For the second

(12)

alternative, the method of Koole et al. was used (27), where the ammonia was replaced with dimethylamine. This was done, because the used ammonia is dispersed in water, which would have lead the CsPbBr3 perovskite NCs to disintegrate. Various

concen-trations of original sample were used. When conducting measurements on the resulting samples in colloid, no emission was detected, indicating no encapsulated perovskites. No further investigation was pursued.

I prepared colloidal samples from the samples in table 1 in quartz cuvettes. The sample is colloidally instable, so they were redispersed using sonication before every measurement. I created thin film samples from the samples in table 1 by method of dropcasting on quartz substrates. The reference samples were obtained using the same method and concentrations of toluene and TMOS, without addition of perovskites and performing the cleaning process, and I measured them in quartz cuvettes.

3.2 Characterization

For the measurements of steady state photoluminescence at several excitation wave-lengths, I used a Jobin Yvon FluoroLog spectrofluorometer (Horiba), equipped with a 450W xenon lamp, coupled to a monochromator was used. To determine the photolumi-nescence quantum yield, I used an integrating sphere, connected with a 150W xenon lamp coupled to a spectrometer (Solar, MSA-130). The integrating sphere is a hollow sphere of which the inside is coated with a material that scatters the light diffusely. Because of this integrating sphere, it is not necessary to take into account the angular dependence of the emission (28). The light in the integrating sphere is detected by a CCD (Hama-matsu, S10141-1108S), coupled to another spectrometer (Solar, M266). I measured the optical density of the samples using a LAMBDA 950 UV/Vis/NIR spectrophotometer (PerkinElmer), which was equipped with an integrating sphere for the thin film sam-ples. To investigate the photoluminesence lifetime, I did time resolved photoluminescence measurements using a Ti:sapphhire lasersystem (Chamelion Ultra, Coherent) with 140fs pulses at λ = 355nm. The emission was then detected using a monochromator (Newport CS260-02) coupled to a PMT (Hamamatsu). A high resolution Scanning Electron Micro-scope (SEM, FEI Verios 460) with a STEM detector was used to take Transmission and Scanning Transmission Electron Microscopy (TEM-STEM) images.

3.3 Heating

To investigate the behaviour of the encapsulated perovskite NCs after heating, I drop-casted samples 1 and 2 in water and the original sample on quartz substrates. For heating these to 50°C, I put the samples in an oven. To heat the samples to temperatures of 100 °C and higher, a silicon oil bath was used, in which I placed an Erlenmeyer containing the samples.

(13)

4 Results and discussion

4.1 Reference samples

(a) Emission of reference samples with an exci-tation wavelength of 360nm. The narrow peaks visible around 400nm are most probably due to Raman scattering. The broad peaks ranging from 375nm to 475nm can most probably be attributed to Stokes scattering.

(b) Absorption of reference samples between 450 and 600nm.

Figure 5: Emission (a) and absorption (b) of reference samples.

Figure 5a shows the emission measurements of the colloidal reference samples, as well as the emission measurements of the solvents on their own and empty sample carriers.

Although there are no compounds in the reference samples that emit light in the visible spectrum, narrow peaks are visible around 400nm and broad peaks are visible ranging from 375 to 475nm in figure 5a. These peaks are believed to be respectively the result of Raman scattering and Stokes scattering, which is the emission and absorption of optic phonons (17). Figure 5b shows the absorption of the reference samples. There are no peaks visible in the area where the emission of CsPbBr3 is expected to be.

4.2 Encapsulation in SiO

2

nanospheres

Like visible in figure 6, the encapsulation in silica nanospheres did not succeed, but there are perovskite NCs inside silica structures. This has lead to the choice to characterize these samples anyway, to research whether the silica structures protect the perovskite NCs at all against external humidity and heat, to investigate the prospect for encapsulation in actual SiO2 nanospheres.

(14)

Figure 6: a) TEM image of sample 2 in toluene. Cubes surrounded by a structure of a different material, presumably silica. b) TEM image of sample 3 in water. A structure in which the cubes are embedded is visible. c) TEM image of sample 1 in water. Cubelike shapes are embedded in another material. d) HAADF-STEM image of sample 2 in water. Cubes and other structures, presumably silica.

4.2.1 Emission and absorption

Figure 7 shows the emission of the different samples in toluene and water, for both the colloidal and thin film samples. Between the different samples, there are slight peak shifts of 6 10 nm. It is most clearly visible when comparing figure 7c and 7d that in sample 2 the smaller perovskite NCs appear to be encapsulated, and the bigger NCs left unencapsulated. This is not the case for the other two samples, where the peaks hardly shift between the samples in water and in toluene.

(15)

(a) The colloidal samples in toluene, integra-tion time of 1s.

(b) The colloidal samples in water, integration time of 30s.

(c) The thin film samples in toluene, integra-tion time of 0.1s.

(d) The thin film samples in water, integration time of 0.1s.

Figure 7: Emission of the different samples. The colloidal samples carry the same con-centrations of perovskite NCs. The same cannot be said about the thin film samples.

The FWHM of the samples in water is around 8nm bigger than the FWHM of the samples in toluene, both for thin film and colloidal samples. This indicates that the spread in perovskite NC sizes in the water samples is bigger than that of the ones in toluene.

In figure 8, one sees the absorption of both colloidal and thin film samples between 450nm and 600nm. For all thin film samples, the expected absorption is visible. This underlines the result from the emission measurements, that there are perovskite NCs inside silica structures, protected from water. For the colloidal samples this is not as clear, even for the samples in toluene, except for sample 1.

(16)

(a) Colloidal samples (b) Thin film samples

Figure 8: Absorption of colloidal samples 8a and thin film samples 8b from 450 to 600nm.

4.2.2 Stability over time

(a) Emission of colloidal samples over time. (b) Emission of thin film samples over time.

Figure 9: Stability over time of colloidal (a) and thin film (b) samples.

In figure 9, the emission over time of the colloidal and thin film samples is shown. The PL peaks of the different samples did not show any blue- or redshift tendencies, with only small differences of 6 4nm.

It is clear from figure 9a that the samples dispersed in water appear more stable than the ones in toluene, which is probably due to the trace amounts water destructurizing the perovskites left unencapsulated slowly in the samples dispersed in toluene, while this happens quickly for the samples dispersed in water.

(17)

The rather big increases in emission over time that are visible in figure 9b are assum-ingly because of differences in positioning of the samples in the setup, due to which the bundle of excitation light being directed onto the sample varied.

The full width at half maximum (FWHM) values of the peaks of the encapsulated samples, both in water and in toluene, only vary with values of 6 1nm, but the FWHM of the unencapsulated sample decreases with ∼ 3nm. This indicates an increased stability of the encapsulated samples over time.

4.2.3 Photoluminescence quantum yields

In tables 2 and 3, one sees the decreases in quantum yields for the samples in colloidal and thin film forms. The quantum yield of the colloidal unencapsulated sample agrees with data from theory, which indicated QYs between 50 − 90%. The QY of the thin film samples are a lot lower. This might be due to exposure to humidity during the evaporation of the solvent after dropcasting.

Table 2: The quantum yields of the different colloidal samples in toluene and water.

Sample QY in toluene QY in water 1 13.33% 0.13% 2 35.40% 0.17% 3 32.85% 0.34% Unencapsulated 64.52% –

Table 3: The quantum yields of the different thin film samples in toluene and in water.

Sample QY in toluene QY in water

1 2.12% 0.15%

2 0.28% 0.16%

3 0.14% 0.06%

Unencapsulated 7.01% –

The quantum yields severely decrease when the samples are dispersed in water, but the fact that it was still possible to measure the QY is positive: it means once again that there are some perovskite NCs being protected by silica structures around them. The decrease for the thin film samples is relatively smaller than the decrease of the colloidal samples. This corroborates with the assumption that the big difference between the colloidal and thin film samples in toluene is due to exposure to humidity during the process, since it appears only unencapsulated perovskite NCs have stopped emitting.

(18)

4.2.4 Photoluminescence lifetimes

From literature, photoluminescence lifetimes of 1-29 ns are expected. As visible in tables 4 and 5, all measured lifetimes except for colloidal sample 1 in toluene are within this window. The lifetimes of the thin film samples in toluene is higher than the same samples in colloidal form, but the lifetimes of the samples in water do not vary to the same extent. Since the lifetimes of bigger perovskite NCs are larger, this might indicate that unen-capsulated perovskite NCs in the toluene samples grew during the dropcasting process, though there is no supporting evidence of this in the emission measurements. Therefore, this increase in lifetimes of the thin film samples, compared to colloidal samples, remains unexplained.

Table 4: The PL lifetimes of the different colloidal samples in toluene and in water.

Sample lifetime in toluene (ns) lifetime in water (ns)

1 0.83 5.71

2 5.66 4.72

3 3.04 5.22

Unencapsulated 2.43 –

Table 5: The PL lifetimes of the different thin film samples in toluene and in water.

Sample lifetime in toluene (ns) lifetime in water (ns)

1 6.91 5.72

2 12.99 2.55

3 10.24 5.07

Unencapsulated 15.16 –

The notion that the lifetime of the samples in water, thus the samples in silica, is still within the expected window, is good news for the possibility of water stability of CsPbBr3

in silica nanospheres.

4.3 Heating

4.3.1 Emission

Figure 10 shows the PL peaks of three different samples after being exposed to higher temperatures. The emission of the unencapsulated sample in figure 10c drops rapidly as expected. More interesting are the PL peaks of the encapsulated samples. In figure 10a, the peak of sample 1 after heating to 50° C is nearly the same heigth as the original peak, which is not the case for the original sample. This shows that sample 1 is more stable than the original sample, up to temperatures of 50° C. Figure 10b does not show a peak

(19)

as high as the one of sample 1 at 50° C, but it does show a less radical emission decrease than the original sample up to 100° C, thus it appears more stable up to this temperature.

(a) Sample 1 (b) Sample 2

(c) Unencapsulated sample

Figure 10: Emission of different samples measured after heating to different temperatures.

Another thing that’s visible in 10b is that the peaks of sample 2 are wider than those of the other samples. Although this already occurs at room temperature, the FWHM increases further when the sample has been heated to higher temperatures. This is more clearly visible in 11. As mentioned before, this indicates a wider spread in size of the perovskite NCs, which indicates that the perovskites have had the chance to change in size, compared to earlier measurements. This might seem counterintuitive, but in the STEM image of the same sample (figure 6d) it looks like there are multiple perovskite NCs within the same silica structure, making it possible for them to fuse together.

(20)

Figure 11: The full width at half maximum (FWHM) of the PL peaks.

4.3.2 Photoluminescence quantum yield

Table 6 indicates the photoluminescence quantum yields for the samples at higher tem-peratures. The setup was not able to detect emission for samples 1 and 2 after they were heated to higher temperatures, and no quantum yields could be calculated. However, as expected from the emission measurements, it was possible to calculate the QY for samples 1 after heating to 50°C, and for sample 2 even after heating to 100 °C.

Table 6: PL quantum yields of the different samples after being exposed to heating. For the encapsulated samples heated to higher temperatures, the setup was not able to detect emission, so no quantum yield could be calculated.

Sample 50 °C 100 °C 150 °C 200 °C

1 1.83% - -

-2 1.54% 0.12% - -Unencapsulated 5.58% 3.44% 0.64% 0.37%

(21)

4.3.3 Photoluminescence lifetime

Figure 12 shows the decay in photoluminescence lifetime of the samples, after they are heated to temperatures of 50°C, 100 °C, 150 °C, and 200 °C. At a temperature of 150 °C, a clear drop in the PL lifetime of the unencapsulated sample is visible, as expected from literature, while the lifetime of the encapsulated samples stays roughly the same. This indicates that the encapsulation of the perovskite NCs in silica spheres has made them more stable when exposed to high temperatures.

Figure 12: Photoluminescence lifetimes of the samples, measured after heating.

5 Conclusion

From the previous, it is possible to draw a careful conclusion that it is useful to pursue the encapsulation of CsPbBr3 perovskites in silica nanospheres, because even the embedding

of the perovskite NCs in silica structures renders them more stable when exposed to water and to temperatures up to 100°C.

Something that could be improved during further research, is the creation and heating of thin film samples. If it is possible to do this under inert atmosphere, the perovskite

(22)

NCs would not suffer as much from the humidity in the surrounding air. Another thing that could be improved, are the PL lifetime measurements. In the used setup, the pulse time was in the same range as the measured lifetimes. It would be better to use a setup with a shorter pulse time.

Further research should be conducted to find a way to encapsulate these perovskite NCs in actual silica nanospheres. For this, one could try to find the optimal ratio of reactants for the encapsulation methods mentioned in this report. Sample 2 in this report seems to be the most fertile, since this sample appears to be the most stable when exposed to higher temperatures. This is a good starting point.

Another interesting prospect for the future is to try encapsulating other cesium lead halide perovskite NCs (CsPbCl3 and CsPbI3), to see whether encapsulation also renders

these more stable. If this works, investigation could be conducted into whether the anion exchange between the perovskites is prevented by the encapsulation. Keeping an eye on sustainability, research could be conducted into materials with the same optical properties and possibilities as cesium lead halide perovskite NCs, that don’t contain lead. For perovskites containing cesium, there are only a few options for this, as the cation replacing lead has to be smaller than cesium and needs to have at least three valence electrons.

6 Acknowledgements

I would like to thank everyone in the Tom Gregorkiewicz Group at the Van der Waals-Zeeman institute within the Institute of Physics, for all the help during this project, and also for making me feel very welcome in the group. Special thanks to Leyre G´omez Navascu´es, for being my supervisor during the project.

References

1. S. Kazim, et al., Angewandte Chemie International Edition 53, 2812 (2014).

2. T. C. Sum, N. Mathews, Energy & Environmental Science 7, 2518 (2014).

3. L. Protesescu, et al., Nano Letters 15, 3692 (2015).

4. A. Swarnkar, et al., Angewandte Chemie 127, 15644 (2015).

5. S. Sun, et al., ACS Nano 10, 3648 (2016).

6. S. Yakunin, et al., Nature Communications 6 (2015).

7. Y. Wang, et al., Advanced Materials 27, 7101 (2015).

(23)

9. F. Palazon, et al., Chemistry of Materials 28, 2902 (2016).

10. M. A. Green, et al., Nature Photonics 8, 506 (2014).

11. P. Ramasamy, et al., Chemical Communications 52, 2067 (2016).

12. M. Kulbak, et al., The journal of physical chemistry 6, 2452 (2015).

13. L. Gomez, et al., Nanoscale 9, 631 (2017).

14. J. Li, et al., RSC Advances 6, 78311 (2016).

15. F. Palazon, et al., Journal of Materials Chemistry C 4, 9179 (2016).

16. J. Hook, H. Hall, Solid state physics, vol. 2 (Wiley, 1991).

17. E. N. Economou, The Physics of Solids: Essentials and Beyond (Springer Science & Business Media, 2010).

18. J. Lin, et al., Nano Letters 16, 7198 (2016).

19. P. Peter Atkins, J. De Paula, Atkins’ Physical Chemistry, vol. 10 (OUP Oxford. ISBN, 2014).

20. A. M. Brouwer, Pure and Applied Chemistry 83, 2213 (2011).

21. D. Jurbergs, et al., Applied Physics Letters 88, 233116 (2006).

22. J. W. Jewett, R. A. Serway, Physics for Scientists and Engineers with Modern Physics (Cengage Learning EMEA, 2008).

23. X. Wang, Z. M. Wang, High-efficiency solar cells (Springer, 2013).

24. M. Fox, Quantum optics: an introduction, vol. 15 (OUP Oxford, 2006).

25. Z. Liang, et al., ACS Applied Materials & Interfaces 8, 28824 (2016).

26. S. Huang, et al., Journal of the American Chemical Society 138, 5749 (2016).

27. R. Koole, et al., Chemistry of Materials 20, 2503 (2008).

Encapsulation of CsPbBr3 Perovskite Nanocrystals in SiO2