Next Article in Journal
Effect of Preparation Method on ZrO2-Based Catalysts Performance for Isobutanol Synthesis from Syngas
Previous Article in Journal
Characteristics of Water and Urea–Water Solution Sprays

Catalysts 2019, 9(9), 751; https://doi.org/10.3390/catal9090751

Article
In Situ Preparation of Pr1-xCaxMnO3 and La1-xSrxMnO3 Catalysts Surface for High-Resolution Environmental Transmission Electron Microscopy
Institute of Materials Physics, University of Goettingen, D-37077, Goettingen, Germany
*
Authors to whom correspondence should be addressed.
Received: 19 July 2019 / Accepted: 29 August 2019 / Published: 6 September 2019

Abstract

:
The study of changes in the atomic structure of a catalyst under chemical reaction conditions is extremely important for understanding the mechanism of their operation. For in situ environmental transmission electron microscopy (ETEM) studies, this requires preparation of electron transparent ultrathin TEM lamella without surface damage. Here, thin films of Pr1-xCaxMnO3 (PCMO, x = 0.1, 0.33) and La1-xSrxMnO3 (LSMO, x = 0.4) perovskites are used to demonstrate a cross-section specimen preparation method, comprised of two steps. The first step is based on optimized focused ion beam cutting procedures using a photoresist protection layer, finally being removed by plasma-etching. The second step is applicable for materials susceptible to surface amorphization, where in situ recrystallization back to perovskite structure is achieved by using electron beam driven chemistry in gases. This requires reduction of residual water vapor in a TEM column. Depending on the gas environment, long crystalline facets having different atomic terminations and Mn-valence state, can be prepared.
Keywords:
perovskites; environmental transmission electron microscopy; surface chemistry

1. Introduction

Rational design of heterogeneous catalysts for high selectivity and turnover of chemical reactions requires detailed knowledge about the activity- and selectivity-determining structural properties, including catalytically active sites. Analysis of atomic and electronic structure of catalyst surfaces and subsurfaces under chemical reaction conditions thus is essential, since catalysts often undergo significant changes in surface and defect structure in their active state [1,2,3]. In situ atomic scale studies of electrocatalysts under working conditions can contribute substantially providing insights into the underlying reaction mechanism [4,5]. Environmental transmission electron microscopy (ETEM) offers unique opportunities in gaining atomic resolution images of surfaces and subsurfaces, formation of surface disorder, as well as spatially resolved spectroscopic information about the electronic structure and oxidation states of catalyst surfaces. In a dedicated ETEM instrument, a standard specimen chamber is replaced with a differently pumped environmental cell (E-cell) equipped with inlet and outlet lines [6]. The E-cell allows performing tomography [7], collecting high resolution images [8] and provides an easy access to use analytical techniques such as energy dispersive X-ray spectroscopy (EDX) and electron energy loss spectroscopy (EELS) [9,10], although the maximal pressure is limited to 20 mbar [11], or 100 mbar [12]. In its turn, dedicated gaseous holders [13] allow performing studies at pressures up to 4–5 bars [14], but field of view, spatial resolution, quality of images and use of analytical tools are limited because of silicon nitride membranes especially at low accelerating voltages [8]. Due to these recent technical improvements, ETEM has become a valuable instrument to reveal structural transformations of nanoparticles and catalyst surfaces in reactive environments [15,16,17,18].
At the same time, atomic resolution studies of catalysts require electron-transparent specimens with a thickness of only a few nanometers. Ideally, for reliable in situ experiments with cross-section specimens one needs to have an original surface of a catalyst showing no traces of preparation process in order to relate the obtained results with ex situ experiments. Conventional preparation methods include mechanical polishing and ion beam milling, where protection against surface damage is achieved by deposition of metal films [19]. In order to avoid sample thinning, Jacobs and Verhoefen [20] developed a method where a catalyst film is deposited onto a very thin Si3N4 membrane. Since the catalyst properties strongly depend on preparation methods and morphology, this method does not allow comparison with the properties of the system used in operando. Nowadays, focused ion beam (FIB) cutting combined with a lift-out technique is often applied for the preparation of ultrathin lamellas of almost any kind of material [21]. This also allows integration of lamella onto micro-electro-mechanical systems (MEMS) used in dedicated in situ holders. In order to minimize ion beam damages, the presence of protection layers such as Pt/C or W/C films is required [22]. However, both metals form clusters and can be very active as catalysts, thus masking the activity of studied material. Moreover, the presence of heavy atoms can change the contrast of high-resolution images collected to observe dynamical changes of catalysts surface. Alternative protection layers containing mainly carbon have been also reported [23].
Recently, it was demonstrated that electron beam driven epitaxial recrystallization of an amorphous surface layer developed during FIB cutting on the preexisting crystalline SrTiO3 is possible with high beam currents in scanning TEM (STEM) mode [24]. However, in high vacuum the resulting oxygen stoichiometry cannot be controlled. In a gaseous environment, the effects caused by the interaction of electron beam with the sample can be combined with a control of chemical changes. For example, STEM EELS imaging of a manganite in 3 mbar of O2 can give rise to sample oxidation by oxygen uptake [25]. This is in contrast to reduction of the same material in high vacuum at the same used electron fluxes due to oxygen knock out.
In this study, we demonstrate a FIB based two-step preparation method of cross-section TEM perovskite oxide specimens, resulting to well-ordered crystalline surface edges with controlled oxygen stoichiometry and avoiding any contamination by metallic impurities. We have selected (001) oriented epitaxial La1-xSrxMnO3 (LSMO) with x = 0.4 and Pr1-xCaxMnO3 (PCMO) with x = 0.1 and 0.33 thin films showing quite high activity as catalyst materials for the oxygen evolution reaction [17,26,27,28]. The first essential step for the lamella preparation is the FIB cutting using a protection layer of alkali resistant positive photoresist. After ion milling and removal of the photoresist by plasma etching thin amorphous layers still can present on materials. The second step is the in situ recrystallization of remaining amorphous surface layers. Epitaxial recrystallization can be achieved in O2 as well as in He and N2 with residual O2, leading to different surface terminations. The partial pressure of residual water in the column of electron microscope is found to be an essential parameter determining whether the epitaxial recrystallization of the perovskite or phase decomposition into a binary oxide evolves. Since the selected materials have significantly different electrical properties we think that observed effects are valid for many other complex oxides as well.

2. Results

The low magnification cross-section image of LSMO (x = 0.4) specimen prepared using photoresist protection layer is shown in Figure 1a. Two windows of different thickness are used to demonstrate the opportunity to control the thickness of lamella. The HRTEM image in Figure 1b and HRSTEM image in Figure 1c show that plasma cleaning during 5 min removes the rest of the photoresist completely.
Both the HRTEM and HRTEM images show that the ~40-50 nm thin lamella has a crystalline structure, despite a thin subunit-cell thick disordered surface layer. EELS spectra presented in Figure 1d show that this disordered surface of LSMO film has lower oxidation state of Mn than the bulk. Three characteristic peaks of O K-edge are visible at 529 eV (A), 534 eV (B) and 543 eV (C). The prepeak (A) due to O 1s transitions to O 2p states hybridized with Mn 3d. It reflects the presence of O 2p holes. The other two peaks are due to transition from O 1s into O states which are hybridized with Ca d / Pr d states (B) and free electron like features (C) [29]. Since oxygen vacancies are electron donors, a decrease in the intensity of the prepeak A indicates an increasing amount of oxygen vacancies. Its lack in the spectrum from the surface points to a reduced state of Mn. This is also confirmed by the analysis of the Mn L-edge. The Mn L edge is due to transition of the spin-orbit split 2p into Mn 3d states and thus shows L3 and L2 subfeatures. The L3/L2 intensity ratio is a measure of the Mn oxidation state [30]. Our analysis yields an oxidation state between +2.6 and +2.8 for the amorphous material. Atomic force microscopy images of the same LSMO 0.4 films show crystalline surface which consist of atomically flat terraces and unit cell height steps [28] as a result of step flow growth mode on vicinal SrTiO3 (STO) (001). There is some evidence that crystalline LSMO films on SrTiO3 can have reduced surfaces after vacuum annealing [31]. Since the LSMO films studied here are annealed in oxygen after growth, complete oxidation of the surface as seen in [31] is expected. We thus attribute surface disorder and Mn reduction of the TEM lamella shown in Figure 1 to preparation-induced damage. Indeed, scattering of Ga ions during FIB thinning can give rise to preferential knock out of oxygen at the LSMO-photoresist interface.
The possibility of a beam stimulated transformation of preparation induced surface disorder to the perovskite crystal structure is demonstrated for a 2–3 nm thick amorphous PCMO layer on top of crystalline PCMO x = 0.33 in different gases, i.e. He, N2, and O2 in Figure 2. The cross-section lamellas have been cut from the PCMO/LSMO/Nb-STO stack; 300 kV electron beam stimulation is performed in TEM mode with a beam of about 100 nm in diameter and ~4 nA current. Such illumination conditions result in a 14,000–20,000 e·s−1·Å−2 dose rate. The process takes about 20–25 min in order to complete the transformation of the amorphous layers in He (Figure 2a) or in N2 (Figure 2c) to a surface facets having of a length of about a few dozen nanometers, Figure 2b,d, respectively. It is noteworthy that the first facets of several nanometers in length were formed already after the first three minutes. The corresponding movies S1, S2 and S3 show the real time changes of structure in N2 gas (pN2 = 5 ubar) at the beginning, after about 20 min and at the end of process, respectively. During all stages of process a pronounced atomic dynamics at the surface is observed similar to Epicier et al. [32]. Also, Figure 2b,d demonstrate that electron beam irradiation can result in different surface terminations. This will be discussed in detail further down. Thus, the process resembles the so-called solid phase epitaxial (SPE) growth, where an initially amorphous Si [33] or complex oxide [34] thin film is epitaxially recrystallized on a crystalline template at elevated temperatures. Remarkably, the electron-beam-induced SPE process in our environmental TEM study is observed at room temperature and depends on a reactive gas environment. The required electron flux is orders of magnitude lower than that reported for the transformation of amorphous STO into perovskite structure during operation the microscope in STEM mode (electron dose rate ≈ 1 × 108 e·s−1·Å−2, high vacuum conditions p = 10−7 mbar) [24].
The crystallization in O2 gas proceeds similarly as in He or N2, however, it takes only about 5–6 min at the same illumination conditions. Snapshots of crystallization in O2 are presented in Figure 3. At the beginning, during several minutes, only mobility of atoms in amorphous layer is visible, similar to observations in He or N2 (Figure 3a–d). Then, the growth of small pyramidal islands is observed (one island is marked with the yellow arrow). These islands grow slowly and over time reach the surface of amorphous layer (Figure 3e). At the same time the thickness of amorphous layer is decreasing. Finally, the amorphous material between crystalline islands also becomes crystalline, and flat facets up to 10 nm in length are formed (Figure 3f–h). It is important to note that in all gases (He, N2 and O2), the final phase of crystallization, i.e. when crystalline islands have sprouted through the amorphous layer, proceeds faster.
In order to verify the chemical composition of original, amorphous and recrystallized PCMO, the EELS spectra were collected at the beginning of experiment and after the recrystallization was completed. In addition, K- edges of oxygen and L-edges of Mn were collected with the better energy resolution to examine the oxidation state of Mn.
The HRSTEM images shown in Figure 4a,b demonstrate an edge of PCMO (x = 0.33) film before and after recrystallization in oxygen at pO2 = 50 ubar, respectively. Clean and atomically sharp steps having a height of few unit cells are commonly observed. The corresponding EELS spectra are shown in Figure 4c. All spectra were aligned using a zero-loss peak (ZLP) as a reference, and then normalized. Note, that the spectrum collected from the amorphous edge looks noisier than other spectra because of smaller thickness of the edge. Based on the reduced intensity of the O K prepeak (A), the crystallization in He results in a reduced state of Mn at the surface. The analysis of L3/L2 ratio of Mn L-edge yields an average value of 3.30 ± 0.06 and a value of 3.24 ± 0.02 close to the surface (see Figure S1). In contrast, when the crystallization was performed in O2 the average value of the Mn oxidation state is 3.36 ± 0.06 and 3.35 ± 0.06 at the surface (see Figure S2). All values are close to the value for the nominal composition as in bulk (+3.33). Thus, the use of oxygen leads to the epitaxial recrystallization of a stoichiometric PCMO, whereas in He, a crystalline edge with increased lattice disorder is developed. The average Mn valence state is the same in He as in O2 within error based on the analysis of the L3/L2 intensity ratio. However, there is a strong reduction of the O K prepeak at the surface, pointing towards a more reduced state in He. The effect of disorder on electronic properties is quite complex. For example, in addition to oxygen vacancies, charge neutral pair defects can contribute to the lattice disorder in He which can maintain the Mn valence while changing the O K edge. Nevertheless, for all three gases used, the lattice spacing up to most top unit cell corresponds to the bulk PCMO material within error.
The chemical composition of the amorphous layer as well as recrystallized material was also verified using EELS, and only Pr, Ca, Mn, and O edges were detected; Figure 5 shows the results for the PCMO, x = 0.33. A HRSTEM image of the recrystallized area is shown in Figure 5a. The EELS spectra reveal a decrease of Ca content in the surface unit cells (Figure 5b). Since the properties of perovskite materials strongly depend on the amount of doping, a careful choice of the illumination conditions and the environment is very essential for control of the chemical composition of the recrystallized perovskite. A long lasting electron beam illumination to fully recrystallize the amorphous layer obviously results in element-specific knock out damage of light elements which is evidenced by chemical analysis (Figure 5). In addition, the lattice disorder of the recrystallized perovskite due to off-stoichiometry is visible by broadening of spots and changing of their intensities in corresponding FFT images calculated from the HRTEM images collected in all the gases used. The effect is notably less pronounced in series of HRTEM images collected from the same TEM lamella in O2 (see Figures S3 and S4), supporting our statement of a less vacancy disordered structure because Mn acquires its nominal valence state.
As it was already described above, that flat facets were formed from the amorphous material at the surface of PCMO in all He, N2 and O2 gases used. Moreover, a careful inspection of the top atomic layer revealed that the termination of the recrystallized edge depends on the gas used. This effect is shown in Figure 6. Both A- and B-cation stable terminations were observed after recrystallization in N2 (Figure 6a), while only the A-cation termination of facets was found after recrystallization in He (Figure 6d). The observation is verified by HRTEM image simulations presented in Figure 6b,c,e,f, respectively. The HRTEM images of PCMO along [100] and [010] look differently because of the symmetry inherent in the Pbnm space group. The Mn-O (B-termination along [010]) is obviously recognized in Figure 6a,b because of a characteristic zigzag motif of Mn-O atomic columns (see Figure S6 for details). In the case of A-termination the difference in HRTEM images along [100] and [010] is almost invisible.
As an example for the recrystallization of PCMO (x = 0.1), Figure 7a–c shows the evolution in He. The first crystalline facets showed up only after 28 min, compared with about 11 min for PCMO (x = 0.33). After such a long irradiation time, one can expect a high density of defects in the material. Indeed, the spots in the FFT image became more smeared after 28 min, which is evidence of the increased structural disorder (see Figure S5). The weak signal of Ca is still visible in the EELS spectra taken from the edge (Figure 7d,e) taking into account that the concentration of Ca is already very low in the original material. In general, the EELS data from PCMO (x = 0.1) sample are very scattered. In some areas the analysis of O-K and Mn-L edges shows that the oxidation state of Mn is in the +3.0–+3.3 range (Figure 7f) pointing against a large concentration of single oxygen vacancies as origin of the lattice disorder. However, very low values of the oxidation state of Mn were also observed (Figure S7).
Usually, the recrystallization of the edge in PCMO (x = 0.1) took much longer for all gases used. This is however consistent with cyclovoltammetry studies, showing a pronounced surface Mn2+/½VO ↔ Mn3+/OO redox couple for PCMO (0.33) which is absent in x = 0.1. As an example, Figure 8 shows cyclovoltagrams of two epitaxial (001) oriented PCMO (x = 0.1 and x = 0.33) films on LSMO/Nb:STO at pH = 13 measured at a sweep rate of 10 mV/s. The electrochemical measurements were performed in 0.1 M potassium hydroxide (KOH) electrolyte prepared by diluting KOH stock solution with deionized water (for neutral pH see [17]). Both films show an exponential increase of currents at positive potential due to oxygen evolution reaction (OER), as verified by measuring the formed O2 by a ring electrode. The pronounced redox peak for x = 0.33 indicates an easier oxygen vacancy formation and healing process compared to x = 0.1 which explains the observation of faster epitaxial recrystallization for x = 0.33 compared to x = 0.1 in the ETEM experiments.
Finally, we note that the epitaxial crystallization only happens at the preexisting crystalline interface, when the partial water pressure is pH2O < 5·10−8 bar (Figure 9). In other words, the residual pressure of water vapor in the column should be below 0.1% of the working gas. Exemplary snap shot from the Quadera™ software (Figure 9a) shows two periods during the experiment in He when the cold trap was not used (Cold trap OFF) and after it was activated (Cold trap ON). At the higher pressure of water vapor (pH2O = 25·10−8 bar, Cold Trap OFF) the amorphous layer was transformed into randomly oriented PrO2 particles as shown in Figure 9b (see also Figure S8). Similar structural transformations of initially amorphous areas were reported earlier for the ETEM experiments performed in water vapor at higher pressures of 5·10−6 and 5·10−5 bar [18], and this is consistent with the drop of OER activity in electrochemical experiments over time due to Mn leaching into the electrolyte [28]. After the cold trap was filled with liquid nitrogen and as soon as the partial pressure of water vapor dropped about an order of magnitude (pH2O = 25·10−9 bar), the epitaxial recrystallization was observed (Figure 9c) similar to results presented in Figure 2, Figure 3 and Figure 4.

3. Discussion

As already mentioned, the observed recrystallization of amorphous PCMO resembles SPE of an amorphous Si which happens via a moving amorphous/crystalline interface, where velocities of a few nm per min requires elevated temperatures above 500 °C [33]. The recrystallization temperature can be significantly reduced to around 200–300 °C, if the recrystallization is assisted by ion irradiation [34]. Although the electron-beam induced recrystallization reported here also happens via the movement of the crystalline/amorphous interface, it however strongly differs from the SPE. First of all, the beam induced temperature increase ΔT in the TEM lamella is irrelevant, excluding a thermally activated growth mechanisms. The upper limit of temperature increase for the used beam parameters is Δ T = 0.015 K (see Appendix A). Furthermore, the epitaxial recrystallization in our experiments strongly depends on the gas environment. In high vacuum conditions, no epitaxial growth is observed. Here, the amorphous layer partially disappears on timescales of tens of minutes due to electron beam-induced sputtering of atoms from damaged areas.
In oxygen both amorphous PCMO x = 0.1 and x = 0.33 layers transform into a crystalline perovskite layer under oxygen uptake during the recrystallization process. This is confirmed by a change of the Mn valence state from about +2.9 in the amorphous to close to the nominal value. There are three possible candidates for beam induced electronic excitations driven the observed solid state chemistry: (i) a photo-chemical process driven by beam induced generation of hole carriers. Since the lifetime of photo-holes (1 ns) is orders of magnitudes below the recrystallization rate (1 s per atom), a photo-chemical process is highly improbable. (ii) An electrochemical process driven by beam induced positive electric potentials due to secondary electron emission. Such an electrochemical mechanism is feasible: The relevant redox potential for oxygen in manganites is of the order of 1 V (see Figure 8), a potential which can be easily generated in semiconductors at dose rates above 10.000 e/Å2s [35]. (iii) Radiolysis, the dissociation of molecules or covalently bound solids by inelastic scattering of primary electrons with bond forming valence electrons. Whereas radiolysis may be relevant for the dissociation of molecular oxygen and activation of O2 for the solid state reaction, it is irrelevant in ionic or metallic solids. Here, this is confirmed by the formation of rather perfect crystalline perovskite structure in O2.
Beam-induced displacements of atoms by nuclear collisions is probable process for activating the growth of crystalline PCMO by a moving amorphous/crystalline interface as observed in O2 as well as in the inert gases N2 and He. However, it cannot explain the oxygen uptake in O2. Maximum energy transfer takes place by backscattering of electron which could displace surface adsorbed oxygen into the amorphous material on the top side of the lamella, but would lead to sputtering on the bottom side and at the edge. This would represent an “inverse sputtering process”, where in contrast to HV, oxygen vacancies are filled by activated surface oxygen. Conversely, in N2, one should then form oxynitrates, in contrast to observation.
The observation that the amorphous phase transforms into crystalline PrOx nanoparticles in the presence of sufficient partial pressure of H2O by preferential leaching of Mn and also Ca give strong hints about the beam induced chemistry involved. Based on the equation provided by Egerton et al. [36], for the energy transfer between a high energy electron and an atom due to elastic scattering, we calculate the upper limit for the different atomic species: Emax is 53.1 eV (O2), 42.5 eV (Ca), 34.0 eV (Mn) and 14.4 eV (Pr). This must be compared to binding energies of Eb = 50 eV (O), 70 eV (A site) [14] and between 50–170 eV (Ti as an exemplary B-site in BaTiO3) [37], depending on the crystal orientation of a perfect perovskite single crystal. This is consistent to the statement that Emax < Eb. However, Eb should be generally reduced in the amorphous phase of the same chemical elements, because of the lower density of packing, open voids and change of bond coordination. Remarkably, our experiments show that the preferential sputtering of atomic species strongly depends on the gas environment. In HV, the amorphous surface layer is partially removed with preferential sputtering of O, Ca and Mn without formation of PrOx. In contrast, in O2 or inert gases, the presence/absence of preferential leaching of Mn and Ca strongly depends on the pH2O. This reflects that the vacancy formation energy close to surfaces depends on the surface chemistry. The presence of adsorbed surface H2O, where Mn2+ and Ca2+ is soluble, clearly facilitates the leaching of displaced Mn and Ca species, giving rise to the formation of PrOx. In contrast, in O2 this process is suppressed and the uptake of oxygen for example by occupation of VO vacancy sites is the dominating process.
Clearly, the crystallization rate and thus the velocity of the amorphous/crystalline interface is slowed down in inert gases He and N2, compared to experiments in O2. This indicates, that beam induced atomic displacements cannot be the only driving force for the movement of the interface. Figure S9 shows that residual molecular and atomic oxygen is also present in an inert gas experiment, e.g., in He. Its relative concentration is only slightly reduced after switching on the cold trap. The epitaxial recrystallization in inert gases may thus be influenced by the presence of oxygen. The analysis of the Mn valence state after He in comparison with the amorphous material at the pristine lamellas does not provide evidence for a large amount of oxygen uptake, since the Mn valence state is the same within error. However, sputtering of oxygen may be still compensated by the electrochemical oxidation, which is slowed down at lower pO2. This could indicate that beam induced epitaxial recrystallization in inert gases is also accompanied by a solid state chemical reaction with oxygen. Consequently, even without net uptake of oxygen, the presence of surface adsorbed oxygen species can alter the beam induced effects from absent in high vacuum, over leaching in H2O to perovskite epitaxy in inert gases.

4. Materials and Methods

Thin films of Pr1-xCaxMnO3 (PCMO) with different Pr/Ca ratio (x = 0.1 and x = 0.33) were grown on Nb-doped SrTiO3 (001) with a 20 nm thick epitaxial La0.6Sr0.4MnO3 (LSMO) (001) interlayer by reactive Xe-ion beam sputtering in an oxygen atmosphere with a partial pressure of 1.4·10−4 mbar at the deposition temperature of Tdep = 1020 K. The PCMO film of thickness of d ≈ 100 nm reveals epitaxial growth with [001]/[110] growth domains, inevitably present due to twinning. The LSMO interlayer was grown in order to adjust the lattice misfit between STO and PCMO. The growth parameters and detailed structural characterization of studied films is published elsewhere [17,18,28].
Specimens for in situ TEM studies were prepared by lift-out technique using a NanoLab 600 (Thermo Scientific™, former FEI, Hillsboro, Oregon, USA) dual beam instrument operated at 30 and 5 kV. Conventional method with FIB lift-out technique usually starts from the deposition of platinum/carbon (Pt/C) protection layers by electron or ion beam [38]. Since Pt is highly active catalyst, its presence at the surface of perovskites should be excluded. All lamellas in this study were prepared using a protection layer of Alkali Resistant Positive Photoresist X AR-P 5900/4 (ALLRESIST GmbH, Strausberg, Germany) which was deposited onto the surface of LSMO and PCMO thin films before FIB cutting procedure. Electron transparent lamella was finally cleaned with Ar+ ions with energies from 500 eV to 100 eV using a Gatan PIPS 691 ion polishing system. The rest of protection layer was removed with a plasma cleaner prior TEM studies. High-resolution TEM (HRTEM), electron diffraction (ED) and STEM studies were performed using an aberration corrected Titan 80-300 environmental microscope (Thermo Scientific™, former FEI, Hillsboro, Oregon, USA) operated at 300 kV. The specimens were imaged along [010] and [110] crystallographic directions in the Pbnm space group. ETEM experiments are performed with N2, He, O2 gases. The chemical composition of gas environment was controlled by the Residual Gas Analyzer (RGA) QMG220 (Pfeiffer Vacuum GmbH, Germany) controlled by Quadera™ software (version 4.3, INFICON Aktiengesellschaft, Liechtenstein, 2010). During the experiments the deviations of pressure from desired values of 5 and 50 microbars did not exceed 5%. Other residual gases are always present in the inlet tubes and electron microscope column. In order to minimize their influence the inlet tubes were purged firstly with the working gas. Then, the specimen chamber was also purged with the working gas during at least half an hour, and the electron beam was blanked. These procedures allowed reducing the presence of other gases to be less than 1%. For example, during the used experimental setting, typical values are H2O: 1% (cold trap off) and 0.05% (cold trap on); O2: 0.2%; N2: 0.04%; hydrocarbons (unspecified) and very low amounts of atomic nitrogen and oxygen. Switching on the cold trap thus changes the main residual gas impurity from H2O to O2 (see SI1). EELS data were collected with a Gatan Imaging Filter (GIF) Quantum 965ER (Gatan, Pleasanton, CA, USA). Atomic models were built using the Vesta software package (version 3.4.7, 2019) [39]. High-angle annular dark field (HAADF) contrast and spectrum imaging was used in order to visualize change in chemical composition. HRTEM images were simulated using QSTEM software (version 2.51, 2019) [40].

5. Conclusions

We successfully demonstrated high-quality preparation of cross-section TEM lamella of complex oxide thin films by FIB, using metal free protection layer, i.e. photoresist. Although beam damage is small, subunit cell disordered layers remain on the surface after plasma etching of photoresist which can be treated in situ by beam driven recrystallization in suitable gas environments. It allows the formation of well-ordered crystalline facets with different surface termination, dependent on the gas. This is an important prerequisite for the in situ atomic scale observation of active states of catalysts. The work also demonstrates the feasibility of direct imaging of solid state chemical and structural transformations from amorphous to crystalline structures on atomic scale. The length of modified edge at TEM lamella varies in a range from several nanometers to several tens or even hundred nanometers. The chemical composition of the modified area depends on the gases used, as well as on the presence of residual water molecules in the column of electron microscope. The presence of water can cause selective leaching of atomic species (Mn and Ca) and thus irreversible chemical changes of surface of PCMO perovskites. By contrast, the reduction of water partial pressure below 5·10−8 bar forms conditions, where crystalline perovskite oxide structure can be grown from a damaged amorphous material by electron-beam-induced solid phase epitaxy controlled by gas regulated solid state chemistry. The observation that the nominal Mn oxidation state can be reached from the reduced amorphous state in pO2 = 50·10−6 bar points to the key role of residual oxygen in N2 and He environments for the healing of preparation induced damage. At this point of our studies, we suggest that beam-induced atomic displacements drive the crystallization front between the amorphous and crystalline phase. However, this process is most probably accompanied by beam-induced electrochemical oxidation which compensates the sputtering of oxygen or can even drive oxygen uptake to the stoichiometric level. This statement is strongly supported by the observed higher speed of epitaxial recrystallization for the material with a pronounced electrochemical redox couple Mn2+/½VO ↔ Mn3+/OO.
The reported results are important both for the design of in situ experiments related to the study of heterogeneous catalysis and battery technology in oxide-based materials. The well controlled design of perovskite oxide surface edges in O2 and inert gases at low pH2O is a vital requirement for the realization of a well-defined starting state of the surface in situ reactions. In order to have better control over size and chemical composition of the modified area, significant improvements of specimen stage stability, inlet and outlet systems of ETEM as well as chemical analysis of gaseous environment is highly desirable.

Supplementary Materials

The following are available online at https://www.mdpi.com/2073-4344/9/9/751/s1: Figure S1: The PCMO (x = 0.33) after electron beam treatment in He at 50 μbars, Figure S2: The PCMO (x = 0.33) after electron beam treatment in O2 at 50 μbars, Figure S3: HRTEM image and rotationally averaged profiles of PCMO in O2 and in He (x = 0.33), Figure S4: HRTEM image and rotationally averaged profiles of PCMO in He (x = 0.1), Figure S5: Analysis of broadening of spots in FFTs, lattice disorder in PCMO (x = 0.1), Figure S6: (1) Through focus series of HRTEM images of PCMO (x = 0.33) for the thickness of 10 nm, (2) Experimental and calculated HRTEM images of PCMO (x = 0.33) for the thickness of 3 nm, Figure S7: The PCMO (x = 0.1) after electron beam treatment in He at 50 μbars, Figure S8: HRSTEM image and EELS spectrum of PrOx particle at the surface of PCMO (x = 0.33), Figure S9: An example of multiple ion detection scan in Quadera™ software, Table S1: Summary of the Mn valence PCMO (x = 0.33) after electron beam treatment in He at 50 μbars, Table S2: Summary of the Mn valence of PCMO (x = 0.33) after electron beam treatment in O2 at 50 μbars. Table S3 (1): Simulation parameters for the through focus series, (2) Simulation parameters for individual experimental images, Table S4. Summary of the Mn valence of PCMO (x = 0.1) after electron beam treatment in He at 50 μbars.

Author Contributions

Conceptualization: C.J. and V.R.; Formal analysis: G.L. and V.R.; Investigation: G.L. and V.R.; Resources: C.J.; Writing—original draft preparation: V.R.; Writing—review and editing: V.R., G.L. and C.J.; Visualization: V.R. and G.L.; Supervision: C.J. and V.R.; Project administration: C.J. and V.R.; Funding acquisition: C.J. and V.R.

Funding

The authors are grateful for financial support by the Deutsche Forschungsgemeinschaft, grant number CRC 1073 (Projects C02 and Z02).

Acknowledgments

The authors wish to thank V. Radisch, V. Moshnyaga and V. Bruchmann-Bamberg for excellent support in TEM specimen preparation.

Conflicts of Interest

The authors declare no conflict of interest.

Appendix A

We calculate the upper limit of TEM lamella heating in the irradiated area under the assumption that all excitations due to inelastic scattering of primary electrons are converted into heat. The energy uptake per unit volume and second (= adsorbed power P per unit volume V) is thus:
P V = Δ E   W   N e ˙ V
where the rate of incidents electrons is given by:
N e ˙ = j   A e
Here, A denotes the illuminated area of the lamella and j the electron current density of primary electrons. In order to estimate the temperature increase due to the absorbed power, we us the Fourier law in 1D:
J q = κ d T d x = P A
where κ is the thermal conductivity of the manganite, Jq is the heat flux density, which is equal to the absorbed power per unit area. From these equations it follows:
Δ T = L κ   P A
where L is the length where the heat must be transported before it hits the interface to the substrate with a much higher thermal conductivity. Heat transfer due to gas convection is negligible at the used pressure range of our experiments.
For the calculation, we use a current of I = 4 nA on a circle of 100 nm diameter. This corresponds to an average dose rate (flux) of r = 3.18 × 104 e/Å2s. In semiconductors, the average energy transfer per inelastic excitation ΔE ~ 3 Eg, where Eg is the bandgap of the material [41]. Since the low loss excitations are the dominating feature in EELS, we use this estimate. It gives ΔE = 3eV per inelastic scattering event in PCMO.
With a measured inelastic mean free path of λ = 70 nm, the scattering probability of an electron in the thin surface regions of t = 10 nm is W = t/ λ = 0.14. Thermal conductivity of PCMO at T = 300 K is k = 1.5 W/mK [42]. The distance of the electron irradiated surface area to the substrate is around L = 100 nm. This yields an upper limit of beam-induced lamella heating of ΔT = 15 mK.

References

  1. Kanan, M.W.; Nocera, D.G. In situ formation of an oxygen evolving catalyst in neutral water containing phosphate and Co2+. Science 2008, 321, 1072–1075. [Google Scholar] [CrossRef]
  2. Vendelbo, S.B.; Elkjær, C.F.; Puspitasari, I.; Creemer, J.F.; Dona, P.; Mele, L.; Morana, B.; Nelissen, B.J.; van Rijn, R.; Kooyman, P.J.; et al. Visualization of oscillatory behaviour of Pt nanoparticles catalysing CO oxidation. Nat. Mater. 2014, 13, 884–890. [Google Scholar] [CrossRef]
  3. Bergmann, A.; Martinez-Moreno, E.; Teschner, D.; Chernev, P.; Gliech, M.; Ferreira de Araújo, J.; Reier, T.; Dau, H.; Strasser, P. Reversible amorphization and the catalytically active state of crystalline Co3O4 during oxygen evolution. Nat. Commun. 2015, 6, 8625. [Google Scholar] [CrossRef]
  4. Arrigo, R.; Hävecker, M.; Schuster, M.E.; Ranjan, C.; Stotz, E.; Knop-Gericke, A.; Schlögl, R. In situ study of the gas-phase electrolysis of water on platinum by NAP-XPS. Angew. Chem. Int. Ed. 2013, 52, 11660–11664. [Google Scholar] [CrossRef]
  5. Zheng, F.; Alayoglu, S.; Guo, J.; Pushkarev, V.; Li, Y.; Glans, P.A.; Chen, J.L.; Somorjai, G. In-situ X-ray absorption study of evolution of oxidation states and structure of cobalt in Co and CoPt bimetallic nanoparticles (4 nm) under reducing (H2) and oxidizing (O2) environments. Nano Lett. 2011, 11, 847–853. [Google Scholar] [CrossRef]
  6. Boyes, E.D.; Gai, P.L. Environmental high resolution electron microscopy and applications to chemical science. Ultramicroscopy 1997, 67, 219–232. [Google Scholar] [CrossRef]
  7. Roiban, L.; Koneti, S.; Morfin, F.; Nguyen, T.S.; Mascunan, P.; Aouine, M.; Epicier, T.; Piccolo, L. Uncovering the 3D structure of combustion-synthesized noble metal-ceria nanocatalysts. ChemCatChem 2017, 9, 4607–4613. [Google Scholar] [CrossRef]
  8. Hansen, T.W.; Wagner, J.B. Environmental transmission electron microscopy in an aberration-corrected environment. Microsc. Microanal. 2012, 18, 684–690. [Google Scholar] [CrossRef]
  9. Hansen, T.W.; Wagner, J.B.; Dunin-Borkowski, R.E. Aberration corrected and monochromated environmental transmission electron microscopy–Challenges and prospects. Mater. Sci. Technol. 2010, 26, 1338–1344. [Google Scholar] [CrossRef]
  10. Hansen, T.W.; Wagner, J.B.; Jinschek, J.R.; Dunin-Borkowski, R.E. The titan environmental transmission electron microscope: Specifications, considerations and first results. Microsc. Microanal. 2009, 15, 714–715. [Google Scholar] [CrossRef]
  11. Thermo Fisher Scientific. Available online: https://www.fei.com/products/tem/themis-etem-for-materials-science/ (accessed on 30 August 2019).
  12. Tanaka, N.; Usukura, J.; Kusunoki, M.; Saito, Y.; Sasaki, K.; Tanji, T.; Muto, S.; Arai, S. Development of an environmental high voltage electron microscope and its application to nano and bio-materials. J. Phys. Conf. Ser. 2014, 522, 012008. [Google Scholar] [CrossRef]
  13. FanWu, F.; Yao, N. Advances in windowed gas cells for in-situ TEM studies. Nano Energy 2015, 13, 735–756. [Google Scholar]
  14. Yokosawa, T.; Alan, T.; Pandraud, G.; Damc, B.; Zandbergen, H. In-situ TEM on (de)hydrogenation of Pd at 0.5–4.5 bar hydrogen pressure and 20–400 °C. Ultramicroscopy 2012, 112, 47–52. [Google Scholar] [CrossRef]
  15. Zhang, S.; Nguyen, L.; Zhu, Y.; Zhan, S.; Tsung, C.K.; Tao, F. In-situ studies of nanocatalysis. Acc. Chem. Res. 2013, 46, 1731–1739. [Google Scholar] [CrossRef]
  16. Han, B.; Stoerzinger, K.A.; Tileli, V.; Gamalski, A.D.; Stach, E.A.; Shao-Horn, Y. Nanoscale structural oscillations in perovskite oxides induced by oxygen evolution. Nat. Mater. 2017, 16, 121–126. [Google Scholar] [CrossRef]
  17. Raabe, S.; Mierwaldt, D.; Ciston, J.; Uijttewaal, M.; Stein, H.; Hoffmann, J.; Zhu, Y.M.; Blochl, P.; Jooss, C. In situ electrochemical electron microscopy study of oxygen evolution activity of doped manganite perovskites. Adv. Funct. Mater. 2012, 12, 3378–3388. [Google Scholar] [CrossRef]
  18. Mierwaldt, D.; Roddatis, V.; Risch, M.; Scholz, J.; Geppert, J.; Abrishami, M.E.; Jooss, C. Environmental TEM investigation of electrochemical stability of perovskite and ruddlesden-popper type manganite oxygen evolution catalysts. Adv. Sustain. Syst. 2017, 1, 1700109. [Google Scholar] [CrossRef]
  19. Newcomb, S.B.; Boothroyd, C.B.; Stobbs, W.M. Specimen preparation methods for the examination of surfaces and interfaces in the transmission electron microscope. J. Microsc. 1985, 140, 195–207. [Google Scholar] [CrossRef]
  20. Jacobs, J.W.M.; Verhoeven, J.F.C.M. Specimen preparation technique for high resolution transmission electron microscopy studies on model supported metal catalysts. J. Microsc. 1986, 143, 103–116. [Google Scholar] [CrossRef]
  21. Giannuzzi, L.A.; Geurts, R.; Ringnalda, J. 2 keV Ga+ FIB milling for reducing amorphous damage in silicon. Microsc. Microanal. 2005, 11, 828–829. [Google Scholar] [CrossRef]
  22. Schaffer, M.; Schaffer, B.; Ramasse, Q. Sample preparation for atomic-resolution STEM at low voltages by FIB. Ultramicroscopy 2012, 114, 62–71. [Google Scholar] [CrossRef]
  23. Park, Y.C.; Park, B.C.; Romankov, S.; Park, K.J.; Yoo, J.H.; Lee, Y.B.; Yang, J.M. Use of permanent marker to deposit a protection layer against FIB damage in TEM specimen preparation. J. Microsc. 2014, 255, 180–187. [Google Scholar] [CrossRef]
  24. Jesse, S.; He, Q.; Lupini, A.R.; Leonard, D.N.; Oxley, M.P.; Ovchinnikov, O.; Unocic, R.R.; Tselev, A.; Fuentes-Cabrera, M.; Sumpter, B.G.; et al. Atomic-level sculpting of crystalline oxides: Toward bulk nanofabrication with single atomic plane precision. Small 2015, 11, 5895–5900. [Google Scholar] [CrossRef]
  25. Kramer, T.; Mierwaldt, D.; Scherff, M.; Kanbach, M.; Jooss, C. Developing an in situ environmental TEM set up for investigations of resistive switching mechanisms in Pt-Pr1-xCaxMnO3-δ-Pt sandwich structures. Ultramicroscopy 2018, 184, 61–70. [Google Scholar] [CrossRef]
  26. Matsumoto, Y.; Sato, E. Oxygen evolution on La1−xSrxMnO3 electrodes in alkaline solutions. Electrochim. Acta 1979, 24, 421–423. [Google Scholar] [CrossRef]
  27. Bockris, J.O.; Otagawa, T. The electrocatalysis of oxygen evolution on perovskites. J. Electrochem. Soc. 1984, 131, 290–302. [Google Scholar] [CrossRef]
  28. Scholz, J.; Risch, M.; Stoerzinger, K.A.; Wartner, G.; Shao-Horn, Y.; Jooss, C. Rotating ring-disk electrode study of oxygen evolution at a perovskite surface: Correlating activity to manganese concentration. J. Phys. Chem. 2016, 120, 27746–27756. [Google Scholar] [CrossRef]
  29. Sotoudeh, M.; Rajpurohit, S.; Blöchl, P.E.; Mierwaldt, D.; Norpoth, J.; Roddatis, V.; Mildner, S.; Ifland, B.; Jooss, C. Electronic structure of Pr1−xCaxMnO3. Phys. Rev. B 2017, 95, 235150. [Google Scholar] [CrossRef]
  30. Varela, M.; Oxley, M.P.; Luo, W.; Tao, J.; Watanabe, M.; Lupini, A.R.; Pantelides, S.T.; Pennycook, S.J. Atomic-resolution imaging of oxidation states in manganites. Phys. Rev. B 2009, 79, 085117. [Google Scholar] [CrossRef]
  31. De Jong, M.P.; Bergenti, I.; Dediu, V.A.; Fahlman, M.; Marsi, M.; Taliani, C. Evidence for Mn2+ ions at surfaces of La0.7Sr0.3MnO3 thin films. Phys. Rev. B 2005, 71, 014434. [Google Scholar] [CrossRef]
  32. Bugnet, M.; Overbury, S.H.; Wu, Z.L.; Epicier, T. Direct visualization and control of atomic mobility at {100} surfaces of ceria in the environmental transmission electron microscope. Nano Lett. 2017, 17, 7652–7658. [Google Scholar] [CrossRef]
  33. Rudawski, N.G.; Jones, K.S.; Morarka, S.; Law, M.E.; Elliman, R.G. Stressed multidirectional solid-phase epitaxial growth of Si. J. Appl. Phys. 2009, 105, 081101. [Google Scholar] [CrossRef]
  34. Evans, P.G.; Chen, Y.; Tilka, J.A.; Babcock, S.E.; Kuech, T.F. Crystallization of amorphous complex oxides: New geometries and new compositions via solid phase epitaxy. Curr. Opin. Solid State Mater. Sci. 2018, 22, 229–242. [Google Scholar] [CrossRef]
  35. Mildner, S.; Beleggia, M.; Mierwaldt, D.; Hansen, T.W.; Wagner, J.B.; Yazdi, S.; Kasama, T.; Ciston, J.; Zhu, Y.; Jooss, C. Environmental TEM study of electron beam induced electrochemistry of Pr0.64Ca0.36MnO3 catalysts for oxygen evolution. J. Phys. Chem. C 2015, 119, 5301–5310. [Google Scholar] [CrossRef]
  36. Egerton, R.F.; Li, P.; Malac, M. Radiation damage in the TEM and SEM. Micron 2004, 35, 399–409. [Google Scholar] [CrossRef]
  37. Gonzalez, E.; Abreu, Y.; Cruz, C.M.; Piñera, I.; Leyva, A. Molecular-dynamics simulation of threshold displacement energies in BaTiO3. Nucl. Instrum. Methods Phys. Res. B 2015, 358, 142–145. [Google Scholar] [CrossRef]
  38. Giannuzzi, L.A.; Stevie, F.A. A review of focused ion beam milling techniques for TEM specimen preparation. Micron 1999, 30, 197–204. [Google Scholar] [CrossRef]
  39. Momma, K.; Izumi, F. VESTA 3 for three-dimensional visualization of crystal, volumetric and morphology data. J. Appl. Crystallogr. 2011, 44, 1272–1276. [Google Scholar] [CrossRef]
  40. Koch, C.T. Determination of Core Structure Periodicity and Point Defect Along Dislocations. Ph.D. Thesis, Arizona State University, Tempe, AZ, USA, May 2002. [Google Scholar]
  41. Klein, C.A. Bandgap dependence and related features of radiation ionization energies in semiconductors. J. Appl. Phys. 1968, 39, 2029–2038. [Google Scholar] [CrossRef]
  42. Thiessen, P.; Roddatis, V.; Rieger, F.; Belenchuk, A.; Keunecke, M.; Moshnyaga, V.; Jooss, C. Effect of charge ordering on cross-plane thermal conductivity in correlated perovskite oxide superlattices. Phys. Rev. B. 2018, 98, 195114. [Google Scholar] [CrossRef]
Figure 1. (a) Low magnification image of La0.6Sr0.4MnO3 (LSMO) (x = 0.4) lamella prepared with the use of photoresist protection layer; (b) High-resolution transmission electron microscope (HRTEM) image of an edge collected from the thin window area; (c) high-resolution scanning TEM (HRSTEM) image, the areas marked with red and blue are used to collect electron energy loss spectroscopy (EELS) data (0.1 eV/channel) presented in (d). Spectral features marked with A, B, C, L2 and L3 are discussed in text.
Figure 1. (a) Low magnification image of La0.6Sr0.4MnO3 (LSMO) (x = 0.4) lamella prepared with the use of photoresist protection layer; (b) High-resolution transmission electron microscope (HRTEM) image of an edge collected from the thin window area; (c) high-resolution scanning TEM (HRSTEM) image, the areas marked with red and blue are used to collect electron energy loss spectroscopy (EELS) data (0.1 eV/channel) presented in (d). Spectral features marked with A, B, C, L2 and L3 are discussed in text.
Catalysts 09 00751 g001
Figure 2. HRTEM images of an edge of Pr1-xCaxMnO3 (PCMO) (x = 0.33) material in He (pHe = 50 ubar) at the beginning (a), after 750 s (b), and in N2 at the beginning (c) and after 700 s (d), respectively.
Figure 2. HRTEM images of an edge of Pr1-xCaxMnO3 (PCMO) (x = 0.33) material in He (pHe = 50 ubar) at the beginning (a), after 750 s (b), and in N2 at the beginning (c) and after 700 s (d), respectively.
Catalysts 09 00751 g002
Figure 3. A series of images showing transformation of amorphous edge of PCMO (x = 0.33) into crystalline structure. The images were collected in O2, pO2 = 50 ubar. (a) During several minutes only mobility of atoms in amorphous layer is visible. (bd) The growth of small island of pyramidal shape is observed (the place is marked with the yellow arrow). (e)These pyramids grow slowly and reach the surface of amorphous layer. Its thickness becomes smaller at the same time. Finally, amorphous areas between pyramids also transform in crystalline perovskite oxide phase, and flat facets up to 10 nm in length are formed (fh).
Figure 3. A series of images showing transformation of amorphous edge of PCMO (x = 0.33) into crystalline structure. The images were collected in O2, pO2 = 50 ubar. (a) During several minutes only mobility of atoms in amorphous layer is visible. (bd) The growth of small island of pyramidal shape is observed (the place is marked with the yellow arrow). (e)These pyramids grow slowly and reach the surface of amorphous layer. Its thickness becomes smaller at the same time. Finally, amorphous areas between pyramids also transform in crystalline perovskite oxide phase, and flat facets up to 10 nm in length are formed (fh).
Catalysts 09 00751 g003
Figure 4. HRSTEM images of PCMO (x = 0.33) before (a) and after (b) crystallization under electron beam in oxygen, respectively. (c) EELS spectra (0.1 eV/channel) collected from the areas marked with different colors in (a) and (b). Spectral features marked with A, B, C, L2 and L3 are discussed in the text.
Figure 4. HRSTEM images of PCMO (x = 0.33) before (a) and after (b) crystallization under electron beam in oxygen, respectively. (c) EELS spectra (0.1 eV/channel) collected from the areas marked with different colors in (a) and (b). Spectral features marked with A, B, C, L2 and L3 are discussed in the text.
Catalysts 09 00751 g004
Figure 5. (a) HRSTEM image of PCMO (x = 0.33) with two selected area used to verify the chemical composition of original PCMO material and recrystallized amorphous layer. The corresponding normalized spectra (1 eV/channel) are shown in (b), and the signal from Ca is remarkably weaker at the last unit cell.
Figure 5. (a) HRSTEM image of PCMO (x = 0.33) with two selected area used to verify the chemical composition of original PCMO material and recrystallized amorphous layer. The corresponding normalized spectra (1 eV/channel) are shown in (b), and the signal from Ca is remarkably weaker at the last unit cell.
Catalysts 09 00751 g005
Figure 6. Experimental HRTEM images of PCMO (x = 0.33) collected in N2 (a) and in He (d) at p = 50 ubar; (b,c,e,f) corresponding calculated images along [100] and [010] directions. Both A- and B-cation stable terminations are visible in N2, whereas only the A-cation termination of facets is visible in He.
Figure 6. Experimental HRTEM images of PCMO (x = 0.33) collected in N2 (a) and in He (d) at p = 50 ubar; (b,c,e,f) corresponding calculated images along [100] and [010] directions. Both A- and B-cation stable terminations are visible in N2, whereas only the A-cation termination of facets is visible in He.
Catalysts 09 00751 g006
Figure 7. (a)–(c) HRTEM mages showing transformation of amorphous edge into crystalline for the PCMO (x = 0.1) in He gas at pressure of 50 ubar. The first crystalline facets (marked with the yellow arrow in b) start to appear after ~28 min (1710 s). (d) High-angle annular dark field (HAADF) image of the recrystallized area. (e) EELS spectra (1 eV/channel) demonstrate the presence of Pr, Ca, Mn and O. (f) EELS spectra (0.1 eV/channel) reveal a slightly reduced Mn oxidation state.
Figure 7. (a)–(c) HRTEM mages showing transformation of amorphous edge into crystalline for the PCMO (x = 0.1) in He gas at pressure of 50 ubar. The first crystalline facets (marked with the yellow arrow in b) start to appear after ~28 min (1710 s). (d) High-angle annular dark field (HAADF) image of the recrystallized area. (e) EELS spectra (1 eV/channel) demonstrate the presence of Pr, Ca, Mn and O. (f) EELS spectra (0.1 eV/channel) reveal a slightly reduced Mn oxidation state.
Catalysts 09 00751 g007
Figure 8. Cyclovoltagramms of (001) oriented PCMO (x = 0.1) and (x = 0.33) films in 0.1 M KOH. The x = 0.33 film shows a redox couple which is overlapping to the onset of OER at the anodic branch and well separated from OER at the cathodic branch. This redox couple is absent for PCMO (x = 0.1).
Figure 8. Cyclovoltagramms of (001) oriented PCMO (x = 0.1) and (x = 0.33) films in 0.1 M KOH. The x = 0.33 film shows a redox couple which is overlapping to the onset of OER at the anodic branch and well separated from OER at the cathodic branch. This redox couple is absent for PCMO (x = 0.1).
Catalysts 09 00751 g008
Figure 9. (a) A snapshot of Quadera™ software during environmental transmission electron microscopy (ETEM) experiment at 50 ubars of He with and without activated cold trap. Experimental HRTEM images in He when the cold trap was OFF and ON are shown in (b) and (c), respectively. The white arrows point out on randomly oriented PrO2 particles, growing in amorphous layer (b), and epitaxially growing pyramids of PCMO (x = 0.33) (c).
Figure 9. (a) A snapshot of Quadera™ software during environmental transmission electron microscopy (ETEM) experiment at 50 ubars of He with and without activated cold trap. Experimental HRTEM images in He when the cold trap was OFF and ON are shown in (b) and (c), respectively. The white arrows point out on randomly oriented PrO2 particles, growing in amorphous layer (b), and epitaxially growing pyramids of PCMO (x = 0.33) (c).
Catalysts 09 00751 g009

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Back to TopTop