The E ﬀ ect of Strontium Doping on LaFeO 3 Thin Films Deposited by the PLD Method

: The aim of the presented investigations was to deposit the thin ﬁlms La 1 − x Sr x FeO 3 (x = 0, 0.1, 0.2) on (100) Si substrate by using the Pulsed Laser Deposition (PLD) method. Structure was exanimated by using XRD, SEM, AFM, TEM and XPS methods. The catalytic properties were analyzed in 4 ppm acetone atmosphere. The doping of Sr thin ﬁlms La 1 − x Sr x FeO 3 (x = 0, 0.1, 0.2) resulted in a decrease in the size of the crystallites, the volume of the elemental cell and change in the grain morphology. In the LaFeO 3 and La 0.9 Sr 0.1 FeO 3 , clusters around which small grains grow are visible in the structure, while in the layer La 0.8 Sr 0.2 FeO 3 , the visible grains are elongated. The TEM analysis has shown that the obtained thin ﬁlms had a thickness in the range 150–170 nm with triangular or ﬂat column ends. The experiment performed in the presence of gases allowed us to conclude that the surfaces (101 / 020) in the triangle-shaped columns and the plane (121 / 200) faces in ﬂat columns were exposed to gases. The best properties in the presence of CH 3 COCH 3 gas were noted for LaFeO 3 thin ﬁlm with triangle columns ending with orientation (101 / 020). raction spectroscopy The chemical composition of the by XPS. The XPS spectra were recorded using a VSW instrument, equipped with a concentric hemispherical 1500 mm electron analyzer and a two-plate 18-channel detector (Galileo). Elemental composition and the chemical states of the elements across the thin coatings were investigated by the X-ray Photoelectron Spectroscopy (XPS) coupled with depth proﬁling by argon ion sputtering. XPS analyses were performed by using the PHI Versa Probe II instrument equipped with scanning electron Al anode X-ray source (Al K α 1,2 with an energy equal to 1486.6 eV) with crystal monochromator. Pass energy of the electrons was equal to 47 eV and the take-o ﬀ angle was 45 ◦ . Binding energy scale was calibrated assuming that the binding energy of electrons of the C 1s level in C–H bonding of the adventitious hydrocarbons (surface contamination) is equal to 284.8 eV. Charge neutralization of the samples was achieved using the argon ion gun and an electron ﬂood gun simultaneously. The ﬁrst XPS analysis of each sample was made on the as-received coating surface. Then, the samples were sputtered (ion etching) several times using the Ar + ion beam for 2 min per 1 cycle and after each sputtering cycle they were immediately analyzed using XPS. The cycles of sputtering and subsequent XPS analyses were repeated until the substrate of the coating was reached. Each cycle of ion sputtering was carried out under the same parameters: in the triangle-shaped columns and the plane (121 / 200) faces in ﬂat columns. The best properties in the presence of CH 3 COCH 3 gas were noted for the LaFeO 3 thin ﬁlm. They were characterized by a fast response time to the presence of 4 ppm acetone and surface regeneration. The LaFeO 3 thin ﬁlm has a grain morphology that ensures higher adsorption and desorption of oxygen; hence, the response times and sensitivity of this ﬁlm are higher than in the case of the La 0.9 Sr 0.1 FeO 3 ﬁlm. These results are a consequence of di ﬀ erent column ends (triangular in LaFeO 3 and ﬂat in La 0.9 Sr 0.1 FeO 3 ). Therefore, the crystallographic orientation of columns involved in gas reactions has a decisive role.


Introduction
Perovskite oxides denoted as ABO 3 are widely studied due to their double-type ionic and electronic conductivity. Perovskites are useful according to their functional properties, showing great potential in several applications such as gas sensors [1][2][3], automotive exhaust catalyst and methane reformers that produce syngas [4][5][6] or cathodes in solid oxide fuel cells (SOFCs) [7,8].
Properties of perovskite oxides result from a large number of cations which can be included in the structure such as A=La, Ca, Sr and B=Ti, Fe, Nb, Ta, Co and Mg, from possible cationic and/or anionic substitutions in this structure (vacancy) and finally from the creation of structural defects caused by cations or oxygen deficiency [9]. LaFeO 3 is a compound which may be classified to the group of perovskites. LaFeO 3 is a p-type semiconductor with orthorhombic structure type Pmna (No. 62) at room temperature. This perovskite is as famous antiferromagnetic materials with high Néel temperature TN = 740 • C. The electrical properties of LaFeO 3 are derived from the ordering antiferroelectric and ferroelectric dipole moments [10]. LaFeO 3 is widely used in the fields of electrical, magnetic sensors and gas sensors due to its chemical stability at elevated temperatures and mixed ion conductivity [10][11][12].
This perovskite can be considered as a potential material for the gas sensors application. Generally, materials used for gas sensors have to present high sensitivity, selectivity, stability, low humidity effect, low detection limit and fast response/recovery times. While using this material in thin film

Structure and Morphology Characterization of Thin Films
The XRD analysis was carried out in Grazing geometry at a constant angle of incidence α = 1° ( Figure 2). The depth of penetration was z = 54 nm. The identification of phases was based on the JCPDS base card numbers 01-088-0641 for LaFeO3, 04-007-6515 for La0.9Sr0.1FeO3 and 04-007-6516 for La0.8Sr0.2FeO3. The analysis of the diffraction patterns did not reveal the presence of other phases such as La2O3, Fe3O4 and SrCoO3 ( Figure 2). Peak positions are consistent with the literature data [18]. As for XRD analysis in targets the peaks in doped thin films are shifted to the right in relation to the peaks in LaFeO3 target. This is due to the substitution of La by Sr and the changing of cell parameters.
To characterize the effect of doping on the crystallographic structure of the La1−xSrxFeO3 thin films, the crystallographic parameters were determined using the CelRef software (Table 1).

Structure and Morphology Characterization of Thin Films
The XRD analysis was carried out in Grazing geometry at a constant angle of incidence α = 1 • (Figure 2). The depth of penetration was z = 54 nm. The identification of phases was based on the JCPDS base card numbers 01-088-0641 for LaFeO 3 , 04-007-6515 for La 0.9 Sr 0.1 FeO 3 and 04-007-6516 for La 0.8 Sr 0.2 FeO 3 . The analysis of the diffraction patterns did not reveal the presence of other phases such as La 2 O 3 , Fe 3 O 4 and SrCoO 3 ( Figure 2). Peak positions are consistent with the literature data [18].

Structure and Morphology Characterization of Thin Films
The XRD analysis was carried out in Grazing geometry at a constant angle of incidence α = 1° ( Figure 2). The depth of penetration was z = 54 nm. The identification of phases was based on the JCPDS base card numbers 01-088-0641 for LaFeO3, 04-007-6515 for La0.9Sr0.1FeO3 and 04-007-6516 for La0.8Sr0.2FeO3. The analysis of the diffraction patterns did not reveal the presence of other phases such as La2O3, Fe3O4 and SrCoO3 ( Figure 2). Peak positions are consistent with the literature data [18]. As for XRD analysis in targets the peaks in doped thin films are shifted to the right in relation to the peaks in LaFeO3 target. This is due to the substitution of La by Sr and the changing of cell parameters.
To characterize the effect of doping on the crystallographic structure of the La1−xSrxFeO3 thin films, the crystallographic parameters were determined using the CelRef software (Table 1). As for XRD analysis in targets the peaks in doped thin films are shifted to the right in relation to the peaks in LaFeO 3 target. This is due to the substitution of La by Sr and the changing of cell parameters.
To characterize the effect of doping on the crystallographic structure of the La 1−x Sr x FeO 3 thin films, the crystallographic parameters were determined using the CelRef software (Table 1). It was observed that with the increase in the content of strontium, the volumes of the elementary cell have decreased. The observed changes in the crystallographic parameters and elementary cell size together with the increase in doping are not caused by La 3+ doping with Sr 2+ strontium due to the similar size of La 3+ (0.13 nm) and Sr 2+ ionic (0.14 nm) rays. Strontium doping leads to the formation of iron in the oxidation state from Fe 3+ to Fe 4+ in the doped thin films which in turn is a consequence of necessity to preserve the electroneutrality. As a result of the formation Fe 4+ the oxygen vacancies are generated. The differences in the ionic radius of iron in different oxidation states explain the decrease in the cell volume, Fe 4+ (0.585 Å) and Fe 3+ (0.645 Å). Similar effects were noted in the system LaFeO 3 doped by Ca and Ti ions, and the unit cell volumes were decreased compared to LaFeO 3 [15,19].
Doping has caused a decrease in the size of crystallites from 36 nm for LaFeO 3 to 18 nm for the La 0.8 Sr 0.2 FeO 3 thin films (Table 1). In nanocrystalline materials such as thin films, interaction between defects (dopants) and grain boundary (GB) plays a key role in determining the structure stability. The interaction between the dopants and surface/grain boundaries may decrease surface energy/grain boundary energy. This leads to the stabilization of the surfaces/grain boundaries [20]. In this case, under the same preparation conditions are more surfaces/grain boundaries are observed. It can, therefore, be concluded that the grain size can decrease with the dopants.
The Rt parameter was determined as the ratio of the intensity of the peak I121/200 to I101/020 (Table 1) after the subtracting the background. The calculations have shown that Rt increases with increasing of Sr content, which indicates the decrease of the I101/020 peak intensity and the increase of the I121/200 peak intensity. The change in intensity of these two peaks suggested preferred orientation.
SEM studies have shown few droplets on the surface of all thin layers ( Figure 3). The cracks visible on Figure 3c may be caused by too high a cooling rate after the PLD process. Cracks are also one of the characteristics of the PLD method. In the deposited thin films tensile stress may appear to GPa [21]. They relaxation during the cooling process can produce the cracks in thin films.  As can be seen the surface of the thin films were built from very fine irregularities (bulges) that constitute the proper thin film. The Ra parameter determined from the tests increases with increasing content of the dopants in the thin films ( Table 1).
The results of AFM confirm the crystalline refinement observed in XRD studies. XPS analyses coupled with the Ar + ion sputterings were used to investigate the oxidation state of elements on the surface and inside the La1−xSrxFeO3 thin films at different depths, as well as at the thin film/substrate. Visible in the Figures 4-7, the cycles correspond to: cycle 1-surface of the thin films; cycle 2-19-thin films; cycle 22-25-surface of the substrate.
On the surface of all thin films the presence of an adventitious carbon was detected. For all thin films after the first sputtering process the spectral line of carbon has disappeared completely ( Figure  4). The analysis revealed the C 1s line on the surface (cycle 1) and its absence at the deeper layers across the thin film till the substrate (cycles 2-19) (Figure 4). The presence of the carbon on the surface thin films is a consequence of the contaminant's adsorption from the environment. Chemical states of the thin film and substrate elements were investigated by registering the following spectral lines: O1s, Fe 2p, La 3d, Sr 3d and Si 2s.  Figure 3e). The structure of flat grains in the vicinity of elongated grains was observed in LaFeO 3 samples produced by magnetron sputtering with different deposition temperature [9]. Similar differences in the structure were observed in strontium doped LaCoO 3 layers [22]. Changes in the structure may be influenced by the temperature of the substrate, the gas pressure in the chamber or doping. The variations in growth kinetics due to the doping process may induce the changes in the structure.
The topography studies of all thin films (Figure 3b,d,f) performed with AFM (tapping mode in air) are in a good agreement with the results obtained at larger scale from SEM analysis ( Figure 3). As can be seen the surface of the thin films were built from very fine irregularities (bulges) that constitute the proper thin film. The Ra parameter determined from the tests increases with increasing content of the dopants in the thin films ( Table 1).
The results of AFM confirm the crystalline refinement observed in XRD studies. XPS analyses coupled with the Ar + ion sputterings were used to investigate the oxidation state of elements on the surface and inside the La 1−x Sr x FeO 3 thin films at different depths, as well as at the thin film/substrate. Visible in the Figures 4-7, the cycles correspond to: cycle 1-surface of the thin films; cycle 2-19-thin films; cycle 22-25-surface of the substrate.
Catalysts 2020, 10, x FOR PEER REVIEW 6 of 14 Figure 4 shows the O 1s line registered on the surface (cycle 1) which was fitted by two lines at 529.22 and 531.93 eV. The peak at BE = 531.93 eV presence is a consequence of the impurities from the surface and is responsible for bonds in C=O organic compounds. The second at BE = 529.22 eV is interpreted as a component of non-metallic oxides. The main pic at BE = 529.8 eV (cycle [4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] is typical for the binding energy of oxygen with rare earth metals like La [23]. The Si 2s spectrum is shown in Figure 4. The peak at BE = 148.9 eV is visible in cycle 22-25. The energy at which the peak occurs is typical for SiO2 and comes from the phase formed of the Si substrate with oxygen. Fe 2p lines have a main peak and a satellite ( Figure 5). These satellites are separated from the main peak by 4 eV. The main peak Fe 2p3/2 is at BE = 706.3 eV and Fe 2p1/2 at BE = 720.12 eV. The determined values are much below those observed in Fe-O oxides. If the iron is present in the third oxidation stage, the peak for Fe 2p3/2 is in the range of BE = 706-711.2 eV. The peak at such a low energy was observed in perovskites in the LaFeO3 produced by the sol-gel method. The presence of the satellite at Fe 2p1/2 is due to the presence of iron on the fourth oxidation state [24]. The intensity of the satellites Fe2p3/2 and Fe2p1/2 have increased with the increasing of the dopant, which indicates a higher Fe 4+ content. The shapes and positions of the La 3d and Sr 3d (Figures 6 and 7) spectra did not change across the thin film and remained the same at the thin film/substrate interface. This meant that the chemical state of the lanthanum and strontium did not change even at the interface.   The shapes and positions of the La 3d and Sr 3d (Figures 6 and 7) spectra did not change across the thin film and remained the same at the thin film/substrate interface. This meant that the chemical state of the lanthanum and strontium did not change even at the interface.    On the surface of all thin films the presence of an adventitious carbon was detected. For all thin films after the first sputtering process the spectral line of carbon has disappeared completely ( Figure 4). The analysis revealed the C 1s line on the surface (cycle 1) and its absence at the deeper layers across the thin film till the substrate (cycles 2-19) (Figure 4). The presence of the carbon on the surface thin films is a consequence of the contaminant's adsorption from the environment. Chemical states of the thin film and substrate elements were investigated by registering the following spectral lines: O1s, Fe 2p, La 3d, Sr 3d and Si 2s. Figure 4 shows the O 1s line registered on the surface (cycle 1) which was fitted by two lines at 529.22 and 531.93 eV. The peak at BE = 531.93 eV presence is a consequence of the impurities from the surface and is responsible for bonds in C=O organic compounds. The second at BE = 529.22 eV is interpreted as a component of non-metallic oxides. The main pic at BE = 529.8 eV (cycle 4-19) is typical for the binding energy of oxygen with rare earth metals like La [23].
The Si 2s spectrum is shown in Figure 4. The peak at BE = 148.9 eV is visible in cycle 22-25. The energy at which the peak occurs is typical for SiO 2 and comes from the phase formed of the Si substrate with oxygen.
Fe 2p lines have a main peak and a satellite ( Figure 5). These satellites are separated from the main peak by 4 eV. The main peak Fe 2p 3/2 is at BE = 706.3 eV and Fe 2p 1/2 at BE = 720.12 eV. The determined values are much below those observed in Fe-O oxides. If the iron is present in the third oxidation stage, the peak for Fe 2p 3/2 is in the range of BE = 706-711.2 eV. The peak at such a low energy was observed in perovskites in the LaFeO 3 produced by the sol-gel method. The presence of the satellite at Fe 2p 1/2 is due to the presence of iron on the fourth oxidation state [24]. The intensity of the satellites Fe2p 3/2 and Fe2p 1/2 have increased with the increasing of the dopant, which indicates a higher Fe 4+ content.
The shapes and positions of the La 3d and Sr 3d (Figures 6 and 7) spectra did not change across the thin film and remained the same at the thin film/substrate interface. This meant that the chemical state of the lanthanum and strontium did not change even at the interface.
The The deposited thin films have a columnar structure with a thickness in the range 150-170 nm (Figures 8-10). The deposited thin films have a columnar structure with a thickness in the range 150-170 nm (Figures 8-10).   The deposited thin films have a columnar structure with a thickness in the range 150-170 nm (Figures 8-10).   The deposited thin films have a columnar structure with a thickness in the range 150-170 nm (Figures 8-10).     In the other thin films, the columns are flat (Figures 9 and 10). The different shape of the columns indicates a different crystallographic orientation of the columns exposed to gases. In the research presented by M. Jędrusik et al. [25], it has been demonstrated that the surfaces exposed to gases in LaFeO3 columnar with triangle shape are (101/020) (Figure 7), and for flat-ended columns it is the plane (121/200). Taking this into account, the results of XRD presented in this paper, as well as the results of M. Jędrusik at al. [21], it can be concluded that the columns in thin films have growth in the direction (121/200).

Gas-Sensing Properties
LaFeO3 thin films are typical p-type semiconductors. The gas sensing mechanism is for those materials based on changes of the resistance while contact with gases. During the exposed La1−xSrxFeO3 sensors to the air, the oxygen is chemically adsorbed on the surface. It can be presented by the following reaction equations [26][27][28]:

O gas ↔ O ads O ads + e ↔ O ads O ads + 2e ↔ 2O ads
It leads to the formation of a thin space-charge layer which decreases the potential barrier. The resistance of the thin films is low. When the thin films are exposed to the acetone the reactions take place on the surface as follows: This process results in the thickening of the space-charge layer, thus increasing the potential barrier and decreasing the current.
The electrons trapped by the adsorbed oxygen are released and the numbers of electrons are annihilated by the holes (h*): In the other thin films, the columns are flat (Figures 9 and 10). The different shape of the columns indicates a different crystallographic orientation of the columns exposed to gases. In the research presented by M. Jędrusik et al. [25], it has been demonstrated that the surfaces exposed to gases in LaFeO 3 columnar with triangle shape are (101/020) (Figure 7), and for flat-ended columns it is the plane (121/200). Taking this into account, the results of XRD presented in this paper, as well as the results of M. Jędrusik at al. [21], it can be concluded that the columns in thin films have growth in the direction (121/200).

Gas-Sensing Properties
LaFeO 3 thin films are typical p-type semiconductors. The gas sensing mechanism is for those materials based on changes of the resistance while contact with gases. During the exposed La 1−x Sr x FeO 3 sensors to the air, the oxygen is chemically adsorbed on the surface. It can be presented by the following reaction equations [26][27][28]: It leads to the formation of a thin space-charge layer which decreases the potential barrier. The resistance of the thin films is low. When the thin films are exposed to the acetone the reactions take place on the surface as follows:  [29].
Considering the above we expected the increasing of the sensitivity with the dopants. For the thin films, La 1−x Sr x FeO 3 resistance tests were carried out in an acetone atmosphere (4 ppm concentration) at different operating temperatures T = 303, 370, 437 • C. The measurements were carried out by using a specially constructed apparatus for measuring the gas sensor response [30]. All the process was controlled by the dedicated application LabView.
Considering the above we expected the increasing of the sensitivity with the dopants. For the thin films, La1−xSrxFeO3 resistance tests were carried out in an acetone atmosphere (4 ppm concentration) at different operating temperatures T = 303, 370, 437 °C. The measurements were carried out by using a specially constructed apparatus for measuring the gas sensor response [30]. All the process was controlled by the dedicated application LabView.
Before measurements, all samples were heated at the initial temperature of T = 230 °C for several hours. During this time, the samples were stabilized. Under stabilized gas chamber conditions (humidity 50%, gas flow, constant working temperature) and when the sensor resistance displayed no drifts, the resistances were measured in air and acetone gas ( Figure 12). For each of the temperatures, the air and the test gas were alternately exchanged in such a way as to maintain a constant gas flow. Figure 12 shows the changes in the resistance of the thin films LaFeO3 and La0.9Sr0.1FeO3 for several operating temperatures in alternating air atmospheres and 4 ppm acetone, with constant gas flow.
Reduction of thin film resistance while increasing the temperature was observed. This phenomenon is characteristic for semiconductor materials. The response value (S) of the sensor was defined as S = Rg/Ra, where Ra and Rg represent the sensor's resistance in the air-Ra and the presence of the gases Rg.
The response time (tres) was the time taken from Ra to Ra+90% ΔR (ΔR= Rg − Ra) under the gas environment and the recovery time (trec) was the time taken from Rg to Rg-90% * ΔR in the air condition [28]. For each of the temperatures, the air and the test gas were alternately exchanged in such a way as to maintain a constant gas flow. Figure 12 shows the changes in the resistance of the thin films LaFeO 3 and La 0.9 Sr 0.1 FeO 3 for several operating temperatures in alternating air atmospheres and 4 ppm acetone, with constant gas flow.
Reduction of thin film resistance while increasing the temperature was observed. This phenomenon is characteristic for semiconductor materials. The response value (S) of the sensor was defined as S = R g /R a , where R a and R g represent the sensor's resistance in the air-R a and the presence of the gases R g .
The response time (tres) was the time taken from R a to R a + 90% ∆R (∆R = R g − R a ) under the gas environment and the recovery time (trec) was the time taken from R g to R g -90% * ∆R in the air condition [28].
The response value (S) of the sensor depends on the operating temperature and is higher in lower temperature. For these thin films the sensitivity was determined in the temperature T = 303 • C for S LaFeO 3 = 3.75 and for S La 0.9 Sr 0.1 FeO 3 = 1.77.
Due to the faster reaction rate at higher temperature, the response times and regeneration time are shortened with increasing temperature. The measured response time to gas at T = 303 • C for LaFeO 3 was t res = 54 s, regeneration time t rec = 66 s and for La 0.9 Sr 0.1 FeO 3 t res = 180 s and t rec = 258 s. Enhanced response time and decreased sensitivity while doping were proved in the research.
A similar effect was observed in the LaFeO3 powders doped with Ag. The contents of 10% and 20% Ag have shown a decrease in sensitivity and the increase only in samples doped by 30% Ag [27].
The presented results are closely related to the structure of thin films. Different crystallographic orientations have different atom arrangements, different electronic properties, such as surface state density, adsorption/desorption energy of interaction with gases and concentration of O ads on the surface.
The research presented by X. Wang et al. [31] allows us to conclude that in LaFeO 3 layers, type surfaces (010) are preferred (more adsorbed oxygen) in adsorption and oxygen desorption processes which determine catalytic properties. In the presented investigations LaFeO 3 thin films dominated a columnar with a triangle shape. The triangular ends of the columns have crystallographic orientation (020) and (101) (Figure 8).
This morphology ensured greater adsorption and desorption of the oxygen. The reaction times and sensitivity of these thin films are higher than in the case of La 0.9 Sr 0.1 FeO 3 which has a flat end of a columnar with orientation (121) and (200). In this case the adsorbed oxygen species reacting with gas molecules is slower than on (020) crystallographic orientation.
With this in mind, crystallographic orientation plays a significant role in reactions with gas. Therefore, despite the use of doping, which caused an increase in the number of vacancies, the sensitivity of thin layers was not increased.

Deposition of the Thin Films
The thin films La 1−x Sr x FeO 3 (x = 0, 0.1, 0.2) were deposited on the substrate (100) oriented Si. The targets with different amounts of strontium were obtained from the Kurt Lesker Company (Hastings, East Sussex, UK). Each target had a dimension d = 1 cal. For deposition of the thin films a laser ablation system Nd-YAG laser Continuum Powerlite DLS (λ = 266 nm), with chamber Neocera was used (Neocera 10000 Virginia Manor Road Ste 300, Beltsville, Maryland, MD, USA). The characteristics of the deposition system were as indicated: a target-substrate distance of 70 mm, the air pressure in the deposition camera was p = 5 Pa. The laser beam hits the target at an incidence angle ξ = 45 • . Substrates and targets were parallel. The deposition conditions were a frequency of f = 10 Hz, energy density on the target ε = 3 J/cm 2 , substrate heated at 570 • C and deposition time t = 1.5 h.

Characterization of the Thin Films and Targets
The phase analysis of the target and thin films were performed by the X-ray diffraction phase analysis method (XRD) using PANanalytical EMPYREAN DY 1061 (Malvern Panalytical Ltd., Malvern, UK) with Cu Kα (λCu = 0.154 nm) radiation in Braga-Brentano and Grazing geometry. The identification of the determined phases was based on the database PDF-4+ product of the ICDD used. For the calculation of the cell parameters software CellRef were applied while crystalline size was calculated using Williamson-Hall (W-H) analysis. The surface morphology and the chemical composition of the target and thin films were observed using a Scanning Electron Microscope (FEI Nova NanoSEM 450, Hillsboro, Oregon, OR, USA) equipped with an Energy Dispersive Spectroscopy Detector (EDAX). The topography of the surface was investigated by using AFM microscopy Veeco Dimension ® Icon™ SPM with NanoScope V (Veeco, Munich, Germany). Transmission electron microscopy (TEM) analyses were carried out by using JEOL JEM-2010 ARP (Jeol company Pleasanton, CA 94588, USA). The TEM investigations were performed on a cross-section of the lamellas obtained from the desired area of the analyses samples by the Focused Ion Beam (FIB) technique using Quanta 3D 200i (FEI company, Hillsboro, Oregon, OR, USA) equipped with an OmniProbe lift-out system. Phase identification was performed by means of selected area electron diffraction (SAED) and it was supplemented by the energy dispersive X-ray spectroscopy (EDS). The chemical composition of the films was studied by XPS. The XPS spectra were recorded using a VSW (Vacuum Systems Workshop Ltd., ULVAC-PHI, Chigasaki, Japan) instrument, equipped with a concentric hemispherical 1500 mm electron analyzer and a two-plate 18-channel detector (Galileo). Elemental composition and the chemical states of the elements across the thin coatings were investigated by the X-ray Photoelectron Spectroscopy (XPS) coupled with depth profiling by argon ion sputtering. XPS analyses were performed by using the PHI Versa Probe II instrument equipped with scanning electron Al anode X-ray source (Al Kα1,2 with an energy equal to 1486.6 eV) with crystal monochromator. Pass energy of the electrons was equal to 47 eV and the take-off angle was 45 • . Binding energy scale was calibrated assuming that the binding energy of electrons of the C 1s level in C-H bonding of the adventitious hydrocarbons (surface contamination) is equal to 284.8 eV. Charge neutralization of the samples was achieved using the argon ion gun and an electron flood gun simultaneously. The first XPS analysis of each sample was made on the as-received coating surface. Then, the samples were sputtered (ion etching) several times using the Ar+ ion beam for 2 min per 1 cycle and after each sputtering cycle they were immediately analyzed using XPS. The cycles of sputtering and subsequent XPS analyses were repeated until the substrate of the coating was reached. Each cycle of ion sputtering was carried out under the same parameters: the Ar + energy-2.5 keV, ion current-3.2 nA and the sputtered area −1 × 1 mm 2 . During the XPS analysis the X-ray microbeam scanned only 500 × 500 µm 2 of the central area of the ion sputtered crater in order to avoid the edge effects. As the sputtering yield was not known the depth scale was expressed in minutes of spraying. The catalytic properties in 4 ppm acetone atmosphere were carried out in a 30 cm 3 gas chamber. The changes of the temperature in the range T = 313-437 • C were measured using a Pt 100 sensor and a digital multimeter Agilent 34970A (Agilent, Santa Clara, California, CA, USA). To measure the change in the resistivity of the samples, a Keithley 6517 electrometer (Keithely, Cleveland, Ohio, OH, USA) was used, operating in the internal voltage source mode. The control device was equipped with LabVIEW software. The measurement system also included a gas flow metering and control system, and a gas atmosphere with a defined composition and humidity (pump, gas cylinder, humidifier, flow controllers with the controller, etc.).

Conclusions
On the grounds of the studies, it was concluded that the doping of Sr layers La 1−x Sr x FeO 3 (x = 0. 0.1, 0.2) result in a decrease in the size of the crystallites and the volume of the elemental cells. The reduction in the size of the elementary cell has been explained by the behavior of electroneutrality and the formation of Fe 4+ . It was observed that the layers of La 1−x Sr x FeO 3 grow in the preferred orientation (121/200), and, as well, the change in the grained morphology was also denoted. Those phenomena are closely connected with the doping of the layers. In the case of LaFeO 3 and La 0.9 Sr 0.1 FeO 3 , clusters around which small grains grow are visible in the structure, while in the layer of x = 0.2 of the content of strontium, the visible grains are elongated. The change in morphology can be explained by the change in the kinetics of the layer growth as a result of doping. Similar effects were observed in the LaFeO 3 layers as a function of the temperature [22]. The roughness parameters increase with the increased Sr content, which was also confirmed by the results of XRD as reducing the size of crystallites. The TEM analyses have shown that the obtained thin films had a thickness in the range 150-170 nm with triangular or flat column ends. For exposed to gases are the surfaces (101/020) in the triangle-shaped columns and the plane (121/200) faces in flat columns. The best properties in the presence of CH 3 COCH 3 gas were noted for the LaFeO 3 thin film. They were characterized by a fast response time to the presence of 4 ppm acetone and surface regeneration. The LaFeO 3 thin film has a grain morphology that ensures higher adsorption and desorption of oxygen; hence, the response times and sensitivity of this film are higher than in the case of the La 0.9 Sr 0.1 FeO 3 film. These results are a consequence of different column ends (triangular in LaFeO 3 and flat in La 0.9 Sr 0.1 FeO 3 ). Therefore, the crystallographic orientation of columns involved in gas reactions has a decisive role.