Study of gadolinium substitution effects in hexagonal yttrium manganite YMnO3

In the present work, gadolinium substitution effects on the properties of yttrium manganite YxGd1−xMn0.97Fe0.03O3 (x from 0 to 1 with a step of 0.2) synthesized by an aqueous sol–gel method have been investigated. Partial substitution of Mn3+ by 57Fe3+ in the manganite was also performed in order to investigate deeper the structural properties of synthesized compounds applying Mössbauer spectroscopy. It was demonstrated that substitution of Y3+ by Gd3+ ions leads to the changes of structural, magnetic and morphological properties of investigated system. The crystal structure gradually transformed from hexagonal to orthorhombic with an increase of Gd3+ content in the crystal lattice. The mixed phase was obtained when x = 0.6, whereas other compounds were determined to be monophasic. Magnetization measurements revealed paramagnetic behavior of all specimens, however magnetization values were found to be dependent on chemical composition of the samples. Solid solutions with orthorhombic structure revealed higher magnetization values compared to those of hexagonal samples. The highest magnetization was observed for pure GdMn0.97Fe0.03O3. Structural properties were investigated by powder X-ray diffraction, Mössbauer, FTIR and Raman spectroscopies. Morphological features of the synthesized specimens were studied by scanning electron microscopy (SEM).

www.nature.com/scientificreports/ which was characterized by enhanced ferromagnetic properties 24 . Furthermore, this compound has high magnetoelectric coefficient and possesses pyroelectric properties 25,26 . For the preparation of GdMnO 3 compound sol-gel 27 , solid-state 28 , sol-gel combustion 29 and co-precipitation 30 methods were previously applied.
There are few studies considering synthesis and characterization of Y x Gd 1−x MnO 3 system [31][32][33][34][35] . Vilarinho et al. 33 and Bos et al. 35 demonstrated the formation of Y x Gd 1−x MnO 3 solid solutions in a whole compositional range, while others investigated dielectric, magnetic and ferroelectric properties of Gd-rich orthorhombic compounds with x ≤ 0.4 31,32,34 . It was observed that partial substitution of Gd 3+ by Y 3+ can lead to the stabilization of ferroelectric phase when x = 0.1. Furthermore, when x > 0.1 weak ferromagnetic character disappears and antiferromagnetic ordering is established 34 . The monophasic Y x Gd 1−x MnO 3 and related systems could be employed to study and understand the charge transport across the interfaces, to investigate deeper spin-disorder state near nonmagnetic impurities, to determine substitutional-driven structure-ferroelectricity relationship and other fundamental physical properties of such complex oxides. These new knowledges obtained could be used in future for the construction of specific functionalities in novel ferroelectrics.
In this work, the solid solutions of Y x Gd 1−x Mn 0.97 Fe 0.03 O 3 (x = 0-1 with a step of 0.2) with partial substitution of Mn 3+ by Fe 3+ (or by 57 Fe 3+ ) ions were prepared for the first time by our best knowledge using environmentally friendly and simple sol-gel technique. Partial substitution of Mn 3+ by 57 Fe 3+ in the manganites was performed in order to investigate deeper the structural and magnetic properties of synthesized compounds applying Mössbauer spectroscopy. The dependence and evolution of structural, magnetic and morphological properties on chemical composition were investigated and discussed herein.

Materials and methods
Synthesis. Synthesis of all samples was performed by sol-gel method using modified previously reported procedure 36 . For the preparation of Y x Gd 1−x Mn 0.97 Fe 0.03 O 3 series by changing x from 0 to 1 with a step of 0.2, yttrium (III) nitrate hexahydrate (Y(NO 3 ) 3 ·6H 2 O, Sigma-Aldrich, 99.9%), gadolinium (III) nitrate hexahydrate (Gd(NO 3 ) 3 ·6H 2 O, Sigma-Aldrich, 99.99%), manganese (II) nitrate tetrahydrate (Mn(NO 3 ) 2 ·4H 2 O, Alfa Aesar, 99.9%) and iron powders (Fe, Carl Roth, 99.5%) were used as the starting materials. Firstly, iron powders were dissolved in 6 M nitric acid (HNO 3 , Carl Roth, 65%) and citric acid monohydrate (C 6 H 8 O 7 ·H 2 O, Chempur, 99.9%) was separately dissolved in 20 ml of distilled water. After the dissolution of citric acid, all metal nitrates and required aliquot of iron solution were added. Next, the obtained mixture was heated on a hot plate at 90 °C under constant stirring until a clear and transparent solution was obtained. After it, an appropriate amount of ethylene glycol (C 2 H 6 O 2 , Sigma-Aldrich, ≥ 99.5%) was added to the above solution (total metal ions to citric acid to ethylene glycol molar ratio was 1:3:10). The obtained liquid precursor was homogenized under constant stirring at 90 °C for 1.5 h. For the formation of the gel the temperature of magnetic stirrer was increased to 150 °C, which led to the evaporation of water. The resulted gel was dried in the oven at 140 °C for 12 h, ground in agate mortar and annealed at 1100 °C for 5 h in air with a heating rate of 5 °C/min. Identical synthesis procedure was applied for the preparation of Y x Gd 1−x Mn 0.97 Fe 0.03 O 3 (x = 0.0, 0.4 and 1.0) samples with 57 Fe. These specimens were used only for Mössbauer spectroscopy measurements.
Characterization. Thermal decomposition of precursor gels was investigated by thermogravimetric and differential scanning calorimetric (TG/DTG/DSC) analysis using PerkinElmer STA 6000 Simultaneous Thermal Analyzer. About 5-10 mg of dried sample was heated from 30 °C to 900 °C at 10 °C/min heating rate in a dry flowing air (20 mL/min). X-ray diffraction (XRD) analysis was performed with Rigaku Miniflex II diffractometer using a primary beam Cu Kα radiation (λ = 1.541838 Å) in 2θ range from 10° to 70° with a step of 0.02° and scanning speed of 2°/min. The obtained diffraction data were refined by the Rietveld method using the Fullprof suite. PerkinElmer FT-IR spectrometer was used for FT-IR analysis of compounds. All spectra were recorded at ambient temperature in the range of 4000-400 cm −1 . Raman spectra were recorded using inVia Raman (Renishaw, United Kingdom) spectrometer equipped with thermoelectrically cooled (− 70 °C) CCD camera and microscope. Raman spectra were excited with 532 nm beam from the CW diode pumped solid state (DPSS) laser (Renishaw, UK). The laser power at the sample was restricted to 0.6 mW to avoid laser-induced sample heating and photodegradation. The 20×/0.40 NA objective was used during all the measurements. The overall integration time was 400 s. Position of the Raman bands on the wavenumber axis was calibrated by the polystyrene film standard spectrum. Parameters of the bands were determined by fitting the experimental spectra with Gaussian-Lorentzian shape components using GRAMS/Al 8.0 (Thermo Scientific, USA) software. The morphology of samples was investigated using a scanning electron microscope (SEM) Hitachi SU-70. Grain size distribution was estimated from SEM micrographs using ImageJ software. Magnetometer consisting of the lock-in amplifier SR510 (Stanford Research Systems), the gauss/teslameter FH-54 (Magnet Physics) and the laboratory magnet supplied by the power source SM 330-AR-22 (Delta Elektronika) was used to record magnetization dependences on applied magnetic field. Mössbauer spectra were measured using 57 Co(Rh) source and Mössbauer spectrometer (Wissenschaftliche Elektronik GmbH) at room (≈296 K) temperature and within 10-70 K temperature range. Closed cycle He cryostat (Advanced Research Systems) was applied for low temperature measurements. The doublets, sextets, quadrupole splitting and hyperfine field distributions, and Hamiltonian method were used to fit to Mössbauer spectra applying WinNormos Site and Dist software. Combined quadrupole and magnetic dipole interactions were described using Hamiltonian method applying the parameters: hyperfine field B, term (main component) of quadrupole interaction eQV zz /2 , asymmetry parameter η, and the angles θ and φ. Q is nuclear quadrupole moment and V zz is z component of electric field gradient (EFG) diagonalized tensor choosing the principal axis system so that |V zz | |V xx | V yy 37 . The asymmetry parameter η = V xx − V yy /V zz . The angle θ is between magnetization direction and EFG z axis while the angle φ is between magnetization projection into the xy plane and EFG x axis. In case of pure quadrupole or magnetic dipole interactions the param- When the quadrupole shifts ε of sextets lines defined by hyperfine magnetic field B are small they can be approximated by first order correction: The outer lines of sextet shift by + ε, i.e. in different direction than inner four lines shifting by -ε. The influence of quadrupole and magnetic dipole interactions was of comparable strength in case of YMn 0.97 Fe 0.03 O 3 . Therefore, direct solution Hamiltonian was used for YMn 0.97 Fe 0.03 O 3 at low temperature. For Mössbauer measurements 3 mol% of Mn was substituted by Fe (90% enriched with 57 Fe). Because of exchange interactions iron atoms generally reflect magnetic ordering and dynamics of Mn spins in studied manganites [38][39][40][41][42][43] .

Results and discussion
The XRD patterns of Y x Gd 1−x Mn 0.97 Fe 0.03 O 3 samples annealed at 1100 °C are represented in Fig. 1. As was mentioned previously, few possible structures can be observed for YMnO 3 . In our case, high annealing temperature resulted in the formation of hexagonal YMn 0.97 Fe 0.03 O 3 with P6 3 cm space group (#185). All diffraction peaks match very well with standard XRD data of hexagonal YMnO 3 (COD #96-153-3979). The same structure was observed for x = 0.8 sample, only with a slight shift of the peaks to lower 2θ values due to the difference in the ionic radii of Gd 3+ and Y 3+ (ionic radius of Gd 3+ in VII-fold coordination is 1.0 Å and for Y 3+ -0.96 Å) 44 . There was no mixture of hexagonal and orthorhombic structures observed as was suggested before for this composition 35 . With increasing the amount of Gd 3+ the phase transition from hexagonal to orthorhombic structure can be clearly seen. In the XRD pattern of x = 0.6 sample the diffraction peaks belonging to both orthorhombic and hexagonal structures were detected. Employing Rietveld refinement for this sample, the ratio between these structures was www.nature.com/scientificreports/ calculated to be around 1 to 9 (10.2%-hexagonal phase, 89.8%-orthorhombic phase). These results show a significant shift towards orthorhombic structure in comparison with previous study. In 33 Figure S4). No secondary phases were identified in the XRD patterns of all synthesized samples.
Rietveld refinement was performed for all synthesized samples. Calculated cell parameters and cell volumes are summarized in Table 1 and Table S1. It is seen that replacement of Y 3+ by Gd 3+ leads to the increase of unit cell parameters in the whole compositional range. Nearly linear dependence between chemical composition and cell parameters can be observed. The cell volume also increased with an increase of Gd 3+ content. On the other hand, only minimal increase of c parameter in hexagonal structure can be seen.   46 . Furthermore, appearance of two peaks can also be seen for the samples with orthorhombic structure. One of them is centered at 405 cm −1 for all orthorhombic samples, it is associated with Mn-O vibrations 47 . The absorption band at 530 cm −1 is ascribed to O-Mn-O bending mode 48 . The observed spectral changes can be explained by the fact that in hexagonal structure the Mn 3+ ions are located in trigonal bipyramid with the coordination number of 5, whereas in orthorhombic structure Mn 3+ and six O 2− anions form octahedra (coordination number 6) 49 . Raman scattering provides detailed molecular level information on short range arrangement or local symmetry which is difficult to acquire by other techniques. The method is also sensitive to structural distortions.  50 . The most intense band of YMn 0.97 Fe 0.03 O 3 compound visible at 681 cm −1 belongs to very small or zero splitting transverse optical (TO) and longitudinal optical (LO) phonons with A 1 symmetry considering the hexagonal structure. This mode is related to displacement of mainly oxygen atoms 50 . The shoulder at lower wavenumber side near 638 cm −1 is associated with TO-LO phonons of E 1 symmetry 50 . Similarly, two low intensity bands located at 450 cm −1 (displacement of mainly oxygen and Mn atoms) and 405 cm −1 (displacement of mainly oxygen atoms) belong to A 1 and E 1 symmetry TO-LO modes, respectively. The low intensity band near 229 cm −1 can be assigned to E 2 symmetry mode associated mainly with deformation vibration of oxygen and Mn atoms. Finally, the intense low frequency band at 135 cm −1 belongs to E 2 symmetry mode related with motion of heavy Y atom. The low intensity feature near 1277 cm −1 is associated with overtone of oxygen stretching vibration 51 . Raman spectrum of GdMn 0.97 Fe 0.03 O 3 differs considerably comparing with YMn 0.97 Fe 0.03 O 3 (Fig. 3). Detailed assignments of Raman bands of orthorhombic perovskite GdMnO 3 are provided in the publications of Illiev et al. 52 and Oliveira et al. 53 . Thus, the most intense band at 609 cm −1 belongs to B 2g (1) Jahn-Teller symmetry in-phase oxygen stretching mode. The A g (1) symmetry MnO 6 bending mode is located at 503 cm −1 . The strong band at 484 cm −1 is associated with A g (3) Jahn-Teller asymmetric stretching vibration of oxygen atoms. Finally, the well-defined band at 368 cm −1 belongs to A g (4) symmetry mode. This mode is associated with out-of-phase rotation of MnO 6 octahedra 52 . The broad and low intensity band at 1301 cm −1 is related with overtone of oxygen stretching vibration 51 . However, drastic spectral changes take place after an additional introduction of Gd 3+ to the level corresponding to Y 0.6 Gd 0.4 MnO 3 composition. Characteristic vibrational bands of hexagonal YMnO 3 completely disappeared. Instead, new bands characteristic to orthorhombic perovskite GdMnO 3 appeared. Peak positions of all the observed bands downshift upon increasing the Gd 3+ content corresponding to x = 0.6. Interestingly, further increase in Gd 3+ amount does not affect the positions of intense bands at 609 or 484 cm −1 (Fig. 4b). However, different behavior was detected for the A g (4) mode near 368 cm −1 (Fig. 4c). Previously, it was demonstrated that phonon frequency of A g (4) mode depends linearly on the MnO 6 hexagon rotation angle; frequency decreases with decreasing the angle 52,54 .  (Fig. 5b,c) show that particles having similar sizes and shape were formed, but slightly narrower size range was observed compared to the undoped sample (derivative maxima were obtained at 384 nm and 415 nm for Y 0.6 Gd 0.4 Mn 0.097 Fe 0.03 O 3 and Y 0.4 Gd 0.6 Mn 0.097 Fe 0.03 O 3 , respectively). Porous structure was also maintained after the Gd was introduced as dopant. The different morphology was observed for the GdMn 0.97 Fe 0.03 O 3 material (Fig. 5d). As seen, GdMn 0.97 Fe 0.03 O 3 powder possesses the smallest grains; and the histogram shows that around 90% of all grains are distributed in 100-500 nm range and around 70% of grains lie in considerably narrower range from 200 to 400 nm (derivative maximum at 320 nm). This shows that with an increase in Gd content the average grain size of Y-Gd-Mn-Fe-O powders becomes smaller and suggests that the surface area and porosity of such ceramic materials could be tailored by changing chemical composition.
Dependence of the magnetization on applied magnetic field strength was studied for all samples and results are presented in Fig. 6. Linear magnetization dependences m = χ H were observed for all Y x Gd 1−x Mn 0.97 Fe 0.03 O 3 solid solutions, which corresponds to paramagnetic state of the materials. Magnetic susceptibility of rare earth manganites is due to both Gd and Mn magnetic moments. The Curie-Weiss law χ mol = N A µ eff 2 (3k B (T − θ )) was used to describe dependence of magnetic (molar) susceptibility, where N A and k B are the Avogadro number and Boltzmann constant on temperature. Application of Curie-Weiss law gave Curie-Weiss temperature θ = -421 K and θ ≈-35 K for YMnO 3 and GdMnO 3 39,55 , respectively. The effective magnetic moment μ eff ≈9.4 μ B of GdMnO 3 was considerably larger than μ eff = 4.98 μ B for YMnO 3 , where μ B is Bohr magneton. Therefore, the inclination magnetization lines increase with amount of Gd 3+ because of both change in θ and μ eff . It can be noticed that more significant decrease in magnetization values was observed along with transformation from orthorhombic structure to hexagonal (between x = 0.6 and 0.8).
Two alternative methods, quadrupole splitting distributions P(∆) with quadrupole splitting ∆ step of 0.1 mm/s and three or four doublets with freely variable parameters, were used for fitting to the room temperature Mössbauer spectra of Y x Gd 1−x Mn 0.97 Fe 0.03 O 3 (Fig. 7, Table 2). Quadrupole splitting distributions P(∆) have peaks approximately at 1.60, 1.58 and 1.95 mm/s for x = 0, 0.4 and 1, respectively. However, the distributions are wide; therefore, they indicate that Fe sites differ significantly by quadrupole splitting. The wider distribution of mixed Y 0.4 Gd 0.6 Mn 0.97 Fe 0.03 O 3 in comparison to other two (Fig. 7b) can be explained by different influence of Y 3+ and Gd 3+ to the local crystal structure. In case of application of separate doublets, the most intense doublet has largest quadrupole splitting ∆ ( Table 2) which value can be explained by significant distortions of bipyramid MnO 5 www.nature.com/scientificreports/ and octahedra MnO 6 in hexagonal and orthorhombic RMnO 3 (R is rare earth) structures evaluating EFG components with application of point charge model 38,40,41 . The doublets of smaller splitting should be attributed to more symmetric Fe sites. According to previous Mössbauer study of hexagonal YFe y Mn 1−y O 3 38 the relative area of doublet with smaller ∆ increased with Mn substitution by Fe. In the studies 39,42 additional doublets at y = 0.1; 0.2 were related with Fe atoms occupying Mn sites in nearest neighborhood. However, at 3% Mn substitution the observed intensity of additional doublet was much larger than could be according to random occupation of    Table 2). Comparing hexagonal and orthorhombic phases it can be noticed that the dependence of isomer shift δ on quadrupole splitting ∆ is of different sign (Fig. 7b). Moreover, it can be noted that the ratio of intensities of doublet lines A 12 is slightly deviated from 1 ( Table 2). Such effect can be caused by Goldanskii-Karyagin effect or sample anisotropy 56 . According to low temperature Mössbauer spectra the magnetic ordering occurs at ≈ 36, 39, 70 K for Y x Gd 1−x Mn 0.97 Fe 0.03 O 3 with x = 0, 0.4 and 1, respectively (Figs. 8 and 9a). It was previously observed for YFe y Mn 1−y O 3 that increasing y from 0.02 to 0.2 magnetic ordering temperature decreased from 73 to 60 K 38 . Crystal structure (hexagonal or orthorhombic) of RMnO 3 and Mn-O-Mn angle in orthorhombic RMnO 3 affect magnetic ordering temperature according to [38][39][40][41][42][43]55,57,58 . Mössbauer spectrum of YMn 0.97 Fe 0.03 O 3 measured at 12 K (Fig. 8a) was fitted to subspectra using Hamiltonian method as the quadrupole shifts were too large to  www.nature.com/scientificreports/  www.nature.com/scientificreports/ consider only first order corrections (Eq. (2)). The term of quadrupole interaction was fixed according to room temperature data, |eQV zz |/2≈ � (Eq. (1), Tables 2 and 3). In case of hexagonal YMn 0.97 Fe 0.03 O 3 the angle θ = 90° (Table 3) corresponds to EFG z axis along crystal c axis and the magnetization in ab plane 39,42,58,59 . The angle φ between magnetization projection into EFG xy plane and x axis had small influence on fitting quality and was fixed to 0 or 90° trying to keep 0 < η < 1 (Table 3). When asymmetry of EFG 37 which expressed by parameter η is small, according to Eq. (2) the spins of Fe lying at different angles in ab plane resulted only in small changes in position of lines of Mössbauer spectrum and can be ascribed to the same subspectrum with slightly larger width of lines. Negative sign of eQV zz /2 for YMn 0.97 Fe 0.03 O 3 was in agreement with point charge calculations of EFG in case of hexagonal structure of rare earth manganites YMnO 3 and YbMnO 3 38,40 . Four subspectra (Table 3) which were fitted to Mössbauer spectrum of YMn 0.97 Fe 0.03 O 3 measured at 12 K correspond to doublets because of fixing eQV zz /2 and area ratios of subspectra. Largest difference in hyperfine field B values of subspectra was ≈3 T indicating that the spectrum is broadened because of variation of both dipole magnetic and quadrupole interactions.
At lowest 11-12 K temperature the lines of Mössbauer spectra of GdMn 0.97 Fe 0.03 O 3 and especially Y 0.4 Gd 0.6 Mn 0.97 Fe 0.03 O 3 (Fig. 8b,c) were broader than those of YMn 0.97 Fe 0.03 O 3 . Four sextets with different parameters B and 2ε were fitted to Mössbauer spectra (Table 3). Quadrupole shifts of subspectra were smaller than those of YMn 0.97 Fe 0.03 O 3 calculated according to Eq. (2) as shown in Fig. 9b, but hyperfine field B of sextets varied from 20 to ≈42 T. Such differences arise because of lower magnetic ordering temperature, crystal structure and spin ordering specifics.
Hyperfine field distributions P(B) were used for fitting to the Mössbauer spectra (Figs. 8 and 10) measured at higher then 11-12 K temperature up to transition to paramagnetic state. The shape of distributions is characterized by the features that are specific of different studied samples. Hyperfine distributions of YMn 0.97 Fe 0.03 O 3 exhibited one peak within 35-45 T hyperfine field range which intensity decreased as temperature increased (Figs. 9c and 10a). Two peaks within 30-45 and 10-20 T hyperfine field regions (Fig. 9c) (Fig. 9d) when the magnetic ordering transition in GdMnO 3 occurs 55,57,58,60 . P(B) within 25-45 T region dominates at low 11-12 K temperature (Fig. 10c, Table 3) when Mn spin order in GdMnO 3 is A type antiferromagnetic. Incommensurate collinear (IC) sinusoidal amplitude modulated spin order (spin lying along a axis of Pnma space group or b axis in Pbnm) at temperature higher than 23 K should lead to P(B) distribution in a wide hyperfine field range as observed for FeVO 4 (Fig. 10b). Two P(B) having different Table 3. Parameters of sextets and Hamiltonian used to fit to Y x Gd 1−x Mn 0.97 Fe 0.03 O 3 Mössbauer spectra measured at 11-12 K: S is relative area, Γ-linewidth, δ-isomer shift relatively to α-Fe at room temperature, 2ε-quadrupole shift, eQV zz 2-term of quadrupole interaction, B-hyperfine field, η-asymmetry parameter, θ is the angle between magnetization and EFG z axis and φ is the angle between magnetization projection and EFG x axis. *Fixed. www.nature.com/scientificreports/ quadrupole shift were needed in 25-45 T range, probably because of different Y and Gd influence on Fe local surrounding. In this case even more complicated spin order may exist. Mössbauer spectra broadening and lines shift to center when temperature increases are explained by the increase in spin relaxation rate. According to Mössbauer spectra line shape theory 62,63 the shape depends on population of stochastic spin states and transition rate between these states. Different stochastic states of Fe occur because of thermal excitations of spins, collective excitations such as magnons. At characteristic time of transition between stochastic states less than 10 -7 s the shape of spectra starts to change (Mössbauer lines broaden) and at less than 10 -9 -10 -10 s averaged states (merged sextet lines to doublet/singlet at paramagnetic state) are observed. The change in peak positions B peak of GdMn 0.97 Fe 0.03 O 3 within 25-45 and 0-25 T hyperfine field regions, when temperature increases, was rather slow similarly to that of one peak position of YMn 0.97 Fe 0.03 O 3 in Figs. 9c and 10a. The gradual probability shift from the higher B region with P(B) peak at B = 30-45 T to another one with 10-20 T peak with increasing temperature is associated with the decrease in hyperfine field because of thermal excitation and averaging of states. The spin order transitions can also contribute to lowering of hyperfine field, however, we do not observe any abrupt changes at 23 K (Fig. 9a).

Conclusions
A series of Y x Gd 1−x Mn 0.97 Fe 0.03 O 3 (x from 0 to 1 with a step of 0.2) powders has been synthesized by an aqueous sol-gel method, and gadolinium substitution effects in yttrium manganite were investigated. Partial substitution of Mn 3+ by 57 Fe 3+ was performed in order to investigate deeper the structural properties of synthesized compounds applying Mössbauer spectroscopy. With increasing the Gd 3+ content in the samples the crystal structure of Y x Gd 1−x Mn 0.97 Fe 0.03 O 3 gradually transformed from hexagonal to orthorhombic. The mixed phase was obtained when x = 0.6, whereas other compounds were determined to be monophasic. It was demonstrated that cell parameters increased almost linearly with increasing amount of Gd 3+ in Y x Gd 1−x Mn 0.97 Fe 0.03 O 3 . The results of FTIR and Raman spectroscopies were in good agreement with ones obtained by XRD analysis. According to SEM micrographs, the most of the samples were composed of porous aggregates which are organized of significantly smaller and mostly uniform particles necked to each other. The particle size varies in the range of approximately 100-600 nm depending on the chemical composition of gadolinium-substituted yttrium manganites. All synthesized compounds were characterized by paramagnetic behavior at room temperature; however, magnetization values were found to be dependent on chemical composition of the samples. Solid solutions with orthorhombic structure revealed higher magnetization values compared to those of hexagonal samples. According to Mössbauer spectroscopy data the magnetic ordering occurs at ≈ 36, 39 and 70 K for Y x Gd 1−x Mn 0.97 Fe 0.03 O 3 with x = 0, 0.4 and 1, respectively. For orthorhombic GdMn 0.97 Fe 0.03 O 3 the change in types of antiferromagnetic ordering at 23 K is www.nature.com/scientificreports/ associated with the increase in hyperfine field probability of distribution within 25-45 T relatively to 10-20 T region which is more intense at higher temperature.