Improvement of the O2 storage/release rate of YMnO3 nanoparticles synthesized by the polymerized complex method

Yusuke Asakura *, Amiko Miyake , Mayu Otomo and Shu Yin *
Institute of Multidisciplinary Research for Advanced Materials, Tohoku University, 2-1-1 Katahira, Aoba-ku, Sendai 980-8577, Japan. E-mail:;

Received 19th October 2019 , Accepted 12th December 2019

First published on 13th December 2019

YMnO3 nanoparticles with a diameter of ca. 100 nm, which were synthesized by the polymerized complex method, exhibited a high O2 storage/release rate, because nanoparticle morphology increased the O2 accessible surface area of the material. This material with a high O2 storage/release rate is promising as a separator of O2 from air.

Oxygen gas is widely used in industrial applications including metallurgical fabrication, chemical syntheses, and glass manufacturing.1 Industrially, oxygen gas has been prepared by air separation using cryogenic distillation or adsorptive materials. Since cryogenic distillation at very low temperature is expensive, separation by using adsorptive materials is more efficient. In air separation using adsorptive materials by a temperature- or pressure-swing adsorption mechanism, zeolites have been conventionally used as adsorptive materials. Because zeolites usually adsorb molecules in a size selective manner, N2 gas, ca. 80% of air, is preferentially adsorbed from air.2 For the preferential adsorption of O2 gas, ca. 20% of air, the separation efficiency should be improved. Therefore, materials with selective O2 adsorption properties should be developed.

Oxides with oxygen nonstoichiometry have been found to be one of the oxygen storage materials or O2 selective adsorbents. Some of this class of materials can release/store O2 by only the alteration of temperature or O2 partial pressure by using the redox reaction of transition metals, meaning that they are suitable for O2 separation by pressure- or temperature-swing adsorption.3 Recently, several examples have been reported, such as YBaCo4O7+δ,4 YCr1−xPxO4−δ,5 Ca0.8Sr0.2MnO3−δ,6 Ca2AlMnO5+δ,7 YMnO3+δ,8 SrCoxFe1−xO3−δ[thin space (1/6-em)]9 and so on. These materials can release/store O2, which result in the topotactic phase transition of their crystal structures by the redox reaction. Because precious elements (Co or Cr) should not be used in practical applications, Mn-based materials are promising. However, the Mn-based materials have some drawbacks, for example, Ca0.8Sr0.2MnO3−δ and Ca2AlMnO5+δ possess high working temperatures,6,7 while YMnO3+δ possesses a low adsorption/desorption rate.8 Therefore, the drawbacks should be solved for practical use as separators of O2 from air.

YMnO3 with hexagonal structure, which is composed of earth–abundant elements, can act as an O2 adsorptive material at a relatively low working temperature (150–300 °C).8 Therefore, YMnO3 is a promising material as a separation medium of O2 from air. However, its slow O2 adsorption/desorption rate is disadvantageous. For increase of the O2 adsorption/desorption rate, the following two procedures can be effective: (i) elemental substitution with other metal ions10 and (ii) decrease of the particle size.8,10b It is reported that substitution of Y in YMnO3 with larger size rare earth metals (Dy10a or Tb10b) increases its oxidation/reduction rate. Decrease of the particle size can facilitate O2 diffusion into the crystal leading to the increase of the O2 adsorption/desorption rate. Up to now, procedure (i) has been frequently attempted, while procedure (ii) has been rarely reported. If YMnO3 nanoparticles can exhibit a high rate of O2 adsorption/desorption, combination with procedure (i) should further improve their performance.

In this study, YMnO3 nanoparticles were synthesized and their O2 adsorption/desorption behaviors were evaluated. A lot of methods for preparation of nanoparticles have been reported such as wet chemical synthesis,11 templating methods12 and so on. Here, we used the polymerized complex method13 for the synthesis of YMnO3 nanoparticles. The polymerized network formed by esterification of citric acid with ethylene glycol can suppress the crystal growth. In the polymerized complex method, the amount of reagents used for polymerization and complex formation can affect the particle size of the obtained oxides. Therefore, the effects of some factors in the synthesis on their O2 adsorption/desorption behavior were also investigated.

The experimental procedures are briefly shown in Scheme 1 (the details are shown in the ESI). (CH3COO)3Y·4H2O and Mn(CO3)2·xH2O (molar ratio = 1[thin space (1/6-em)]:[thin space (1/6-em)]1) were added into an aqueous citric acid (CA) solution and stirred on a hot plate at 40 °C. After the dissolution of the two reagents, ethylene glycol (EG) was further added and heated with stirring at 150 °C until the formation of a gel. The obtained gel was calcined to form a black powder. The experimental conditions including the amount of CA and EG and the temperature of calcination were altered. The sample conditions and the sample names are summarized in Table 1.

image file: c9dt04095a-s1.tif
Scheme 1 Procedure for YMnO3 nanoparticle synthesis by using the polymerized complex method.
Table 1 Sample names and their synthetic conditions
Sample name CA/mol EG/mol Calcination temperature/°C Calcination time/h
YMO_700_12 0.064 0.105 700 12
YMO_700_24 0.064 0.105 700 24
YMO_800_1 0.064 0.105 800 1
YMO_800_3 0.064 0.105 800 3
YMO_800_6 0.064 0.105 800 6
YMO_900_6 0.064 0.105 900 6
YMO_onlyEG_800_6 0 0.105 800 6
YMO_onlyCA_800_6 0.105 0 800 6
YMO_CA_800_6 0.105 0.105 800 6

The XRD patterns of the obtained samples are shown in Fig. 1 with the peak positions of the standard hexagonal YMnO3 (JCPDS no. 25-1079). In the case of CA[thin space (1/6-em)]:[thin space (1/6-em)]EG = 0.064[thin space (1/6-em)]:[thin space (1/6-em)]0.105 as a composition for gel formation (Fig. 1a–f), calcination at 800 °C or 900 °C for 6 h led to the formation of a single hexagonal YMnO3 phase (YMO_800_6 and YMO_900_6). The pattern of YMO_700_12 (Fig. 1a) showed broad peaks and different peak intensity ratios from those of the hexagonal YMnO3, although the peak positions were similar. Considering that the pattern of YMO_700_24 (Fig. 1b) obtained by calcination for a longer time showed a mixture of Y2O3 and Mn3O4 with a very small amount of hexagonal YMnO3, YMO_700_12 may possess an amorphous phase or metastable orthorhombic YMnO3.14 These behaviors are consistent with those in the calculated phase diagram of Y–Mo–O in which YMnO3 is thermodynamically stable over 789 °C.15 YMO_800_1 (Fig. 1c) and YMO_800_3 (Fig. 1d) exhibited a mixture of the three phases of Y2O3, M3O4, and YMnO3 instead of the formation of hexagonal YMnO3 at the adequate temperature (800 °C) from the thermodynamic aspect. This is probably because the phase did not reach the equilibrium state at this temperature due to the short calcination time.

image file: c9dt04095a-f1.tif
Fig. 1 XRD patterns of (a) YMO_700_12, (b) YMO_700_24, (c) YMO_800_1, (d) YMO_800_3, (e) YMO_800_6, (f) YMO_900_6, (g) YMO_onlyEG_800_6, (h) YMO_onlyCA_800_6, and (i) YMO_CA_800_6 with the peak positions of the standard YMnO3 (JCPDS no. 25-1079).

In addition, the ratio of CA and EG in gel formation was altered for investigation of the effects of CA and EG on the final products (CA[thin space (1/6-em)]:[thin space (1/6-em)]EG = 0[thin space (1/6-em)]:[thin space (1/6-em)]0.105, 0.105[thin space (1/6-em)]:[thin space (1/6-em)]0, and 0.105[thin space (1/6-em)]:[thin space (1/6-em)]0.105). In the case of the use of only EG for gel formation, gelation was difficult. The XRD pattern of YMO_onlyEG_800_6 (Fig. 1g) after calcination shows a mixture of Y2O3, Mn3O4, and YMnO3. On the other hand, in the case of the use of only CA for gel formation, although no gel was also formed, a pure YMnO3 phase was obtained after calcination (Fig. 1h). This means that CA plays an important role in the successful formation of the pure YMnO3 phase, probably because the complexation of the metals with CA in gel formation led to the homogeneous distribution of the metals in the precursor to facilitate the formation of YMnO3 during calcination. The slight increase of CA for gel formation did not largely affect the XRD pattern of the product (Fig. 1i).

The TEM images of the obtained samples are shown in Fig. 2. The image of YMO_700_12 shows very small nanoparticles with a size of 20–30 nm. The TEM image of YMO_800_1 shows slightly larger particles of size 20–50 nm. Judging from the TEM images of YMO_800_3 and YMO_800_6, the particles sizes (ca. 100 nm) were larger than that of YMO_800_1. Although the particle sizes of YMO_800_3 and YMO_800_6 were almost the same, the aggregation degree of YMO_800_6 was higher. The particle size of YMO_900_6 was further larger (ca. 150 nm). These results indicate that calcination for a longer time and/or at higher temperature can also lead to crystal growth. For investigation of the effect of the CA and EG ratio on the obtained morphology, the TEM images of YMO_800_6, YMO_onlyCA_800_6, and YMO_CA_800_6 (Fig. 2d, f and g) obtained under the same calcination conditions were compared. The particle sizes were not so different between one another, while the higher EG ratio in the gelation led to the dispersed particles. Particularly, the particles in YMO_onlyCA_800_6 were highly aggregated unlike the case of YMO_800_6. Although some particles in YMO_CA_800_6 seem to be more dispersed than those in YMO_800_6, some of them were aggregated (Fig. 2h). This is probably because the polymerization of CA with EG can physically separate the crystal nucleus. Consequently, both the complexation of the metals with CA and polymerization of CA with EG were important for the formation of the pure YMnO3 phase and the dispersed particle morphology. More precise control of the CA/EG ratio may lead to the formation of particles with various degrees of aggregation.

image file: c9dt04095a-f2.tif
Fig. 2 TEM images of (a) YMO_700_12, (b) YMO_800_1, (c) YMO_800_3, (d) YMO_800_6, (e) YMO_900_6, (f) YMO_onlyCA_800_6, and (g and h) YMO_CA_800_6.

The crystallite sizes and the specific surface areas of the samples calculated from the XRD patterns and the N2 adsorption results, respectively, are summarized in Table 2. The crystallite sizes were smaller than the particle sizes observed in the TEM images. This means that the observed particles were composed of primary crystal grains. The BET specific surface areas of all the samples were comparable. Because YMO_900_6 possessed a larger particle size in the TEM image (Fig. 2e; the diameter is ca. 150 nm), the specific surface area of YMO_900_6 should be lower than those of the samples obtained by calcination at 800 °C (Fig. 2b–d and f–h; the diameters are ca. 100 nm). This is probably because the surface roughness and/or the aggregation degree of the particles affected the specific surface area.

Table 2 Crystallite sizes and specific surface areas of the samples
Sample name Crystallite sizea/nm Specific surface areab/m2 g−1
a The crystallite sizes were calculated using the Scherrer equation on the basis of the XRD results. b The specific surface areas were calculated using the BET equation on the basis of the N2 absorption results.
YMO_800_1 28 19
YMO_800_3 29 17
YMO_800_6 29 19
YMO_900_6 42 19
YMO_onlyCA_800_6 24 17
YMO_CA_800_6 21 19

For evaluation of O2 adsorption/desorption properties, we measured the weight alteration of the obtained samples on heating from r.t. to 500 °C and subsequent cooling to r.t. with a ramping/decreasing rate of 1 °C min−1. The weight at 500 °C was assumed as that of YMnO3 without excess oxygen, and the oxygen content (3 + δ) in YMnO3+δ was based on the weight alteration (ΔW) observed in the TG curves. Fig. 3a shows the TG curves of YMO_800_1, YMO_800_3, YMO_800_6, YMO_onlyCA_800_6, and YMO_CA_800_6. The oxygen storage capacity of YMO_800_6 was the highest among the five samples, and the maximum value of δ on rising the temperature was ca. 0.32. The reported YMnO3 samples without any elemental substitution exhibited a δ value of below 0.05 under the same evaluation conditions (1 °C min−1).8 This means that the YMnO3 samples obtained by the polymerized complex method in this study possessed a much higher O2 adsorption/desorption rate. In addition, the value was even larger than that measured at a ramping/decreasing rate of 0.1 °C min−1 in the previously reported YMnO3.8 The homogeneously small nanoparticles should contribute to the high oxygen storage capacity at a higher ramping/decreasing rate, or the high oxygen storage/release speed, indicating that nanostructurization is effective for increment of the oxygen diffusion rate. The XRD pattern of YMO_800_6 after TG evaluation (Fig. S1) shows the presence of the oxygen-adsorbed and the pristine phase with no other decomposed phases like Mn3O4, Mn2O3 and MnO2. This indicates that the crystal structure of the obtained YMnO3 is stable under the oxidative conditions.

image file: c9dt04095a-f3.tif
Fig. 3 TG curves of the samples under pure O2 flow (100 mL min−1) with a ramping/decreasing rate of 1 °C min−1; (a) YMO_800_1, YMO_800_3, YMO_800_6, YMO_onlyCA_800_6, and YMO_CA_800_6, and (b) YMO_700_12, YMO_700_24, and YMO_900_6.

YMO_800_1, YMO_800_3, YMO_onlyCA_800_6, and YMO_CA_800_6 possessed lower oxygen storage capacity, although their specific surface areas are comparable to that of YMO_800_6. Because YMO_800_1 and YMO_800_3 included some impurities observed in the XRD patterns (Fig. 1c and d), such impurities can prevent easy access of O2 into the YMnO3 crystal. On the other hand, YMO_onlyCA_800_6 and YMO_CA_800_6 possessed no impurity phase as judged from their XRD patterns (Fig. 1f and h). However, the TEM images of the two samples (Fig. 2f and h) show more aggregated particles compared with the case of YMO_800_6. This may be the reason for the slightly lower oxygen storage capacity of YMO_ onlyCA_800_6 and YMO_CA_800_6. The TG curve of YMO_800_6 at a ramping/decreasing rate of 5 °C min−1 (Fig. 4a) shows lower oxygen storage capacity than that at a ramping/decreasing rate of 1 °C min−1. This indicates that the O2 diffusion-limitation inside the crystal was dominant for the O2 adsorption/desorption process in the YMnO3 nanoparticles, although the oxygen migration energy barrier in the YMnO3 lattice has been calculated to be very low.16 Furthermore, a smaller particle size of YMnO3 can lead to higher rates of O2 adsorption/desorption.

image file: c9dt04095a-f4.tif
Fig. 4 TG curves of YMO_800_6 under (a) pure O2 flow (100 mL min−1) with a ramping/decreasing rate of 5 °C min−1 and (b) atmospheric air with a ramping/decreasing rate of 1 °C min−1.

In the cases of YMO_700_12 and YMO_700_24 (Fig. 3b), the oxygen storage capacities were lower than that of YMO_800_6. The TG curve of YMO_700_12 shows loose adsorption/desorption of oxygen at over 300 °C. Because the difference between storage and release temperatures should be lessened for practical application, YMO_700_12 is not so promising in spite of the relatively high capacity. Although the reason for the slow uptake/release is not exactly understood at the present moment, the presence of the amorphous phase or the different degrees of aggregation may be related to the slow storage rate. YMO_700_24 showed very low oxygen storage capacity. This is probably because of a lot of impurities.

YMO_900_6 exhibited lower oxygen storage capacity (Fig. 3b) than that of YMO_800_6. Although the particle size of YMO_900_6 (Fig. 2e; ca. 150 nm) was larger than that of YMO_800_6 (Fig. 2d; ca. 100 nm), their specific surface areas were comparable (Table 2) probably because of the effect of the surface roughness and/or the aggregation degree of the particles. Therefore, we cannot simply explain the lower oxygen storage capacity of YMO_900_6 by the larger particle size. The large difference between them was observed in the crystallite sizes calculated from the XRD peaks using the Scherrer equation; 29 nm in YMO_800_6 and 42 nm in YMO_900_6 (Table 2). This means that the volume ratio of the grain boundary in YMO_800_6 is much larger than that in YMO_900_6. For some oxides, oxygen diffusion in the grain boundary is faster than that inside the crystal.17 Therefore, the oxygen diffusion in the grain boundary of YMnO3 might be higher than that inside the crystal leading to the increase of the oxygen storage/release rate, or the overall oxygen storage capacity. Because we now do not possess any evidence for such hypothesis, it should be investigated in the near future.

Consequently, the synthetic conditions altered the particle size and the aggregation degree of the obtained YMnO3 to diversify the oxygen storage capacities. The control of time and temperature of heating in gel formation and/or the following calcination can further improve the activity.

We also evaluated the oxygen storage capacity of YMO_800_6 under atmospheric air with a ramping/decreasing rate of 1 °C min−1 (Fig. 4b). The oxygen storage capacity was drastically decreased to a δ value of 0.13 which is ca. 40% of that under a pure O2 atmosphere. Such decrement in the evaluation under air was also observed in a previous report,10b meaning that the oxygen adsorption/desorption of YMO-based materials are sensitive to O2 partial pressure.

Here, the oxygen storage capacities at ramping rates of 1 and 5 °C min−1 of YMO_800_6 are compared with those of Y0.7Tb0.3MnO3 reported previously.10b The Y0.7Tb0.3MnO3 material synthesized by the sol–gel method (YTMO) exhibited the highest oxygen storage capacity among the YMnO3-based materials reported previously. Their maximum δ values and maximum weight increase ratios on rising the temperature are listed in Table 3. The δ value and weight increase ratio correspond to the adsorbed oxygen molar ratio and weight ratio, respectively. In the case of a ramping rate of 1 °C min−1, the maximum δ value of YMO_800_6 (0.32) was lower than that of YTMO (0.36), meaning that the amount of the adsorbed oxygen per one Mn atom in YMO_800_6 was smaller than that in YTMO. However, the maximum weight increase ratios were very close (2.67 wt% in YMO_800_6; 2.72 wt% in YMTO), because of the heavy atomic weight of Tb. This indicates that their performances per weight are comparable at a ramping/decreasing rate of 1 °C min−1. On the other hand, in the case of a ramping rate of 5 °C min−1, both the maximum δ value and maximum weight increase ratio of YMO_800_6 were much higher than those of YTMO. For practical application in pure O2 gas production based on the temperature-swing system, high swing speed of temperature should lead to effective production, and O2 adsorbers must possess high oxygen storage capacity at a high ramping/cooling rate. From this viewpoint, YMO_800_6 is superior as an O2 adsorber to YTMO. This means that the polymerized complex method for the synthesis can lead to a high oxygen storage/release rate compared with the sol–gel method reported previously. Considering the improvement by elemental substitution like in the case of YTMO, a combination of the polymerized complex method shown here with the elemental substitution method will provide oxygen storage materials with further higher oxygen storage/release rates.

Table 3 Maximum δ values and maximum weight increase ratios on rising the temperature for YMO_800_6 and Y0.7Tb0.3MnO3 (YTMO)
  Maximum δ value/— Maximum weight increase ratio/wt%
  1 °C min−1 5 °C min−1 1 °C min−1 5 °C min−1
a The values of Y0.7Tb0.3MnO3 (YTMO) are based on a previous report.10b
YMO_800_6 0.32 0.18 2.67 1.50
Y0.7Tb0.3MnO3 (YTMO)a 0.36 0.09 2.72 0.65


YMnO3 nanoparticles were synthesized by the polymerized complex method, and the obtained nanoparticles exhibited high oxygen storage capacity. To the best of our knowledge, the storage/release rate of the obtained YMnO3 by calcination at 800 °C for 6 h was the highest among YMnO3 without any elemental substitution reported previously. In addition, the adsorbed oxygen weight ratio at a ramping rate of 5 °C min−1 was the highest among YMnO3-based materials. These facts strongly indicate that nanostructurization of materials with oxygen storage ability is effective for increment of the oxygen diffusion rate. In addition, the heating temperature for gel formation, the subsequent calcination temperature and time largely, and the CA/EG ratio in the gel formation affected oxygen storage capacity and storage/release rates. This is probably because such conditions can induce different amounts of impurities and a distinct degree of particle aggregation. The oxygen storage/release rate of the obtained YMnO3 nanoparticles is still slow for O2 production in practical situations. Therefore, combination of this synthetic method with the improvement of oxygen diffusion by elemental substitution reported previously10 will lead to more practical YMnO3-based separation media of O2 from air. Conventionally, such elemental substitutions have been achieved by using rare metals including Dy and Tb. Earth abundant elements should be utilized as substitution elements. In addition, the obtained YMnO3 nanoparticles here and the YMnO3-based materials reported previously10b possessed lower oxygen storage capacity under atmospheric air. For practical O2 production from air, we must seek materials which can effectively work at low O2 partial pressure.

Conflicts of interest

There are no conflicts to declare.


The authors would like to acknowledge Prof. T. Motohashi (Kanagawa Univ.), Prof. M. Kakihana, and Prof. H. Kato (Tohoku Univ.) for their helpful discussions. This work was supported by the Council for Science, Technology and Innovation (CSTI), Cross-ministerial Strategic Innovation Promotion Program (SIP), “Energy systems toward a decarbonized society” (Funding agency: JST) and Dynamic Alliance for Open Innovation Bridging Human, Environment and Materials in Network Joint Research Center for Materials and Devices.


  1. W. F. Castle, Int. J. Refrig., 2002, 25, 158 CrossRef CAS .
  2. R. V. Jasra, N. V. Choudary and S. G. T. Bhat, Sep. Sci. Technol., 1991, 26, 885 CrossRef CAS .
  3. B. Bulfin, J. Vieten, C. Agrafiotis, M. Roeb and C. Sattler, J. Mater. Chem. A, 2017, 5, 18951 RSC .
  4. M. Karppinen, H. Yamauchi, S. Otani, T. Fujita, T. Motohashi, Y. H. Huang, M. Valkeapaa and H. Fjellvag, Chem. Mater., 2006, 18, 490 CrossRef CAS .
  5. Y. Aoki, K. Kuroda, S. Hinokuma, C. Kura, C. Zhu, E. Tsuji, A. Nakao, M. Wakeshima, Y. Hinatsu and H. Habazaki, J. Am. Chem. Soc., 2017, 139, 11197 CrossRef CAS PubMed .
  6. B. Bulfin, J. Vieten, D. E. Starr, A. Azarpira, C. Zachaus, M. Havecker, K. Skorupska, M. Schmucker, M. Roeb and C. Sattler, J. Mater. Chem. A, 2017, 5, 7912 RSC .
  7. T. Motohashi, Y. Hirano, Y. Masubuchi, K. Oshima, T. Setoyama and S. Kikkawa, Chem. Mater., 2013, 25, 372 CrossRef CAS .
  8. A. Klimkowicz, K. Swierczek, S. Kobayashi, A. Takasaki, W. Allahyani and B. Dabrowski, J. Solid State Chem., 2018, 258, 471 CrossRef CAS .
  9. H. Ikeda, A. Tsuchida, J. Morita and N. Miura, Ind. Eng. Chem. Res., 2016, 55, 6501 CrossRef CAS .
  10. (a) S. Remsen and B. Dabrowski, Chem. Mater., 2011, 23, 3818 CrossRef CAS ; (b) A. Klimkowicz, K. Cichy, O. Chmaissem, B. Dabrowski, B. Poudel, K. Swierczek, K. M. Taddei and A. Takasaki, J. Mater. Chem. A, 2019, 7, 2608 RSC .
  11. A. V. Nikam, B. L. V. Prasad and A. A. Kulkarni, CrystEngComm, 2018, 20, 5091 RSC .
  12. (a) V. Malgras, J. Henzie, T. Takei and Y. Yamauchi, Angew. Chem., Int. Ed., 2018, 57, 8881 CrossRef CAS PubMed ; (b) V. Malgras, J. Henzie, T. Takei and Y. Yamauchi, Chem. Commun., 2017, 53, 2359 RSC .
  13. K. Bergum, H. Okamoto, H. Fjellvag, T. Grande, M. A. Einarsrud and S. M. Selbach, Dalton Trans., 2011, 40, 7583 RSC .
  14. H. W. Brinks, H. Fjellvag and A. Kjekshus, J. Solid State Chem., 1997, 129, 334 CrossRef CAS .
  15. M. Chen, B. Hallstedt and L. J. Gauckler, J. Alloys Compd., 2005, 393, 114 CrossRef CAS .
  16. S. H. Skjaervo, E. T. Wefring, S. K. Nesdal, N. H. Gaukas, G. H. Olsen, J. Glaum, T. Tybell and S. M. Selbach, Nat. Commun., 2016, 7, 13745 CrossRef PubMed .
  17. (a) E. Navickas, T. M. Huber, Y. Chen, W. Hetaba, G. Holzlechner, G. Rupp, M. Stoger-Pollach, G. Friedbacher, H. Hutter, B. Yildiz and J. Fleig, Phys. Chem. Chem. Phys., 2015, 17, 7659 RSC ; (b) J. M. Polfus, B. Yildiz and H. L. Tuller, Phys. Chem. Chem. Phys., 2018, 20, 19142 RSC .


Electronic supplementary information (ESI) available: the experimental details and XRD patterns of the samples. See DOI: 10.1039/C9DT04095A

This journal is © The Royal Society of Chemistry 2020