Phase Transition Behavior and Mechanical Properties of 9 Mol% CaO-PSZ with MnO₂ Doping Under Thermal Stress

Abstract

MnO₂-doped 9 mol% CaO-stabilized zirconia (CSZ) was investigated in terms of phase stability, microstructure, and mechanical properties before and after thermal cycling. As the MnO₂ content increased from 2 to 4 mol%, the monoclinic phase fraction decreased significantly (from 32.6% to 2.5%), while the tetragonal phase fraction increased (from 58.2% to 90.3%), indicating an enhanced phase stability comparable to fully stabilized ZrO₂. The cubic phase fraction decreased from 9.2% to 3.4% with 2–3 mol% MnO₂, but increased to 7.2% at 4 mol%. The 9 mol% CSZ showed a mixture of grains around 2 μm and 10 μm, while the MnO₂-doped CSZ exhibited only grains larger than 30 μm, suggesting that MnO₂ acted as a sintering aid. After thermal cycling, increasing the MnO₂ content from 2 to 4 mol% led to an increase in the monoclinic phase fraction (from 7.8% to 17.2%) and a decrease in the tetragonal phase fraction (from 53.6% to 21.8%). The Vickers hardness and wear resistance of MnO₂-doped CSZ were superior to those of undoped 9-CSZ, and improved as the MnO₂ doping level increased. These mechanical properties were maximized in the CSZ doped with 3 mol% MnO₂, and this trend persisted after thermal cycling. These results demonstrate that MnO₂ doping effectively enhances the phase stability and mechanical performance of CaO-partially stabilized zirconia under thermal stress cycling conditions.

1. Introduction

Zirconia (ZrO₂) exhibits excellent mechanical, thermal, and chemical properties even in harsh environments, making it a widely studied material. Despite being a ceramic material, it possesses high toughness comparable to metals and is therefore applied in various material development fields, including structural and functional applications [1,2,3,4]. Zirconia is frequently utilized under ultra-high-temperature conditions because of its high melting point and chemical durability; however, its crystal structure and mechanical properties change significantly with temperature. Zirconia is stable below 1170 °C in the monoclinic phase, between 1170 °C and 2370 °C in the tetragonal phase, and above 2370 °C in the cubic phase. The phase transition from tetragonal to monoclinic is accompanied by volume expansion and results in the formation of cracks and grain boundary fractures, which rapidly decrease the strength [5,6,7]. The dopant species and its concentration determine whether zirconia is classified as partially stabilized zirconia (PSZ) or fully stabilized zirconia (FSZ). The phase transformation to the monoclinic phase induces crack formation, but PSZ and FSZ enhance their mechanical properties by absorbing the stress generated during this process. Thus, the stabilization of zirconia phases is governed by the type and amount of the dopant [8,9,10]. Various doping strategies have been studied to improve the high-temperature mechanical properties of zirconia through the use of stabilizers such as Y₂O₃, MgO, Al₂O₃, Bi₂O₃, and CaO [11,12]. Among the various stabilizers, CaO-stabilized zirconia demonstrates enhanced hardness and oxidation resistance at high temperatures compared with zirconia stabilized with other oxides [13,14,15]. Although CaO-PSZ exhibits excellent mechanical and thermal behavior, extended exposure to temperatures above 1500 °C can result in phase destabilization due to the inherent instability of Ca as a dopant. Therefore, ongoing research is necessary to address the destabilization of CaO-PSZ and to further enhance its high-temperature mechanical and thermal performance.

Previous studies have examined the trend in the mechanical properties of CaO-PSZ as the doping amount of CaO increased from 5 to 10 mol%, and it was found that 9 mol% CaO-stabilized zirconia (CSZ) showed superior mechanical properties [16]. To further enhance the phase stability and mechanical properties of CSZ under harsh thermal conditions, MnO₂, which can exhibit various valence states and substitute the cations in the lattice, was doped into 9 mol% CSZ. This trend is expected to be different from that observed when ZrO₂ is doped solely with CaO or MnO₂. The effects of MnO₂ doping on the structural and mechanical properties of CSZ were then investigated before and after thermal shock to assess its durability under thermal stress.

2. Materials and Methods

In this study, 9 mol% CSZ was synthesized using ZrO₂ (99%, ≤ 45 μm, Daejung Chemicals & Metals Co., Siheung-si, Republic of Korea) and CaO (98%, Junsei Chemical Co., Tokyo, Japan) as starting materials through a solid-state method, and the synthesized 9-CSZ was subsequently mixed with MnO₂ (90%, ≤ 500 nm, Junsei Chemical Co., Tokyo, Japan), as shown in

Table 1. Compositions and fabrication conditions of 9-CSZ specimens with varying MnO₂ content.

Name	Composition		Fabrication Conditions
	Raw Material	MnO2 Addition (mol%)	Calcination Temp (°C)	Holding Time (h)
9-CSZ	9 mol% CSZ	-	1200	2
9-CSZ_2 MnO2		2
9-CSZ_3 MnO2		3
9-CSZ_4 MnO2		4

. The mixture was subjected to ball milling using zirconia balls and ethanol for 24 h, dried at 90 °C for 24 h, and then calcined at 1200 °C for 2 h with a heating rate of 5 °C/min.

The specimens utilized in this research were fabricated by uniaxially compacting the synthesized raw powders at a pressure of 10 tons. This process resulted in cylindrical samples with a diameter of 20 mm and rod-shaped compact samples measuring 5 mm in width and 35 mm in length. These green bodies were subsequently sintered in a box-type electric furnace, where the temperature was increased at a rate of 5 °C/min, held at 1600 °C for 6 h, and then cooled in the furnace. For the thermal shock evaluation, the sintered samples were heated in air to 1300 °C at a rate of 5 °C/min, maintained at this temperature for 5 min, and then cooled naturally to 900 °C. This heating and cooling process was repeated 40 times to induce thermal shock conditions.

Phase analysis of the specimens before and after thermal shock was performed using X-ray diffraction (XRD; Rigaku, Ultima-IV, Hitachi, Japan). Measurements were taken from 20° to 80° with a step size of 0.02° (2θ) and a scanning rate of 2°/min. The resulting diffraction patterns and phase fractions were interpreted using the Rietveld refinement technique in Highscore Plus software (version 3.0c), referencing the ICSD patterns for monoclinic ZrO₂ (m-ZrO₂, ICSD 98-006-0900), tetragonal ZrO₂ (t-ZrO₂, ICSD 98-007-0014), cubic ZrO₂ (c-ZrO₂, ICSD 98-064-7689), and MnO₂ (ICSD 98-000-0393). Surface microstructural features were examined using high-resolution scanning electron microscopy (HR-SEM; SU8230, Hitachi, Japan). High-resolution transmission electron microscopy (JEOL JEM-ARM 200F, 200 kV accelerating voltage) was employed to obtain detailed TEM images. Mechanical properties were assessed using a micro-Vickers hardness tester (Wilson, VH1102, Lake Bluff, IL, USA) with a 20 N load and a 15 s dwell time. Each sample was measured ten times, and the average value was reported as the Vickers hardness.

Wear resistance was evaluated using a ball-on-disk operating method, following the ISO 20808 standard [17], with a sliding distance of 1000 m, an applied load of 9.6 N, and a sliding speed of 0.1 m/s, to determine the specific wear rate (mm³/N·m). Flexural strength was determined by conducting three-point bending tests on a universal testing machine from the United States, using a support span of 20 mm and a crosshead speed of 0.5 mm/min.

3. Results and Discussion

The results of the phase analysis for doping 9 mol% CaO-stabilized zirconia (9-CSZ) with 2–4 mol% MnO₂ are shown in Figure 1 and

Table 2. Rietveld results of 9-CSZ powder calcinated at 1200 °C.

		9-CSZ	2MnCSZ	3MnCSZ	4MnCSZ
m-ZrO2	volume %	32.6	3.8	6.5	2.5
	a (Å)	5.149	5.137	5.471	5.223
	b (Å)	5.203	5.234	5.662	5.154
	c (Å)	5.319	5.269	5.143	5.219
	α = γ	90°	90°	90°	90°
	β	99.198	99.093	99.439	99.156
	lattice volume (Å3)	140.6650712	139.8875853	157.1567065	138.7020022
	chemical composition	ZrO1.987 ± 0.003	ZrO1.982 ± 0.018	ZrO1.978 ± 0.006	ZrO1.976 ± 0.010
t-ZrO2	volume %	58.2	88.2	90.1	90.3
	a = b (Å)	3.626	3.613	3.616	3.616
	c (Å)	5.122	5.129	5.126	5.111
	α = β = γ	90°	90°	90°	90°
	lattice volume (Å3)	67.34342087	66.9527812	67.02478746	66.82865562
	chemical composition	ZrO1.973 ± 0.005	ZrO1.966 ± 0.018	ZrO1.966 ± 0.003	ZrO1.964 ± 0.010
c-ZrO2	volume %	9.2	8	3.4	7.2
	a = b = c (Å)	5.134	5.115	5.113	5.094
	α = β = γ	90°	90°	90°	90°
	lattice volume (Å3)	135.3217461	133.8248959	133.6679779	132.1833706
	chemical composition	ZrO1.995 ± 0.007	ZrO1.992 ± 0.014	ZrO1.985 ± 0.009	ZrO1.980 ± 0.010
MnO2	volume %	-	0	0	0
Rwp		22.474	18.498	17.342	16.886
Rp		18.458	14.549	13.594	13.564
χ2		4.0256	2.604	2.191	1.958

. The monoclinic fraction (V_m) of 9-CSZ was 32.6%. In contrast, with the addition of 2, 3, and 4 mol% MnO₂, V_m decreased to 3.8%, 6.5%, and 2.5%, respectively, which indicates the progression of phase stabilization to the level of fully stabilized ZrO₂. The tetragonal fraction (V_T) for 9-CSZ was 58.2%, whereas MnO₂-doped CSZ showed increased values, ranging from 88.2% to 90.3%. The cubic phase fraction (Vc) of 9-CSZ was 9.2%, and it decreased to 8.0% and 3.4% with the addition of 2 and 3 mol% MnO₂, respectively. Notably, with the addition of 4 mol% MnO₂, Vc increased to 7.2% compared with the 3.4% observed for the 3 mol% MnO₂-doped CSZ. The doping limit of CaO in ZrO₂ is established to be 15 mol%; however, it is evident that the doping limit for MnO₂ is relatively lower. Ca²⁺ ions (1.00 Å) are larger than Zr⁴⁺ ions (0.84 Å), whereas Mn²⁺ (0.83 Å) and Mn³⁺ (0.65 Å) ions are smaller than Zr⁴⁺ ions. These differences in ionic radii lead to lattice contraction and distortion, which play a significant role in phase stabilization. MnO₂ doping in ZrO₂ not only generates additional oxygen vacancies but also enhances phase stabilization through the lattice distortion caused by the contraction of the lattice due to the smaller dopant ions [18]. Therefore, both the increased concentration of oxygen vacancies and the size effect of the dopant ions synergistically contribute to the stabilization of the tetragonal phase.

The Mn 2p_3/2 XPS spectra of 9-CSZ doped with 2–4 mol% MnO₂ are shown in Figure 2. The peaks at 640.8, 641.8, and 642.7 eV correspond to Mn²⁺, Mn³⁺, and Mn⁴⁺, respectively. It was observed that Mn predominantly exists as Mn²⁺ in all samples [19]. This indicates that manganese oxide, initially introduced as MnO₂ (Mn⁴⁺), is mainly present as Mn²⁺ within the zirconia lattice, implying a change in valence state [20]. The predominant presence of Mn²⁺ leads to the formation of more oxygen vacancies due to charge compensation, compared with Mn³⁺ and Mn⁴⁺. This results in a greater reduction in the monoclinic phase upon doping, which correlates with the phase fractions observed in the XRD analysis. The fractions of Mn³⁺ and Mn⁴⁺ gradually decreased as the MnO₂ doping level increased, reaching 0% in the sample with 4 mol% MnO₂. This result is known to be due to the high crystal field stabilization energy (CFSE) of the tetragonal phase, which makes Mn⁴⁺ unstable and leads to its reduction to the lower valence states of Mn²⁺ or Mn³⁺ for stabilization [21]. Therefore, Mn primarily exists as Mn²⁺ in the samples, which is associated with the decrease in the monoclinic phase and the increase in the tetragonal phase.

Figure 3 shows the microstructural images of 9-CSZ and 9-CSZ doped with 2–4 mol% MnO₂. The 9-CSZ sample displayed a mixture of small grains approximately 2 μm in size and relatively larger grains of around 10 μm. In contrast, the MnO₂-doped 9-CSZ specimens contained exclusively large grains exceeding 30 μm, indicating significant grain growth. This pronounced grain growth suggests that MnO₂-doped 9-CSZ has enhanced sinterability compared with undoped 9-CSZ.

Based on these microstructural observations, and according to the three-dimensional critical grain size model proposed by Potter and Heuer, the tetragonal-to-monoclinic phase transformation is governed by the relationship between the grain size and the critical grain size (r_c).

$$r_c={\displaystyle \frac{2\gamma }{∆G_v}}$$

where r_c is the critical grain size (radius), γ is the interfacial energy, and ΔG_v is the difference in volumetric free energy.

MnO₂ doping not only significantly increased the average grain size (from 2 to 11 μm to 36–50 μm), but also dramatically reduced the monoclinic phase fraction (from 32.6% to 2.5–6.5%). This indicates that MnO₂ doping increases the critical grain size, thereby stabilizing the tetragonal phase even at much larger grain sizes. The observed decrease in monoclinic phase fraction with increasing MnO₂ content is therefore attributed to the increase in r_c rather than a reduction in grain size.

The MnO₂-doped CSZ samples were subjected to thermal shock treatment by rapid cooling from 1300 °C to 900 °C, corresponding to a temperature difference of 400 °C. The phase analysis results obtained after this treatment are shown in Figure 4 and

Table 3. Rietveld refinement results of 9-CSZ powder after thermal cycle (Δ400 °C).

		9-CSZ	2MnCSZ	3MnCSZ	4MnCSZ
m-ZrO2	volume %	27.1	7.8	17.2	9.3
	a (Å)	5.145	5.212	5.171	6.084
	b (Å)	5.194	4.928	5.903	4.989
	c (Å)	5.309	5.55	5.26	5.324
	α = γ	90°	90°	90°	90°
	β	99.174	99.693	99.137	100.791
	lattice volume (Å3)	140.058358	140.5152436	158.5211616	158.7421548
	chemical composition	ZrO1.995 ± 0.008	ZrO1.992 ± 0.003	ZrO1.993 ± 0.011	ZrO1.988 ± 0.020
t-ZrO2	volume %	65.9	53.6	21.8	17
	a = b (Å)	3.626	3.619	3.633	3.657
	c (Å)	5.11	5.129	5.129	5.101
	α = β = γ	90°	90°	90°	90°
	lattice volume (Å3)	67.18564636	67.17533877	67.69607588	68.21898355
	chemical composition	ZrO1.983 ± 0.013	ZrO1.979 ± 0.012	ZrO1.977 ± 0.009	ZrO1.968 ± 0.021
c-ZrO2	volume %	7	38.7	61	73.7
	a = b = c (Å)	5.142	5.109	5.107	5.113
	α = β = γ	90°	90°	90°	90°
	lattice volume (Å3)	135.9553233	133.35451	133.19796	133.6679779
	chemical composition	ZrO1.998 ± 0.003	ZrO1.995 ± 0.017	ZrO1.991 ± 0.008	ZrO1.983 ± 0.016
MnO2	volume %	-	0	0	0
Rwp		16.599	17.205	19.18	17.156
Rp		12.902	13.28	14.79	12.973
χ2		1.423	1.557	1.806	1.547

. After thermal shock, the monoclinic phase fraction was 30.7% for 9-CSZ, whereas it was significantly lower for 2–4 mol% MnO₂-doped CSZ, ranging from 1.1% to 2.9%. This trend was consistent with the results observed prior to thermal shock. However, the proportion of the tetragonal phase in 2–4 mol% MnO₂-doped CSZ was lower, ranging from 17% to 53.6%, compared with 65.9% in 9-CSZ. Additionally, the fraction of the cubic phase increased dramatically in 2–4 mol% MnO₂-doped CSZ, from 38.7% to 73.7%, compared with only 7% in 9-CSZ. These results suggest that the increase in Mn²⁺ ions resulting from MnO₂ doping in 9-CSZ leads to an increase in oxygen vacancies due to the difference in valence states. The increased concentration of oxygen vacancies promotes the stabilization of the cubic phase, which may result in a rapid increase in the cubic phase fraction during thermal shock.

The morphological characteristics of all ZrO₂ samples with varying amounts of MnO₂ were examined using high-resolution TEM, and the resulting images are presented in Figure 5, and the related data are summarized in

Table 4. Summary of interplanar spacings, angles, and estimated phases obtained from HRTEM images (Figure 5).

	Length of a (Å)	Length of b (Å)	Angle Between a and b(°)	Estimated Phase
9-CSZ	2.91	3.81	87.58	Monoclinic
2MnCSZ	3.01	3.16	69.96	Tetragonal
3MnCSZ	2.99	3.13	71.32	Tetragonal
4MnCSZ	2.98	2.96	59.93	Cubic

. The (111) plane of tetragonal ZrO₂ exhibits an interplanar spacing in the range of 3.00–3.10 Å. In the case of the cubic phase, the interplanar spacing of the (111) plane is specifically 2.97 Å. For the tetragonal phase, the O–Zr–O bond angle is about 71.05–71.32°, while in the cubic phase, the Zr–Zr atomic angle on the (111) plane is 60°, and the Zr–O–Zr bond angle is approximately 109.5° [22,23]. The interplanar spacings of 2.91 Å and 3.81 Å observed in Figure 5a are similar to those of the ($\overline{1}11$) and (011) planes of the monoclinic phase, respectively (ICSD 98-001-5983). The measured angle between the atomic planes is 87.58°, which differs from the theoretical value of 91°. However, this value is considered as consistent when possible interface angle distortions during TEM analysis are taken into account. Furthermore, such a combination of interplanar spacing and angle cannot be observed in the cubic or tetragonal phases; thus, the structure is identified as monoclinic. Figure 5b,c show interplanar spacings in the range of 2.99–3.16 Å, with atomic arrangement angles of 69.96° and 71.32°, respectively. These values are very similar to the interplanar spacing and O–Zr–O bond angles of the tetragonal phase; therefore, these regions are presumed to correspond to the (111) plane of the tetragonal phase. Figure 5d exhibits an interplanar spacing of 2.96–2.98 Å and an atomic arrangement angle of 59.93°. These values are nearly identical to the interplanar spacing (2.97 Å) and atomic angle (60°) of the (111) plane of the cubic phase, indicating that this region is likely to correspond to the (111) plane of the cubic phase. The TEM image analysis provides indirect evidence that MnO₂-doped CSZ retains the tetragonal and cubic phases even after thermal shock.

The Vickers hardness values of each specimen before and after thermal shock are shown in Figure 6. The Vickers hardness of 9-CSZ before the thermal shock was 10,674 MPa. In contrast, the Vickers hardness of MnO₂-doped CSZ substantially increased with doping levels of 2–4 mol% MnO₂, and were 24,662 MPa, 29,437 MPa, and 27,013 MPa, respectively. The highest value was observed in the 3 mol% MnO₂-doped CSZ, while the 4 mol% MnO₂-doped CSZ exhibited only a slight decrease. After thermal shock, the Vickers hardness decreased for all specimens, with values of 6247 MPa, 14,305 MPa, 20,294 MPa, and 18,637 MPa, respectively. However, the overall trend remained the same as that observed before thermal shock.

Figure 7 shows the specific wear amount of each specimen before and after the thermal shock. The specific wear amount of 9-CSZ before thermal shock was 1.59 × 10⁵ mm³/Nm, whereas for 2 to 4 mol% MnO₂-doped CSZ, it was 0.28 × 10⁵ mm³/Nm, 0.17 × 10⁵ mm³/Nm, and 0.38 × 10⁵ mm³/Nm, respectively. This demonstrates a significant reduction in the wear amount compared with 9-CSZ. After thermal shock, the specific wear amounts were 3.77 × 10⁵ mm³/Nm, 0.82 × 10⁵ mm³/Nm, 0.42 × 10⁵ mm³/Nm, and 0.99 × 10⁵ mm³/Nm, respectively.

The flexural strength before and after thermal shock for 9-CSZ and MnO₂-doped CSZ is shown in Figure 8. The flexural strength of 9-CSZ before thermal shock was 100.2 MPa, whereas those of 2–4 mol% MnO₂-doped CSZ were 387.8 MPa, 411.8 MPa, and 382.4 MPa, respectively. Thus, the flexural strength of MnO₂-doped CSZ increased sharply by more than 386% compared with 9-CSZ. After thermal shock, the flexural strength of all samples declined; the value for 9-CSZ dropped to 88.18 MPa, while those for 2–4 mol% MnO₂-doped CSZ were 195.2 MPa, 270.7 MPa, and 261.5 MPa, respectively, showing the same trend as observed before thermal shock. The flexural strength of 4 mol% MnO₂-doped CSZ was slightly lower than that of 3 mol% MnO₂-doped CSZ. This pattern was similar to the trends found in the Vickers hardness and specific wear amount.

MnO₂ doping enhances the sinterability of CSZ, as demonstrated by the microstructural results. This improvement likely contributes to the superior mechanical properties of MnO₂-doped CSZ compared with 9-CSZ. However, the fact that the mechanical properties did not improve proportionally with increasing MnO₂ doping levels implies that factors beyond sinterability, particularly the phase fraction of CSZ, also play a crucial role. Consequently, doping 9 mol% CSZ with MnO₂ enhanced the phase stability of CSZ and promoted sintering, resulting in excellent mechanical properties. Specifically, doping with 3 mol% MnO₂ fully stabilized the CSZ phase. However, the addition of more than 3 mol% MnO₂ led to an increase in the cubic phase fraction of CSZ, which in turn caused a deterioration in the mechanical properties. Xuemei Song et al. investigated the mechanical properties of yttrium-stabilized zirconia (YSZ) according to its phases and found that the cubic phase possesses a lower elastic modulus, greater ductility, and superior plasticity compared with the monoclinic and tetragonal phases [24]. Materials with a high elastic modulus and low plastic deformation typically exhibit high hardness. Therefore, it can be inferred that the hardness and wear resistance of the cubic phase are lower than those of the tetragonal phase.

The microstructure images of each specimen after thermal stress are shown in Figure 9. MnO₂-doped CSZ displayed relatively uniform and larger grain sizes compared with 9-CSZ, consistent with the observations before thermal shock. These results indicate that the decline in the properties after thermal stress is not related to changes in the microstructure.

4. Conclusions

We investigated the effects of MnO₂ doping on the microstructure, crystal structure, and tribological properties of CSZ under thermal stress. The addition of more than 2 mol% MnO₂ to 9 mol% CSZ stabilizes the ZrO₂ phase, resulting in a tetragonal phase fraction exceeding 88%. The cubic phase fraction of zirconia initially decreases with increasing MnO₂ content, but increases again a doping level of 4 mol% MnO₂. As the MnO₂ doping level increases, the concentrations of Mn³⁺ and Mn⁴⁺ ions in CSZ gradually decline, reaching 0% at 4 mol% MnO₂. This phenomenon is attributed to greater lattice distortion caused by the higher concentration of Mn²⁺ ions, since Mn²⁺ ions have a lower valence state than Zr⁴⁺. MnO₂-doped CSZ maintains a higher phase stability under thermal stress than 9 mol% CSZ; however, a rapid surge in the cubic phase is observed in MnO₂-doped CSZ after thermal stress, which is also considered to result from enhanced lattice distortion due to Mn²⁺.

The tribological properties, as indicated by the Vickers hardness and specific wear amount, are superior in MnO₂-doped CSZ compared with 9 mol% CSZ, primarily due to the enhanced sinterability imparted by MnO₂ doping. However, CSZ doped with 4 mol% MnO₂ exhibits inferior mechanical properties compared with the sample doped with 3 mol% MnO₂, both before and after thermal shock. This deterioration is attributed to the increased cubic phase fraction, which has inferior mechanical properties compared with the tetragonal phase. These results demonstrate that both the microstructural changes induced by MnO₂ doping and the phase composition of CSZ significantly influence its tribological properties. In particular, doping 3 mol% MnO₂ into 9 mol% CSZ results in the highest mechanical performance by fully stabilizing the ZrO₂ phase and maintaining a stable microstructure under thermal stress. These findings are expected to contribute to improving the performance and lifetime of CSZ in high-temperature applications requiring superior mechanical properties.