The effect of chemical composition and heat parameter on Nano-sized precipitate dispersion characterization in ODS steels

Oxide dispersion strengthened (ODS) supersaturated solid solution alloy powders with different Y/Ti atomic ratios were formed by mechanical alloying (MA) method and then heat treated under different conditions. The morphology, phase composition, and particle size distribution were researched. The precipitate distribution densities and hardness values of alloy powders after heat treatment were measured at the same time by the SAXS method and hard tester. The results show that the morphology of MA alloy powders is irregular in shape and only a small amount of MA alloy powders is bigger than 30 μm. During the MA process, alloy elements are solid solutions in the matrix. The work-harden effect of alloy powders formed in the MA process would be eliminated during the heat treatment process. The chemical composition and heat treatment parameters are crucial effects on nano-sized precipitate distribution characterization in ODS steels. The ODS sample with a Y/Ti atomic ratio of 0.99 has the highest distribution density of oxide precipitates at 1.89×1022/m3 resulting in the highest hardness value during the heat treatment process. The sample with a Y/Ti atomic ratio of 0.42 has the highest distribution density of nanoclusters 7.32×1023/m3.


Introduction
Oxide dispersion-strengthened (ODS) steel microstructure, irradiation properties, and mechanical properties are controlled mainly by high-density nano-sized precipitates [1][2].So ODS steels are the one of leading structure materials of IV Generation fission reactors [3][4].The excellent properties of ODS steels make materials work in the core of reactors.
To ensure the nano-sized precipitates dispersed in the microstructure, powder metallurgy methods were used for the solidification of ODS steels.Among all factors, heating parameters (including heat temperature and heat time) and chemical composition are crucial factors.The alloy elements also affect the kind of nano-sized precipitates in ODS steels like Y 2 O 3 , Y 2 Ti 2 O 7 , Y 2 TiO 5 , Y 4 Zr 3 O 12 , YAlO 3 , nanoclusters, and so on.Hirata et al. illustrated that nano-sized precipitates dispersed in ODS steels had NaCl structure [5].The solidification temperatures and time also have key effects on nano-sized precipitates dispersed characterization of ODS steels.Odette et al cleared that nano-sized precipitate distribution density decreased as the solidification temperature rose [6].For now, most research is focused on microstructure, nano-sized precipitates distribution density, and mechanical and irradiation properties of ODS steels [7][8][9].Due to the small size of precipitates in ODS steels, it is difficult to characterize the structure of precipitates by original materials research methods.So SAS (Small Angle Scattering) method was used in ODS steels to characterize the distribution characterization of nano-sized precipitates.
In this research, we use alloy powders with different chemical compositions of ODS steel powder in order to study the effect of chemical composition on microstructure.Meanwhile, heating parameters also are changed to simulate the solidification parameters.Firstly, SEM and XRD are used for observing the milled powder surface feature and phase composition.Then SAXS method is used for characterizing the nano-sized precipitate dispersion characterization.At last, the hardness values of alloy powders are measured after heated or milled.

Raw Materials and experiment
The mechanical alloying (MA) method was used for the formation of Fe-based 9Cr-ODS steels supersaturated solid solution alloy powders.Alloy powders were added to different Y 2 O 3 contents to realize different Y/Ti atomic ratios.The basic information on raw materials used in this research is listed in Table 1.
Table 1 The first step was MA by using Fritsch P5 miller (shown in Figure 1(A)) for the formation of supersaturated solid solution alloy powders.High-purity Ar gas was filled with the mill can to protect alloy powders to avoid population and oxidation.The weight ratio of mill media and alloy powders was 10:1.The milling time was 50 h under room temperature.The alloy powders after milled were sealed in quartz tubes with high-purity Ar gas.And then alloy powders were heated at different temperatures (900℃, 1000℃, 1100℃, 1200℃) for different times (10 min, 30 min, 1 h, 2 h, 4 h) to simulate the solidification conditions.After milled, the morphology of alloy powders was observed in JEOL-7100F SEM, meanwhile, backscatter images of alloy powders were also obtained.The XRD was used in studying the phase composition of milled alloy powders.The XRD scanning mode was a continuous scan with a step of 0.02°.

A B
In order to measure the distribution characterization of nano-sized precipitates in ODS alloys during the heat process, the SAXS method was applied.SAXS experiments were carried out at Shanghai Synchrotron Radiation Facility beamline BL16B1 as shown in Figure 1(B).Before SAXS experiments, the measure detector was corrected by a standard sample.The distance from the sample to the detector was 2 m.Fit 2D software was applied to transform scattering rings into scattering curves [10].The scattering vector q was defined as below: where  was the X-ray wavelength, 2 was the scattering angle.IRENA software was used to fit the SAXS curves after transformation [11].The IRENA software fit process is described in [12].The fit results also contained the distribution density N(r) of nano-sized precipitates with radius r.The distribution density N(r) was assumed to have a log-normal distribution.
The alloy powders after milled or heated were inlaid into resin and then mechanically ground, to make sure alloy powders could expose cross-section.The 401MVDTM hardness tester was used to measure the hardness values of alloy powders.Each sample was measured five times and the average value was calculated.The measure lord, temperature, and lord time were 50 g, room temperature, and 10 s, respectively.

Experimental results and analysis
Figure 2 is the surface morphology characterization and XRD result of milled alloy powders.It is clear that milled alloy powders are irregular in shape because mill media would crash the alloy powders repeatedly during the mill process.So huge deformities will happen in alloy powders under the impact of mill media.A small amount of flat alloy powders are observed in milled alloy powders and the average size of milled alloy powders is smaller than 10 μm.The contrast of the backscatter image is uniform, which means all elements have solid solutions in the matrix and are uniformly distributed because big atomic-weight elements will present light contrast in backscatter images.XRD result also illustrates that only Fe diffraction peaks exist in the curve.There are no other peaks appearing in the XRD curve, which means that all alloy elements have a solid solution in the matrix and no phases containing alloy elements formation in milled alloy powders.We can see from Table 1 that some of the raw materials powders are close to 75 μm, but most alloy powders are smaller than 10 μm after milled.The phenomenon means that mill technology can break the alloy powders accompanied with alloy elements solid solution.It is also noted that the grain size of alloy powders would decrease during the milling process based on the Scherrer equation.The high density of dislocations would form during the milling process as a result of huge deformation.These dislocations could work as atoms diffuse channels, which can promote the alloy elements' diffusion.Suryanarayana [13] illustrated that dislocations could transform into grain boundary finally, so grains would be cut continually until the limitation of the miller.because the nano-sized precipitates in ODS steels are very small, the SAXS method is used for measuring the distribution characterization.Figure 3(A) depicts the SAXS ring obtained from the A B experiment directly.The SAXS ring can be transformed into a curve as shown in Figure 3(B).Because the scattering vector q has an opposite relationship to the size of scatter objects, the curve can be cut into different ranges corresponding to different kinds of precipitates.The ranges cut in this study are like Figure 3(B) [14].The range I is q from 0.3 to 1.0 corresponding to the oxide precipitates with a fixed crystalline structure.The range II is q from 1.0 to 2.0 corresponding to the nanoclusters in ODS steels.From Figure 3(B), the fit lines are perfectly fitted to the experimental line.Figure 4 shows the distribution characterizations of precipitates in ODS steels with a Y/Ti atomic ratio equal to 0.14.The precipitates with fixed crystalline structures and coherent nanoclusters have no certain crystalline structure dispersed in ODS steels.The distribution density change curves of two kinds of precipitates dispersed in ODS steels are shown in Figure 4.When the Y/Ti atomic ratio is 0.14, the distribution density of precipitates with fixed crystalline structures increases with the heating time prolonging at 900℃.If the heating temperature rises to 1000℃, the distribution density of precipitates with fixed crystalline structures increases first with the heating time prolonging, and until heating for 2 h, it then decreases.High temperature will accelerate alloy elements' atom diffusion based on Fick's law, and promote precipitate formation.The distribution density peak of precipitates appears after heating for 2 h, which is earlier than the heating at 900℃.The distribution densities change tendency of precipitates in ODS steels are similar to the heating at 1100℃ and 1200℃.Another characterization is that the maximum value of the distribution density of fixed crystalline structure precipitates increases first when the heating temperature reaches 1000℃.The distribution densities of fixed crystalline structure precipitates are decreasing all the time when heating at 1100℃ and 1200℃.Figure 4(B) lays out the distribution density change tendency of nanoclusters dispersed in ODS steels with different heating conditions.It is clear that the distribution density of nanoclusters rises all the time with the heating time prolonged at 900℃.At 1000℃, the peak of nanocluster distribution density appears at 30 min and then decreases.Raising the heating temperature to 1100℃ and 1200℃, the distribution densities of nanoclusters decrease all the time with the heating time prolonging.In one word, the maximum value of nanocluster distribution density appears earlier if the heating temperature is higher.The nanoclusters are coherent with the matrix and are forming during the solidification process, which is restricted by element diffusion.Nanoclusters would form continuously as a result of element diffusion with heating time prolonging.So the distribution density of nanoclusters is rising all the time at 900℃.When the heating temperature reaches 1000℃, the distribution density of nanoclusters peaks after 30 minutes of heating because higher temperature promotes element atoms to diffuse.Raising the heating temperature, the nanoclusters would merge with each other, resulting in a distribution density decrease.The nanocluster formation would decrease the material's system energy, resulting in materials being more stable.Figure 5 is the precipitate distribution characterization in ODS steels with a Y/Ti atomic ratio equal to 0.42. Figure 5(A) illustrates that the distribution density of precipitates with fixed crystalline structures will increase as the heating time prolongs at 900℃.But under 1000℃, 1100℃, and 1200℃ heating conditions, the distribution density of precipitates with fixed crystalline structures increases first until heating for 1 h.The higher heating temperature results in a lower distribution density of fixed crystalline structure precipitates.Figure 5(B) is the distribution density of nanoclusters dispersed in ODS steels.The distribution densities rise first and then decrease as the heating time is prolonged at all heating temperatures.However, the maximum value of nanocluster distribution density appears after heating for 2 h at 900℃, which is different from other heating temperatures.Compared to the sample with a Y/Ti atomic ratio equal to 0.14, the nanocluster distribution density maximum value is higher.Figure 6 is the precipitate distribution characterization in ODS steels with a Y/Ti atomic ratio equal to 0.99.It is clear that only under 900℃ heating conditions, the distribution density of precipitates with fixed crystalline structures increases first as the heating time prolongs, as shown in Figure 6(A).There is a formation process under 900℃ heating conditions, which means that precipitates with fixed crystalline structures will form continuously during the heating process.So the distribution density of precipitates with fixed crystalline structures rises in earlier stages and then decreases.At other heating temperatures, the distribution density of precipitates with fixed crystalline structures will decrease all the time.The distribution density of precipitates with fixed crystalline structures decreases as the heating temperature rises.It should be noted that the sample with a Y/Ti atomic ratio equal to 0.99 has the highest distribution density of precipitates with fixed crystalline structures, no matter how the heating conditions change.The distribution density of precipitates with fixed crystalline structures in the samples with a Y/Ti atomic ratio equal to 0.99 is one order higher than other samples.But the distribution density peak value of nanoclusters is lower than other samples.Because Y, and Ti elements, which are the main elements composed of nanoclusters, are consumed by forming precipitates with fixed crystalline structures.It is clear that chemical compositions and heating conditions have a crucial effect on precipitate distribution characterization.It is usually accepted that alloy elements are solid solutions in the matrix and form supersaturated solid solution alloy powders during the MA process.The precipitates would precipitate during the forming process, because element atoms would diffuse alone with dislocation at high temperatures based on Fick law, and longer heat time means a farther diffusion distance.The Y/Ti atomic ratio of nanoclusters that exist in ODS steels is close to 0.6, which is near to the sample with a Y/Ti atomic ratio of 0.42.So the sample with a Y/Ti atomic ratio of 0.42 has the highest distribution density of nanoclusters.Yang [15] manifested that the sample with a Y/Ti atomic ratio of 0.42 had a higher distribution density of precipitates by means of TEM.The sample with a Y/Ti atomic ratio of 0.99 has the highest distribution density of precipitates with fixed crystalline structures because the Y/Ti atomic ratio of the sample is close to Y 2 Ti 2 O 7 , which is usually formed in ODS steels.Chinnappan [16] illustrated that Y 2 Ti 2 O 7 would precipitate preferentially in ODS steels due to the most standard formation of enthalpy.So Y 2 Ti 2 O 7 precipitation could decrease the system energy and result in system stabilization.
Figure 7 is the hardness sample under the optical microscope.The alloy powders in resin show a bright contrast.So the pressure head could work accurately at alloy powders.Each sample is measured many times at different alloy powders and the average value is calculated as the hardness value of the sample.The references illustrate that oxides like Y 2 O 3 , Y 2 Ti 2 O 7 , and Y 2 TiO 5 dispersed in ODS steels are hard phases that can pin dislocations and improve the hardness values of the sample [17].The oxides usually have different crystallite structures from the Fe matrix, so the more the amount of oxides is, the higher the hardness values of samples are. Figure 7 illustrates that the oxide distribution density of the sample with a Y/Ti atomic ratio of 0.99 is the highest.So the hardening effect of oxides in the sample with a Y/Ti atomic ratio of 0.99 is most obvious.The heat treatment would decrease the hardness values of MA alloy powders by means of removing the work-harden effect, resulting in the hardness value decreasing.So most heat treatment alloy powders are softer than MA alloy powders.

Conclusion
In this research ODS alloy powders with different Y/Ti atomic ratios are prepared by MA method and heat treated by different parameters to simulate the forming technology parameters.The hardness values of alloy powders after heat treatment are also measured at the same time.The following conclusions can be obtained: (1) The alloy elements are solid solutions in matrix and forming supersaturated solid solution alloy powders.
(2) The sample with a Y/Ti atomic ratio of 0.42 has the most nanoclusters after heat treatment, and the sample with a Y/Ti atomic ratio of 0.99 has the most oxides after heat treatment.
(3) The sample with a Y/Ti atomic ratio of 0.99 has the highest hardness value.(4) Raising heat temperature or prolonging heat time will promote precipitates to grow up and decrease distribution density.
(5) The ODS steels should have a Y/Ti atomic ratio close to 0.42.

Figure 2 .
Figure 2. The results of MA alloy powders.(A) second electron image (B) backscatter electron image (C) XRD result of MA alloy powders

Figure 4 .
Figure 4.The distribution characterization of precipitates in samples with a Y/Ti atomic ratio of 0.14 (A) fixed crystalline structure precipitates (B) nanoclusters.

Figure 5 .
Figure 5.The distribution characterization of precipitates in samples with a Y/Ti atomic ratio of 0.42 (A) fixed crystalline structure precipitates (B) nanoclusters.

Figure 6 .
Figure 6.The distribution characterization of precipitates in samples with a Y/Ti atomic ratio of 0.99 (A) fixed crystalline structure precipitates (B) nanoclusters.

Figure 7 .
Figure 7.The hardness sample of alloy powders.

Figure 8
Figure 8 is the hardness value of alloy powders with different Y/Ti atomic ratios under different heat conditions.The hardness values of MA powders with different compositions are different because composition difference means different solid solution atoms in the matrix, which results in different

Figure 8 .
Figure 8.The hardness values of alloy powders under different heat conditions.(A) Y/Ti atomic ratio 0.14 (B) Y/Ti atomic ratio 0.42 (C) Y/Ti atomic ratio 0.99 It is clear that MA powders have the highest hardness value except for samples heated at 900℃ for 10 min due to the vast deformation during the MA process.All samples become soft at 1000℃, 1100℃, and 1200℃ heating conditions, and the heating time is longer for the softer samples because the work-harden phenomenon is dispersed during the heating process.The higher the temperatures are, the faster the work-harden phenomenon disperses.All the hardness values of samples heated at 900℃ for 10 min increase slightly as MA alloy powders.The hardness values of alloy powders heating at 900℃ express an increased tendency with the heating time prolonging, which is different from samples heating at 1000℃, 1100℃ and 1200℃.Figure8(B)shows the hardness of the Y/Ti atomic ratio equal to 0.42 as heating time prolongs under different temperatures.Though the hardness values of samples change as heating time changes, the values are still close to the value of the sample heated at 900℃ for 10 min.It should be noted that the difference of the hardness values of samples with a Y/Ti atomic ratio equal to 0.99 (Figure8(C)) between 900℃, 1000℃, 1100℃ and 1200℃ are most obvious.The alloy powders with a Y/Ti atomic ratio equal to 0.99 have the highest hardness values during heat treatment.