MA Benyuan, PU Chunying, HE Junbao, ZHOU Dawei, FU Yü,WEI Nian, LU Tiecheng＊

(1. College of Physics and Electronic Engineering, Nanyang Normal University, Nanyang 473061, China; 2. Department of Physics and Key Laboratory for Radiation Physics and Technology of Ministry of Education, Sichuan University, Chengdu 610064, China)

Abstract: We investigated the impure phase problem and summarized its two formation mechanisms of YAG powders synthesized via the co-precipitation method. The ions loss problem caused by high concentration reaction solution in the titration process was emphatically studied, and the corresponding thermodynamic explanation was carried out. In addition, influence of powder crystallinity and its new qualitative and quantitative standards were studied. One reason of impure phase is the local nonuniform mixture of Y and Al elements in precursor, which easily causes intermediate phases during calcination and difficulty of high pure powders at low temperatures. The other reason is the precipitation dissolution during titration and then the Y3+/Al3+ loss, caused by high concentration of reaction solution. The powder crystallinity can be promoted by increasing calcination temperature or holding time of precursor. Besides the routine XRD method, the TEMEDX method should be also introduced to directly determine the quality of crystallinity.

Key words: transparent ceramics; phase transition; crystallinity; co-precipitation; activity coefficient

1 Introduction

Yttrium aluminum garnet (Y3Al5O12, YAG)transparent ceramic is an important optical gain medium, with compatible properties of both single crystal of physicochemical properties and glass material of shape-controllable structures. It attracts much attention thanks to its applications as laser material, scintillation material and fluorescent material after doped with the rare-earth ions of Nd3+, Yb3+,Er3+, Gd3+, Ce3+, Ho3+, Gd3+,etc[1-23]. For instance, Nd3+doped YAG (Nd:YAG) transparent ceramics show good laser performances[24-27]. Ce:YAG transparent ceramics can substitute Ce:YAG phosphors and epoxy resin for the ceramic LED equipment[28]. It is especially worth mentioning that rare-earth doped YAG transparent ceramics have huge potential application on the gain of nuclear fusion ignition device and tactical laser weapons.

YAG powder is the fundamental raw material for transparent ceramics fabrication, thus powder characteristics have a crucial influence on the phase structure, microstructure and optical performance of final ceramics. The sequent failure of ceramic fabrication and then a cost waste can be easily caused if powder problems are not clearly understood. Powder characteristics can be summarized into two factors. One is powder TEM morphology, the other is its intrinsic XRD phases and crystallinity. However, most attention is only paid to powder morphology,eg, controlling of particle sizes, optimization of size distribution and elimination of powder agglomeration. Though high pure YAG powders have been obtained in many ways for high transparent ceramics, the formation mechanism of impure phase problem of powders is not clearly opened to public. It is interesting that there is no special literature to analyze and summarize its formation reasons even impure phase problem usually happen in routine fabrication process. As to crystallinity, its original purpose is to characterize organic polymer powders with a definition of the ratio of crystallized powder weight to total powder weight. For optical level required transparent ceramics, which requires a nearly full XRD phase purity of powders, powder crystallinity may lose its original meaning because the crystallized powder weight calculated from XRD results basically equal to the total powder weight. Furthermore, it is not accurate enough to characterize powders only with XRD phase purity, because the amorphism and lattice defects cannot be well determined by routine XRD method, and these poorly crystallized particles still can escape XRD detection and sneak into raw powders,causing a bad influence to final ceramics similar as the role of XRD impure phases. Thus, it needs a new qualitative and quantitative standard for the powder crystallinity of transparent ceramics.

There are many methods adopted to synthesize YAG powders, such as solid-state reaction, coprecipitation, sol-gel process, hydrothermal reaction[29-37]. Compared to other methods, co-precipitation is a potential and useful method for obtaining well-dispersed nanopowders with good chemical homogeneity and narrow particle size distribution. It has been revealed that by choosing suitable precipitant,adjusting reactive solvent or controlling reactive pH,this method can ensure the co-precipitation of Y3+and Al3+precipitations, which is further beneficial to the uniform mixture of reactive elements, the reducing of calcination temperature and the fine nanoparticles[38-41].However, it is short of deep understanding and literatures on powder phase formation mechanism and powder crystallinity of co-precipitation method. In this paper, impure phase problem and its corresponding two formation mechanisms were studied during the synthesis of YAG powders via the co-precipitation method. The ions loss during titration of the coprecipitated process was emphatically studied, and the corresponding thermodynamic explanation was carried out. In addition, powder crystallinity and its influence on TEM morphology and surface elements composition were investigated. The qualitative and quantitative standards were re-summarized.

2 Experimental

Y(NO3)3·6H2O (99.99%) and Al(NO3)3·9H2O(99.99%) were mixed in the molar ratio of 3:5 according to Y3Al5O12(YAG) in 600 mL distilled water, with Al3+concentration of 0.08-1 mol/L. The precipitant solution was prepared by dissolving NH4HCO3(analytical grade) in 2 500 mL distilled water, obtaining a NH4HCO3concentration of 0.5-3 mol/L. In the pH experiment, 0-8 g NH4NO3was added into NH4HCO3solution with the mixed solution as precipitant solution. Then the salt solution was dripped into the precipitant solution at a speed of 3 mL/min under 200 rpm/min stirring at 20 ℃. After aging for 24 h, the precipitation suspension was vacuum filtered.The obtained precipitation was washed with distilled water and ethanol, dried in a drier. Finally, the obtained precursor was divided into several heaps and calcined at 700-1 250 ℃ for 2-6 h to obtain various YAG powders.

The pH values of precipitant solution were monitored per 5 min during titration process with a pH meter (PHS-3C, Shanghai precision and scientific instrument Co., LTD, China). To investigate ion loss problem, the ion concentration in the filtrate of precipitation suspension during vacuum filtration process was detected by inductively coupled plasma atomic emission spectrometer (ICP-OES, Spectro Arcos, Germany). The chemical bond compositions of precursor were investigated by Fourier transformation infrared spectroscope (FT-IR, Perkin-Elmer Company,Spectrum one). The phase compositions, lattice constants and full width at half maximum (FWHM)were identified by X-ray diffractometer (XRD, Cu-Kα radiation 1.54 Å, D/max-rA, Rigaku, Japan).The powders morphologies were observed by a transmission electron microscopy (TEM, JEM-100CX II, Tokyo, Japan). The lattice observation and surface element analysis of powder particles were performed on a high resolution electron microscopy (JEOL-3011,JEOL, Japan) and energy dispersive analysis system of X-ray (EDX, Quantax400, Bruker, Germany).

3 Results and discussion

3.1 Chemical reaction mechanism during titration

The precipitant of NH4HCO3could be hydrolyzed in distilled water as

When the mixed salt solution of Y3+and Al3+ions was dripped into the precipitant solution in titration process, the mainly precipitation components of precursor were generated as follows:

Under mechanical stirring, the ideal result was that these precipitation components were coprecipitated and mixed in molecular level. Y and Al elements should be mixed uniformly according to the 3:5 ratio of Y3Al5O12, because this uniform mixture was very beneficial to reducing the calcination temperature of precursor and promoting the sphericity degree of powder particles.

3.2 Influence of local element mismatch in precursor on XRD phases of powders

In fact, it was very difficult to keep the 3:5 ratio mixture of Y and Al elements. The influence factors on the titration process of co-precipitation method mainly were reaction solution pH, solution concentration,ambient temperature, titration rate,etc[40-44]. When an improper control on these factors, Y and Al elements could not be mixed uniformly according to the 3:5 ratio of Y3Al5O12at each micro-region of precursor,and elements mismatch might be caused. Subsequently,during the calcination process of precursor, XRD intermediate phases might be generated in elements mismatch micro-region. To eliminate intermediate phases, powder calcination temperature must be rose to promote lattice diffusion between intermediate phases to form YAG phase. If the calcination temperature was not high enough, or the calcinations time was not long enough, the intermediate phases could not be eliminated completely and then residual in final powders as XRD impure phases.

Fig.1 investigated one influence factor of reaction solution pH on powder calcination temperature and powder morphology[45]. The S1 curve in Fig.1(a)showed the pH change of NH4HCO3precipitant solution during titration process. It could be seen that pH values varied intensely in a wide range from 7.4 to 9.3. To avoid this intense pH change, NH4NO3was added to form buffer solution as shown in S2 curve.It could be observed that pH change of S2 curve was controlled in a narrow range with an alleviated variation trend. As shown in equation (2), the OHconsumption by the formed precipitations during titration process would cause the equation (2) turn right.Meanwhile, the addition of NH4NO3would increase the NH4+concentration and promote the equation (2) turn left. There was a balance between them to inhibit the intense change of OH-concentration, so the pH change of S2 curve was controlled in a narrow range.

Fig.1 (a) pH changes of different precipitant solutions during titration process, S1: NH4HCO3 precipitant solution, S2: NH4HCO3 precipitant solution added with 8 g NH4NO3; (b) FT-IR curves comparison of the corresponding precursors. XRD phase evolution comparison of the precursors during calcining process; (c) NH4HCO3; (d) NH4HCO3 with 8 g NH4NO3. TEM comparison of final powders: (e)NH4HCO3 and (f) NH4HCO3 with 8 g NH4NO3. Reproduced with permission from Ref. [45], © Springer 2010

Fig.1(b) shows the FT-IR spectra of precursors with and without NH4NO3. The main peak positions in two curves roughly consisted with each other except for transmittances, indicating that the main compositions and chemical bonds varied little in the two pH ranges of precursor precipitations. However, pH changes had an important impact on the XRD phase structure of powders. Because the needed pH environments for initially generating precipitations of Y3+and Al3+were different (eg, pH values of 3.3, 4.5, and 6.9 for Al(OH)3, Y2(CO3)3, Y(OH)3, respectively[41]), and the deposition velocities of Y3+and Al3+precipitations might have distinct deviation at different pH values. As shown in S1 curve in Fig.1(a), pH values of precipitant solution without NH4NO3declined drastically. This might cause unfavorable precipitating sequence or deposition velocities, which may be not beneficial to the mixture Y3+and Al3+precipitations towards the 3:5 ratio at local region. Though it was mechanical stirred at the same rate of 200 rpm/min in titration process, these precipitation components might not be mixed sufficiently in molecular level because of the strong adsorption interaction of colloid precipitations.The local nonuniform mixture of Y and Al elements in precursor might cause intermediate phases during calcination process. As could be observed in Fig.1(c),there remained many intermediate phases until 1 100oC during the calcinations process of S1 precursor.This result indicated the elements mismatch of Y and Al in the local micro-region of S1 precursor.In contrast, no intermediate phase appeared in the calcination process of S2 precursor in Fig.1(d), and pure phase YAG powder could be obtained at a lower temperature of 900 ℃. These results indicated that the uniform mixture of Y3+and Al3+precipitations was very important to reduce the final calcination temperature of YAG powders. Though the addition of NH4NO3could alleviate the intense pH change, the adding amount should be controlled at appropriate value. It was found that impure phases would be caused as the adding amount exceeded 9 g[45]. The reason was attributed to the increasing ion concentration and unfavorable charge environment which might cause the Y3+/Al3+loss as discussed in section 3.3.

Fig.2 shows the schematic of the formation process of YAG phase from intermediate phases in the calcination process of precursor. At the micro-region of precursor with uniform mixture of Y and Al elements,Y, Al and O could directly react according to 3:5:12 to form Y3Al5O12(YAG) phase, while at the element mismatch micro-region (Y:Al ≠ 3:5) Y, Al, and O had to react to form other intermediate phases, such as YAlO3(YAP) and Y2Al4O9(YAM). Subsequently, YAG only could be formed by increasing the calcination temperature or prolonging the calcination time to promote the reaction between these intermediate phases. Obviously, one reason of the XRD impure phases in final powders was the insufficient elements diffusion between intermediate phases if calcination temperature or time was not high or long enough.

Fig.2 Schematic of the formation process of YAG phase from intermediate phases in the calcination process of precursor, which has local element mismatch of Y and Al

In addition, the intermediate phases had an adverse impact on the morphology of powder particles.During calcination, the crystallites at the micro-region having formed YAG phase could continue the growth to form large YAG grains, while the formation of YAG grain was delayed at the micro-region of intermediate phases before it was eliminated. As could be observed in Fig.1(e), the S1 particles present significant size difference, while the particles of S2 powders in Fig.1(f)show a relatively better size uniformity. Obviously,to obtain spherical powder particles, one effective method was to keep the uniform mixture of Y and Al elements in precursor to prevent the formation environment of intermediate phases. Similarly, the sufficient ball milling of oxide raw materials of the solid-phase reaction method, or the benefit of the mixture of reaction solution of the sol-gel method, had the same purpose of keeping the uniform mixture of Y and Al elements in precursor to low the calcination temperature or better the particles morphology.

3.3 Influence of ions loss during titration on XRD phases of powders

Table 1 shows the influence of reaction solution concentration during titration on the XRD phase structure of final powders (1 100 ℃ 2 h calcination). It could be observed that, as the concentration of Al3+salt solution (keep Y3+:Al3+= 3:5) or NH4HCO3precipitant solution increased, more impure phases such as Y2O3,YAP and YAM appeared. This result indicates that high concentration of reaction solution was not beneficial to obtaining pure phase YAG powders. Fig.3 shows the impure phase structure of the YAG powder from 4#precursor by calcining at higher temperature of 1 200℃ and for longer time of 4 h. It was reported (or saw the results in Fig.1) that YAG phase could be formed at 900 ℃. Thus, 1 200 ℃ should be high enough for easily obtaining pure phase YAG powders even if concerning the appearance of intermediate phases. In the experiment, 1.8 ppm Al3+ions had been detected in the filtrate of 4# precursor precipitation during vacuum filtration process, indicating Al3+ions loss.Fig.4 showed the schematic of the formation process of XRD impure phases of YAG powders calcined from 4# precursor. At the element mismatch micro-region(Y:Al ≠ 3:5), Y, Al and O firstly reacted with each other to form intermediate phases. Continuously increasing the calcination temperature, only part of intermediate phases could be transformed into YAG phase because of the Y3+/Al3+ions missing had been caused by ions loss during vacuum filtration process. Hence, the other reason of XRD impure phases in final powders was the precipitation ions loss of Y3+/Al3+.

Fig.3 XRD pattern of the powder calcined at higher temperature of 1 200 ℃ for 4 h from 4# precursor in Table 1. 1.8 ppm ICP concentration of Al3+ ions had been detected in the filtrate of precursor precipitation during vacuum filtration,indicating Al3+ ions loss

Fig.4 Schematic of the formation process of impure phases of YAG powders calcined from S4 precursor, which concerned Y3+/Al3+ ions loss(the total element ratio of Y: Al is not equal to 3:5 in precursor)

Table 1 XRD phase structure of YAG powders as the reaction solution concentration varied

The Y3+/Al3+ions loss was deeply related to the high concentration of reaction solution during titration process. For example, few part of the generated precipitation of Al(OH)3could be redissolved as the solution environment (eg, reaction temperature, electric field, pH,etc) changed[47]. As shown in equation (3),free Al3+and OH-ions in reaction solution collide to generate Al(OH)3molecules which are then adsorbed onto the surface of colloid precipitation. Meanwhile,some Al(OH)3molecules are redissolved because of the attraction force of both Al3+by ambient anions and OH-by cations[47]. Precipitation could exist steadily only this precipitation-dissolution came into balance.In the high-concentration reaction solution, the ionic action of Al3+and OH-on the right side of equation(3) is strongly limited by ambient anions (the limited Al3+and OH-ions could be considered as dead ions or nonmoving ions). The equation (3) has to turn right to maintain the chemical equilibrium because actually the living ions of Al3+and OH-are scarce, thus causing the Al(OH)3dissolution. In this way, some Al3+ions would be lost as the precipitation suspension was filtrated.

The ions loss could be theoretically explained by principle of solubility product. The effective ion concentrationa(living ions) and ambient ion strengthIcould be shown as follows:

where,c,f,d, and A/B are theoretical concentration,activity coefficient, smallest ion distance, and Debye-Huchel constants, respectively. The variables ofmiandziare mass molar fraction and valence charge ofi-kind ions, respectively[47]. It could be observed that effective ion concentrationa(or activity coefficientf) would decline with increasing ion strengthI, which is proportional to mass molar fractionmiand valence chargezi. For the high concentration of reaction solution of YAG during titration process, more the ambient ions (NO3-, CO32-,NH4+, OH-, Al3+, Y3+, and H+) contribute a high value of ion strengthI, which strongly reduces the effective ion concentrations of Al3+and OH-on the right side of equation (3).

When equation (3) turns right, the corresponding change of Gibbs free energy is:

where,R,T, andKөare universal constant,thermodynamic temperature and solubility product constant (Kө= 1.3×10-33for Al(OH)3), espectively.aAl3+andaOH-,cAl3+andcOH-,fAl3+andfOH-represent the effective ion concentration, theoretical concentration,activity coefficient of Al3+and OH-, respectively.According to thermodynamic principle, chemical reaction proceeds as Gibbs free energy decreases. Thus,the precipitation-dissolution relation of Al(OH)3in equation (3) could be shown as follows:

AsaAl3+oraOH-declines, ΔG＜0, equation (3)turns right, Al(OH)3is dissolved.

AsaAl3+oraOH-remains, ΔG=0, equation (3)keeps nonmoving, precipitation-dissolution balance.

AsaAl3+oraOH-increases, ΔG＞0, equation (3)turns left, Al(OH)3is precipitated.

It could be concluded from these results that high concentration of reaction solution limits the free action of Al3+and OH-in solution and then causes a decline of effective ion concentration, which promotes the dissolution of Al(OH)3precipitation in the titration process. This dissolution causes the ions loss of Y3+/Al3+and XRD impure phases in final powders.

3.4 Investigation of powder crystallinity

Crystallinity of YAG powders concerned the disordered lattice structures of particles. It was found in our previous literature that poor powder crystallinity contributed more the inhomogeneous ceramic sintering and introduced more the defects in final ceramics,eg,impure phase inclusions and dislocations[48]. Except for relating to the uniform mixture of Y and Al elements in precursor, which affected the homogeneous powder sintering and element diffusion, crystallinity of YAG powders mainly related to calcination schedule of precursor. Figs.5(a) and 5(b) show the XRD patterns of 2at% Nd:YAG powders calcined from precursor at 1 100-1 250 ℃ for 2-6 h. Powder crystallinity could be promoted by increasing calcination temperature(Fig.5(b)) or holding time (Fig.5(a))[48]. As powder crystallinity gradually promotes, the XRD peak intensity increases and the FWHM becomes narrow (Fig. (c)). The actual main diffraction peak (420), which has position deviation (caused by lattice distortion), gradually shifts toward the standard position of 33.36 in JCPDS card in Fig.5(d), causing the lattice constant gradually reduced toward the standard value of 12.01 Å[48].

Fig.5 Crystallinity comparison from XRD results: XRD patterns of powders calcined from precursor (a) at 1 100 ℃ for varied time, S1-S2-S3: 2-4-6 h; (b) for 4 h but at different temperatures, S2-S4-S5-S6: 1 100-1 150-1 200-1 250 ℃; (c) XRD peak relative intensity(red line) and FWHM (blue line) comparison; (d) Lattice constants comparison (black line), actual XRD (420) peak positions comparison (purple line). Reproduced with permission from Ref. [48], © Elsevier B.V. 2017

Fig.6 shows the crystallinity comparison between S1 and S6 powders in Fig.5 from TEM-EDX results[48].It could be observed that poor crystallinity of S1 powder shows inadequate growth, wide sintering neck and with lattice defect layer on particle (Figs.6(a), 6(c),and 6(e)). The element molar ratio of (Y+Nd):Al of the defect layer seriously deviates from the true value 3:5 of the stoichiometric ratio of Y3Al5O12(Fig.6(g)).In contrast, good crystallinity of S6 powder shows adequate growth and narrow sintering neck, without lattice defect layer on particle (Figs.6(b), 6(d), and 6(f)). The element molar ratio of (Y+Nd):Al shows a negligible deviation considering measuring error(Fig.6(h)).

The difference of various crystallinities could be roughly compared by the XRD method in Fig.5, but powder crystallinity of one could not be compared or determined directly. Furthermore, the detected results by routine XRD method were only the crystallized powder particles, with amorphism and lattice defects not being well determined. Hence, besides of the XRD method in Fig.5, the TEM-EDX method in Fig.6 should be introduced to directly determine the quality of crystallinity. From the above results, the new qualitative and quantitative standards for the good powder crystallinity of transparent ceramics could be summarized as follows: XRD results: pure phase,high XRD peak intensity, narrow FWHM and standard lattice constant. TEM-EDX results: without obvious lattice defects in particle or defect layer on particle,without obvious deviation of element molar ratio on particle surface.

Fig.6 Crystallinity comparison from TEM-EDX results: TEM analysis of (a)(c)(e) S1 powder, (b)(d)(f) S6 powder. EDX analysis near the surface layers of (g) S1 powder, (h) S6 powder. Reproduced with permission from Ref. [48], © Elsevier B.V. 2017

4 Conclusions

We investigated the formation mechanisms of impure phase problem of YAG powders synthesized via the co-precipitation method. In addition, influence of powder crystallinity and its new qualitative and quantitative standards were studied. The results show that two formation mechanisms are responsible for powder impure phase. One is the local nonuniform mixture of Y and Al elements in precursor, which easily causes intermediate phases during calcination and difficulty on obtaining high pure powders at low temperature. The other mechanism is that high concentration of reaction solution limits the free action of precipitation ions in solution, for example Al3+and OH-, and then causes a decline of effective ion concentration which promotes the dissolution of Al(OH)3precipitation in the titration process. This dissolution causes the ions loss of Y3+/Al3+and then the XRD impure phases in final powders. Crystallinity of YAG powders mainly concerns the disordered lattice structures of particles. Poor powder crystallinity contributes more the inhomogeneous ceramic sintering and introduces more the defects in final ceramics.Both routine XRD method and the required TEMEDX method should be introduced to well analyze the quality of crystallinity.