Reaction mechanism and kinetic analysis of the formation of Sr2SiO4 via solid-state reaction

(1)

Reaction mechanism and kinetic analysis of the formation

of Sr

2

SiO

4

via solid-state reaction

Chung-Hsin Lu

∗

, Po-Chi Wu

Electronic and Electro-optical Ceramics Lab, Department of Chemical Engineering, National Taiwan University, Taipei, Taiwan, ROC

Received 7 August 2007; received in revised form 12 November 2007; accepted 16 November 2007 Available online 22 November 2007

Abstract

The kinetics for the formation of Sr2SiO4during solid-state reaction is investigated using an isothermal method. TG/DTA and XRD analysis

suggest a direct reaction between SrCO3 and SiO2 powders. The conversion ratios of the Sr2SiO4 starting materials are calculated from the

weight loss. Based on the reaction kinetic isothermal analysis, Sr2SiO4 formation is corroborated as controlled by the Brounshtein-Ginstling’s

diffusion-controlled model. The formation process activation energy is estimated to be 139.6 kJ/mol. According to microscopic observations, the microstructures vary drastically at 830◦C, implying a reaction initiation involving reactive SrCO3and SiO2. In view of the diffusion controlled

mechanism and microstructural observations, a reaction model for the formation of Sr2SiO4has been established.

Keywords: Sr2SiO4; Kinetics; Solid-state; Phosphor

1. Introduction

White light illumination applications have increased in recent years. Traditionally, white light is produced by a phosphor layer coated onto the inner side of lamp tubes under excita-tion around 280 nm. However, white light has the drawbacks of low power efficiency, reliability and lifetime. In comparison, the light emitting diode (LED) could be considered the next generation light illumination source due to its high reliability and low energy consumption. The first high brightness GaN-based blue light emitting diode prototype was developed in 1993

[1].

To obtain white light through GaN-based LEDs, lumines-cent materials coated on the top of an LED chip is a promising approach. However, the excitation source provided by blue emitting LEDs (400–490 nm) is different from conventional excitation sources. Consequently, phosphors with different exci-tation characteristics are required to achieve efficient light conversion. YAG:Ce3+ (Y3Al5O12:Ce3+) with a highly stable

structure is a commercially available yellow phosphor used

∗_{Corresponding author.}

E-mail address:chlu@ntu.edu.tw(C.-H. Lu).

for white LEDs [2–4]. However, it has problems of low color stability with increasing applied current, low color rendering index, and low color reproducibility [5,6]. Another new type of phosphor-Sr2SiO4:Eu2+ has attracted researchers’ attention [7–10]. Sr2SiO4provides the broadband absorption in UV/Blue

region due to low symmetry of the crystallographic sites. In addition, the Sr2SiO4:Eu2+phosphor has a higher luminous

effi-ciency, CRI and color stability than YAG[10,11], giving rise to a new phosphor approach for white LED applications. There-fore, Sr2SiO4 is a suitable host lattice for phosphor

applica-tions.

Sr2SiO4powder is usually prepared in a solid-state reaction

by heating mixed strontium and silicon salts at elevated temper-atures. Although the photoluminescent properties of Sr2SiO4

have been explored, the formation mechanism and reaction kinetics have not been studied. Knowledge of the fundamental reaction kinetics and mechanism are important when optimizing the solid-state process for phosphor applications.

In this study, Sr2SiO4powders were prepared via solid-state

reaction by heating mixed precursors at elevated temperatures. The purpose of this study is to elucidate the reaction mechanism and reaction kinetics of formation of Sr2SiO4 in a solid-state

reaction. The precursors were examined using thermal and X-ray diffraction analysis to determine the optimum reaction

(2)

range. The isothermal analysis was adopted to understand the reaction mechanism and kinetics. Using the microstructural observations accompanied by the kinetics, a reaction model for the formation of Sr2SiO4 via the solid-state reaction is

proposed.

2. Experimental

Analytical grade strontium carbonate (SrCO3, Aldrich Chemicals, 99.9%), and silicon dioxide (SiO2,−325 mesh, Aldrich Chemicals, 99.6%) were mixed in their stoichiometric ratio according to the Sr2SiO4 chemical formula. The mixture was ball milled with ethyl alcohol as a dispersing agent and zirconia (ZrO2) ball for 24 h. The slurry was subsequently dried in a vacuum-rotation dryer.

Differential thermal analysis (DTA) and thermogravimetry analysis (TGA) were applied for tracing the reaction processes. The heating rate was 10◦C/min with alumina powder used as a reference. For isothermal heating experiments, the heating temperatures were set at 700◦C, 750◦C, and 800◦C. The heated samples were soaked after different heating time at the above three temperatures. The phase and purity of the heated powders were examined using the powder X-ray diffraction method with a X-ray diffractometer (MAC Science MXP3). The morphology, grain size, and microstructure of the products were analyzed using a scanning electron microscope (SEM, Hitachi S-800).

3. Results and discussion

3.1. Formation process of Sr2SiO4

The powder derived from mixing SrCO3and SiO2was soaked

between 600◦C and 1000◦C, at 100◦C intervals. The corre-sponding XRD patterns are illustrated inFig. 1. According to the XRD results shown inFig. 1, it was found that after quench-ing at 600◦C, no difference was exhibited between the sample and the precursor. This implied that SrCO3 and SiO2did not

Fig. 1. X-ray diffraction patterns of (a) raw materials of Sr2SiO4, and the samples quenched at (b) 600◦C, (c) 700◦C, (d) 800◦C, (e) 900◦C, and (f) 1000◦C.

react at below 600◦C. As the firing temperature increased, more Sr2SiO4product was formed, with a corresponding decrease in

SrCO3and SiO2. When the quenching temperature was raised to

900◦C, the diffraction peaks of SrCO3and SiO2decreased

dra-matically and nearly all of the major diffraction peaks conformed with the standard powder diffraction profile of Sr2SiO4 in the

ICDD database, No. 761494[12]. Sr2SiO4formation was

com-pleted when the temperature reached 1000◦C, and no SrCO3

and SiO2could be detected. Because no intermediate phase was

detected at temperatures ranging from 600◦C to 1000◦C, the Sr2SiO4formation process was confirmed to be a direct reaction

between SrCO3and SiO2.

Fig. 2illustrates the DTA/TGA curves for the Sr2SiO4

precur-sors heated at rate of 10◦C/min. The TG-DTA analysis showed two stages of weight loss accompanied by two endothermic peaks. One small endothermic peak at 450◦C in DTA curve, corresponding to the weight loss shown in TG, was due to pre-cursor dehydration. An apparent weight loss occurred at around 700◦C, and no further weight loss was found at temperatures higher than 980◦C. To explain the broad endothermic peak at around 820◦C, the DTA/TG analysis for pure strontium carbon-ate and silicon dioxide (SiO2) were also performed. According

to our experimental results, the silicon dioxide seemed to be stable in comparison with strontium carbonate in the range of 25◦C to 1000◦C. The results are not shown here. As a result, the endothermic peak at 820◦C shown inFig. 2was attributed to reactive strontium carbonate reacting with stable silicon oxide, thereby leading to the formation of Sr2SiO4. The total weight

loss measured from TG experiment amounted to 24.7%, which was closed to the theoretical weight loss of this reaction. The net equation of the reaction involving SrCO3and SiO2is given

in Eq.(1).

2SrCO3+ SiO2→ Sr2SiO4+ 2CO2 (1)

Fig. 2. (a) Differential thermal analysis and (b) thermogravimetry analysis for the starting materials of Sr2SiO4at a heating rate of 10◦C/min.

(3)

Table 1

Stoichiometric table for the formation of Sr2SiO4

Species Initially Change Remaining

SrCO3 2X 2αX 2X(1− α)

SiO2 X αX X(1− α)

Sr2SiO4 0 0 αX

CO2 0 0 2αX

X: specific moles.α: fraction converted to Sr2SiO4.

3.2. Reaction kinetics analysis

During the reaction indicated in Eq. (1), the conversion ratio of Sr2SiO4can be calculated from the weight loss of the

samples. The detailed calculations are described as follows. According toTable 1, the weight before the reaction (Winitial)

is 2XMSrCO3 + XMSiO2; the weight after the reaction (Wfinal) is

2X(1 − α)MSrCO3+ X(1 − α)MSiO2+ aXMSr2SiO4. SetW%

as the percentage of weight loss after the quench experiment. Make mass balance as follows:

[2XMSrCO3+ XMSiO2]· W%

= Winitial− Wfinal

= 2αXMSrCO3+ αXMSiO2 − αXMSr2SiO4. (2)

Rearranging Eq.(2)gives

α = (2MSrCO3+ MSiO2)· W%

(2MSrCO3+ MSiO2− MSr2SiO4) =

W%

24.77%. (3) where X is specific moles,MSrCO3 is the molecular weight of

SrCO3,MSiO2 is the molecular weight of SiO2,MSr2SiO4 is the

molecular weight of Sr2SiO4, andα is the fraction converted to

Sr2SiO4. Take the dehydration of precursors into consideration;

the weight loss was 0.92% in TG experiment, hence, the exact conversion of the reaction is revised as

α = W% − 0.92%

24.77% . (4)

As deduced fromFigs. 1 and 2, for the reaction to proceed, the temperature should be higher than 700◦C. Therefore, the reac-tion temperatures were chosen at 700◦C, 750◦C and 800◦C, respectively. The mixed Sr2SiO4precursors were isothermally

heated at these temperatures for various periods of time and then quenched to room temperature. The weight differenceW of each specimen before and after the heating process was recorded. The conversion ratio of Sr2SiO4formation under each heating

condition was calculated by Eq.(4).Fig. 3shows the relation between the conversion ratios and reaction conditions. At 700◦C and 750◦C, the conversion ratio monotonously rose with reac-tion time. The reacreac-tion was nearly completed at 800◦C after 60 min. In addition, it was noted that for the same reaction period, the conversion increased with a rise in the heating tem-perature. Heating specimens at 700◦C for 120 min increased conversion ratio to about 71%. After reacting for 120 min, the conversion ratios at 750◦C and 800◦C were 95% and 97%, respectively.Fig. 3indicates that the conversion was enhanced with an increase in reaction temperature and time.

Fig. 3. Conversion ratio of Sr2SiO4versus reaction time.

Fig. 4. X-ray diffraction patterns of the starting materials of Sr2SiO4heated at (a) 700◦C for (I) 20 min, (II) 40 min, and (III) 60 min, and (b) 750◦C for (IV) 20 min, (V) 40 min, and (VI) 60 min, and (c) 800◦C for (VII) 20 min, (VIII) 40 min, and (IX) 60 min.

(4)

Fig. 4 illustrates XRD patterns of the obtained powders quenched at 700◦C for various reaction time. There was no reaction between SrCO3 and SiO2after heating at 700◦C for

20 min, since only the XRD diffraction peaks of SrCO3and SiO2

were observed. As the reaction time was prolonged to 40 min, the (1 2 1) and (1 0 3) planes of Sr2SiO4appeared, indicating a

small amount of Sr2SiO4began to form. With the increase in

reaction time, the intensity of XRD peaks of Sr2SiO4increased

rapidly; whereas the intensities of the peaks of SrCO3and SiO2

decreased correspondingly. The XRD results for the samples heated at 750◦C and 800◦C are also observed in Fig. 4. As the heating temperature was increased, the diffraction peaks of Sr2SiO4 were found at less reaction time and those peaks

of SrCO3 and SiO2 diminished after heating for 40 min at

800◦C.

To analyze the Sr2SiO4reaction kinetics, the Hancock and

Sharps’ method based on the Johnson-Mehl-Avrami equation was adopted [13,14]. The Johnson-Mehl-Avrami equation is presumed as:

α = 1 − exp(−rtm) (5)

where r is the reaction rate, t is the reaction time, and m is a constant which will vary with the system geometry. With proper linearization processes, Eq.(5)can be written as:

ln[−ln(1 − α)] = ln(r) + m ln(t) (6) According to the m values in Eq. (6), the solid-state reac-tions can be divided into three groups: a diffusion mechanism for m = 0.54–0.62, a first-order or phase boundary mechanism for m = 1.0–1.24, and a nucleation or growth mechanism for m = 2.0–3.0[15]. By substituting the conversion ratio data from 0 min to 60 min at all three temperatures inFig. 3into Eq.(6), one can make a plot of the regression lines of Eq.(6), as shown inFig. 5. The m values at 700◦C, 750◦C, and 800◦C were esti-mated to be 0.578, 0.594, and 0.577, respectively. They were

Fig. 5. Plot of ln((−ln(1 − α)) versus ln(t) for the Sr2SiO4reaction process.

Fig. 6. Plot of 1− 2α/3 − (1 − α)2/3_{versus reaction time for the formation of} Sr2SiO4.

very close to each other, suggesting the reactions occurred in the studied temperature range were guided by a single reac-tion mechanism. Comparing the m values with the reacreac-tion mechanisms collected by Hancock et al., it is reasonable to conclude that the formation mechanism of Sr2SiO4is a

three-dimensional diffusion controlled process. Therefore, the relation between the conversion factor and reaction time for this mech-anism can be expressed by the Brounshtein-Ginstling model

[16]:

1− 2α/3 − (1 − α)2/3= kt (7) where k is the reaction rate constant. By substitutingα into Eq.(7)and plotting conversion ratio against reaction time gave the reaction rate at specific temperatures from its slope. From

Fig. 6, it can be observed that the values of reaction rate con-stant k at 700◦C, 750◦C, and 800◦C were 0.00074 min−1, 0.00186 min−1, and 0.00368 min−1, respectively. The activation energy of the reaction was estimated by the Arrhenius equation: k = k0exp −E RT , (8)

where E is the activation energy, R is the gas constant, and T is the absolute temperature.Fig. 7illustrates the plot of ln(k) versus 1/T for Ginstling-Brounshtein model. From the slope of the line inFig. 7, the activation energy for Sr2SiO4formation

was calculated to be 139.6 kJ/mol. These results were similar to the other ceramic reactions[17,18].

3.3. Microstructural observation and reaction model

The surface microstructures of the specimens are shown in

Fig. 8. After ball milling for 24 h, the mixtures appeared to be uniform with grain size around 150 nm as shown inFig. 8(a). When the specimen heated at 800◦C were quenched instantly (Fig. 8b), increase in the grain size of starting materials was observed.Fig. 8(c) shows the sample quenched at 830◦C,

(5)

reveal-Fig. 7. Plot of ln(k) versus 1/T for the formation of Sr2SiO4.

ing drastically varied microstructures having many small islands (∼20 nm) dispersed on the surface of the grains. This suggested that the reactive strontium carbonate decomposed and reacted instantly with the SiO2solid particles. When the specimens were

quenched at 870◦C (Fig. 8d), the number of islands decreased and the grain growth occurred in the sample, providing evidence of the diffusion process. In addition, from XRD analysis, the crystal structure of these grains was found to primarily consist of Sr2SiO4.

In view of the three-dimensional diffusion controlled mech-anism with TG/DTA, XRD and SEM results, a microscopic reaction model is proposed and illustrated inFig. 9. Before the reaction involving SrCO3 and SiO2, no intermediate product

was detected in the samples. As the calcination temperature was increased, SrCO3reacted with SiO2to form the shell of Sr2SiO4

on the surface, accompanied by release of carbon dioxide during the reaction. With increasing calcination time, reactive SrCO3

diffused into the core of SiO2to thicken the Sr2SiO4shell. The

diffusion process was completed with the exhaust of raw mate-rials. Eventually, the pure phase of Sr2SiO4was formed in the

samples.

(6)

Fig. 9. A microscopic reaction model of Sr2SiO4.

4. Conclusions

The reaction mechanism and kinetic analysis for the for-mation of Sr2SiO4are investigated in this study. The process

of formation of Sr2SiO4 is assumed to be a direct reaction

between SrCO3 and SiO2 via the TG/DTA and XRD

anal-ysis. The conversion of Sr2SiO4 from the starting materials

increases with an increase in the heating temperature and heat-ing time. For the ceramic reaction involvheat-ing SrCO3and SiO2,

the three-dimensional solid-state reaction model is considered. The formation of Sr2SiO4 is confirmed to be governed by a

diffusion controlled mechanism via reaction kinetic isothermal analysis. According to the Brounshtein-Ginstling model, the activation energy for the formation of Sr2SiO4is calculated to

be 139.6 kJ/mol. A microscopic reaction model was also pro-posed to describe the formation of Sr2SiO4, which is a promising

phosphor host for use in white light emitting diodes.

References

[1] S. Nakamura, T. Mukai, M. Senoh, Appl. Phys. Lett. 64 (1994) 1687–1689. [2] Z. Yang, X. Li, Y. Yang, X. Li, J. Lumin. 122/123 (2007) 707–709. [3] T. Tamura, T. Setomoto, T. Taguchi, J. Lumin. 87/89 (2000) 1180–1182. [4] D. Jia, Y. Wang, X. Guo, K. Li, Y.K. Zou, W. Jia, J. Electrochem. Soc. 154

(2007) J1–J4.

[5] J.S. Kim, P.E. Jeon, J.C. Choi, H.L. Park, Appl. Phys. Lett. 84 (2004) 2931–2933.

[6] H. Zhang, T. Horikawa, H. Hanzawa, A. Hamaguchi, K.I. Machida, J. Electrochem. Soc. 154 (2007) J59–J61.

[7] G. Blasse, P.E. Wanmaker, J. Vrugt, Philips Res. Rep. 23 (1968) 189–200. [8] J.S. Yoo, S.H. Kim, W.T. Yoo, G.Y. Hong, K.P. Kim, J. Rowland, P.H.

Holloway, J. Electrochem. Soc. 152 (2005) G382–G385. [9] A. Nag, T.R.N. Kutty, J. Mater. Chem. 14 (2004) 1598–1604.

[10] J.K. Park, M.A. Lim, C.H. Kim, H.D. Park, J.T. Park, S.Y. Choi, Appl. Phys. Lett. 82 (2003) 683–685.

[11] J.S. Kim, J.Y. Kang, P.E. Jeon, J.C. Choi, H.L. Park, T.W. Kim, Jpn. J. Appl. Phys. 43 (2004) 989–992.

[12] M. Catti, G. Gazzoni, G. Ivaldi, G. Zanini, Acta Cryst. B39 (1983) 674–679. [13] J.D. Hancock, J.H. Sharp, J. Am. Ceram. Soc. 55 (1972) 74–77. [14] W.A. Johnson, R.F. Mehl, Trans. AIME 135 (1939) 416. [15] K.W. SEO, J.K. OH, J. Ceram. Soc. Jpn. 108 (2000) 691–696.

[16] A.M. Ginstling, B.L. Brounshtein, J. Appl. Chem. USSR 23 (1950) 1327–1338.

[17] C.H. Lu, C.C. Tsai, J. Mater. Res. 11 (1996) 1219–1227. [18] C.H. Lu, W.T. Hsu, J.T. Lee, J. Mater. Res. 19 (2004) 2956–2963.