Preparation of Apatite-Wollastonite-Phlogopite Glass-Ceramic Composites by Powder Sintering Method

An Apatite-Wollastonite-Phlogopite glass-ceramic composite, was developed by sintering and crystallization of the powdered glass. The non-isothermal and isothermal sintering kinetics were studied for this glass-ceramic. Hot-stage microscopy (HSM) measurements demonstrated that it is possible to sinter and crystallize this glass-ceramic with 80% relative density. The activation energy of sintering was analyzed using previously reported model of sintering and it was obtained Q=193.83 KjmolK. Also it was shown that the microstructure of sample is a function of particle size distribution.


Introduction
Apatite-Wollastonite (A/W) and Phlogopite (ph) are well known glass-ceramics [1,2] which were applied as acceptable biomaterials.Nevertheless, Phlogopite has low biocompatibility and A/W has low toughness [3], this work is an attempt to improve the properties by mixing the two compositions and focus on the sintering behaviour of A/W and Phlogopite glass-ceramic composite, which was not studied systematically before.
A/W glass was taken as a matrix and Phlogopite glass was added as a second phase.During heat treatment crystalline phase which have different properties will precipitate.The purpose is to use Phlogopite crystals as a reinforce agent, because of Phlogopite morphology, crack deflection will occur [4], so the toughness of A/W could be increased.The composite was prepared with sintering of two mixed frits, and the sintering conditions were studied systematically.
Glass powder compacts densify through viscous flow sintering if crystallization occurs before densification is completed [5] .The increase in viscosity can prevent further densification [6].In order to obtain a well crystallized material with minimal porosity, glass particles must sinter before crystallization starts.Therefore, glass particles can deform and coalesce, reducing voids, before crystals start to nucleate and grow [6].

Experimental procedure
Apatite-Wollastonite (A/W) and Phlogopite (Ph) glasses were prepared, from reagentgrade MgCO The A-W and Ph glass compositions were melted at 1450°C and 1550°C, respectively, which were kept in an electric furnace for 2 hr using Pt crucibles.To obtain frits, these melts were poured in to cold water.In order to see the effect of ph glass composition on sintering behaviour of A/W glass, 5%wt of Phlogopite (ph) glass composition added to A/W glass-ceramics which termed C595.

Hot-stage microscopy (HSM)
Hot-stage microscopy is considered a suitable experimental technique to study sintering phenomena in glasses.Not only qualitative observations but also quantitative studies of sintering kinetics can be under taken with this technique.
A side -view hot stage microscope equipment EM [201] equipped with image analysis system and electrical furnace Leica 1750/15 was used.The microscopy projects the image of sample through a quartz window and onto the recording device.The computerized image analysis system automatically records and analyzes the geometry changes of the samples during heating.The image analyzer takes into account the thermal expansion of the alumina substrate while measuring the height of the samples during firing.The HSM software calculates the percentage of decrease in height, width and area of the samples images.
The measurements were conducted in air at different heating rate of (5,10,15, 20)°C /min.The samples, cylinder shapes of approximately 3 mm in height and placed on a 10×15×1 mm alumina [>99.5 wt %AL 2 O 3 ] support, were prepared with uniaxial pressing with 200 MPa pressure.The HSM software calculates the percentage of decrease in height, width and area of the sample images with respect to the initial dimensions of the samples, such that the densification as a function of temperature or time may be calculated.The relative density of cylindrical samples, ρ, (density of the sample divided by the bulk glass density) during sintering can be obtained by ρ = (h r /A r 2 ).ρ o in which h r is the relative height and A r is the relative area of the samples at any temperature or time, and ρ o is the initial relative density of the green sample (relative to the bulk glasses density, which were 2.96 g/cm 3 for A/W glass and 2.55 g/cm 3 for Phlogopite glass) [7].The initial relative green density of the samples was around 0.60.The sintering data obtained by HSM are compared with the density of samples densified in an electrical furnace.Cylindrical samples (1 cm) in height and (0.8cm) of diameter from the different glass powder fractions were prepared by isostatic pressing at 200MPa.The samples were weighted and their axial and radial dimensions measured.The compacts were isothermally treated at temperatures between 800 and 1200 °C for 1h.The density of the initial and sintered samples was measured by Archimedes method employing Hg.The error in the determination of the density is 0.02.The sintering process of the powder fractions for each glass was studied from the density variations and dimensional modifications of the samples as a function of temperature.The obtained results indicate that a good agreement was found between measured and calculated samples density [7].Thermal analyses TG-DTA were performed with an EXSTAR 6300 [Seiko] equipment on glass powder samples [φ<60 μm, 1mm] employing a heating rate of (5,10,15, 20)° C.min −1 .α-Al 2 O 3 calcinated at 1200°C was used as inert reference.The treated samples were studied by X-ray diffraction [XRD] with a Siemens model D5000 diffractometer with kristalloflex 710 equipment.Scanning electron microscopy (SEM) employing a Zeiss-DSM950 microscope has been used for characterizing the microstructures of the glass-ceramics.

Results and discussion
The DTA traces of two glasses show intensive Apatite-Wollastonite crystallization peaks in the temperature interval of 850-950•C for A/W glass and 700°C for Ph glass [8,9] .Figs. 1. and 2. depict the crystalline phase of heat treated glasses at crystallization temperature.In Fig. 3., the densification of A/W and C595 starts at 760°C and 790°C, respectively, then are completed at 850 • C; after 850 • C relative density remains constant (86% relative density for C595 and 95% for A/W composition).Compared to A/W and C595, in the case of Ph some delay in the densification is observed and sintering starts at 1030°C, which can be due to the higher crystallization tendency [10] of this compound.This justifies the use of identical sintering conditions and explains the crystallization trends by the differences in the heating rate with respect to the sintering kinetic.The obtained sintering results in Fig. 3 were in Non-Isothermal circumstance, in order to evaluate the isothermal condition on sintering of C595, composite heat treated inside sintering temperature range (820°C -850°C) individually for 1 hr.The sintering results presented in Fig. 5.
It can be seen that the shrinkage of C595 at 850°C is higher than the 820°C, this comment can be related to fast diffusion of elements as well as sintering of composite.Since it has been shown previously that, densification behaviour of glass ceramics systems can be described by liquid phase sintering [9], it was proposed [11] the kinetic equation derived by Kissinger to analysis isothermal sintering data in Fig. 5.
(ΔL/L 0 ) n = K t /T (1) Where ΔL/L 0 is the fractional shrinkage in a time t, n the exponent, K the temperaturedependent rate constant, and T the absolute temperature.The rate constant, K, can be described by the Arrhenius equation [11] K=K 0 exp (-Q/RT) (2) Where K 0 is a pre-exponential term, Q the apparent activation energy of sintering, and R the gas constant.The value of n can be determined by rearranging Eq.( 1) as follows: nln(ΔL/L 0 )=Ln(K/T)+ln(t) (3) the data shown in Fig. 5. can be replotted as Ln (ΔL/L 0 ) versus ln (t) at various sintering temperature from 820°C and 850˚C as shown, using the least-squares fit method, the value of n can be calculated from the slopes (1/n) in Fig. 6., the mean value of 0.03 is used in Eq. 1.for the following analyses.2).forT=820˚C and T=850˚C

Sintering with different heating rate
In order to see the effect of heating rate on sintering behaviour of C595, this composite was heat treated at 5, 10, 15, 20°C/min heating rate.The obtained results are measured by HSM presented in Fig. 7.The densification curves in Fig. 7. have a sigmoid shape.At first, the density increases more and more rapidly with temperature because of the change in viscosity.However, when the specimen approaches theoretical density, the densification rate begins to level off, resulting in a plateau at high temperature.The results in Fig. 7. also show that the densification curves move to the right as the heating rate increase.This is reasonable because densification depends on time and temperature.At the faster heating rate there is less time for densification beside this; ultimate density of samples did not show significance differences which can be associated to slightly differences in green density.
The differential of the shrinkage -temperature plots are calculated, and the results are shown as derivative dimension change [d(ΔL/L 0 /dT] as function of temperature at various linear heating rates in Fig. 8. Having thus the kinetic expression and activation energy of isothermal sintering, it is instructive to extend these results to analyze quantitatively the shrinkage data obtained at constant rates of heating.To do that, we assume Eq. ( 1) used for isothermal heating to be applicable to constant -rate heating as demonstrate in the literature.Differentiation of Eq.( 3) with respect to time gives the linear shrinkage rate as follows : d(ΔL/L 0 )/dt=1/n×K×t -1 (ΔL/L 0 ) (4) In which n=0.03 is used.For the condition here temperature (T) is a linear function of time (T>500°C)we propose that T=at where a is the heating rate.Combining Eqs. ( 1), ( 2) and (3) yields d(ΔL/L 0 )/dt=1/n×T -1 ×a -1/n ×K 0 1/n exp(-Q/nRT) (5) The above equation can be further rearranged and expressed in logarithmic form Ln[Td(ΔL/L 0 )/dt]=ln(1/nk 0 1/n )-1/nLn(a)-Q/nRT (6) the data shown in Fig. 5. can be replotted as Ln (ΔL/L 0 ) versus ln (t) at various sintering temperature from 820° C and 850˚C as shown, using the least-squares fit method , the value of n can be calculated from the slopes (1/n) in Fig. 6., the mean value of 0.03 is used in Eq. 1.for the following analyses using the n value of Fig. 6. and equation ( 6) the Q value obtained: Q=193.83K.j.mol.K -1 .By comparing the activation energy of A/W (Q=530 K.j/mol) [8] and Phlogopite crystallization (Q=164.20 K.j/mol) [10] along with the C595 sintering energy, it appears that Phlogopite crystallization will delay the sintering of A/W-Phlogopite composite.

Effect of particle size
The variation of relative density and sintering time of C595 obtained at 850°C and 10°C/min using different particle size ((60<a<100 µm) and (a<60 micron)) of Ph glass are shown in Fig. 9.It is well known that ph glass shows, surface crystallization, which usually prevents the complete densification [10], in this case by decreasing particle size, crystallization will be promoted.It sounds that in the case of C595 with coarse particle size of ph glass, the impediment of crystallization, will result in successful sintering, comparing to sample with fine particles of ph glass.Some porosity among crystals were remarked which can be demonstrated to inter particle porosity [13].Consequently surface crystallization of Phlogopite [8], requires high surface area and will be enhanced by decreasing the particle size, on the other hand sintering development needs fine particles.In this case it sounds that in competition involving sintering and crystallization, the latter will succeed and consequently sintering will be delayed, as well in agreement with previous work [6].Phlogopite glass composition accelerate A/W crystallization whichever can be an other explanation for the obtained consequences.

Conclusion
The investigated apatite-wollastonite-Phlogopite glass composition is characterized by activation energies of sintering (Q=193.83K.j.mol.K -1 ).That increase as a function of the crystallization ability.Due to the surface crystallization of phlogopite glass, the densification and crystallization will have competition leading to low relative density product, also the formation of the porosity increase with increasing of the crystallization.
The presence of Phlogopite glass composition in A/W composition increase the crystallization and inhibits the sintering.In C595 composites the densification stops after about 80% relative density.Using of controlled particle size of frits may be improve sintering.Investigation showed that although gradual addition of Phlogopite to the A/W glass increases the crystallinity, however causes an unfired composite.Simple HSM experiments can be used to analyze the behavior of samples at different conditions of heating rate, sintering time, particle size, etc., thereby considerably reducing the experimental work.

Fig. 3 .
Fig. 3. Variation of relative density vs. temperature of A/W, Ph and C595 at 10°C/min

Fig. 8 .
Fig. 8. Derivative dimension change with respect to temperature [d(ΔL/L 0 /dT] versus temperature obtained at constant heating rate from 400 to 1000˚C

Fig. 9 .Fig. 10 .
Fig. 9. Relative density of C595 vs. soaking time at sintering temperature 850°C with coarse ■and fine▲particle size of ph glass 3, K 2 CO 3 , MgF 2 , H 2 [PO 4 ] 3, Al 2 O 3 and SiO 2 sand, the batches were calcinated at 900ºC for 24 hr.Tab.I and II summarizes the nominal and analytical compositions of glasses.The nominal / analytical compositions of A/W glass (parts by weight).