Powder Metallurgy and Metal Ceramics, Vol. 49, Nos. 9-10, 2011
NANOSTRUCTURED MATERIALS PRESSING AND SINTERING OF BINARY MIXTURES BASED ON NANOSIZED HYDROXYAPATITE V. V. Skorokhod,1 S. M. Solonin,1 V. A. Dubok,1 L. L. Kolomiets,1,2 T. V. Permyakova,1 and A. V. Shinkaruk1 UDC 621.762 The paper examines the pressing, sintering, and structure of binary (hydroxyapatite–pyroceram and hydroxyapatite–tricalcium phosphate) powder mixtures. It is shown that the compactibility of these powder mixtures is an additive function and depends on pure components. The sinterability of the HA–TCP mixture obeys the same law. The HA–pyroceram powder mixture is sintered in the presence of a liquid phase whose amount depends on the particle size of HA, as established experimentally. The use of nanosized powder intensifies diffusion even in solid state due to the great interface area and leads to additional crystallization of the amorphous phase. Keywords: compactibility and sinterability, hydroxyapatite–tricalcium phosphate, hydroxyapatite– pyroceram, amorphous phase Hydroxyapatite (HA, Ca10(PO4)6(OH)2) and β-tricalcium phosphate (β-TCP, Ca3(PO4)2) are the best known calcium–phosphate ceramics. Hydroxyapatite is stable and tricalcium phosphate rapidly dissolves in the human body. Many researchers note that the dissolution rate of well-crystallized HA after implantation in the human body is inadequate to reach the optimum result. On the other hand, the dissolution rate of β-TCP is too high to form bones. To achieve the optimum dissolution rate, binary calcium–phosphate HA–TCP ceramics has been proposed on the grounds that its biodegradation can be controlled by a simple variation in the TCP amount. Bioglass crystalline materials (biopyrocerams) and bioglasses are close to ceramics. Bioglass crystalline materials are actually ceramics with a higher content of the glass phase and bioglasses commonly contain or produce a crystalline phase after thermal treatment. Bioactive pyrocerams contain, in mol.%: 5–17 P2O5, 20–50 CaO, 20–55 SiO2, and 10–50 Na2O. The introduction of B2O3 instead of some part of SiO2 (to 15 wt.%) and the introduction of up to 12% CaF2 instead of CaO or increase in the amount of the crystalline phase through variation in crystallization conditions hardly affect the bioactivity of the material. In addition, biopyrocerams are of permanent interest due to their remarkable mechanical properties. It is generally clear that additions of HA, as a more refractory phase, may be used to control the melting temperature of less refractory pyrocerams.
1Frantsevich
Institute for Problems of Materials Science, National Academy of Sciences of Ukraine, Kiev,
Ukraine. 2To
whom correspondence should be addressed; e-mail:
[email protected].
Translated from Poroshkovaya Metallurgiya, Translated from Poroshkovaya Metallurgiya, Vol. 49, No. 9– 10 (475), pp. 119–125, 2010. Original article submitted May 21, 2009.
588
1068-1302/11/0910-0588 ©2011 Springer Science+Business Media, Inc.
90 80 70
Porosity,%
60 50 40 1 2 3 4 5
30 20 10 0
200
400 600 800 1000 Pressure, МPа
Fig. 1. Porosity as a function of compaction pressure of HA (1), HA–50 TCP (2), TCP (3), HA–50 pyroceram (4), and pyroceram (5) powders
TABLE 1. Bulk Porosity of Powders and Porosity of Compacts after Pressing at 500 MPa Powder HA HA + TCP TCP HA + pyroceram Pyroceram
Bulk porosity, %
Porosity of compacts, %
83.0 75.8 70.4 69.7 44.2
38.5 35.0 33.0 28.0 20.0
In the above regard, it is a highly relevant task for implantology to produce and analyze binary HA–TCP and HA–pyroceram composites. This paper focuses on the pressing, sintering, and structure of the binary powder mixtures since powder metallurgy methods are most promising for producing structural bioceramics. We used nanosized HA powder with a specific surface area of 70 m2/g, β-TCP powder with a specific surface area of 4.5 m2/g, and biopyroceram powder with a specific surface area of 0.12 m2/g. The powders were mechanically mixed to obtain, in wt.%: 50 HA–50 TCP and 50 HA–50 pyroceram. The compactibility of the powder was examined using a mold 10 mm in diameter in a continuous mode, i.e., without taking the samples out of the mold. Figure 1 shows compaction curves for pure powders and their binary mixtures. The maximum compactibility (porosity 18.5%) is shown by pyroceram, the minimum (33.5%) by HA, and the intermediate (26%) by TCP. As would be expected, the binary mixtures show compactibility that is intermediate between pure powders: that of HA + TCP is between HA and TCP and that of HA + pyroceram is between HA and pyroceram. To understand why the compactibility of these powders differs, their bulk porosities are needed. The bulk porosities and porosities of compacts pressed at 500 MPa are shown in Table 1. The HA powder is nanosized and, thus, has the greatest bulk porosity. The TCP powder is the coarsest and, thus, has the lowest bulk porosity. The bulk porosity of their mixture is intermediate. The pyroceram powder consists of coarse crystals (as will be shown below) and has the minimum bulk porosity. Its mixture with HA has the bulk porosity intermediate between HA and pyroceram. The compactibility noticeably correlates with bulk
589
30 1
Volume shrinkage, %
25
3
20 4
15
2
10 5 0 800
900
1000
Т, С О
Fig. 2. Volume shrinkage as a function of sintering temperature of HA (1), TCP (2), HA–50 TCP (3), and HA–50 pyroceram (4) powders porosity: the lower the bulk porosity, the better the compactibility of the powders. This is because the layer of looser powders is higher in a mold, which increases pressure losses in side friction under pressing. Therefore, the results show that the use of binary compositions substantially decreases the porosity of a material containing poorly compactable nanosized HA powder. To examine the sintering process, we pressed samples 15 mm in diameter and 8 mm in height with an initial porosity of 37.5–40%. The samples were sintered in air from HA + 50 wt.% β-TCP and HA + 50 wt.% pyroceram powder mixtures and HA, β-TCP, and pyroceram powders at temperatures from 800 to 1100°C with 1 h holding at each temperature; the same samples were sintered at different temperatures, with weighing and measuring interruptions. Figure 2 shows how volume shrinkage depends on sintering temperature. We will first describe how pure pyroceram behaves during sintering. Volume changes of pyroceram are not shown in the figure for the following reasons. The pyroceram sample did not change its volume at 800°C; it became longer, bubbles appeared at its ends, and it softened at 900°C; the sample melted and lost its shape at 1000°C; and pyroceram completely melted and flowed at 1100°C. This behavior testifies that biopyroceram has no melting point but softens gradually, like amorphous glass. Nevertheless, the x-ray pattern of pyroceram shows diffraction lines (Fig. 3), which are indicative of a crystalline phase. Hence, pyroceram contains two phases: amorphous (low-melting) and crystalline (refractory). The amorphous glass phase shows no diffraction and is evident as a diffuse halo, whose size can determine the amount of the low-melting phase to a certain extent. Biopyroceram has the following initial composition, in %: 41 CaO, 22 SiO2, 16 P2O5, 12 B2O3, 3 MgO, 5 Na2O, and 1 CaF2. Thus, the crystalline phase of the biopyroceram is represented by silicate α′-Ca2SiO4 and all oxides. The pyroceram continues crystallizing during sintering at 900 and 1000°C to form wollastonite CaSiO3 and silicate α-Ca2SiO4. The x-ray diffraction pattern of the samples sintered at 900 and 1000°C differs from that of the initial pyroceram in a substantially smaller (by 38%) halo, which is due to a smaller amount of the amorphous glass phase as sintering involved additional crystallization. Noteworthy is that the diffraction lines disappear after complete melting at 1100°C and the halo increases to its initial size. Unlike pure pyroceram, composite samples (HA–50% pyroceram) are sintered regularly and do not melt at 1100°C. Figure 4 shows x-ray diffraction patterns of the samples sintered at 900–1100°C. These x-ray diffraction patterns are difficult to interpret since the most intensive lines of HA and pyroceram match, and the HA lines are much more intensive than those of pyroceram. Hence, the presence of wollastonite’s marker line in pyroceram at 2θ = 30 deg, which does not match with any HA lines, is practically the only sign of pyroceram in the binary mixture.
590
CaSiO3 (wollastonite) α -Ca2SiO4 α-Ca2SiO4
HA CaSiO3 (wollastonite) 4
4
3
3
2
2
1
1
20
25
30
35
40
45
2θ, deg
Fig. 3. X-ray diffraction patterns of biopyroceram powders: initial (1) and sintered at 900 (2), 1000 (3), and 1100°C (4)
20
25
30
35
45
40
2θ, deg
Fig. 4. X-ray diffraction patterns of binary HA–50 pyroceram mixtures: initial (1) and sintered at 900 (2), 1000 (3), and 1100°C (4)
As is the case with pure pyroceram, this line shows up only after sintering of composites. The other difference between the x-ray diffraction patterns of sintered and initial samples is a 29% smaller halo, which is evidence of a smaller amount of the low-melting phase in pyroceram after sintering. To find out why pyroceram does not melt in the composite with HA but does this without HA, we need to assess the crystallization of pyroceram with indirect indicators: volume changes, melting, variation in micro- and macrostructure, minimum HA content of composites, and different HA particles sizes. Two samples (pyroceram–20% HA) were sintered at 1050°C using HA powders of two types: nanosized with a surface area of 70 m2/g and coarse with 4 m2/g. The sample with nanosized HA was smooth, showed no deformation, and somewhat shrank along the diameter. The sample with coarse HA was rough, lost its shape, and had large bubbles inside resulting from the expansion of the liquid phase (Fig. 5a). This difference in sintering behavior may be because of the only factor: a large amount of the liquid phase in the powder with coarse HA and no liquid phase in the sample with nanosized HA. This is ascertained by x-ray diffraction patterns of the samples(Fig. 5b).
2
1 20
a
25
30
35
40
45
2θ, deg
b
Fig. 5. Photographs (a) and x-ray diffraction patterns (b) of samples made of binary pyroceram– 20 HA mixtures sintered at 1050°C: 1) nanosized HA, 2) coarse HA
591
a
b
d
c
e
Fig. 6. Microphotographs of fractured surface of bioceramic samples sintered at 1100°C (×1000): TCP (a), nanosized HA (b), biopyroceram (c), HA–50 TCP (d), HA–50 pyroceram (e) The amorphous halo of the sample with nanosized HA is twice as small as that of the sample with coarse HA. Therefore, the amount of the amorphous low-melting phase in pyroceram in the sample with fine HA is much smaller than that in the sample with coarse HA. This is because the presence of nanosized HA in a pyroceram compact leads to its intensive diffusion even in solid state due to a great interface area. This results in additional crystallization of the amorphous phase, sharply reduces the amount of the low-melting phase, and increases the amount of the refractory crystalline phase. The introduction of coarse HA powder neither provides a great interface area nor intensifies solid-state diffusion; hence, there is no additional crystallization of the glass phase. As a result, the sample melts at a temperature that nanosized HA undergoes without any sacrifice. The crystallization processes in pyroceram may be controlled not only through sintering conditions but also through the intensification of solidstate diffusion using the interface area factor. Turning back to Fig. 2, we should note that the HA–pyroceram composite shows the greatest volume shrinkage probably because of liquid-phase sintering with a small amount of the residual liquid phase in incompletely crystallized pyroceram. The minimum shrinkage during sintering was shown by β-tricalcium silicate, which is natural, taking into account its particle size. The maximum shrinkage was exhibited by nanosized HA powder. The shrinkage of the HA–TCP composite, as would be expected, was close to intermediate between that of HA and TCP.
592
X-ray diffraction of sintered HA–TCP composites revealed structurally activated transformation of hydroxyapatite in contact with β-tricalcium phosphate. It was established that hydroxyapatite transformed to βtricalcium phosphate according to the dehydration reaction through contact interface activation during sintering of mechanical mixtures of HA and TCP powders. These results are described in our previous paper [1]. Figure 6 shows microphotographs of fractured surface of nanosized HA, TCP, and pyroceram powders and their mixtures sintered at 1100°C. Biopyroceram is represented by coarse elongated crystals about 3 to 10–15 µm in size and TCP by about 5 µm particles. With this magnification, HA represents a continuous dark field with agglomerates about 10 µm in size. This is due to the high absorbing capacity of the nanosized powder. Two phases are seen in the binary HA–50% TCP mixture: coarse TCP (pores to 1 µm) and fine (to the right of the photograph center) HA (pores to 0.3 µm). The crystalline phase of pyroceram and dark HA phase along the edges of the photograph are present in the structure of the binary HA–pyroceram mixture.
CONCLUSIONS The compactibility of the binary HA–pyroceram and HA–TCP mixtures and the sinterability of the binary HA–TCP mixture show additive dependence on the compactibility and sinterability of pure components. The sintering of the binary HA–pyroceram mixture is not subject to this dependence as it occurs in the presence of the liquid phase. The halo in the x-ray diffraction patterns of pyroceram correlates with the amount of the amorphous lowmelting glass phase, which additionally crystallizes during sintering. The sintering temperature of low-melting biopyroceram can be increased by 300°C with addition of 20% nanosized HA. The crystallization of pyroceram in the HA–pyroceram composite can be controlled not only through sintering conditions but also through the intensification of solid-state diffusion using the interface area factor.
REFERENCES 1.
V. V. Skorokhod, S. M. Solonin, V. A. Dubok, et al., “Decomposition activation of hydroxyapatite in contact with β-tricalcium phosphate,” Powder Metall. Met. Ceram., 49, No. 5–6, 324–329 (2010).
593