Paper
Journal of the Ceramic Society of Japan 118 [6] 462-466 2010
Preparation of submicrometer-sized porous spherical hydroxyapatite agglomerates by ultrasonic spray pyrolysis technique Kiyoshi ITATANI,³ Tomoki TSUGAWA, Tomohiro UMEDA, Yoshiro MUSHA,* Ian J. DAVIES** and Seiichiro KODA Department of Materials and Life Sciences, Faculty of Science and Engineering, Sophia University, 7–1 Kioi-cho, Chiyoda-ku, Tokyo 102–8554 *2nd Department of Orthopaedic Surgery, School of Medicine, Toho University, 2–17–6 Oohashi, Meguro-ku, Tokyo 153–8515 **Department of Mechanical Engineering, Curtin University of Technology, GPO Box U1987, Perth, WA 6845, Australia
Microstructural changes during the heating of submicrometer-sized spherical hydroxyapatite (Ca10(PO4)6(OH)2: HAp) agglomerates were examined. The starting powder was prepared by the spray pyrolysis of calcium phosphate (Ca/P ratio = 1.67) solution containing 0.01671.5 mol·dm¹3 Ca(NO3)2, 0.01000.9 mol·dm¹3 (NH4)2HPO4 and concentrated HNO3 at 600°C using an ultrasonic vibrator. The resulting powders were composed of spherical agglomerates; the agglomerate diameter was reduced with decreasing concentration of spraying solution. The mean diameter of agglomerates obtained by the spray pyrolysis of solution with the lowest concentrations, i.e., 0.0167 mol·dm¹3 Ca(NO3)2 and 0.0100 mol·dm¹3 (NH4)2HPO4 (abbreviated as 0.0167/0.0100), was as small as 0.34 µm. The spherical agglomerates became hollow following a heat treatment at 800°C for 10 min. The incorporation of 0.015 mol·dm¹3 citric acid into 0.0167/0.0100 solution contributed to the formation of hollow spherical agglomerates following spray pyrolysis. Furthermore, the heat treatment of these agglomerates at 800°C for 10 min increased the amount of pores with mean diameter of 0.2 µm, due to the burn-out of residual carbon. ©2010 The Ceramic Society of Japan. All rights reserved.
Key-words : Hydroxyapatite, Ultrasonic spray pyrolysis, Citric acid addition, Carbon burnout technique, Porous spherical agglomerates, Submicrometer scale [Received February 15, 2010; Accepted March 23, 2010]
1.
Introduction
Living bone is composed of inorganic materials (i.e., hydroxyapatite (Ca10(PO4)6(OH)2; HAp) as a main component) and organic materials (e.g., collagen, protein, etc.). Thus HAp is known to be a suitable bone and tooth implant material.1) The dense form of HAp has applications including a space filler material for bioactive fixation whereas the porous form is currently used as a bone substitute for biological fixation.1),2) In order to meet the requirements of these applications, researchers have paid attention to the preparation of HAp particles with various shapes, using many chemical routes. The present authors have investigated the properties of various kinds of calcium phosphate powders, i.e., HAp, ¢-calcium orthophosphate (¢Ca3(PO4)2), calcium diphosphate (Ca2P2O7) and calcium metaphosphate (Ca(PO3)2) prepared using the spray pyrolysis technique.3)6) This technique involves the simultaneous spray pyrolysis of solutions containing the desired types and amounts of metal ions into the “hot zone” of a furnace. The resulting powders have the following characteristics: (i) homogeneous chemical composition and submicrometer-sized particles due to flash pyrolysis and solid-state reaction, (ii) strict control of the chemical composition (provided that one can control the chemical composition of the starting solution) and (iii) formation of hollow spherical agglomerates, reflecting the ³
Corresponding author: K. Itatani; E-mail:
[email protected]
462
outward form of the starting droplets.7) In particular, this technique is noted for the preparation of hollow agglomerates and, as such, is a key method for the preparation of porous agglomerates. In order to further increase the porosity of such agglomerates, the present authors have also paid attention to the carbon burnout technique, i.e., pore formation due to the burnout of carbon dispersed within the agglomerates. The nano-sized carbon particles, homogeneously dispersed in the agglomerates through the spray pyrolysis of calcium phosphate solution containing an organic compound, may be eliminated by burnout to form porous agglomerates. Amongst suitable water- and acid-soluble organic compounds, we employed citric acid (C6H8O7) on the basis of (i) comparatively high molecular weight (=192), i.e., large carbon content (37.5%), and (ii) comparatively high solubility in water (750 g·dm¹3 at 20.0°C).8) On the basis of this background information, the present paper describes research results concerning the morphological and microstructural changes of agglomerates that were varied due to several factors, such as the concentration of spraying solution, heat treatment conditions, and the effect of organic compound (i.e., citric acid) addition, with the prospect that the spherical agglomerates may be utilized as bioresorbable materials. The incorporation of the organic compound (citric acid) in the spraying solution also encourages the enlargement of pore sizes following the elimination of free carbon in the spray-pyrolyzed powder after the heat treatment in air.9) ©2010 The Ceramic Society of Japan
JCS-Japan
Journal of the Ceramic Society of Japan 118 [6] 462-466 2010
Sample No.
Ca(NO3)2
(NH4)2HPO4
HNO3
mol·dm¹3
mol·dm¹3
cm3
1 2 3 4
1.5 0.5 0.05 0.0167
0.9 0.3 0.03 0.0100
12 12 12 12
2.1
Abbreviation 1.5/0.9 0.5/0.3 0.05/0.03 0.0167/0.0100
Evaluation
Phase identification of the powders was conducted using an Xray diffractometer (Model RINT2100V/P, Rigaku, Tokyo; 40 kV, 40 mA; XRD) with monochromatic Cu K¡ radiation. The phase changes during heating in the range of room temperature to 1200°C were examined using thermogravimetry (TG; Model Thermo Plus TG8120, Rigaku, Tokyo); the amount of powder used for each measurement was approximately 25 mg. The agglomerate morphologies were observed using a fieldemission scanning electron microscope (FE-SEM: Model S4500, Hitachi, Tokyo; accelerating voltage, 5 kV). The agglomerate diameters were determined from at least 200 individual agglomerates. Microstructures of the agglomerates were investigated using a transmission electron microscope (TEM: Model JEM-2011, JEOL, Tokyo; accelerating voltage, 200 kV), together with electron diffraction analysis. The specific surface area of the powder was measured by a BrunauerEmettTeller (BET) technique, using nitrogen (N2) as the adsorption gas. In addition, pore sizes for the powder were measured using a mercury porosimeter (Model AutoPore 9420, Micromeritics Instrument, Norcross, GA, USA).
3.1
50 nm
2. Experimental Powder preparation
The starting calcium phosphate solutions (1 dm3), whose chemical composition corresponded to that of HAp (Ca/P ratio = 1.67), were prepared using Ca(NO3)2, (NH4)2HPO4 and concentrated HNO3. The concentrations of these compounds in order to prepare the solutions are listed in Table 1, together with their abbreviations. In addition, the calcium phosphate solutions containing 0.010.03 mol·dm¹3 citric acid were prepared in order to homogeneously disperse the carbon in the spray-pyrolyzed powder. The calcium phosphate solution was sprayed into the “hot zone” (60 mm (inner diameter) © 800 mm (length) mullite tube) in an electric furnace heated at 600°C using an ultrasonic vibrator (frequency: 2.4 MHz); the spray pyrolysis temperature was monitored using a chromelalumel thermocouple. The resulting spray-pyrolyzed powder was collected using a testtube type filter, whereas the other main products, i.e., water vapor containing various salts, were condensed using a Liebig condenser. The spray-pyrolyzed powder was further heat-treated in air at the desired temperature for 10 min; the heating rate from room temperature up to the desired temperature was fixed at 10°C·min¹1.
2.2
X-ray intensity / a. u.
Table 1. Preparation conditions of the starting calcium phosphate solutions
3. Results and discussion Properties of the spray-pyrolyzed and heattreated powders
As mentioned in the previous section, calcium phosphate powders were prepared by the spray pyrolysis of solutions containing Ca(NO3)2, (NH4)2HPO4 and concentrated HNO3 at
10
20
30
40
50
2 θ / ° (Cu Kα) Typical XRD pattern of the powder obtained by spraypyrolyzing 0.0167/0.0100 solution at 600°C, together with a typical TEM micrograph. : HAp. Fig. 1.
600°C. A typical XRD pattern of the spray-pyrolyzed powder has been shown in Fig. 1, together with a typical TEM micrograph. The XRD pattern indicated the presence of poorly crystalline HAp.10) Such poorly crystalline HAp powder may often be formed when the droplets have passed through the hot zone of the electric furnace in a short time (generally within one minute, assuming that the droplets passed smoothly through the hot zone, i.e., laminar flow). TEM observations showed that the solid spherical agglomerate was composed of closely-packed particles with approximate sizes of 30 nm. The original spray-pyrolyzed powder mentioned above was noted to have some issues related to: (i) poorly crystalline HAp phase and the presence of amorphous materials, (ii) the existence of residual nitrates and water, and (iii) relatively small amount of pores present on the surfaces and insides of the agglomerates. In this section the spray-pyrolyzed powder was heat-treated in an attempt to overcome these problems. The change in mass during heating of spray-pyrolyzed HAp powder examined using the TG method is shown in Fig. 2, together with typical FE-SEM micrographs. The TG curve indicated step-wise mass losses in the ranges of room temperature to 100°C, 100 to 400°C and 400 to 800°C. These step-wise mass losses may be attributed to the elimination of residual water and nitrates.3) On the other hand, FE-SEM micrographs of HAp powder heattreated at 800°C for 10 min showed the presence of numerous pores (typical diameter ¯1 ¯m) on the agglomerate surfaces. In contrast to this, FE-SEM micrographs of the HAp powder heat-treated at 1000°C for 10 min indicated that the smaller agglomerates (typical diameter ¯1 ¯m) had bonded together, and that no obvious evidence of surface pores was noted. The marked difference in amount of pores between agglomerates heat-treated at 800°C and 1000°C is explained in terms of (i) the coalescence of pores due to particle rearrangement (heating up to 800°C) and (ii) the elimination of pores due to the sintering of particles (further heating up to 1000°C).6) On the basis of the TG and FE-SEM results, the optimum heat treatment temperature for the preparation of porous HAp agglomerates was selected to be 800°C, not only due to the mass loss being almost completed by 800°C, but also due to the reduction in amount of surface pores upon further heating to 1000°C. The specific surface area of the HAp powder heat463
JCS-Japan
Itatani et al.: Preparation of submicrometer-sized porous spherical hydroxyapatite agglomerates by ultrasonic spray pyrolysis technique
50
0
(a)
5
1 µm
1 µm
10
Frequency / %
Mass loss / %
40 (b) 30 (c) 20
(d)
10 15
0
800
400
1000
1200
Temperature / °C
0
TG curve of the HAp powder obtained by the spray pyrolysis of 0.0167/0.0100 solution at the heating rate of 10°C·min¹1, together with typical FE-SEM micrographs. Note that the powders shown in FE-SEM micrographs were heat-treated at each temperature for 10 min.
0
Fig. 2.
1
10
Agglomerate diameter / µm HAp powder obtained by spray pyrolysis and subsequently heat treatment at 800°C for 10 min. Concentration of the starting solutions: (a) 0.0167/0.0100, (b) 0.05/0.03, (c) 0.5/0.3, (d) 1.5/0.9. Fig. 4.
Ca(NO3)2 / (NH4)2HPO4 0.0167 / 0.0100
0.05 / 0.03
0.5 / 0.3
1.5 / 0.9
2 µm
2 µm
2 µm
2 µm
0.5 µm
0.5 µm
0.5 µm
0.5 µm
Fig. 3. FE-SEM (above) and TEM (below) micrographs of the HAp powders heat-treated at 800°C for 10 min.
treated at 800°C for 10 min was as high as 15.6 m2·g¹1, although this was lower than that of the spray-pyrolyzed powder (approximately 23 m2·g¹1). Microstructural changes taking place in the spherical HAp agglomerates heat-treated at 800°C for 10 min are shown in Fig. 3, as a function of the Ca(NO3)2 and (NH4)2HPO4 concentrations in the spraying solution. Spray pyrolysis and subsequent heat treatment of the 0.5/0.3 and 1.5/0.9 solutions resulted in the formation of dense spherical agglomerates containing closelypacked particles. In contrast to this, the spray pyrolysis of 0.05/ 0.03 (or lower concentration) solution resulted in the formation of hollow spherical agglomerates. The phenomenon regarding the formation of such pores on the agglomerate surfaces is explained in terms of (i) the formation of (small) pores due to the release of a larger amount of water vapor in the dilute solutions (i.e., 0.05/0.03 or lower concentration) and (ii) the coalescence of small pores into larger pores. The noted formation and simultaneous coalescence of pores is also supported by the fact that the specific surface area (as will be 464
shown later) increased with decreasing concentrations of Ca(NO3)2 and (NH4)2HPO4. Agglomerate diameters for these spray-pyrolyzed and heattreated HAp powders have been examined quantitatively and presented in Fig. 4. The agglomerate diameter distribution curve for the spray-pyrolyzed powder was shifted towards the lower diameter-side with decreasing concentrations of Ca(NO3)2 and (NH4)2HPO4. The mean diameter of the spherical agglomerates derived from the 0.0167/0.010 solution was the smallest (0.34 ¯m). The presence of submicrometer-sized diameters is not only due to the smaller droplet size, but also due to the elimination of a larger amount of solvent (water) during spray pyrolysis. The specific surface areas of these spray-pyrolyzed and heat-treated powders were measured and plotted against the concentrations of Ca(NO3)2 and (NH4)2HPO4 in the spraying solution with the results being shown in Fig. 5. The specific surface areas of the powders derived from the spray pyrolysis of 1.5/0.9 and 1.0/0.6 solutions were approximately 5 m2·g¹1. The specific surface area increased with decreasing concentrations of Ca(NO3)2 and (NH4)2HPO4 and reached 15.6 m2·g¹1 for the case of 0.0167/ 0.010 solution. Since hollow spherical agglomerates with a mean diameter of 0.34 ¯m could be prepared from the spray pyrolysis of 0.0167/0.0100 solution and subsequent heat-treatment, the pore diameters for this powder were further investigated by mercury porosimetry (as will be shown later). Most of the pore diameters were noted to be in the range of 0.07 to 0.35 ¯m with a frequency peak at 0.18 ¯m. On the basis of FE-SEM and TEM micrographs, therefore, the diameters of pores in the agglomerates were considered to be typically 0.2 ¯m. The following phenomena due to the heat treatment of spraypyrolyzed agglomerates are thus noted: (i) the spray pyrolysis of 0.5/0.3 (or higher concentration) solution; no significant amount of pores could be found on the agglomerate surface even following heat treatment, and (ii) the spray pyrolysis of 0.05/ 0.03 (or lower concentration) solution; the agglomerates were hollow following heat treatment, chiefly due to the coalescence of pores formed by the spray pyrolysis of dilute solution.
JCS-Japan
Journal of the Ceramic Society of Japan 118 [6] 462-466 2010
Concentration of (NH4)2HPO4 / mol•dm-3 0
0.3
0.6
(a)
(b)
0.9
1 µm
15
10
X-ray intensity / a. u.
Specific surface area / m2 • g -1
20
5
1 µm
(c)
0 0
0.5
1.0
1.5
Concentration of Ca(NO3) 2 / mol•dm-3 Relationship between concentrations of Ca(NO3)2/(NH4)2HPO4 and specific surface area. Note that the spray-pyrolyzed powders were heat-treated at 800°C for 10 min. Fig. 5.
3.2
Effect of citric acid addition on properties of HAp agglomerates
As described above, hollow spherical HAp agglomerates could be prepared by the spray pyrolysis of dilute calcium phosphate solution. In order to increase the porosity in the agglomerates, the burn-out of residual carbon formed by the spray pyrolysis of solution containing citric acid was investigated. Firstly, the properties of agglomerates prepared using 0.0167/ 0.0100 solution with 0.15 mol·dm¹3 citric acid addition were examined. The color of the spray-pyrolyzed powder was brownish, suggesting the presence of residual carbon due to the pyrolysis of citric acid. A typical FE-SEM micrograph, TEM micrograph and XRD pattern for the spray-pyrolyzed agglomerates are presented in Fig. 6. The resulting agglomerates were found to be spherical and submicrometer-sized (Fig. 6(a)). The agglomerates appeared to be hollow, whereas the electron diffraction pattern indicated broad rings (Fig. 6(b)). Moreover, no distinct X-ray reflections were found in the XRD pattern (Fig. 6(c)). It would appear that the spherical agglomerates in the spraypyrolyzed powder are already hollow without heat treatment, but that the crystallization of amorphous phase to HAp may be retarded. As for the previous case, the hollow agglomerates are believed to be formed due to the coalescence of pores through the rearrangement of particles. This process seems to be promoted, compared to the case without citric acid addition, by the pyrolysis of citric acid and subsequent partial burn-out of residual carbon. One possible scenario may be that the highly exothermic residual carbon burn-out provides sufficient energy to accelerate this process. Moreover, the residual carbon formed by the pyrolysis of citric acid seems to be restricted by the solidstate reaction and subsequent crystallization to HAp. In order to examine the temperature required for burn-out of the residual carbon in the spray-pyrolyzed powder, DTATG of the resulting powder was conducted (data not shown here). DTA
10
20
30
40
50
2 θ / ° (Cu Kα) Typical (a) FE-SEM micrograph, (b) TEM micrograph and (c) XRD pattern of the powder obtained by the spray pyrolysis of 0.0166/ 0.0100 solution with 0.015 mol·dm¹3 citric acid at 600°C. Fig. 6.
curve indicated that an exothermic event started to occur at approximately 650°C. This exothermic event is attributed to the combustion of free carbon originating from the citric acid.9) On the other hand, the TG curve indicated that a step-wise mass loss occurred in the range of room temperature up to 800°C. On the basis of the DTATG results, the spray-pyrolyzed powders were heat-treated at 800°C for 10 min. The XRD patterns of these heat-treated powders indicated the presence of well-crystalline HAp (data not shown here). The specific surface areas of the heat-treated powders have been shown in Fig. 7, together with typical FE-SEM and TEM micrographs. The specific surface area of the heat-treated powder exhibited a maximum (21 m2·g¹1) at 0.015 mol·dm¹3 citric acid addition (Fig. 7(a)). This powder was comprised of submicrometer-sized hollow spherical agglomerates with many pores on the surface and inside of the agglomerates (Fig. 7(b) and (c)). The increase in specific surface area with increasing amount of citric acid addition from 0.01 to 0.015 mol·dm¹3 is interpreted as the formation of pores resulting from the burn-out of residual carbon in air. On the other hand, the decrease in specific surface area, upon further increase in the amount of citric acid addition from 0.015 to 0.30 mol·dm¹3, is explained in terms of (i) a significant exothermic effect due to the rapid burn-out of residual carbon and (ii) simultaneous particle growth due to sintering. In addition, the effect of citric acid addition on agglomerate diameter was determined from FE-SEM micrographs. The mean agglomerate size without citric acid addition was 0.34 ¯m, whereas that with 0.015 mole·dm¹3 citric acid addition increased to 0.48 ¯m. The increase in agglomerate size due to the addition of citric acid to the spraying solution indicates that an evolved gas, e.g., H2O and CO2, may have contributed to swelling the agglomerates during the reaction, as well as the increase in amount of pores. Porosities of the heat-treated powder were further examined by mercury porosimetry and have been shown in Fig. 8. Pore 465
JCS-Japan
Itatani et al.: Preparation of submicrometer-sized porous spherical hydroxyapatite agglomerates by ultrasonic spray pyrolysis technique
volume for the agglomerates may be clearly enhanced by the addition of citric acid, which corresponds well with the results of FE-SEM and TEM observations. It should be noted that the number of pores with diameters of approximately 0.2 ¯m was significantly enhanced by the addition of citric acid in the spraying solution.
(c)
(b)
1 µm
1 µm
Specific surface area / m 2 •g -1
25 (a)
20 15 10 5 0 0
0.010
0.030
0.020
Amount of citric acid / mol•dm-3 Fig. 7. Effect of citric acid addition on the (a) specific surface area of powder, together with (b) FE-SEM and (c) TEM micrographs. Note that the powders obtained by the spray pyrolysis of 0.0166/0.0100 solution containing with citric acid were further heat-treated at 800°C for 10 min.
5 Intra-agglomerate
Inter-agglomerate
Incremental intrusion / cm3 •g -1
4 (b) 3
4.
Microstructural changes during the heating of submicrometersized spherical HAp agglomerates were examined. The starting powder was prepared by the spray pyrolysis of calcium phosphate (Ca/P ratio = 1.67) solution containing 0.0167 1.5 mol·dm¹3 Ca(NO3)2, 0.01000.9 mol·dm¹3 (NH4)2HPO4 and concentrated HNO3 at 600°C, using an ultrasonic vibrator. The results obtained were summarized as follows: (i) The spray-pyrolyzed powders were composed of spherical agglomerates. The agglomerate diameter was reduced with decreasing concentration of spraying solution; the mean diameter of agglomerates obtained by the spray pyrolysis of solution with the lowest concentrations (0.0167 mol·dm¹3 Ca(NO3)2 and 0.0100 mol·dm¹3 (NH4)2HPO4, i.e., 0.0167/0.0100) was as low as 0.34 ¯m. In addition, the spherical agglomerates became hollow following additional heat treatment. (ii) In order to increase the porosity of spherical agglomerates, the 0.0167/0.0100 solutions containing 0.01 to 0.03 mol·dm¹3 of citric acid were spray-pyrolyzed at 600°C. The spray-pyrolyzed agglomerates were hollow and this was attributed to the burn-out of residual carbon. When the powder obtained by the spray pyrolysis of 0.0167/0.0100 solution containing 0.015 mol·dm¹3 citric acid was heattreated at 800°C for 10 min, a significant amount of pores was formed on the surfaces of the agglomerates. The mean particle and pore diameters of this powder were 0.48 and 0.2 ¯m, respectively. Acknowledgments The authors wish to express their thanks to Professor Y. Yokogawa of Osaka City University for measuring the pore diameters of the powders.
2 (a)
References 1) 2)
1
3)
0 0.01
0.1
1
10
100
Pore diameter / µm Fig. 8. Pore diameter distributions of the powders heat-treated at 800°C for 10 min, followed by the spray pyrolysis of 0.0167/0.0100 solution (a) without and (b) with citric acid addition.
4) 5) 6) 7)
diameters for the powders were centered in the ranges of 0.05 to 0.5 ¯m and 0.5 to 200 ¯m. The intrusion peaks of the powders without and with citric acid addition appeared at approximately 0.2 ¯m. On the basis of FE-SEM and TEM micrographs, the pore populations with diameters of 0.5 to 200 ¯m are considered to be inter-agglomerate pores, whereas those with diameters of 0.05 to 0.5 ¯m correspond to intra-agglomerate pores, i.e., the pores present on the surfaces and insides of agglomerates. The pore
466
Conclusions
8)
9)
10)
L. L. Hench, J. Am. Ceram. Soc., 81, 17051728 (1998). D. S. Metger, M. R. Rieger and D. W. Foreman, J. Mater. Sci. Mater. Med., 10, 917 (1999). K. Itatani, O. Takahashi, A. Kishioka and M. Kinoshita, Gypsum & Lime, 213, 1927 (1988) [in Japanese]. M. Aizawa, K. Itatani, Y. Miyamoto, A. Kishioka and M. Kinoshita, Gypsum & Lime, 237, 2230 (1992) [in Japanese]. K. Itatani, T. Nishioka, S. Seike, F. S. Howell, A. Kishioka and M. Kinoshita, J. Am. Ceram. Soc., 77, 801805 (1994). K. Itatani, M. Abe, T. Umeda, I. J. Davies and S. Koda, China Particuol., 2, 200206 (2004). K. Itatani and M. Aizawa, J. Soc. Inorg. Mater. Japan, 10, 285292 (2003) [in Japaense]. M. L. Richardson and S. Gangolli, Eds. “The Dictionary of Substances and Their Effects,” The Royal Society of Chemistry, Cambridge (1993) pp. 527528. K. Itatani, H. Itokazu, T. Umeda, I. J. Davies, Y. Musha, K. Mizutani and S. Koda, Key Eng. Mater., 309–311, 129132 (2005). Joint Committee on Powder Diffraction Standards, The International Centre for Diffraction Data No. 09-0432, Newtown Square, PA, USA (1959).
