Materials Letters 106 (2013) 373376

Contents lists available at SciVerse ScienceDirect

Materials Letters
journal homepage: www.elsevier.com/locate/matlet

Effect of the partial substitution of K2O, MgO, B2O3 for CaO on

crystallization, structure and properties of Na2OCaOSiO2P2O5
system glass-ceramics
H.C. Li a,b, D.G. Wang a,b,n, J.H. Hu a,b, C.Z. Chen a,b,n
a
Key Laboratory for LiquidSolid Structural Evolution & Processing of Materials, Ministry of Education, Shandong University, Shandong,
Jinan 250061, People's Republic of China
b
School of Materials Science and Engineering, Shandong, Jinan 250061, People's Republic of China

art ic l e i nf o a b s t r a c t

Article history: Na2OCaOSiO2P2O5 (45S5) system ceramics, in which CaO is partly replaced by K2O, MgO or B2O3, have
Received 23 January 2013 been obtained by the solgel method. The crystallization, structure, mechanical properties and in vitro
Accepted 2 May 2013 bioactivity of glass-ceramic were investigated. Results indicated that the doping ions in glass have
Available online 11 May 2013
remarkably improved the volume density, bending strength, fracture toughness and microhardness of
Keywords: the glass-ceramics. The addition of K2O and B2O3 decreased the integrity of glass network, promoted
Ceramic glass crystallization and crystallization did not inhibit glass bioactivity. The addition of MgO made the
Solgel preparation glass network much denser and decreased the ability of apatite induction.
Crystallization & 2013 Elsevier B.V. All rights reserved.
Mechanical properties
Bioactivity

1. Introduction
Na2OCaOSiO2P2O5 (45S5) is the most famous glass because
of its excellent bioactivity and biocompatibility [1,2], but the low
mechanical strength restricts its clinical application [3]. Mechan-
ical properties and bioactivity can be controlled by modifying the
glass network [4,5], so it is necessary to modify the bioactive
glasses by changing its composition to obtain glass-ceramics with
high mechanical strength and excellent bioactivity.
In the preparation process of solgel glass, doping ions have a
very small effect on the glass formation [6]. K abound in body
uids, and Mg and B is the trace element in human body. These
ions have different valence and they are widely used in previous
biological materials. This study chooses 45S5 bioglass as base
material and the glasses doped with K2O, MgO and B2O3 were
carried out by solgel method, and the inuences of doping ions
on glass properties were studied.
2. Experimental
Solgel derived 45S5 bioglass with the composition of 24.4 mol%
Na2O46.1 mol% SiO226.9 mol% CaO2.6 mol% P2O5 used as the base
glass, and the substitution of 6 mol% K2O, MgO and B2O3 respectively
n
Corresponding authors at: Jing Shi Road # 17923, Jinan 250061, Shandong,
Peoples Republic of China. Tel.: +86 531 88395991; fax: +86 531 88393538.
E-mail addresses: wangdg@sdu.edu.cn (D.G. Wang),
czchen@sdu.edu.cn (C.Z. Chen).
for CaO in 45S5 bioglass was conducted. The glasses were prepared
by hydrolysis and polycondensation of tetraethyl orthosilicate,
triethyl phosphate, Ca(NO3)2
4H2O, KNO3, Mg(NO3)2
6H2O, and
H3BO3. Nitric acid was used as catalyst. Each reactant was consecu-
tively added under continuous stirring. The sol was introduced into
polyethylene containers at room temperature till the gel appeared.
The gel was kept at 60 1C for aging 3 days and then drying at 120 1C,
followed by ball milling to obtain glass powders. After that, glass
powders stabilized at 600 1C for 2 h. For preparation of compact
samples, the mixture of glass powders after stabilization and poly-
vinyl alcohol water solution binders was put into the corresponding
mold and uniaxially pressed at 200 MPa, and then the compact
samples were sintered in high temperature furnace according
to the suitable heat treatment, which was determined by
thermogravimetric and differential thermal analysis (TG/DTA)
of the dried gel.
TG/DTA were carried out in air at a heating rate of 10 1C/min (SDT
Q600 of TA Instruments). Crystalline phases were analyzed by X-ray
diffraction (XRD, Bruker, D8 Advance). Microstructure was performed
on a scanning electron microscope (SEM, JEOL, JSM 6380LA). The
volume density () was evaluated by the Archimedes method. The
bending strength (sf) and fracture toughness (KIC) was measured by
the three-point bending method (RGD-5 type electronic tensile
machine). The Microhardness (HV) were determined by indentation
test (HV-1000 Huayin micro hardness tester). The bioactivity of
ceramics was evaluated in simulated body uid (SBF) at 37 1C in
sterile polyethylene containers, and the SBF solution was replaced
every 2 days. The change in pH values of SBF solution was measured
by pH-meter (PB-10, Sartorius Stedim Biotech (Beijing) Company Ltd.).

0167-577X/$ - see front matter & 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.matlet.2013.05.013
374 H.C. Li et al. / Materials Letters 106 (2013) 373376
Fig. 1. TG/DTA curves of the base dry gels prior to heat treatment.
3. Results and discussion
Fig. 1 shows the TG/DTA curves of the base gels after being
dried at 120 1C. Two obvious weight losses were observed on the
TG curve. The rst process between room temperature and
275.93 1C was attributed to the volatilization of residual water
and ethanol. The second process occured at 729.30 1C, due to the
decomposition of nitrates. From the DTA curve, glass transition
temperature (Tg) was observed at 521.55 1C and crystallization
peaks were observed at 709.83 1C and 847.24 1C. According to the
results, 600 1C was selected as the stabilization temperature. The
heat treatment was performed by heating the glasses 30 min at
550 1C for nucleation followed by 2 h at 710 1C and 1 h at 850 1C
for crystal growth and the heating rate was 3 1C/min [7].
Fig. 2 shows the XRD patterns of the glasses under diffe-
rent conditions and the change in pH values of SBF solution with
immersion time. All samples showed the characteristic of amor-
phous material with a wide band at 20381 (2) in XRD patterns
of the glasses before crystallization treatment (Fig. 2a), indicating
the glass powder is mainly amorphous. NaNO3 was detected,
because the stabilization temperature 600 1C is lower than the
temperature of NO3- completly decomposition. Besides, Na2Ca-
Si3O8 were observed in 45S5-K and 45S5-B, which means K2O and
B2O3 can promote initial glass crystallization. K+ as a glass network
regulator, has more damage to glass structure than Ca2+. K2O will
reduce oxygen concentration in glass, destroy network structure
and make glass network more loose [8]. BO is a high combination
bond and it can form three coordinated BO or tetra-coordinated
BO. B2O3 increases the number of trihedron in glass structure and
reduces the connection degree of network [9]. The more the
fracture of glass network, the easier the crystallization.
After crystallization treatment (Fig. 2b), NaNO3 was not observed.
Amorphous diffraction band disappeared and some new sharp crystal
peaks appeared, indicating glass crystallization. Some patterns were
agreed with the standard XRD patterns for Na6Ca3Si6O18, Na4Ca4[-
Si6O18], and Na4.24Ca3.8[Si6O18]. The structures of all crystal phases are
similar, so the chemical formula can be denoted as NaxCa60.5xSi6O18
(x46), and Na6Ca3Si6O18 was the principal phase. In addition,
Na2CaSi3O8 was detected in all samples except the glass containing
Mg, which can be explained by the fact that Dietzel's ionic eld
strength of Mg2+ (2.04 2) is larger than that of Ca2+ (4.73 2), the
modifying character of Mg2+ is smaller than Ca2+. MgO acts as a
network modier like CaO, and it can enter into glass structure. When
CaO was replaced by MgO, glass network became tighter [10]. The
measuring results of mechanical properties of the glasses after crystal-
lization treatment are shown in Table 1. The volume density (),
bending strength (sf) and microhardness (HV) were signicantly
increased and the fracture toughness (KIC) was slightly reduced. KO
and MgO bond energy is larger than CaO, and BO is a high
combination bond. When CaO was substituted by K2O, MgO and
B2O3, the density, bending strength and micro-hardness will increase.
Glass-ceramic fracture belongs to brittle fracture. Many factors cause
the reduction of fracture toughness, for example: particle size, shape,
crystallinity, porosity distribution and residual strains. In addition, the
presence of more than one crystalline phase in glass-ceramic can also
lead to complex mechanical behavior [11].
After soaking 7 days in SBF (Fig. 2c), the characteristic peaks of
carbonated hydroxyapatite (Ca10(PO4)3(CO3)3(OH)2) were obviously
detected in all samples, and hydroxyapatite (Ca10(PO4)6(OH)2) were
also observed in 45S5-Mg. Crystallization did not inhibit the bioactivity
of samples. This is similar to the conclusion of Peitl's group [12], they
suggested that the apatite layer can be formed when glass-ceramic
was soaked in SBF, even glass-ceramic was fully crystallized. Besides,
CaCO3 appeared in 45S5 and 45S5-B samples. Fig. 2d shows the
change in pH values of SBF solution for this four glass-ceramics with
immersion time. Change in pH values of different glass-ceramics
basically followed the same pattern. In the beginning of immersion
time, pH value raised rapidly, which reason is that the rapid ion
exchange between Na+ , Ca
2+
or Mg2+ in material and H+ or H3O+ in
SBF solution took place. The rapid ion exchange lead to the formation
of SiO2-rich layer on glass surface, and then Ca2+ and PO43- were
migrated onto SiO2-rich layer followed by nucleation and growth of
apatite [13]. After soaking 6 h, pH value decreased, because the apatite
layer has been generated on the sample surface. The SBF solution was
replaced every 2 days, so all kinds of ions can be added in time and the
apatite layer grows continuously.
Fig. 3 is the SEM micrographs of the glass-ceramic surface before
and after soaking in SBF for 7 days. Before soaking, the surface of 45S5
 H.C. Li et al. / Materials Letters 106 (2013) 373376 375

Fig. 2. XRD patterns of (a) the glasses before crystallization treatment, (b) the glasses after crystallization treatment, (c) the glass-ceramics after soaking 7 days in SBF
solution, and (d) the change in pH values of SBF solution for the glass-ceramics with immersion time.

Table 1 MgO is larger than that of CaO, glass doped with MgO is not
Test results of the properties of the glass-ceramic.
lively in solution and the rate of ions exchange might be slower
[15]. So MgO causes a decrease of apatite deposition rate. The
Sample (g/cm3) sf (Mpa) KIC (MPa ml/2) HV(MPa)
incorporation of K2O and B2O3 reduces the integrity of glass
45S5 1.63 3.97 2.01 147.59 network and increases the network breakpoint, then promotes
45S5-K 2.14 7.06 1.65 187.50 the release of si4+, so that the ions constituted glass network are
45S5-Mg 2.74 10.48 1.43 294.89
easily leached from the SBF, and the dissolution rate of Ca and P
45S5-B 9.87 14.10 1.32 314.21
increases. Consequently, the deposition rate of apatite layer accel-
erates and the bioactivity accordingly increases.

was loose and agglomerated particles were observed (Fig. 3a). When
K2O (Fig. 3b), MgO (Fig. 3c) and B2O3 (Fig. 3d) were added, the surface
of samples became compact and smooth. After soaking, the surface of
45S5 was fully covered by numerous apatite particles (Fig. 3e). When
K2O was incorporated, the apatite layer became more close (Fig. 3f).
When MgO was incorporated, some small particles that were not fully
connected generated on 45S5-Mg surface, and the original glass was
still observed (Fig. 3g). Different from the previous one, the surface of
the containing B2O3 sample was fully covered by closely arranged
spherical particles (Fig. 3h).
Arcos's report [14] suggests that the release rate of Si4+
determines the bioactivity of the glass. The bond energy of
4. Conclusion
The partial substitution of K2O, MgO or B2O3 for CaO in the
Na2OCaOSiO2P2O5 system was conducted by the solgel
method. The incorporation of K2O and B2O3 decreases the integrity
of glass network and promotes glass crystallization. Crystallization
does not inhibit the glasses bioactivity. The incorporation of MgO
makes the network much denser and causes a decrease of apatite
deposition rate. The doping ions in glass adjust glass network and
remarkably improve the mechanical properties.
376 H.C. Li et al. / Materials Letters 106 (2013) 373376

Fig. 3. SEM micrographs of the glass-ceramic surface before(ad) and after(eh) soaking in SBF for 7 days: (a) and (e) 45S5, (b) and (f) 45S5-K, (c) and (g) 45S5-Mg, and
(d) and (h) 45S5-B.

Acknowledgment [3] Siqueira RL, Zanotto ED. Mater Sci Eng C 2011;31:17919.
[4] Ma J, Chen CZ, Wang DG, Hu JH. Mater Lett 2011;65:1303.
This work was nancially supported by the Department of [5] Guo HW, Gong YX, Gao SY. Mater Lett 2010;64:9979.
[6] Sun T, Xiao HN, Cheng Y, Liu HB. Ceram Int 2009;3:10515.
Science and Technology of Shandong Province (Grant no. [7] Sun YY, Liu HY, Huang SQ, Zhou YC, Ren XQ, Zhang R, et al. Foshan Ceramics
2009GG10003032). 2008;9:125.
[8] Thakur OP, Kumar D, Parkash O, Pandey L. Mater Lett 1995;23:25360.
[9] Fu Q, Rahaman MN, Fu HL, Liu XJ. Biomed Mater Res 2010;1:16471.
References [10] Watts SJ, Hill RG, ODonnell MD, Law RV. J Non-Cryst Solids 2010;9:51724.
[11] Denry IL, Holloway JA. Dent Mater 2004;20:2139.
[1] Brando SM, Schellini SA, Moraes AD, Padovani CR, Pellizzon CH, Peitl O, et al. [12] Peitl O, Zanotto ED, Hench LL. J Non-Cryst Solids 2001;1:11526.
Orbit 2012;3:1439. [13] Ma J, Chen CZ, Wang DG, Meng XG, Shi JZ. Ceram Int 2010;36:19116.
[2] Bretcanu O, Misra S, Roy I, Renghini C, Fiori F, Boccaccini AR, et al. J Tissue Eng [14] Arcos D, Greenspan DC, Vallet-Regi. J Biomed Mater Res 2003;3:34451.
Regen M 2009;2:13948. [15] Vallet-Reg M, Salinas AJ, Roman J, Gil M. J Mater Chem 1999;2:5158.