Bioactive glass preparation and use

Abstract

A process of preparing a glass comprising: (a) heating a mixture of precursor chemicals to a melt temperature to form a melt, the melt being characterized in that quenching the melt at or above a threshold temperature results in a spinodal phase separation, and quenching the melt below the threshold temperature results in a droplet phase separation; and (b) quenching the melt at or above the threshold temperature in a preheated mold to form the glass composition having the spinodal phase separation.

Claims

1. A method of using a glass to promote bioactivity, said method comprising: disposing said glass composition in a cellular environment, said glass having spinodal separation and being essentially free of pores; and facilitating cellular attachment and proliferation on said glass composition.

2. The method of claim 1, wherein said cellular attachment and/or proliferation is greater than that on glass having formed from the same mixture but with droplet phase separation or no phase separation.

3. The method of claim 1, wherein said cellular environment is a body of an animal.

4. The method of claim 3, wherein said cellular environment is a human body.

5. The method of claim 3, wherein facilitating cellular attachment comprises maintaining the life of the cells of said animal for an extended period of time.

6. The method of claim 5, wherein said animal is a human.

Description

BRIEF DESCRIPTION OF FIGURES

(1) FIG. 1(a) shows 45S5 glasses phase separated to different degrees.

(2) FIGS. 1(b) and (c) show SEMs of droplet-like phase separation and spinodal phase separation, respectively.

(3) FIG. 2 shows FT-IR spectra of droplet-like and spinodally phase separated 45S5 glasses, with spectrum of fused silica shown for comparison.

(4) FIGS. 3(a) and (b) show XRD spectra of spinodally and droplet-like phase separated 45S5 glasses, and obtained ceramics, respectively.

(5) FIG. 4(a)-(d) show typical DSC heating curves for spinodally (a, b) and droplet-like (c, d) phase separated 45S5 glasses measured for the bulk (a, c) and <32 m powder (b, d) samples.

(6) FIGS. 5(a) and (b) show DSC heating curves for spinodally and droplet-like phase separated 45S5 glasses, respectively, measured for different particle size at 5 K/min heating rate.

(7) FIG. 6 shows Ozawa plots for the estimation of the crystallization activation energy for spinodally phase-separated glass of different particle size.

(8) FIG. 7 shows Ozawa plots for the estimation of the crystallization activation energy for droplet-like phase-separated glass of different particle size.

(9) FIG. 8 shows a typical polycrystal formed at the surface of 45S5 sample.

(10) FIGS. 9(a) and (b) show Z() functions for spinodally and droplet-like phase separated 45S5 glasses, respectively, wherein open symbols correspond to bulk samples, full symbols stand for <32 m size powder.

(11) FIGS. 10(a) and (b) show positron annihilation lifetime (PAL) spectra obtained for spinodally and droplet-like phase-separated 45S5 glasses and ceramics, respectively, with example of fitting components for glass samples.

(12) FIG. 11 is a schematic of one embodiment of the method of the present invention.

(13) FIGS. 12(a) and (b) illustrate droplet and spinodal nanostructures, respectively, formed by quenching from varying temperatures.

(14) FIGS. 13(a) and (b) are SEM photos showing droplet and spinodal nanostructures, respectively.

(15) FIGS. 14(i)-(iv) illustrate nanostructure affecting cellular density and adhesion.

(16) FIGS. 15(i)-(iv) shows results confirming that cells respond to differences in the nanostructure of 45S5 bioglass.

DETAILED DESCRIPTION

(17) In one embodiment, the present invention involves a process of preparing a glass comprising heating a mixture of chemicals to a melt temperature to form a melt, the melt being characterized in that quenching the melt at or above a threshold temperature results in a spinodal phase separation, and quenching the melt below the threshold temperature results in a droplet phase separation; and quenching the melt at or above the threshold temperature to form the glass composition having the spinodal phase separation.

(18) The mixture may be any mixture for producing a glass having spinodal phase separation. Suitable mixtures include, for example, the mixture of high purity (99.99% or better) CaCO.sub.3, Na.sub.2CO.sub.3, silicon dioxide SiO.sub.2 and calcium phosphate tribasic Ca.sub.5OH(PO.sub.4).sub.3 powders, for 45S5 glass (45 wt. % SiO.sub.2-24.5 wt. % Na.sub.2O-24.5 wt. % CaO-6 wt. % P.sub.2O.sub.5), and other bioactive silicate glasses that are prone to phase separation, for example, the derivatives of 45S5 composition doped with the oxides of silver, zinc, strontium, boron, titanium, etc., or different ratios of the constituents of 45S5 composition.

(19) The melt temperature can vary based on the mixture and other process parameters including the temperature of the mold and quenching temperature. For example, suitable results have been obtained for 45S5 having a melt temperature of about 1400 to about 1600 C. As mentioned above, this melt temperature is generally higher than that used to prepare conventional 45S5. Applicants recognize that in the conventional process, the nucleation and growth mechanism, which produces isolated droplets in an otherwise continuous phase, overrules spinodal type phase separation when the melt is cast, for example, from 1250-1400 C. into unheated molds. In contrast, the glass produced from the method of the present invention has interconnected phases in a spinodal type distribution. A different degree of spinodal phase separation with two interpenetrating phases can be obtained by varying the melt temperature within 1400-1550 C. range and casting into preheated molds. In one embodiment, the melt temperature is about 1450 to about 1550 C., and, in a more particular embodiment, the melt temperature is about 1550 C.

(20) The threshold temperature will vary depending on the composition of the mixture, among other variables. For example, suitable results have been obtained in which the threshold temperature is no less than about 86% of the melt temperature. In a more particular embodiment, the threshold temperature is no less than about 88.5% of the melt temperature, and in a more particular embodiment, the threshold temperature is no less than about 89% of the melt temperature. For example, in one embodiment, in which the molds are heated above 100 C. and a 45S5 mixture is used, the threshold temperature is no less than about 1375 C. In a more particular embodiment, the threshold temperature ranges from 1375 to about 1550 C., and in a still more particular embodiment, the threshold temperature ranges from about 1375 to about 1380 C.

(21) In one embodiment, the quenching comprises pouring the melt into a preheated mold. Again, the degree to which the mold is preheated will depend on the mixture and other heating parameters, although suitable results have been achieved with a mold preheated above 100 C. In a more particular embodiment, the mold is preheated to about 100 to about 300 C.

(22) It should be understood that the process of the preparing the spinodal phase separated glass may involve other steps which are well known in the art. For example, in one embodiment, the mixture is calcined before forming the melt. Suitable results have been achieved when calcining at about 800 C. for 4-7 hours.

(23) Additionally, it may be preferable to cool the melt from its melt temperature prior to quenching. For example, in one embodiment, the melt is allowed to cool gradually from 1550 C. before being quenched. In one particular embodiment, the melt is allowed to cool to 1380 C. before it reaches the threshold temperature.

(24) Another aspect of the invention is the product made from the process described above.

(25) Still another aspect of the invention is a method of using the glass composition to promote cellular attachment and proliferation. As mentioned above, an aspect of the invention is the recognition and discovery that the nanoscale structure of glass is important for its biomedical performance. In particular, Applicants have found that the in vitro performance of glass is superior when its structure comprises of interconnected spinodally phase separated nanostructure as opposed to the same glass with isolated droplets in a matrix. Cells respond more favorably to the morphology and composition of the phases in the former structure than in the latter structure. Applicants note that, within the broad classification of spinodal or droplet type nanostructures, significant variations of the distribution of two phases may also impact the cell response. In other words, the conditions of glass fabrication may be further optimized for improving the cell response, but the basic premise of superior performance of the spinodal nanostructure will remain. This advantage of spinodal nanostructure applies to other silicate compositions as well, such as those derived from standard 45S5 composition for instance.

(26) In one embodiment, the method comprises disposing the glass composition made from the process described above in a cellular environment, and then facilitating cellular attachment and proliferation on the glass composition. The cellular environment may be any fluid environment having living cells. Such an environment may be, for example, in a Petri dish or in an animal body. Facilitating attachment and proliferation means broadly maintaining the environment to enable the cells to live for a significant period. Such cellular environments and conditions for maintaining them are well known in the art.

EXAMPLE 1

(27) The following example demonstrates the different phase separations that can be formed depending on whether the melt is quenched at or above the threshold temperature, or below the threshold temperature. Further, these examples illustrate the different properties and influences the different phase separations have. Note these results are published in Golovchak et al. Influence of phase separation on the devitrification of 45S5 bioglass, Acta Biomater (2014), http://dx.doi.org/10.1016/j.actbio.2014.07.024 hereby incorporated by reference. Note that these are just examples, and that, one may find other combination of process parameters to obtain the two types of nanostructures.

(28) Referring to FIG. 11, a schematic of one embodiment of the method of preparing the glass composition of the present invention is shown. Specifically, a 45S5 glass (45SiO.sub.2-24.5Na.sub.2O-24.5CaO-6P.sub.2O.sub.5 by wt. %) was synthesized using melt-quenching and casting in stainless steel molds. High purity (99.99% or better) carbonates CaCO.sub.3, Na.sub.2CO.sub.3, silicon dioxide SiO.sub.2 and calcium phosphate tribasic Ca.sub.5OH(PO.sub.4).sub.3 powders were used as raw precursors, and melted in a Pt crucible. The powder of a given sample was obtained by ball milling (1 hour) of bulk pieces followed by sieving through a collection of 5 sieves (500 m, 300 m, 150 m, 75 m, 32 m) to separate particles by size.

(29) The mixture of starting precursor powders was calcined at 800 C. for 6 hours and then slowly (2 K/min) heated till the maximum melt temperature of 1550 C., where it was maintained at least 3 hours to homogenize the melt. To induce spinodal phase decomposition the melt was quenched from 1550 C. into preheated (200-400 C.) molds. Different degrees of spinodal phase separation with two interpenetrating phases are obtainable by varying the melt temperature within 1400-1550 C. range and casting into preheated molds. The nucleation and growth mechanism, which produces isolated droplets in an otherwise continuous phase, overrules spinodal type phase separation when melt is cast from 1250-1400 C. into unheated molds. Thus the nanostructure of 45S glass, which can be controlled by varying the temperature at which the melt is maintained before casting, the temperature of the mold and subsequent cooling routine.

(30) For comparison a commercial 45S5 Bioglass was used, which had pronounced droplet-like phase separation. The two types of phase-separation (i.e., spinodal and drop-like) arise from the difference in the melt temperature just prior to casting, and the quench rate during solidification.

(31) The 45S5 glasses were transformed into glass-ceramics by heating in two stepsannealing of the material at 650-670 C for 3 hours to induce crystal nucleation, followed by another heat treatment at 730-750 C. for 6 hours to facilitate their growth.

(32) The average composition of the prepared glasses was checked with X-ray photoelectron spectroscopy (XPS). No significant difference was found between the droplet-like and spinodally phase separated glasses as set forth in Table 1.

(33) TABLE-US-00001 TABLE 1 Composition of the investigated spinodally and droplet-like phase separated 45S5 glasses and derived ceramics. determined from XPS data. Composition, at. % Sample O Si Ca Na P Theoretical 55.2 16.3 9.5 17.2 1.8 Spinodally phase-separated glass 50.1 18.9 8.8 20.5 1.6 Ceramics out of spinodally phase- 49.4 19.3 9.4 19.9 1.9 separated glass Droplet-like phase-separated glass 49.8 17.9 9.0 21.3 1.9 Ceramics out of droplet-like phase- 50.7 19.0 8.8 19.4 2.3 separated glass

(34) The XPS spectra were recorded in a normal emission mode on sample surfaces freshly fractured inside the ultra-high vacuum of the Scienta ESCA-300 spectrometer using monochromatic Al K X-rays (1486.6 eV). The XPS data consisted of survey scans over the entire binding energy (BE) range and selected scans over the core-level photoelectron peaks of interest. The surface charging from photoelectron emission was neutralized using a low energy (<10 eV) electron flood gun. The experimental positions of core levels for all of the investigated samples were adjusted by referencing to the 1 s core level peak of adventitious carbon at 284.6 eV. Data analysis of the XPS spectra was conducted using the standard CASA-XPS software package.

(35) Positron annihilation lifetime (PAL) spectra were recorded with the fast coincidence system of 230 ps resolution (FWHM of a single Gaussian, determined by measuring 60Co isotope) at the temperature, T=22 C. and relative humidity, RH=35%. Each PAL spectrum was measured with a channel width of 6.15 ps (total number of channels 8000) and contained 3.Math.10.sup.6 coincidences in total. Isotope .sup.22Na (activity 50 kBq) was used as source of positrons (prepared from aqueous solution of .sup.22NaCl, wrapped with Kapton foil of 12 m thickness and sealed), which was sandwiched between two identical samples. All the PAL spectra of the investigated samples (dried at 120 C. for 4 hours in vacuum before the measurements) were decomposed into four discrete exponentials s(t)=(I.sub.i/.sub.i)exp(t/.sub.i) with average positron lifetime .sub.i and intensity I.sub.i of i.sup.th positron decay component (i=1 to 4) using standard LT 9.0 program. The additional peaks into the envelope of fitted curve were added only if they significantly improved goodness of the fit. The uncertainties in the determination of lifetimes (.sub.i) and corresponding intensities (I.sub.i) were 0.005-0.5 ns (increasing with increasing .sub.i) and 0.2-1%, respectively.

(36) FT-IR measurements on samples polished to high optical quality were performed in a reflection mode (nearly normal incident angle), using Varian 7000e spectrometer.

(37) Rigaku MiniFlex II diffractometer was used for X-ray powder diffraction (XRD) studies. The XRD patterns were recorded within 15-60 angular range, 0.02 scan step and 1 s integration time.

(38) DSC data were acquired for bulk and powder samples using NETZSCH 404/3F microcalorimeter pre-calibrated with a set of standard elements. DSC curves were recorded in a nitrogen atmosphere with 2, 5, 10, 15 and 20 K/min heating rates. Three independent DSC measurements were performed in each case to confirm the reproducibility of the obtained results. Raw DSC data were processed using NETZSCH PC software package.

(39) Spinodally phase-separated 45S5 glass appears slightly purple in the reflected light in comparison to the glass with droplet-like phase separation (FIG. 1a). This color originates from the scattering of light caused by the co-existence of two interconnected phases with close refractive indexes; UV-VIS spectroscopy testified that no characteristic color lines were present in UV-VIS absorption spectra. High-resolution SEM images in FIG. 1b,c show the difference in the structure of the colorless and purple tinged glasses at the nanoscale. Two interconnected phases can be clearly seen in FIG. 1c, which is characteristic of spinodal type phase decomposition, whereas fingerprints of droplet-like phase separation with sharp compositional boundaries are visible on the surface of colorless samples (FIG. 1b). In the latter case, the droplets are too big (1 m) to cause color as a result of scattering. However, in both cases the matrix of the investigated glasses consists of two immiscible amorphous phases: one is supposed to be silica-rich and the other containing more sodium/calcium ions and phosphorus.

(40) FT-IR reflection spectra of the two types of 45S5 glasses recorded in the region of fundamental vibrations absorption show only subtle differences (FIG. 2), which is not surprising owing to the integral nature of the measured spectra (averaged through a relatively large macroscopic surface area). As compared to fused silica, where the bending vibrations of SiO.sub.4 tetrahedral units at 800 cm.sup.1, asymmetric stretch modes at 1123 cm.sup.1 and 1220 cm.sup.1 are observed (FIG. 2), the 45S5 glass has these optical modes shifted towards lower frequencies (700 cm.sup.1 for bending and 1050-1100 cm.sup.1 for asymmetric stretching vibrations, respectively), which is typical for any other alkali and alkaline earth containing glasses. Also, new bands at 900 and 850 cm.sup.1 emerge, which are associated with the stretching vibrations of SiO.sub.4 tetrahedra having one or two non-bridging oxygens (NBO) in the nearest surrounding of Si atoms (so-called Q.sup.3 and Q.sup.2 groups). The band at 600 cm.sup.1 is usually attributed to phosphate complexes, having PO bending vibrations in this range.

(41) In principle, viscosity of the phase separated silicate glass should be higher than the viscosity of homogeneous glass of the same composition, because the separated silica-rich matrix has high viscosity and dominates flow behavior. Indeed, the activation energy of viscous flow for the phase separated glasses, calculated using Ozawa plot for glass transition temperature determined at various heating rates, is found to be dependent on the degree and type of phase separation. Thus, the value of the activation energy of viscous flow for spinodally phase separated glass was found to be 100 kJ/mol lower than the one for droplet-like phase separated 45S5 glass. We can speculate then that in the case of droplet-like phase separation more alkaline ions are removed from the glass matrix into the droplets, leading to higher viscosity of the residual silica-rich phase.

(42) The 45S5 glass transforms into a glass-ceramic by heating the material at >700 C. for more than 0.5 h, resulting in several possible crystalline phases viz. Na.sub.2Ca.sub.2Si.sub.3O.sub.9, Na.sub.2CaSi.sub.2O.sub.6-combeite and Na.sub.2Ca.sub.4(PO.sub.4).sub.2SiO.sub.4-silicorhenanite. XRD data of as-prepared glasses and fully crystallized samples (after DSC scans) are shown in FIG. 3. Predominant crystalline phase identified in both ceramics prepared from droplet-like and spinodally phase-separated parent glasses is combeite Na.sub.2CaSi.sub.2O.sub.6. For a spinodally phase-separated glass the additional reflections in FIG. 3b are caused by crystalline Si, which was added to the powder for calibration purposes.

(43) From comparison of typical DSC scans for the investigated samples (FIGS. 4 and 5) it can be ascertained that crystallization kinetics differs for glasses with different types of phase separation and it also depends on whether the sample is in bulk or powder form. In particular, the activation energy of crystallization calculated according to the Ozawa method is found to be higher for the bulk glass with droplet-like phase separation (compare curves for bulk samples in FIGS. 6 and 7). This can be explained by the fact that crystallization of this glass starts at 30 K lower temperatures than that of the spinodally phase separated sample, where the viscosity of supercooled liquid is generally higher. The higher viscosity means more constraints for structural rearrangements needed for crystallization to occur. In the case of glass with spinodal type of phase separation, crystallization starts at higher temperatures where the viscosity of the supercooled liquid is lower (FIGS. 6 and 7) and structural rearrangements are easier. Accordingly, Applicants have determined that that big droplets of Na/Ca/P-rich phase provide more nuclei for crystallization, initiating the process at lower temperatures.

(44) Decrease in the particle size of powder samples causes significant broadening of the crystallization peak and its shift towards lower temperatures (see e.g., FIGS. 4 and 5). The activation energy for crystallization becomes much higher than for the bulk samples and almost equal in magnitude for the droplet-like and spinodally phase-separated glasses when the particle size is less than 150 m (see e.g. FIGS. 6 and 7). The onset temperature of crystallization also becomes almost the same for both types of the material and simultaneously shifts by 30-50 K towards lower values in comparison to bulk crystallization. Therefore, Applicants have determined that extended surface area of powdered samples provides a high enough concentration of nucleation sites that the entire crystallization is governed by the processes on the surface.

(45) A typical SEM picture of polycrystal formed at the surface of 45S5 BG is shown in FIG. 8. The value of the crystallization activation energy for powdered samples of <300 m sizes (350-390 kJ/mol) agrees well with activation energy values 350 kJ/mol obtained earlier for 45S5 powder prepared by tape casting technique (particle sizes <10 m).

(46) The crystallization kinetics as studied with DSC are usually analyzed with Johnson-Mehl-Avrami (JMA) nucleation-growth model. However, the JMA equation for non-isothermal conditions is valid only if a certain number of criteria are satisfied: the entire nucleation process takes place during the early stages of the transformation, and becomes negligible afterward; the overall crystallization rate is defined only by the temperature and does not depend on the previous thermal history. Fundamental kinetic equations for non-isothermal crystal growth from preexisting nuclei have been developed by Ozawa and a simple method of kinetic analysis of DSC data for these processes has been proposed:

(47) $\begin{array}{cc}\frac{d}{dt}=\mathrm{Af}\left(\right)e^{\left(-\frac{E_a}{\mathrm{RT}}\right)}& \left(1\right)\end{array}$
where is fraction of crystallized volume

(48) $\begin{array}{cc}=\frac{1}{H_c}\underset{0}{\stackrel{T}{}}dT.& \left(2\right)\end{array}$
Here is the specific heat flow measured with DSC (W/g) and H.sub.c is the total enthalpy change associated with the crystallization process; the pre-exponential factor A and activation energy E.sub.a are kinetic parameters that should not depend on the temperature T and ; and
f()=m(1)[ln(1)].sup.11/m(3)
is an algebraic expression of the JMA model.

(49) It has been demonstrated that the JMA exponent m is a characteristic parameter linked to crystal forming morphology. In particular, m1 means predominant surface crystallization, while m3 corresponds to three-dimensional bulk crystallization. If one applies JMA model straight to the investigated 45S5 glasses, then m1 is typically obtained for powdered samples with <300 m particle sizes and m3-4 for the bulk samples. However, the validity of JMA equation for the investigated glasses should be demonstrated first. A simple test for the applicability of JMA model is proposed by Malek. It is based on the analysis of probe functions:

(50) $\begin{array}{cc}y\left(\right)=e^{\left(-\frac{E_a}{\mathrm{RT}}\right)}& \left(4\right)\\ z\left(\right)=T^2& \left(5\right)\end{array}$

(51) For JMA equation to be valid the maximum of the z() function should occur around =0.630.02 value. As seen from FIG. 9, the maximum for z() function is around =0.55-0.56 for bulk (and powder with particle sizes >300 m) samples of both spinodally and droplet-like phase-separated samples. Accordingly, we may conclude that the JMA equation is not applicable to describe the crystallization kinetics of bulk 45S5 BG, which is consistent with previous studies. The same is true for powdered samples with particle sizes <300 m of droplet-like phase separated glass, which also demonstrate significant deviation from =0.63 value expected for the validity of the JMA model (FIG. 9b); the values are significantly higher than predicted. However, in the case of spinodally phase-separated glass (FIG. 9a), the JMA model can be applied for the analysis of crystallization kinetics of powder samples with small particle sizes (<300 m). Accordingly, the mean value of m (about 1) obtained for small particles (<300 m) of spinodally phase-separated glass is consistent with surface initiated crystallization mechanism proposed by Clupper and Hench for tape cast 45S5 glass (particle size less than 10 m). Therefore, Applicants have determined that for small particle sizes (when the crystallization is mostly surface-driven) and spinodal-like phase separation, the crystallization kinetics can still be approximated with the JMA model.

(52) The evolution of nanopores during phase separation and crystallization is studied with PAL spectroscopy. The main potential of the PAL spectroscopy method lies in its ability to characterize the local free volumes (including both open and closed pores) in materials on a sub-nanometer scale. The PAL method is particularly effective when positronium (Ps) is formed (ortho-Ps and para-Ps), because the energetic and geometric characteristics of this electron-positron bound state (hydrogen-like atom) are well determined, which allow quantification of free volume dimensions. Thus, it is possible to estimate the average hole size from the ortho-Ps lifetime in a given material. Assuming approximately spherical shape of the free volume, the ortho-Ps lifetime (.sub.0) can be related to the average radius of holes (R) by a semi-empirical Tao-Eldrup equation:

(53) $\begin{array}{cc}{}_o={\left[2\left(1-\frac{R}{R+R}+\frac{1}{2}\sin \left(\frac{2R}{R+R}\right)\right)+0.007\right]}^{-1},& \left(6\right)\end{array}$
where R is an empirically determined parameter (in the classical case R0.1656 nm), describing effective thickness of the electron layer responsible for the pick-off annihilation of ortho-Ps in the hole.

(54) Typical PAL spectra for 45S5 glasses and glass-ceramics are shown in FIG. 10, giving the best-fit parameters as listed in Table 2. Four discrete exponentially decaying components can be distinguished in these spectra (FIG. 10), using iterative curve fitting procedure of the LT program. The first component (.sub.1, I.sub.1) includes free annihilation, para-Ps decay and is related also to the positrons' bulk lifetime (.sub.b) in the sample. The second component (.sub.2, I.sub.2) is usually caused by positrons that are trapped before annihilation in the free volume devoid of significant electron density in the glass structure, where formation of Ps is impossible for any other reason (geometric, energetic, inhibition of Ps formation, etc.). However, this component may convolute with greater bulk lifetime component of another phase, if it is present in the material.

(55) The remaining two components (.sub.3, I.sub.3), (.sub.4, I.sub.4) can be directly associated with ortho-Ps formation (Table 2). These lifetimes can be related to corresponding pores via Tao-Eldrup relation (Eq. (6)) assuming ortho-Ps is in the ground (often denoted as 1 s) state, which is usually satisfied at low temperatures and for relatively small pores. As seen from Table 2, the as-prepared 45S5 glasses both spinodally and droplet-like phase-separated contain almost the same amount of voids R.sub.31.7 in radius (estimated from .sub.3 and I.sub.3 values). However, spinodally phase-separated glass contains more voids of larger radius R.sub.43.6 (Table 2) in comparison to the droplet-like phase-separated material (estimated by .sub.4 and I.sub.4 values).

(56) TABLE-US-00002 TABLE 2 Fitting results of PAL measurements in phase separated 45S5 bioglass and derived ceramics: .sub.1 in ns, I.sub.1 in % R.sub.3 in . Component fit Voids Sample .sub.1 (0.005) I.sub.1 (1) .sub.2 (0.005) I.sub.2 (1) .sub.3 (0.01) I.sub.3 (0.4) .sub.4 (0.05) I.sub.4 (0.2) R.sub.3 R.sub.4 Spinodally phase separated Glass 0.180 50 0.420 44 1.026 4.7 2.859 1.0 1.73 3.58 Ceramics 0.180 55 0.420 41 1.467 2.8 3.762 0.6 2.33 4.16 Droplet-like phase separated Glass 0.180 47 0.420 47 1.035 4.7 3.444 0.3 1.73 3.97 Ceramics 0.180 52 0.420 45 1.383 2.9 4.240 0.3 2.22 4.43

(57) These void sizes normally are associated with fine pores within the oxide building blocks (individual coordination polyhedra) of the silica network and their interconnection. After droplet-like phase separation or ceramization process the voids agglomerate together, forming voids of larger dimensions (R.sub.33.2-3.3 and R.sub.44.2-4.4 in radii, see Table 2), which most probably reside within the boundaries between the devitrified crystallites and residual glassy phase, or different glassy phases. At the same time structure densifies at a finer scale, as indicated by the increase of I.sub.1 intensity after ceramization of both types of 45S5 phase-separated parent glasses (Table 2). This behavior can be understood if one takes into account the fact that bulk lifetime of positrons is generally lower in crystalline material (formed crystallites) than in the corresponding glass.

(58) The example shows that type of phase separation (spinodal vs. droplet-like) has a pronounced effect on the devitrification characteristics of 45S5 BG. In particular, it appears that activation energy of viscous flow for a spinodally phase separated 45S5 glass is lower than that for the droplet-like phase separated glass. Crystallization starts at lower temperatures and the activation energy of crystallization is higher for droplet-like phase separated glass, whereas the spinodally phase separated glass crystallizes at higher temperatures and, therefore, has a lower activation energy of crystallization. The JMA equation is found to be not applicable to the crystallization kinetics analysis for both types of bulk 45S5 BGs. However, for small particle sizes (<300 m), where crystallization processes are mostly surface driven, JMA equation still works under certain conditions, viz. for spinodally phase separated 45S5 glass powder with small particle size. The nature of phase separation also affects the pore distribution at the nanoscale as shown by PAL spectroscopy. As-prepared 45S5 glasses both spinodally and droplet-like phase-separated contain almost the same amount of fine voids R.sub.31.7 in radius, whereas spinodally phase-separated glass contains more voids of larger dimensions R.sub.43.6 . This difference in sub-nanoscale structure is another possible mechanism for the difference in how the proteins and cells respond to them. After devitrification these voids show the tendency to agglomeration in both types of materials.

EXAMPLE 2

(59) The following example demonstrates the bioactivity of a glass composition having a spinodal distribution. In this example, 45S5 bioglass is used which is characterized by its complex composition as follows: 24.4 mol % Na.sub.2O (sodium oxide) 26.9 mol % CaO (calcium oxide) 2.6 mol % P.sub.2O.sub.5 (phosphorus pentoxide) 46.1 mol % SiO.sub.2 (silicon dioxide).

(60) Referring to FIGS. 11 and 12, quenching from varying temperatures can alter 45S5 bioglass nanostructure as described above. Specifically, the melt-quench process is characterized by melting precursors at high temperatures followed by quenching (rapid cooling) the melt to room temperature, producing solid bioglass.

(61) Here, three glass batches were rapidly cooled from 1550, 1380, and 1370 C. to room temperature. As discussed above, cooling the batches at these different temperatures resulted in different phase separation. The differences in phase separation result in varying nanostructure in, either isolated droplets (1370 C.) or spinodal interconnected (1380 & 1550 C.) morphology according to SEM observations. Specifically, referring to FIG. 13, nanostructure of 45S5 glass quench at 1370 C. with droplet (a), and at 1380 & 1550 C. with spinodal nanostructures (b) observed by SEM (scale bar: 500 nm).

(62) Each cylindrical glass column produced from the three different samples was cut into a set of 2mm discs, polished to achieve an optical finish. All samples were re-cut, autoclaved, and sterilized through acetone washes. All samples were pre-incubated in PBS for 72 hrs at 37 C. All samples were seeded with 30,000 cells/cm.sup.2 in 3.5 cm dishes and two samples from each batch were incubated for 2, 6, 12, and 24 hrs. Samples were then subsequently fixed, and stained for DAPI, A488-Phalloidin, and Vinculin. All samples were then imaged for cell density, morphology, and attachment.

(63) It should be understood that, according to current research, upon contact with protein-rich body fluids, the surfaces of 45S5 bioglass samples are instantaneously coated and saturated with numerous extra-cellular matric (ECM) proteins, The cells of the body, therefore, do not actually contact the biomaterial itself, but rather attach to the molecular architecture of the surface-adsorbed proteins.

(64) Referring to FIG. 14, the effects of nanostructure on cellular density and adhesion are shown. Specifically, in (i) and (ii), cell density was visualized with DAPI for the 1370 C. and 1550 C. samples, respectively, and, in (iii) and (iv), Actin was visualized with Alexa-488-Phalloidin for the 1370 C. and 1550 C. samples, respectively. MC3T3-E1 pre-osteoblasts fixed and imaged 2 hours after seeding onto 45S5 bioactive glass samples that were prepared by quenching the melt at 1370 and 1550 C. Note significant increase in cell number and attachment on the sample quenched at 1550 C.

(65) Referring to FIG. 15, results confirm that cells are able to respond to differences in the nanostructure of 45S5 bioglass. Average cell density of MC3T3-E1 preosteoblasts on polished 45S5 bioglass samples prepared by quenching the melt at 1370 C., 1380 C., and 1550 C. are shown. From the results above, it is evident that cells preferred 45S5 bioglass samples exhibiting spinodal nanostructure over the samples exhibiting droplet nanostructure. (i) 2H: We observed statistically significant differences between the 1370 C. and 1380 C. samples in five experiments, while statistically significant differences between 1370 C. & 1550 C. samples were observed in 3 experiments; (ii & iii) 6H & 12H: Statistically significant differences between the 1370 C. & 1380 C. samples and between 1370 C. & 1550 C. samples were observed once; (iv) 24H: Statistically significant differences between the 1370 C. & 1380 C. samples in 2 experiments were observed, while statistically significant differences between 1370 C. & 1550 C. samples were observed in 1 experiment.

(66) Based on these results, a strong cellular preference for samples exhibiting spinodal nanostructure was observed. Specifically, the results show that cells are able to react to subtle architectural variations in 45S5 bioglass as cells attached to samples exhibiting spinodal nanostructure in larger numbers, consistently.

(67) In conclusion, Applicants have discovered that a glass comprising spinodal type nanostructure is biomedically superior to the same glass comprising droplet type nanostructure. These two types of nanostructures can be obtained by controlling the melt cooling process, which may include the melt temperature, mold temperature, casting procedure, batch size, subsequent heat treatment, etc., among other process parameters that affect thermal history during glass formation. Applicants note that within the broad classification of spinodal or droplet type nanostructures, significant variations of the distribution of two phases exist, which also impact cell response. In other words, the conditions of glass fabrication may be further optimized for improving the cell response, but the basic premise of superior performance of the spinodal nanostructure will remain.