Materials Science and Engineering C 72 (2017) 202–211

Contents lists available at ScienceDirect

Materials Science and Engineering C

journal homepage: www.elsevier.com/locate/msec

A novel tantalum-containing bioglass. Part I. Structure and solubility

Adel MF. Alhalawani a, Mark R. Towler a,b,⁎
a
Department of Mechanical & Industrial Engineering, Ryerson University, Toronto, Ontario, Canada
b
Li Ka Shing Knowledge Institute, St. Michael's Hospital, Toronto, Ontario, Canada

a r t i c l e i n f o a b s t r a c t

Article history: Bioglasses are employed for surgical augmentation in a range of hard tissue applications. Tantalum is a bioactive
Received 17 June 2016 and biocompatible transition metal that has been used as an orthopedic medical device. It has a range of biolog-
Received in revised form 2 October 2016 ical and physical properties that make its incorporation into ionic form into bioactive glass systems promising for
Accepted 17 November 2016
various clinical applications. The work herein reports the characterization and properties of novel tantalum-con-
Available online 18 November 2016
taining glasses. A series of glasses based on the system 48SiO2-(36-X)ZnO-6CaO-8SrO-2P2O5-XTa2O5 with X
Keywords:
varying from 0 mol% (TA0) to 0.5 mol% (TA2) were synthesized. The addition of small amounts of Ta2O5 did
Bioactive glass not cause crystallization of the glasses but increasing Ta2O5 content at the expense of ZnO was found to result
Solubility in an increased number of bridging oxygens (BOs). This, along with the data recorded by differential thermal
Glass characterization analysis (DTA) and magic angle spinning-nuclear magnetic resonance (MAS-NMR), confirms that Ta acts as a
Silicate glasses glass former in this series. Solubility experiments showed that minor changes in the glass structure caused by
Tantalum Ta incorporation (0.5 mol%) exhibited greater cumulative % weight loss, pH values and cumulative Zn2+ and
Sr2+ ion concentration over a period of 30 days of maturation, when compared to Ta2O5-free glasses. The results
presented in this article confirm that replacing ZnO with Ta2O5 in silicate glasses results in the formation of stron-
ger bonds within the glass network without any adverse effects on the solubility of the glasses prepared from
them.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction flow from the less to the more electronegative atom. Further, it is ac-
cepted that silica glasses undergo modification in response to the addi-
Bioactive glasses are candidate materials for a wide variety of bio- tion of other cations/atoms [8]. As an example, the alkali ions locate
medical applications as they can bond to bone and be formulated to re- themselves in the structure near the non-bridging oxygen (NBO)
lease bioactive ions into the local environment, resulting in when added to silica glasses resulting in the formation of meta, pyro
antimicrobial activity and enhanced cell response [1,2]. and ortho-silicates. [SiO4/2]0, [SiO3/2O]−, [SiO2/2O2]2 −, [SiO1/2O3]3 −
Silicate-glasses are inorganic amorphous solids composed of SiO3− 4 and [SiO4]4 −, which are present in silicate glasses, are designated as
tetrahedral units. In other words, silicon is coordinated to 4 oxygen Q4, Q3, Q2, Q1 and Q0 respectively, where the superscripts indicate the
atoms and each oxygen atom is coordinated to 2 silicon units so that number of bridging oxygens (BOs) centered on the given Si atom
the structure is a three-dimensional (3-D) network of corner connected through which it is connected to other Si atoms in the glass structure [9].
[SiO4/2] tetrahedra [3]. These SiO3−
4 tetrahedral units form the backbone The solubility of a bioglass network is related to alkali ion content
of the glass structure while modifying cations charge balance the silicate [9]; the addition of glass former cations will result in a systematic de-
chains. The ionicity of the Si\\O bond, resulting from the difference in crease in the solubility of these systems. Readers are referred to the
the electronegativity of Si and O, allows for the formation of Si\\O\\Si work of Hoppe et al. [1] for detailed information on the degradation ki-
bonds [4], forming the backbone of various bioglass systems. Si can netics of these biomaterials and the specific effect of the released ionic
also bond to other atoms depending on the glass composition [5,6]. In- dissolution products, for example Sr2+ and Zn2+ ions, impart on biolog-
deed, bond formation corresponds to a state of electronegativity equal- ical performance.
ization stated by Sanderson [7]. When a bond is formed between two Transition metals can play a dual role in oxide glasses [10]. In some
atoms, X and Z, with different electro-negativities, there is an electron concentrations the transition metal may enter the network structure
while in other concentration amounts, they may allow the oxygen
ions of their former cation to break the oxygen bridges in the system,
⁎ Corresponding author at: Department of Mechanical and Industrial Engineering,
Faculty of Engineering and Architectural Science, 350 Victoria Street, Toronto M5B 2K3,
therefore acting as a glass modifier. Tantalum (Ta) is a transition
ON, Canada. metal that has been used as a bone implant [11–13] due to its physical
E-mail address: mtowler@ryerson.ca (M.R. Towler). and biological properties. The Ta ion is reported to be bioactive and

http://dx.doi.org/10.1016/j.msec.2016.11.066
0928-4931/© 2016 Elsevier B.V. All rights reserved.
 A.M.F. Alhalawani, M.R. Towler / Materials Science and Engineering C 72 (2017) 202–211 203

biocompatible due to the formation of a stable tantalum pentoxide at room temperature (23 ± 1 °C). Glass powder samples were packed
(Ta2O5) component on its surface [14,15]. Studies [11,16] have shown into stainless steel sample holders. With the X-ray generator set at
that Ta surfaces exhibit lower contact angles and higher surface energies 30 kV and 30 mA, a copper anode was used to produce a divergent
than titanium (Ti) or hydroxyapatite (HA) surfaces offering a favorable beam with an average Kα wavelength of 1.541874 Å. The range of 10–
biological environment for adhesion, growth and differentiation of 80° 2θ with a step size of 0.02° 2θ and a count time of 10 s per step
human cells. Despite some processing challenges [13,17], the inclusion were used for the measurements. X'Pert Highscore™ data analysis soft-
of Ta and other transition metals into ionomer glasses can improve ware version 1.0 d (PANalytical, Almelo, The Netherlands) was
their thermal and chemical stability [6,18–20]. employed to find peak parameters.
In a previous article [6], the authors synthesized and characterized a
wholly new silicate-glass series in which ZnO was substituted with up 2.2.2. Particle size analysis (PSA)
to 8 mol% Ta2O5. In that work, Alhalawani and Towler showed that Ta The particle size distribution (PSD) of each glass series was recorded
incorporation into silicate-based glasses was possible by the melt- using a Multisizer 4 Particle size analyzer (Beckman Coulter, Fullerton,
quenching process. They also confirmed that Ta behaved as a glass for- CA, USA). The glass powder samples (n = 5) were evaluated in the
mer whereas Zn acted as a glass intermediate, depending on its content, range of 2 to 60 μm with a run length of 60 s. A background analysis
in that particular glass system. These novel glasses were designed spe- was performed and subtracted for accurate results. The fluid used in
cifically to formulate a series of glass polyalkenoate cements (GPCs) this case was a sodium chloride (NaCl) electrolyte solution at a temper-
for use in sternal fixation. Initial, unpublished data has confirmed that ature range of 10–37 °C. The relevant volume statistics were calculated
high Ta-containing glasses have rheology (setting and working times) on each glass composition. The average diameters (n = 5) at the 10%,
that are deemed unsuitable for sternal applications. The work herein ex- 50%, and 90% of the cumulative volume distribution (d10, d50 and d90, re-
pands the understanding of this particular bioglass system where a new spectively) were recorded.
series of the previously formulated glass [6] was synthesized containing
lower Ta2O5 contents (0.0 to 0.5 mol%). The manuscript also aims to 2.2.3. Scanning electron microscopy-energy dispersive spectroscopy (SEM-
characterize the structure and solubility of the glass system under study. EDS)
Sample imaging was carried out with an FEI Co. Quanta 200F Envi-
2. Experimental methods ronmental Scanning Electron Microscope equipped with an EDAX Gen-
esis Energy-Dispersive Spectrometer (Oxford Instruments X-max,
2.1. Glass synthesis Netherlands). Secondary electron (SE) and backscattered electron
(BSE) images were taken on glass particles and polished disc surfaces.
Three glasses were proposed for this study (Table 1), a Ta2O5-free All EDS spectra were collected at 20 kV using a beam current of 26 nA.
SiO2-ZnO-CaO-SrO-P2O5 glass (TA0) and two Ta2O5-containing glasses Quantitative EDS spectra was subsequently converted into relative con-
(TA1 and TA2). Appropriate amounts of analytical grade silica, zinc centration data (n = 3).
oxide, calcium carbonate, strontium carbonate, ammonium dihydrogen
phosphate and tantalum oxide (Fisher Scientific, Ottawa, ON, Canada; 2.2.4. Differential thermal analysis (DTA)
Sigma-Aldrich, Oakville, ON, Canada) were weighed out and hand- A combined differential thermal analyzer–thermal gravimetric ana-
mixed using a spatula. Platinum (Pt.) crucibles and a Lindberg/Blue M lyzer (DTA–TG; SDT 2960 Simultaneous DSC-TGA, TA Instruments, DW,
model furnace (Lindberg/Blue M, Asheville, NC USA) with a UP550 con- USA) was used to study the thermal properties of the glasses. A heating
troller were used for melting the sieved powders (1650 °C, 1.5 h). The rate of 20 °C/min was employed using an air atmosphere with alumina
melts were shock quenched in water to obtain frit which was then in a matched platinum crucible as a reference and then cooled to room
dried in an oven (100 °C, 1 h), ground using a ball mill (400 rounds temperature at the same rate. Sample measurements were carried out
per minute, 15 min), and sieved once more through a 45 μm mesh. every 6 s between 30 °C and 1200 °C. Data analysis was performed
TA0, TA1 and TA2 were then annealed at 670 °C, 666 °C and 677 °C, re- using NETZSCH Proteus software, V. 6 (Netzsch-Gerätebau GmbH,
spectively for 12 h to relieve internal stresses within the glass network. Selb, Germany).
The furnace (Lindberg/Blue M, Asheville, NC, USA) was programmed to
reach to the annealing temperature in 3 h and to cool down to the room 2.2.5. X-ray photoelectron spectroscopy (XPS)
temperature (25 ± 2 °C) in 3 h. The glass powders of the selected com- The powders' chemical compositions as well as local chemical envi-
positions were then sieved through a 45 μm mesh and utilised for sub- ronment were analyzed using PHI Quantera Scanning X-ray photoelec-
sequent characterization. tron Microprobe (XPS). The XPS data sets were collected with Al Kα X-
rays (monochromatic, beam size = 100 μm) at an output power of 26.2
2.2. Glass structural and thermal characterization watts, with a photon energy of 1486.6 eV and a step size of ~0.025 eV.
Survey scans (~ 0.5 eV step size) were performed with a pass energy
2.2.1. X-ray diffraction (XRD) of 140 eV to gain qualitative information such as peak identification
A Bruker D2 Phaser desktop X-ray diffractometer (Bruker AXS Inc., and peak position. Peaks identified in all survey scans were used to ad-
WI, USA) was used to obtain X-ray diffraction patterns from the glasses just high resolution scan binding energy range, pass energy (26 eV) and
beam dwelling time (~100 ms). The beam sweeps for each high resolu-
tion scan was adjusted to yield a signal-to-noise ratio of N 100:1. The an-
Table 1 alyzed area was 1–2 mm in diameter.
Composition of the glass series.

SiO2 ZnO CaO SrO P2O5 Ta2O5 2.2.6. Magic angle spinning-nuclear magnetic resonance (MAS-NMR)
29
mol% Si MAS-NMR spectra were recorded at 7.05 T (Tesla) on a Varian
TA0 48.0 36.0 6.0 8.0 2.0 0.0 Unity Inova 300 FT-NMR spectrometer (Palo Alto, CA, USA), equipped
TA1 48.0 35.8 6.0 8.0 2.0 0.2 with a cross polarization-magic angle spinning (CP-MAS) probe. The
TA2 48.0 35.5 6.0 8.0 2.0 0.5 glass samples were placed in a zirconia sample tube with a diameter
wt% of 7 mm. The sample spinning speed at the magic angle to the external
TA0 35.8 36.4 7.5 14.7 5.7 0.0 magnetic field was 5 kHz. 29Si MAS NMR spectra were taken at
TA1 35.5 35.8 7.4 14.5 5.7 1.1 59.59 MHz with 7.0-ls pulse length (pulse angle, p/2), 100-s recycle de-
TA2 35.0 35.1 7.3 14.3 5.6 2.7
lays, where the signals from 2126, 1837 and 1880 pulses were
204 A.M.F. Alhalawani, M.R. Towler / Materials Science and Engineering C 72 (2017) 202–211

accumulated for TA0, TA1 and TA2, respectively. 29Si NMR chemical
shifts are reported in ppm, with PDMS (polydimethyl silane) as the ex-
ternal reference (−34 ppm vs. TMS 0 ppm). All NMR spectra were re-
corded in a room for exclusive use of NMR, where the room
temperature was kept at 300 K by means of an air-conditioner. Data
analysis of the NMR spectra was performed by nonlinear curve-fitting
using ORIGIN software (Microcal Software Inc., Northhampton, MA,
USA).

2.3. Effect of glass structure on ion release and solubility

2.3.1. Disc sample preparation and degradation analysis

Disc samples were prepared by weighing 0.1 g powder into a stain-
less steel die (sample diameter 1.5 × 6 mm ϕ) which was pressed under
2.5 tons of pressure for 30 s. Disc samples were kept amorphous by an-
nealing at Tg ± 10 °C for 12 h. The surface area of each glass disc was
then calculated from the dimensions measured using an electronic pre-
cision caliper (Cedarlane Laboratories Ltd., Hornby, ON, Canada). Disc
samples were then weighed and immersed in measured quantities
(10 ml) of de-ionized (DI) water. All samples were maintained at
37 °C. At various time points (1, 7 and 30 days), the DI water was re- Fig. 1. XRD traces for the formulated glasses.

moved for pH and ion release analysis. Then the discs were removed,
dried in an incubator for 24 h, and weighed before being immersed in Particle size analysis (PSA) was conducted for each glass composi-
fresh volumes of DI water. This study was conducted in triplicate, and tion and the results are presented in Table 2. The PSA results were com-
the data plotted as cumulative degradation (percentage weight loss parable for all glasses under study implying that any changes through
per unit area, as a function of time). Eq. (1) was then used to obtain the series would be related to chemistry, not physicality, of the glasses.
the % weight loss per unit area: SEM was employed to provide compositional contrast images that
result from different atomic number elements and their distribution
within the glasses. EDS analysis was also performed to provide qualita-
Mo −Mt
%of weight loss ¼  100 ð1Þ tive spectra and quantitative relative proportions (wt%) of the particular
A
elements from the SEM backscattered images. SEM-EDS results of the
glass series are presented in Fig. 2a–c which show similar morphology
where Mo is the initial weight in g, Mt is the weight at time t in g and A is for the glasses through the series. Thus, the incorporation of Ta2O5 did
the surface area in cm2. not cause any obvious changes in the morphology or the mean particle
size of the glass series. The white square in the SEM images (Fig. 2a–c)
2.3.2. pH analysis indicate the interfacial regions used to identify the corresponding EDS
The pH measurements were collected using a Corning 430 pH meter spectra as well as the chemical composition (n = 3) of the glass series.
(Corning Life Sciences, Acton, MA). Prior to testing, the pH electrode was Qualitative EDS spectra showed that TA0 contains Si, Zn, Ca, Sr, and P,
calibrated using pH buffer solutions 4.00 ± 0.02 and 7.00 ± 0.02 (Fisher while TA1 and TA2 were found to have the same elements but with
Scientific, Pittsburgh, PA). Sterile DI water (pH = 6.0) was used as a con- the addition of Ta, thus confirming the starting formulation of each
trol and was measured at each time period. glass. Further, it was found, as expected, that Ta increases from 0.0 to
1.9 wt% while Zn decreases from 31.8 to 29.9 wt%, with increasing
2.3.3. Ion release profiles Ta2O5 content from 0.0 to 2.7 wt%, respectively. The Si fraction however
Each sample (n = 3) was immersed in 10 ml of DI water for 1, 7 and show a significant discrepancy. Si:Zn ratio is ~ 1:1 in the original glass
30 days prior to testing. The ion release profile of each specimen was (wt%, Table 1), however the EDS results show a 1:2 rate. This could be
measured using atomic absorption spectroscopy (AAS) on a Perkin- attributed to the high signal present for O and/or the ion diffusion
Elmer Analyst 800 (Perkin Elmer, MA, USA). AAS calibration standards through the glass. Initial quantitative analysis of the glass composition
for Sr and Zn elements were prepared from a stock solution (Sigma-Al- by using EDS can lead to the following assumptions:
drich, Oakville, ON, Canada) on a gravimetric basis. Three target calibra-
tion standards were prepared for each ion and DI water was used as a – The EDS results are usually collected at low vacuum, therefore the
control. Owing to the much greater expected concentration of ions, oxygen content recorded by EDS represents BO and NBO and may
samples were diluted in DI water at 1:10 ratio. The final cumulative con- also represent oxygen in the surrounding environment. This may re-
centration was calculated from the results of the measurements taking sult in a significant discrepancy in predicting the elemental bulk
into account the dilution factor. composition.
– EDS provides the quantitative relative proportions of the particular
3. Results and discussion elements but not the oxides, therefore EDS results cannot be solely
used for comparing the chemical composition of the processed and
3.1. Glass structural and thermal characterization formulated glass.

X-ray diffraction patterns were recorded for each of the formulated

Table 2
glasses and are presented in Fig. 1. XRD confirmed that all fired glasses Particle size analysis data for the glass series.
were fully amorphous; that no crystalline species are present in either
Average (μm) d10 (μm) d50 (μm) d90 (μm)
TA0, TA1 or TA2 during glass forming. The results presented herein indi-
cate that any changes in the properties of the glasses will be attributed TA0 11.5 6.4 8.8 20.2
to Ta2O5 incorporation rather than phase changes/separation in the TA1 11.1 6.4 8.6 18.9
TA2 10.3 6.4 8.4 16.1
glasses.
 A.M.F. Alhalawani, M.R. Towler / Materials Science and Engineering C 72 (2017) 202–211 205

Thermal profiles of the glasses are presented in Fig. 3 DTA curves also
shows the recorded thermal events and their assignments. The glass
transition temperature was observed at 670 °C, 666 °C and 677 °C for
TA0, TA1 and TA2, respectively. Previous studies [6,21] have shown
that the addition of transition metals, such as Ta2O5, to bio-glasses in-
creases the glass transition temperature (Tg) and facilitates incorpora-
tion of the transition metal oxide inside the glass network. The shift in
Tg implies increased glass stability, which may be attributed to the for-
mation of BO groups. The glass transition is followed by an exothermic
peak caused by glass crystallization (Tc). TA1 and TA2 show an ideal exo-
thermic crystallization reaction at 874 °C and 866 °C, respectively while
TA0 shows a broad peak around that region, observed at 897 °C. The
slow crystallization of TA0 can be attributed to ‘interfering’ nucleation
and oxidation transitions as well as the slow diffusion rates of the reac-
tants in TA0. This is in good agreement with our previous study where
higher levels of Ta2O5 resulted in a greater glass forming tendency and
a delay in the nucleation process [6]. Finally the melting temperature
for TA0 appears at 1094 °C. For TA1 and TA2, endothermic peaks ap-
peared at 1085 °C and 1098 °C, respectively. This last endothermic pro-
cess for TA1 and TA2 was assigned to initial decomposition and melting
of some of the glass elements since the temperature difference does not
decrease significantly, compared to that of TA0 at 1094 °C, over the
course of the peak. The glass melting temperature of TA1 and TA2 was
not observed in this study, agreeing with our previous work which indi-
cated that substituting ZnO with Ta2O5 increased melting temperature
[6]. This may indicate increased stability and homogeneity of the glass
reactants. This assumption agrees with the previous study on similar
glass compositions [6]. In that article, results from simultaneous thermal
analysis (STA) showed that increasing Ta2O5 content at the expense of
ZnO resulted in higher glass stability resulting from increased glass tran-
sition, crystallization and melting temperatures.
X-ray photoelectron spectroscopy was employed to derive informa-
tion on the elemental composition and speciation of matter by assessing
the electronic structure of the atoms residing within the surface region of
the matter being analyzed. The survey spectrum of the glasses formulat-
ed is shown in Fig. 4. Besides the expected Si2p, Zn2p3, P2p, Ca2p, Sr3d5,
Ta4d and O1s peaks, a C1s peak can be seen in the survey scans which is
attributed to ‘adventitious carbon’ present due to the adsorption of im-
purities during the glass firing process. It is, however, important to
note that the presence of this peak is common and does not affect the in-
terpretation of our results. The XPS survey scan results (Fig. 4) are in
good agreement with EDS data. TA0 was found to contain Si2p, Zn2p3,
P2p, Ca2p, Sr3d5, and O1s, while TA1 and TA2 contain each of these ele-
ments in addition to Ta4d, reflecting the initial glass formulation.
Elemental compositions of the O1s, Si2p, Zn2p3, P2p, Ca2p, Sr3d5,
and Ta4d peaks are presented in Table 3. Presenting the elemental com-
position of the C1s peak can make it difficult to compare relative chang-
es between EDS and XPS results, therefore the elemental composition of
the C1s peak is not presented. The elemental composition of all other
peaks was adjusted accordingly. Comparing the glass composition ob-
tained from both EDS and XPS with the initial batch formulation
(Table 1), it is clear that XPS gives better approximation, particularly
when comparing the Ta, Zn and Si content. The wt% of Si and Zn are al-
most equal in the expected glass compositions (Table 1) to those from
the XPS results whereas the EDS quantitative analysis showed a Zn con-
tent that is almost twice as large as that of the Si. Further, EDS and XPS
Fig. 2. SEM images of the glass series and the corresponding EDS qualitative spectra and show that Ta and O content increases while the content of Zn decreases
quantitative elemental composition (wt%). as a function of Ta2O5, thus they present a similar trend to the precursor
glass formulations. It is important to note that XPS is a surface technique
and therefore explanations offered around the glass composition are
subject to the assumption that the bulk of the glass is similar in compo-
– The penetration depth of the EDS is ~2–5 μm, hence the results also sition of the surface. However, although EDS quantitative analysis con-
include bulk composition data. This advantage of EDS analysis may siders the bulk composition of the glass, XPS results record
however be associated with masking and overlapping issues compositions closer to those from the initial batch calculation.
resulting in significant discrepancy and compositional High resolution O1s spectra were also obtained from XPS to deter-
heterogeneity. mine the effect of Ta2O5 substitution. The O1s spectra were curve fitted
206 A.M.F. Alhalawani, M.R. Towler / Materials Science and Engineering C 72 (2017) 202–211

Fig. 3. DTA curves of the glass series.

 A.M.F. Alhalawani, M.R. Towler / Materials Science and Engineering C 72 (2017) 202–211 207

with respect to BO and NBO contributions and are presented in Fig. 5. It

is clear from Fig. 5 that the binding energy of the O1s spectrum shifts
slightly from 531.8 to 532.1 eV, as a function of Ta2O5 content. This is in-
dicative of increasing the BO content in the glass, further suggesting that
tantalum acts as a network former in these glasses. Table 4 presents the
peak positions for the BO and the NBO and their corresponding atomic %
(at%). BO and NBO remained at 531.3 eV and 532.5 eV regardless of
Ta2O5 content. However increasing Ta2O5 content was found to increase
the at% of BO peaks on the expense of NBO content, thus increasing the
BO/NBO ratio. This aligns with our explanations in the previous article
[6], in which we show that Ta acts as a glass former, resulting in in-
creased network connectivity. These results also suggest decreased bio-
activity, as the Ta2O5 content increases, resulting from the formation of
additional Si\\O\\BO that are known to have a negative effect on the
ion exchange process. The solubility section of this article will provide
more detailed information with regard to this hypothesis.
MAS-NMR was employed to further investigate the structural effects
of Ta2O5 incorporation. Chemical shifts in MAS-NMR represents struc-
tural changes around the Si atom which lie in the region of − 60 to
−120 ppm for SiO4 tetrahedra [22]. Fig. 6 shows the MAS-NMR spectra
Fig. 4. XPS survey scan of the glass series. of the glass series (TA0, TA1 and TA2). Fig. 7a–c are the expanded ver-
sions of the NMR spectra of the glass series shown in Fig. 6. Fig. 7d–f
are the corresponding curve fitted (simulated) spectra. All glass samples
showed similar broad resonances at ~− 80 ppm. It can be seen that
Table 3
Elemental composition (wt%) of the glass series as determined by XPS. there are slight chemical shift differences with the chemical shift of
TA0 (−80.1) N TA1 (−82.4) N TA2 (−83.5). A shift in ppm in a negative
O1s Si2p Zn2p3 Ca2p Sr3p1 P2p Ta4d
direction, as presented with TA1 and TA2, is indicative of an increase in
TA0 37.7 23.3 26.9 3.0 7.8 1.4 0.0 BO species attached to the silicon, within the glass. Thus confirming
TA1 38.1 23.5 24.9 2.9 7.5 1.1 1.6 with the XPS results presented earlier in this manuscript. Previous stud-
TA2 38.3 23.9 23.2 2.8 7.1 1.1 3.0
ies have indicated that chemical shifts in the region between −60 and

Fig. 5. Curve fitting of the O1s spectra for the glass series with respect to BO and NBO contributions.
208 A.M.F. Alhalawani, M.R. Towler / Materials Science and Engineering C 72 (2017) 202–211

Table 4
Peak positions (eV) for the BO and the NBO peaks and their corresponding at%, obtained from the curve fitting of the O1s peak, of the glass series.

TA0 TA1 TA2

O1s (NBO) 531.3 531.3 531.3

at% 45.2 45.2 39.6
O1s (BO) 532.5 532.5 532.5
at% 54.8 54.8 60.4

− 120 ppm represent structural changes around the Si atom in a four and between TA0 and TA2 measured at 7 and 30 days. Ion release studies
coordinate state and suggested the presence of Q1, Q2 and Q3 species of both Zn2+ and Sr2+ showed significant differences (p b 0.05) when
at − 78, − 85 and − 95 ppm respectively [22]. All glasses show a results were compared with respect to aging time as well as Ta2O5 con-
broad peak around −80 ppm. This suggests that the formulated glasses tent (see Tables 5 and 6).
contain both Q1/Q2. However the broadness of the spectral envelope in Considering the former role of Ta2O5 and its substitution with ZnO in
all peaks suggests the presence of multiple Q-species and indicates that the formulated glasses, the solubility properties were expected to de-
silicon is present in distorted environments within the glass structure. crease. This assumption can be attributed to the fact that dissolution
rates must decrease with additional cross-links formed between the sil-
3.2. Glass solubility properties icate groups and tantalum ions. The solubility behavior of the glasses
under study however shows that water can diffuse into the glass struc-
As discussed, substituting Ta2O5 with ZnO resulted in increased BO/ ture causing some cations to release into the surrounding medium,
NBO ratio which correlated with a shift in the thermal events in these resulting in increased % weight loss and consequently higher pH and
glasses to higher temperatures. However, incorporation of Ta2O5, re- ion release profiles. Previous studies [23,24] have shown that solubility
placing ZnO, was found to increase the cumulative % weight loss per of a glass system strongly depends on the glass composition. It can be
unit area of the glasses under study (see Fig. 8a). The results of pH mea- generalized that the addition of network modifiers disrupts bonds with-
surements (Fig. 8b) as well as ion release studies of Zn2+ and Sr2+ ions in the glass network resulting in increased number of NBOs and subse-
(Fig. 9a, b) reflected the degradation behavior of the glasses. It can be quently increased hydration/solubility when aged in a medium such as
noted that the TA2 formulation exhibits greater cumulative % weight water or simulated body fluid. Vice versa, the addition of a glass former
loss, pH values and cumulative Zn2+ and Sr2+ ion concentration over results in the formation of additional cross-linking within the glass
a period of 30 days of maturation, when compared to TA1 and/or TA0 structure resulting in increased network connectivity, reducing solubil-
glasses. Tables 5 and 6 show the statistics around these studies, which ity. In this study, the results obtained give a basis for assuming that in-
consider both the effect of Ta2O5 content and aging on the obtained re- creasing the Ta2O5 content in the formulated silica-based glasses is
sults. Significant differences (Table 5, p b 0.05) in the cumulative % accompanied by a rapid dissolution of the unstable residual glass
weight loss could only be observed when comparing 1 day with phase at the initial stage of the interaction. This could happen due to
30 day results for TA0 and TA2 samples. However, no significant differ- the physical and chemical characteristics of Ta, which is a basic metal
ence (Table 6, p N 0.05) in the cumulative % weight loss was observed that has a highly reactive surface. The surface of the Ta2O5 is protected
when results were compared with respect to Ta2O5 content. Significant by a thin oxide layer [25], thus preventing its reaction with water. The
changes (Table 5, p b 0.05) in the pH measurements were obtained authors assume that when Ta2O5 is incorporated in a bioglass system
when results were compared with respect to all time modalities. With and soaked in water, Ta acts in the same way as Ca or Sr, meaning that
regard to Ta2O5 content, there were significant differences in pH mea- it acts as an unstable residual glass particle. This results in its quick dis-
surements (Table 6, p b 0.05) between TA0 and TA1 measured at day 7 solution in the water upon immersing. Further, the solubility behavior
of the glass system under study could have resulted from the fact that
Ta is more electropositive than Zn, as predicted from the periodic
table of elements. According to the generalized solubility rules, the
more electropositive the central atom, the more basic the oxide. The re-
sults obtained herein confirm with the general rules of solubility and
show that Ta when compared to Zn is a more electropositive metal
that increases pH of the medium in which it is immersed due to its
rapid dissolution in the surrounding medium. This initial burst of the
cations could be favorable for both cell viability and osseo-integration
as well as for fighting against bacterial species.

4. Conclusions

The work herein has shown that the synthesis of amorphous Ta2O5-
containing glasses was possible via the melt-quenching process. The in-
corporation of up to 0.5 mol% Ta2O5 at the expense of ZnO resulted in
structural changes resulting from the insertion of TaO units into the sil-
icate network. Glass solubility experiments showed that minor amounts
of Ta incorporation altered the glass solubility. The ability to control
glass solubility by minor compositional modifications offers great prom-
ise for the clinical applications of such bioglasses where the coordina-
tion of material solubility with bone remodelling/formation are of
Fig. 6. 29Si MAS-NMR spectra of the glass series. paramount importance.
 A.M.F. Alhalawani, M.R. Towler / Materials Science and Engineering C 72 (2017) 202–211 209

Fig. 7. Curve fitting of the 29Si NMR spectra: (a) expanded spectrum of TA0; (b) expanded spectrum of TA1; (c) expanded spectrum of TA2; (d) simulated (curve fitted) spectrum of (a); (e)
simulated spectrum of (b) and (f) simulated spectrum of (c).

(a) (b)
1 Day 7 Days 30 Days 1 Day 7 Days 30 Days
0.25 7.2
Cumulative degradation (% mass

7.1
0.2 7
loss/surface area)

6.9
0.15
6.8
pH

6.7
0.1
6.6
0.05 6.5
6.4
0 6.3
TA0 TA1 TA2 TA0 TA1 TA2

Fig. 8. (a) Percentage weight loss of the glass series in deionized water as a function of time, (b) pH measurements during glass solubility in deionized water. Error bars represent standard
deviation from the mean.
210 A.M.F. Alhalawani, M.R. Towler / Materials Science and Engineering C 72 (2017) 202–211

(a) Zn 2+ (b) Sr 2+
1 Day 7 Days 30 Days 1 Day 7 Days 30 Days
6 35

Cumulative concentration (ppm)

Cumulative concentration (ppm)

5 30
25
4
20
3
15
2
10
1 5

0 0
TA0 TA1 TA2 TA0 TA1 TA2

Fig. 9. Ion release profiles of (a) Zn2+ and (b) Sr2+ ions during glass solubility in deionized water. Error bars represent standard deviation from the mean.

Table 5
Glass solubility statistics (with respect to aging time).

TA0 TA1 TA2

% weight loss 1 day vs. 7 day 0.082 0.615 0.059

7 day vs. 0.536 0.921 0.984
30 day
1 day vs. 0.006⁎ 0.099 0.009⁎
30 day
pH 1 day vs. 7 day 0.018⁎ 0.000⁎ 0.000⁎
7 day vs. 0.000⁎ 0.000⁎ 0.000⁎
30 day
1 day vs. 0.000⁎ 0.000⁎ 0.000⁎
30 day
Sr ion release 1 day vs. 7 day 0.000⁎ 0.000⁎ 0.000⁎
7 day vs. 0.000⁎ 0.000⁎ 0.000⁎
30 day
1 day vs. 0.000⁎ 0.000⁎ 0.000⁎
30 day
Zn ion release 1 day vs. 7 day 0.001⁎ 0.000⁎ 0.000⁎
7 day vs. 0.018⁎ 0.000⁎ 0.000⁎
30 day
1 day vs. 0.000⁎ 0.000⁎ 0.000⁎
30 day
⁎ The mean difference is significant at the 0.05 level.

Acknowledgment References
[1] A. Hoppe, N.S. Guldal, A.R. Boccaccini, A review of the biological response to ionic
The authors gratefully acknowledge the financial assistance of the dissolution products from bioactive glasses and glass-ceramics, Biomaterials 32
Collaborative Health Research Project, CIHR/NSERC (315694-DAN). (2011) 2757–2774, http://dx.doi.org/10.1016/j.biomaterials.2011.01.004.
[2] J.R. Jones, Review of bioactive glass: from hench to hybrids, Acta Biomater. 9 (2013)
4457–4486, http://dx.doi.org/10.1016/j.actbio.2012.08.023.
Table 6 [3] W.H. Zachariasen, The atomic arrangement in glass, J. Am. Chem. Soc. 54 (1932)
Glass solubility statistics (with respect to Ta2O5 content). 3841–3851, http://dx.doi.org/10.1021/ja01349a006.
[4] S. Grabowsky, M.F. Hesse, C. Paulmann, P. Luger, J. Beckmann, How to make the
TA0 vs. TA1 TA1 vs. TA2 TA0 vs. TA2 ionic Si\\O bond more covalent and the Si\ \O\\Si linkage a better acceptor for hy-
drogen bonding, Inorg. Chem. 48 (2009) 4384–4393, http://dx.doi.org/10.1021/
% weight loss 1 day 1.000 1.000 1.000 ic900074r.
7 day 1.000 1.000 0.965 [5] S.-P. Szu, L.C. Klein, M. Greenblatt, Effect of precursors on the structure of
30 day 1.000 1.000 1.000 phosphosilicate gels: 29Si and 31P MAS-NMR study, J. Non-Cryst. Solids 143 (1992)
pH 1 day 1.000 0.417 0.280 21–30, http://dx.doi.org/10.1016/S0022-3093(05)80548-4.
7 day 0.038⁎ 0.716 0.004⁎ [6] A.M.F. Alhalawani, M.R. Towler, The effect of ZnO ↔ Ta2O5 substitution on the struc-
30 day 0.212 0.124 0.003⁎ tural and thermal properties of SiO2-ZnO-SrO-CaO-P2O5 glasses, Mater. Charact. 114
Sr ion release 1 day 0.000⁎ 0.000⁎ 0.000⁎ (2016) 218–224, http://dx.doi.org/10.1016/j.matchar.2016.03.004.
7 day 0.000⁎ 0.001⁎ 0.000⁎ [7] R.T. Sanderson, An interpretation of bond lengths and a classification of bonds, Sci-
30 day 0.000⁎ 0.001⁎ 0.000⁎ ence 114 (1951) 670–672, http://dx.doi.org/10.1126/science.114.2973.670.
Zn ion release 1 day 0.001⁎ 0.005⁎ 0.000⁎ [8] H. Darwish, S. Ibrahim, M.M. Gomaa, Electrical and physical properties of Na2O–
CaO–MgO–SiO2 glass doped with NdF3, J. Mater. Sci. Mater. Electron. 24 (2012)
7 day 0.000⁎ 0.000⁎ 0.000⁎
1028–1036, http://dx.doi.org/10.1007/s10854-012-0873-8.
30 day 0.000⁎ 0.000⁎ 0.000⁎
[9] M. Eigen, Structural Chemistry of Glasses, Elsevier, 2002, http://dx.doi.org/10.1016/
⁎ The mean difference is significant at the 0.05 level. B978-008043958-7/50030-3.
 A.M.F. Alhalawani, M.R. Towler / Materials Science and Engineering C 72 (2017) 202–211 211

[10] G. Calas, L. Cormier, L. Galoisy, P. Jollivet, Structure–property relationships in multi- [19] L. Cordeiro, R.M. Silva, G.M. de Pietro, C. Pereira, E.A. Ferreira, S.J.L. Ribeiro, et al.,
component oxide glasses, C. R. Chim. 5 (2002) 831–843, http://dx.doi.org/10.1016/ Thermal and structural properties of tantalum alkali-phosphate glasses, J. Non-
S1631-0748(02)01459-5. Cryst. Solids 402 (2014) 44–48, http://dx.doi.org/10.1016/j.jnoncrysol.2014.05.015.
[11] V.K. Balla, S. Bodhak, S. Bose, A. Bandyopadhyay, Porous tantalum structures for [20] K.M. Wetherall, P. Doughty, G. Mountjoy, M. Bettinelli, A. Speghini, M.F. Casula, F.
bone implants: fabrication, mechanical and in vitro biological properties, Acta Cesare-Marincola, E. Locci, The atomic structure of niobium and tantalum contain-
Biomater. 6 (2010) 3349–3359, http://dx.doi.org/10.1016/j.actbio.2010.01.046. ing borophosphate glasses, J. Phys. Condens. Matter. 21 (2009) 37.
[12] K.B. Sagomonyants, M. Hakim-Zargar, A. Jhaveri, M.S. Aronow, G. Gronowicz, Porous [21] T. Kosuge, Y. Benino, V. Dimitrov, R. Sato, T. Komatsu, Thermal stability and heat ca-
tantalum stimulates the proliferation and osteogenesis of osteoblasts from elderly fe- pacity changes at the glass transition in K2O–WO3–TeO2 glasses, J. Non-Cryst. Solids
male patients, J. Orthop. Res. 29 (2011) 609–616, http://dx.doi.org/10.1002/jor.21251. 242 (1998) 154–164, http://dx.doi.org/10.1016/S0022-3093(98)00800-X.
[13] V.K. Balla, S. Bose, N.M. Davies, A. Bandyopadhyay, Tantalum—a bioactive metal for [22] A. Stamboulis, R.V. Law, R.G. Hill, Characterisation of commercial ionomer glasses
implants, JOM 62 (2010) 61–64, http://dx.doi.org/10.1007/s11837-010-0110-y. using magic angle nuclear magnetic resonance (MAS-NMR), Biomaterials 25
[14] T. Miyaza, H.-M. Kim, T. Kokubo, C. Ohtsuki, H. Kato, T. Nakamura, Mechanism of (2004) 3907–3913, http://dx.doi.org/10.1016/j.biomaterials.2003.10.074.
bonelike apatite formation on bioactive tantalum metal in a simulated body fluid, [23] D.S. Brauer, C. Rüssel, J. Kraft, Solubility of glasses in the system P2O5–CaO–MgO–
Biomaterials 23 (2002) 827–832. Na2O–TiO2: experimental and modeling using artificial neural networks, J. Non-
[15] Y.-Y. Chang, H.-L. Huang, H.-J. Chen, C.-H. Lai, C.-Y. Wen, Antibacterial properties and Cryst. Solids 353 (2007) 263–270, http://dx.doi.org/10.1016/j.jnoncrysol.2006.12.
cytocompatibility of tantalum oxide coatings, Surf. Coat. Technol. 259 (2014) 005.
193–198, http://dx.doi.org/10.1016/j.surfcoat.2014.03.061. [24] N.Y. Mikhailenko, E.E. Stroganova, N.V. Buchilin, Solubility of calcium phosphate
[16] M. Roy, V.K. Balla, S. Bose, A. Bandyopadhyay, Comparison of tantalum and hydroxy- glasses and glass ceramic materials in water and physiological media, Glas. Ceram.
apatite coatings on titanium for applications in load bearing implants, Adv. Eng. 70 (2013) 158–163, http://dx.doi.org/10.1007/s10717-013-9531-8.
Mater. 12 (2010) B637–B641, http://dx.doi.org/10.1002/adem.201080017. [25] G. Mohandas, N. Oskolkov, M.T. McMahon, P. Walczak, M. Janowski, Porous tanta-
[17] J. Black, Biologic performance of tantalum, Clin. Mater. 16 (1994) 167–173, http:// lum and tantalum oxide nanoparticles for regenerative medicine, Acta Neurobiol.
dx.doi.org/10.1016/0267-6605(94)90113-9. Exp. (Wars) 74 (2014) 188–196.
[18] F. ElBatal, G. El-Bassyouni, Bioactivity of Hench bioglass and corresponding glass-ce-
ramic and the effect of transition metal oxides, SILICON 3 (2011) 185–197, http://
dx.doi.org/10.1007/s12633-011-9095-6.