127

Review Paper

Mechanical properties of bioactive glasses,

glass-ceramics and composites

I D Thompson and L L Hench

Imperial College of Science, Technology and Medicine, London

Abstract: The application of bioactive glass and glass-ceramics has been widely documented over
the past twenty years but the high modulus and low fracture toughness has made them less applicable
for clinical, load bearing, applications. The development of non-resorbable polyethylene and poly-
sulphone matrices for these materials has improved the mechanical properties. However, the primary
concern of whether the bioactivity of the composites is reduced is still unresolved. The more recent
development of resorbable carrier systems, dextran and collagen, for bioactive glasses does not intro-
duce such problems, hence making this form of composite suitable for novel soft tissue applications.
The development of a simple quality index has enabled some of the materials described within this
paper to be ranked by their ability to replace bone, thus enabling possible new research directions
to be emphasized.

Keywords: glass, glass-ceramic, composite, bioactive, resorbable

1 INTRODUCTION AND CLASSIFICATION

During the last 20 years ceramics have come to be widely

used in medical applications (1). The range of modern
medical ceramics, or bioceramics, exhibits all of the four
possible implant/tissue interactions. The four types of
implant interactions along with examples of typical
medical ceramic are:
1. Nearly inert Fixation by pure mechanical inter-
action, morphological fixation (alumina and zirconia
femoral heads).
2. Porous Fixation by biological ingrowth into implant
pores, biological fixation [hydroxyapatite (HA) and
HA coated porous metals used for femoral stems].
3. Bioactive Fixation by chemical bonding with im-
plant and tissues (bioactive glasses, HA and bioactive
glass-ceramics used for dental and orthopaedic
devices). Fig. 1 Tissue reaction spectrum for various bioceramic
4. Resorbable Fixation by replacement of implant with implants (1). (a) Relative rates of bioactivity. (b) Time
biological tissues (tricalcium phosphates). dependence of bone bonding at an implant interface,
(A=45S5 BioglassA, B=Mina 13 CeravitalA, C=
The relative bioactivity of each of the four types of 54S4.3 BioglassA, D=A/W Glass-Ceramic, E=HA,
implant–tissue reactions is shown in Fig. 1 (1, 2). The F=Kgy213 CeravitalA)
concept of bioactive ceramics was introduced in 1969
(3) by Hench et al. This led to the standard definition of bioactivity, ‘A bioactive material is one that elicits
a specific biological response at the interface of the
The MS was received on 23 June 1997 and was accepted for publication material which results in the formation of bond between
on 10 November 1997. tissues and the material’ (1). In this paper, only types 3
H03797 © IMechE 1998 Proc Instn Mech Engrs Vol 212 Part H
128 I D THOMPSON AND L L HENCH

and 4 (bioactive and resorbable) implants will be con- Table 1 Types of bioactive materials and typical trade names
sidered. In this area there are four principal varieties of
Material type Trade name/s Reference
‘bioactive’ materials. Table 1 details the materials types
and the trade names of various commercial products. Glass 45S5 BioglassA (1)
Early work did not distinguish between degrees of bio- Glass-ceramic CeraboneA (4)
CeravitalA (5)
activity but it is now known that materials such as Ceramic HA (6)
HA and bioactive glasses have very different levels of Composite HAPEXA (7)
bioactivity.
A further set of classifications of bioactive materials been produced for the class A bioactive materials (1).
was suggested and discussed in detail by Hench (2). Two Table 2 shows some of the more common mechanical
distinct classes of bioactive behaviour are described, properties of bioactive glasses and glass-ceramics com-
class A and class B. Class A bioactivity is ‘The process pared with cortical and cancellous bone (1). The bioac-
whereby a biological surface is colonised by osteogenic tive ceramic A/W (CeraboneA) has substantially greater
stem cells free in the defect environment as a result of mechanical properties than bioactive glass (45S5
surgical intervention’. An extra- and an intra-cellular BioglassA). A/W glass-ceramic is used for vertebral
response is elicited by a class A bioactive material at the replacements where a large compressive strength is
interface. Such materials are said to be osteoproductive. required.
Class B bioactivity is a bioconductive pathway that Successful use of bulk bioactive implants is based
allows bone to grow along it, thus the material only upon development of a strong interfacial bond with tis-
exhibits an extracellular response at the interface. Such sues. The failure stresses at the implant–tissue interfaces
materials are said to be osteoconductive. are rarely documented as the precise area of contact
It is from this A/B class system that the differences cannot be accurately calculated. Consequently, inter-
between bioactive glasses and some glass-ceramics facial failure loads are often quoted for a specific set
(class A) and the synthetic HA materials (class B) can of materials ( Table 3); much of this work has been
be clearly defined. Class A materials are both osteoprod- pioneered by Kokubo et al. (15–19).
uctive and osteoconductive. It should be noted that none The values quoted in Table 2 show that BioglassA fails
of these materials are osteoinductive, as this requires the at the lowest interfacial load. However, there appears
presence of bone morphogenic proteins, BMP, and other to be little statistical difference between BioglassA and
factors which are only present in natural bone. BMPs the glass-ceramic, CeravitalA. The comparatively low
cause bone cells to multiply, hence increasing rate of interfacial strength exhibited by BioglassA in the ‘pull-
bone matrix formation by effecting precursor cells to off test’ is due to failure within the silica gel layer. In
differentiate into osteogenic cells. Osteoproductive mate- other models of interfacial testing BioglassA coated
rials cause the existing bone cells to produce matrix at metal implants exhibited very high levels of interfacial
an increased rate and enhance the rate of proliferation strength (20). The data obtained from flame sprayed
of already existing osteoprogenitor cells. BioglassA metal implants suggested that the interfacial
The Index of Bioactivity (1) is a simple method of shear strength was 75 per cent of the strength of natural
determining the rate at which a bioactive material pro- monkey bone (20).
duces a biological bond between the natural and artificial The strength variation between bioactive ceramics is
materials. The index of bioactivity (I ) value is 100/(time of less importance than the problem of low toughness.
b
taken to achieve 50 per cent of interface to be bonded ). 45S5 BioglassA does not resist cyclic loading or crack
The values can be calculated from Fig. 1, but are shown growth. The stronger glass-ceramics also exhibit lower
in Table 6 (see Section 4). Class A bioactive glasses, with toughness values than those of natural, load-bearing,
a very high I , >8, also have the ability to bond to cortical bone ( Table 2). The future use of all present
b
soft tissues. bioactive ceramics in load-bearing applications is thus
limited.
Two primary solutions to the problem of brittle bio-
2 MECHANICAL PROPERTIES OF BIOACTIVE active materials are:
GLASSES AND GLASS-CERAMICS
(a) to adapt the glass and glass-ceramic to have better
The biological interactions and interfacial bonding mechanical properties;
mechanisms of BioglassA, apatite-wollastonite (A/W ) (b) to develop alternative types of material that retain
glass-ceramics and other bioceramics have been well bioactivity, but have improved mechanical
documented over the last 30 years (8). Clinical uses of properties.
bioactive implants are based upon formation of a strong Production of composites is the most likely to achieve
interfacial bond with both hard and soft tissues (1, 9). optimal mechanical properties and still retain bioactivity
However, the mechanical properties of the bioactive (21–23).
glasses are poor, and few mechanical property data have Whichever route is taken to optimize mechanical
Proc Instn Mech Engrs Vol 212 Part H H03797 © IMechE 1998
 MECHANICAL PROPERTIES OF BIOACTIVE GLASSES 129

Table 2 Mechanical properties of bone, glass and glass-ceramic artificial bone materials (estimated figures are underlined )

Density Hardness Bending strength Fracture toughness Young’s modulus

Material (g/cm3) ( Vickers, HV ) (MPa) (K ) (MPa m1/2) (GPa)
1C
Cortical bone (1) 1.6–2.1 — 50–150* 2–12 7–30
Cancellous bone (1) 1.0 — 10–20 (10) 0.1 0.05–0.5
Glasses (11)
BioglassA (45S5) 2.7 4.58±9.4 42 (tensile) 0.6 (12) 35
Glass-ceramics
CeravitalA (13) — — — — 1000–150
Cerabone A-WA (14) 3.1 680 680 2.0 118

* Bone soaked in Ringers solution during test.

Table 3 Failure loads of bone implant materials after 8 weeks rate of hydroxy-carbonate apatite (HCA) formation in
of implantation (15–19) various simulated biological fluids (SBF ). It was reported
by Kokubo that A/W glass-ceramics did not form the
Material Failure load (kg) Type of fracture
HCA layer when exposed to tris-buffer solution. However,
Glasses when A/W glass-ceramic was exposed to SBF, a polycrys-
BioglassA (45S5) 2.75±1.80 Within material talline HCA layer was formed, thus bioactivity was proven
Glass-ceramics
CeravitalA 3.52±1.48 Within material in vitro as well as in vivo (4). During detailed investigations
CeraboneA A-W 7.43±1.19 Within bone of the A/W system Li et al. also reported that growth of
a crystalline phase within a glass phase degraded the bioac-
tivity, as the mechanical properties were enhanced (24).
performance, the following design criteria are vitally In contrast, Filho, La Torre and Hench (25), exploring
important for a successful implant: the crystallization of the BioglassA system, have quantified
(a) elastic modulus must match that of bone; the effect of crystallization on the rate of apatite formation
(b) increased toughness; on partially crystallized, class A, bioactive glass-ceramics.
(c) increase of fatigue resistance; They concluded that approximately 40 per cent crystal-
(d ) increased strength; linity in the glass-ceramic had little statistical effect on
(e) maintain class A bioactivity. bioactivity. In fact, bioactivity was still observable up to
100 per cent crystallinity, the rate of HCA formation was
slightly slowed, but was rapid enough for a class A
2.1 Ceramic phases within bioactive glasses response (25).
The crystalline phase is produced in the glassy
Glass-ceramic materials generally have higher strength BioglassA by heat treatment. The process involves
and improved mechanical properties when compared with holding the glass at 550 °C for 150 h, thus producing
bioactive glasses (Table 2), however, the effects on bioac- nucleating crystals and then heating at 680 °C from 66
tivity of a dispersed crystalline phase in a glass matrix are to 113 min to produce the required crystal growth. The
of concern. The bioactivity of a glass is measured by the crystal phase produced is Na Ca Si O . The strength
2 2 3 9

Fig. 2 Mechanical properties of ceramic phased bioactive glass (strength and elastic properties) (12)
H03797 © IMechE 1998 Proc Instn Mech Engrs Vol 212 Part H
130 I D THOMPSON AND L L HENCH

Fig. 3 Mechanical properties of ceramic phase Bioglass (fracture toughness) (12)

and elastic modulus of SSP6 (12) glass-ceramics with Composites with a more elastically compliant matrix
varying ceramic volume fractions of crystal phases are have been of interest for this application.
shown in Fig. 2. The fracture toughness of the SSP6 To date there have been few ceramic/polymer com-
bioactive glasses with varying volume fractions of crys- posites formed. All of the designs have the same design
tals is shown in Fig. 3. The compositions of 45S5 and criteria, increase of strength and a modulus matching
SSP6 are shown in Table 4; both show class A that of the bone to be replaced. However, the biological
bioactivity. requirements have further divided this approach into
two areas; the distinguishing difference being the devel-
opment of (a) resorbable matrices and (b) non-resorbing
3 BIOACTIVE COMPOSITES matrices.

The original concept of reinforcing bioactive ceramics

3.1 Non-resorbing matrices
was to improve fracture toughness; this can be achieved
by two methods; (a) use of a high fracture tough phase, The concept of a polymeric bioceramic composite was
e.g. metal fibres, or (b) use of toughened ceramic par- first developed by Bonfield (28), who developed a mate-
ticles. It has been shown that BioglassA with 60 per cent rial made from two materials that matched the natural
volume fraction of stainless steel fibres increases bending components of bone, hydroxyapatite and collagen. The
strength from 42 MPa to 340 MPa (27). A/W glass cer- collagen was replaced by high-density polyethylene and
amics with a dispersed 0.5 volume fraction of tetragonal the HA was replaced by synthetically produced HA.
zirconia have improved toughness to 4 MPa m1/2 and Thus the first compliant material with properties com-
increased strength from 680 MPa to 703 MPa (27). parable to those of bone was developed (28). The same
However, despite being bioactive, all of these materials theory has been applied to BioglassA and bioactive glass-
have elastic moduli that are much higher than bone, as ceramics. This type of composite is important since it
illustrated in Fig. 4, consequently they will lead to some can provide long-term mechanical support/stability to
stress shielding. A possible use of this type of composite the fracture while possessing properties that allow it to
could be in higher load-bearing applications such as aid in the fracture healing process. There are two princi-
small joint prostheses. An obvious application is apply- pal polymeric matrices in current research; polyethylene
ing reinforcement to bioactive glass dental implants (21). and polysulphone. The structures of these polymers are
shown in Fig. 5. Composites have used BioglassA with
Table 4 Composition of SSP6 and Bioglass 45S5 (12, 26) polyethylene (29) and polysulphone (30) and A/W glass-
ceramic with polyethylene (31). The mechanical proper-
Additive SSP6 (%) Bioglass 45S5 (%) ties of these three types of material are shown in Table 5.
SiO 47.58 45.0 As with most composites, the principal variable is the
2
Na O
2
24.1 24.5 volume fraction of glass to matrix. Testing of the com-
CaO 22.77 24.5 posites has produced some unusual results, the addition
P O 5.53 6.0
2 5 of the more stiff and high-strength material should
Proc Instn Mech Engrs Vol 212 Part H H03797 © IMechE 1998
 MECHANICAL PROPERTIES OF BIOACTIVE GLASSES 131

Fig. 4 Mechanical properties of various biomaterials

increase the strength of the composite, however, the interfacial bonding between the bioceramics and the
overall strength has been reduced. The inverse relation- polymer matrices by calculating the theoretical strength
ship between volume fraction of glass/polysulphone and by a simple ‘rule of mixtures’ type calculation. Figures
strength (Table 5) is due to the angular particles forming 7 and 8 show the theoretical moduli of the materials as
crack initiation sites in the polymeric matrix (30). This if there were perfect bonding between the composite
effect is also noted with the polyethylene matrix (29). phases.
The bioactivity of all glass composites is still being The perfectly bonded materials (Figs 7 and 8) show
investigated. However, the BioglassA/polysulphone that they are stiffer than natural bone. Only the standard
(BG/PS) composite shows considerable enhancement of glassy BioglassA, in either a polyethylene or polysul-
modulus over the HA/polyethylene, BG/polyethylene phone matrix, has a modulus comparable to bone and
(BG/PE) and A/W glass-ceramic/PE systems (Fig. 6). has the added advantage of having class A bioactivity.
The BG/PS system thus has a closer match to the modu- Therefore, research should be directed at improving
lus of cortical bone, with an equivalent strain to failure. interfacial strength rather than changing the properties
The modulus of natural bone is shown in Fig. 6 as a of the BioglassA, i.e. the ceramic phase.
function of varying volume fractions of apatite present However, controlling interfacial strength is complex
in the bones from ten different animals (33). There and solutions are unlikely in the short term. Hence,
clearly is a failure of the synthetic composite materials research to reduce crack initiation sites is important.
to actually match the natural bone modulus and depen- Rounded and smaller BioglassA particles may be an
dence on the apatite phase. The reasons for the difference improvement (30) as they reduce crack initiation.
are discussed by Hench (34). They include much smaller
size HA crystallites in the natural bone, large aspect ratio
3.2 Resorbable composites
of HA crystals in bone, strong ionic bonding between
defect sites in natural HA crystals with collagen fibrils The resorbable matrices have a very specific role in
and anisotropic orientation of the HA crystals in natural modern surgery. Resorbable matrices can be used as a
bone. The present generation of synthetic bioactive com- transient carrier system. The resorbable carrier allows
posites lack each of these features and therefore offer a BioglassA particles to be delivered into a specific site.
limited compromise in mechanical properties compared Urinary incontinence is a medical condition that
with living bone. The interface between phases in the BioglassA, with its soft tissue bonding ability, could help.
synthetic composite is one of the most important limiting Traditionally polytetrafluoroethylene (PTFE ) has been
factors in long-term performance. used as a treatment. The bulking of the peri-urethral
Thus, a possible failure mechanism within the com- tissues adjacent to the sphincter increases resistance to
posites is the interface between polymer and bioceramic, overflow and increases voluntary control. This material
interfacial bonding is a typical weak link for most com- (PTFE based) has been withdrawn from the market
posites. It is possible to investigate the effect of the because of potential migration problems (9, 34).
H03797 © IMechE 1998 Proc Instn Mech Engrs Vol 212 Part H
132 I D THOMPSON AND L L HENCH

Fig. 5 Structures of polymers used in bioactive composites

Table 5 Mechanical properties of polysulfone and polyethylene BioglassA composites and A/W glass-ceramic in polyethylene
composites (particle size=45.7 mm in polyethylene, 38.45 mm in polysulphone)

Polyethylene (29) Polysulfone (30) A/W glass-ceramic polyethylene (23)

Volume
fraction Young’s Fracture Tensile Young’s Fracture Tensile Young’s Fracture Tensile
of modulus strain strength modulus strain strength modulus strain strength
Bioglass (GPa) (%) (MPa) (GPa) (%) (MPa) (GPa) (%) (MPa)

0 0.65 >360 17.89 2.6 5 (yield) 107 0.65 >360 17.89

10 1.05 105.1 14.34 0.96 >180 17.32
20 1.12 64.0 12.69 4.65 2.5 2.5 1.34 130 16.67
30 1.83 28.7 14.68
40 2.54 8.5 10.15 6.7 1.5 1.5 2.84 5.3 14.87

Proc Instn Mech Engrs Vol 212 Part H H03797 © IMechE 1998
 MECHANICAL PROPERTIES OF BIOACTIVE GLASSES 133

Fig. 6 Plot of experimental data for polymer/bioceramic composites (23, 29–32)

Fig. 7 Modulus of Bioglass, Bioglass-ceramic 40 per cent and Bioglass-ceramic 100 per cent in polyethylene
matrix (12, 31, 32)

A recent clinical study (35) used purified bovine col- sugar formed from sucrose by Leuconostoc mesenteroides
lagen with a gluteraldehyde cross-linking agent as a bacteria. The hyaluronate study (34, 36) showed that
material to attempt to cure incontinence in 32 women injections of 120–355 mm diameter BioglassA particles
above the age of 65. The results showed that improve- suspended in sodium hyaluronate (medical grade) could
ment was 79 per cent after one year and 69 per cent after be injected around pig urethra with 66.6 per cent success
two years (35). Collagen is known to be hydrolysed and of narrowing the urethra. It is believed that the 33.3 per
gradually resorbed, thus an improved, more persistent cent failure was due to either (a) the BioglassA being
material is sought. retained in the needle due to large particle diameter, thus
Studies using injectable BioglassA have been reported only sodium hyaluronate was being injected, or (b) the
by Wilson et al. (36, 37). The carrier systems used were BioglassA being injected into the bladder lumen (34).
(a) hyaluronate and (b) dextran, a polymerized glucose The temperature sensitivity and need for refrigeration of
H03797 © IMechE 1998 Proc Instn Mech Engrs Vol 212 Part H
134 I D THOMPSON AND L L HENCH

Fig. 8 Moduli of Bioglass, Bioglass-ceramic 40 per cent and Bioglass-ceramic 100 per cent in polysulphone
matrix (12, 31, 32)

the hyaluronate made it unstable. No mechanical data

on the viscosities of this type of material are available.
An alternative to the sodium hyaluronate has also
been proposed by Wilson et al. (37). The carrier material
proposed is dextran. A comparative study of a
BioglassA/dextran composite versus ZyplastA collagen
injections examined the augmentation produced by the
two materials in vivo (37). The dextran chosen was of
low molecular weight, hence enabling the composite to
have a low viscosity and to be injected into the site. Low
molecular weight dextran also has the advantage of
being rapidly metabolized, thus leaving the BioglassA to
become exposed to the biological fluids and hence
become bonded to the soft tissues. The BioglassA chosen
in the dextran experiment was of a lower particle size,
90–150 mm, than in the sodium hyaluronate experiment, Fig. 9 Persistence of BioglassA/dextran composite versus
in order to improve the injectability. The persistence ZyplastA Collagen based implant materials (37)
results of the BG/dextran composite versus collagen-
based injectable materials are shown in Fig. 9. No
mechanical data of this report have been published to approach to producing bioactive load-bearing pros-
date. theses. The non-resorbable polymer matrix materials
provide materials with a range of properties depending
upon the volume fraction of BioglassA. Due to the poly-
4 SUMMARY sulphone’s ability to match the modulus of both cortical
and cancellous bone, it is possible to make a functionally
With low strengths (moduli in the range of 32–150 GPa graded material, which allows the implant material to
and K values of 0.6–1), it is evident that BioglassA have a varying modulus throughout the device, mimick-
1C
and bioactive ceramics are brittle and do not pro- ing the natural tissue. A typical application for a func-
vide sufficient strength for load-bearing applications. tionally graded material is a dental tooth implant, where
Although these properties have improved, it is unlikely cancellous bone overlies cortical bone in the mandible.
that the BioglassA and A/W glass-ceramic are generally Here the implant could have a high modulus base with
applicable in load-bearing orthopaedic applications. a reduced modulus composite top, on which the abut-
Thus, the composite method is currently the only feasible ment for the crown can be fitted.
Proc Instn Mech Engrs Vol 212 Part H H03797 © IMechE 1998
 MECHANICAL PROPERTIES OF BIOACTIVE GLASSES 135

Table 6 Calculation of quality index (1, 7, 12, 23, 29, 30) (estimated values are underlined )

A B D (A×C×D)/B
Fracture Young’s C U. tensile Quality I (material ) I (material )
q q
Ceramic Polymeric toughness modulus I strength index
b
phase phase K lower (GPa) value (MPa) I I (cortical ) I (cancellous)
1C q q q
Cortical 6.0 15 13 100 500 1.00
Cancellous 0.1 1 13 3 8 0.02 1.0
HA / 1.0 85 3 80 3 0.01 0.4
Bioglass / 0.6 35 13 42 9 0.02 1.2
A/W GC / 2.0 118 6 215* 20 0.04 2.7
BG-C 40% / 1.0 68 13 210 37 0.07 5.0
BG-C 100% / 0.8 80 13 200 24 0.05 3.2
HA (0.4 Vf ) PE 3.0 4 3 23 49 0.10 6.5
Bioglass (0.4 Vf ) PE 1.2 3 13 10 60 0.12 8.0
Bioglass (0.4 Vf ) PS 1.2 7 13 52 116 0.23 15.5
Bioglass (0.4 Vf ) PS (modified) 1.2 5 13 103 303 0.61 40.4

Note: All values are median values from a range of mechanical properties data
* Bending strength value as UTS is not quoted.

With this evolving range of bone replacement mate- Table 6 shows that the BioglassA/modified polysulphone
rials it is important to develop a system to evaluate the material (0.4V ) has an I value higher than any other
f q
quality of a material when compared to that of natural material. However, when compared to the 500 value of
bone. There is no agreement on which material proper- cortical bone, the failure of synthetic materials becomes
ties are the most important to mimic and different apparent. The best materials at present are 60 to 10
research groups have different properties. The authors per cent of the quality of cortical bone. However, the
believe the following properties are important when eval- cancellous I value, 7.5, is more comparable to a number
q
uating artificial bone replacement materials: of the synthetic materials. The considerably lower value
is due to the fracture toughness being less than unity.
(a) elastic modulus;
Actual bioactivity testing and accurate fracture tough-
(b) strength;
ness data are required to verify these simplistic theoreti-
(c) fracture toughness;
cal calculations. The interface between the polysulphone
(d ) bioactivity.
and BioglassA is of primary concern in raising the I to
q
Using the four variables it is possible to try to produce that of the cortical bone. The modified polysulphone is
a quality index (I ) of a material’s performance com- the most likely material to reach the target values. This
q
pared to natural bone ranking identifies that improvements of fracture tough-
ness for the BG/PS composite is a potential research
fracture toughness×index of bioactivity
path to follow in order to match cortical bone. Only an
×tensile strength increase in K to 2.00 MPa m1/2 will produce an I value
I = 1C q
q Young’s modulus equal to that of bone. Rounding of particles and con-
The division of the two mechanical and one biological trolling size distribution and interface bonding could
components by the modulus is an attempt to represent make such an improvement possible.
the high-strength low-modulus properties of cortical
bone. 5 CONCLUSIONS
Table 6 details the main materials covered in this
paper and ranks them in order of quality when compared
The I data suggest that BG/PS composite material has
to cortical and cancellous bone. A material with the goal q
properties closest to those of cortical bone. This material
to reproduce cortical bone should have an I value
q also has a class A bioactivity and a very high I value.
#500, while a material aiming to reproduce cancellous b
This combination of properties suggests that it may be
bone should have a value #7.5. The cancellous bone
a valuable material in orthopaedic applications.
rankings show that the index does not produce ideal
replacement materials, therefore it may be assumed that
if any future ranking distributions are produced, all REFERENCES
materials with moduli above that of the bone are
excluded from any theoretical ranking systems. 1 Hench, L. L. and Wilson, J. In Introduction to Bioceramics,
The index of quality (I ) assumes that the bioactivity 1993, Ch. 1 ( World Scientific).
q
for a composite, 40 per cent ceramic/60 per cent polymer, 2 Hench, L. L. Bioactive ceramics: theory and clinical appli-
is equivalent to that of the pure bioceramic and that cations. In Bioceramics 7 ( Eds O. H. Anderson and A. Yli-
composites with lower volumes of ceramic are less. Urpo), 1994, pp. 3–14 (Butterworth-Heinemann, Oxford ).
H03797 © IMechE 1998 Proc Instn Mech Engrs Vol 212 Part H
136 I D THOMPSON AND L L HENCH

3 Hench, L. L., Splinter, R. J., Allen, W. C. and Greenlee, for the Advancement of Medical Instrumentation, New
T. K. Bonding mechanisms at the interface of ceramic pros- Orleans, Louisianna, 19–20 April 1974.
thetic materials. J. Biomed. Mater. Res., 1972, 2, 117–141. 21 Ducheyne, P., Marcolongo, M. and Schepers, E. In
4 Kokubo, T. In Introduction to Bioceramics (Eds L. L. Hench Introduction to Bioceramics (Eds L. L. Hench and
and J. Wilson), 1993, Ch. 5 ( World Scientific). J. Wilson), 1993, Ch. 15 ( World Scientific).
5 Gross, U. M., Muller-Mai, C. and Voigt, C. In Introduction 22 Bonfield, W. In Introduction to Bioceramics (Eds L. L.
to Bioceramics ( Eds L. L. Hench and J. Wilson), 1993, Hench and J. Wilson), 1993, Ch. 16 ( World Scientific).
Ch. 7 ( World Scientific). 23 Wang, M., Kokubo, T. and Bonfield, W. A-W glass-ceramic
6 Shors, E. C. and Holmes, R. E. In Introduction to Bio- reinforced polyethylene for medical applications. In
ceramics (Eds L. L. Hench and J. Wilson), 1993, Ch. 10 Bioceramics ( Eds T. Kokubo, T. Nakamoro and
( World Scientific). F. Miyaji), Vol. 9 (Pergamon, Japan).
7 Tanner, K. E. and Bonfield, W. Mater. World, 1997, 5(1), 24 Li, P., Yang, Q., Zhang, F. and Kokubo, T. J. Biomed.
18–20. Mater. Res., 1992, 3, 452–456.
8 Hulbert, S. F., Bokros, J. C., Hench, L. L., Wilson, J. 25 Filho, O. P., Latorre, G. P. and Hench, L. L. Effect of
and Heimke, G. Ceramics in clinical applications: past, crystallization on apatite-layer formation of bioactive glass
present and future. In High Technology Ceramics 45S5. J. Biomed. Mater. Res., 1996, 30, 509–514.
( Ed. P. Vincenzini), 1987 ( Elsevier, Amsterdam, The 26 Hench, L. L. and Andersson, O. In Introduction to
Netherlands). Bioceramics (Eds L. L. Hench and J. Wilson), 1993, Ch. 3
9 Hench, L. L. Bioceramics from concept to clinic. J. Am. ( World Scientific).
Ceramics Soc., 1991, 74(7), 1487–1510. 27 Cao, W. and Hench, L. L. Bioactive materials. Ceramics
10 Bronzino, J. D. (Ed.) Biomedical Engineering Handbook, Int., 1996, 22, 493–507.
1995 (CRC Press). 28 Bonfield, W. et al. Hydroxyapatite reinforced polyethyl-
11 Hench, L. L. and Ethridge, E. C. Biomaterials, an Interfacial ene—a mechanically compatible implant. Biomaterials,
Approach, 1982, p. 137 (Academic Press, New York). 1981, 2, 185.
12 Filho, O. P. Virto-ceramica bioativa de alto desempenho
29 Wang, M., Bonfield, W. and Hench, L. L. BioglassA/high
mechanico. PhD thesis, Federal University of Sao Carlos,
density polyethylene composite as a new soft tissue bonding
Brazil, 1995.
material. In Bioceramics (Eds J. Wilson, L. L. Hench and
13 Bromer, H., Deutscher, B., Blenke, B., Pfeil, E. and
D. G. Greenspan), Vol. 8 (Pergamon, Florida, USA).
Strunz, V. Properties of the bioactive implant material
30 Orefice, R. L., LaTorre, G. P., West, J. K. and Hench, L. L.
CeravitalA. In Science of Ceramics, Vol. 9, 1977, pp.
Processing and characterisation of bioactive polysulphone
219–223.
BioglassA composites. In Bioceramics (Eds J. Wilson, L. L.
14 Kokubo, T. Mechanical properties of new types of glass-
Hench and D. G. Greenspan), Vol. 8 (Pergamon,
ceramic for prosthetic applications. In Multiphase
Florida, USA).
Biomedical Materials (Eds T. Tsuruta and A. Nakajima),
31 Evans, G. P., Behiri, J., Currey, J. D. and Bonfield, W.
1989 ( VSP, Utrecht, The Netherlands).
Microhardness and Young’s modulus in cortical bone
15 Nakamura, T., Yamamuro, T., Higashi, S., Kokubo, T. and
exhibiting a wide range of mineral volume fractions and in
Ito, S. A new glass-ceramic for bone replacement: evalu-
a bone analogue. J. Mater. in Medicine, 1990, 1, 38.
ation of its bonding to bone tissue. J. Biomed. Mater. Res.,
1985, 19, 685–698. 32 Bonfield, W. Composite materials. In Bioceramics (Eds
16 Nakamura, T., Yamamuro, T., Higashi, S., Katutani, Y., T. Kokubo, T. Nakamoro and F. Miyaji), Vol. 9
Kitsugi, T., Kokubo, T. and Ito, S. A new bioactive glass- (Pergamon, Japan).
ceramic for artificial bone. In Treatise on Biomedical 33 Hench, L. L. Biochemical processing of materials: a review.
Materials (Ed. T. Yamamuro), 1985, pp. 109–117 In Better Ceramics Through Chemistry VI ( Eds A. K.
(Research Centre for Medical Polymers and Biomaterials, Cheetham, J. Brinker, M. L. Mecartney and C. Sanchez),
Kyoto University, Kyoto, Japan). Vol. 346, 1994, pp. 993–1004 (Materials Research Society,
17 Kotani, S., Yamamuro, T., Nakamura, T., Kisugi, T., Pittsburgh, Pennsylvania).
Fujmita, Y., Kawanabe, K., Kokubo, T. and Ohtsuki, C. The 34 Walker, R. D., Wilson, J. and Clarke, A. E. Injectable bio-
bone-bonding behaviour of two glass-ceramics ( KGS and glass as a potential substitute for injectable polytetrafluoro-
A-W GC ). In Bioceramics (Ed. G. Heimke), Vol. 2, 1990, ethylene. J. Urology, 1992, 148, 645–647.
pp. 105–112 (German Ceramic Society, Cologne). 35 Stanton, S. L. and Monga, A. K. Incontinence in elderly
18 Kotani, S., Fujita, Y., Kitsugi, T., Nakamura, T., women: is periurethral collagen an advance? Br. J. Obstet.
Yamamuro, T., Ohtsuki, C. and Kokubo, T. Bone bonding Gynaec., 1997, 104, 154–157.
mechanisms of b-tricalcium phosphate. J. Biomed. Mater. 36 Wilson, J., Clark, A. E., Walker R. D. and Ramer, M.
Res., 1991, 25, 1303–1315. Development and use of an injectable form of BioglassA.
19 Fujita, Y., Yamamuro, T., Nakamura, T., Kotani, S., In Bioceramics ( Eds W. Bonfield, G. W. Hastings and K. E.
Ohtsuki, C. and Kokubo, T. The bonding behaviour of cal- Tanner), Vol. 4, 1991 (Butterworth-Heinemann, Oxford ).
cite to bone. J. Biomed. Mater. Res., 1991, 25, 991–1003. 37 Wilson, J., Procter, D., Brownlee, R. and Elsberg, L. Tissue
20 Hench, L. L., Paschall, H. A., Allen, W. C. and augmentation with injectable BioglassA: A 12 month study
Piotrowski, G. Interfacial behavior of ceramic implants. in rabbits. In Transactions of the 5th World Biomaterials
National Bureau of Standards Special Publication 415, Congress, 1996, Vol. 1, p. 4, Vol. 2, p. 671.
Biomaterials Proceedings of a Symposium held in conjunc- 38 Orefice R. Polysulphone–BioglassA composites. PhD
tion with the Ninth Annual Meeting of the Association thesis, University of Florida, USA, 1997.
Proc Instn Mech Engrs Vol 212 Part H H03797 © IMechE 1998
 MECHANICAL PROPERTIES OF BIOACTIVE GLASSES 137

H03797 © IMechE 1998 Proc Instn Mech Engrs Vol 212 Part H