6

CERAMIC IMPLANT MATERIALS

TEM photographs of hydroxyapatite crystals synthesized hydrothermally at 200C under 2 MPa

for 5 hr: (a) without additives, (b) KOH (10 wt%) added, (c) K3PO4 (10 wt%) added, and (d)
EDTA (5 wt%) added. Reprinted with permission from Yoshimura and Suda (1994). Copyright
1994, Chemical Rubber Co.

139
140 CH. 6: CERAMIC IMPLANT MATERIALS

Ceramics are refractory, polycrystalline compounds, usually inorganic, including silicates,

metallic oxides, carbides, and various refractory hydrides, sulfides, and selenides. Oxides such
as Al2O3, MgO, SiO2, etc. contain metallic and nonmetallic elements. Ionic salts (NaCl, CsCl,
ZnS, etc.) can form polycrystalline aggregates, but soluble salts are not suitable for structural
biomaterials. Diamond and carbonaceous structures like graphite and pyrolized carbons are
covalently bonded. The important factors influencing the structure and property relationship of
the ceramic materials are radius ratios (2.2.2) and the relative electronegativity between posi-
tive and negative ions.
Recently ceramic materials have been given a lot of attention as candidates for implant
materials since they possess some highly desirable characteristics for some applications. Ce-
ramics have been used for some time in dentistry for dental crowns by reason of their inertness
to body fluids, high compressive strength, and good aesthetic appearance in their resemblance
to natural teeth.
Some carbons have also found use as implants, especially for blood interfacing applica-
tions such as heart valves. Due to their high specific strength as fibers and their biocompatibil-
ity, they are also being used as a reinforcing component for composite implant materials
and tensile loading applications such as artificial tendon and ligament replacements. Al-
though the black color can be a drawback in some dental applications, this is not a problem
if they are used as implants. They have such desirable qualities as good biocompatibility and
ease of fabrication.

6.1. STRUCTUREPROPERTY RELATIONSHIP OF CERAMICS

6.1.1. Atomic Bonding and Arrangement
As discussed earlier in Chapter 2 (Horwitz et al., 1993), when atoms such as sodium (metal)
and chlorine (nonmetal) are ionized, the sodium will lose an electron and the chlorine will gain
an electron:
Na Na +  e , e  Cl Cl . (6-1)
Thus, the sodium and chlorine can make an ionic compound by the strong affinity of the posi-
tive and negative ions. Again, soluble salts are not suitable for structural biomaterials. The
negatively charged ions are much larger than the positively charged ions due to the gain and
loss of electrons, as given in Table 6-1. The radius of an ion varies according to the coordina-
tion numbers: the higher the coordination number, the larger the radius. For example, the oxy-
gen ion (O2) has a radius of 0.128, 0.14, and 0.144 nm for coordination numbers 4, 6, and 8,
respectively.

Table 6-1. Atomic and Ionic Radii of Some Elements

Group I Group II Group IV Group VI I

Atomic Ionic Atomic Ionic Atomic Ionic Atomic Ionic

a a a a
Element radius radius Ele. radius radius Ele. radius radius Ele. radius radius

Li+ 0.152 0.068 Be++ 0.111 0.031 O 0.074 1.40 F 0.071 0.130
Na+ 0.186 0.095 Mg++ 0.160 0.065 S-- 0.102 1.84 Cl 0.099 0.181
K+ 0.227 0.133 Ca++ 0.197 0.099 Se-- 0.116 1.98 Br 0.114 0.195
a
Covalent. Units are in nm. Reprinted with permission from Starfield and Shrager (1972). Copyright 1972,
McGraw-Hill.
BIOMATERIALS: AN INTRODUCTION 141

Figure 6-1. AX structures of ceramics. The dark spheres represent positive ions (A+) and the

circled ones represent negative ions (X ).

Table 6-2. Selected AmXn Structures

CN of Available
Prototype Lattice A (or X) sites Minimum Other
compound of A (or X) sites filled rA/Rx compounds

CsCl Simple cubic 8 All 0.732 CsI

NaCl fcc 6 All 0.414 MgP, MnS, LiF
ZnS fcc 4 1/2 0.225 E-SiC, CdS, AlP
Al2O3 hcp 6 2/3 0.414 Cr2O3, Fe2O3

Ceramics can be classified according to their structural compounds, of which AmXn is an

example. The A represents a metal and X represents a nonmetal element, and m and n are inte-
gers. The simplest case of this system is the AX structure, of which there are three types (see
Figure 6-1). The difference between these structures is due to the relative size of the ions
(minimum radius ratio). If the positive and negative ions are about the same size (rA/RX >
0.732), the structure becomes a simple cubic (CsCl structure). A face-centered cubic structure
arises if the relative sizes of the ions are quite different since the positive ions can be fitted in
the tetragonal or octagonal spaces created among larger negative ions. These are summarized
in Table 6-2. The aluminum and chromium oxide belong to the A2X3 type structure. The O2
ions form hexagonal close packing, while the positive ions (Al3+, Cr3+) fill in 2/3 of the octahe-
dral sites, leaving 1/3 vacant.

6.1.2. Physical Properties

Ceramics are generally hard; in fact, the measurement of hardness is calibrated against ceramic
materials. Diamond is the hardest, with a hardness index on the Mohs scale of 10, and talc
(Mg3Si4O10COH) is the softest (Moh's hardness 1), and others such as alumina (Al2O3; hardness
9) quartz (SiO2; hardness 8) and apatite (e.g., fluorapatite, Ca5P3O12F; hardness 5) are in be-
tween. Other characteristics of ceramic materials are their high melting temperatures and low
conductivity of electricity and heat. These characteristics come about as a result of the nature
of the chemical bonding in ceramics.
Unlike metals and polymers, ceramics are difficult to shear plastically due to the ionic na-
ture of bonding, as shown in Figure 6-2. In order to shear, the planes of atoms should slip past
each other. However, for ceramic materials the ions with the same electric charge repel each
142 CH. 6: CERAMIC IMPLANT MATERIALS

other; hence moving the plane of atoms is very difficult. This makes the ceramics brittle (non-
ductile); moreover, creep at room temperature is almost zero. Some ceramics such as cova-
lently bonded diamond have similar properties as those of ionic ceramics due to the breakages
of theses primary bonds when deformed beyond their elastic limit. The ceramics are also very
sensitive to notches or microcracks since instead of undergoing plastic deformation (or yield)
they will fracture elastically once the crack propagates. This is also the reason why the ceram-
ics have low tensile strength compared to the compressive strength, as discussed in 3.1.2. If
the ceramic is made flaw free, then it becomes very strong even in tension. Glass fibers made
this way have tensile strengths twice that of a high-strength steel (7 GPa).

Figure 6-2. Schematic two-dimensional illustration of slips in ionic and non-ionic bond materials.

Example 6-1
A piece of window glass fails at 70 MPa (104 psi). Calculate the largest size elliptic crack
which is responsible for the low strength. The stress concentration factor (V/Va) can be ex-
pressed as 2 c / r , where c is the crack depth (2c if away from surface) and r is the crack tip
radius.

Answer
Assuming the crack tip radius has the dimension of an oxygen ion (0.14 nm) and the theoreti-
cal strength of glass is 7 GPa,

r (V / V a ) 2
c
4
(1.4)(7000 / 70)2

4
= 3.5 x 102 nm or 0.35 Pm.
So, even a small microcrack significantly weakens the glass.

6.2. ALUMINUM OXIDES (ALUMINA)

Alpha-alumina (D-Al2O3) has a hexagonal close-packed structure (a = 0.4758 nm and c =
1.2991 nm). Natural alumina is known as sapphire or ruby (depending on the types of impuri-
ties that give rise to color). The single-crystal form of alumina has been used successfully to
make implants. Single-crystal alumina can be made by feeding fine alumina powders onto the
surface of a seed crystal, which is slowly withdrawn from an electric arc or oxy-hydrogen
BIOMATERIALS: AN INTRODUCTION 143

flame as the fused powder builds up. Alumina single crystals up to 10 cm in diameter have
been grown by this method.
The main source of high-purity alumina (aluminum oxide) is bauxite and native corun-
dum. The commonly available (alpha, D) alumina can be prepared by calcining alumina tri-
hydrate resulting in calcined alumina. The chemical composition and density of commercially
available pure calcined alumina are given in Table 6-3. The American Society for Testing
and Materials (ASTM) specifies 99.5% pure alumina and less than 0.1% of combined SiO2 and
alkali oxides (mostly Na2O) for implant use.

Table 6-3. Chemical Composition of Calcined Aluminas

Chemicals Composition (weight %)

Al2O3 99.6
SiO2 0.12
Fe2O3 0.03
Na2O 0.04

Aluminum Company of America. Reprinted with permission from Gitzen

(1970). Copyright 1970, American Ceramic Society.

Table 6-4. Physical Property Requirements of

Alumina Implants (ASTM, 2000)

Properties Values

Flexural strength > 400 MPa (58,000 psi)

6
Elastic modulus 380 GPa (55.1 q 10 psi)
3
Density (g/cm ) 3.83.9

The strength of polycrystalline alumina depends on porosity and grain size. Generally the
smaller the grains and porosity, the higher the resulting strength. The ASTM standards (F603-
78) require a flexural strength of greater than 400 MPa and an elastic modulus of 380 GPa, as
given in Table 6-4.
Alumina in general is a quite hard material (Mohs number of 9); the hardness varies from
2,000 kg/mm2 (19.6 GPa) to 3,000 kg/mm2 (29.4 GPa). This high hardness permits one to use
alumina as an abrasive (emery) and as bearings for watch movements. The high hardness is
accompanied by low friction and wear; these are major advantages of using the alumina as
joint replacement material in spite of its brittleness.

6.3. ZIRCONIUM OXIDES (ZIRCONIA)

Zirconium oxides or zirconia (ZrO2) have been tried for application in fabricating implants.
Zirconia is called fake diamond or cubic zirconia since it has a high refractive index (as
does diamond) and some zirconia single crystals can be made gem grade. Some mechanical
properties are as good or better than alumina ceramics. Zirconia is highly biocompatible, as are
other ceramics, and can be made in the form of large implants such as the femoral head and
acetabular cup in total hip joint replacement. These materials are strengthened by phase trans-
144 CH. 6: CERAMIC IMPLANT MATERIALS

formation and control of grain sizes. A major drawback is that they may be weakened signifi-
cantly under stress in the presence of moisture; this weakening occurs at a much faster rate at
elevated temperature such as occurs during steam sterilization (autoclaving).

Figure 6-3. Cubic structure of zirconia that belongs to the fluorite structure. Modified with
permission from Kingery et al. (1976). Copyright 1976, Wiley.

Figure 6-4. Partial phase diagram of ZrO2CaO: Css denotes cubic, Tss tetragonal, and Mss
monoclinic solid solution phase. Reprinted with permission from Drennan and Steele (1986).
Copyright 1986, Pergamon Press.

6.3.1. Structure of Zirconia

The zirconia is allotropic and the transition from monoclinic (a z b z c, D= J= 90 z E) to
tetragonal (a = b z c, D= J= E= 90) at 1000~1200C and tetragonal to cubic (a = b = c, D=
J= E= 90) structure at 2370C. The monoclinic-to-tetragonal phase transition is a diffu-
sionless transformation accompanying a volume reduction of 7.5%. The cubic structure of the
zirconia belongs to the fluorite (CaF2) structure, as shown in Figure 6-3. The crystallographic
parameters of the unit cell structures are given in Table 6-5. The partial phase diagram of
BIOMATERIALS: AN INTRODUCTION 145

Table 6-5. Physical Properties of Zirconia

Property Values
a,b
Polymorphism
Monoclinic tetragonal 12731473 (K)
Tetragonal cubic 2643 (K)
Cubic liquid 2953 (K)
Crystallography
Monoclinic
a 5.1454
b 5.2075
c 5.3107
9914'
Space group P21/c
Tetragonal
c
a 3.64
c 5.27
Space group P42/nmc
Cubic
a 5.065
Space group Fm3m
3
Density (g/cm )
Monoclinic 5.6
Tetragonal 6.10
a
Cubic 6.29
c 6
Thermal expansion coefficient (10 /K)
Monoclinic 7
Tetragonal 12
Heat of formationc (kJ/mol) 1096.7
Boiling point (K) 4549
Thermal conductivity (W /m/K)
at 100C 1.675
at 1300C 2.094
Mohs hardness 6.5
Refractive index 2.15
a
Calculated value (see Ex. 6-2).
Reprinted with permission from Drennan and Steele (1986). Copyright 1986, MIT
Press.

ZrO2CaO is shown in Figure 6-4. The CaO acts as a stabilizing oxide, where Css is the cubic
solid solution, called fully stabilized zirconia, which is resistant to most molten metals; thus it
is used to make crucibles. Partially stabilized zirconia (PSZ) is resulted in the two-phase region
of [Tss + Css]. These materials have enhanced mechanical properties. Another oxide com-
monly used to stabilize cubic zirconia is yttrium oxide (Y2O3), as shown in Figure 6-5. It is
critical that the precipitates of tetragonal phase remain small (<0.2 Pm) in the cubic zirconia
matrix to enhance its mechanical properties. If the tetragonal precipitates become large the
phase transforms into monoclinic, causing cracks in the material. To control the phase trans-
formation MgO is used along with Y2O3 during sintering and the aging process. Figure 6-6
shows the microstructures of yttrium- and magnesium-stabilized zirconia. The tetragonal pre-
cipitates strengthen the structure of the cubic zirconia matrix due to the volume difference dur-
ing the phase transformation.
146 CH. 6: CERAMIC IMPLANT MATERIALS

Figure 6-5. Phase diagram of ZrO2Y2O3. Reprinted with permission from Burger and Will-
mann (1993). Copyright 1993, Pergamon Press.

Figure 6-6. Yttria- and magnesium-stabilized zirconia (A) and tetragonal precipitates (B) in
cubic matrix grains. Reprinted with permission from Burger and Willmann (1993). Copyright
1993, Pergamon Press.

6.3.2. Properties of Zirconia

Properties of various zirconia are summarized in Table 6-6. The strength data for the partially
stabilized zirconia with yttrium oxide showed the highest flexural strength and fracture tough-
ness. However, the Weibull modulus was lower than the yttrium magnesium oxide-stabilized
BIOMATERIALS: AN INTRODUCTION 147

Table 6-6. Properties of Various Zirconia

Properties CSZ YMgPSZ Y-TZP

Youngs modulus (GPa) 210 210 210

Flexural strength (MPa) 200 600 950
Hardness (Vickers, HV0.5) 1250 1250 1250
1/2
Fracture toughness (MPa m ) 5.8 10.5
Weibull modulus 8 25 18
Density 6.1 5.85 6

Reprinted with permission from Burger and Willmann (1993). Copyright 1983, Pergamon.

Figure 6-7. Schematic representation of the toughening of zirconia with partially stabilized zir-
conia: (a) crack before phase transformation; (b) crack arrestment due to phase transformation
of the dispersed PSZ particles. Reprinted with permission from Callister (1994). Copyright
1994, Wiley.

zirconia. It is also interesting that the increased fracture toughness is due to a phase transfor-
mation that operates by arresting the propagation of cracks, as shown in Figure 6-7. Small par-
ticles of partially stabilized ZrO2 are dispersed in the matrix materials, which could be zirconia
itself. This partial stabilization enables the retention of a metastable tetragonal structure at am-
bient temperature. During crack propagation the tetragonal particles in the crack tip region
undergo phase transformation, increasing its volume, which sets up a compressive field sur-
rounding the particles and closes the crack opening, resulting in a stronger material. The proc-
ess is similar to the precipitation of a tetragonal structure in cubic grains.
The yttrium-stabilized zirconia has been used for fabricating the femoral head of total hip
joint prostheses and has two advantages over the alumina. One is the finer grain size and a
well-controlled microstructure without any residual porosity of the Y-TZP, making it better
tribological material than the alumina. The other is higher fracture strength and toughness due
to the phase transformation toughening process.
As mentioned, zirconia has many salient features in comparison with alumina. A compari-
son of the properties is given in Table 6-7. The biocompatibility of zirconia is about the same
as alumina ceramic, but its tribological properties are quite different. In one study the friction
148 CH. 6: CERAMIC IMPLANT MATERIALS

and wear properties of zirconia, alumina, and 316L stainless steel against ultra-high-molecular-
weight polyethylene (UHMWPE) were evaluated by using a uni- and bidirectional wear testing
machine in bovine serum, saline, and distilled water. Table 6-8 shows the results of wear of
UHMWPE. The wear factor was estimated by the following equation:

Wear volume (mm 3 )

Wear factor  . (6-2)
Load (N)/Sliding distance (m)

Table 6-7. Comparison of Properties of Alumina and Zirconia

Property Alumina Zirconia

Chemical composition A1203 + MgO ZrO2 + MgO + Y203

Purity (%) 99.9 95~97
3
Density (g/cm ) > 3.97 5.74~6.0
Porosity (%) < 0.1 < 0.1
Bending strength (MPa) > 500 500~1 000
Compression strength (MPa) 4100 2000
Young's modulus (GPa) 380 210
Poisson's ratio 0.23 0.3
1/2
Fracture toughness (MPa m ) 4 up to 10
6
Thermal expansion coefficient (q10 /K) 8 11
Thermal conductivity (W/m/K) 30 2
Hardness (HV0. 1) up to 2200 1200
Contact angle () 10 50

Reprinted with permission from Willmann (1993). Copyright 1993, Pergamon.

Table 6-8. Wear of UHMWPE on Two Different Wear Devices

3 9
*Wear factor (mm /N-m) q 10

Medium Bovine serum Saline Distilled water

Counterfaces Unidirectional Reciprocate Unidirectional Reciprocate Unidirectional Reciprocate

Zirconia (3) 10.7 12 0.56 14 7.5 3 0.45 5 8.61 11 0.38 6
Alumina (3) 18.2 6 1.01 8 32.7 7 0.57 2 11.8 4 0.68 4
316L SS (2) 27.7 30 1.81 4 90.5 40 3.89 8 37.1 10 1.12 10

( ) = number of specimens tested. *Average and range.

Reprinted with permission from Kumar et al. (1991). Copyright 1991, Wiley.

The wear factor for the yttrium oxide-stabilized (Y-PSZ) zirconia showed a smaller value to
alumina and 316L stainless steel in all test conditions and modes. Also, the unidirectional wear
test showed a greatly higher wear volume than the bidirectional (reciprocating) tests. The ac-
tual wear volume versus number of cycles in unidirectional tests is shown in Figure 6-8. The
wear factor is the slope of the curve divided by the load (3.45 MPa).
The friction coefficient also showed a lower value for the zirconia (0.0280.082) than alu-
mina (0.0440.115) or 316L stainless steel (0.0610.156). As with the wear factor, the bidirec-
tional reciprocating mode showed somewhat lower friction than the unidirectional arrange-
ment, although it was not as drastic as for wear. Also, the types of lubricant did not influence
the friction. One reason for the excellent wear and friction characteristic of the zirconia is at-
tributed to the fact that zirconia has less porosity, as shown in Figure 6-9. Also, the average
BIOMATERIALS: AN INTRODUCTION 149

Figure 6-8. Wear volume versus number of cycles for unidirectional test (one cycle = 50 mm)
for zirconia (a), alumina (b), and 316L stainless steel (c). Reprinted with permission from
Kumar et al. (1991). Copyright 1991, Wiley.

Figure 6-9. Scanning electron microscopic picture of polished surfaces of zirconia (A) and
alumina (B). Note the porosity in the alumina. Reprinted with permission from Kumar et al.
(1991). Copyright 1991, Wiley.
150 CH. 6: CERAMIC IMPLANT MATERIALS

grain size of zirconia (0.3 Pm) was about one-tenth that of the alumina (2.5 Pm), although the
surface roughness was about the same for both (0.0050.013 Pm Ra, the average root mean
square value of surface roughness).
Some researchers evaluated the use of zirconia for a hemiarthroplasty femoral head im-
plant and found it suitable due to its low friction with articular cartilage and its excellent bio-
compatibility. On the other hand, the wear rate of zirconiazirconia is many times that of the
aluminaalumina combination, preempting its use for the femoral head and socket.

Figure 6-10. Relationship between the bending strength and amount of phase transformation
aged in water at 95 and 121C. Reprinted with permission from Shimizu et al. (1993). Copy-
right 1993, Wiley.

Figure 6-11. Fourier transform IR spectroscopy of zirconia before (a) and after (b) aging in
water at 121C for 960 hr. Reprinted with permission from Shimizu et al. (1993). Copyright
1993, Wiley.

There is a direct one-to-one relationship between the amount of phase transformation and
bending strength of zirconia, as shown in Figure 6-10, indicating that only the amount of phase
transformation influences the mechanical properties. The moisture has an effect on the zirconia
by forming ZrOH bonding that precedes the phase transformation and was detected by infra-
red (IR) spectroscopy, as shown in Figure 6-11. The yttria-stabilized zirconia is a good candi-
BIOMATERIALS: AN INTRODUCTION 151

date to replace the alumina ceramic for orthopedic applications in spite of the effect of aging
on the mechanical properties on zirconia. Even after aging, zirconia is a much stronger mate-
rial than alumina, which has a strength of about 400 MPa.

Example 6-2
Calculate the density of cubic zirconia and compare that with the other forms of zirconia given
in Table 6.5.

Answer

Mass 4 q 91  8q16
gmol1
Density    6.29 g/cm 3 .
Volume 5.065q108 cm
3 6.02 q10 23 mol1

This value seems very reasonable compared to the density of the tetragonal structure, 6.10
g/cm3. However, it is somewhat odd that the density increases with increased temperature since
the cubic structure exists at higher temperature than the tetragonal zirconia.

6.3.3. Manufacture of Zirconia

Zircon (ZrSiO4) is a gold-colored silicate of zirconium; zircon is a mineral (baddeleyite) found
in igneous and sedimentary rocks and occurring in tetragonal crystals colored yellow, brown,
or red, depending on impurities. The zircon is first chlorinated to form ZrCl4 in a fluidized bed
reactor in the presence of petroleum coke. A second chlorination is required for high-quality
zirconium. Zirconium is precipitated with either hydroxides or sulfates, then calcined to its
oxide.
The zirconia is partially stabilized above 1700C in the cubic phase, which results in large
grain sizes (50-70 Pm). When it is cooled, a phase transformation takes place and tetragonal
precipitates can be formed in the cubic matrix. Combined cubic and tetragonal phase results in
enhanced mechanical properties.

Example 6-3
Calculate the wear constant of UHMWPE with zirconia as a mating material for a joint re-
placement. Use the data in Figure 6-8.

Answer
From Figure 6-8, the average wear volume for zirconia in bovine serum is about 0.6 mm3 after
105 cycles, with a sliding distance of 50 mm. Therefore, the wear constant can be calculated,
since the wear constant is defined as ('V is wear volume, 'l is total sliding distance, P is load,
and H is hardness)

0.6 mm 3 q3q100 MPa

Wear constant (K )   8.3 q 107.
43.35 N q 50 mm q10 5

Assume the hardness of the UHMWPE is about 100 MPa, and the load applied 43.4 N. This
corresponds to a stress of 3.45 MPa.
The wear constant for the UHMWPE with 316L stainless steel would be about 3 times
larger according to the wear volume at 100,000 cycles(~2 mm3)
152 CH. 6: CERAMIC IMPLANT MATERIALS

6.4. CALCIUM PHOSPHATE

Calcium phosphate has been used to make artificial bone. Recently, this material has been syn-
thesized and used for manufacturing various forms of implant as well as for solid or porous
coatings on other implants. There are mono-, di-, tri-, and tetra-calcium phosphates, in addition
to the hydroxyapatite and E-whitlockite, which have ratios of 5/3 and 3/2 for calcium and
phosphorus (Ca/P), respectively. The stability in solution generally increases with increasing
Ca/P ratios. Hydoxyapatite is the most important among the calcium compounds since it is
found in natural hard tissues as mineral phase. Hydroxyapatite acts as a reinforcement in hard
tissues and is responsible for the stiffness of bone, dentin, and enamel.

Figure 6-12. Hydroxyapatite structure projected down the c-axis on the basal plane. Reprinted
with permission from Posner et al. (1958). Copyright 1958, Munksgaard International.

6.4.1. Structure of Calcium Phosphate (Hydroxyapatite)

Calcium phosphate can be crystallized into the salts mono-, di- tri-, and tetra-calcium phos-
phate, hydroxyapatite, and E-whitlockite, depending on the Ca/P ratio, presence of water, im-
purities, and temperature. The most important is the hydroxyapatite due to its presence in natu-
ral bone and teeth. In a wet environment and at lower temperature (<900C), it is more likely
that the (hydroxyl or hydroxy) apatite will form while in a dry atmosphere and at higher tem-
perature the E-whitlockite (3CaOP2O3) will be formed. Both forms are very tissue compatible
and are used for bone substitute in granular form or as a solid block. We will consider the apa-
tite form of the calcium phosphate since it is considered more closely related to the mineral
phase of bone and teeth.
The mineral part of bone and teeth is made of a crystalline form of calcium phosphate
similar to hydroxyapatite [Ca10(PO4)6(OH)2]. The apatite family of minerals, A10(BO4)6X2, crys-
BIOMATERIALS: AN INTRODUCTION 153

tallizes into hexagonal rhombic prisms and has unit cell dimensions a = 0.9432 nm and c =
0.6881 nm. The atomic structure of hydroxyapatite projected down on the c-axis onto the basal
plane is given in Figure 6-12. Note that the hydroxyl ions lie on the corners of the projected
basal plane, and they occur at equidistant intervals along half of the cell (0.344 nm), along col-
umns perpendicular to the basal plane and parallel to the c-axis. Six of the ten calcium ions in
the unit cell are associated with the hydroxyls in these columns, resulting in strong interac-
tions.
The ideal Ca/P ratio of hydroxyapatite is 10/6 and the calculated density is 3.219 g/cm3. It
is interesting to note that the substitution of OH with F will give greater chemical stability due
to the closer coordination of F (symmetric shape) as compared to the hydroxyl (nonsymmetric,
two atoms) by the nearest calcium. This is one of the reasons for the better caries resistance of
teeth following fluoridation.

6.4.2. Properties of Calcium Phosphates (Hydroxyapatite)

There is a wide variation of the mechanical properties of synthetic calcium phosphates, as
given in Table 6-9. The wide variations of properties are due to the variations in the structure
of polycrystalline calcium phosphates due to variations in the manufacturing processes. De-
pending on the final firing conditions, the calcium phosphate can be calcium hydroxyapatite or
Ewhitlockite. In many instances, however, both types of structure exist in the same final
product.

Table 6-9. Physical Properties of Synthetic Calcium Phosphates

Properties Values

Elastic modulus (GPa) 40117

Compressive strength (MPa) 294
Bending strength (MPa) 147
Hardness (Vickers, GPa) 3.43
Poisson's ratio 0.27
3
Density (theoretical, g/cm ) 3.16

And other sources.

Polycrystalline hydroxyapatite has a high elastic modulus (40117 GPa). Hard tissues
such as bone, dentin, and dental enamel are natural composites that contain hydroxyapatite (or
a similar mineral) as well as protein, other organic materials, and water. Enamel is the stiffest
hard tissue with an elastic modulus of 74 GPa, and it contains the most mineral. Dentin (E = 21
GPa) and compact bone (E = 12~18 GPa) contain comparatively less mineral. The Poisson's
ratio for the mineral or synthetic hydroxyapatite is about 0.27, which is close to that of bone
( 0.3).
Among the most interesting properties of hydroxyapatite as a biomaterial is its excellent
biocompatibility. Indeed, it appears to form a direct chemical bond with hard tissues. In an
experimental trial, new lamellar cancellous bone was formed around implanted hydroxyapatite
granules in the marrow cavity of rabbits after 4 weeks, as shown in Figure 6-13.
154 CH. 6: CERAMIC IMPLANT MATERIALS

Figure 6-13. X-ray microradiographic picture showing the hydroxyapatite granules and bony
tissues surrounding them after 4 weeks in a rabbit marrow cavity (40q). The mottled areas are
regions of new bone deposition and the white areas are implants. Reprinted with permission
from Niwa et al. (1980). Copyright 1980, Springer-Verlag.

6.4.3. Manufacture of Calcium Phosphates (Hydroxyapatite)

Many different methods have been developed to make precipitates of hydroxyapatite from an
aqueous solution of Ca(NO3)2 and NaH2PO4. One method uses precipitates that are filtered and
dried to form a fine particle powder. After calcination for about 3 hours at 900C to promote
crystallization, the powder is pressed into final form and sintered at about 10501200C for 3
hours. Above 1250C the hydroxyapatite shows a second phase precipitation along the grain
boundaries.

Example 6-4
Calculate the theoretical density of hydroxyapatite crystal [Ca10(PO4)6(OH)2].

Answer
From Figure 6-12 one can see that there are 10 Ca atoms in the hexagonal unit cell prism, 4
inside, (2 for 1/2, 2 for 1/4, 3/4 position), 2 for top and bottom (0 and 1 position), and 4 for
sides (1/4, 3/4 position). Therefore,

(10 q 40  6 q31  26 q16  2 q1)

U
3
9.432 q q 9.432 q6.881q108 q 6.01q10 23
2
 3.16 g/cm3.
(This is very close to the value given in the literature; McConell, 1963).
BIOMATERIALS: AN INTRODUCTION 155

6.5. GLASS-CERAMICS
Glass-ceramics are polycrystalline ceramics made by controlled crystallization of glasses.
They were originally developed by S.D. Stookey of Corning Glass Works in the early 1960s.
They were first utilized in photosensitive glasses in which small amounts of copper, silver and
gold are precipitated by ultraviolet light irradiation. These metallic precipitates help to nucleate
and crystallize the glass into a fine grained ceramic which possess excellent mechanical and
thermal properties. Bioglass and Ceravital are two glass-ceramics developed for implants.

6.5.1. Formation of Glass-Ceramics

The formation of glass-ceramics is influenced by the nucleation and growth of small (<1-Pm
diameter) crystals as well as the size distribution of these crystals. It is estimated that about
1012 to 1015 nuclei per cubic centimeter are required to achieve such small crystals. In addition
to the metallic agents mentioned (Cu, Ag, and Au), Pt groups, TiO2, ZrO2, and P2O5 are widely
used for this purpose. The nucleation of glass is carried out at temperatures much lower than
the melting temperature. During processing the melt viscosity is kept in the range of 1011 and
1012 Poise for 1 to 2 hours. In order to obtain a larger fraction of the microcrystalline phase, the
material is further heated to an appropriate temperature for maximum crystal growth. Defor-
mation of the product, phase transformation within the crystalline phases, and redissolution of
some of the phases are to be avoided. Crystallization is usually more than 90% complete with
grain sizes of 0.1 to 1 Pm. Grains smaller than one micron are called nanocrystalline. These
grains are much smaller than those of the conventional ceramics. Figure 6-14 shows a sche-
matic representation of the temperaturetime cycle for a glass-ceramic.

Figure 614. Temperature-time cycle for a glass ceramic. Reprinted with permission from
Kingery et al. (1976). Copyright 1976, Wiley.

The glass-ceramics developed for implantation are SiO2CaONa2OP2O5 and Li2OZnO

SiO2 systems. There are two different groups experimenting with the SiO2CaONa2OP2O5
glass-ceramic. One group varied the compositions (except for P2O5) as given in Table 6-10 in
order to obtain the best composition to induce direct bonding with bone. The bonding is related
to simultaneous formation of a calcium phosphate and an SiO2-rich film layer on the surface,
as exhibited by 46S5.2 type Bioglass. If an SiO2-rich layer forms first and a calcium phos-
phate film develops later (4655 mol% SiO2 samples) or no phosphate film is formed (60
mol% SiO2), then no direct bonding with bone is observed. The approximate region of the
SiO2CaONa2O system for the tissue-glass ceramic reaction is shown in Figure 6-15. As can
be seen, the best region (Region A) for good tissue bonding is the composition given for
46S5.2 type Bioglass (Table 6-10).
156 CH. 6: CERAMIC IMPLANT MATERIALS

Table 6-10. Compositions of Bioglass and Ceravital Glass-Ceramics

Type Code SiO2 CaO Na2O P2O5 MgO K2O

Bioglass
42S5.6 42.1 29.0 26.3 2.6
(45S5)46S5.2 46.1 26.9 24.4 2.6
49S4.9 49.1 25.3 23.0 2.6
52S4.6 52.1 23.8 21.5 2.6
55S4.3 55.1 22.2 20.1 2.6
60S3.8 60.1 19.6 17.7 2.6

Cervital*
Bioactive 40.050. 30.035.0 5.010.0 10.015.0 2.55.0 0.53.0
**Nonbioactive 30.035.0 25.030.0 3.57.5 7.512.0 1.02.5 0.52.0

*The Ceravital composition is in weight % while the Bioglass compositions are in mol %.
**In addition Al2O3 (5.015.0), TiO2(1.05.0) and Ta2O5 (5.015.0) are added.

Figure 6-15. The SiO2CaONa2O phase diagram. Region A: bonding in 30 days with bone.
Region B: nonbondingtoo low reactivity. Region C: nonbondingtoo high reactivity. Re-
gion D: bonding but does not form glass. Reprinted with permission from Hench and Ethridge
(1982). Copyright 1982, Academic Press.

The composition of Ceravital is similar to the Bioglass in terms of SiO2 content but
differs somewhat in others, as given in Table 6-10. In addition, Al2O3, TiO2, and Ta2O5 are used
for the Ceravital glass-ceramic in order to control the dissolution rate. The mixtures were
melted in a platinum crucible at 1500C for 3 hours and annealed, and then cooled. The nuclea-
BIOMATERIALS: AN INTRODUCTION 157

tion and crystallization temperatures were 680 and 750C, respectively, for 24 hours each.
When the size of crystallites was about 0.4 nm and the crystals did not exhibit the characteris-
tic needle structure, the process was stopped to obtain a fine grain structure.

6.5.2. Properties of Glass-Ceramics

Glass-ceramics have several desirable properties compared to glasses and ceramics. The ther-
mal coefficient of expansion is very low, typically 107 to 105 per degree C, and in some cases
it can be made even negative. Due to the controlled grain size and improved resistance to sur-
face damage, the tensile strength of these materials can be increased by at least a factor of
two, from about 100 to 200 MPa. The resistance to scratching and abrasion are close to that of
sapphire.
In an experimental trial, Bioglass glass-ceramic was implanted in the femur of rats for 6
weeks. Transmission electron micrographs showed intimate contacts between the mineralized
bone and the Bioglass, as given in Figure 6-16. The mechanical strength of the interfacial
bond between bone and Bioglass ceramic is the same order of magnitude as the strength of
2
the bulk glass-ceramic (850 kg/cm or 83.3 MPa), which is about three-fourths that of the host
bone strength.

Figure 6-16. Transmission electron micrograph of well-mineralized bone (b) juxtaposed to the
glass-ceramic (c), which was fractured during sectioning (q51,500). Insert (a) is the diffraction
pattern from the ceramic area and (b) is from bone area. Reprinted with permission from
Beckham et al. (1971). Copyright 1971, Springer-Verlag.
158 CH. 6: CERAMIC IMPLANT MATERIALS

The main drawback of the glass-ceramic is its brittleness, as is the case with other glasses
and ceramics. Additionally, due to restrictions on the composition for biocompatibility (or os-
teogenicity), mechanical strength cannot be substantially improved as for other glass-ceramics.
Therefore, they cannot be used for making major load-bearing implants such as joint implants.
However, they can be used as fillers for bone cement, dental restorative composites, and coat-
ing material.

Example 6-5
From the phase diagram of Al2O3SiO2, answer the following:
a. Determine the exact w/o of Al2O3 for mullite, which has a 3Al2O32Si02 composition.
b. Determine the amount of liquid in 50 w/o Al2O350 w/o SiO2 at 1588C.

Answer
a.

3Al 2 O3 6(27)  9q16


3Al 2 O3 + 2SiO2 6 q 27q 9q16  2 q 28  4 q16

162  144

306  56  64

306

426

 0.718 (71.8%).
b. Using the lever rule [see 5.1.1]:

71.8  50 21.8
%L    0.329 (32.9%).
71.8  5.5 66.3

6.6. OTHER CERAMICS

There are many other ceramic materials studied as well, including titanium oxide (TiO2), bar-
ium titanate (BaTiO3), tricalcium phosphate (Ca3(PO4)2), and calcium aluminate (CaOAl2O3).
Titanium oxide was tried for use in a component of bone cement or as a blood-interfacing ma-
terial. Porous calcium aluminate was used to induce tissue ingrowth into pores with the aim of
achieving better implant fixation. However, this material loses its strength considerably after in
vivo and in vitro aging, as shown in Figure 6-17. Tricalcium phosphate together with calcium
aluminate were tried as biodegradable implants in the hope of regenerating new bone.
Barium titanate with a textured surface has been used in experimental trials to achieve im-
proved fixation of implants to bone. This material is piezoelectric (following a polarization
procedure). Therefore, mechanical loads on the implant will generate electrical signals that are
capable of stimulating bone healing and ingrowth. These loads on the implant arise during use
of the implanted limb. Alternatively, the implant can be exposed to ultrasound to generate
electrical signals.
BIOMATERIALS: AN INTRODUCTION 159

Figure 6-17. Aging effect on the strength of calcium aluminate in vitro and in vivo. Reprinted
with permission from Schnittgrund et al. (1973). Copyright 1973, Wiley.

6.7. CARBONS
Carbons can be made in many allotropic forms: crystalline diamond, graphite, noncrystalline
glassy carbon, and partially crystalline (now referred to as icosahedral) pyrolytic carbon.
Among these, only pyrolytic carbon is widely utilized for implant fabrication; it is normally
used as a surface coating. It is also possible to coat surfaces with diamond-like carbon (DLC).
This technique has the potential to improve performance of such medical devices as surgical
knives, scissors, and articulating surfaces of joint implants; however, it is not, as of this writ-
ing, commercially available. This DLC coating is now used to coat razor blades.

6.7.1. Structure of Carbons

The crystalline structure of carbon as used in implants is similar to the graphite structure
shown in Figure 6-18. The planar hexagonal arrays are formed by strong covalent bonds in
which one valence electron per atom is free to move, resulting in high but anisotropic electric
conductivity. The bonding between layers is stronger than the van der Waals force; therefore,
crosslinks between them are considered to be present. Indeed, the remarkable lubricating prop-
erty of graphite cannot be realized unless the crosslinks are eliminated.
The poorly crystalline carbons are thought to contain unassociated or unoriented carbon
atoms. The hexagonal layers are not perfectly arranged, as shown in Figure 6-19. The strong
bonding within layers and the weaker bonding between layers cause the properties of individ-
ual crystallites to be highly anisotropic. However, if the crystallites are randomly dispersed,
then the aggregate becomes isotropic.
160 CH. 6: CERAMIC IMPLANT MATERIALS

Figure 6-18. Crystal structure of graphite. Reprinted with permission from Shobert (1964).
Copyright 1964, Academic Press.

Figure 6-19. Schematic representation of poorly crystalline carbon: (a) single-layer plane; (b)
parallel layers in a crystalline; (c) unassociated carbon; (d) an aggregate of crystallites, single
layers and unassociated carbon. Reprinted with permission from Bokros (1972). Copyright
1972, Marcel Dekker.

6.7.2. Properties of Carbon

The mechanical properties of carbon, especially pyrolytic carbon, are largely dependent on
density, as shown in Figures 6-20 and 6-21. The increased mechanical properties are directly
related to the increased density, which indicates the properties depend mainly on the aggregate
structure of the material.
Graphite and glassy carbon have much lower mechanical strength than pyrolytic carbon,
as given in Table 6-11. However, the average modulus of elasticity is almost the same for all
carbons. The strength of pyrolytic carbon is quite high compared to graphite and glassy carbon.
This is again due to the lesser amount of flaws and unassociated carbons in the aggregate.
A composite carbon that is reinforced with carbon fiber has been considered for implants.
The properties are highly anisotropic, as given in Table 6-12. The density is in the range of
3
1.41.45 g/cm , with a porosity of 3538%.
Carbons exhibit excellent compatibility with tissues. In particular, compatibility with
blood has made pyrolytic carbon deposited heart valves and blood vessel walls a widely ac-
cepted part of the surgical armamentarium.
BIOMATERIALS: AN INTRODUCTION 161

Figure 6-20. Fracture stress versus density for unalloyed LTI pyrolite carbons. Reprinted with
permission from Kaae (1971). Copyright 1971, Elsevier Science.

Figure 6-21. Elastic moduli versus density for unalloyed LTI pyrolite carbons. Reprinted with
permission from Kaae (1971). Copyright 1971, Elsevier Science.

6.7.3. Manufacture of Implants

Pyrolytic carbons can be deposited onto finished implants from hydrocarbon gas in a fluidized
bed at a controlled temperature and pressure, as shown in Figure 6-22. The anisotropy, density,
crystallite size, and structure of the deposited carbon can be controlled by temperature, compo-
sition of the fluidizing gas, bed geometry, and residence time (velocity) of the gas molecules in
the bed. The microstructure of deposited carbon should be particularly controlled since the
formation of growth features associated with uneven crystallization can result in a weaker ma-
terial, as shown in Figure 6-23. It is also possible to introduce various other elements into the
fluidizing gas and codeposit them with carbon. Usually silicon (1020 w/o) is codeposited (or
162 CH. 6: CERAMIC IMPLANT MATERIALS

alloyed) to increase hardness for applications requiring resistance to abrasion, such as heart
valve discs.

Table 6-11. Properties of Various Types of Carbon

Types of carbon

Properties Graphite Glassy Pyrolytica

3
Density (g/cm ) 1.5-1.9 1.5 1.5-2.0
Elastic modulus (GPa) 24 24 28
a
Compressive strength (MPa) 138 172 517 (575 )
3 b
Toughness (mN/cm ) 6.3 0.6 4.8
a
1.0 w/o Si-alloyed pyrolytic carbon, Pyrolite (Carbomedics, Austin, TX)
b 3 3 3
1 m-N/cm = 1.45 q 10 in-lb/in .

Table 6-12. Mechanical Properties of Carbon Fiber-Reinforced Carbon

Fiber lay-up
Property Unidirectional 090 Crossply

Flexural modulus (GPa)

Longitudinal 140 60
Transverse 7 60
Flexural strength (MPa)
Longitudinal 1,200 500
Transverse 15 500
Interlaminar shear strength (MPa) 18 18

Reprinted with permission from Adams and Williams (1978). Copyright 1978, Wiley.

Pyrolytic carbon was deposited onto the surfaces of blood vessel implants made of poly-
mers. This is called ultra-low-temperature isotropic (ULTI) carbon, instead of LTI (low-
temperature-isotropic) carbon. The deposited carbon is thin enough not to interfere with the
flexibility of grafts yet exhibits excellent blood compatibility.
The vitreous or glassy carbon is made by controlled pyrolysis of polymers such as phe-
nolformaldehyde, rayon (Glasser et al., 1992), and polyacrylonitrile at high temperature in a
controlled environment. This process is particularly useful for making carbon fibers and tex-
tiles, which can themselves be used or as components of composites.

6.8. DETERIORATION OF CERAMICS

It is of great interest to know whether inert ceramics such as alumina undergo significant static
or dynamic fatigue. In one study it was shown that above a critical stress level the fatigue
strength of alumina is reduced by the presence of water. This is due to delayed crack growth,
which is accelerated by the water molecules. However, another study showed that a reduction
in strength occurred if evidence of penetration by water was observed under a scanning elec-
tron microscope (SEM). No decrease in strength was observed for samples that showed no
watermarks on the fractured surface, as shown in Figure 6-24. It was suggested that the pres-
ence of a minor amount of silica in one sample lot may have contributed to permeation of the
BIOMATERIALS: AN INTRODUCTION 163

Figure 6-22. Schematic diagram showing particles being coated with carbon in a fluidized bed.
Reprinted with permission from Bokros (1972). Copyright 1972, Marcel Dekker.

Figure 6-23. Microstructures of carbons deposited in a fluidized bed: (a) granular carbon with
distinct growth features; (b) isotropic carbon without growth features. Both under polarized
light, 240q. Reprinted with permission from Bokros (1972). Copyright 1972, Marcel Dek-
ker.

water molecules, which is detrimental to strength. It is not clear whether the same static fatigue
mechanism operates in single-crystal alumina or not. It is, however, reasonable to assume that
the same static fatigue will occur if the ceramic contains flaws or impurities, which will act as
the source of crack initiation and growth under stress.
A study of the fatigue behavior of vapor-deposited pyrolytic carbon fibers (400500 nm
thick) onto a stainless steel substrate showed that the film did not break unless the substrate
164 CH. 6: CERAMIC IMPLANT MATERIALS

underwent plastic deformation at 1.3 u 102 strain and up to one million cycles of loading.
Therefore, the fatigue is closely related to the substrate, as shown in Figure 6-25. A similar
substratecarbon adherence is the basis for the pyrolytic carbon-deposited polymer arterial
grafts, as mentioned earlier.

Figure 6-24. Flexural strength of dense alumina rods after aging under stress in Ringer's solu-
tion base indicating standard deviation. Lots 1 and 2 are from different batches of production.
Reprinted with permission from Krainess and Knapp (1978). Copyright 1978, Wiley.

Figure 6-25. Strain versus number of cycles to failure: F = absence of fatigue cracks in carbon
film; M = fracture of carbon film due to fatigue failure of substrates; Q = data for substrate de-
termined in single-cycle tensile test. Reprinted with permission from Shim and Haubold (1980).
Copyright 1980, Marcel Dekker.

The fatigue life of ceramics can be predicted by assuming that fatigue fracture is due to
the slow growth of preexisting flaws. Generally, the strength distribution (si) of ceramics in an
inert atmosphere can be correlated with the probability of failure F, by the following equation;

1 s
ln ln  m ln i , (6-3)

1  F s
0
BIOMATERIALS: AN INTRODUCTION 165

in which m and s0 are constants. The m is called the Weibull modulus, which indicates the dis-
tribution of fracture strength: the higher the value, the narrower the distribution. The metals
and polymers have values of 50, while most ceramics and glasses have values less than 20.
Figure 6-26 shows a good fit for Bioglass-coated alumina.

Figure 6-26. Plot of ln [1/(1 F)] versus ln s for Bioglass-coated alumina in a tris hy-
droxyaminomethane buffer and liquid nitrogen. F is the probability of failure and V is strength.
Reprinted with permission from Ritter et al. (1979). Copyright 1979, Wiley.

A minimum service life (tmin) of a specimen can be predicted by means of a proof test
wherein it is subjected to stresses greater than those expected in service. Proof tests also elimi-
nate weaker pieces. This minimum life can be predicted from the following equation:

tmin  BV Np 2 Va N , (6-4)

in which VP is the proof test stress, Va is the applied stress, and B and N are constants. Rear-
ranging Eq. (6-3), we obtain
2
V p
2
tmin V a  B . (6-5)
V
a

Figure 6-27 shows a plot of Eq. (6-5) for alumina on a logarithmic scale.

Example 6-6
Calculate the proof stress of an alumina sample if it is to last for 20 years at 100 MPa in air and
in Ringer's solution.
166 CH. 6: CERAMIC IMPLANT MATERIALS

Answer

tmin V2a  20 yr q3.15q10 7 s/yr q(100 MPa)2 ,

log tmin V2a  12.8.

This value is comparable to the dotted line for 80 years at 69 MPa in Figure 6-27. Therefore,
Vp/Va = 2.0 in air and 2.55 in Ringer's solution. So Vp = 200 MPa in air, and Vp = 255 MPa in
Ringer's solution. As one might expect, it is necessary to test the sample more rigorously in
Ringer's solution than in air.

Figure 6-27. Plot of Eq. (6-45) for alumina after proof testing. N = 43.85, m = 13.21, and V0 =
55,728 psi. Reprinted with permission from Ritter et al. (1979). Copyright 1979, Wiley.

PROBLEMS
6-1. A ceramic is used to fabricate a hip joint. Assume a simple ball-and-socket configura-
tion with a surface contact area of 1.0 cm2 and continuous static loading in a simu-
lated condition similar to Figure 6-17 (extrapolate the data if necessary). Observe that
the actual contact area is smaller than the total surface area.
a. How long will it last if the loading is a force due to 70 kg (mass), in water and in
blood?
b. Will the implant last a longer or shorter time with dynamic loading? Give reasons.
6-2. From the phase diagram of Al2O3SiO2,
a. Determine the exact mole % of Al2O3SiO2 for mullite, which has a 3Al2O32SiO2
composition.
b. Determine the amount of mullite in 50 wt % Al2O350 wt% SiO2 at 1469C.
BIOMATERIALS: AN INTRODUCTION 167

6-3. Calculate the proof stress of an alumina sample if it is to last for 40 years at 100 MPa
in air and in Ringers solution.
6-4. One is trying to coat the surface of an orthopedic implant with hydroxyapatite to en-
hance compatibility with tissues. List the possible problems associated with this tech-
nique. Give two methods of applying the coating.
6-5. Discuss the advantages of coating the femoral stem of a hip prosthesis with hy-
droxyapatite, alumina, and diamond-like carbon. Give two general advantages to all
ceramic coatings and, in addition, one or more specific advantage(s) for each coating.
Show them in table form.
6-6. Zirconia (ZrO2) is often stabilized with calcium to provide an important refractory.
The basic cell is ZrO2 with 1 Ca2+ ion present for every 10 Zr4+ ions. Will the vacant
sites be anion or cation? What percentage of the total number of all sites will be va-
cant?
6-7. Calculate the density of monoclinic and tetragonal zirconia.
6-8. Calculate the amount of volume change when the zirconia changes from monoclinic
to tetragonal structure.
6-9. Calculate the amount of volume change by adding 3 mole% of Y2O3 into cubic zirco-
nia. Make assumptions of ideal mixing.
6-10. Consider the following mechanical properties.
Properties Bone Vitreous carbon
Ultimate tensile strength (MPa) 100 120
Tensile modulus (GPa) 1218 2.8
This comparison suggests that vitreous carbon would be an excellent material for
bone replacement. What is wrong with this idea?
6-11. Discuss the advantages of coating the femoral stem of a hip prosthesis with hy-
droxyapatite, alumina, and diamond-like carbon. Give two general advantages to all
ceramic coatings and, in addition, one or more specific advantage(s) for each coating.
Show them in a table form.
6-12. CsCl has a simple cubic structure and the following data are given:
168 CH. 6: CERAMIC IMPLANT MATERIALS

Element Ion radius (nm) amu

Cl 0.181 35.4
Cs 0.165 132.9
a. Calculate the minimum radius ratio of a simple cubic cell.
b. What is the coordination number of the Cs?
c. What is the density of CsCl?
d. Why is this material brittle?
6-13. NaCl has a face-centered cubic structure and the following data are given:
Element Ion radius () amu
Cl 1.81 35.4
Na 0.97 23
a. Draw the positions of Cl and Na ions in one of the cubic faces.
b. What is the coordination number of the Na?
c. Calculate the lattice parameter (a) of the unit cell.
d. Calculate the density of NaCl.
f. Which direction will have the highest linear density of ions: [100], [110], or [111]?
e. Why is this material brittle?
6-14. From the list of ceramics choose the most appropriate one:
A. Al2O3 B. Hydroxyapatite C. Tricalcium phosphate
D. Carbon E. Glass-ceramic
a. Resorbable in vivo.
b. Has best blood compatibility
c. A large single crystal can be made and sometimes is called ruby or sapphire
d. Has a direct bone-bonding ability
e. Bone mineral has the similar structure
6-15. From the following ceramic materials select the most appropriate one for the ques-
tions:
A. Al2O3 B. Ca3(PO4)2 C. Ca10(PO4)6(OH)2 D. 45S5
E. LTI F. ZrO2 G. DLC
a. Osteogenic properties.
b. Used to make dental implants.
c. Hardest ceramic among those listed.
 d. E-whitlockite forms in dry condition.
e. Readily resorbed in vivo.
f. AmXn structure and 2/3 available lattice sites are filled.
g. Fluidized bed is used to coat the surface of heart valve discs.
h. Glass ceramic and has capacity to bond directly with bone.
BIOMATERIALS: AN INTRODUCTION 169

i. Similar to the diamond, coated on a surface.

j. Single crystal is used for making jewels.
6-16. Collagen and TGF-E (tissue growth factor) are mixed with hydroxyapatite (Ha et al.,
1993) for making synthetic bone [BioME] implants. A 10-mm diameter 20-mm
long BioME rod is made using 49492% by weight [2% is TGF-E] of the materials
to test its properties. Using the following data, answer the following questions:
Young's
Material modulus (GPa) Strength (MPa) Density (g/cc)
Collagen 1 10 (yield) 1.0
HA 100 100 (fracture) 3.2
TGF-E 0 0 1.0
a. Calculate the density of the BioME.
b. Calculate the yield strain of the BioME.
c. Calculate the Young's modulus of the BioME.
d. Calculate the maximum strength of the BioME.
e. Give at least one reason why you would not use TGF-E to make the BioME.

SYMBOLS/DEFINITIONS
Words
Activation energy of phase transformation: Thermal energy required to overcome an energy
barrier. Eactivation = RT ln(transformation rate), where R is the gas constant and T is tempera-
ture (K).
Alumina: Aluminum oxide (Al2O3), which is very hard (Mohs hardness of 9) and strong. Sin-
gle crystals are called sapphire or ruby depending on color. Alumina is used to fabricate
hip joint socket components or dental root implants.
Baddeleyite: Mineral containing zircon.
Biolox: Trade name of alumina ceramic.
Calcium phosphate: A family of calcium phosphate ceramics including hydroxyapatite E-
whitlockite, mono-, di-, tri-, and tetra-calcium phosphate, which are used to make substi-
tute or augment artificial bone substitutes.
Cubic zirconia: Partially stabilized zirconia in cubic structure to prevent fracture during cool-
ing.
Electronegativity: Potential of an atom to attract electrons, especially in the context of form-
ing a chemical bond.
Fluorite (CaF2) structure: AX2 structure of ceramic, where A is a metal and X a nonmetal.
Glass-ceramics: A glass crystallized by heat treatment. Some of those have the ability of
forming chemical bonds with hard and soft tissues. Bioglass and Ceravital are well-
known examples.
Hydroxyapatite [Ca10(PO4)6(OH)2]: A calcium phosphate ceramic with a calcium-to-
phosphorus ratio of 5/3 and nominal composition. It has good mechanical properties and
excellent biocompatibility. Hydroxyapatite is the mineral constituent of bone.
170 CH. 6: CERAMIC IMPLANT MATERIALS

LTI carbon: A silicon alloyed pyrolytic carbon deposited onto a substrate at low temperature
with isotropic crystal morphology. Highly blood compatible and used for cardiovascular
implant fabrication, such as artificial heart valves.
Maximum radius ratio: The ratio of atomic radii computed by assuming the largest atom or
ion which can be placed in a crystal's unit cell structure without deforming the structure.
Mohs scale: A hardness scale in which 10 (diamond) is the highest and 1(talc) is the softest.
Monoclinic structure: One crystal system having a z b z c, D= J= 90
Partially stabilized zirconia: See cubic zirconia.
Phase transformation toughening process: By transforming phases under stress the material
becomes stronger due to the volume expansion of the transformed phase.
Tetragonal structure: One crystal system having a = b z c, D E J 90.
Wear factor: Similar to the wear constant; wear volume generated by given load and sliding
distance .
Weibull modulus: Slope of the Weibull plot. Larger values indicate predictable failure stress.
Ceramics are 10~20, metals are >50.
Weibull plot: Plot of the logarithm of probability of failure (F) [actually, ln ln 1/(1 F)] and
test stress/strength fraction.
E-whitlockite (3CaOP2O3): One of the calcium phosphate compounds, similar to tricalcium
phosphate, 3CaPO4.
Yttrium oxide (Y2O3): Oxide used to partially stabilize zirconia.
Zircon (ZrSiO4): Crystalline mineral, a silicate of zirconium with tetragonal structure.
Zirconia (ZrO2): Zirconium oxide, which is hard and strong.

BIBLIOGRAPHY
Adams D, Williams DF. 1978. Carbon fiber-reinforced carbon as a potential implant material. J Biomed
Mater Res 12:3542.
ASTM. 2000. Annual book of ASTM standards. West Conshohocken, PA: ASTM.
Beckham CA, Greenlee Jr TK, Crebo AR. 1971. Bone formation at a ceramic implant interfacing. Calcif
Tissue Res 8:165171.
Blencke BA, Bromer H, Deutscher KK. 1978. Compatibility and long-term stability of glass-ceramic
implants. J Biomed Mater Res 12(3):307316.
Bokros JC, LaGrange LD, Schoen GJ. 1972. Control of structure of carbon for use in bioengineering. In
Chemistry and physics of carbon, pp. 103171. Ed PL Walker. New York: Dekker.
Bokros JC, Atkins RJ, Shim HS, Haubold AD, Agarwal MK. 1976. Carbon in prosthetic devices. In Pe-
troleum derived carbons, pp. 237265. Ed ML Deviney, TM O'Grady. Washington, DC: American
Chemical Society.
Burger W, Willmann G. 1993. Advantages and risks of zirconia ceramics in biomedical applications. In
Bioceramics, Vol. 6, pp. 299-304. Ed P Ducheyne, D Christiansen. Oxford: Pergamon.
Callister JWD. 1994. Materials science and engineering: an introduction, 3rd ed. New York: Wiley.
Drennan J, Steele BCH. 1986. Zirconia and hafnia. In Encyclopedia of materials science and engineer-
ing, pp. 55425545. Ed MB Beaver. Oxford: Pergamon/MIT Press.
Gill RM. 1972. Carbon fibres in composite materials. London: Butterworths.
Gilman JJ. 1967. The nature of ceramics. Sci Am 218(3):113.
Gitzen WH, ed. 1970. Alumina as a ceramic material. Westerville, OH: American Ceramic Society.
Glasser WG, Hatakeyama H, eds. 1992. Viscoelasticity of biomaterials. Washington, DC: American
Chemical Society.
BIOMATERIALS: AN INTRODUCTION 171

Ha S-W, Mayer J, Wintermantel E. 1993. Micro-mechanical testing of hydroxyapatite coatings on carbon

fiber reinforced thermoplastics. In Bioceramics, Vol. 6, 489493. Ed P Ducheyne, D Christiansen.
Oxford: Pergamon.
Hastings GW, Williams, DF, eds. 1980. Mechanical properties of biomaterials, Part 3. Chichester:
Wiley.
Hench LL, Ethridge EC. 1982. Biomaterials: an interfacial approach. New York: Academic Press.
Horwitz BR, Rockowitz NL, Goll SR, Booth Jr RE, Balderston RA, Rothman RH, Cohn JC. 1993. A
prospective randomized comparison of two surgical approaches to total hip arthroplasty. Clin Or-
thop Relat Res 291:154163.
Hulbert SF, Young FA, eds. 1978. Use of ceramics in surgical implants. New York: Gordon & Breach.
Hulbert SF, King FW Jr, Klawitter JJ. 1972. Initial surface interaction of blood and blood components
with Al2O3 and TiO2. J Biomed Mater Res Symp 2:6989
Kaae JL. 1971. Structure and mechanical properties of isotropic pyrolytic carbon deposited below 1600
degrees. J Nucl Mater 38:4250.
Kawahara H, Hirabayashi M, Shikita T. 1980. Single crystal alumina for dental implants and bone
screws. J Biomed Mater Res 14:597606.
Kingery WD, Bowen HK, Uhlmann DR. 1976. Introduction to ceramics, 2nd ed. New York: Wiley.
Krainess FE, Knapp WJ. 1978. Strength of a dense alumina ceramic after aging in vitro. J Biomed Mater
Res 12:241246.
Kumar P, Shimizu K, Oka M. 1991. Low wear rate of UHMWPE against zirconia ceramic (Y-PSZ) in
comparison to alumina ceramic and SUS 316L alloy. J Biomed Mater Res 25:813828.
McConell D. 1973. Apatite: its crystal chemistry, mineralogy, utilization, and biologic occurrence. Ber-
lin: Springer-Verlag.
McMillan PW. 1979. Glass-ceramics. New York: Academic Press.
Niwa S, Sawai K, Takahashie S, Tagai H, Ono M, Fukuda Y. 1980. Experimental studies on the implan-
tation of hydroxyapatite in the medullary canal of rabbits. In Proceedings of the first world biomate-
rials congress (Baden, Austria). Berlin: Springer-Verlag.
Norton F. 1974. Elements of ceramics. Reading, MA: Addison-Wesley.
Ogino M, Ohuchi F, Hench LL. 1980. Compositional dependence of the formation of calcium phosphate
film on Bioglass. J Biomed Mater Res 14:5564.
Posner AS, Perloff A, Diorio AD. 1958. Refinement of the hydroxyapatite structure. Acta Crystallogr
11:308309.
Ritter Jr JE, Greenspan DC, Palmer RA, Hench LL. 1979. Use of fracture of an alumina and Bioglass-
coated alumina. J Biomed Mater Res 13:251263.
Schnittgrund GS, Kenner GH, Brown SD. 1973. In vivo and in vitro changes in strength of orthopedic
calcium aluminate. J Biomed Mater Res Symp 4:435452.
Shim HS, Haubold AD. 1980. The fatigue behavior of vapor deposited carbon films. Biomater Med De-
vices Artif Organs 8:333344.
Shimizu K, Oka M, Kumar P, Kotoura Y, Yamamuro T, Makinouchi K, Nakamura T. 1993. Time-
dependent changes in the mechanical properties of zirconia ceramic. J Biomed Mater Res 27:729
734.
Shobert EI. 1964. Carbon and graphite. New York: Academic Press.
Starfield MJ, Shrager AM. 1972. Introductory to materials science. New York: McGraw-Hill.
Willmann G. 1993. Zirconia: a medical-grade material? In Bioceramics, Vol. 6. pp. 271276. Ed P
Ducheyne, D Christiansen. Oxford: Pergamon.
Yoshimura M, Suda H, eds. 1994. Hydrothermal processing of hydroxyapatite: past, present, and future.
Hydroxyapatite and related materials. Boca Raton, FL: CRC Press.