The effect of a silane coupling agent on the bond strength

of bone cement and cobalt–chrome alloy

Scott A. Yerby,1,2 Adam F. Paal,3 Phillip M. Young,4 Gary S. Beaupré,1,2,4 Kevin L. Ohashi,4
Stuart B. Goodman2
1
Rehabilitation Research and Development Center, Veterans Affairs, Palo Alto Health Care System,
Palo Alto, California 94304
2
Department of Functional Restoration, Division of Orthopaedic Surgery, Stanford University,
Stanford, California 94305
3
Department of Orthopaedics, Health Sciences Center, T18, SUNY Stony Brook, Stony Brook, New York 11794
4
Mechanical Engineering Department, Biomechanical Engineering Division, Stanford University,
Stanford, California 94305

Received 3 November 1998; accepted 16 July 1999

Abstract: Debonding of the cement–implant interface has
been hypothesized to be the leading initial indicator of failed
total hip prostheses. Many attempts have been made to in-
crease the bond strength of this interface by precoating
the implant, increasing the implant’s surface roughness,
and creating macro-grooves or channels on the implant.
However, each of these approaches introduces new compli-
cations. This study introduces a unique silane coupling
agent used to chemically bond the bone cement to the im-
plant. Cylindrical cobalt–chrome samples were treated with
the silane coupling agent, bonded to polymethylmethacry-
late, and pushed out to failure. The mean shear strengths
were compared to the failure strengths of untreated
samples. Half of the specimens were tested immediately fol-
lowing cement curing, and the other half were tested after
immersion in saline solution for 60 days. The mean shear
strength of the silane-coated samples ranged from 18.2 to
24.1 MPa, and the mean shear strength of the uncoated
samples ranged from 7.6 to 15.0 MPa. The increase in
strength following silane coating noted in this study may
increase the longevity of the implant by decreasing debond-
ing at the interface and, therefore, subsequent failure due to
loosening. © 2000 John Wiley & Sons, Inc. J Biomed Mater
Res, 49, 127–133, 2000.
Key words: polymethylmethacrylate; silane; cobalt–chrome;
interface; shear strength

INTRODUCTION cations such as excess wear debris, loosening, and ce-

Debonding of the implant–cement interface has
been cited as the initial event of loosening in cemented
hip prostheses. Jasty et al. examined sixteen post-
mortem femora with total hip arthroplasties and de-
scribed debonding at the implant–polymethylmethac-
rylate (PMMA) interface to be the most common early
feature.1 In all cases the PMMA–bone interface was
intact. Early debonding at the cement–metal interface
can lead to progressive interfacial motion and compli-
Correspondence to: S. Yerby; e-mail: yerby@bones.
stanford.edu
Contract grant sponsor: Rehabilitation Research and De-
velopment Center of the Department of Veterans Affairs
Contract grant sponsor: Kulzer GmbH, Wehrheim, Ger-
many
© 2000 John Wiley & Sons, Inc. CCC 0021-9304/00/010127-07
ment fracture.
Although the loosening rates have declined dra-
matically with the advent of improved cementing
techniques and implant designs, failure due to
debonding still is of primary concern.1–6 The implant–
cement interface is inherently weak in tension and
shear since PMMA adheres poorly to metals. There-
fore it is believed that improving the bond strength of
the implant–PMMA interface will increase the longev-
ity of the implant and improve the patient’s quality of
life.
Previous investigations of the shear strength of the
interface between the PMMA and the implant have
concentrated on variations of the: (1) implant materi-
als,7–11 (2) implant surface treatment,11–15 (3) im-
mersion environment and immersion duration,8,9,11–16
and (4) bonding agent or precoat used to enhance
the bond strength.7,9,10,12–15,17 Implant materials com-
monly tested include cobalt–chrome, stainless steel,
128 YERBY ET AL.

and titanium alloys. Common testing environments

include ambient air, saline immersion at 37°C, and in
vivo animal models. The precoating of implants with
a thin layer of meticulously applied PMMA is done
in an attempt to improve the interfacial bond strength
between the surgically prepared PMMA and the
precoated implant. Other methods include applying
a silane coupling agent to the implant surface or
adding 4-methacryloyloxyethyl trimellitate anhydride
(4-META) to the PMMA mixture during prepara-
tion.18,19
This study investigates the influence of a new silane
bonding agent on the shear strength of the implant–
PMMA interface. We hypothesized that (1) the addi-
tion of a silane coating will significantly increase the
strength of the implant–cement interface and (2) im-
mersion in saline solution for 60 days will lead to a
significant reduction in the interface strength of un-
coated specimens, but it will not lead to a reduction in
the interface strength of silane-coated specimens.

MATERIALS AND METHODS

Cobalt–chrome (CoCr) samples were grouped into 60 si-

lane-coated and 60 uncoated samples. These groups were
equally sub-divided into samples immersed in saline solu-
tion for 60 days (saline) and those tested immediately after
preparation (dry). Finally, each sub-group included three
available surgical bone cements. This resulted in 10 samples
of each treatment.
Cylindrical specimens measuring 38.1 mm in length and
6.35 mm in diameter were grit blasted using 250 micron
aluminum oxide (AlO2) grit sprayed at a pressure of 0.4 MPa
from a regulated grit blaster (Basic Mobil, Model 2914-3250,
Renfert GmbH, Hilzingen, Germany). The resulting average
surface roughness (Ra) following the grit blasting was 1.10
microns, as measured by a surface profilimeter (Alpha Step,
Tencor Inst., Mountain View, California). This surface
roughness corresponds to a matte finish, as found on a num-
ber of commercial femoral hip prostheses.20 After the CoCr
specimens were grit blasted, each was blown clean with
pressurized air in an attempt to remove all of the AlO2 resi-
due. The specimens then were placed in a Teflon mold,
which positioned the CoCr specimen concentric and axial to
a phenolic tube measuring 12.7 mm in height, 25.4 mm in
outer diameter (OD), and 19.1 mm in inner diameter (ID).
After placement, the space between the phenolic ID and the
CoCr specimen was filled with curing PMMA (Fig. 1). The
PMMA was one of three formulations that included Sim-
plex-P (Howmedica, Rutherford, New Jersey), Palacos E
(Kulzer GmbH, Wehrheim, Germany), and Palacos R (Kul-
zer GmbH, Wehrheim, Germany). Simplex-P and Palacos R
are standard viscosity cements, and Palacos E is a low vis-
cosity cement. Each cement was mixed in a vacuum mixer at
50 kPa below atmospheric pressure and according to the
manufacturer’s specifications. The monomer was chilled to
0°C in order to increase the working time.3,21–23 After the
PMMA was placed, a conforming aluminum plate was
Figure 1. A cross-sectional schematic of the test sample
and loading scenario. The CoCr rods were 38.1 mm in length
and 6.35 mm in diameter. The phenolic used to constrain the
curing PMMA was 12.7 mm in height, 25.4 mm in outer
diameter (OD), and 19.1 mm in inner diameter (ID).
placed atop each polymerizing specimen and pressed down-
ward to induce pressure on the curing cement and reduce
the number of voids incurred during PMMA application.
Prior to cement introduction, each CoCr specimen in the
silane group was treated with an application of a prelimi-
nary coating of Siloc Pre (Kulzer GmbH, Wehrheim, Ger-
many) followed by heating to 340°C. The specimens were
allowed to cool to room temperature and then were treated
with an application of Siloc Bond (Kulzer GmbH, We-
hrheim, Germany), which was allowed to cure for 5 min in
air. The specimens then were placed in the Teflon mold,
co-axial with the phenolic tubes, and PMMA was applied as
described above.
Once the PMMA was cured, half of the specimens were
immersed in a 37°C saline bath for 60 days and then tested
in shear to failure. The other half were tested to failure 24 h
after preparation.
The testing consisted of applying an axial load to the CoCr
and constraining the PMMA and phenolic (Fig. 1). All of the
specimens were displaced at a rate of 0.01651 mm/s, which
is equivalent to an imposed interfacial shear strain rate of
0.0013/s8,24. Failure of the CoCr–cement interface was de-
noted by a dramatic decrease in load followed by a relatively
constant low load. The interfacial shear strength values were
calculated by dividing the measured failure load by the al-
loy/cement contact surface area.
SILANE BONDING OF CEMENT–IMPLANT INTERFACE 129

A three-factor analysis of variance with a level of signifi-

18.2 (1.7)
Silane
cance of 0.05 was applied to compare mean strength values.
Cement type, coating, and immersion environment were the

Palacos R
factors. A Scheffe’s post hoc test was used to determine sig-
nificant differences.

Uncoated

8.5 (0.8)
RESULTS
Mean Failure Strengths of the Test Samples For the Two Environments, Three Cements, and Two Surface Treatments

The mean shear strength of the silane-coated

20.5 (3.8) samples ranged from 18.2 to 24.1 MPa, and the mean
Silane
Saline-Immersed

shear strength of the uncoated samples ranged from

Palacos E

7.6 to 15.0 MPa (Table I). The mean shear strength of

the silane-coated samples was 1.6 to 2.7 times the un-
coated samples (Fig. 2). There was no significant dif-
Uncoated

7.56 (1.4)

ference between the mean shear strength of the dry

samples (Fig. 2) and the saline-immersed samples (Fig.
3) for any given cement and coating except when com-
paring the uncoated Simplex-P samples. Also, for a
given immersion environment and surface coating, no
21.6 (3.4)
Silane

cement produced a significantly different mean shear

strength than any other cement except when compar-
Simplex P

ing the uncoated, dry Simplex-P to both Palacos R and

Palacos E (Fig. 2).
Uncoated

10.3 (0.6)

DISCUSSION
TABLE I

This study demonstrated an immediate improve-

21.7 (2.6)
Silane

ment of the mean interfacial bond strength following

Palacos R
Uncoated

9.7 (0.9)
22.7 (2.0)
Silane
Palacos E
Dry

Uncoated

9.5 (0.6)

Figure 2. A bar graph representing the mean shear

24.1 (2.2)
Silane

strength of the CoCr samples not immersed in saline solu-

tion. The two groups depicted are the silane-coated and the
Mean (STDEV)

uncoated samples. For each cement, the mean failure

Simplex P

strength of the silane-treated samples was significantly

greater than that of the untreated sample. In addition, the
mean failure strength of the uncoated Simplex-P samples
Uncoated

15.0 (1.7)

was significantly greater than the mean failure strength of

either of the Palacos cements. The error bars represent ±1
standard deviation.
130
Figure 3. A bar graph representing the mean shear
strength of the CoCr samples immersed in saline solution.
The two graphs depicted are the silane-coated and the un-
coated samples. For each cement, the mean failure strength
of the silane-treated samples was significantly greater than
that of the untreated sample. The error bars represent ±1
standard deviation.
application of a silane bonding agent. However, the
expected degradation of shear strength of the un-
coated samples following immersion in saline solution
was noted only for the Simplex-P cement. The increase
in failure strength as a result of silane bonding is not
surprising and follows a similar trend reported by oth-
ers.9,13,17 In a dental application, May et al. reported a
48% (16.1 to 23.8 MPa) increase in the push-out
strength of silane-coated CoCr bonded to PMMA com-
pared to uncoated samples.17 Earlier, they reported a
69% (15.7 to 26.3 MPa) increase in push-out strength in
grit-blasted specimens with similar surface treat-
ments.13 Similarly, Raab et al., tested silane-coated
CoCr samples in shear to failure using a push-out
model.9 Samples were tested 24 h after preparation
and after 30 and 60 days of saline solution immersion
at 37°C. There was no significant difference in mean
push-out strength among any of the treatments
(range: 24.8–25.7 MPa), yet each was significantly
greater than the mean strength of uncoated specimens
(6.9 MPa).
Precoating with PMMA is another commonly used
method for enhancing the bond strength between the
implant material and the bone cement. Price et al.
demonstrated a 124% (3.8 to 8.5 MPa) increase be-
tween uncoated and precoated samples; however, the
mean failure strengths were quite low.15 Also, they
showed no significant degradation in mean failure
strength between PMMA precoated samples tested
immediately after preparation (8.5 MPa) and those
tested after 30 days of saline immersion (8.4 MPa) and
60 days of saline immersion (8.3 MPa). As demon-
YERBY ET AL.
strated in previous studies by Raab et al. and Price et
al., and in the present study, saline immersion does
not seem to degrade the properties of silane- or
PMMA-precoated implant samples.
The lack of interfacial strength degradation follow-
ing saline immersion of the uncoated samples in the
present study currently is unexplained for the Palacos
cements. One possible explanation is that the curing
cement was injected between the phenolic mold and
the implant and was pressurized following injection.
This may have allowed the curing cement to fill the
interstices of the grit-blasted implant and, with the
additional pressure, to form a much better fit/bond
than if the implant had been inserted into the curing
cement. By inserting the implant into the cement, the
potential for large voids at the bone cement/implant
interface is high, especially if the cement is not pres-
surized while it cures. This lack of degradation also
may be due to the difference in fluidity between the
Simplex and Palacos cements. The Palacos cements
may allow for a closer micro-approximation between
the cement and the implant and restrict microleakage
at the interface. However, any theory is difficult to
defend without thoroughly examining the chemical
and mechanical interactions and differences among
the three cement formulations since the Simplex-P be-
haved as expected and the interfacial shear strength of
the two Palacos formulations did not degrade with
immersion.
When comparing push-out results, one must be
aware that the measured failure strength is sensitive to
loading and boundary conditions. An analytical study
by Dhert et al. demonstrated that push-out tests are
very sensitive to the implant-support distance.25
Wang et al. reported on the push-out strength of the
bone–PMMA interface and concluded that this test is
sensitive to the diameter of the pin and the thickness
of the substrate.26 However, a substantial amount of
the variability between results may have been due to
unpredicted porosity and resulting contact surface. Fi-
nally, Harrigan et al. reported on a finite element
analysis of a push-out test and concluded that the
stress patterns along the interface are nonuniform,
with the peak stresses occurring at the bottom of the
implant–cement interface near the support.27 Al-
though the current model is not intended to mimic in
the in situ loading conditions, it does allow for com-
parisons between treatments within this study and
previous studies. The model used in the current study
was similar to that used by Ahmed et al. and Raab et
al., and this allowed for a direct comparison between
the data from the current study and previous stud-
ies.9,24
A model that mimics the in vivo environment one
step further than the previously mentioned push-out
models is an in vivo animal model. Barb et al. bilater-
SILANE BONDING OF CEMENT–IMPLANT INTERFACE
ally implanted uncoated and PMMA-precoated CoCr
femoral prostheses in canine femora for 1, 2, and 6
months.12 The mean push-out strength of the pre-
coated specimens after 1, 2, and 6 months was 17.2,
12.1, and 12.6 MPa, respectively, while the mean
strength of the unocated specimens was 8.4, 4.8, and
7.4 MPa, respectively. On the other hand, Park et al.
showed no significant increase in the mean push-out
strength of uncoated and precoated rods implanted in
dogs for 3 weeks (1.65 and 1.71 MPa, respectively), yet
these were relatively low strengths.14
Previous studies and the current experiment have
demonstrated the increased interface shear strength
achieved by precoating the implant with either a si-
lane bonding agent or a meticulously applied layer of
PMMA. The current study used three different PMMA
formulations and achieved mean failure strengths
from 21.7 to 24.1 MPa 24 h after preparation and from
18.2 to 21.6 MPa after 60 days of saline immersion.
There was no significant degradation of the silane-
coated samples after saline immersion. In addition, the
mean failure strengths achieved after silane coating
were comparable to the highest strengths achieved by
Raab et al. and May et al.9,13,17 The primary difference
between the current study and the previous studies
lies in the surface preparation of the implant surface.
Raab et al. described an elaborate surface prepara-
tion that involved: (1) degreasing the metal specimen
in 1N of boiling NaOH for 10 min, (2) rinsing the
specimen in H2SO4 for 10 s, (3) passivating it in HNO3
for 10 min, and (4) ultrasonically rinsing it in distilled
water twice for 5 min between each step.9 Finally, the
specimens were coated with hydrolized gammameth-
acryloxypropyltrimethoxysilane (A-174). Earlier Sili-
coating methods involved ultrasonically cleaning the
metal specimens with Siliclean (Kulzer, Inc.), drying
the specimens, and pyrolytically depositing a silica
coating before the silane coating was applied.28 To the
contrary, the silane application technique presented in
the current study is relatively simple and easy to per-
form. The specimens are grit-blasted, blown clean,
pre-coated, heated, cooled, and coated again. The ini-
tial coating (Siloc Pre) is a solution of isopropanol,
water, acetone, tetramethoxysilane, soluble silica, and
chromium tetrachloride. The final coating (Siloc Bond)
is a solution of isopropanol, water, acetone, gamma-
methacryloxypropyltrimethoxysilane, and a methac-
rylate acid ester. The distinct advantage of the current
technique compared with previous silane techniques
are the decreased processing time and the lack of
elaborate equipment or caustic materials needed to
prepare the specimens.
When comparing cement formulations, no signifi-
cant difference in mean failure strength was noted,
However, in all silane-coated samples, there was a
trend toward the Simplex-P having the highest mean
131
failure strength, followed by the Palacos E and then
the Palacos R. This trend also was apparent in the
uncoated specimens except that the mean failure
strength of the Palacos E and R were nearly identical.
In another comparison of cement formulations, Stone
et al. reported that there was no significant difference
between the mean shear strengths of Palacos R and
CMW 1.10 However, in every comparison, they found
a trend toward Palacos R providing the greatest shear
strength. There has been a great deal of work con-
ducted to determine and compare the material prop-
erties of different cement formulations.29–34 The ulti-
mate shear strengths of the bulk cement measured by
several investigators range from 20.1 to 41 MPa.35–37
These values are similar to the bone cement/implant
interfacial shear strengths measured in the current
study as well as in many other studies involving either
silane coupling agents or PMMA precoats. Therefore it
would be expected that with the use of these adhesion
promoters, implant failures at the cement/implant in-
terface would be reduced to nearly equal the number
of failures seen originating within the cement mantle.
All cemented femoral implants suffer from bone ce-
ment/implant debonding to some degree. In 1984,
Ahmed et al. stated that “an unbonded stem is likely
to be the predisposing factor for cement fracture.”24
Later, Harris wrote, “The mechanism of initiation of
loosening of cemented femoral components is now
known. It is debonding at the cement metal inter-
face.”5 And Davies et al. stated, “It is becoming in-
creasingly apparent that debonding at the cement–
metal interface is a common occurrence and is a criti-
cal precursor to failure of the total hip replacement.”3
Finally, in 1997, Mann et al. wrote: “The stem–cement
interface is considered to be the primary initiation site
leading to eventual clinical loosening of cemented
femoral hip components.”38 All of these conclusions
were supported by the retrieval work of Jasty et al.1
They studied the cement/implant interface of 16 ce-
mented femoral implants at autopsy. The femurs were
sectioned and microscopically examined for cement
cracks and loosening. They concluded that cement/
stem debonding occurred in all 16 specimens and that
the majority of all the cement mantle fractures seemed
to originate at the cement/stem interface—most nota-
bly at the corners of the implant. The cracks seemed to
occur initially at the distal tip and then secondarily in
the proximal/lateral region.
It is apparent that the previously cited studies are
referring to implants that are in situ for a given
amount of time and cyclically loaded throughout their
lifetime. On the other hand, the present study ana-
lyzes quarter-cycle load to failure, which provides
data representative of the maximum interfacial shear
strength. These data give insight into the added bond
strength provided by the silane coating and allow
132
comparison with previously published data. The next
logical step is to cyclically test similar samples and
compare the fatigue properties for given treatments,
followed by an in vivo animal model to investigate the
long-term in vivo properties and biocompatibility of
the treatments.
CONCLUSIONS
Long-term fixation failure of cemented femoral im-
plants is primarily mechanical and is thought to begin
with debonding at the cement–implant interface. The
increase in strength following silane coating noted in
this study may increase the longevity of the implant
by decreasing debonding at the interface and thereby
subsequent failure due to loosening. We conclude that
a silane bonding agent used in conjunction with a ce-
mented femoral implant will provide an excellent
bond between the cement and the implant. The bond
reduces micromotion at the cement/implant interface
and prevents subsequent three-body wear between
the implant and cement.
References
1. Jasty M et al. The initiation of failure in cemented femoral
components of hip arthroplasties. J Bone Joint Surg Br 1991;73:
551–558.
2. Barrack RL, Mulroy RD Jr., Harris WH. Improved cementing
techniques and femoral component loosening in young pa-
tients with hip arthroplasty. A 12-year radiographic review. J
Bone Joint Surg Br 1992;74:385–389.
3. Davies JP, Harris WH. Tensile bonding strength of the cement–
prosthesis interface. Orthopedics 1994;17:171–173.
4. Gardiner RC, Hozack WJ. Failure of the cement–bone interface.
A consequence of strengthening the cement–prosthesis inter-
face? J Bone Joint Surg Br 1994;76:49–52.
5. Harris WH. Is it advantageous to strengthen the cement–metal
interface and use a collar for cemented femoral components of
total hip replacements? Clin Orthop 1992;285:67–72.
6. Jasty M, Maloney WJ, Bragdon CR, Haire T, Harris WH. His-
tomorphological studies of the long-term skeletal responses to
well fixed cemented femoral components. J Bone Joint Surg
Am 1990;72:1220–1229.
7. Morita S, Yamamoto H, Furuya K, Ishihara K, Nakabayashi N.
Enhanced strength in cemented stem fixation using adhesive
acrylic cement as a metal coating material. J Biomed Mater Res
1997;34:171–175.
8. Raab S, Ahmed AM, Provan JW. The quasistatic and fatigue
performance of the implant/bone–cement interface. J Biomed
Mater Res 1981;15:159–182.
9. Raab S, Ahmed AM, Provan JW. Thin film PMMA precoating
for improved implant bone–cement fixation. J Biomed Mater
Res 1982;16:679–704.
10. Stone MH, Wilkinson R, Stother IG. Some factors affecting the
strength of the cement–metal interface. J Bone Joint Surg Br
1989;71:217–221.
YERBY ET AL.
11. Welsh RP, Pilliar RM, Macnab I. Surgical implants. The role of
surface porosity in fixation to bone and acrylic. J Bone Joint
Surg Am 1971;53;963–977.
12. Barb W, Park JB, Kenner GH, von Recum AF. Intramedullary
fixation of artificial hip joints with bone cement-precoated im-
plants. I. Interfacial strengths. J Biomed Mater Res 1982;16:447–
458.
13. May KB, Russell MM, Razzoog ME, Lang BR. The shear
strength of polymethyl methacrylate bonded to titanium par-
tial denture framework material. J Prosthet Dent 1993;70:410–
413.
14. Park JB, von Recum AF, Gratzick GE. Pre-coated orthopedic
implants with bone cement. Biomater Med Devices Artif Org
1979;7:41–53.
15. Price H, Hawkins M, Bock D, Parr J. PMMA precoating. A
study of the effects of pre- and post-coating treatments on bone
cement adhesion. Trans Orthop Res Soc 1985;31:91.
16. May KB, Van Putten MC, Bow DA, Lang BR. 4-META poly-
methyl methacrylate shear bond strength to titanium. Oper
Dent 1997;22:37–40.
17. May KB, Fox J, Razzoog ME, Lang BR. Silane to enhance the
bond between polymethyl methacrylate and titanium. J Pros-
thet Dent 1995;73:428–431.
18. Ishijima T, Caputo AA, Mito R. Adhesion of resin to casting
alloys. J Prosthet Dent 1992;67:445–449.
19. Ishihara, K. Nakabayashi N. Adhesive bone cement both to
bone and metals: 4-META in PMMA initiated with tri-n-butyl
borane. J Biomed Mater Res 1989;23;1475–1482.
20. Crowninshield RD, Jennings JD, Laurent ML, Maloney WJ.
Cemented femoral component surface finish mechanics. Clin
Orthop 1998;355:90–102.
21. Molino LN, Topoleski LD. Effect of BaSO4 on the fatigue crack
propagation rate of PMMA bone cement. J Biomed Mater Res
1996;31:131–137.
22. Stone JJ, Rand JA, Chiu EK, Grabowski JJ, An KN. Cement
viscosity affects the bone–cement interface in total hip arthro-
plasty. J Orthop Res 1996;14:834–837.
23. Topoleski LD, Ducheyne P, Cuckler JM. A fractographic analy-
sis of in vivo poly(methyl methacrylate) bone cement failure
mechanisms. J Biomed Mater Res 1990;24:135–154.
25. Dhert WJA et al. A finite element analysis of the push-out test:
Influence of test conditions. J Biomed Mater Res 1992;26:119–
130.
26. Wang X, Subramanian A, Dhanda R, Agrawal CM. Testing of
bone–biomaterial interfacial bonding strength: A comparison
of different techniques. J Biomed Mater Res 1996;33:133–138.
27. Harrigan TP, Kareh J, Harris WH. The influence of support
conditions in the loading fixture on failure mechanisms in the
push-out test: A finite element study. J Orthop Res 1990;8:678–
684.
28. Caeg C, Leinfelder KF, Lacefield WR, Bell W. Effectiveness of
a method used in bonding resins to metal. J Prosthet Dent
1990;64:37–41.
29. de Wijn JR, Slooff TJ, Driessens FC. Characterization of bone
cements. Acta Orthop Scand 1975;46:38–51.
30. Hansen D, Jensen JS. Mixing does not improve mechanical
properties of all bone cements. Manual and centrifugation–
vacuum mixing compared for 10 cement brands. Acta Orthop
Scand 1992;63:13–18.
31. Hansen D, Steen Jensen J. Additional mechanical tests of bone
cements. Acta Orthop Belg. 1992;58:268–271.
32. Saha S, Pal S. Mechanical properties of bone cement: A review.
J Biomed Mater Res 1984;18:435–462.
33. Smeds S, Goertzen D, Ivarsson I. Influence of temperature and
vacuum mixing on bone cement properties. Clin Orthop
1997;334:326–334.
SILANE BONDING OF CEMENT–IMPLANT INTERFACE
34. Weber SC, Bargar WL. A comparison of the mechanical prop-
erties of Simplex, Zimmer, and Zimmer low viscosity bone
cements. Biomater Med Devices Artif Org 1983;11:3–12.
35. Astleford WJ, Asher MA, Lindholm US, Rockwood CA Jr.
Some physical and mechanical factors affecting the simple
shear strength of methylmethacrylate. Clin Orthop
1975;108:145–148.
36. Lee AJ, Wrighton JD. Some properties of polymethylmethac-
133
rylate with reference to its use in orthopedic surgery. Clin
Orthop 1973;95:281–287.
37. Stubbs BE, Matthews LS, Sonstegard DA. Experimental fixa-
tion of fractures of the femur with methylmethacrylate. J Bone
Joint Surg Am 1975;57:317–321.
38. Mann KA, Edidin AA, Ordway NR, Manley MT. Fracture
toughness of CoCr alloy–PMMA cement interface. J Biomed
Mater Res 1997;38:211–219.