J. Biomedical Science and Engineering, 2010, 3, 525-528
doi:10.4236/jbise.2010.35073 Published Online May 2010 (http://www.SciRP.org/journal/jbise/ JBiSE
).
Published Online May 2010 in SciRes. http://www.scirp.org/journal/jbise
In vitro evaluation of a new resilient, hard-carbon, thin-film
coating as a bearing material for ventricular assist devices
——In Vitro Bearing Evaluation of Bi oMedFlex
Nicole A. Mielke1, Alex L. Massiello1, David J. Horvath1, Stephen M. Benefit1, Darren Burgess2,
Leonard A. R. Golding1, Kiyotaka Fukamachi1
1Department of Biomedical Engineering/ND20, Lerner Research Institute, Cleveland Clinic, Cleveland, Ohio, USA;
2BioMedFlex, LLC, Denver, North Carolina, USA.
Email: fukamak@ccf.org
Received 1 December 2009; revised 14 December 2009; accepted 20 December 2009.
ABSTRACT
Our aim was to evaluate the potential use of Bio-
MedFlex® (BMF), a new resilient, hard-carbon, thin-
film coating, as a blood journal bearing material in
Cleveland Heart’s continuous-flow left and right ven-
tricular assist devices (VADs). BMF is not classified
as a diamond-like carbon (DLC) and differs from
other thin-film carbon coatings by its high flexural
strength, radiopacity, and wear resistance. A 2- to
4-μm-thick BMF adhesion layer was deposited on the
VAD journal bearing surfaces. A commercial DLC
coating used in other clinical blood pump applica-
tions was used as a control. Durability and reliability
of the BMF coating was verified in severe pump
start/stop testing using 20 BMF-coated journal bear-
ing pairs. The BMF-coated surfaces showed no coat-
ing failures, whereas 57% of the DLC bearing pairs
developed scratches through the carbon coating, do-
cumenting that BMF can provide a durable coating
in our blood journal bearing application. In conclu-
sion, BMF has shown qualities that support its sig-
nificant advantages as an alternative journal bea-
ring material in Cleveland Heart pumps. Our plan
includes biocompatibility testing with ongoing animal
studies, endurance testing with submerged pumps
running in saline, and assessment of batch coating
processing capability.
Keywords: Heart Assist Device; Diamond-Like Carbon;
Materials Testing; Wear Resistance; Journal Bearing
1. INTRODUCTION
Implantable ventricular assist devices (VADs) are a
reliable treatment option for patients with terminal-
stage heart failure who are unresponsive to conven-
tional therapies. Although current designs demonstrate
adequate performance, reliability remains a significant
issue, as these devices are implanted for extended dura-
tions (well beyond 1 year) and may likely be intention-
ally implanted in patients as permanent (“destination”)
therapy. Because components of implantable pumps
must withstand difficult in vivo environments and run-
ning conditions, carbon materials (diamond-like carbon
[DLC] and pyrolytic carbon) have emerged as promising
coating materials for pump components to enhance du-
rability and biocompatibility characteristics. Some of
these carbon coatings are chemically inert, wear and
corrosion resistant, and bio- and hemocompatible. These
coatings may also minimize platelet adhesions and acti-
vation, prevent thrombogenicity, and improve perform-
ance by decreasing power usage or increasing pump
output.
The continuous-flow VADs of Cleveland Heart (Char-
lotte, NC) have been adapted from the CorAide LVD-
4000 Left Ventricular Assist System (Arrow Interna-
tional, Reading, PA), developed at Cleveland Clinic.
Detailed descriptions of the pump have previously been
published [1,2]. The left and right ventricular assist de-
vices (LVADs and RVADs, respectively) consist of three
subassemblies: the volute housing, the rotatin g assembly
(RA), and the stator assembly (Figure 1). The RA con-
tains a cylindrical four-pole magnet and spins aroun d the
titanium stator housing post, which contains the motor
windings. The RA is supported axially by permanent
magnets and radially by a thin film of blood, forming a
patented blood journal bearing which, except for startup
and shutdown, makes no mechanical contact during use.
The blood journal bearing materials, dimensions, clear-
ances and geometry are shared by both VADs. The vo-
lute housings and RA impeller blades are optimized for
the different afterload conditio ns that occur in the RVAD
526 N. A. Mielke et al. / J. Biomedical Science and Engineering 3 (2010) 525-528
Copyright © 2010 SciRes. JBiSE
Figure 1. Depiction of an LVAD subassembly. RVADs differ
only in RA impeller and in volute housing design.
vs. LVAD.
During European clinical trials of the CorAide LVAD,
it was found that the original fluorinated ethylene propyl-
ene (FEP) coating on the journal bearings delaminated
from its titanium substrate. The authors have since sea-
rched for the best alternative bearing material to use
on both the LVAD and RVAD; this article focuses on
the preliminary in vitro evaluation of the BioMedFlex
(BMF) thin-film coating (BioMedFlex, Denver, NC).
2. MATERIAL AND METHODS
2.1. BioMedFlex Coating
The BMF coating is deposited on substrates by a pro-
prietary plasma-assisted chemical vapor deposition
process in a high-vacuum environment at temperatures
below 200. Because this coating is created from layers
of nano-crystalline diamond and nano-crystalline silicon
carbide in a matrix of the noncrystalline forms of both
compounds, BMF’s unique characteristics are set apart
from other thin-film carbon coatings used in biomedical
applications [3]. Its high flexural strength, its radiopacity,
and its wear resistance are surpassed only by pure dia-
mond thin films. Several other material properties of the
BMF coating that make it a good candidate for use in
blood journal bearing applications include chemical re-
sistance, dimensional stability, and low coefficient of
friction. BMF does not absorb water and does not cor-
rode in saline. These two key qualities prevent dimen-
sional changes due to absorption or corrosion in the
high-tolerance clearance and profile of the journal bear-
ing-conditions that occurred with the FEP coating in the
CorAide clinical trial. BMF is a clean, reliable, high-
technology coating process that is not line-of-sight and
does not require parts to be rotated in the chamber. The
BMF coating is 2-4 m thick and aggressively adheres
to titanium surfaces, requiring sandblasting to remove it
from the surface of coated parts. For reference, a matrix
comparing the material properties of BMF vs. titanium
nitride, a commonly used material in the field, is pro-
vided bel ow as Table 1.
2.2. Journal B ea r in g Starting Conditions
Once running, the bearing allows the RA to revolve
levitated from the stator housing post surfaces on a
film of blood, deriving lift from its own motion and
thereby eliminating any surface contact between mov-
ing parts until the pump is shut down (Figure 2).
However, there is transient (300 ms) sliding and vi-
brating contact (800-900 g) load between the RA and
stator assembly at start-up and again during stopping
(Figure 3); the only mechanical loading the bearing
undergoes is during its start/stop cycles. The worst-
case expected loading sustained by the bearing over its
lifetime was established based on the fact that, on av-
erage, the pump is started and stopped six times during
manufacturing and three times during clinical implanta-
tion. During its service life with the patient, the pump
may be started and stopped (in a worst case) eight times
per year due to external component changes, software
upgrades, or simple patient connection mistakes. These
values translate to about 50 starts and stops in a 5-year
period. Multiplying by a safety factor of 2, the number
of starts and stops that will approximately simulate the
worst-case service life is 100.
Table 1. Comparison matrix of material properties: Bio-
MedFlex vs. Titanium nitride.
Typical values BioMedFlex Titanium nitride
Inert Yes No
Flexural strength High triaxial Low single axis,
brittle
Optical transparency 50Å Yes, 2 μm No No
Mechanical wear Low High
Conformal Yes No: Line of sight
Hardness (V) 3800 2500
Young’s modulus (GPa) 130-180 600
Adhesion (PSI) > 8500 < 2000
Coefficient o f friction 0.10-0.15 0.65
Thickness (μm ) 0.1-4 0.25-12
Surface Ra (μm) 0.10 0.20
Figure 2. Journal bearing. Le ft: Inner surfaces of t he RA are
in contact with stator when pump is not running. Right: After
starting, the RA is separated from the stator and rides on a
film of blood dur ing operation. Assy, Asse mbly.
N. A. Mielke et al. / J. Biomedical Science and Engineering 3 (2010) 525-528 527
Copyright © 2010 SciRes.
the DLC-coated parts limited the number of DLC test
pieces evaluated in this study, as the primary goal of the
project was to develop new journal bearing materials.
The success criterion was the completion of at least 100
start/stop cycles per bearing pair without damage to the
coated surfaces.
3. RESULTS
BMF-coated pumps showed no coating failures (Table
2), whereas four of seven commercial DLC bearing pairs
developed scratches through the carbon coating (Table
3), documenting that BMF can provide a durable coating
in our blood jour nal beari n g application. A representative
scratch is shown in Figure 4; scratches were present on
the stator post only, with no damage to the RA on com-
mercial DLC bearing pairs.
Figure 3. Pump vibration and motor current at start-up and
shut-down of a Cleveland Heart VAD.
2.3. Start/Stop Test
To evaluate the durability and reliability of the coating
on pump journal bearing mated surfaces, pump assem-
blies consisting of a BMF-coated stator and RA were
exposed to 100 start/stop cycles at the most challenging
bearing load conditions: maximum pump speed and
wide-open flow conditions in de-ionized water. The
pumps were disassembled, and the stator and RA were
visually inspected for scratches, delamination, or other
damage after 1, 5, 25, and 100 start/stop cycles.
4. DISCUSSION
Although titanium alloy has been the material of choice
for implantable rotary blood pumps [4], given the bear-
ing loads seen in our pumps at startup and shutdown, it
is a very poor choice as a bearing material. It has been a
priority to determine the optimum coating material and
coating parameters to obtain the bearing properties
needed for the Cleveland Heart LVAD and RVAD.
BMF’s unique thin-film carbon formulation and its lay-
ered matrix of diamond and silicon carbide have pro-
vided a new biomedical material that combines very
high hardness, aggressive substrate adhesion and much
higher flexural properties. BMF has proven to be more
wear resistant than the commercially available DLC
control.
A commercial DLC coating used in other blood pump
applications was also applied to another set of bearing
surfaces as a control. Twenty different combinations of
BMF-coated journal bearing pairs and seven different
DLC-coated pairs were exposed to start/stop testing to
compare the two coatings for durability and reliability.
The high component costs, long lead time for fabrication
and coating of parts, and the early and frequent failure of Previous studies of this bearing design have documented
Figure 4. BMF-coated RVAD RA (left) and stator assembly (middle); commercial DLC-
coated stator assembly scratches post-test (right).
Table 2. BMF bearing testing results.
BMF Coated RA
BMF Bearing Testing RA BMF-01 RA BMF-02 RA BMF-03 RA BMF-04
Stator BMF-01 Passed: No damagePassed: No damagePassed: No damage Passed: No damage
Stator BMF-02 Passed: No damagePassed: No damagePassed: No damage Passed: No damage
Stator BMF-03 Passed: No damagePassed: No damagePassed: No damage Passed: No damage
Stator BMF-04 Passed: No damagePassed: No damagePassed: No damage Passed: No damage
BMF-Coated Stator
Stator BMF-05 Passed: No damagePassed: No damagePassed: No damage Passed: No damage
BMF, BioMedFlex®; RA, Rotating Assembly.
JBiSE
528 N. A. Mielke et al. / J. Biomedical Science and Engineering 3 (2010) 525-528
Copyright © 2010 SciRes. JBiSE
Table 3. Commercial DLC bearing testing results.
Commercial DLC-Coated Rotating Assembly
Commercial DLC Bearing TestingRA DLC-01 RA DLC-02 RA DLC-03 RA DLC-04 RA DLC-05
Stator DLC-01 Passed: No damagePassed: No damageFAILED FAILED -
Stator DLC-02 - - - Passed: No damage FAILED
Commercial
DLC-Coated
Stator Stator DLC-03 - - FAILED - -
DLC, diamond-like coating; RA, Rota ting Assembly.
that the blood film and journal bearing remain stable
once the start-up period is over [5]. Because of this,
damage to the bearing can occur only dur ing startin g and
stopping; it is not a matter of total running time, but the
number of start/stops that is critical. According to the
test procedure, pump disassembly and component in-
spection was performed only after the 1st, 5th, 25th and
100th start/stop cycle. Generally, bearing damage oc-
curred early on in the testing (within the first 5 start/stop
cycles) for the DLC-coated parts and was progressive
and cumulative. Representative bearing wear (Figure 4)
for the commercial DLC-coated stator is compared to
undamaged BMF-coated RA and stator parts post-bear-
ing test; bearing wear was only on the stator post of the
DLC-coated parts and not the RA.
With start/stop reliability demonstrated, endurance tests
of indefinite duration in 37 saline and glycerin blood
analog have been started to verify compatibility in a
saline environment. After 275 and 70 days, respectively,
there has been no bearing wear on two pump assemblies
with BMF paired journal bearings undergoing mock
circulatory in vitro endurance testing in an aggressive,
pulsatile environment.
BMF was successfully deposited onto titanium pump
surfaces and endured at least 100 start/stop cycles for
each bearing pair without damage; over 57% of the
commercial DLC-coated pairs failed at the same test
parameters. Five of the BMF stators accumulated up to
800 start/stop cycles with no wear. Neither the testing
duration nor number of start/stop cycles were increased,
as the test was deemed to be adequately aggressive given
the “failure” of the commercial DLC pairs. These tests
demonstrated that BMF can provide a durable coating in
our application, along with a viable solution to bearing
FEP coating adhesion problems seen in the initial Cor-
Aide clinical trials.
Biocompatibility is being validated with ongoing in
vivo studies of the performance of LVADs, RVADs, and
biventricular assist devices in animals [6].
5. CONCLUSIONS
BMF has many qualities that support its significant ad-
vantages as an alternative journal bearing material in
both Cleveland Heart pumps: 1) demonstrated bearing
reliability, 2) a thin-film coating that still offers a depth
of several micrometers, and 3) the capability for batch
coating processing. To further test durability, BMF-
coated pumps have been placed on mock circulatory
endurance test with the pumps running and su b merged in
a saline/glycerin blood analog fluid at body temperature.
6. ACKNOWLEDGEMENTS
This project was supported by the Global Cardiovascular Innovation
Center and the State of Ohio under the Third Frontier project. The
authors also acknowledge support from grant BRP 5 R01 HL074896
(to K.F.) from the National Heart, Lung, and Blood Institute of the
National Institutes of Health.
REFERENCES
[1] Fukamachi, K., Horvath, D.J., Massiello, A.L., Ootaki, Y.,
Kamohara, K., Akiyama, M., Zahr, F., Kopcak, M.W.,
Dessoffy, R., Chen, J.F., Benefit, S. and Golding, L.R.
(2005) Development of a small implantable right ven-
tricular assist device. American Society for Artificial
Internal Organs Journal, 51(6), 730-735.
[2] Ootaki, Y., Kamohara, K., Akiyama, M., Zahr, F., Kopcak,
M.W., De ssoffy, R., Massiello, A.L., Horvath, D.J., Chen,
J.F., Benefit, S., Golding, L.R. and Fukamachi, K. (2005)
Initial in vivo evaluation of the DexAide right ventricular
assist device. American Society for Artificial Internal
Organs Journal, 51(6), 739-742.
[3] BioMedFlex, LLC, Accessed November 27, 2009. http://
www.BioMedFlex.com
[4] Yamazaki, K., Litwak, P., Tagusari, O., Mori, T., Kono,
K., Kameneva, M., Watach, M., Gordon, L., Miyagi-
shima, M., Tomioka, J., Umezu, M., Outa, E., Antaki, J.F.,
Kormos, R.L., Koyanagi, H. and Griffith, B.P. (1998) An
implantable centrifugal blood pump with a recirculating
purge system (Cool-Seal system). Artificial Organs,
22(6), 466-474.
[5] Malanoski, S.B., Belawski, H., Horvath, D.J., Smith,
W.A. and Golding, L.R. (1998) Stable blood lubricated
hydrodynamic journal bearing with magnetic loading.
American Society for Artificial Internal Organs Journal,
44(5), M737-M740.
[6] Takaseya, T., Fumoto, H., Shiose, A., Arakawa, Y., Rao,
S., Horvath, D.J., Massiello, A.L., Mielke, N.A., Chen,
J.F., Zhou, Q., Dessoffy, R., Kramer, L., Benefit, S.,
Golding, L.R. and Fukamachi, K. (2009) In vivo bio-
compatibility evaluation of a new resilient, hard-carbon,
thin-film coating for ventricular assist devices. Artificial
Organs (in Press).