Shape-memory anchoring system for bladder sensors

Iaci M. Pereira,1,2,3 Fabrice Axisa,3 Rodrigo L. Oréfice,2 Jan Vanfleteren,3 Hercules P. Neves4
1
Brazilian Army Technological Center, Rio de Janeiro, Brazil
2
Department of Metallurgical and Materials Engineering, Federal University of Minas Gerais, Belo Horizonte, Brazil
3
Centre for Microsystems Technology, Technology Park, Ghent University/Interuniversity Micro-Electronics Center,
Zwijnaarde, Belgium
4
Interuniversity Microelectronics Centre, Leuven, Belgium

Received 20 July 2010; revised 5 October 2010; accepted 17 October 2010

Published online 28 December 2010 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jbm.b.31780

Abstract: In this work, we propose the use of shape-memory
polymer as an anchoring system for a bladder sensor. The
anchoring system was designed from a biomedical biodegrad-
able water-based poly(ester-urethane) produced in an aqueous
environment by using isophorone diisocyanate/hydrazine (hard
was simulated and measured in synthetic urine. Results indicated
that shape recovery can occur at body temperature and expul-
sion of the sensor by the body along with urine may occur
through the combined effect of urine hydrolytic attack and
compression exerted by the bladder walls. V
C 2010 Wiley Periodicals,

segment) and poly(caprolactone diol)/2,2-bis (hydroxymethyl) Inc. J Biomed Mater Res Part B: Appl Biomater 96B: 369–375, 2011.
propionic acid (soft segment) as the main reagents. Tensile
strength and elongation-at-break deterioration upon degrada- Key Words: shape-memory, bladder sensor, biodegradable
tion in synthetic urine were investigated. In-body shape recovery waterborne

INTRODUCTION stress for the patient. However, as the sensor become

The clinical importance of monitoring physical and chemi-
cal parameters in the body has led to a significant motiva-
tion to develop sensors for biomedical use. Moreover,
presently a convergence of factors is resulting in rapid
development of the field, for instance: (i) advances in sili-
con microprocessor technology, (ii) advances in biomateri-
als, and (iii) the advent of minimally invasive surgery.1
Benefiting from these advances, a new generation of elec-
tronic devices fabricated on bendable lightweight miniatur-
ized substrates is widely believed to have great potential
for applications in developing new in vivo tools for continu-
ous monitoring of health indicators.1 These devices can be
inserted in the body through natural body openings or
small artificial incisions and should be functional and
stable during the length of time for which monitoring is
required. Noninvasive sensors have a huge potential for
disease prevention and for diagnosis in medicine2 and
have recently been investigated in the particular context of
pressure sensing in the bladder.
Sensing pressure in the bladder is an important topic in
urology research. Bladder diseases can be prevented or
predicted by observing abnormal variations in bladder urine
pressure. Therefore, periodic evaluation of patients has been
recognized as one of the most important research topics in
clinical medical investigations.3 Furthermore, sensor minia-
turization has enabled the pressure sensor to be implanted
through catheters into the bladder with minimal pain and
smaller, it has become more difficult to keep the sensor
implanted and to control its tendency to be expelled along
with urine.
In this work, we propose the use of shape-memory (SM)
polymer as an anchoring system for the bladder sensor. SM
polymer are thermo-responsive functional material which
possess the ability to memorize a permanent shape that can
substantially differ from their initial temporary form.4,5 Dur-
ing this study, it was employed a biomedical biodegradable
waterborne poly(ester-urethane) (PU) which presents
SM properties. The built device can be introduced into the
bladder in a compressed temporary shape through mini-
mally invasive surgery. Once implanted into the bladder, at
body temperature, the polymer expands to its permanent
shape. The expanded larger volume helps to prevent the
sensors expulsion and stabilizes the sensor, anchoring it in
the urinary bladder. Moreover, since it is biocompatible,
minimal foreign body reaction is expected and since it is
designed to undergo hydrolytic degradation, the sensor
would be eliminated by the body after an appropriate time
has passed, which would eliminate the need for a second
surgical procedure.
MATERIALS AND METHODS
Sensor design
Sensors may be built over a stretchable polyimide (PI) film
and encapsulated in silicone. In addition to being

Correspondence to: I. M. Pereira; e-mail: iacipere@yahoo.com.br

Contract grant sponsors: National Council for Scientific and Technological Development (CNPq); The State of Minas Gerais Research Foundation
(FAPEMIG)

V
C 2010 WILEY PERIODICALS, INC. 369
 FIGURE 1. Sensor production scheme.
biocompatible, silicone encapsulation protects the sensor
from chemical attack by urine. The encapsulated sensor
(pill) is then inserted in the body. The pill’s shape and small
size facilitate implantation; however, it may also ease the
pill’s ejection by the body. In the present work, the elec-
tronic system was simulated by a PI film containing copper
patterning. Figure 1 shows the simulated sensor production
scheme.
Initially the water-based PU was produced, and cast in
molds. The PU film obtained was cut in the appropriate
shape. To guarantee a strong interface between PU and PI
films, the PU area that interfaces the PI film was subjected
to surface modification and attached to the PI film, i.e.,
the area was wiped with acetone and attached to the PI
film by pressure at a temperature of 80
C. Finally, the
adherence between the ensemble and the silicone is
enhanced by adherence promoter. Afterwards, the system
was encapsulated in silicone by injection molding.
Polymer synthesis
Poly(caprolactone diol) (PCL1250Mn ¼ 1250 g/mol and
PCL2000Mn ¼ 2000 g/mol), isophorone diisocyanate
(IPDI), 2,2-bis(hydroxymethyl) propionic acid (DMPA), and
dibutyl tin dilaurate (0.01 wt %, DBDLT) were obtained
from Aldrich (St. Louis, MO). Triethylamine (TEA, 98%) and
hydrazine (HZ, 25%) were purchased from Vetec (RJ,
Brazil). All these chemicals were used throughout this work
without any previous treatment.
PUs were synthesized by the conventional prepolymer
method using a 250-mL three-necked glass flask equipped
with a heating mantel, a mechanical stirrer, and a ther-
mometer with nitrogen atmosphere. In the first step,
prepolymer was obtained by reacting PCL1250 (23.7 wt
%), PCL2000 (37.9 wt %), DMPA (3.8 wt %), and IPDI
(29.6 wt %) at 2 NCO/OH ratio for about 3.5 h. During this
370 PEREIRA ET AL.
time, DBDLT was added twice. The reaction was carried
out at 70–75
C. After cooling to 40
C, the carboxylic acid
groups were neutralized by adding TEA (2.9 wt %). The
mixture was then gently stirred for 40 min. Dispersion in
water and PU chain extension were achieved by adding
HZ (2.1 wt %) and deionized water to the neutralized pre-
polymer in high-speed stirring. To ensure that the reaction
was completed, the mixture was stirred for 60 min. This
chemical procedure was very successful in producing PU
dispersions in water with about 30% solid content. The
weight fraction (%) of hard segment W (HZ) was 32.6%
(IPDI þ HZ).
To produce a film from the water-based polymeric dis-
persions, the water dispersion was cast in an open mold
and allowed to dry. The films were then annealed at 80
C
for 72 h.
Laser machine
Figure 2 shows the anchor system draft; the shape is
obtained by laser cutting. A custom-built laser set-up inte-
grating a CO2 laser (GSIL Impact SSM 2150) was used to
cut the PU film. The laser operated with pulse energy of
17.8 J/cm2, pulse repetition rate of 100 Hz, and spot size of
FIGURE 2. Anchor system draft (thickness ¼ 0.4 mm).
SHAPE-MEMORY ANCHORING SYSTEM FOR BLADDER SENSORS

FIGURE 3. (a) Tensile specimen (sample thickness 0.4 mm) and (b)
peel specimen.
4
5.4  10 2
cm . The sample was mounted on the fixed
working table and the machining process was completed by
moving the projection lens at 1 mm/s. Velocity and laser
displacement were computer controlled.
Infrared spectroscopy
Infrared spectra were collected in a Fourier transform infra-
red spectrophotometer (Spectrum 1000 model Perkin-Elmer
FTIR). Measurements were carried out using attenuated
total reflectance (ATR) technique. Each spectrum was a
result of 64 scans with a resolution of 4 cm1.
Mechanical tests
Tensile strength (rm) and elongation at break (em) of PU
samples shown in Figure 3(a) were measured using a uni-
versal testing machine (INSTRON 5543) at 10 mm/min
crosshead speed. Sample length between the clamps was
30 mm. Tests were performed at room temperature. Results
represent the average of three samples. To evaluate adhe-
sion of PU and PI film, 180
peel tests were performed. Peel
specimens consisted of strips with dimension of 10 mm 
5 mm  0.4 mm that were bonded over 5 mm. Bonded
specimens were placed in an INSTRON 5543 tensile testing
machine with the PU portion in the upper clamp and PI
film in the lower one. A peeling angle of 180
was achieved
with the specimen in the position described in Figure 3(b).
The PU film was then peeled from the PI by separating the
two clamps with a crosshead speed of 2 mm/min. Bond
strength was obtained by Eq. (1):
F Tests were performed at room temperature. The average of
Bond Strength ¼
A
in which F is maximum force applied and A is cross-
sectional area of the bonded specimen. Results represent
the average of three.
To estimate adhesion of the ensemble on silicone, two
pills were submitted to tension tests. The pills were placed
in an INSTRON 5543 tensile testing machine with the PU
portion in the upper clamp and silicone in the lower one.
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | FEB 2011 VOL 96B, ISSUE 2
ORIGINAL RESEARCH REPORT
The clamps were pulled apart with a crosshead speed of
0.5 mm/min. Results represent the average of three samples.
SM properties
Shape memory experiments were done using an INSTRON
5543 tensile testing machine. Six strips with dimension of
28 mm  3.5 mm  0.4 mm were tested. Sample length
between clamps was 20 mm and crosshead speed was
10 mm/min. The shape recovery cycle consisted of the
following steps: (1) samples were deformed to 55 mm at
room temperature; (2) samples were removed from the
clamps and immediately stored in a refrigerator at 20
C;
(3) the frozen sample was divided into two pieces; (4)
divided samples were subsequently immersed in different
mediums at 37
C for 15 min to recover its original shape.
The first sample was immersed in artificial urine obtained
from SURINEV (LENEXA, KS), and the second one in
R
deionized water. Figure 4 illustrates a SM thermomechanical
cycle test.
SM properties of the materials were described by strain
recovery ratio, RR, and strain fixity ratio, Rf. Both can be
determined according to Eqs. (2) and (3):
Lm  LRes
Rr ð%Þ ¼  100 (2)
Lm  Lo
LRet  Lo
Rf ð%Þ ¼  100 (3)
Lm  Lo
in which: Lo is original length, Lm is deformed length, Lret is
retained length, and Lres is length after recovery.
Shape recovery body simulation
To simulate shape recovery of implanted pills, two pills
were subjected to a shape-memory thermomechanical cycle.
Initially, the arms were stretched, folded, and stored in a
refrigerator at 20
C. The recovery was simulated at 37
C,
immersing the deformed pills in artificial urine (SURINEV).
R
Figure 5 illustrated the shape memory cycle.
Hydrolytic degradation tests
Hydrolytic degradation was investigated by immersing
specimens in synthetic urine (SURINEV) at 37
C, shown in
R
Figure 3(a). Samples were removed from the media at 1
day, 7 days, 16 days, and 25 days. Wet sample mechanical
properties rm and em were measured using a universal test-
ing machine (INSTRON 5543). Before the experiments,
specimens were gently wiped. Sample length between
clamps was 30 mm and crosshead speed was 10 mm/min.
(1) two samples was obtained. rm and em were calculated as
fractions of the non-degraded sample, Eq. (4).
 
Pt  P0
DPt ð%Þ ¼  100 (4)
P0
in which P represents measured mechanical property rm
or em and indexes 0 and t are the initial value and the
property measured at time t, respectively.
371
 FIGURE 4. Shape memory thermomechanical cycle.
After mechanical tests, the specimens were dried to a
constant weight at 37
C for 1 week and weighed using an
analytical balance to determine weight loss. Weight loss of
PU films after degradation was evaluated as residual weight
(%), which was defined by Eq. (5).
 
w 0  wt
Residual Weight ð%Þ ¼ 100   100 (5)
w0
In which w0 and wt were the initial weight and the
weight at time t, respectively.
RESULTS AND DISCUSSION
The characteristic absorption bands of poly(ester-urethanes)
are assigned in Figure 6. Primary amine stretching modes
appear at 3600–3150 cm1. Carbonyl group (C¼ ¼O) stretch-
ing vibrations appear at 1760–1600 cm1. Secondary amide
absorption, >NAH, appears at 1640–1540 cm1. The band
at 1150 cm1 is assigned to the stretching of the
ACAOACA group.6–9
FIGURE 5. Shape memory recovery in vivo simulation.
372 PEREIRA ET AL.
Figure 7 shows photographs of (a) PU film, (b) ensemble,
(c) pill after silicone encapsulation by injection molding, and
(d) simulated sensor. To produce a film from waterborne
polymeric dispersions, the water dispersion was cast in an
open mold and allowed to dry. After application on the mold,
the water evaporated quickly from the system. Polymer
particles moved and stabilized the molecule. Eventually,
hydrogen bonding formed between the molecules. Since the
water content of the dispersion was high, during film forma-
tion there were frequent shrinkage defects, i.e., a depression
in the surface or an internal void. Because the problem
occurred in the region that dried last, the mold was designed
with thickness variations so the region of interest dried first
without any shrinkage defect, Figure 7(a). The thicker region
of the mold contains more dispersion volume and thus
FIGURE 6. FTIR spectra of PU produced.
SHAPE-MEMORY ANCHORING SYSTEM FOR BLADDER SENSORS
 ORIGINAL RESEARCH REPORT

FIGURE 7. Photographs from the sensor manufacturing process: (a) PU film, (b) ensemble, (c) pill after encapsulation by injection molding, and (d)
simulated sensor.

has longer drying time; this region also works as a reservoir
supplying the surrounding regions with liquid.
Molding defects during silicone encapsulation were
avoided by controlling surface roughness of the mold. It was
also necessary to design a mold with two cavities; one was
used to inject silicone into the mold and the other one was
used to release air from the mold and bubbles trapped in
the injected silicone, Figure 7(c). After encapsulation, excess
molding material could penetrate into mold gaps. This flash
was removed to assure the final surface quality of the pills,
Figure 7(d).
Typical mechanical properties of the films are illustrated
in Figure 8. PU is characterized by high ductility and good
mechanical properties, rm ¼ 19.0 6 1.9 MPa and em ¼ 691
6 9.6%. High adherence strengths were observed between
PU and PI film. Bond strength between the two films was
26.1 6 5.3 KPa. Tearing of PI film was observed as fracture
mode for some specimens, which might indicate that higher
average bond strength could be observed if the PI film were
thicker. Although adherence forces between PU and silicone
have been reported as being poor, pill tension tests showed
good maximum load, 7.6 6 2.7 N; however, the test may
have measured the adherence between the PI and silicone
rather the adherence between the PU and silicone.
RR in synthetic urine and DI water are 92.7% 6 1.6%
and 95.2% 6 2.3%, respectively, results represent average
of six samples. During shape memory experiments, upon
removing the clamps in Step 2, a significant recovery from
strain occurs and thus low values of Rf are observed, 68.3
6 12.0%. According to Ping et al.,6 when deformation tem-
perature is too low (30
C below the melting temperature of
the soft segment), shape recovery starts at a lower tempera-
ture and takes place over a wide temperature range. In this
study, this feature guarantees recovery at low temperatures
similar to body temperature. However, higher Rf could be
obtained if the metastable structure developed during strain
were frozen before the mechanical clamps were removed.
Figure 9 shows photographs of the shape recovery body
simulation. Figure 9(a) shows the pill before deformation,
Figure 9(b) shows the stretched arms, and Figure 9(c)
shows the folded pill in the refrigerator. The restored shape
can be seen in Figure 9(d). As recovery starts at low tem-
perature, it is important to handle the pill with appropriate
tools because the temperature of the hand may initiate the
recovery process.
Previous works9 showed that the hydrolytic degradation
process is characterized by hydrolytic attack on the soft seg-
ment ester, increasing crystallinity of the soft segments. The

FIGURE 8. Mechanical test and peel test of obtained specimens.

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | FEB 2011 VOL 96B, ISSUE 2 373
 FIGURE 9. (a) Pill before deformation, (b) pill after deformation, (c) pill stored in the refrigerator, and (d) restored shape.

FIGURE 10. (a) Stress–strain curve of untreated specimen and degraded samples; (b) mechanical property comparison.

chemi-crystallization process10 is accompanied by progressive CONCLUSION

reduction of the hydrogen-bonded fraction of carbonyl groups. Biodegradable shape memory polymer is an appropriate
The tensile properties of plastics, above all tensile strength anchoring system material for a miniaturized bladder sen-
and elongation-at-break, are sensitive to material degrada- sor. The sensor can be implanted during a minimally inva-
tion.11 Stress–strain curves of untreated and hydrolytic sive surgery and after an appropriate time has passed,
degraded specimens are given in Figure 10(a). Mechanical
properties rm and em of degraded samples were compared
with mechanical properties of untreated specimens. The
results are summarized in Figure 10(b).
Hydrolytic attack is primarily associated with reduction
in tensile strength and also with a strong reduction in elon-
gation-at-break until specimens become completely brittle.
Mechanical tests indicate that sensor expulsion by the body
may occur due to erosion of highly degraded material. The
combined effect of hydrolytic attack by urine and the
compression cycles exerted by the bladder walls can cause
the anchoring system to disintegrate. The pieces and the
sensor will be expelled by the body along with urine. As
observed in Figure 11, weight loss after 25 days of degrada-
tion is small, thus the combined effects are more important
than weight loss alone.
After 25 days, the material has become too brittle to
determine its mechanical properties. Complete degradation
is observed after 48 h of hydrolytic attack in synthetic urine
at 80
C. FIGURE 11. Degradation mass changes of polyurethanes.

374 PEREIRA ET AL. SHAPE-MEMORY ANCHORING SYSTEM FOR BLADDER SENSORS


hydrolytic attack by urine and bladder wall compression
cycles can cause the system to disintegrate; the pieces and
sensor would be eliminated along with urine without a
second surgical procedure. Shape recovery in a synthetic
urine medium was observed at body temperature.
REFERENCES
1. Ratner BD, Hoffmann AS, Schoen FJ, Lemons JE. Biomaterials
Science: An Introduction to Biomaterials in Medicine, 2nd ed.
Elsevier; 2004.
2. Axisa F, Schmitt PM, Gehin C, Delhomme G, McAdams E, Dittmar
A. Flexible technologies and smart clothing for citizen medicine,
home healthcare, and disease prevention. IEEE Trans Inf Technol
Biomed 2005;9:325–336.
3. Wang CC, Huang CC, Liou JS, Ciou YJ, Huang IY, Li CP, Lee YC,
Wu WJ. An implantable long-term bladder urine pressure
measurement system with a 1-atm canceling instrumentation
amplifier. IEEE Int Symp Digital Object Identifier 2007;2383–2386.
4. Lendlein A, Langer R. Biodegradable, elastic shape-memory poly-
mers for potential biomedical applications. Science 2002;296:
1673–1676.
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | FEB 2011 VOL 96B, ISSUE 2
ORIGINAL RESEARCH REPORT
5. Lendlein A, Kelch S. Shape memory polymers. Angew Chem Int
Ed 2002;41:2034–2057.
6. Ping P, Wang W, Chen X, Jing X. Poly(e-caprolactone) Poly-
urethane and Its Shape-Memory Property. Biomacromolecules
2005;6:587–592.
7. Chattopadhyay DK, Sreedhar B, Raju KVSN. The phase mixing
studied of moisture cured crosslinked poly-urethane-ureas during
cure. Polymer 2006;47:3814–3825.
8. Bao H, Zhang Z, Ying S. Effect of ionic group on the
behaviour of polyurethaneurea emulsion. Polymer 1996;37:
2751–2754.
9. Pereira IM, Gomide VS, Zonari AAC, Leite MF, Orefice RL, Goes
AM. Proliferation of human mesenchymal stem cells derived from
adipose tissue on polyurethanes with tunable biodegradability.
Polı́meros 2010;20:1–8.
10. Gorna K, Gogolewski S. In vitro degradation of novel medical
biodegradable aliphatic polyurethanes based on e-caprolactone
R V
and Pluronics with various hydrophilicities. Polym Degrad Stab
2002;75:113–122.
11. Pretsch T, Jakob I, Müller W. Hydrolytic degradation and functional
stability of a segmented shape memory poly(ester urethane). Polym
Degrad Stab 2009;94:61–73.
375