Biodegradable nanofiber-based

piezoelectric transducer
Eli J. Currya,1, Thinh T. Leb,1, Ritopa Dasa, Kai Kec, Elise M. Santorellad, Debayon Pauld, Meysam T. Chorsib,
Khanh T. M. Trana, Jeffrey Baroodya, Emily R. Borgesa, Brian Koe, Asiyeh Golabchif, Xiaonan Xing, David Roweg,
Lixia Yueh, Jianlin Fengh, M. Daniela Morales-Acostai, Qian Wuj, I-Ping Chenk, X. Tracy Cuif, Joel Pachterd,
and Thanh D. Nguyena,b,i,2
a
Department of Biomedical Engineering, University of Connecticut, Storrs, CT 06269; bDepartment of Mechanical Engineering, University of Connecticut,
Storrs, CT 06269; cCollege of Polymer Science and Engineering, Sichuan University, Chengdu 610065, Sichuan, China; dBlood–Brain Barrier Laboratory,
Department of Immunology, University of Connecticut Health Center, Farmington, CT 06030; eDepartment of Agricultural and Biological Engineering,
Mississippi State University, Mississippi State, MS 39762; fDepartment of Bioengineering, University of Pittsburgh, Pittsburgh, PA 15260; gCenter for
Regenerative Medicine and Skeletal Development, University of Connecticut Health Center, Farmington, CT 06030; hDepartment of Cell Biology, University
of Connecticut Health Center, Farmington, CT 06030; iInstitute of Materials Science, University of Connecticut, Storrs, CT 06269; jPathology and Laboratory
Medicine, University of Connecticut Health Center, Farmington, CT 06030; and kDepartment of Oral Health and Diagnostic Sciences, School of Dental
Medicine, University of Connecticut Health Center, Farmington, CT 06030

Edited by Samir Mitragotri, Harvard University, and accepted by Editorial Board Member James J. Collins November 26, 2019 (received for review June
16, 2019)

Piezoelectric materials, a type of “smart” material that generates been extensively studied and shown to be safe and the most ef-
electricity while deforming and vice versa, have been used exten- fective tool (11–13). However, the use of external US is limited to
sively for many important implantable medical devices such as small animals with thin skull bones. Since the human skull is thick
sensors, transducers, and actuators. However, commonly utilized and absorbs more than 90% of US energy, it requires a large and
piezoelectric materials are either toxic or nondegradable. Thus, bulky array of external US transducers, a complicated energy-

ENGINEERING
implanted devices employing these materials raise a significant focusing operation, and a tedious MRI (magnetic resonance im-
concern in terms of safety issues and often require an invasive aging) monitoring procedure (11, 12). This extensive process would
removal surgery, which can damage directly interfaced tissues/
be useful for a single treatment like viral gene delivery-based ap-
organs. Here, we present a strategy for materials processing, de-
proaches (14). However, in certain applications such as chemo-
vice assembly, and electronic integration to 1) create biodegrad-
able and biocompatible piezoelectric PLLA [poly(L-lactic acid)]
therapy, research has shown the opening of the BBB requires
nanofibers with a highly controllable, efficient, and stable piezoelec- repetitive treatment (15). As such, implanted US transducers (e.g.,
tric performance, and 2) demonstrate device applications of this nano- Sonocloud) have emerged as an excellent alternative, which can
material, including a highly sensitive biodegradable pressure sensor repeatedly induce low-intensity sonication deep inside brain tissue
for monitoring vital physiological pressures and a biodegradable ultra- at a precise location to open the BBB without causing any damage
sonic transducer for blood–brain barrier opening that can be used to to the surrounding brain tissue (10, 15). Unfortunately, all of the
facilitate the delivery of drugs into the brain. These significant appli- aforementioned pressure sensors and US transducers rely on
cations, which have not been achieved so far by conventional piezo- conventional piezoelectric materials such as PZT (lead zirconate
electric materials and bulk piezoelectric PLLA, demonstrate the PLLA titanate), PVDF (polyvinylidene fluoride), ZnO (zinc oxide), etc.,
nanofibers as a powerful material platform that offers a profound which are either toxic and/or nondegradable. Thus, these piezo-
impact on various medical fields including drug delivery, tissue engi- electric devices pose significant concerns regarding safety after
neering, and implanted medical devices.
Significance
|
biodegradable piezoelectric | PLLA | pressure sensors |
ultrasound transducer
This paper describes a powerful biodegradable and biocompatible
piezoelectric nanofiber platform for significant medical implant

T he ability to create biodegradable and biocompatible piezo-

electric materials and small-scale devices (1, 2) could bring
about significant biomedical applications, ranging from medical
applications, including a highly sensitive, wireless, biodegradable
force sensor for the monitoring of physiological pressures,
and a biodegradable ultrasonic transducer for the delivery of
sensing (3), tissue engineering (4), and cellular manipulation (5), drugs across the blood–brain barrier. Built upon materials
to controlled drug delivery (6). commonly utilized in medical implants, the devices can self-
Piezoelectric materials, a type of “smart” material that gen- degrade, causing no harm to the body, and avoid any invasive
erates electricity while deforming and vice versa (1), are used in removal surgeries.
many important implantable medical devices such as sensors,
transducers, and actuators. Piezoelectric sensors have been used Author contributions: E.J.C., T.T.L., and T.D.N. designed research; E.J.C., T.T.L., R.D., K.K.,
E.M.S., D.P., M.T.C., K.T.M.T., J.B., E.R.B., B.K., A.G., X.X., L.Y., J.F., M.D.M.-A., and I.-P.C.
along with medical catheters inside the body to monitor important
performed research; D.R., L.Y., X.T.C., J.P., and T.D.N. contributed new reagents/analytic
physiological pressures such as intracranial pressure (7), blood tools; E.J.C., T.T.L., Q.W., X.T.C., J.P., and T.D.N. analyzed data; and E.J.C., T.T.L., and T.D.N.
pressure (8), bladder pressure (9), etc. More recently, researchers wrote the paper.
have developed implanted piezoelectric ultrasonic transducers to The authors declare no competing interest.
disrupt the blood–brain barrier (BBB) and facilitate the delivery of This article is a PNAS Direct Submission. S.M. is a guest editor invited by the
drugs into the brain (10). The BBB, which is composed of tight Editorial Board.
Downloaded at UNIVERSITEIT TWENTE on December 24, 2019

junctions between the endothelial cells in the blood vessels of the Published under the PNAS license.
brain, prevents most therapeutics from accessing the brain tissue 1
E.J.C. and T.T.L. contributed equally to this work.
and thus is a major hurdle for the treatment of brain diseases (e.g., 2
To whom correspondence may be addressed. Email: nguyentd@uconn.edu.
cancers). There are several established methods for opening the This article contains supporting information online at https://www.pnas.org/lookup/suppl/
BBB, which include solvent, adjuvant, acoustic wave, lipidization, doi:10.1073/pnas.1910343117/-/DCSupplemental.
and osmostic pressure; ultrasound (US) or acoustic waves have

www.pnas.org/cgi/doi/10.1073/pnas.1910343117 PNAS Latest Articles | 1 of 7

 implantation and require a removal surgery, which is invasive and
deleterious to directly interfaced organs or tissues.
PLLA [poly(L-lactic acid)], a biocompatible and biodegradable
medical polymer (16–18), has been shown to exhibit piezoelec-
tricity when appropriately processed, thereby offering an excellent
platform to construct safer, biodegradable piezoelectric implants,
which can avoid problematic removal surgeries (3, 19). Previously,
our group has employed thermally stretched, compression-molded
PLLA bulk films to create a biodegradable piezoelectric force
sensor (3). However, stretched PLLA bulk films pose several
problems, including low reproducibility, film rigidity, and modest
piezoelectric constants (∼5 to 12 pC/N) (20, 21), which render
the bulk PLLA films useless for actuators, transducers or highly
sensitive pressure sensors. Recently, biodegradable amino acid
crystals (e.g., glycine) have been reported with an excellent pie-
zoelectric constant (22). However, it is challenging to fabricate
these powder-based materials into functional films and orient the
crystals in a repeatable manner to obtain a controllable piezo-
electric performance for device applications. A few researchers
have utilized electrospinning to create flexible PLLA piezo-
electric nanofiber films (23–25), but the reported works struggle
with major limitations. First, these reports lack appropriate
material processing to stabilize the nanomaterial or utilize the
shear-piezoelectric mode (i.e., d14) of PLLA for an optimal pi- Fig. 1. PLLA nanofibers with highly controllable and excellent piezoelectric
ezoelectric performance. Consequently, the PLLA nanofibers performance for biodegradable implanted piezoelectric devices. The image
can only produce small, unstable electrical signals under applied at Top is a simplified schematic of the treated piezoelectric PLLA nanofibers.
force (23). Second, the measured electrical signals are often The image at Bottom Left is the schematic of a biodegradable pressure
mixed with other noises caused by friction between the rough sensor and ultrasound (US) transducer. The image at Bottom Right is a
nanofiber film and metal electrodes, commonly known as the schematic illustrating the biodegradable US transducer, implanted inside the
triboelectric effect (24). Third, there is no report on the ability to brain, which can repeatedly induce US to open the blood–brain barrier (BBB)
and facilitate the delivery of drugs into the brain.
control the piezoelectric performance of the PLLA nanofibers.
These major drawbacks collectively restrict applications of this
nanomaterial. As a result, there are only a few reported appli- seen by the DSC (differential scanning calorimetry) (SI Appen-
cations of piezoelectric PLLA nanofibers for nondegradable and dix, Fig. S2). Therefore, the samples were carefully annealed and
nonimplantable force sensors or energy harvesters. slowly cooled down in two serial steps at 105 and 160.1 °C to
Here, we present a strategy for materials processing, device
improve the crystallinity (SI Appendix, Methods). After these
assembly, and electronic integration to 1) achieve biodegradable
annealing processes, the crystallinities of the processed nanofiber
and biocompatible piezoelectric PLLA nanofibers with a highly
samples appear to be in about the same range of 70 to 88% (see
controllable, efficient, and stable piezoelectric performance, and
DSC data of Fig. 2A). The nanofiber films, collected at smaller
2) demonstrate biodegradable, safe piezoelectric devices built
spin speeds, have lower levels of fiber alignment. Therefore, we
upon this powerful nanomaterial (Fig. 1). First, we show that a
chose the 300 rpm electrospun PLLA sample as a negative
biodegradable force sensor, made with the PLLA nanofiber film,
possesses higher sensitivity and flexibility than that of the control due to its lower crystallinity and poor fiber orientation
reported thermally stretched PLLA bulk film (3) and can be used (SI Appendix, Fig. S3), which results in little-to-no piezoelectric
to wirelessly monitor vital physiological pressures. Second, we effect. XRD (X-ray diffraction) data in SI Appendix, Fig. S4,
demonstrate the same PLLA nanofiber sensor acts as a bio- show that all of the samples are predominantly (200) and (110)
degradable ultrasonic transducer that can be implanted into the crystal phases, indicating the presence of β-form crystal struc-
brain to open the BBB and safely self-degrade, causing no harm to tures (33), which is the piezoelectric phase of PLLA (3). Addi-
the body. Despite several achievements in the field of biodegrad- tionally, as seen in the 2D XRD images of Fig. 2B and SI
able electronics (26–30), this report introduces a biodegradable, Appendix, Fig. S3B, electrospinning with a faster collector speed
highly efficient piezoelectric US transducer, which is only made of improves the orientation of the crystal domains in each nano-
materials commonly utilized in medical implants to facilitate the fiber. The fiber alignment over the entire film also appeared to
BBB opening for the delivery of drugs into the brain. increase with faster collector speeds, as seen in the scanning
In order to improve the piezoelectric response of PLLA, the electron microscopy (SEM) images (Fig. 2C and SI Appendix,
two major material properties that need to be addressed are the Fig. S3C). However, macroscopic orientation of the PLLA fibers
crystallinity and orientation of the polymer chains. By improving and the molecular alignment are also related to the jet speed
these properties, the carbon–oxygen double bonds (C=O) pre- (dictated by applied voltage used for electrospinning) (SI Ap-
sent in the helical PLLA backbone become aligned resulting in pendix, Fig. S5). By tuning the jet speed to match the drum
an inherent net polarization, and a well-documented shear pie- speed, the optimal piezoelectric PLLA film can be generated.
zoelectric response under applied force (31, 32). The PLLA Estimation of the crystallinity (using DSC) and Herman’s ori-
nanofibers are made using a custom electrospinning setup, as entation factor (using 2D XRD) for the electrospun PLLA
seen in SI Appendix, Fig. S1. The speed of the rotating drum was samples is described in Fig. 2D, which shows that improving the
Downloaded at UNIVERSITEIT TWENTE on December 24, 2019

varied from 300 to 4,000 rpm, while other parameters such as collector drum speed generally results in higher crystallinity and
the voltage applied to the needle, distance to collector, needle crystal alignment in the nanofibers. Thus, by tailoring the col-
gauge, flow rate, and solution concentrations were held constant. lector and jet speeds, we can control the piezoelectricity of the
This resulted in PLLA nanofiber mats with different levels of nanofibers.
fiber orientation. The nanofiber samples initially made by the We then assessed the piezoelectric performance of the PLLA
electrospinning setup are highly amorphous and unstable, as nanofiber films through an impact test (i.e., generation of voltage

2 of 7 | www.pnas.org/cgi/doi/10.1073/pnas.1910343117 Curry et al.

 have the same area of 161.29 mm2 and thicknesses in the range
of 19 to 28 μm (SI Appendix, Methods). Additionally, prior to
fabrication of the sensors, all of the films are soaked in deionized
water to minimize the influence of the triboelectric effect (35).
Fig. 3A illustrates charge outputs from the PLLA samples sub-
jected to a 30-N impact force. The signals generated from the 6
treated PLLA samples clearly show the electrospun sample
collected at 4,000 rpm has the largest charge output of about
0.9 nC while the 300 rpm sample exhibits little-to no charge output
(∼0.1 nC). The highly aligned nanofiber film, collected at 3,000
and 4,000 rpm drum speeds, noticeably outperforms the bulk pie-
zoelectric PLLA film [annealed and stretched with a 3.5 draw
ratio (DR), following our previous report (3)]. All open-circuit
voltage outputs for these piezoelectric nanofiber films were
reversible when the electrode connections were swapped (SI
Appendix, Fig. S6), indicating that the PLLA is truly polarized,
and that the measured signal is minimally influenced by tri-
boelectricity. The impact measurement was also repeated using
dry films, and the resulting data was used to estimate the shear
piezoelectric coefficient (d14) for all of the samples (SI Appendix,
Fig. S7). Using the measured mechanical properties of the PLLA
films (SI Appendix, Fig. S7A), we can roughly estimate the piezo-
electric constant of the samples, based on our previously reported
model (3); the 4,000 rpm sample appears to exhibit a d14 of ∼19 pC/N,
while the conventional bulk PLLA film only exhibits a d14 of ∼12

ENGINEERING
pC/N (SI Appendix, Fig. S7B). This indicates the processing of
PLLA nanofibers significantly improves the material’s shear pie-
zoelectric response. Furthermore, our strategy of cutting the PLLA
films at 45° angles to utilize shear piezoelectricity was also justified
by comparing the charge outputs of a 0° and 45° cut film under the
same applied force (SI Appendix, Fig. S8). For the actuation mea-
surement, a treated PLLA film (1.27 cm × 1.27 cm) was sandwiched
in the center of aluminum foil electrodes (9.53 mm × 9.53 mm). A
controlled voltage waveform was then applied to the sensor, and the
displacement in the exposed right corner of the sample was mea-
sured using a laser displacement sensor (SI Appendix, Methods). As
seen in Fig. 3B, the treated PLLA nanofiber samples vibrate with
the same frequency (1 Hz) as the applied sinusoidal voltage wave-
form (20 Vpp). The 4,000 rpm electrospun sample again exhibits the
greatest displacement (∼4.5 μm), while the stretched 3.5 DR
bulk film and 300 rpm electrospun samples exhibited no
measurable displacement. This result confirms the superior
Fig. 2. Material characterization of the electrospun PLLA. (A) Results from piezoelectric performance of the highly aligned nanofiber film. In ad-
differential scanning calorimetry (DSC) of electrospun PLLA nanofiber films dition, as the amplitude of the applied voltage increases, the amplitude
collected at different spin speeds. (B) The 2D X-ray diffraction (2D XRD) of displacement for the electrospun films also increases, and the
images show orientation of crystal domains inside the electrospun PLLA
displacement is frequency dependent (Fig. 3C and SI Appendix,
nanofibers, made with different collection speeds. (C) Scanning electron
microscopy (SEM) images show PLLA nanofiber alignment with different
Fig. S9). Piezoelectric performance in the treated PLLA nanofiber
collection speeds. (Scale bars, 40 μm.) (D) Graphical summary illustrating the film is also stable. This advantage is significant as there has been
trend that, as the PLLA nanofibers are collected at faster speeds, the Herman little research to avoid depolarization of the PLLA nanofibers
orientation factor (i.e., crystal alignment) and crystallinity percentage gen- over time (25, 32, 34). Indeed, as seen in Fig. 3D, only an elec-
erally increase. trospun sample (3,000 rpm) that underwent our full annealing
processes (i.e., annealed at both 105 and 160.1 °C) has a stable
piezoelectric output under the same applied force (∼30 N) for
under impact force) and an actuation test (i.e., displacement 7 d, with a marginal loss (∼6%) in signal at 14 d. In contrast, the
under an applied electric field) (3, 20). To create the PLLA untreated (i.e., not annealed) and partially treated (i.e., annealed
sensor for these tests, we fully annealed and cut the PLLA film at only at 105 °C) samples rapidly lose their performance and are
a 45° angle relative to the fiber direction to utilize shear piezo- therefore not stable for long-term implant applications.
electricity by maximizing shear force under an applied normal After verifying the piezoelectric effect of the PLLA nano-
force. The fully treated and cut PLLA films possess a stable, fibers, we create a biodegradable force sensor by using the
efficient, and highly controllable piezoelectric performance, nanofibers, molybdenum (Mo) electrodes, and encapsulating
which has not been achieved by previous reports for the PLLA untreated PLLA layers (Fig. 1). PLLA or PLA are common
nanofibers (25, 32, 34). We then sandwiched the films between biodegradable polymers used in Food and Drug Administration
Downloaded at UNIVERSITEIT TWENTE on December 24, 2019

aluminum foil electrodes and Kapton tape (SI Appendix, Meth- (FDA)-approved implanted tissue scaffolds, bone screws, and
ods). For impact testing, the PLLA sensor is subjected to a drug carriers (36). Molybdenum is a common nutrient (37) and a
consistent force induced by an actuator, which is integrated with biodegradable metal, used extensively in FDA-approved hip and
a dynamic force sensor and driven by a defined voltage wave- joint implants (38). Although we previously reported on a bio-
form. The charge output from the PLLA sample is measured degradable piezoelectric force sensor, the device is based on the
with an electrometer (SI Appendix, Methods). All of the sensors stretched PLLA bulk film, which is less flexible, exhibits much

Curry et al. PNAS Latest Articles | 3 of 7

 lower piezoelectric performance, and consequently offers lower sen-
sitivity for force detection. Fig. 4A clearly illustrates this by showing
that the slope of a calibration curve for a biodegradable sensor made
with a treated 4,000 rpm electrospun PLLA film is 1.8 times steeper
than that of a sensor, using a conventional thermally stretched, bulk
PLLA film (DR = 3.5), reported in our previous work (3). However,
when compared to a sensor made of a common nondegradable
piezoelectric PVDF-TrFE film (which exhibits a higher d33 of
approximately −34 pC/N), the biodegradable 4,000 rpm electro-
spun PLLA sensor appears to produce lower-amplitude signals
under the same applied force (SI Appendix, Fig. S10). We also
show that the charge output from the same 4,000 rpm sensor is stable
for over 10,000 cycles of the same impact force (10 N). Not only is
the 4,000 rpm electrospun film more sensitive, but its higher crys-
tallinity does not appear to result in any significant changes to the
degradation rate when compared to the bulk PLLA piezoelectric
film (SI Appendix, Fig. S11). Without being encapsulated, the PLLA
films exhibit a reduction in piezoelectricity after being exposed to
aqueous environments due to plastic deformation under an applied
load (SI Appendix, Fig. S12). We then demonstrate the use of the
nanofiber sensor to monitor intraabdominal pressure in a mouse.
The sensor (5 × 5 mm) is fully implanted into the abdominal cavity of
a mouse and connected to a small printed circuit board (PCB) via a
subcutaneous (s.c.) biodegradable wire made of Mo and coated in
PLLA (SI Appendix, Methods). The PCB contains a charge ampli-
fying circuit, a wireless near-field communication (NFC) chip and a
commercial antenna (SI Appendix, Methods). The entire PCB is
sealed inside an 18 × 14-mm PDMS (polydimethylsiloxane) box and
s.c. implanted at the back of the animal (Fig. 4C). Thus, the ab-
dominal sensor and the connecting wires can self-degrade, while the
nondegradable PCB could be easily removed at the end of the
sensor’s lifetime in a minimally invasive manner. The mouse’s ab-
domen, filled with saline solution, is then manually stimulated to
generate an internal fluid pressure, which mimics a change in
intraabdominal pressure (SI Appendix, Methods). A clearly distin-
guishable signal (Fig. 4D) is wirelessly measured while the mouse’s
abdomen is periodically depressed and relaxed. The measured
pressure signal was then compared to the signal generated by a
300 rpm PLLA sensor (negative control) to verify the signal was not
generated by triboelectricity and motion artifacts of the wires (Fig.
4D). These results clearly demonstrate the potential of the bio-
degradable PLLA sensor for monitoring vital physiological pressures
inside the body.
In addition to monitoring intraabdominal pressure, we dem-
onstrate that the same PLLA nanofiber sensor can also be used
as a biodegradable US transducer. We first tested the PLLA
nanofibers’ ability to transmit or receive ultrasonic waves. During
US transmission testing (Fig. 5 A, Inset), a capsule hydrophone
was used to measure the acoustic pressure. The PLLA device was
driven by a function generator to produce a continuous ultra-
sonic wave at 1 MHz (SI Appendix, Methods). As seen in Fig. 5A,
there was no signal detected when the function generator is
“off.” When the generator was “on,” all PLLA transducers
generated distinct acoustic waves while the 300 rpm sample
(negative-control sample, nonpiezoelectric) resulted in only
noise (SI Appendix, Fig. S13). The trend is similar in the US
receiving test, as illustrated in SI Appendix, Fig. S14; in all of
these experiments, the highly aligned 4,000 rpm sample provided
the highest conversion signals. Interestingly, the PLLA trans-
ducers can act as speakers to generate audible sounds and even
Fig. 3. Piezoelectric characterization of the treated PLLA nanofiber films. play music (Movies S1–S3). We also conducted a degradation
(A) Charge output from stretched, bulk piezo-PLLA (yellow) and treated experiment and demonstrated that a transducer, using encap-
Downloaded at UNIVERSITEIT TWENTE on December 24, 2019

electrospun PLLA, collected at different speeds, under the same impact

sulating layers of untreated PLLA (100 μm thick), can have a
force. (B) Displacement of stretched, bulk PLLA (yellow) and treated elec-
trospun PLLA, collected at different speeds, under the same voltage (20 Vpp)
lifetime of up to 8 d in phosphate buffer saline (PBS) at 37 °C
at 1 Hz. (C) Displacement of 300 rpm PLLA negative-control sample (red) and (Fig. 5B) (SI Appendix, Methods). Longer functional lifetimes can
4,000 rpm PLLA (black), under increasing magnitudes of voltage at 1 Hz. (D) certainly be achieved by engineering the properties [i.e., thick-
Comparison of the piezoelectric performance for 3,000 rpm electrospun ness or molecular weight (MW)] of the encapsulating PLLA
PLLA samples annealed under different conditions over a 14-d period. layers (3) or using other biodegradable encapsulating polymers.

4 of 7 | www.pnas.org/cgi/doi/10.1073/pnas.1910343117 Curry et al.

 ENGINEERING
Fig. 4. Wireless, biodegradable PLLA-nanofiber force sensor. (A) Comparison of calibration curves for a biodegradable sensor using stretched, bulk piezo-
PLLA film (black) and a 4,000 rpm electrospun PLLA nanofiber film (red). Inset shows the optical image of the biodegradable and flexible force sensor, made
from the PLLA nanofibers. (Scale bar, 5 mm.) (B) Output from a charge amplifying circuit connected to a 4,000 rpm electrospun, biodegradable PLLA sensor
that is subjected to 10,000 cycles of a 10-N force. (C) Simplified schematic of the implanted, wireless pressure sensor in a mouse (Left) and optical image of a
mouse receiving the wireless PLLA sensor implanted (Right). NFC, near-field communication chip. (D) Comparison of the simulated abdominal pressure signals,
wirelessly recorded from an implanted biodegradable PLLA nanofiber sensor using a 300 rpm negative control (black) and a 4,000 rpm film (red).

Importantly, for a proof-of-concept on a potential application of biodegradable device for delivering drugs through the BBB,
the biodegradable transducer, we employ the PLLA device, made another in vivo animal model was performed. The procedure of
of 4,000 rpm nanofiber samples, for disruption of the BBB in vivo. this experiment is similar to the previous experiment except
The experiment is illustrated in Fig. 6A. A 5 mm × 5 mm bio- that the dextran (3 kDa, FITC, Lysine Fixable; Thermo Fisher)
degradable US transducer, which is connected to flexible, bio- as a drug model was retro-orbitally injected into the mice after
degradable PLLA-encapsulated Mo wires, was placed on a the sonication process (SI Appendix, Methods and Fig. S20).
craniotomy defect in a mouse skull (SI Appendix, Methods). The Additionally, another control group in which mice did not re-
spatial pressure field of the biodegradable transducer is provided ceive the microbubbles before sonication was added to this
in SI Appendix, Fig. S15. The transducer was operated at 1 MHz to experiment in order to validate the effect of microbubbles in
generate an acoustic pressure of 0.3 MPa (rarefaction pressure the BBB opening. As seen in Fig. 6 E, Left, a remarkable level
value) in a series of 2 shots lasting 30 s, with a 30-s break between of green signal (FITC) was found around the microvessels in
each shot (SI Appendix, Fig. S16). The device functioned well in its the brain of mice that received the treatments by the 4,000 rpm
predefined lifetime and eventually self-degraded (Fig. 6B). We transducer and microbubbles. It is noticeable that the intensity
processed the brains for fluorescence analysis of bloodborne ele- of the FITC signal is reduced at deeper areas of the brain. On
ments to gauge leakage of the BBB (SI Appendix, Methods). Two the other hand, no green signal was detected from the same
indicators of leakage were intentionally chosen in order to reflect
the relative degree of BBB disruption. Tissue autofluorescence at
488 nm is associated with the presence of the 64.5 kDa (in MW)
blood protein hemoglobin (39, 40), which has been suggested to
leak across a disrupted BBB (41). As seen in Fig. 6C, a noticeable
halo of autofluorescence (green stain) could be seen around var-
ious microvessels (red stain) in the brains of mice sonicated by the
4,000 rpm transducer. In contrast, no similar signal was observed
from the same coronal sections (C2) of the control mouse, soni-
cated by the 300 rpm nonpiezoelectric control sample. Additional
brain sections of the control mouse (receiving the nonpiezoelectric
device) are provided in SI Appendix, Fig. S17. As further illus-
trated in Fig. 6D and SI Appendix, Fig. S18, for the mice that
received US treatment, the closer the coronal sections are to the
Downloaded at UNIVERSITEIT TWENTE on December 24, 2019

implanted transducer, the more disrupted vessels are associated

Fig. 5. US characterization of the biodegradable PLLA-nanofiber trans-
with autofluorescent signals. This serves as an internal control and
ducer. (A) The output pressure from the transducer with different electro-
clearly shows the local US-induced BBB opening. The BBB spinning speeds under the same input voltage. The Inset is the simplified
opening was again confirmed by immunofluorescence analysis on schematic of the experiment. (B) Output pressure from a biodegradable US
the leakage of the serum protein IgG (∼150 kDa) (SI Appendix, transducer made from 4,000 rpm electrospun PLLA under the same input
Fig. S19). To further certify the potential application of the voltage after different days in PBS at 37 °C.

Curry et al. PNAS Latest Articles | 5 of 7

 Fig. 6. In vivo experiment to demonstrate the application of PLLA nanofiber transducer for the BBB opening and drug delivering. (A) The schematic (Left)
and optical image (Right) of the in vivo experiment. (B) The optical images of a typical biodegradable US transducer at different days in the buffered solution
at an accelerated-degradation temperature of 70 °C. (Scale bars, 5 mm.) (C) Representative images showing the autofluorescent signal of blood protein at the
coronal section (C2) from the brains of mice that received US from the 4,000 rpm PLLA transducer (Left) and the 300 rpm PLLA negative-control transducer
(Right). (D) Representative images show the blood protein signal at different coronal sections of the same mouse brain receiving the US treatment. Section C3
(Right) is closer to the implanted transducer, while section C1 (Left) is far away from the implanted US transducer, serving as an internal control. (Scale bars in
C and D, 30 μm.) (E) Representative images show the signal of dextran (FITC) at the coronal sections from the brains of mice that received different treatments
and samples. The dashed lines show the boundary between the brain and the biodegradable device. The asterisk (*) shows the position of the implanted
device. (Scale bars, 50 μm.)

coronal sections of the two control samples, Fig. 6 E, Center of mice and the intracranial cavity of rats (SI Appendix, Methods)
and Right. Additional representative images of the brain sec- for histology analysis. The histological images from both exper-
tions in the experimental group, as well as the control groups, iments showed that the device elicits minimal fibrosis and immune
are provided in SI Appendix, Figs. S21 and S22. If higher output response after implantation for 2 and 4 wk, as seen in SI Appendix,
acoustic pressure and wireless powering are needed for the US Figs. S28–S30. Collectively, these results illustrate that the bio-
Downloaded at UNIVERSITEIT TWENTE on December 24, 2019

transducer, we can fabricate the device with multiple layers of degradable PLLA transducer can be implanted safely into the brain
PLLA nanofiber films and utilize the inductive coupling effect, to locally and effectively open the BBB, which could facilitate the
as demonstrated in SI Appendix, Figs. S23–S25, and SI Ap- delivery of drugs into the brain for the treatment of various brain
pendix, Figs. S26 and S27, respectively. diseases or disorders. Built upon materials commonly utilized in
Finally, to demonstrate the biocompatibility of the PLLA medical implants, the transducer can self-degrade, causing no harm
nanofiber devices, we implanted these devices s.c. into the backs to the body, and avoid any invasive brain surgery for removal.

6 of 7 | www.pnas.org/cgi/doi/10.1073/pnas.1910343117 Curry et al.

 Conclusions engineering and drug delivery to medical implanted sensors/
We have presented a strategy for materials processing, device transducers/actuators.
assembly, electronic integration, and piezoelectric stimulation
to 1) create biodegradable piezoelectric PLLA nanofibers with Materials and Methods
stable, effective, and highly controllable piezoelectric perfor- Details of fabrications and characterizations of the PLLA nanofiber films, the
mance, and 2) demonstrate implanted-device applications of force sensor, and the US transducer along with in vitro and in vivo experiments
this powerful nanomaterial, including a biodegradable, wireless, all appear in SI Appendix. Animal procedures are approved and performed
following Institutional Animal Care and Use Committee guidelines at
highly sensitive pressure sensor and a biodegradable ultrasonic
the University of Connecticut and University of Connecticut Health Center.
transducer. All of these piezoelectric implants employ only common
biodegradable and biocompatible materials that can safely self- Data Availability Statement. All data supporting the findings of this study are
degrade, avoiding undesired invasive removal surgeries. In future available within the article and/or its supplementary materials.
applications, these piezoelectric devices could be used to harvest
energy from normal body motions, thereby eliminating the need for ACKNOWLEDGMENTS. The project is in part funded by NIH Grant 1R21AR075196-
batteries often required by conventional medical implants. The 01. We thank Mark Bouley and Peter Glaude for their help with the de-
PLLA nanofibers could also serve as a tissue scaffold that can sign and fabrication of the electrospinning box; Dr. Laura Pinatti for
generate electrical stimulation under body motions to promote assistance with DSC; Adam Wentworth for assistance with mechanical
testing; Sanja Novak for her help with the retro-orbital injection; Lisa
tissue regeneration. We therefore anticipate the PLLA nanofibers, Chubba and Shaelyn Killoh for their assistance with the tail vein injec-
and the biodegradable piezoelectric devices described herein can tion; and Dr. Yusuf Khan for help with our initial characterization of
profoundly impact various fields of medicine, ranging from tissue acoustic pressure.

1. M. T. Chorsi et al., Piezoelectric biomaterials for sensors and actuators. Adv. Mater. 31, 21. M. Ando et al., “New human machine interface devices using a piezoelectric
e1802084 (2019). poly(L-lactic acid) film” in 2013 IEEE International Symposium on the Applications
2. T. D. Nguyen et al., Piezoelectric nanoribbons for monitoring cellular deformations. of Ferroelectric and Workshop on the Piezoresponse Force Microscopy (ISAF/PFM)
Nat. Nanotechnol. 7, 587–593 (2012). (IEEE, 2013), pp. 236–239.
3. E. J. Curry et al., Biodegradable piezoelectric force sensor. Proc. Natl. Acad. Sci. U.S.A. 22. S. Guerin et al., Control of piezoelectricity in amino acids by supramolecular packing.

ENGINEERING
115, 909–914 (2018). Nat. Mater. 17, 180–186 (2018).
4. Y. Ikada, Y. Shikinami, Y. Hara, M. Tagawa, E. Fukada, Enhancement of bone for- 23. G. Zhao et al., Electrospun poly(L-lactic acid) nanofibers for nanogenerator and di-
mation by drawn poly(L-lactide). J. Biomed. Mater. Res. 30, 553–558 (1996). agnostic sensor applications. Macromol. Mater. Eng. 302, 1600476 (2017).
5. Y. Tajitsu, M. Kanesaki, M. Tsukiji, K. Imoto, M. Date, E. Fukada, Novel tweezers for 24. A. Sultana et al., Human skin interactive self-powered wearable piezoelectric bio-e-
biological cells using piezoelectric polylactic acid fibers. Ferroelectrics 320, 133–139 skin by electrospun poly-L-lactic acid nanofibers for non-invasive physiological signal
(2005). monitoring. J. Mater. Chem. B 5, 7352–7359 (2017).
6. Q. Cui, C. Liu, X. F. Zha, Study on a piezoelectric micropump for the controlled drug 25. S. J. Lee, A. P. Arun, K. J. Kim, Piezoelectric properties of electrospun poly(L-lactic acid)
delivery system. Microfluid. Nanofluidics 3, 377–390 (2007). nanofiber web. Mater. Lett. 148, 58–62 (2015).
7. C. Li, P.-M. Wu, L. A. Shutter, R. K. Narayan, Dual-mode operation of flexible piezo- 26. C. M. Boutry et al., Biodegradable and flexible arterial-pulse sensor for the wireless
electric polymer diaphragm for intracranial pressure measurement. Appl. Phys. Lett. monitoring of blood flow. Nat. Biomed. Eng. 3, 47–57 (2019).
96, 053502 (2010). 27. C. M. Boutry et al., A stretchable and biodegradable strain and pressure sensor for
8. C. Dagdeviren et al., Conformable amplified lead zirconate titanate sensors with orthopaedic application. Nat. Electron. 1, 314–321 (2018).
enhanced piezoelectric response for cutaneous pressure monitoring. Nat. Commun. 5, 28. C. M. Boutry et al., A sensitive and biodegradable pressure sensor array for cardio-
4496 (2014).
vascular monitoring. Adv. Mater. 27, 6954–6961 (2015).
9. R. Woltjer et al., (2016) “Optimization of piezo-MEMS layout for a bladder monitor”
29. J. K. Chang et al., Biodegradable electronic systems in 3D, heterogeneously in-
in 2016 IEEE International Ultrasonics Symposium (IUS) (IEEE, 2016), pp. 1–4.
tegrated formats. Adv. Mater. 30, 1704955 (2018).
10. A. Idbaih et al., Safety and feasibility of repeated and transient blood–brain barrier
30. S.-K. Kang et al., Bioresorbable silicon electronic sensors for the brain. Nature 530, 71–
disruption by pulsed ultrasound in patients with recurrent glioblastoma. Clin. Cancer
76 (2016).
Res. 25, 3793–3801 (2019).
31. M. Yoshida, T. Onogi, K. Onishi, T. Inagaki, Y. Tajitsu, High piezoelectric performance
11. N. McDannold, N. Vykhodtseva, S. Raymond, F. A. Jolesz, K. Hynynen, MRI-guided
of poly (lactic acid) film manufactured by solid-state extrusion. Jpn. J. Appl. Phys. 53,
targeted blood-brain barrier disruption with focused ultrasound: Histological findings
09PC02 (2014).
in rabbits. Ultrasound Med. Biol. 31, 1527–1537 (2005).
32. M. Smith, Y. Calahorra, Q. Jing, S. Kar-Narayan, Direct observation of shear piezo-
12. M. Kinoshita, N. McDannold, F. A. Jolesz, K. Hynynen, Noninvasive localized delivery
electricity in poly-L-lactic acid nanowires. APL Mater. 5, 074105 (2017).
of Herceptin to the mouse brain by MRI-guided focused ultrasound-induced blood-
33. J.-F. Ru et al., Dominant β-form of poly(L-lactic acid) obtained directly from melt under
brain barrier disruption. Proc. Natl. Acad. Sci. U.S.A. 103, 11719–11723 (2006).
shear and pressure fields. Macromolecules 49, 3826–3837 (2016).
13. N. Vykhodtseva, N. McDannold, K. Hynynen, Progress and problems in the application
34. V. Sencadas et al., Local piezoelectric activity of single poly(L-lactic acid) (PLLA) mi-
of focused ultrasound for blood-brain barrier disruption. Ultrasonics 48, 279–296
crofibers. Appl. Phys. A 109, 51–55 (2012).
(2008).
35. T. Wartzek, T. Lammersen, B. Eilebrecht, M. Walter, S. Leonhardt, Triboelectricity in
14. J. O. Szablowski, A. Lee-Gosselin, B. Lue, D. Malounda, M. G. Shapiro, Acoustically
targeted chemogenetics for the non-invasive control of neural circuits. Nat. Biomed. capacitive biopotential measurements. IEEE Trans. Biomed. Eng. 58, 1268–1277
Eng. 2, 475–484 (2018). (2011).
15. C. Horodyckid et al., Safe long-term repeated disruption of the blood-brain barrier 36. H. H. Peltoniemi, D. Hallikainen, T. Toivonen, P. Helevirta, T. Waris, SR-PLLA and SR-
using an implantable ultrasound device: A multiparametric study in a primate model. PGA miniscrews: Biodegradation and tissue reactions in the calvarium and dura ma-
J. Neurosurg. 126, 1351–1361 (2017). ter. J. Craniomaxillofac. Surg. 27, 42–50 (1999).
16. D. da Silva et al., Biocompatibility, biodegradation and excretion of polylactic acid 37. J. A. Novotny, J. R. Turnlund, Molybdenum intake influences molybdenum kinetics in
(PLA) in medical implants and theranostic systems. Chem. Eng. J. 340, 9–14 (2018). men. J. Nutr. 137, 37–42 (2007).
17. S. Lee et al., Lactic acid assisted fabrication of bioactive three-dimensional PLLA/β-TCP 38. W.-C. Witzleb, J. Ziegler, F. Krummenauer, V. Neumeister, K.-P. Guenther, Exposure to
fibrous scaffold for biomedical application. Chem. Eng. J. 347, 771–781 (2018). chromium, cobalt and molybdenum from metal-on-metal total hip replacement and
18. S. Farah, D. G. Anderson, R. Langer, Physical and mechanical properties of PLA, and hip resurfacing arthroplasty. Acta Orthop. 77, 697–705 (2006).
their functions in widespread applications—a comprehensive review. Adv. Drug Deliv. 39. C. G. Atkins, K. Buckley, M. W. Blades, R. F. B. Turner, Raman spectroscopy of blood
Rev. 107, 367–392 (2016). and blood components. Appl. Spectrosc. 71, 767–793 (2017).
19. Y. Tajitsu, Fundamental study on improvement of piezoelectricity of poly(ι-lactic acid) 40. T. Kobayashi, K. Nakata, I. Yajima, M. Kato, H. J. B. Tsurui, Label-free imaging of
and its application to film actuators. IEEE Trans. Ultrason. Ferroelectr. Freq. Control melanoma with confocal photothermal microscopy: Differentiation between malig-
60, 1625–1629 (2013). nant and benign tissue. Bioeng. 5, 67 (2018).
20. M. Ando et al., Pressure-sensitive touch panel based on piezoelectric poly (l-lactic 41. A. Lewin et al., Free serum haemoglobin is associated with brain atrophy in secondary
acid) film. Jpn. J. Appl. Phys. 52, 09KD17 (2013). progressive multiple sclerosis. Wellcome Open Res. 1, 10 (2016).
Downloaded at UNIVERSITEIT TWENTE on December 24, 2019

Curry et al. PNAS Latest Articles | 7 of 7