Access to this full-text is provided by Frontiers.
Content available from Frontiers in Neuroscience
This content is subject to copyright.
fnins-15-771980 December 16, 2021 Time: 12:57 # 1
ORIGINAL RESEARCH
published: 22 December 2021
doi: 10.3389/fnins.2021.771980
Edited by:
Wei Tong,
The University of Melbourne, Australia
Reviewed by:
Flavia Vitale,
University of Pennsylvania,
United States
Noe Alvarez,
University of Cincinnati, United States
Elissa Welle,
University of Michigan, United States
*Correspondence:
Xiaojie Duan
xjduan@pku.edu.cn
Specialty section:
This article was submitted to
Neural Technology,
a section of the journal
Frontiers in Neuroscience
Received: 07 September 2021
Accepted: 29 November 2021
Published: 22 December 2021
Citation:
Fu X, Li G, Niu Y, Xu J, Wang P,
Zhou Z, Ye Z, Liu X, Xu Z, Yang Z,
Zhang Y, Lei T, Zhang B, Li Q, Cao A,
Jiang T and Duan X (2021)
Carbon-Based Fiber Materials as
Implantable Depth Neural Electrodes.
Front. Neurosci. 15:771980.
doi: 10.3389/fnins.2021.771980
Carbon-Based Fiber Materials as
Implantable Depth Neural Electrodes
Xuefeng Fu1, Gen Li1, Yutao Niu2,3 , Jingcao Xu4, Puxin Wang1,5 , Zhaoxiao Zhou1,5 ,
Ziming Ye4, Xiaojun Liu1, Zheng Xu1, Ziqian Yang1,5, Yongyi Zhang2,3, Ting Lei4,
Baogui Zhang6, Qingwen Li2,3 , Anyuan Cao4, Tianzai Jiang6and Xiaojie Duan1,5,7*
1Department of Biomedical Engineering, College of Future Technology, Peking University, Beijing, China, 2School
of Nano-Tech and Nano-Bionics, University of Science and Technology of China, Hefei, China, 3Key Laboratory
of Multifunctional Nanomaterials and Smart Systems, Advanced Materials Division, Suzhou Institute of Nano-Tech
and Nano-Bionics, Chinese Academy of Sciences (CAS), Suzhou, China, 4School of Materials Science and Engineering,
Peking University, Beijing, China, 5Academy for Advanced Interdisciplinary Studies, Peking University, Beijing, China,
6Brainnetome Center, Institute of Automation, Chinese Academy of Sciences (CAS), Beijing, China, 7National Biomedical
Imaging Center, Peking University, Beijing, China
Implantable brain electrophysiology electrodes are valuable tools in both fundamental
and applied neuroscience due to their ability to record neural activity with high
spatiotemporal resolution from shallow and deep brain regions. Their use has been
hindered, however, by the challenges in achieving chronically stable operations.
Furthermore, implantable depth neural electrodes can only carry out limited data
sampling within predefined anatomical regions, making it challenging to perform
large-area brain mapping. Minimizing inflammatory responses and associated gliosis
formation, and improving the durability and stability of the electrode insulation layers
are critical to achieve long-term stable neural recording and stimulation. Combining
electrophysiological measurements with simultaneous whole-brain imaging techniques,
such as magnetic resonance imaging (MRI), provides a useful solution to alleviate
the challenge in scalability of implantable depth electrodes. In recent years, various
carbon-based materials have been used to fabricate flexible neural depth electrodes
with reduced inflammatory responses and MRI-compatible electrodes, which allows
structural and functional MRI mapping of the whole brain without obstructing any brain
regions around the electrodes. Here, we conducted a systematic comparative evaluation
on the electrochemical properties, mechanical properties, and MRI compatibility of
different kinds of carbon-based fiber materials, including carbon nanotube fibers,
graphene fibers, and carbon fibers. We also developed a strategy to improve the
stability of the electrode insulation without sacrificing the flexibility of the implantable
depth electrodes by sandwiching an inorganic barrier layer inside the polymer insulation
film. These studies provide us with important insights into choosing the most suitable
materials for next-generation implantable depth electrodes with unique capabilities for
applications in both fundamental and translational neuroscience research.
Keywords: brain activity mapping, multi-modal neural interfacing, soft bioelectronics, carbon nanomaterials,
biocompatibility
Frontiers in Neuroscience | www.frontiersin.org 1December 2021 | Volume 15 | Article 771980
fnins-15-771980 December 16, 2021 Time: 12:57 # 2
Fu et al. Carbon Neural Electrodes
INTRODUCTION
Implantable depth neural electrodes constitute the basis for
a wide range of applications, including deciphering how
information is encoded inside the brain (Buzsáki, 2004;Stanley,
2013), treating various neurological diseases (Borchers et al.,
2012;Lozano and Lipsman, 2013;Cash and Hochberg, 2015),
and realizing brain-machine interfaces (BMIs) (Bensmaia and
Miller, 2014;Shenoy and Carmena, 2014;Choi et al., 2016;
Lebedev and Nicolelis, 2017). The capability of spatiotemporal
mapping at the single-neuron level is advantageous over
electroencephalography (EEG) or electrocorticography (ECoG)
surface probes (Kim et al., 2010;Buzsáki et al., 2012), or non-
invasive brain imaging methods such as functional magnetic
resonance imaging (fMRI) (Logothetis et al., 2001;Weiskopf
et al., 2007;Gosselin et al., 2011;Figee et al., 2013) or
functional near infrared spectroscopy (fNIR) (Izzetoglu et al.,
2005;Bunce et al., 2006;Ayaz et al., 2007, 2009;Harrison
et al., 2014). Despite this advantage, single neuronal recordings
with implantable depth electrodes remain limited in several
aspects, including the limited number of sampling sites and
challenges in achieving chronically stable operation (Vitale
et al., 2015). Combining electrophysiological measurements with
simultaneous whole-brain imaging techniques, such as fMRI,
provides a useful solution to alleviate the challenge in scalability
of implantable depth electrodes (Zhao et al., 2020). However,
many commonly used metals for implantable depth electrodes
elicit significant magnetic resonance imaging (MRI) artifact
due to the mismatch in magnetic susceptibility between metal
and water/tissues, which obstructs functional and structural
mapping of a large volume of brain tissues surrounding the
electrodes (Zhao et al., 2016). Implantable depth electrodes
with high MRI compatibility are important for combining high-
resolution electrophysiological measurements with more global
MRI mapping of brain activity for fundamental neuroscience
studies, as well as clinical evaluation and monitoring.
Under mechanical mismatch between the implantable
electrodes and brain tissue, the natural micromotion of the host
brain tissue induces intense stress at the electrode-brain interface.
This stress causes repetitive mechanical stimulation and injury
on tissues and results in sustained inflammatory responses,
leading to neuronal loss and glial scar formation around the
electrodes (Williams et al., 1999;Seymour and Kipke, 2007;
Zhong and Bellamkonda, 2008;Jeong et al., 2015). Minimizing
the chronic inflammatory tissue responses of the electrodes
is thus critical to achieve chronically stable neural recording
and stimulation. Recent research has shown that increasing
the mechanical compliance of the implantable electrodes can
effectively alleviate the chronic inflammatory responses and
reduce gliosis formation (Xie et al., 2015;Fu et al., 2016;Liu,
2018;Lu et al., 2019). The flexibility of the implantable depth
electrodes, characterized by bending stiffness Kwhich is the ratio
between the longitudinal loading force and the displacement, is
strongly dependent on the electrode size. Materials with excellent
interfacial electrochemical properties, including low electrode
impedance and high charge injection capability, are thus highly
desirable for flexible electrodes fabrication because they allow for
further electrode miniaturization while maintaining reasonable
electrode electrochemical performance.
Structural failures of neural electrodes upon implantation
constitute another prevalent failure mode of neural recording and
stimulation, in addition to neuronal degeneration and glial scar
encapsulation around the implanted electrodes. The stress at the
electrode–tissue interface can accelerate the material degradation
which results in cracking, blistering, and delamination of
the electrode insulation layers, eventually contributing to the
failure of neural recording and stimulation (Prasad et al., 2012,
2014;Gilgunn et al., 2013;Kozai et al., 2015). Thin polymer
films are commonly used insulation layers for implantable
depth electrodes due to their softness, good biocompatibility,
and readiness for conformal coating. However, these polymer
encapsulation layers showed intrinsic limitations in water
permeability and extrinsic effects associated with localized defects
(arising from the growth process, i.e., pinholes, cracks, and
grain boundaries) (Li et al., 2010, 2019;Minnikanti et al., 2014),
which may strongly affect the dielectric properties of the material
and lead to electrical leakage under long-term degradation in
physiological conditions. Increasing the thickness of the electrode
insulation layer is an effective way to improve its durability
and stability under physiological environment. However, the
thickness increase of the insulation layer is accompanied by
a significant reduction of the mechanical compliance of the
electrodes. Exploring strategies of improving the durability
and stability of the insulation under physiological conditions
without sacrificing the mechanical compliance of the electrodes
is critical to achieve long-term, stable chronic neural recording
and stimulation at high spatiotemporal resolution.
Carbon-based materials have emerged as new candidates
for implantable neural interfaces. Carbon fibers (CFs) were
used to fabricate electrodes with the assistance of poly
(3,4-ethylenedioxythiophene) (PEDOT) coating, which enabled
single-neuron recording in acute and early chronic experiments
in rats (Kozai et al., 2012;Patel et al., 2016). Electrodes made
from carbon nanotube fibers (CNTFs) were demonstrated to be
capable of continuously detecting and isolating single neuronal
units from rats in early chronic scale (Vitale et al., 2015)
and for up to 4–5 months (Lu et al., 2019) without electrode
repositioning, with greatly reduced brain inflammatory responses
as compared to their stiff metal counterparts. Graphene fibers
(GFs) electrodes were found to have a charge-injection-limit
(CIL) of ∼10 mC/cm2, which is higher than most commonly
used electrode materials for neural stimulation (Apollo et al.,
2015;Zhao et al., 2020), and capable of detecting neuronal
activity (Apollo et al., 2015) with a high signal-to-noise ratio
(SNR) of 9.2 dB (Wang et al., 2019). The CFs were also used
extensively in electrochemical neurotransmitter detection both
in vitro and in vivo (Huffman and Venton, 2009;Özel et al.,
2011;Khan et al., 2014;Durairaj et al., 2018). In addition,
electrodes made of carbon-based fibers showed greatly reduced
MRI artifacts compared with electrodes made of metals such
as PtIr (Lu et al., 2019;Zhao et al., 2020). These results
demonstrated the advantages of carbon-based fibers in building
multi-modal and chronically stable neural interfaces. In this
work, we aim to provide a systematic comparative evaluation on
Frontiers in Neuroscience | www.frontiersin.org 2December 2021 | Volume 15 | Article 771980
fnins-15-771980 December 16, 2021 Time: 12:57 # 3
Fu et al. Carbon Neural Electrodes
the electrochemical properties, mechanical properties, and MRI
compatibility of different kinds of carbon-based fiber materials,
and to explore solutions to improve the durability and stability
of the insulation layer without sacrificing the flexibility of the
implantable depth electrodes. These studies will provide us with
important insights into choosing the most suitable materials
for next-generation implantable depth electrodes with unique
capabilities for applications in both fundamental and applied
neuroscience research.
MATERIALS AND METHODS
Electrode Fabrication
In this work, we fabricated neural depth implantable electrodes
from four types of carbon-based fiber materials, including two
types of CNTFs, one type of GFs, and one type of CFs. Array CNT
fibers (aCNTFs) were dry-drawn from a vertically super-aligned
array of CNTs grown by chemical vapor deposition (CVD) on a
silicon substrate (Jia et al., 2011) (see Supplementary Methods
for details). The second type of CNTFs are floating catalyst CNT
fibers (fCNTFs), which were spun from the CNT membrane
grown with a floating catalyst CVD method using a liquid source
of carbon and an iron nanocatalyst (Zhou et al., 2021) (see
Supplementary Methods for details). Compared to aCNTFs,
fCNTFs have higher content of iron contamination and also
higher electrical conductivity (Bulmer et al., 2021). The aCNTFs
and fCNTFs used in our work had twisted angles of 25◦and
10◦, respectively. The GFs were prepared through a dimension-
confined hydrothermal process from aqueous graphene oxide
(GO) suspensions (Dong et al., 2012). This gave GFs with
excellent electrical conductivity and mechanical robustness.
Market available CFs (C3005, World Precision Instruments,
United States) were used to fabricate the CF electrodes. All
these carbon-based fibers had a diameter of 30 µm unless
specified otherwise.
For electrode fabrication, the fibers were first coated with
Parylene-C film of 3–5 µm thickness in a Parylene coater (PDS
2010 Labcoter, Specialty Coating Systems, United States). The
thickness of the Parylene-C coating layers was confirmed via
scanning electron microscopy (SEM). One end of the insulated
fibers was soldered onto a printed circuit board or a metal
pin connector used to connect to the external electronics. The
soldering points were sealed and stabilized with a thin layer of
epoxy. The active sites of the electrodes were exposed by blade
cut. The same method described above was used to fabricate the
platinum (Pt) electrodes from Pt microwires (30 µm diameter,
XYϕ0.03, XIYU Mechanical and Electrical Technology Co.,
China) for a comparison.
Electrode Characterization
An electrochemical workstation (CHI660e, CH Instruments,
United States) was used to perform all electrochemical
characterization. Impedance spectroscopy (EIS) and cyclic
voltammetry (CV) were measured in 1x phosphate-buffered
saline (PBS, pH 7.4) at room temperature in a three-electrode
cell comprising an Ag| AgCl served as the reference electrode, a
large area platinum foil (50 mm ×50 mm, 0.15 mm thickness,
GF29782312, Merk, Germany) as the counter electrode, and the
tested electrode as the working electrode. EIS was performed in
the range of 10∼100 kHz. CV was scanned between potentials of
−0.6 and 0.8 V at a rate of 50 mV/s, beginning at open-circuit
potential and sweeping in the positive direction first. Each sample
was swept for two cycles. The cathodic charge-storage-capacity
(CSCc) was calculated as the time integral of the cathodic current
recorded in the second cycle. For water window testing, CV
was performed at a scan rate of 1 mV/s. The water window
was determined as the water oxidation and reduction potential
obtained from CV measurements, where a steep increase in
current was observed (Supplementary Figure 1).
For CIL measurement using voltage transient experiments,
a three-electrode cell (the same as above) was used. Biphasic,
symmetric, and charge-balanced square-wave current pulses
of 60 µs duration were delivered to the tested samples at
a frequency of 130 Hz with a stimulator (Model 2100, A-M
Systems, United States). Voltage transients under the current
pulses were recorded with an oscilloscope (DSO5202P, Hantek,
China), and the negative potential excursion was calculated by
subtracting the initial access voltage due to solution resistance
from the total voltage (Zhao et al., 2020). The CIL was calculated
by multiplying the current amplitude and pulse duration at which
the negative potential excursion reaches the water reduction
limit (−1.5 and −0.6 V for carbon-based fiber electrodes and
Pt electrodes, respectively), divided by the geometric surface
area (cross-sectional area) of the electrodes. All the electrodes
used in the electrochemical characterization were made from
carbon-based fibers or Pt microwires insulated with ∼5µm thick
Parylene-C film.
The Young’s modulus and tensile strength of each type of
material were measured from the stress–strain curves on a
single-column testing instrument (Instron 5843, Instron Corp.,
United States) (Supplementary Figure 2). The bending stiffness
K, which is the ratio between the longitudinal loading force and
the displacement, represents the mechanical characteristics of
the tissue–electrode interface. For a cylindrical beam, Kcan be
estimated as (Steif, 2012):
K=Eπd4
64
where Eis Young’s modulus of the material and dis the
diameter of the beam.
Insulation Stability Improvement and
Test
We explored the strategy of using alternating polymer/inorganic
multilayers for insulation to improve the long-term stability of
the carbon-based fiber electrodes. The aCNTFs with a diameter
of 15 µm were first coated with 2 µm thick Parylene-C (PDS 2010
Labcoter, Specialty Coating Systems, United States). Then atomic
layer deposition (ALD) of 10 nm HfO2/20 nm Al2O3/10 nm
HfO2triple layer was carried out at 100◦C at 0.1 nm/min using
the ALD System (Savannah S200, Ultratech, United States).
Finally, another 2-µm-thick Parylene-C layer was coated to finish
Frontiers in Neuroscience | www.frontiersin.org 3December 2021 | Volume 15 | Article 771980
fnins-15-771980 December 16, 2021 Time: 12:57 # 4
Fu et al. Carbon Neural Electrodes
FIGURE 1 | SEM characterization. (A–D), SEM images of the side view of an aCNTF (A), fCNTF (B),GF(C), and CF (D). The twisted angle is marked as θ. Insets of
panels (A–D), SEM images of the tip of an aCNTF, fCNTF, GF, and CF electrode. Scale bars of 10 µm apply to all panels and insets.
the insulation. These electrodes are labeled as “Test 1” samples.
“Test 2” samples were prepared same way except for that for
each layer of Parylene-C, 1 µm thickness was used. Electrodes
insulated with 4 and 2 µm Parylene-C were labeled as “Control 1”
and “Control 2” samples, respectively, and used for comparison
of the electrode stability. “Test 1” and “Control 1,” “Test 2” and
“Control 2” samples were placed side by side, respectively, for
Parylene-C deposition to avoid the variance in film thickness and
quality from batch to batch.
Accelerated aging test (AAT) was performed to test the
stability of the above aCNTF electrodes. The electrodes were
completely submerged in 1x PBS at 60◦C in an electro-
thermostatic water bath (CU-420, Shanghai Yiheng Instrument
Co., China) throughout the test. The use of 60◦C instead of
the physiological temperature of 37◦C was to accelerate the
degradation of the electrode coatings. ASTM F1980 (American
Society for Testing and Materials Standard guide for accelerated
aging of sterile medical device packages) recommends that aging
temperature do not exceed 60◦C to avoid non-linear variations
in the rate of reaction (Bierwagen et al., 2003), therefore
we maintained a constant 60◦C during our experiments.
Temperature variations were below ±0.5◦C throughout the
experiment. PBS solutions in the vials was replaced every day
to keep a stable concentration and pH value. The electro-
thermostatic water bath was sealed to avoid the solution
evaporation. The aCNTF electrodes was soldered onto a single-
row, multi-pins straight male headers tip which was bridged on
the opening of the vial. This way the electrodes were immersed in
PBS while the connectors were kept out of the solution.
We used the following formula and assumptions to extrapolate
the simulated age at body temperature (Age37◦C) from that at the
AAT temperature (Age60◦C) (Hukins et al., 2008):
Age37 ◦
C=(Age60 ◦
C)×Q10(TAA −TRS)/10
Where Q10 = 2 (a 10◦C increase in temperature doubles
rate of the chemical reaction), TAA = 60◦C (accelerated aging
Frontiers in Neuroscience | www.frontiersin.org 4December 2021 | Volume 15 | Article 771980
fnins-15-771980 December 16, 2021 Time: 12:57 # 5
Fu et al. Carbon Neural Electrodes
FIGURE 2 | Electrochemical properties. (A–C) representative impedance magnitude (A), phase (B), and CV (C) of various electrodes. (D) Calculated CSCc.
(E) Representative voltage transient of various electrodes (upper curves) in response to a current pulse of 150 µA amplitude (lower curve). (F) CIL of different
electrodes. All electrodes were made of carbon-based fibers or Pt wires with 30 µm diameter insulated with 5 µm thick Parylene-C. The same color codes in (A) are
used in (B,E). Error bars in (D,F) show SEM (n=8forDand n=5forF).
temperature) and TRS = 37◦C (recommended shelf temperature-
body temperature). From this calculation, the age of 153 days at
60◦C corresponds to an age of 753 days at body temperature.
Magnetic Resonance Imaging
Compatibility Studies
All procedures for handling the animals were approved
by the Institutional Animal Care and Use Committees of
Peking University (#COE-DuanXJ-1). All MRI experiments were
performed using a 9.4 T animal MRI scanner (Bruker BioSpin
94/20USR MRI, Germany) with Bruker’s 86 mm volume coil
for transmission and a four-channel rat head surface coil for
receiving (ParaVision Version 6.0.1 for MRI acquisitions). The
electrodes made from carbon-based fibers or Pt wires coated with
∼5µm Parylene-C were implanted in rat brains as described
previously (Lu et al., 2019). For each animal used, all the five types
of electrodes were implanted. A total of five rats were used for the
MRI study. The connectors were not included in the electrodes
to avoid their influence on MRI. The electrodes were tethered
and cemented to the skull. Two craniotomies (∼2×3 mm2
each) were performed directly above two areas of interest (AP:
−2.0∼−4.0 mm; ML: ±1.0∼3.0 mm; DV: −4.0∼−6.0 mm
from dura). The five types of electrodes were implanted in
Frontiers in Neuroscience | www.frontiersin.org 5December 2021 | Volume 15 | Article 771980
fnins-15-771980 December 16, 2021 Time: 12:57 # 6
Fu et al. Carbon Neural Electrodes
these areas with a ∼0.8 mm distance between the adjacent
electrodes, with positions shown as in Figure 5. Craniotomies
were sealed with a silicone elastomer (Kwik-Sil, World Precision
Instruments, United States). MRI scans were conducted 1–3 days
after the electrode implantation surgery. After anesthesia with 4%
isoflurane, the animal’s head was fixed in the MRI scanning coil
with the body axis at the centerline to perform T2- and echo-
planar imaging (EPI)-weighted horizontal and coronal plane
scans following the calibration scan. During MRI scanning, 1.5%
isoflurane with 100% medical air delivered via a nose cone was
used to maintain anesthesia. Animal temperature, respiration,
and blood oxygen saturation were all monitored and within
normal ranges (Model 1025, SA Instruments, United States).
Body temperature was maintained at 37 ±0.5◦C using a
circulated hot water bed and a hot air blower.
T2-weighted anatomical images were acquired with
parameters as follows: TR/TE = 2200/33 ms, RARE factor = 8,
field-of-view (FOV) = 28 ×28 mm2, matrix size = 512 ×512,
and slice thickness = 0.6 mm. EPI-weighted images were
acquired with parameters as follows: TR/TE = 500/13 ms,
FOV = 30 ×30 mm2, matrix size = 80 ×80, flip angle = 55◦,
slice thickness = 0.6 mm, and segment = 4. To quantify the MRI
artifact size of the electrodes, raw coronal images with the largest
electrode artifact were selected. EPI images were upsampled
from 0.37 ×0.37 ×0.6 mm3to 0.05 ×0.05 ×0.6 mm3voxel
resolution. No upsampling was done for T2images. The edges
of artifacts were detected with Canny Edge Detector using
Matlab (R2018b, Mathworks, United States). The artifact size
in the medial-lateral direction was then calculated from the
pixel numbers and averaged over different animal subjects
(Supplementary Figure 3).
RESULTS AND DISCUSSION
Comparison on Electrochemical
Performance
Different carbon-based fibers show distinct surface morphology
due to their different microstructure. As shown in Figure 1, side
view of SEM images of aCNTFs and fCNTFs shows aligned CNT
bundles twisted around the fiber axis at characteristic angles.
While for the GFs, the SEM image shows a porous structure with
rough surface. CFs showed a much smoother surface morphology
than that of the CNTFs and GFs. The SEM images of the exposed
cross sections (inset of Figure 1) of various electrodes showed
a rough and porous microstructure for aCNTF, fCNTF, and GF
electrodes, which is important to achieve a large specific surface
area, thus helpful to obtain better electrochemical performance
for neural electrode application.
The magnitude of impedance at 1 kHz is a commonly
used metric for neural electrodes. Figures 2A,B show
representative EIS of various electrodes. EIS measurements
gave similar impedance magnitude of 129.49 ±5.19 kand
162.56 ±20.06 k(mean ±SEM, n= 8. Same for below) at
1 kHz for electrodes made from the aCNTFs and fCNTFs. This
value was lower than that of Pt electrodes (536.58 ±55.53 k
at 1 kHz), consistent with previous results that the CNTFs have
superior interfacial electrochemical properties than Pt or PtIr
wires (Vitale et al., 2015;Ganji et al., 2017;Lu et al., 2019;Zhao
et al., 2020). GF electrodes showed the lowest impedance value of
50.44 ±5.21 kat 1 kHz. And CF electrodes showed impedance
values of 2.11 ±0.40 Mat 1 kHz, higher than all other
electrodes. Compared to aCNTF and CF electrodes, the phase
of the fCNTF electrodes showed a less capacitive characteristics.
This may be related with the high iron content of the fCNTFs
which provided extra redox sites. The GF electrodes showed
largest CSCcof 832.03 ±50.18 mC/cm2(mean ±SEM, n= 8.
Same for below) among all tested electrodes (Figures 2C,D). The
aCNTF and fCNTF electrodes exhibited slightly lower CSCcof
223.86 ±12.78 and 156.21 ±29.37 mC/cm2, respectively. And
the Pt and CF electrodes showed lower CSCcof 5.88 ±0.78 and
0.91 ±0.19 mC/cm2, respectively.
Voltage transient measurements were carried out to estimate
the CIL, which is defined as the maximum charge that can
be injected in a current-controlled stimulation pulse without
polarizing an electrode beyond the potentials for water reduction
or oxidation (Cogan, 2008;Zhao et al., 2020). Compared to
Pt electrodes, the carbon-based fiber electrodes showed larger
water window (Supplementary Figure 1). Based on the water
window measurement, we calculated the CIL using −1.5 and
−0.6 V as the water reduction limit for carbon-based fiber
electrodes and Pt electrodes, respectively. The aCNTF and GF
electrodes showed comparable CIL of 9.43 ±0.45 mC/cm2
(mean ±SEM, n= 5. Same for below) and 9.96 ±0.39 mC/cm2
(Figures 2E,F) respectively, which is slightly higher than that
of the fCNTF electrodes (3.74 ±0.29 mC/cm2). The CF and
Pt electrodes exhibited much lower CIL of 0.12 ±0.01 and
0.15 ±0.01 mC/cm2, respectively. A comparison of our results
with literature was included as Supplementary Table 1.
The low CIL and CSCcindicate that when using CF or
Pt electrodes to inject charge safely for neural stimulation, a
much larger electrode size is required. The significantly larger
impedance and lower charge injection capability indicate poor
interfacial electrochemical property of the CFs, consistent with
previous reports (Kozai et al., 2012;Pancrazio et al., 2017).
The GF, aCNTF, and fCNTF microelectrodes showed excellent
electrochemical performance which arises from the high surface
area of these fibers accessible to ions due to the interstitial
spaces between the aligned CNTs, CNT bundles, and GO sheets
constituting the fibers (Dong et al., 2012;Lu et al., 2019). These
superior interfacial electrochemical properties of GF, aCNTF, and
TABLE 1 | Mechanical properties of different materials.
Material Twisted angle (◦) Young’s
modulus
(GPa)
Tensile
strength
(MPa)
Bending
stiffness K
(nN.m2)
aCNTF 25 10.89 ±1.22 741.21 ±46.05 0.43
fCNTF 10 22.26 ±0.61 925.55 ±23.35 0.88
GF – 12.49 ±1.04 168.39 ±18.78 0.50
CF – 51.86 ±1.86 899.28 ±39.93 2.06
Pt 81.72 ±4.43 622.22 ±13.95 3.25
Error bars show SEM (n = 6).
Frontiers in Neuroscience | www.frontiersin.org 6December 2021 | Volume 15 | Article 771980
fnins-15-771980 December 16, 2021 Time: 12:57 # 7
Fu et al. Carbon Neural Electrodes
FIGURE 3 | Insulation stability test. Change and statistical comparison analysis of impedance magnitude (A) and phase (B) at 1 kHz, and CSCc(C) of different
electrode samples upon time under AAT. The shaded regions represent the SEM (n= 10 for “Test 1” and “Control 1” electrodes, n= 6 for “Test 2” and “Control 2”
electrodes). For all comparisons, Shapiro–Wilk was used to test normality and Brown-Forsythe was used to test the homogeneity of variance. For comparison
between Day 1 and Day 153 of the same type of insulation layer, paired sample T-test or paired Wilcoxon signed-rank test (when test for normality or equal variance
failed) was used. For comparison between test and control group on Day 153, two sample T-test or Mann–Whitney test (when test for normality or equal variance
failed) was used. * corresponds to comparison between test and control group on Day 153; # corresponds to comparison between Day 1 and Day 153 of the same
samples. *, #, p≤0.05; **, ##, p≤0.01; ***, ###, p≤0.001.
fCNTF electrodes allow a much smaller size to be used, which can
effectively increase the mechanical compliance of the electrodes,
thus reducing the inflammatory responses and helping to achieve
long-term stable neural recording and stimulation.
Comparison on Mechanical Properties
Table 1 compares the mechanical characteristics of each
type of material. The aCNTFs, fCNTFs, and GFs are more
mechanically compliant with smaller Young’s modulus compared
to CFs and Pt wires. The bending stiffness Kscales with
material rigidity (Young’s modulus) Elinearly but scales with
the beam diameter dto the fourth power. The improved
electrochemical properties of the aCNTFs, fCNTFs, and GFs
permit a much smaller diameter to be used. This can drastically
reduce the bending stiffness of electrodes, thus mitigating
mechanical stress at the electrode–tissue interface and reducing
the inflammatory responses (Muthuswamy et al., 2011;Lu
et al., 2019). Compared to other carbon-based fibers, the GFs
showed lower tensile strength. A high tensile strength helps to
achieve a high success rate in electrode fabrication, handling,
transportation, and implantation. The low tensile strength
of the GFs makes them less mechanically strong especially
when the diameter is decreased. The CNTFs, with combined
improved electrochemical performance, mechanical compliance,
Frontiers in Neuroscience | www.frontiersin.org 7December 2021 | Volume 15 | Article 771980
fnins-15-771980 December 16, 2021 Time: 12:57 # 8
Fu et al. Carbon Neural Electrodes
FIGURE 4 | AAT results. Representative impedance spectra (A,B) and CV (C) of different electrode samples at day 1, day 51, and day 153 in ATT.
and tensile strength, are ideal candidates for fabricating flexible
implantable depth electrodes.
Stability Improvement
To improve the structural durability and stability of the
neural electrodes without sacrificing the flexibility of the
electrodes, we explored the strategy of using alternating stacks
of polymer/inorganic materials as insulation layers. We carried
out AAT to characterize the structural and functional stability
of different insulation strategies. Figure 3 shows the change
of impedance amplitude and phase at 1 kHz, and CSCcupon
time from the AAT, with some representative impedance spectra
and CV curves shown in Figure 4. The impedance amplitude
at 1 kHz of “Test 1” samples decreased by less than 20% after
153 days of AAT. Nevertheless, “Control 1” samples which have
the same thickness of Paylene-C as “Test 1” samples but with
no sandwiching inorganic layer in the encapsulation exhibited a
large drop of over 70% in impedance amplitude at 1 kHz after
153 days of AAT. Meanwhile, the impedance phase of “Test 1”
samples exhibited a slight increase while that of “Control 1”
samples exhibited a significant change from −38◦to −11◦after
153 days of AAT. The CSCcof the “Test-1” samples showed no
statistically significant difference from that on Day 1, distinct
from that of the “Control 1” samples which showed a ∼26-
fold increase. The significant decrease in impedance amplitude,
increase in impedance phase and CSCcof “Control 1” samples
indicates a leakage in the insulation layer of the electrodes.
The “Test 2” samples also showed significant improvement in
stability compared to “Control 2” samples, manifested as smaller
change in impedance amplitude, impedance phase, and CSCc. By
extrapolating the simulated age at body temperature from the
age at the AAT temperature, we conclude that by using stacks
of polymer/organic layers for electrode insulation, the aCNTF
electrodes could be stable at physiological conditions at least
for up to 753 days.
These results indicated that by sandwiching an inorganic
HfO2/Al2O3/HfO2ALD film inside the Parylene-C layer as a
barrier layer, the durability and stability of the insulation can
be significantly improved. Inorganic materials including Al2O3,
SiNx, etc., which are grown directly by techniques such as ALD
and CVD, offer superior insulation properties (Minnikanti et al.,
2014;Ahn et al., 2019;Li et al., 2019). Nevertheless, in most
Frontiers in Neuroscience | www.frontiersin.org 8December 2021 | Volume 15 | Article 771980
fnins-15-771980 December 16, 2021 Time: 12:57 # 9
Fu et al. Carbon Neural Electrodes
FIGURE 5 | In vivo MRI artifact assessment. (A–C) Coronal sections of the T2-weighted images of a rat implanted with various electrodes. (D,E) Horizontal sections
of T2-weighted (D) and EPI (E) images of a rat implanted with various electrodes. (F) T2artifact size of different electrodes. (G,H) Coronal sections of the
EPI-weighted images of a rat implanted with various electrodes. (I) EPI artifact size of different electrodes. All electrodes were made from carbon-based fibers or Pt
wires of 30 µm diameter insulated with ∼5µm thick Parylene-C. The insets are zoomed-in photographs of the dashed boxes. Scale bar, 1 mm. Error bars in (F,I)
show SEM. For each type of electrode, n= 5 from five animals.
cases, especially those encountered in research laboratories, these
films are often suffered from extrinsic limitations associated with
heterogeneities in the growth processes and/or contaminants
which leads to micro/nanoscale material defects including
pinholes, cracks, nanopores, and grain boundaries, etc. (Li
et al., 2010, 2019;Minnikanti et al., 2014). By sandwiching
an inorganic barrier layer inside the polymer encapsulation,
the potential of such types of defects to extend throughout
the whole thickness of the insulation layer can be effectively
reduced, thus improving the durability and stability of the
insulation. The Al2O3film from ALD is reported to possess
low water permeability (Xie et al., 2012), but easily dissolves
through hydrolysis when contacting with aqueous solutions
(Abdulagatov et al., 2011;Correa et al., 2015;Nehm et al., 2015;
Daubert et al., 2017;Kim et al., 2017). On the other hand, HfO2is
chemically inert and insoluble in aqueous solutions which makes
it suitable as a capping layer of Al2O3to increase its barrier
performance in a liquid water environment (Kim et al., 2017;
Ahn et al., 2019). The use of the nanometer thick inorganic layers
from ALD doesn’t add too much in electrode diameter, thus will
not compromise the mechanical compliance of the electrodes.
Although the final validation of this insulation strategy will
require chronic implantation and in vivo neural recording from
large number of animal models, we think our work here provides
a possible way to achieve neural recording and stimulation at
high spatiotemporal resolution over time scales spanning the
Frontiers in Neuroscience | www.frontiersin.org 9December 2021 | Volume 15 | Article 771980
fnins-15-771980 December 16, 2021 Time: 12:57 # 10
Fu et al. Carbon Neural Electrodes
lifetime of many animal models for biomedical research. In
addition, we believe that the strategy of using alternating stacks
of polymer/inorganic materials as insulation layers could work
for other types of neural electrodes, including metal microwire
electrodes, silicon probes, and thin film electrode array too.
Comparison on Magnetic Resonance
Imaging Compatibility
We compared the MRI image artifacts of different electrodes
implanted into rat brains in a high-field 9.4T MRI scanner.
All electrodes were made from carbon-based fibers or Pt
microwires insulated with 5-µm-thick Parylene-C. As shown
in Figure 5, GF electrodes showed smallest T2artifacts of
0.075 ±0.005 mm (mean ±SEM, n= 5. Same for below) and
were undetectable in EPI coronal images. The aCNTF and CF
electrodes showed comparable artifact sizes of 0.281 ±0.004 mm
and 0.240 ±0.006 mm, respectively, for T2sequence and
0.838 ±0.012 mm and 0.814 ±0.012 mm, respectively, for
EPI sequence. The fCNTF electrodes elicited largest artifact
with 1.045 ±0.019 mm T2artifact and 2.150 ±0.049 mm
EPI artifact. This is consistent with the fact that fCNTFs
have high content of iron contamination from CNTs growth
process (Zhou et al., 2021). The artifact size of the Pt
electrodes was 0.856 ±0.011 mm for T2sequence and
1.556 ±0.026 mm for EPI sequence, which is slightly smaller
than those from fCNTF electrodes and much larger than other
carbon-based electrodes.
The results indicate that GFs are the most suitable material
for fabricating highly MRI-compatible neural electrodes. But the
relatively low tensile strength limited their miniaturization
for flexible electrodes fabrication. The aCNTF and CF
electrodes could be alternative choices for MRI-compatible
electrodes. However, due to their poor interfacial electrochemical
performance as described above, a much larger diameter
will have to be used when using CF electrodes for neural
recording and stimulation. This will further increase the MRI
artifact size and makes them less desirable for MRI-compatible
neural electrodes. The aCNTF electrodes have excellent
electrochemical and mechanical properties, which permits
the fabrication of small diameter neural electrodes. Together
with their reasonably small MRI artifact, the aCNTF will be
a suitable choice for fabricating flexible and MRI-compatible
neural electrodes for stable longitudinal studies involving
simultaneous electrophysiological measurements and anatomical
or functional MRI.
CONCLUSION
Our study here demonstrated that the nanocarbon-based
fibers, including aCNTFs, fCNTFs, and GFs, have superior
electrochemical performance for neural electrodes fabrication.
The low Young’s modulus and high tensile strength of aCNTFs
and fCNTFs, combined with their superior electrochemical
performances, made them especially suitable for fabricating
implantable depth electrodes with small size and high flexibility,
which is expected to be associated with reduced inflammatory
responses. Combining the strategy of sandwiching inorganic
barrier layers using ALD inside the polymer insulation to
improve the structural stability will provide us a promising way
to achieve long-term stable neural recording and stimulation
spanning the lifetime of the animal models for biomedical
research. The GF electrodes showed highest MRI compatibility.
But the aCNTFs are more suitable for fabricating flexible and
MRI-compatible electrodes due to their excellent electrochemical
performance, mechanical compliance, and strength, as well as
high MRI compatibility.
DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be
made available by the authors, without undue reservation.
ETHICS STATEMENT
The animal study was reviewed and approved by the Institutional
Animal Care and Use Committees of Peking University (#COE-
DuanXJ-1).
AUTHOR CONTRIBUTIONS
XD and XF conceived and designed the experiments. XF, YN,
PW, ZZ, XL, ZX, ZQY, YZ, and QL fabricated the electrodes
and performed the electrochemical characterization. XF, GL,
BZ, and TJ performed the MRI studies. XF, JX, and TL
conducted the insulation stability improvement research. XF,
ZMY, and AC did the mechanical characterization. XD and
XF wrote the manuscript. All authors discussed the results and
commented on the manuscript.
FUNDING
This work was supported by grants from the Beijing Natural
Science Foundation (JQ20008), the National Natural Science
Foundation of China (No. 21972005), the National Basic
Research Program of China (No. 2016YFA0200103), and the
Beijing Graphene Innovation Program (Z191100000819001).
ACKNOWLEDGMENTS
The authors would like to thank the National Center for Protein
Sciences at Peking University in Beijing, China, for assistance
with magnetic resonance imaging and Xiaohang Liang for help
with the operation.
SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found online
at: https://www.frontiersin.org/articles/10.3389/fnins.2021.
771980/full#supplementary-material
Frontiers in Neuroscience | www.frontiersin.org 10 December 2021 | Volume 15 | Article 771980
fnins-15-771980 December 16, 2021 Time: 12:57 # 11
Fu et al. Carbon Neural Electrodes
REFERENCES
Abdulagatov, A. I., Yan, Y., Cooper, J. R., Zhang, Y., Gibbs, Z. M., Cavanagh, A. S.,
et al. (2011). Al2O3and TiO2atomic layer deposition on copper for water
corrosion resistance. ACS Appl. Mater. Interfaces 3, 4593–4601. doi: 10.1021/
am2009579
Ahn, S.-H., Jeong, J., and Kim, S. J. (2019). Emerging encapsulation technologies
for long-term reliability of microfabricated implantable devices. Micromachines
10:508. doi: 10.3390/mi10080508
Apollo, N. V., Maturana, M. I., Tong, W., Nayagam, D. A. X., Shivdasani, M. N.,
Foroughi, J., et al. (2015). Soft, flexible freestanding neural stimulation and
recording electrodes fabricated from reduced graphene oxide. Adv. Funct.
Mater. 25, 3551–3559. doi: 10.1002/adfm.201500110
Ayaz, H., Izzetoglu, M., Bunce, S., Heiman-Patterson, T., and Onaral, B. (2007).
“Detecting cognitive activity related hemodynamic signal for brain computer
interface using functional near infrared spectroscopy,” in Proceedings of the 2007
3rd International IEEE/EMBS Conference on Neural Engineering, Kohala Coast,
HI, 342–345. doi: 10.1109/CNE.2007.369680
Ayaz, H., Shewokis, P. A., Bunce, S., Schultheis, M., and Onaral, B. (2009).
“Assessment of cognitive neural correlates for a functional near infrared-based
brain computer interface system,” in Proceedings of the International Conference
on Foundations of Augmented Cognition, (Berlin: Springer), 699–708. doi: 10.
1007/978-3-642-02812- 0_79
Bensmaia, S. J., and Miller, L. E. (2014). Restoring sensorimotor function through
intracortical interfaces: progress and looming challenges. Nat. Rev. Neurosci. 15,
313–325. doi: 10.1038/nrn3724
Bierwagen, G., Tallman, D., Li, J., He, L., and Jeffcoate, C. (2003). EIS studies
of coated metals in accelerated exposure. Prog. Org. Coat. 46, 149–158. doi:
10.1016/S0300-9440(02)00222-9
Borchers, S., Himmelbach, M., Logothetis, N., and Karnath, H.-O. (2012). Direct
electrical stimulation of human cortex—the gold standard for mapping brain
functions? Nat. Rev. Neurosci. 13, 63–70. doi: 10.1038/nrn3140
Bulmer, J. S., Kaniyoor, A., and Elliott, J. A. (2021). A meta-analysis of conductive
and strong carbon nanotube materials. Adv. Mater. 33:2008432. doi: 10.1002/
adma.202008432
Bunce, S. C., Izzetoglu, M., Izzetoglu, K., Onaral, B., and Pourrezaei, K. (2006).
Functional near-infrared spectroscopy. IEEE Eng. Med. Biol. Mag. 25, 54–62.
doi: 10.1109/MEMB.2006.1657788
Buzsáki, G. (2004). Large-scale recording of neuronal ensembles. Nat. Neurosci. 7,
446–451. doi: 10.1038/nn1233
Buzsáki, G., Anastassiou, C. A., and Koch, C. (2012). The origin of extracellular
fields and currents—EEG, ECoG, LFP and spikes. Nat. Rev. Neurosci. 13,
407–420. doi: 10.1038/nrn3241
Cash, S. S., and Hochberg, L. R. (2015). The emergence of single neurons in clinical
neurology. Neuron 86, 79–91. doi: 10.1016/j.neuron.2015.03.058
Choi, S., Lee, H., Ghaffari, R., Hyeon, T., and Kim, D. (2016). Recent
advances in flexible and stretchable bio-electronic devices integrated
with nanomaterials. Adv. Mater. 28, 4203–4218. doi: 10.1002/adma.20150
4150
Cogan, S. F. (2008). Neural stimulation and recording electrodes. Annu. Rev.
Biomed. Eng. 10, 275–309. doi: 10.1146/annurev.bioeng.10.061807.160518
Correa, G. C., Bao, B., and Strandwitz, N. C. (2015). Chemical stability of titania
and alumina thin films formed by atomic layer deposition. ACS Appl. Mater.
Interfaces 7, 14816–14821. doi: 10.1021/acsami.5b03278
Daubert, J. S., Hill, G. T., Gotsch, H. N., Gremaud, A. P., Ovental, J. S., Williams,
P. S., et al. (2017). Corrosion protection of copper using Al2O3, TiO2, ZnO,
HfO2, and ZrO2atomic layer deposition. ACS Appl. Mater. Interfaces 9, 4192–
4201. doi: 10.1021/acsami.6b13571
Dong, Z., Jiang, C., Cheng, H., Zhao, Y., Shi, G., Jiang, L., et al. (2012). Facile
fabrication of light, flexible and multifunctional graphene fibers. Adv. Mater.
24, 1856–1861. doi: 10.1002/adma.201200170
Durairaj, S., Sidhureddy, B., Cirone, J., and Chen, A. (2018). Nanomaterials-based
electrochemical sensors for in vitro and in vivo analyses of neurotransmitters.
Appl. Sci. 8:1504. doi: 10.3390/app8091504
Figee, M., Luigjes, J., Smolders, R., Valencia-Alfonso, C.-E., Van Wingen, G., De
Kwaasteniet, B., et al. (2013). Deep brain stimulation restores frontostriatal
network activity in obsessive-compulsive disorder. Nat. Neurosci. 16, 386–387.
doi: 10.1038/nn.3344
Fu, T.-M., Hong, G., Zhou, T., Schuhmann, T. G., Viveros, R. D., and Lieber, C. M.
(2016). Stable long-term chronic brain mapping at the single-neuron level. Nat.
Methods 13, 875–882. doi: 10.1038/nmeth.3969
Ganji, M., Elthakeb, A. T., Tanaka, A., Gilja, V., Halgren, E., and Dayeh,
S. A. (2017). Scaling effects on the electrochemical performance of poly
(3, 4-ethylenedioxythiophene (PEDOT), Au, and Pt for electrocorticography
recording. Adv. Funct. Mater. 27:1703018. doi: 10.1002/adfm.201703018
Gilgunn, P. J., Ong, X. C., Flesher, S. N., Schwartz, A. B., and Gaunt, R. A. (2013).
“Structural analysis of explanted microelectrode arrays,” in Proceedings of the
2013 6th International IEEE/EMBS Conference on Neural Engineering (NER),
San Diego, CA, 719–722. doi: 10.1109/NER.2013.6696035
Gosselin, N., Bottari, C., Chen, J.-K., Petrides, M., Tinawi, S., De Guise, É,
et al. (2011). Electrophysiology and functional MRI in post-acute mild
traumatic brain injury. J. Neurotrauma 28, 329–341. doi: 10.1089/neu.2010.
1493
Harrison, J., ˙
Izzeto˘
glu, K., Ayaz, H., Willems, B., Hah, S., Ahlstrom, U., et al. (2014).
Cognitive workload and learning assessment during the implementation of a
next-generation air traffic control technology using functional near-infrared
spectroscopy. IEEE Trans. Hum. Mach. Syst. 44, 429–440.
Huffman, M. L., and Venton, B. J. (2009). Carbon-fiber microelectrodes for in vivo
applications. Analyst 134, 18–24. doi: 10.1039/B807563H
Hukins, D. W. L., Mahomed, A., and Kukureka, S. N. (2008). Accelerated aging
for testing polymeric biomaterials and medical devices. Med. Eng. Phys. 30,
1270–1274. doi: 10.1016/j.medengphy.2008.06.001
Izzetoglu, M., Izzetoglu, K., Bunce, S., Ayaz, H., Devaraj, A., Onaral, B., et al. (2005).
Functional near-infrared neuroimaging. IEEE Trans. Neural Syst. Rehabil. Eng.
13, 153–159. doi: 10.1109/TNSRE.2005.847377
Jeong, J.-W., Shin, G., Park, S. I., Yu, K. J., Xu, L., and Rogers, J. A. (2015). Soft
materials in neuroengineering for hard problems in neuroscience. Neuron 86,
175–186. doi: 10.1016/j.neuron.2014.12.035
Jia, J., Zhao, J., Xu, G., Di, J., Yong, Z., Tao, Y., et al. (2011). A comparison of the
mechanical properties of fibers spun from different carbon nanotubes. Carbon
N. Y. 49, 1333–1339. doi: 10.1016/j.carbon.2010.11.054
Khan, A., Khan, A. A. P., Asiri, A. M., Rub, M. A., Rahman, M. M., and Ghani,
S. A. (2014). In vitro studies of carbon fiber microbiosensor for dopamine
neurotransmitter supported by copper-graphene oxide composite. Microchim.
Acta 181, 1049–1057. doi: 10.1007/s00604-014-1202-0
Kim, D.-H., Viventi, J., Amsden, J. J., Xiao, J., Vigeland, L., Kim, Y.-S., et al.
(2010). Dissolvable films of silk fibroin for ultrathin conformal bio-integrated
electronics. Nat. Mater. 9, 511–517. doi: 10.1038/nmat2745
Kim, L. H., Jang, J. H., Jeong, Y. J., Kim, K., Baek, Y., Kwon, H., et al. (2017). Highly-
impermeable Al2O3/HfO2moisture barrier films grown by low-temperature
plasma-enhanced atomic layer deposition. Org. Electron. 50, 296–303. doi: 10.
1016/j.orgel.2017.07.051
Kozai, T. D. Y., Catt, K., Li, X., Gugel, Z. V., Olafsson, V. T., Vazquez, A. L., et al.
(2015). Mechanical failure modes of chronically implanted planar silicon-based
neural probes for laminar recording. Biomaterials 37, 25–39. doi: 10.1016/j.
biomaterials.2014.10.040
Kozai, T. D. Y., Langhals, N. B., Patel, P. R., Deng, X., Zhang, H., Smith, K. L.,
et al. (2012). Ultrasmall implantable composite microelectrodes with bioactive
surfaces for chronic neural interfaces. Nat. Mater. 11, 1065–1073. doi: 10.1038/
nmat3468
Lebedev, M. A., and Nicolelis, M. A. L. (2017). Brain-machine interfaces: from basic
science to neuroprostheses and neurorehabilitation. Physiol. Rev. 97, 767–837.
doi: 10.1152/physrev.00027.2016
Li, C., Cauwe, M., Yang, Y., Schaubroeck, D., Mader, L., and Op de Beeck, M.
(2019). Ultra-long-term reliable encapsulation using an atomic layer deposited
HfO2/Al2O3/HfO2triple-interlayer for biomedical implants. Coatings 9:579.
doi: 10.3390/coatings9090579
Li, W., Rodger, D. C., Meng, E., Weiland, J. D., Humayun, M. S., and Tai, Y.-
C. (2010). Wafer-level parylene packaging with integrated RF electronics for
wireless retinal prostheses. J. Microelectromech. Syst. 19, 735–742. doi: 10.1109/
JMEMS.2010.2049985
Liu, J. (2018). “Syringe injectable electronics,” in Biomimetics Through
Nanoelectronics, (Cham: Springer), 65–93. doi: 10.1007/978-3-319-68609-7_5
Logothetis, N. K., Pauls, J., Augath, M., Trinath, T., and Oeltermann, A. (2001).
Neurophysiological investigation of the basis of the fMRI signal. Nature 412,
150–157. doi: 10.1038/35084005
Frontiers in Neuroscience | www.frontiersin.org 11 December 2021 | Volume 15 | Article 771980
fnins-15-771980 December 16, 2021 Time: 12:57 # 12
Fu et al. Carbon Neural Electrodes
Lozano, A. M., and Lipsman, N. (2013). Probing and regulating dysfunctional
circuits using deep brain stimulation. Neuron 77, 406–424. doi: 10.1016/j.
neuron.2013.01.020
Lu, L., Fu, X., Liew, Y., Zhang, Y., Zhao, S., Xu, Z., et al. (2019). Soft and MRI
compatible neural electrodes from carbon nanotube fibers. Nano Lett. 19,
1577–1586. doi: 10.1021/acs.nanolett.8b04456
Minnikanti, S., Diao, G., Pancrazio, J. J., Xie, X., Rieth, L., Solzbacher, F.,
et al. (2014). Lifetime assessment of atomic-layer-deposited Al2O3–Parylene
C bilayer coating for neural interfaces using accelerated age testing and
electrochemical characterization. Acta Biomater. 10, 960–967. doi: 10.1016/j.
actbio.2013.10.031
Muthuswamy, J., Anand, S., and Sridharan, A. (2011). Adaptive movable neural
interfaces for monitoring single neurons in the brain. Front. Neurosci. 5:94.
doi: 10.3389/fnins.2011.00094
Nehm, F., Klumbies, H., Richter, C., Singh, A., Schroeder, U., Mikolajick, T., et al.
(2015). Breakdown and protection of ALD moisture barrier thin films. ACS
Appl. Mater. Interfaces 7, 22121–22127. doi: 10.1021/acsami.5b06891
Özel, R. E., Wallace, K. N., and Andreescu, S. (2011). Chitosan coated carbon
fiber microelectrode for selective in vivo detection of neurotransmitters in live
zebrafish embryos. Anal. Chim. Acta 695, 89–95. doi: 10.1016/j.aca.2011.03.057
Pancrazio, J. J., Deku, F., Ghazavi, A., Stiller, A. M., Rihani, R., Frewin, C. L., et al.
(2017). Thinking small: progress on microscale neurostimulation technology.
Neuromodulation Technol. Neural Interface 20, 745–752. doi: 10.1111/ner.
12716
Patel, P. R., Zhang, H., Robbins, M. T., Nofar, J. B., Marshall, S. P., Kobylarek,
M. J., et al. (2016). Chronic in vivo stability assessment of carbon fiber
microelectrode arrays. J. Neural Eng. 13:66002. doi: 10.1088/1741-2560/13/6/06
6002
Prasad, A., Xue, Q.-S., Dieme, R., Sankar, V., Mayrand, R., Nishida, T., et al.
(2014). Abiotic-biotic characterization of Pt/Ir microelectrode arrays in chronic
implants. Front. Neuroeng. 7:2. doi: 10.3389/fneng.2014.00002
Prasad, A., Xue, Q.-S., Sankar, V., Nishida, T., Shaw, G., Streit, W. J., et al. (2012).
Comprehensive characterization and failure modes of tungsten microwire
arrays in chronic neural implants. J. Neural Eng. 9, 56015. doi: 10.1088/1741-
2560/9/5/056015
Seymour, J. P., and Kipke, D. R. (2007). Neural probe design for reduced tissue
encapsulation in CNS. Biomaterials 28, 3594–3607. doi: 10.1016/j.biomaterials.
2007.03.024
Shenoy, K. V., and Carmena, J. M. (2014). Combining decoder design and neural
adaptation in brain-machine interfaces. Neuron 84, 665–680. doi: 10.1016/j.
neuron.2014.08.038
Stanley, G. B. (2013). Reading and writing the neural code. Nat. Neurosci. 16,
259–263. doi: 10.1038/nn.3330
Steif, P. S. (2012). Mechanics of Materials. Upper Saddle River, NJ: Pearson.
Vitale, F., Summerson, S. R., Aazhang, B., Kemere, C., and Pasquali, M. (2015).
Neural stimulation and recording with bidirectional, soft carbon nanotube
fiber microelectrodes. ACS Nano 9, 4465–4474. doi: 10.1021/acsnano.5b0
1060
Wang, K., Frewin, C. L., Esrafilzadeh, D., Yu, C., Wang, C., Pancrazio,
J. J., et al. (2019). High-performance graphene-fiber-based neural recording
microelectrodes. Adv. Mater. 31:1805867. doi: 10.1002/adma.201805867
Weiskopf, N., Sitaram, R., Josephs, O., Veit, R., Scharnowski, F., Goebel, R.,
et al. (2007). Real-time functional magnetic resonance imaging: methods and
applications. Magn. Reson. Imaging 25, 989–1003. doi: 10.1016/j.mri.2007.02.
007
Williams, J. C., Rennaker, R. L., and Kipke, D. R. (1999). Long-term neural
recording characteristics of wire microelectrode arrays implanted in cerebral
cortex. Brain Res. Protoc. 4, 303–313. doi: 10.1016/S1385-299X(99)00034-3
Xie, C., Liu, J., Fu, T.-M., Dai, X., Zhou, W., and Lieber, C. M. (2015). Three-
dimensional macroporous nanoelectronic networks as minimally invasive brain
probes. Nat. Mater. 14, 1286–1292. doi: 10.1038/nmat4427
Xie, X., Rieth, L., Merugu, S., Tathireddy, P., and Solzbacher, F. (2012). Plasma-
assisted atomic layer deposition of Al2O3and parylene C bi-layer encapsulation
for chronic implantable electronics. Appl. Phys. Lett. 101:93702. doi: 10.1063/1.
4748322
Zhao, S., Li, G., Tong, C., Chen, W., Wang, P., Dai, J., et al. (2020). Full activation
pattern mapping by simultaneous deep brain stimulation and fMRI with
graphene fiber electrodes. Nat. Commun. 11:1788. doi: 10.1038/s41467-020-
15570-9
Zhao, S., Liu, X., Xu, Z., Ren, H., Deng, B., Tang, M., et al. (2016). Graphene
encapsulated copper microwires as highly MRI compatible neural electrodes.
Nano Lett. 16, 7731–7738. doi: 10.1021/acs.nanolett.6b03829
Zhong, Y., and Bellamkonda, R. V. (2008). Biomaterials for the central nervous
system. J. R. Soc. Interface 5, 957–975. doi: 10.1098/rsif.2008.0071
Zhou, T., Niu, Y., Li, Z., Li, H., Yong, Z., Wu, K., et al. (2021). The synergetic
relationship between the length and orientation of carbon nanotubes in direct
spinning of high-strength carbon nanotube fibers. Mater. Des. 203:109557. doi:
10.1016/j.matdes.2021.109557
Conflict of Interest: The authors declare that the research was conducted in the
absence of any commercial or financial relationships that could be construed as a
potential conflict of interest.
Publisher’s Note: All claims expressed in this article are solely those of the authors
and do not necessarily represent those of their affiliated organizations, or those of
the publisher, the editors and the reviewers. Any product that may be evaluated in
this article, or claim that may be made by its manufacturer, is not guaranteed or
endorsed by the publisher.
Copyright © 2021 Fu, Li, Niu, Xu, Wang, Zhou, Ye, Liu, Xu, Yang, Zhang, Lei,
Zhang, Li, Cao, Jiang and Duan. This is an open-access article distributed under the
terms of the Creative Commons Attribution License (CC BY). The use, distribution
or reproduction in other forums is permitted, provided the original author(s) and
the copyright owner(s) are credited and that the original publication in this journal
is cited, in accordance with accepted academic practice. No use, distribution or
reproduction is permitted which does not comply with these terms.
Frontiers in Neuroscience | www.frontiersin.org 12 December 2021 | Volume 15 | Article 771980
Available via license: CC BY
Content may be subject to copyright.
