POLYCRYSTALLINE DIAMOND BASED NEURAL INTERFACE FOR OPTOGENETICS
AND NEUROTRANSMITTER DETECTION
By
Bin Fan

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Electrical Engineering - Doctor of Philosophy
2017

ABSTRACT
POLYCRYSTALLINE DIAMOND BASED NEURAL INTERFACE FOR OPTOGENETICS
AND NEUROTRANSMITTER DETECTION
By
Bin Fan
Neural interface forms a communication bridge between a human brain and external
circuitries, which enables promising bioelectronics medicines for diseases treatments, such as
inflammatory bowel disease, Alzheimer's disease, and restore sensorimotor function lost due to
traumatic brain, spinal cord injury, and amputations. Neurons in the central nervous systems
communicate with each other electrically along the axon from soma to dendrite and chemically
between neuron to neuron in the synapses through release and uptake of neurotransmitters. In
particular, dopamine (DA) is one of the most important neurotransmitters, which associates with
many aspects of the neurophysiological processing, such as stress, memory, and addiction.
External stimulation is desired to study the dynamics of DA release and uptake and its correlation
to the animal behavioral changes. Previously, electrical stimulation was used as a neuromodulation
technique for such purpose, which can cause a significant amount of nondopaminergic system
activation and result in consequential neurological activities or dynamics not related to DA
release[1]. Recent advances in optogenetics provide a unique neuromodulation technique,
allowing optical control of genetically targeted specific neurons that express light-sensitive opsin
proteins with sub-millisecond temporal precision. Utilizing the cell-type specificity of
Optogenetics, researchers can have a more controlled manipulation of the dopaminergic system
and have an unbiased study on DA related neurological diseases.
The current engineering tools for Optogenetics use laser and micro light emitting diodes
(ÎźLEDs) as the light sources, where ÎźLEDs show great promises with respect to device

miniaturization, simplicity, low power and low cost of system implementation. However, using
ÎźLEDs as a light source can cause potential thermally-induced tissue damage due to ÂľLED Joule
heating. To address the localized Joule heating issue, a ÎźLED based optrode was developed in this
thesis using polycrystalline diamond as a heat spreader due to its very high thermal conductivity.
Compared with an SU8 probe with the same dimensions, a diamond probe can reduce the
maximum temperature variations by ~90% at 3.6V 100ms duration pulses. The functionality of
the probe was tested in vivo, where light-evoked action potentials were successfully detected.
Besides the very high thermal conductivity, diamond has unique features for neurotransmitter
sensing, such as a larger potential window, low background current and resistance to surface
fouling. In addition, diamond is a biocompatible and chemically inert material, which enables
long-term device implantation. Therefore, above mentioned properties make diamond a promising
candidate for Optogenetics and neurotransmitter detection. However, diamond is a rigid material
and the micromotion-induced strain has been hypothesized to be the main cause of harmful
immune responses and even irreversible tissue damage. Due to the process temperature intolerant,
diamond cannot synthesis onto polymer substrates directly. To address this issue, a wafer-level
substrate transfer process is first time proposed to transfer all diamond macro/micro patterns from
a diamond growth substrate, silicon, onto a flexible Parylene substrate. The electrochemical
properties of the transferred diamond-polymer electrodes were evaluated (i) using an outer sphere
redox couple to study the electron transfer process and (ii) quantitative and qualitative studies of a
neurotransmitter redox dopamine/dopamine-o-quinone. A linear response of the BDD sensor to
dopamine concentrations of 0.5 ÂľM to 100 ÂľM was observed (R2 = 0.999) with a sensitivity of
0.21 ¾A/cm2¡¾M.

Copyright by
BIN FAN
2017

ACKNOWLEDGEMENTS

First and foremost, I would like to express my sincere gratitude to Dr. Wen Li for her gracious
support throughout my entire Ph.D. program and the opportunity to prove myself in the
challenging multidisciplinary field of research under her expert guidance. Dr. Li’s creativity,
enthusiasm, problem-solving acumen and vision of bioelectronics inspired me on pursuing
innovative, efficient solutions to various real life device changelings. As a mentor, she always
offered her guidance and help whenever I needed it not only for the scientific research but also my
life. I am very grateful for her never-ending help, patience, understanding and encouragements
through a hard time in my life.
Besides, I would like to thank my committee members: Dr. Arthur J. Weber for his mentorship
and invaluable advice on the neuroscience study and surgical skills, Dr. Prem Chahal for his
guidance and help on RF sensor development and Dr. Thomas SchĂźlke for his support on the
development of diamond devices.
Then, I would like to extend my appreciation to our collaborators: Michael F. Becker, Robert
Rochenberg and Cory Rusinek from Fraunhofer USA, Center for Coating and Diamond
Technologies on diamond based neural interfaces; Dr. Maysam Ghovanloo and Yaoyao Jia from
Georgia Institute of Technology on the development of wireless neural interfaces. None of my
work can be conducted without the help from our excellent collaborators.
Furthermore, I would like to thank my colleagues from Microtechnology lab: Kiyong Kwon,
Xiaopeng Bi, Wasif Khan, Brain Crum, Haider Almumen, Yue Guo, Tian Xie and colleagues
outside the Microtechnology lab: in Li, Xianbo Yang, Liangliang Chen, Mingquan Yuan, Liang
Zhou, Heyu Yin, Hao Wan Yue Huang, Yaoyao Jia, Steven Leung and Yiyan Li for their technical
assistance on microfabrication, circuit design, testing and analysis. At the same time, I would like
v

to show my special gratitude to Dr. Baokang Bi from Keck Microfabrication Facility and Brian
Wright from ECE shop for their supports on device microfabrication.
In addition, I would like to special thank Xiaofeng Zhao, Yiqun Yang, Pedro Nariyoshi and
Jie Li. I appreciate your company and support for the past years.
Finally, and most importantly, I would like to sincerely thank my family and Yan Zhu for
their endless love and support of every step of my way.
The completion of my Ph.D. program at Michigan State University is just a start of another
new journey of my way. I would like to quote what Andy said in the movie of Shawshank
Redemption at the beginning of this new journey: “Hope is a good thing, maybe the best of things,
and no good thing ever dies.”

vi

TABLE OF CONTENTS

LIST OF TABLES .......................................................................................................................... x
LIST OF FIGURES ....................................................................................................................... xi

Chapter 1. Introduction ............................................................................................................... 1
1.1 Background ......................................................................................................................... 1
1.2 Current challenges on hardware development ...................................................................... 3
1.2.1Thermal challenges of putting ÂľLED near tissue ............................................................ 3
1.2.2 Material long-term compatibility and safety .................................................................. 5
1.2.3 Electrochemical performance of the electrodes.............................................................. 8
1.3 Solutions to the challenges and objective of this work ......................................................... 9
1.4 Layout of the Dissertation ................................................................................................... 12
Chapter 2. A review of Optogenetics and light delivery methods............................................ 13
2.1 Microbial Opsins ................................................................................................................. 13
2.1.1 Fast Excitation .............................................................................................................. 14
2.1.2 Fast Inhibition ............................................................................................................... 15
2.1.3 Step-function opsin (SFO) ............................................................................................ 17
2.1.4 Biochemical Modulation .............................................................................................. 18
2.2 Optogenetic Neural Implants .............................................................................................. 19
2.2.1 Laser-coupled Optical Neural Implants ........................................................................ 21
2.2.1.1 Glass-sharpened optical fibers ............................................................................... 22
2.2.1.2 Out-of-plane microwaveguide arrays ............................................................... 23
2.2.1.3. In-plane microwaveguide probe ........................................................................... 27
2.2.2 LED-Based Optical Neural Implants............................................................................ 28
2.2.2.1. Utah-type optical arrays ........................................................................................ 29
2.2.2.1.1 Surface-mounted ÂľLED arrays ........................................................................ 30
2.2.2.1.2 Optical fiber/waveguide-coupled ÂľLED arrays ............................................... 31
2.2.2.2 Michigan-type optical probes ................................................................................ 32
Chapter 3. BDD neurotransmitter detection sensor .................................................................. 39
3.1 Theory of Electrochemistry ............................................................................................. 39
3.1.1 Nernst equation ............................................................................................................. 39
3.1.2 Kinetics of electrode reactions ..................................................................................... 40
3.1.3 Mass transport............................................................................................................... 40
3.2 Techniques for study of electrode reactions .................................................................. 41
3.2.1 Controlled potential methods (Potential step) .............................................................. 42
vii

3.2.2 Potential sweep methods .............................................................................................. 44
3.3 Polycrystalline diamond deposition and characteristics ..................................................... 45
3.4 BDD devices for chemical sensing ..................................................................................... 51
Chapter 4. An implantable, miniaturized SU-8 optical probe for Optogenetics-based deep
brain stimulation ........................................................................................................................... 55
4.1 Motivation ........................................................................................................................... 55
4.2 Optical probe device design fabrication .............................................................................. 56
4.2.1 Device design ............................................................................................................... 56
4.2.2 Fabrication process ....................................................................................................... 57
4.3 Results and discussions ....................................................................................................... 59
4.3.1 Electrical properties ...................................................................................................... 59
4.3.2 Optical properties ......................................................................................................... 60
4.3.3 In-vivo LFP signal recordings....................................................................................... 61
4.4 Conclusion ........................................................................................................................... 63
Chapter 5. A hybrid neural interface optrode with a polycrystalline diamond heat spreader for
Optogenetics ............................................................................................................................... 64
5.1 Motivation ......................................................................................................................... 64
5.2 Methodology ....................................................................................................................... 66
5.2.1 Device fabrication......................................................................................................... 66
5.2.2 FEM Simulation of Device Thermal Properties ........................................................... 68
5.2.3 Device Characterization ............................................................................................... 70
5.2.3 In-vivo Animal Experiments ......................................................................................... 71
5.3 Results and discussions .................................................................................................... 72
5.3.1 Fabricated devices ........................................................................................................ 72
5.3.2 Optical and Electrical Properties .................................................................................. 74
5.3.3 Thermal Properties ....................................................................................................... 76
5.3.4 In-vivo Optical Neuromodulation and Recording ........................................................ 80
5.4. Conclusion ........................................................................................................................ 81
Chapter 6. Large-scale, all polycrystalline diamond structures transferred on flexible ParyleneC films for electrochemical sensing .............................................................................................. 83
6.1 Motivation ......................................................................................................................... 83
6.2 Methodology ....................................................................................................................... 85
6.3 Results................................................................................................................................. 88
6.3.1 Characteristics of the BDD films ................................................................................ 88
6.3.2 Mechanical properties of the transferred BDD-Parylene structure .............................. 89
6.3.3 Characteristics of the BDD-Parylene electrochemical sensors .................................... 91
6.3.3.1 Potential window ................................................................................................... 91
6.3.3.2 Double-layer capacitance ....................................................................................... 92
6.3.3.3 Chemical redox characteristics .............................................................................. 93
viii

6.3.3.4 Dopamine sensing characteristics .......................................................................... 95
6.4 Conclusions ......................................................................................................................... 97
Chapter 7.

Conclusion .............................................................................................................. 98

BIBLIOGRAPHY ....................................................................................................................... 100

ix

LIST OF TABLES
Table 1-1 Thermal conductivity of noble metal and polymers [65], [66], [67], [68], [69],
[70], [71], [72] .............................................................................................................................. 10
Table 2-1 Summary of the specification of miniaturized, laser-based Optogenetic neural implants
....................................................................................................................................................... 25
Table 2-2 Summary of the specification of miniaturized, ÂľLED-based Optogenetic neural
implants ......................................................................................................................................... 36
Table 4-1 Dimensions of the SU-8 probe and ÂľLED chips (L: length, W: width, H: height) ...... 56
Table 4-2 Typical parameters using for optical stimulation ......................................................... 60
Table 5-1 Single/dual-shank probe dimensions ............................................................................ 65
Table 5-2 Simulation Parameters in COMSOLÂŽ .......................................................................... 69
Table 6-1 Summary of BDD film Characteristics ........................................................................ 89
Table 6-2 CV with different concentrations ................................................................................. 94

x

LIST OF FIGURES
Figure 1-1 Concept diagram of a diamond based opto-electro-chemical hybrid neural interface.11
Figure 2-1 Major classes of single-component Optogenetics tool. (Reprinted from[83]) ............ 13
Figure 2-2 Kinetic and spectral attributes of Optogenetic tool variants. The variant refers to ChR2
mutation if not specified (Reprinted from [99]) ........................................................................... 19
Figure 2-3 Examples of laser-based optical neural interfaces: (a) A dual-core optical fiber system
with one optical core for optical stimulation and one hollow core filled up with 1-3M NaCl for
electrical recording. (Reprinted from [67]) (b) A multimode optical fiber with four tetrode bundles
attached for electrophysiological recording. (Reprinted from [69]) (c) A dual-mode optrode array
adapted from a Utah multielectrode array, where one recording shank was replaced with a
multimode optical fiber. (Reprinted from [72]) (d) An in-plane neural probe adapted from
conventional Michigan neural probe with embedded dielectric waveguides and microfluidic
channels. (Reprinted from [79]) (e) A 3D multiwaveguide array consisting of a set of waveguide
combs assembled on a base plate-holder through two alignment and fixation pieces. (Reprinted
from [81]) ...................................................................................................................................... 24
Figure 2-4 Examples of ÂľLED-based neural interfaces: (a) A high density ÂľLED array fabricated
by conventional silicon-based microfabrication technology. Reprinted from[90] (b) A 4×4 OptoµECoG array with a transparent microelectrode array and a µLED array on a flexible Parylene-C
substrate for epidural optical stimulation and electrical recording of cortical activity. Reprinted
from [94]. (c) A ÂľLED-coupled optical fiber array with a miniaturized Si housing plate for optical
fiber alignment and fixation. Reprinted from [96] (d) A ÂľLED-coupled SU-8 microwaveguide
array fabricated by a droplet backside exposure method, where an ITO-Parylene-gold-Parylene
sandwich clay was used to minimize light-induced artifacts. Reprinted from [99]. (e) A custom
designed ÂľLED probe fabricated from an epitaxial GaN/sapphire substrate by semiconductorbased microfabrication technology. Reprinted from [105]. (f) A flexible, multifunctional neural
probe with integrated temperature sensor, neural recording microelectrodes, light intensity sensor,
and ÂľLED. Reprinted from [107]. ................................................................................................ 35
Figure 2-5 A spider schematic compares several main specifications of the laser- and LED-based
microdevices surveyed in this paper, in terms of size, density, multiple functions, wireless
capability, maximum light delivery efficiency, and light delivery efficiency. The performance is
rated on a scale of 1 to 5, with 5 being the best. ........................................................................... 37
Figure 3-1 Simplified block diagram for potentiometric measurements. ..................................... 42
Figure 3-2 A demonstration of applied potential (a) and recorded current (b) for potential step
technique. (Figures are adopted from www.studyblue.com) ........................................................ 43
Figure 3-3 A demonstration of cyclic voltammetry: (a) Input potential between WE and RE (b)
Recorded current-time plot. (Figure 3-2 (b) is adopted from http://urrjaa.blogspot.com/) .......... 45

xi

Figure 3-4 A diagram of diamond CVD reactor: (a) hot filaments CVD reactor (b) Microwave
plasma CVD. (Figure 3-3 (a) is adopted from [169], Figure 3-3 (b) is adopted from [170]. ....... 46
Figure 3-5 (a) schematic (b) scanning electron microscopy (SEM) image of the cross section of
diamond film of MC BDD. SEM imagines of (c) diamond nucleation side and (d) diamond growth
side. (a) and (b) are adopted and reprinted from [168]. (c) and (d) are taken using the films and
devices reported in Chapter 6. ...................................................................................................... 47
Figure 3-6 Voltammograms of Au, BDD growth side and nucleation side in 1M KCL solution
(WE: BDD / Au, CE: Pt, RE: Ag/Agcl), Scan rate: 0.1V/s. This experiment is done using the films
and devices reported in Chapter 6. ................................................................................................ 49
Figure 3-7 Raman spectroscope of heavily boron doped polycrystalline diamond with different
B/C ration in the gas phase. Reprinted from [175] ....................................................................... 50
Figure 3-8 Raman spectroscopy of BDD nucleation side and growth side. This experiment is done
using the films and devices reported in Chapter 6. ....................................................................... 51
Figure 3-9 (a) BDD coated metal thin wires with Polypropylene pipet tip as insulation layer.
Reprinted from [177]. (b) A thinned down BDD neural probe with thickness of 3Âľm. Reprinted
from [181]. (c) A diamond polymer interface fabricated by transfer diamond from growth rigid
substrate onto spin-cast epoxy (only pads are made of diamond). Reprinted from [182]. (d) A BDD
microelectrode array fabricated using wafer transfer technology. Reprinted from [183]. ........... 53
Figure 4-1 Conceptural diagram of the optical probe: (a) an overall view (b) a cross-section view.
....................................................................................................................................................... 55
Figure 4-3 Fabrication process...................................................................................................... 58
Figure 4-2: (a) A optical probe with thick photoresist residue at the edge. (b) A fabricated optical
probe with a ÂľLED mounted on the tip. ....................................................................................... 58
Figure 4-4 I-V curve of the ÂľLEDs .............................................................................................. 59
Figure 4-5 Electrical and optical characteristics of SU-8 probe. (a) Spectrum of light emitting from
ÂľLED chip with a ~280Âľm SU-8 coating. (b) A comparison of light intensity of optical probe
coated with and without SU-8. (c)SEM image and profilometry of the SU-8 probe (d) Light
scattering property measured in gelatin. ....................................................................................... 61
Figure 4-6 Demonstration of the efficacy of the deep brain optical stimulation using the fabricated
probe in the rat’s brain: (a) the experiment setup and (b) – (d) the recorded LFP with the µLED
input voltage of 3.0V, 3.2 V and 3.4V, respectively. ................................................................... 62
Figure 5-1 Concept diagram of the proposed neural interface optrode with a PCD heat spreader.
....................................................................................................................................................... 65

xii

Figure 5-2 Fabrication process for making the proposed PCD probe. (a) PCD growth on
molybdenum (Mo). (b) PCD release from Mo substrate. (c) Metal deposition and patterning. (d)(f) ÂľLEDs assembly. (g)-(h) Parylene C coating and patterning. (i) Probe shaping. ................... 67
Figure 5-3 (a)Simulation model in COMSOLÂŽ. (b) Mesh generated for FEM simulation. ........ 69
Figure 5-4 Diamond growth side (rough side) of a PCD substrate. (b) Diamond nucleation (smooth
side) of the same PCD substrate. (c) An AFM imagine shows the surface morphology of the
nucleation side of the PCD substrate. (d) Fabricated PCD and SU-8 probe for heat dissipation
measurement. (e) A fabricated single-shank PCD probe with Parylene ribbon cable for
interconnection. (f) Fabricated single/dual-shank probes. Inset shows the ÂľLED, recording
electrode, and metal interconnects. ............................................................................................... 73
Figure 5-5Figure 5-5 The electrical and optical properties of a fabricated PCD probe. (a) Light
spectrum of the ÂľLED. (b) Light intensity and power consumption of the ÂľLED. (c) Light
scattering property in gelatin. (d) Normalized blue light intensity (both amplitude and color
indicate the light intensity)............................................................................................................ 75
Figure 5-6 Heat distribution of the SU-8 and PCD probes with 3.4V, 1Hz, 100ms duration pulses:
(a)-(b) from the high-resolution infrared camera (c)-(d) from the FEM simulation in COMSOLÂŽ
(Enlarged tip areas are shown in dash squares.) ........................................................................... 77
Figure 5-7 Cooling curves of the probes after activating ÂľLED for 60sec with different inputs of
(a) an SU-8 probe and (b) a PCD probe. (c) Instantaneous and steady state temperature variations
of the SU-8 probe and the PCD probe. (d) and (e) shows close-up view of instantaneous
temperature variations of the diamond probe in (c). ..................................................................... 78
Figure 5-8 (a) In-vivo testing setup.(b) Probe schematic shows the location of 4 recording channels.
....................................................................................................................................................... 80
Figure 5-9 (a) Recorded signals with the applied voltage of 3.2V, 3.4V, and 3.6V, respectively. (b)
Action potentials recorded from different channels with 40 trials stacking at the input voltage of
3.6 V.............................................................................................................................................. 81
Figure 6-1. Illustration of the fabrication processes for I. pre-substrate transfer patterning; II.
Substrate transfer; and III. Post-substrate transfer processing. ..................................................... 85
Figure 6-2 Device design of ÂľLED probes and chemical sensors used to demonstrate the BDD
transfer process. The black and grey colors represent BDD and Parylene, respectively.............. 87
Figure 6-3 (a) A SEM image shows the surface morphology of the BDD nucleation side. (b) A
Raman spectrum shows both boron and diamond bands from the BDD nucleation side. ............ 89
Figure 6-4 The custom designed KOH etching kit (a) before and (b) after assembly. (c) BDD
patterns on a wafer-scale flexible Parylene-C substrate after removing the Si substrate. Microscope
images shows contact pads and ÂľLED electrode probes with 200 Âľm and 50 Âľm trace width. (d)
A SEM image shows the mesh structure of the contact pads, where Parylene anchors, Parylene
substrate and BDD are highlighted in white, green, and blue, respectively. (e) and (f) Flexibility
xiii

of a BDD ÂľLED probe wrapped around the tip of a micro punch, with ÂľLED off and on. (g) and
(h) Scotch tapeÂŽ testing before(g) and after (h) peeling a BDD-Parylene sensor off the tape, with
the BDD side facing down. ........................................................................................................... 90
Figure 6-5 (a) A fabricated BDD-Parylene sensor with three electrodes. (b) Voltammogram of Au
and BDD electrode in 1.0M KCL solution vs. Ag/AgCl at a scan rate of 0.1V/s (CE: Pt electrode,
RE: Ag/AgCl). (c) Voltammograms of BDD electrodes vs. BDD at various scan rates (CE: BDD,
RE: BDD). The voltammograms are offset for better visibility. .................................................. 92
Figure 6-6 Voltammograms of the BDD sensor with various concentrations of Ru(NH3)62+/3+ in
1.0M KCl solution vs. BDD at a scan rate of 0.1V/s. (CE: BDD, RE: BDD) (b) Fitting curve of
the oxidation peak current density versus different concentrations of Ruhex. (c) Voltammograms
of BDD electrodes vs. BDD at various scan rates (CE: BDD, RE: BDD) in 2 M Ru(NH3)62+/3+ with
1.0 M KCl solution. (d) Fitting curve of the oxidation peak current density versus square root of
scan rates. ...................................................................................................................................... 94
Figure 6-7 (a) Dopamine / Dopamine-o-quinone redox. (b) Voltammograms of the BDD sensor in
various concentrations of DA with 0.1 M PBS vs. BDD. The scan rates are 1.0 V/s (c)
Chronoamperograms of various concentrations of DA in PBS buffer solution vs. BDD at an
applied potential of 1.0 V. (d) Fitting curve of the background corrected current versus the
concentrations of DA. ................................................................................................................... 96

xiv

Chapter 1.

Introduction

1.1 Background
The complex brain networks comprise billions of interconnected neurons with diverse
types, shapes, sizes, and activity patterns. Targeted access to specific neural populations
with high spatiotemporal resolution enables the study of neural circuits and cellular
conditions, for both fundamental understandings of brain functions and development of
therapeutic strategies for many brain injuries and disorders. While well-established
microelectrophysiological methods have been successfully used to record neural activity at
single-cell resolution[2], neuromodulation with electrical modality, which initiates neural
functional response by injecting a biphasic current to depolarize the membranes of nerve
cells[3], suffers from indiscriminate stimulation of cell components (somas, dendrites, and
axons) as well as poor spatial resolution due to unpredictable current pathways[4]. Recent
advances in optogenetics provide a unique neuromodulation technique, allowing optical
control of genetically targeted specific neurons that express light-sensitive opsin proteins,
such as light-sensitive ion channels, Channelrhodopsin-2 (ChR2)[5], optically activated
chloride pumps, Halorhodopsin (NpHR)[6] and proton pumps, Archaerhodopsin (Arch)[7]
The cell-type specificity of Optogenetics is achieved by selecting appropriate promoters,
for example, CamKIIÎą for targeting excitatory neurons, glial fibrillary acidic protein for
targeting astroglia, and ppHcrt promoter for targeting hypocretin neurons in rodents[8].
At the same time, neurons in the central nervous systems communicate with each other
electrically along the axon from soma to dendrite and chemically between neuron to neuron in the
synapses through release and uptake of neurotransmitters. In particular, dopamine (DA) is one of
the

most

important

neurotransmitters.

Dopamine
1

(DA)

contracted

from

3,4-

dihydroxyphenethylamine associated with many aspects of the neurophysiological processing,
such as stress[9], memory[10], and addiction[11]. There are three major dopaminergic pathways
in the brain: Mesolimbic pathway from ventral tegmental area (VTA) to the nucleus accumbens
(NAc), Nigrostriatal pathway from substantia nigra (SN) to the striatum, Mesocortical pathway
from VTA to the prefrontal cortex[12]. Abnormal activities of DA storage, release and reuptake
are the main cause of several neural disorders in the central nervous system, such as Parkinson’s
diseases and schizophrenia [1],[11]. Besides, dysregulated DA is found to be an important factor
affecting cardiovascular and renal systems. For example, a DA dose between 32 to 64 Âľg/kg caused
an increase in heart contractile force, heart rate, and arterial pressure of an anesthetized dog[13].
Low-dose DA has been commonly used to increase renal blood flow and reduce the risk of renal
failure[14]. Hence, real-time monitoring of dynamic changes in DA concentration is very
important for understanding the functionality of the brain and other organs. Because of the low
concentration and dynamically changing levels of DA in the brain, in-situ detection without sample
treatment is desired, which requires implanted sensors with a wide working potential window,
resistance to molecular adsorption and corrosion, biocompatibility, mechanical flexibility, high
sensitivity and selectivity of the target analyte. Of well-established chemical sensing approaches,
electrochemical sensors show unique advantages of low cost, fast dynamic response (up to the
millisecond range), miniaturized geometry, and high spatial resolution.
External stimulation is desired to study the dynamics of DA release and uptake and its
correlation to the animal behavioral changes. Previously, electrical stimulation was used as
a neuromodulation technique for such purpose, which can cause a significant amount of
nondopaminergic system activation and result in consequential neurological activities or
dynamics not related to DA release[1]. Utilizing the cell-type specificity of Optogenetics,

2

researchers can have a more controlled manipulation of the dopaminergic system and have
an unbiased study on DA related neurological diseases. Specifically, Threlfell et al. [15],
have demonstrated that light-induced activation of cholinergic interneurons can trigger
dopamine release in mouse striatum through the activation of nicotinic receptors. Bass et
al. [16] have studied optically evoked DA release in rat substantia nigra with different
quantities and light pulse width, where dopamine release is found to be more sensitive to
the changes of the optical pulse width. Melchior et al. [17] have compared electrical and
optical stimulation on dopamine terminals in the nucleus accumbens. Lu et al.[1] have
studied different optical light stimulation parameters on direct manipulation of DA release
and dynamic in the nucleus accumbens, such as light intensity, pulse width, and the shape
of stimulation waveforms

1.2 Current challenges on hardware development
In order to fully realize the remarkable potential of studying DA dynamics using
Optogenetics, engineering tools are in demand to achieve simultaneous light delivery,
electrophysiological and neurochemical recording. Despite the significant development of
a wide variety of Optogenetics and Neurotransmitter detection tools, several challenges still
remain such as localized heating due to LED activation, material compatibility and safety,
and electrochemical performance of the electrodes. The following sections will discuss
these challenges in the current approaches and envision possible solutions to the identified
problems.

1.2.1Thermal challenges of putting ÂľLED near tissue
Recently, advanced microfabrication techniques have been investigated to construct
and miniaturize optical neural implants capable of multi-site, localized light stimulation of
3

three-dimensional (3D) brain networks with fine spatial resolution. These devices can be
categorized into two major groups based on different light sources: laser, including laser
diodes and diode-pumped solid-state (DPSS) laser diodes, and LEDs, including bulk LEDs
and microscale LEDs (ÂľLEDs). Although lasers and laser diodes provide several benefits,
including high light intensity, low beam divergence, and narrow spectral bandwidth, laserbased optical systems have the following drawbacks such as high power consumption with
typical several tens of mW per channel, difficulty for integrating with wireless telemetry
and restriction of natural behavior of the subjects by tethered optical fibers and
commutation systems[18]. Compared to laser and laser diodes, LEDs provide unique
advantages, including low power consumption, illumination stability, and fast lightswitching ability[19]. More importantly, electronically driven LEDs are particularly
suitable for integration with wireless telemetries to enable fully implantable systems for
applications in freely behaving animals[20]. However, they are not without significant
concerns. These include potential thermally-induced tissue damage due to ÂľLED heat
deposition

in

the

brain,

particularly

for

high-density

neural

implants

where

microelectronics are in direct contact with large-area brain tissues. To prevent tissue damage
and consequent behavioral and physiological changes, the temperature perturbation induced by
optical neural implants should be less than 1 °C [21],[22]. Therefore, there are several important
considerations that should be taken into account when designing LED-based optical devices. First,
device layout and ÂľLED array configuration can be optimized to minimize electrical heat
generated from ÂľLEDs. Second, the proper selection of substrate materials can potentially reduce
localized heating effects by dissipating the LED heat into surrounding brain tissue. The high
thermal capacity of brain tissue can counteract the temperature variation. Third, optical stimulation

4

parameters should be optimized to enable effective opsin activation, while preventing the
overheating of brain tissue.
In order to reduce electrical heat generated during the operation of ÂľLEDs, the thermal
performance of ÂľLEDs has been explored analytically and experimentally[23],[24],[25],[26].
LEDs with different dimensions were fabricated[25] on a poly(ethyleneterephthalate) (PET)
substrate. The thermal performance was quantified by measuring the maximum temperature
change upon activating the LEDs under different conditions using a thermal imager. The following
findings are derived from these studies. First, increasing the LED size can lead to an increase in
the maximum temperature change and a decrease in the overall energy efficiency. Second, when
designed in an array configuration, increasing the separation between ÂľLEDs can effectively
decrease the maximum temperature change. Finally, decreasing the pulse duty cycle can also
reduce the maximum temperature rise.
In addition, analytical and finite element method (FEM) simulations[23] have been conducted
to predict the thermal behavior of ÂľLEDs and ÂľLED arrays in tissues. Both approaches imply that
the maximum temperature change in tissues can be reduced by lowering the peak power and
decreasing the duty cycle and period of LED activation. For a µLED array, a larger 𝒓𝒅 /√𝑨 will
result in a smaller temperature change, where 𝒓𝒅 is the distance between the centers of two
adjacent ÂľLEDs and A is the total surface area of the ÂľLED. To further reduce the temperature
variation during optical stimulation, especially the localized hot spots, a substrate material with
high thermal conductivity should be carefully selected such as polycrystalline diamond, which has
a thermal conductivity (up to 1800 W/(m∙K)) [27].

1.2.2 Material long-term compatibility and safety
One of the major challenges of fiber- or waveguide-coupled optical systems is to obtain high

5

optical coupling efficiency from the fiber (or waveguide) to the stimulation site. Microfabricated
fibers and waveguides are normally made of polymers, such as SU-8, or dielectric materials, such
as oxynitride. While polymers provide excellent mechanical flexibility and fabrication simplicity,
the absorption of water could negatively affect the long-term optical properties of the polymerbased devices. Deterioration of mechanical properties of polymer waveguides is also observed
during aging of the devices in buffered saline solution (PBS). Moreover, commonly used
photosensitive polymers, such as SU-8, have a high absorption loss near 473 nm[28][29], which
significantly reduces their light-guiding quality. Finally, the biocompatibility of SU-8 has not been
fully evaluated in chronic studies. Dielectric materials are considered to be more appropriate than
polymers because of their biocompatibility, low water permeation and absorption rates, and optical
clarity over a broad spectral region. However, thick dielectric waveguides are difficult to construct
due to stress and extended plasma-etching time. As a result, the coupling efficiency between thin
dielectric waveguides and multi-mode fiber optics can be significantly affected by the large
coupling loss at the fiber-waveguide junction. Further modification and optimization of fabrication
techniques are necessary to improve the coupling efficiency. Furthermore, silicon-based dielectric
films have shown increased dissolution in water at elevated temperatures and may require
additional encapsulating barriers for chronic applications[30].
Another major challenge of chronic neural implants is the mechanical property mismatch
between rigid implanted devices and soft brain tissue, which increases the possibility of negative
neural

response,

glial

scar

formation,

inflammation,

and

mechanically

induced

trauma[31],[32],[33]. While the mechanical rigidity can be alleviated by the use of polymer
substrates, the surgical insertion of such flexible devices into deep brain regions will be
challenging. To address this issue, a temporary coating that can stiffen the probe during the

6

insertion and be dissolved by body fluids afterward has been adopted to facilitate the implantation
of the flexible optical neural implants. Among different biodegradable polymers, silk fibroin, a
biopolymer obtained from cocoons, has been widely used in bio-integrated electronics[34]. Silk
fibroin can be dissolved by most aqueous solutions with a programmable rate of dissolution
controlled by the ratio of solvent and silk concentrations. Tae-il Kim et al. [26] successfully
demonstrated the use of silk to temporarily bond a flexible ÂľLED probe to a thick and rigid epoxy
carrier during probe insertion. The silk fully dissolved in an artificial cerebrospinal fluid (ACSF)
solution 15 min after the insertion was made. Another dissolvable adhesive, polyethylene glycol
(PEG), was used by Falk Barz et al. [35]. In this study, the PEG with a molecular weight (MW) of
1500 g/mol was quickly dissolved in electrically conducting agar-based gel in 1 min. As the
melting point of PEG with different MW can range from 4–8 °C (MW=400) to 55–62 °C
(MW=8000) [36], a careful selection should be conducted to match the temperature range of the
target implantation sites.
Furthermore, as implantable devices get miniaturized, the amount of water needed to increase
the humidity of the encapsulated environment decreases accordingly, which takes a shorter time
for implanted materials to reach corrosive levels [37]. Therefore, encapsulating materials and
techniques should be carefully considered in order to achieve long-term stability of implantable
devices. Although traditional processes such as glass-to-metal seal, ceramic-to-metal seal, and
fusion welding can provide real hermetic sealing for implantable devices, the high processing
temperature may not be compatible with polymer-based implantable devices. Recently, polymer
encapsulations, such as Parylene, polyimide, silicone and epoxy, have been widely used as a
barrier coating for electronics. Although the biostability of these materials is questionable because
polymers tend to degrade due to hydrolytic, oxidative, and enzymatic mechanisms[38], recent

7

studies have shown that Parylene encapsulation of CMOS circuitry can survive at 55 °C for five
months. Besides, metal coated Parylene barriers may further reduce the permeability of moisture
and can remain intact in vivo for over 10 years[39]. Atomic-layer-deposited alumina-Parylene
bilayer encapsulation has also been studied, where a Utah electrode array (UEA) with an ASIC
chip survived for 228 days of soaking testing at 37 °C[40].
Finally, biocompatibility has always been an important criterion of all the implantable devices
to prevent glial formation and other foreign body reactions that present significant risks for devices
and host tissue. Particularly for Optogenetics applications, glial encapsulation can increase the
backscattering and attenuate light delivered to host tissue[41]. Common strategies for minimizing
foreign body responses include careful selection of biomaterial coatings, surface modification, and
optimization of device design to reduce the size and mechanical mismatch. Considerable work on
biomaterials and biocompatibility issues for neural implants has been compiled in [42],[43],[44]

1.2.3 Electrochemical performance of the electrodes
For effective recording and sensing both electrophysiological (electrical) and
neurophysiological (chemical) signals, an ideal electrode should have the following
important features[45],[46],[47], [48], [49]: Enough potential windows for sensing the target
analyte, smaller double-layer capacitance and high sensitivity to achieve higher signal-tobackground-noise ratio, resistance to surface biofouling and miniaturized size for largescale sensing with high resolution. Many efforts have been put onto microfabrication of
miniaturized neural probes for electrophysiological recording at single-cell resolution in
vivo[50]. However, there are still limitations on device development of extracellular
concentrations of neurotransmitters in vivo[51]. Traditional polymer-based Micro-ElectroMechanical-Systems (MEMS) devices are often based on metal electrodes[52]. However,
8

metal electrode requires surface modification or pre-treatment methods to increase the
sensing surface and facilitate electron transfer process[53]. For example, Aneliya et al. [54]
was reported a platinum sensor modified with conducting polymer poly-(3,4ethylenedioxythiophene, PEDOT) and inorganic Cu crystals of appropriate size to achieve
selective detection of the neurotransmitter dopamine in the presence of ascorbic acid.
Besides, the commonly used electrode materials, such as platinum, gold, iridium, iridium
oxide, cannot survive from long-term implantation. They often failed due to corrosion,
astrogliosis and fibrotic encapsulation[55], [56], [57]. Conductive polymers such as PEDOT
are seen to modify the metal surface for better biocompatibility but the stability of such
method is still unknown.

1.3 Solutions to the challenges and objective of this work
Carbon materials have been widely used in many biosensing applications[58],
especially sp3 carbon, i.e. diamond. Diamond is a unique material with complete sp3
hybridization carbon, which results in an extensive tetrahedral bonding between carbons
throughout the lattice and leads to many exceptional properties[59]. Diamond has very high
thermal conductivity (up to 1800 W/m·K), which is much higher than most noble metals
and biocompatible polymers. A comparison of thermal conductivity of the different
material is listed in Table 1-1. Besides, diamond is a biocompatible material without any
cytotoxic or hemolytic [60],[61] and is resistant to corrosion and surface adsorption and
deactivation[62],[63], which enables device long-term reliability and stability. In addition,
diamond can be doped with a dopant (boron for p-type and phosphate for n-type) to conduct
electricity. For example, the electrical resistivity of boron doped polycrystalline diamond
(BDD) can reach 1.69×10-3Ω·cm with a doping level of 6000ppm[64]. More importantly,

9

BDD has relative smaller double layer capacitance and wider potential window, which
gives low background current and a wider range of potentials for electrical neural recording
and chemical sensing. Therefore, all of the properties that diamond featured above have
been proved to be favorable to implanted neural interfaces for Optogenetics.
Table 1-1 Thermal conductivity of noble metal and polymers [65], [66], [67], [68], [69], [70]
[71], [72]

Thermal
Material

Thermal
Material

conductivity

conductivity

Silicon[65]

149 (W/m·K)

Copper[66]

385 (W/m·K)

Gold[67]

314 (W/m·K)

Parylene[68]

0.084 (W/m·K)

Iridium[69]

147 (W/m·K)

SU8[70]

0.3 (W/m·K)

Platinum[71]

71.6 (W/m·K)

Polyimide[72]

0.12 (W/m·K)

Despite its many advantages, BDD is a rigid material with Young’s modulus of ~1000 GPa
[73], which is several orders of magnitude higher than that of the brain tissues (~103 to 105 Pa
[74]). The micromotion-induced strain between rigid implants and surrounding soft tissues has
been hypothesized to cause a harmful immune response and even irreversible tissue damage[75].
Recently, mechanically flexible, polymer-based neural implants have shown promises as the next
generation of implanted devices[43],[76],[77],[78]. For those devices, electrodes and
interconnecting traces made of noble metals were constructed on soft polymeric substrates with
low Young’s moduli, such as polydimethylsiloxane (PDMS) (360-870 KPa [79]), polyimide (2.5
GPa [72]), SU-8 (SU-8 2000, 2.0 GPa [80]) and Parylene (2.8 GPa [81]). Consequently, the overall
effective Young’s modulus can be significantly reduced to minimize the mechanical mismatches
between rigid metal and soft tissues. Unfortunately, unlike noble metals, BDD cannot be fabricated
directly on a polymer substrate due to its high synthesis temperature (500 – 900 °C [82]) exceeding

10

the glass transition temperatures of polymers. To address this issue, a wafer transfer process is
required to transfer diamond from BDD growth substrates, such as silicon, onto flexible polymer
substrates.
The ultimate goal of this thesis work is to fabricate an integrated diamond neural
interfacing system using the diamond on a flexible Parylene substrate, which combines
three modules- a ÂľLED for optical stimulation with diamond traces as a heat spreader,
neurophysiological and neurochemical sensing capability with diamond electrodes. A
concept diagram is shown in Figure 1-1, where only one probe is shown in the figure to
emphasize the functionality of each module. The work presented in this thesis will be the
design and implement of each module, which includes a hybrid neural interface optrode.
with a polycrystalline diamond (PCD) heat spreader and a novel large-scale wafer transfer
process for fabrication of diamond-polymer chemical sensors.

Figure 1-1 Concept diagram of a diamond based opto-electro-chemical hybrid neural interface.

11

1.4 Layout of the Dissertation
Chapter 2 summarizes the state-of-art of Optogenetic technology and light delivery
strategy. Chapter 3 gives a review of the state of art diamond-based devices and the theory
of electrochemistry for chemical sensing. Chapter 4 introduces an SU8 based ÂľLED probe
for Optogenetics, which serves as a comparison for evaluating the high thermal
conductivity of diamond probe. Chapter 5 reports a hybrid neural interface with a diamond
as a heat spreader for optical stimulation and neurophysiological recording. The heat
spreading performance is evaluated using a high-resolution infrared camera and compared
with an SU8 probe with same dimension and layout reported in Chapter 4. The results show
that the maximum temperature of the diamond probe is ~90% lower than that of the SU8
probe. Besides, the functionality of the diamond probe was tested in vivo where lightevoked action potentials were successfully detected. Chapter 6 shows a novel wafertransfer process of transferring BDD patterns from diamond growth onto a flexible Parylene
substrate. The electrochemical properties of the transferred BDD-polymer electrodes are
evaluated using (i) an outer sphere redox couple Ru(NH3)62+/3+ to study electron transfer process
and (ii) quantitative and qualitative studies of neurotransmitter redox dopamine/dopamine-oquinone. A linear response of the BDD sensor to dopamine concentrations of 0.5 ÂľM to 100 ÂľM
was observed (R2 = 0.999) with a sensitivity of 0.21 ¾A/cm2¡¾M. Finally, conclusions and future
works are given in Chapter 7
.

12

Chapter 2.

A review of Optogenetics and light delivery
methods

2.1 Microbial Opsins
The core components used in Optogenetics are light-sensitive microbial opsin. There
are four major types of opsins: Channelrhodopsin (ChR), Halorhodopsin (HR),
Bacteriorhodopsin (BR) and Opsin-receptor chimaeras OptoXRs as demonstrated in Figure
2-1. Channelrhodopsins are light-activated cation channels. The direction of net
photocurrent due to ChR activation is down the electrochemical gradient, which polarizes
membranes and enables action potentials. Halorhodopsin (HR) is a chloride pump, which
pumps chloride ion from extracellular into intracellular space. Similar to Halorhodopsin
(HR), Bacteriorhodopsin (BR) is a proton pump, which pumps protons from cytoplasm to
extracellular

medium.

Both

Halorhodopsin

(HR)

and

Bacteriorhodopsin

(BR)

hyperpolarizes membranes and inhibits neural activities. OptoXRs refers to opsin-receptor
chimaeras, which can initiate light-activated G protein-coupled biochemical signaling
cascades in targeted neurons. If we categorize the opsins according to the functionality,
they can be grouped as excitation, inhibition, bi-stable modulation and modulation of

Figure 2-1 Major classes of single-component Optogenetics tool. (Reprinted from[83])
13

intracellular biochemical signaling. The details of each functional group will be described
in the following sections.

2.1.1 Fast Excitation
Channelrhodopsins(ChRs) were first identified by Nagel et al. in 2002[84], where a
protein encoded by one of the genomic sequences from green algae Chlamydomonas
reinhardtii show light-modulated ion-flux property. Then the initial demonstration for
neuroscience application was done by Boyden et al.[85]. However, there are several
limitations for those early stage wild-type ChRs. Firstly, the light spectrum of the early
stage ChRs is in the blue light wavelength, which has limited penetration depths into neural
tissue as light strongly absorbed and scattered compared with higher wavelength lights such
as yellow – red wavelengths [86]. Secondly, the wide-type ChRs do not have fast enough
off-kinetics to achieve reliable spiking above 40Hz, which is required by many neuronal
cell types and physiological processes[83]. Lastly, wild-type ChR2s have relative small
photocurrent. The evoked photocurrent under illumination is the current for depolarizing
the neurons, which depends upon many factors, including the properties of the opsin being
expressed, the wavelength, intensity and duration of the incident light, and even recent
illumination history[83], [87]. Small photocurrent requires high light intensity to evoke
reliable action potentials, especially when stimulating through thick layers of tissue[88]. To
address these limitations, the molecular modification has been extensively used for
engineering ChR to achieve certain desired improvements of opsins such as larger peak
activation wavelength, faster off-kinetics or higher photocurrent. To date, most of ChR used
for Optogenetics have been molecular modified and those variants with different kinetic
and spectral attributes are summarized in the blue circle of Figure 2-2. Although the

14

successful improvement introduced by gene mutation, often times, an improvement of one
aspect of the opsin will impair other aspects of the opsin. For example, the introduction of
H134R to ChR2 can increase photocurrent magnitude by 2-fold at the expense of slower
channel closure kinetics by 2-fold, which gives a poorer temporal precision[89].
Enormous interests have been shown in the development of a red-shifted opsin that can
manipulate two isolated neuronal populations with either pre-existed ChR2 or red-shifted
opsin expressed under the same volume of tissue. Such opsin requires having a red-shift of
more than 50nm on the peak activation wavelength to that of ChR2 in order to avoid any
interferences or crosstalk between each population. However, current molecular
modification of ChR2 can only have red-shifts of ~30nm, which is between the safety
spectrum separations. Zhang et al.[90]. discovered an opsin (VChR1) from Volvox carteri.,
which has ~75nm red-shift to ChR2 and sufficient to achieve independent manipulation.
However, the VChR1 expressed neurons show low expression in mammalian neurons with
small photocurrents. Then a new family of chimaeric red-shifted opsins was created by
Yizhar et al.[91], which comprises of sequences of ChR1 and VChR1 fragments and has a
peak activation wavelength between 535 to 545 nm. ChR1/VChR1 chimaeras (C1V1) show
enhanced expression and photocurrent in HEK cells compared with only VChR1 expressed
cells. A summary of VChR1, chimaeras C1V1 and its variants with different peak
activation wavelength and kinetics is shown in the green circle of Figure 2-2.

2.1.2 Fast Inhibition
Above mentioned excitatory opsins like ChR2 can only test the sufficiency of the
contribution of the targeted cells to the neural circuitry or behavioral property[92].
However, multiple different cell populations could be involved and give rise to the same

15

consequences. An ideal solution to test the necessity of targeted cell is to complement
excitatory opsins with inhibitory opsins to permit both excitation and inhibition to the
targeted neural population with independent light control[92]. Light-driven chloride pump
- bacteriorhodopsin and proton pump - Halorhodopsins are two types of opsins that
homologues to ChR2 and proved to be capable of inhibiting neural activity.
An initial study of chloride pump was conducted by Zhang et al.[93], which focused
on Natronomonas pharaonic (NpHR) and Halobacterium salinarum (HsHR). Both NpHR
and HsHR expressed cells lead to rapid outward currents under illumination of light with
maximum activation wavelength of 580nm, which is red-shifted enough from the maximum
wavelength of ChR2, 460nm. This gives the capability of activation independently of
excitatory and inhibitory opsins. Theoretically, a chloride pump should have a low affinity
on the intracellular regime for neuron activity inhibition, where Cl 2 ions are released.
Further investigation found that HsHR has a lower extracellular Cl2 affinity and showed
rapid rundown of current at low extracellular Cl2 concentration and does not recover
without light illumination. On the contrary, NpHR showed higher extracellular Cl2 affinity
and stability and was chosen as inhibitory opsins for optical inhibition study.
Chow et al. (Mac, Arch) [7] and Gradinaru et al. (eBR)[94] explored the use of proton
pumps as Optogenetic tools for neural activity inhibition. Opsin archaerhodopsin-31
(Arch)[7] from Halorubrum sodomense can achieve almost 100% silencing of neurons in
the awake mouse cortex under yellow-green light illumination (566nm) and can mediate
current up to several hundred pico-amps with low light intensity and shows very fast
kinetics of recovering from light-dependent inactivation. Opsin Mac[7] from fungus
Leptosphaeria maculans provides an opportunity to silence neural activity using blue light,

16

which enables the potential possibility of silencing two neural populations with different
colors of light (blue versus red). Opsin eBR[94] is adopted from BR for optimal membrane
trafficking, which can deliver 50pA of outward photocurrent and 10mV hyperpolarizations
with optimal wavelength light of 560nm, which is sufficient to silence spiking in the
hippocampal pyramidal neuron. A summary of fast inhibition opsins with different peak
activation wavelength and kinetics is shown in the red circle of Figure 2-2.

2.1.3 Step-function opsin (SFO)
Step function opsins are a group of ChR mutants that the photocurrents of which can
be precisely turned on and off with different colors of light. The SFO opsins have vastly
longer time constants than the wild-type ChR2. For example, the time constants of C128T,
C128A, and C128S mutants are 2s, 42s, and 100s[95], compared with wild-type ChR2 of
~10ms. With the additional and combinatorial mutagenesis of these early SFO opsins,
Yizhar et al.[91] reported stabilized SFOs with time constant of 30min (mutant
128S/156A), which can be used to “step” targeted neurons to a stable depolarized potential
with blue light (470nm), followed by removing the light source and starting behavior or
physiological study. Then, the SFOs can be deactivated with a yellow light pulse (590nm)
illumination. The benefit of SFOs is to rule out the photocurrent artifacts caused by light
illumination, which always mingles with the recorded neural signals upon optogenetic
stimulations. Besides, ChRs with SFO mutants is responsive to the light intensity at least
300-fold lower than the wild-type ChR[95]. This can lead to excite a larger population of
neurons at a given light intensity[83] and prevents the potential tissue over – heating effect
by the light sources[83]. A summary of SFO and its variants with different peak activation
wavelength and time constants is shown in the pink circle of Figure 2-2, where the

17

activation wavelength of ChR2 SFO and VChR1-SFOs are 470nm and 560 nm,
respectively and deactivation wavelength are 590nm and 390 nm, respectively[95], [46],
[91].

2.1.4 Biochemical Modulation
The microbial opsin genes described above are defined as type I, which primarily
modulate ion flow to control the excitation or inhibition of a neuron by manipulating the
membrane potential to either depolarize or hyperpolarize the cells. Another type of opsin –
vertebrate rhodopsin, defined as type II, is a tool for modulating intracellular biochemical
signaling. Vertebrate rhodopsin is both a type II opsin and a G protein-coupled receptor
(GPCR), which can modulate the G protein signaling on the intracellular side by adsorption
of photons. Optogenetic modulation of biochemical signaling can be achieved by
constructing chimeras[96], which are referred to as OptoXRs (opsin–receptor chimaeras).
In those OptoXRs, the intracellular loops of vertebrate rhodopsins are replaced with
conventional ligand-gated GPCRs of the host cells, such as dopaminergic, serotonergic and
adrenergic receptors for optical control of intracellular signaling in freely moving mice[97].
The engineering tools designed for type I opsins can be adapted into type II opsins for
biochemical signaling. Airan et al.[97] reported Opto-β2AR and Opto-ι1AR chimaeras for
modulating adrenergic Gs-protein signaling and Gq-protein signaling, respectively with a
laser diode-coupled fiber optic devices in nucleus accumbens. A Rh-CT(5-HT1A)
chimaeras was published by Oh et al.[98] as a suitable proxy of agonist-induced 5-HT1A
receptor activation, which is linked to Gi/o signaling pathway upon light stimulation.
Besides, microbial photoactivated adenylyl cyclase (AC) shows low cyclase activity
in darkness but much higher in the light, which catalyzes adenosine triphosphate (ATP) to

18

3’,5’ –cyclic AMP (cAMP) and pyrophosphate. The cAMP produced by AC is then served
as a second messenger for intracellular signaling. Therefore, by manipulating light
illumination, photoactivated AC can be used to modulate the activity of the second message
in the intracellular signaling. Stierl et al.[100] published that photoactivated adenylyl
cyclase (bPAC) mediate light-dependent cAMP increases in Drosophilia central nervous
system. A Blac gene encoded photoactivated adenylyl cyclase was designed by Ryu et
al.[101] with blue light sensitivity to modulate second cAMP and cGMP levels in vivo. A
summary of as mentioned chimaeras and photoactivated AC is shown in Figure 2-2 as
highlighted in orange.

Figure 2-2 Kinetic and spectral attributes of Optogenetic tool variants. The variant refers to ChR2
mutation if not specified (Reprinted from [99])

2.2 Optogenetic Neural Implants
In order to fully realize the remarkable potential of these opsins, engineering tools for
simultaneous light delivery and electrophysiological recording is needed. For in vitro light
19

delivery, in 2005, Boyden et al.[102] demonstrated reliable, millisecond, single-component,
genetically targeted optical neuromodulation, where ChR2-expressing hippocampal neurons were
excited using an incandescent lamp (450–490 nm, 300W) with a chroma excitation filter, and the
light-induced neural activity was recorded using a whole-cell patch clamp. Following that,
Ishizuka et al.[103] utilized a surface- mounted, blue-light-emitting diode (LED) (470–490 nm) to
quantify the relationship between the light-gated current and the intensity of blue light illumination
on ChR2-expressing hippocampal cell cultures. Other in vitro optical instruments have also been
reported, such as a focused laser beam using acousto-optic deflectors[104] and digital micro-mirror
devices (DMDs)[105],[106]. Although these in vitro approaches can successfully activate neural
activity in both cultured neuronal and acute slice preparation, they are not suitable for in vivo
stimulation in the intact brain or for study in freely behaving animals.
The first demonstration of functional optical control of intact animal brains was reported in
2007 by Dr. Deisseroth’s group [107]. In their studies, the motor cortex of living rodents was
stimulated through an intracranial, multimode, optical fiber coupled to a solid-state laser diode
system, with an output light intensity of ~380 mW/mm2. Since then, many implantable light
delivery systems have been implemented by coupling a thick optical fiber of a few hundred
microns to a laser or LED light source. Such systems have been used to study the light-evoked
neural activity as well as behavioral changes in commonly used animal models, both small
(mice/rats) [108],[109],[110],[111],[112] and large (non-human primates) [113],[114], [115],
[116]. These systems, however, inevitably activate many uninterested neurons and are impractical
in the spatial control of multi-site stimulation in large-scale neural networks. Therefore, there has
been an increased need for the development of implantable, reliable light delivery and recording
interfaces with high spatiotemporal resolution and spectral control ability[83].

20

Recently, advanced microfabrication techniques have been investigated to construct and
miniaturize optical neural implants capable of multi-site, localized light stimulation of threedimensional (3D) brain networks with fine spatial resolution. These devices can be categorized
into two major groups based on different light sources: laser, including laser diodes and diodepumped solid-state (DPSS) laser diodes, and LEDs, including bulk LEDs and microscale LEDs
(ÂľLEDs). Optical fibers, microwaveguides, channel waveguides, and tapered optrodes are most
commonly used to guide light from sources to target neurons. Microfabricated probes with ÂľLEDs
mounted directly at the tip of the probe shaft have also been implemented by several groups.
Furthermore, monolithic integration of miniaturized optical elements with multi-electrodes and
wireless interfaces enables spatially-confined optical stimulation and simultaneous recording of
light-evoked neural activity in freely moving animals.
This section reviews some of the representative microimplants for Optogenetic applications
and their related fabrication technologies. Section 2.2.1 summarizes microscale optical implants
based on lasers or laser diodes. Section 2.2.2 is devoted to microimplants based on LED light
sources.

2.2.1 Laser-coupled Optical Neural Implants
Effective photostimulation of Optogenetic opsins requires the minimum irradiance of
1 (or 7) mW/mm2 for neural excitation (or inhibition)[117]. The practical requirement of
irradiance is also affected by the high degree of light scattering and absorption in neural
tissue[83]. For these reasons, fiber-coupled lasers with high power are being widely used
as light sources for many Optogenetic experiments[109],[117],[87],[118],[119]. A laser can
generate coherent light with unique characteristics: low divergence to focus the light beam
over a long distance and high temporal coherence on confining the bandwidth of emitted

21

light within a narrow spectrum. The former characteristic allows light to be steered through
optical fibers to target cells with lower loss than with incoherent light sources (e.g., LEDs).
This results in more efficient coupling between light sources and fibers with thin core
diameters of 50 microns or less. The latter characteristic enables the high efficiency of
optical stimulation since the majority of irradiance will fall into the peak activation spectra
of microbial opsins and contribute to optical stimulation. To deliver laser light into target
cells, waveguiding structures must be used and are typically implemented by several
fabrication

techniques,

including

glass-sharpened

optical

fibers,

out-of-plane

microwaveguide arrays, and in-plane microwaveguide probes. In the following sections, I
will discuss the device configurations and fabrication techniques of different laser-coupled, optical
neural interfaces. Representative prototypes are presented in Figure 2-2, and their specifications
are summarized in Table 2-1 [76], [120], [121], [122], [123], [124], [125]. [107], [109], [118],
[126], [127].

2.2.1.1 Glass-sharpened optical fibers
Single site glass-sharpened optical fibers are typically made of commercially available
multimode optical fibers with core diameters of ~200Âľm. To reduce the thickness of a
multimode fiber for localized optical stimulation, in some approaches the plastic cladding
layer of the fiber was stripped and the bare glass core with a minimum diameter of 100 Âľm
was guided into a rodent brain through an implanted cannula[128],[87]. Wet chemical
etching is often employed to sharpen the tip of the glass core in order to further improve
spatial resolution and minimize the tissue damage during device insertion. Figure 2-2(a)
shows combine a multimode optical fiber attached with four tetrode bundles for
electrophysiological recording[109]. Figure 2-2(b) shows a dual-core optical fiber system.

22

One optical core for optical stimulation and one hollow core filled with 1-3M NaCl for
electrical recording[119].

2.2.1.2 Out-of-plane microwaveguide arrays
These devices are normally micromachined, employing thin out-of-plane waveguide
shanks with tapered tips to improve spatial resolution and reduce implant invasiveness. The
light illuminated by laser light sources is butt-coupled to the waveguide shank and then
emitted from the tip for neural stimulation. The optical waveguides can be readily
integrated with silicon Utah multielectrode probes for simultaneous stimulation and
recording of neural activity. One such device is a SiO2 Utah waveguide array capable of
optical stimulation with both visible and infrared (IR) light. This device consists of 10×10
arrays of optrodes 0.5 mm to 2 mm long at a 400 Âľm pitch, constructed by bulk
micromachining fused silica or quartz dices of 3 mm thickness and 50 mm diameter. A
dicing saw with a bevel blade was used to shape the pyramidal tips with a precisely
controlled taper slope[127]. Furthermore, Zhang et al. reported a dual-modal optrode
array[126],[129],[130],[131] (Figure 2-2(c)) modified from a previously developed silicon
Utah multielectrode array. In their design, one of the 100 silicon shanks was replaced with
a multimode optical fiber, by removing a shank, drilling a hole using ablative laser
machining, inserting the fiber through the hole mechanically, and then bonding the fiber
with adhesive epoxy.

23

Figure 2-3 Examples of laser-based optical neural interfaces: (a) A dual-core optical fiber system
with one optical core for optical stimulation and one hollow core filled up with 1-3M NaCl for
electrical recording. (Reprinted from [67]) (b) A multimode optical fiber with four tetrode bundles
attached for electrophysiological recording. (Reprinted from [69]) (c) A dual-mode optrode array
adapted from a Utah multielectrode array, where one recording shank was replaced with a
multimode optical fiber. (Reprinted from [72]) (d) An in-plane neural probe adapted from
conventional Michigan neural probe with embedded dielectric waveguides and microfluidic
channels. (Reprinted from [79]) (e) A 3D multiwaveguide array consisting of a set of waveguide
combs assembled on a base plate-holder through two alignment and fixation pieces. (Reprinted
from [81])

24

Table 2-1 Summary of the specification of miniaturized, laser-based Optogenetic neural implants
Optical Neurostimulation Components

Light source

# of
channels

Dimensions

Output light
intensity (max. or
used)

Electrical Recording Components
Light
delivery
efficiency

Other
capabilities

Substrate
material

Ref

1.37MΊ

No

Si

[76]

# of
channe
ls

Dimensions

1kHz
Impedance

8

143Âľm in
diameter
20Âľm
separation

Optical fiber
coupled
waveguide
(Oxynitride core)

1

70Âľm wide

7mW/mm2

-10.5 Âą 1.9
dB

Bare Laser chip
coupled
waveguide
(SU8 core)

1×2

15 Âľm wide
13Âľm long

29.7mW/mm2
@ 659nm

--

2×4

20Âľm in
diameter

1.54Âą
0.06MΊ

No

Si

[120]

Optical fiber
coupled
waveguide
(SU8)

1

0.15mm in
width

60mW/mm2

-12dB

8

--

280KΊ 350KΊ

Micro
-fluidic
Channel

Polyimide

[121]

Optical fiber
coupled
waveguide
(SU8)

1

≤150µm wide

0.9mW

--

16

20µm×20µm

0.8MΊ

Micro
-fluidic
Channel

Si

[122]

Optical fiber
coupled multiwaveguide
(Oxynitride core)

12

60~360Âľm
wide
1cm long
1mm
separation

--

-10dB

0

--

--

No

Quartz

[123]

Tapered optical
fiber with multiopenings

Max. 7

600nm in
diameter

3.5mW

--

0

--

--

No

-

[124]

Lasercoupled fiber

1

200 Âľm in
diameter

~380mW / mm2.

--

0

--

--

No

--

[107]

25

Table 2-2 (cont’d)

9µm×60µm

#1 scheme:
148Âą56mW/mm2
#2 scheme:
200mW/mm2

#1 scheme:
17.3Âą1.8dB
#2 scheme:
11.9Âą2.5dB

0

--

--

No

Si

[125]

1

200 Âľm in
diameter

60~160mW/mm2
@ 473nm
160~260mW/mm2
@ 561nm:

--

4

Diameter
~25Âľm

--

No

--

[109]

Lasercoupled fiber

1

Tip diameter of
6-20Âľm
Fiber diameter
of 4 Âľm

≤10mW/mm2

--

1

0.7Âľm in
diameter

--

No

--

[118]

Lasercoupled optrode

1

50-62.5Âľm in
diameter

916mW/mm2

-1.55dB

99

1mm long
400Âľm
separation.

112KΩ 671KΩ

No

--

[126]

100

0.5-2mm long
150Âľm wide
400Âľm
separation

--

-1.49dB

0

-

-

No

SiO2

[127]

Laser3-D waveguide

Lasercoupled fiber

Lasercoupled optrode

192

26

2.2.1.3. In-plane microwaveguide probe
Compared to the out-of-plane arrays, in-plane microwaveguide probes benefit more
from modern microelectromechanical system (MEMS) technology evolved from the
process technology in the conventional semiconductor device fabrication. Most of these
probes share a similar configuration: an in-plane microwaveguide for light delivery carried
by a silicon or polymer shaft with electrophysiological recording and/or microfluidic
modalities. Several combinations of dielectric materials used for microwaveguides include:
oxynitride core (refractive index: 1.51) with oxide clad (refractive index: 1.46)[132], and
SU-8 core with either silicon oxide[120], tungsten-titanium alloy (10% titanium)[121], or
glass clad[122]. For the two designs with microfluidic modality, integrated microchannels
are constructed by either photopatterning of SU-8[121] or reflow of borosilicate glass
followed by chemical mechanical polishing (CMP)[122] (Figure 2-2(d)). Light coupling
between laser light sources and planar microwaveguides is typically achieved through
optical fibers.
Despite their significant advantages, the aforementioned devices are limited to
delivering light to a single target, and therefore not suitable for applications that require
delivering patterned light independently to distributed targets in 3D brain circuits, such as
in the rhesus macaque cortex[123]. From a fabrication perspective, a straightforward approach
to increase the spatial density of optical stimulation is to assemble 3D arrays with planar, multishank waveguide probes, using possible methods originally developed for 3D Michigan type
multielectrode arrays. Such methods include backbone stacking and bonding[133], folded
Parylene cable[55], and orthogonal insertion of planar probes into a carrying platform[134].
Making planar waveguide probes with multiple shafts can be achieved simply by modifying

27

photomask designs without increasing the complexity of device fabrication. However, the spatial
density is still limited by the number of shafts. Therefore, the ability to deliver light through
multiple sites along a single probe shaft will provide a major breakthrough for high-resolution
spatial photostimulation. Recently, optical probes with spatially distributed emitting sites
along a single probe shaft were reported by Zorzos et al.[123]. In this approach, twelve
varying-length dielectric microwaveguides were lithographically patterned on the same
shaft. As the waveguides can be separately coupled to different light sources, this device
enables independently addressable optical stimulation at each output with adjustable
wavelength. To expand the spatial resolution into three dimensions, a 3D multiwaveguide
array[135] (Figure 2-2(e)) was implemented, consisting of a set of waveguide combs
inserted orthogonally into a base plate-holder with the assistance of two alignment and
fixation pieces. These devices, while successfully demonstrated, still have a large footprint
due to the requirements of multiple waveguides and light sources.

2.2.2 LED-Based Optical Neural Implants
Although lasers and laser diodes provide several benefits, including high light
intensity, low beam divergence, and narrow spectral bandwidth, laser-based optical systems
have the following drawbacks. First, lasers are power hungry with typical power
consumption of several tens of mW per channel. Second, when used with freely behaving
animals, lasers require the use of tethered optical fibers and commutation systems, which
greatly restrict the natural behavior of the subjects, require costly optical commutators, and
may bias the outcomes[18]. Third, the activation of laser diodes may require relatively high
voltage/current, and the possibility of localized heat generation may damage surrounding
tissue. Compared to laser and laser diodes, LEDs provide unique advantages, including low

28

power consumption, illumination stability, and fast light-switching ability[19]. More
importantly, electronically driven LEDs are particularly suitable for integration with
wireless telemetries to enable fully implantable systems for applications in freely behaving
animals[20]. A variety of ÂľLED processing approaches has been developed by many
researchers, such as Jeon et al.[136], Zhang et al.[137], and Kim et al.[25]. The fabrication
of ÂľLED-coupled optical probes relies on two basic stereotypes of neural probes being used
for electrical stimulation: Utah-type[138] and Michigan-type[139]. In the following sections,
the device configurations and fabrication techniques of different ÂľLED-coupled optical neural
interfaces are discussed in detail. Representative prototypes are presented in Figure 2-3 and their
specifications are summarized in Table 2-2 [140], [141], [142], [143], [144], [145], [76], [26],
[146]

2.2.2.1. Utah-type optical arrays
The Utah neural probes, which were made by bulk micromachining thick boron-doped
silicon substrates, have been widely used for electrical stimulation and chronic neural
recordings[147],[148]. Compared with the Michigan probe, the Utah-probe topology
enables the arrangement of high-density shanks in a 3D configuration. Taking this
advantage, the Utah-probe topology has been adopted to make LED-coupled optical probes
for Optogenetic applications. Two main designs of Utah-type optical probes include planar,
surface-mounted LED arrays and 3D arrays with ÂľLEDs coupled to optical fibers or
waveguides. Probes based on the former design are primarily used in in vitro studies with
cell culture and brain slice preparations, while the latter design targets in vivo studies in the
deep cortical layers and brain regions of living animals. The following two sections will

29

discuss in more detail the surface-mounted ÂľLED arrays and optical fiber/waveguidecoupled ÂľLEDs array.

2.2.2.1.1 Surface-mounted ÂľLED arrays
Grossman et al.[136],[149],[142],[150] (Figure 2-3(a)) used conventional silicon-based
microfabrication technology to build the first custom designed, high-power ÂľLED array,
which can generate arbitrary optical excitation patterns with micrometer and millisecond
resolution. Despite the successful demonstration of optical modulation of neural activity
with high density and high spatial resolution, this type of probe suffers from disadvantages
related to the integration of neural recording capability and heat generation due to highdensity ÂľLED illumination. In the in vitro study, because a whole cell patch clamp was
used for neural signal recording, it was difficult to record signals from multiple neurons
simultaneously. In addition, these ÂľLED arrays would have difficulties in thermal
management, especially when operated at high frequency and long duration, due to their
ultra-high density (a 64×64 µLED array with a small pitch of 50µm). The excessive heating
may cause tissue damage as well as physiological and behavioral change[151], which may
bias the outcomes of Optogenetics.
Beside custom made ÂľLED arrays, commercially available die-form LED chips are
employed in the fabrication of surface-mounted optical arrays. Polymers, such as ParyleneC, SU-8, and polyimide[43], have been used as carrying substrates and insulating layers of
the LED chips, due to the mechanical flexibility, biocompatibility, chemical resistance, and
stability of the polymers. As an example, Kwon et al.[152],[153] (Figure 2-3(b)) reported a
multichannel Opto-ÂľECoG array, which combines a transparent microelectrode array and
a ÂľLED array on a flexible Parylene-C substrate for epidural optical stimulation and

30

electrical recording of cortical activity. The ÂľECoG array featured a Parylene-indium tin
oxide

(ITO)-gold-Parylene

sandwich

structure

for

electrical

recording

without

compromising the optical throughput. The ÂľECoG array and the ÂľLED array were
fabricated separately, and then aligned and bonded together using adhesive polymers. It is
of note that the maximum temperature variance of this device was ~9 °C and ~1 °C when
the LED was driven by 100 ms voltage pulses of 3.2 V and 2.7 V, respectively. Hence, the
application of this probe is limited to short pulses with low applied voltage.

2.2.2.1.2 Optical fiber/waveguide-coupled ÂľLED arrays
The surface-mounted ÂľLED arrays have limited stimulation depths due to the
scattering and absorption of LED light by neural tissue[151]. In view of such shortcomings,
significant developments have been made to couple ÂľLED light into waveguiding
structures, such as optical fibers, microwaveguides, and optrodes, for delivering light into
deep brain regions. For example, a 3×3, LED-coupled, optical fiber array was reported by
Schwaerzle et al.[140],[154] (Figure 2-3(c)). In this design, commercially available ÂľLED
were flip-chip bonded onto a polyimide substrate with patterned metal interconnects. DRIE
was used to machine a miniaturized silicon housing plate that carried the ÂľLED chips and
aligned them to optical fibers with high precision. The equivalent coupling efficiency
between LEDs and fibers was measured to be ~ 0.88-1.27%. When the LEDs were driven
by 30 mA current pulses, the maximum temperature variations on the silicon housing were
5 °C, 10 °C, and 15 °C, with 5 %, 10 % and 15 % pulse duty cycles
Recently, LED-coupled SU-8 microwaveguide arrays have been developed by Kwon
et al.[141],[155],[156],[157],[158],[159],[160] (Figure 2-3(d)), using polymer-based MEMS
techniques. These 3D arrays were equipped with single-length or varying-length slanted

31

microwaveguides to deliver light to the same or different layers of the cortex. Backside UV
exposure of an SU-8 photoresist was used to create taper-shaped waveguides with the tip
and base diameters controlled by adjusting the mask designs and the separation between
the mask and the resist layer. Building on the 3D waveguide arrays, an ITO-Parylene-goldParylene sandwich clay was coated on the outer sidewalls of the SU-8 waveguides to serve
several functions. In particular, the ITO layer was used to eliminate photoelectrical artifacts
without compromising the light transmission. The opaque gold layer was used to block out
the light leakage and record light-evoked neural activity. The coupling efficiency was
estimated to be ~10 %, which is close to the maximum coupling efficiency of LED-fiber
butt coupling[161]. The maximum temperature variation of the LED was ~ 9 °C with an
applied voltage of 3.2 V and 100 ms activation duration.
While the above fiber/waveguide-coupled ÂľLED arrays have been successfully tested
in vivo, the main limitation of these devices is the low efficiency of the LED-fiber butt
coupling, which is typically less than 10 %. In order to maintain sufficient light intensity
for effective photostimulation of opsins, high applied voltage and long duration of driving
pulses are necessary, which will inevitably increase the local temperature. The low thermal
conductivity of the polymer substrate and package also prevents heat transfer from the
devices to the surrounding biological environment, therefore increasing the risk of
thermally-induced tissue damage.

2.2.2.2 Michigan-type optical probes
In order to achieve highly efficient light coupling, an alternative light delivery strategy
has been explored by inserting ÂľLEDs directly into deep brain targets of interest. Wellestablished Michigan-type probes provide an ideal platform where both custom designed

32

and commercially available ÂľLED chips can be mounted onto the probe tip as light sources
for optical neuromodulation. Michigan-type optical probes constructed with commercial
ÂľLED chips have been reported by many research groups.
Although using commercially available ÂľLED chips can reduce the complexity of
device fabrication and ÂľLED assembly, the dimensions of ÂľLEDs are limited by factory
specifications, which make it difficult to miniaturize the devices. A custom designed ÂľLED
chip will provide an opportunity for the miniaturization of Michigan-type probes to increase
the spatial resolution of photostimulation, reduce device invasiveness, and prevent
unnecessary

tissue

damage.

Both

semiconductor-based

and

polymer-based

microfabrication technologies have been explored. A representative prototype built using
traditional semiconductor technology was reported by McAlinden et al.[162],[163] (Figure
2-3(e)). In this case, a blue ÂľLED probe was fabricated from a commercial LED wafer with
epitaxial GaN structures grown on a sapphire substrate. A 7-mm-long probe carried five
LEDs (with a diameter of 40 Âľm and a pitch of 250 Âľm) on a 1.3-mm-long tip shaft for
optical neuromodulation. Laser dicing was used to shape the sapphire probe, followed by
mechanical thinning of the probe to 100 Âľm thick from the backside of the substrate. With
a 200 ms driving pulse (resulting in 600mW/mm2 light intensity), the probe had a maximum
temperature rise of less than 2 °C, which benefits from the relatively high thermal
conductivity of sapphire (23 W∙m-1∙°C-1). Despite the low heating benefit, this device
suffers from the mechanical rigidity of sapphire, which may promote a neuro-inflammatory
response, induce mechanical strains in the surrounding tissue, and lead to irreversible tissue
damage during chronic applications in freely moving subjects.
As a solution, mechanically compliant polymeric substrates can be used to carry the

33

LEDs. However, most fabrication processes for making inorganic LEDs, particularly for
blue GaN LEDs, are incompatible with polymers, due to the requirement of high processing
temperatures. Therefore, novel techniques are needed to transfer pre-fabricated ÂľLED chips
from a solid substrate (sapphire or silicon carbide) to a flexible polymer substrate. Kim et
al. [25] demonstrated a ÂľLED substrate transfer technology, by which ÂľLEDs with lateral
dimensions ranging from 1mm×1mm to 25µm×25µm can be transferred from a foreign
substrate such as sapphire to a target polymer substrate. Using this technology, Kim et
al.[26] (Figure 2-3(f)) implemented a flexible, multifunctional, neural interface probe,
which consisted of a platinum microelectrode array for neurophysiological recording, a
ÂľLED array for optical neuromodulation, a precision temperature microsensor layer for
real-time monitoring of local temperature variation, and a microscale, ultrathin silicon
photodiode array for measuring the light intensity of ÂľLEDs. Each sensor/array layer was
constructed on a flexible polyester substrate, and then stacked and bonded together with
UV-curable epoxy. A releasable base was fabricated from epoxy and bonded to the asfabricated hybrid neural probe with bio-resolvable adhesive silk to facilitate the insertion
of the mechanically flexible probe into brain tissue. The time-average temperature change
with various duty cycles at 17.7 mW/mm2 peak light output was within 1 °C when the
device was inserted 0.3mm into brain tissue. As the state of the art, these multifunctional
optoelectronics present significant advantages they are not without significant concerns for
experiments with freely moving animals. The devices were tested in vivo to demonstrate their
functionalities for optical stimulation, electrical recording, and dissecting complex neurobiology
and behavior of freely moving animals. The time-average temperature change with various duty
cycles at 17.7 mW/mm2 peak light output was within 1 °C when the device was inserted 0.3mm

34

Figure 2-4 Examples of ÂľLED-based neural interfaces: (a) A high density ÂľLED array fabricated
by conventional silicon-based microfabrication technology. Reprinted from[90] (b) A 4×4 OptoµECoG array with a transparent microelectrode array and a µLED array on a flexible Parylene-C
substrate for epidural optical stimulation and electrical recording of cortical activity. Reprinted
from [94]. (c) A ÂľLED-coupled optical fiber array with a miniaturized Si housing plate for optical
fiber alignment and fixation. Reprinted from [96] (d) A ÂľLED-coupled SU-8 microwaveguide
array fabricated by a droplet backside exposure method, where an ITO-Parylene-gold-Parylene
sandwich clay was used to minimize light-induced artifacts. Reprinted from [99]. (e) A custom
designed ÂľLED probe fabricated from an epitaxial GaN/sapphire substrate by semiconductorbased microfabrication technology. Reprinted from [105]. (f) A flexible, multifunctional neural
probe with integrated temperature sensor, neural recording microelectrodes, light intensity sensor,
and ÂľLED. Reprinted from [107].
into brain tissue. As the state of the art, these multifunctional optoelectronics present significant
advantages, including spatially precise and cellular-scale light delivery, highly effective thermal
management, mechanical flexibility, and integration with wireless components to eliminate the
need for fiber optics, tethers, and commutators. While such devices have shown great
improvements over the previous instruments, they are not without significant concerns. These

35

Table 2-3 Summary of the specification of miniaturized, ÂľLED-based Optogenetic neural implants
Optical Neurostimulation Components

Electrical Recording Components
1kHz
Impedance

In vivo or
in vitro
validation

Ref

Light source

# of
channels

Dimensions

Output light
intensity
(max. or used)

ÂľLED coupled
optic fiber

3×3

5mm long,
105Âľm in diameter,
550Âľm separation

1.28mW/mm2

-20.56dB –
-18.97dB

0

--

--

None

[140]

ÂľLED coupled
optic optrode

32

Base size: 300Âľm,
Tip size: 30Âľm

10mW/ mm2

-10 dB

32

Electrode size:
30Âľm

10~500kΊ

In vivo

[141]

Surface mounted
ÂľLED array

64×64

20Âľm in diameter
50Âľm separation

250mW/mm2

--

0

--

--

In vitro

[142]

3

200Âľm in diameter, 700Âľm
separation

10mW/mm2

--

16

200Âľm in
diameter
700Âľm
separation

1~5kΊ

In vivo

[143]

5×1

1.3 mm long
80Âľm wide
ÂľLED diameter: 40Âľm,
250Âľm separation

600mW/mm2

2%

0

--

--

None

[144]

Penetrating probe
with ÂľLED

1

12mm long
900Âľm wide.
µLED: 1×0.6×0.2mm3

0.7mW/mm2

--

3

50µm×100µm

--

In vivo

[145]

Penetrating probe
with ÂľLED

1

4.2mm×0.86mm×0.28mm
µLED: 0.55×0.29×0.1mm3

1mW/mm2

--

0

--

--

In vivo

[76]

Penetrating probe
with ÂľLED

1

~400Âľm wide, 20Âľm thick
µLED: 50×50×6.45µm3

7mW/mm217.7mW/mm2

--

1

20µm×20µm

1MΊ

In vivo

[26]

Penetrating probe
with ÂľLED

2

8×6.8×0.25mm2
µLED: 0.55×0.29×0.1mm3

1.5mW/mm2

--

4

120Âľm in
diameter

20~90kΊ

In vivo

[146]

Surface mounted
ÂľLED array

Penetrating probe
with ÂľLED

Light
delivery
efficiency

# of
channels

Dimensions

36

Figure 2-5 A spider schematic compares several main specifications of the laser- and LED-based
microdevices surveyed in this paper, in terms of size, density, multiple functions, wireless
capability, maximum light delivery efficiency, and light delivery efficiency. The performance is
rated on a scale of 1 to 5, with 5 being the best.
include potential thermally-induced tissue damage due to LED heat deposition in the brain as well
as electronic failure, particularly for high-density neural implants where microelectronics are in
direct contact with large-area brain tissues. Furthermore, the very limited adaptation of this
exciting method by other researchers in this field could be mainly due to the labor-intensive and
costly process of device fabrication
As Optogenetic studies have rapidly expanded into in vivo applications using freely

37

moving animals, there is an increasing demand for the development of untethered optical
tools for wirelessly controlled optical stimulation and neural readouts of large-scale brain
circuitry. Two major types of optical microdevices based on laser and LED light sources
have been surveyed. An overview of the advantages and disadvantages of both major types
is summarized using a spider web diagram as shown in Figure 2-4 from the perspectives of
size, functionality, maximum light delivery capability, light delivery efficiency, density and
wireless capability. The performance is rated on a scale of 1 to 5, with 5 being the best.

38

Chapter 3.

BDD neurotransmitter detection sensor

3.1 Theory of Electrochemistry
3.1.1 Nernst equation
From a thermodynamic standpoint, the Nernst equation is derived from the standard changes
in Gibbs free energy in a chemical reaction. Considering a half-reaction at anode, where 𝝂𝒐 and
𝝂𝑹 are stoichiometric coefficients,
𝝂𝒐 𝑶 + 𝒏𝒆− ⟶ 𝝂𝑹 𝑹

(3-1)

and another half-reaction at the cathode,
𝑯𝟐 ⟶ 𝟐𝑯+ + 𝟐𝒆−

(3-2)

𝝂𝑯𝟐 + 𝝂𝒐 𝑶 𝝅 ⟶ 𝝂𝑹 𝑹 + 𝝂𝑯+

(3-3)

the cell reaction is then,

The free energy is given by thermodynamic, considering 𝒂𝑯+ = 𝒂𝑯𝟐 = 𝟏, where 𝒂𝒊 is the activity
of the species i,
△ 𝑮 = △ 𝑮𝒐 + 𝑹𝑻 𝐥𝐧

𝒂𝑹 𝝊𝑹
𝒂𝟎 𝝊𝟎

(3-4)

Since △ 𝑮 = −𝒏𝑭𝑬 and △ 𝑮𝒐 = −𝒏𝑭𝑬𝒐 , where 𝑬𝒐 is the cell potential at standard conditions.
The potential E (also called half-cell potential) at the anode or cathode electrode can be calculated,
𝑬 = 𝑬𝒐 +

𝑹𝑻
𝒏𝑭

𝐥𝐧

𝒂𝟎 𝝊𝟎
𝒂𝑹 𝝊𝑹

(3-5)

which is called Nernst equation. The cell potential can be calculated using
𝑬𝒄𝒆𝒍𝒍 = 𝑬𝒄𝒂𝒕𝒉𝒐𝒅𝒆 − 𝑬𝒂𝒏𝒐𝒅𝒆

(3-6)

However, the 𝑬𝒐 of above-mentioned Nernst equation is under the standard condition and it is
inconvenient to deal with activates under non-standard condition. A solution to this is to introduce
′

formal potential 𝑬𝒐𝒆 , where 𝜸𝒊 is the activity coefficient:
39

′

𝑬𝒐𝒆 = 𝑬𝒐𝒆 +

𝑹𝑻
𝒏𝑭

𝜸

𝐥𝐧 ( 𝒐 )

(3-7)

𝜸𝑹

Then the Nernst equation can be written as
𝟐.𝟑𝑹𝑻

′

𝑬𝒆 = 𝑬𝒐𝒆 +

𝒏𝑭

𝐥𝐨𝐠

𝑪𝒐

(3-8)

𝑪𝑹

3.1.2 Kinetics of electrode reactions
In the previous section, it was shown that at equilibrium, the potential of an inert electrode at
anode or cathode in the solution of Redox (O and R) and supporting electrolyte is determined by
Nernst equation. At the situation that away from the equilibrium potential, the kinetics of electron
transfer is determined by Butler-Volmer equation:
𝜶𝒏𝑭𝜼

(𝟏−𝜶)𝒏𝑭𝜼

𝑹𝑻

𝑹𝑻

𝒊𝒏𝒆𝒕 = 𝒏𝑭𝑨𝒌𝒐 (𝑪𝒐 (𝟎, 𝐭)𝐞𝐱𝐩 (

) − 𝑪𝑹 (𝟎, 𝒕) 𝐞𝐱𝐩 (−

))

(3-9)

where 𝒊𝒏𝒆𝒕 is the net current of forward and backward reaction (A) , 𝒏 is the number of electrons
involved in the electrode reaction, 𝑭 is Faraday’s constant (C mole e-1), 𝑹 is universal gas
constant(J mol -1 k -1), 𝑻 is absolute temperature (k) , 𝜶 is the transfer coefficient (diemsionless),
𝑪𝒐 (𝟎, 𝐭) is the concentration at distance x=0 from the surface at time t (mol/cm3), 𝜼 is the
overpotential (v) and 𝒌𝒐 is the standard rate constant (cm/s). The standard rate constant is a
measure of the kinetic facility of a redox couple. A system with large rate constant will achieve
equilibrium faster than a system with smaller rate constant.

3.1.3 Mass transport
Mass transport phenomenon describes the supply of reactant and the removal of the product
from the electrode. There are three forms of mass transport: Diffusion, Migration, and Convection.
Diffusion describes the movement of a chemical species due to the concentration gradient.
Diffusion will move the species from high concentration regions to low concentration regions.
Migration is the movement of charged species in the presence of an electric gradient. Migration is

40

an electrostatic phenomenon and not the focus of this thesis. In the experiment, high concentration
supporting electrolytes will be used to avoid the influence of the migration in an electrochemical
cell. Convection refers the movement of a chemical species due to external mechanical forces.
Convection may be introduced by stirring the solution or rotating disk electrode. The mass
transport is determined by Nernst-Planck Equation:
𝝏𝑪𝒙 (𝒙)

𝑱𝒙 = −𝑫𝒙 (

𝝏𝒙

)−

𝒛𝒙 𝑭
𝑹𝑻

𝝏𝝓(𝒙)

𝑫𝒙 𝑪𝒙 (

𝝏𝒙

) + 𝑪𝒙 𝝊(𝒙)

(3-10)

where 𝑱 is the flux of the spices (mol cm-2 s-1), 𝑫 is the diffusion coefficient of the solution species
(cm2/s), 𝑪 is the concentration of the species (mol/cm3), 𝝓 is the electrostatic potential (v) and
𝝊(𝒙) is the fluid velocity, cm s-1, 𝒛 is the charge of species, F is Faraday’s constant (C mole e-1),
R is gas constant value (J mol

-1

k -1), T is temperature (k). The three terms of Nernst-Planck

Equation describes diffusion, migration and convection, respectively.

3.2 Techniques for study of electrode reactions
In this section, basic electrochemical techniques will be introduced for studying electrode
reactions. The mass transfer process of the systems considered in this section will occur only by
diffusion. An excess of the inert electrolyte is added so migration does not contribute to the mass
transfer process. No rotating disk electrode or stirring is applied to prevent any convection. The
techniques can be categorized into controlled potential technique (potential step) and potential
sweep technique. Both controlled potential method and potential sweep method share a same block
diagram of measurement as shown in Figure 3-1. A potential step is applied for a controlled
potential technique between a working electrode (WE) and a reference electrode (RE) and
measures the current flow from working electrode to counter electrode (CE). The recorded current
is a combination of Faraday current and double layer charging current.

41

Figure 3-1 Simplified block diagram for potentiometric measurements.

3.2.1 Controlled potential methods (Potential step)
A demonstration of applied potential and recorded current is shown in Figure 3-2. Potential
step experiments often use to obtain quantitative data when a redox system is understood
qualitatively by a potential sweep technology such as cyclic voltammetry(CV). This kind of
experiment is also called chronoamperometry since the current is recorded as a function of time.
The recorded Faraday current follows Cottrell equation:
𝒊𝑭 (𝒕) =

𝟏/𝟐

𝒏𝑭𝑨𝑫𝒐 𝑪∗𝒐
𝝅𝟏/𝟐 𝒕𝟏/𝟐

(3-11)

, where𝒊𝑭 (𝒕) is measured Faraday current (A), 𝒏 is the number of electrons involved in the
electrode reaction, 𝑭 is Faraday’s constant (C mole e-1), 𝑨 is the area of the planar electrode (cm2),
𝑪∗𝒐 is the bulk concentration of the analyte (mol/cm3), 𝑫𝒐 is the diffusion coefficient of the species
(cm2/s), t is the time (s). The double-layer charging current caused by double-layer capacitance is
determined as below:

42

𝒊𝒄 (𝒕) =

𝑬
𝑹𝒔

𝒆−𝒕/𝑹𝒔 𝒄𝒅

(3-12)

, where 𝑬 is applied potential (V), 𝑹𝒔 is the solution resistance (Ω) and 𝒄𝒅 is the double layer
capacitance (F). If the electrode is a microelectrode arrays, where the diffusion layer thickness for
each array electrode is less the spacing between each electrode, the Cottrell equation is written as:
𝟏

𝒊𝑭 (𝒕) = 𝒏𝑭𝑨𝑫𝑪∗ [

𝟏

(𝝅𝑫𝒕)𝟐

+ 𝟏/𝒓𝒐 ]

(3-13)

, where 𝒓𝒐 is the radical distance from the electrode center. Therefore, 𝐥𝐢𝐦 𝒊𝑭 (𝒕) = 𝒏𝑭𝑨𝑫𝑪/𝒓𝒐 ∗ ,
𝒕⟶∞

the recorded current reaches a steady-state condition.

Figure 3-2 A demonstration of applied potential (a) and recorded current (b) for potential step
technique. (Figures are adopted from www.studyblue.com)

43

3.2.2 Potential sweep methods
The complete electrochemical behavior of a Redox system can be studied by applying a series
of potential steps and recording the corresponding current-time curves. One of such methods is
described is called cyclic voltammetry (CV), as shown in Figure 3-2, where a saw-tooth wave is
applied between WE and RE of a chemical sensor and the current-time curve is recorded between
WE and CE. The voltammogram has one oxidation peak potential Epa and one reduction peak
potential Epc with corresponding oxidation peak current (anodic current) ipa and reduction peak
current (cathodic current) ipc., respectively. The recorded current contains both background current
(double-layer charging current) and Faraday current. For a reversible system (kinetic fast system),
the peak current contributed by Redox reaction under semi-infinite linear diffusion condition can
be determined by
𝟑

𝟏

𝟏

𝒊𝒑 = (𝟐. 𝟔𝟗 × 𝟏𝟎𝟓 )𝒏𝟐 𝑨𝑫𝒐 𝟐 𝑪𝒐 ∗ 𝝊𝟐

(3-14)

, where 𝒊𝒑 is the peak Faraday current (A), 𝒏 is the number of electrons involved in the reaction,
𝑨 is the area of the planar electrode (cm2), 𝑪∗𝒐 is the bulk concentration of the analyte
(mol/cm3), 𝑫𝒐 is the diffusion coefficient of the species (cm2/s), 𝝊 is scan rate (V/s). The peak
double layer charging current can be determined by
|𝒊𝒄 | = 𝑨 𝒄𝒅 𝝊

(3-15)

, where 𝒄𝒅 is the double layer capacitance (F). The difference between peak potential Ep and half
peak potential Ep/2, where the peak current is reduced by half, can be used to determine whether a
Redox system is reversible or not. For a kinetic fast system (reversible system),
|𝑬𝒑 − 𝑬𝒑/𝟐 | = 𝟐. 𝟐𝟎 𝑹𝑻/𝒏𝑭
, which equates 56.5/n mV at 25°C.

44

(3-16)

Figure 3-3 A demonstration of cyclic voltammetry: (a) Input potential between WE and RE (b)
Recorded current-time plot. (Figure 3-2 (b) is adopted from http://urrjaa.blogspot.com/)

3.3 Polycrystalline diamond deposition and characteristics
Diamond and graphite are both allotrope of carbon. At room temperature and pressure,
graphite is the stable allotrope and it is more stable than diamond, which is the metastable allotrope.
Therefore, nature diamond occurs at depths of around 200km down to the earth, where the pressure
and temperature are in the range of 70-80 kbar and 1400-1600°C, respectively[164], [165]. The
first industrialized diamond synthesis methodology was high pressure, high temperature (HPHT)
method proposed by Bundy in 1955 [166]. Later, the chemical vapor deposition (CVD) technology
was introduced into diamond synthesis, which enables the processes to happen at low pressures.
In this case, diamond growth is driven by kinetics in a high growth temperature rather than

45

Figure 3-4 A diagram of diamond CVD reactor: (a) hot filaments CVD reactor (b) Microwave
plasma CVD. (Figure 3-3 (a) is adopted from [169], Figure 3-3 (b) is adopted from [170].
thermodynamics[59]. The industrialized processes of CVD diamond were reported in the
1980s[167]. The CVD growths rely on a very small fraction of carbon (typical <5%) and relatively
high concentrations of dissociated hydrogen (H) with a growth temperature greater than 700°C to
provide surface kinetics[168]. During CVD, a plasma can be created by two major methods: hot
filaments (HF)[169] or microwaves (MW)[170]. Reactors based on both methods are shown in
Figure 3-4 (a) and (b). MW-CVD can generate higher plasma temperatures than HF reactors, which
results in the higher concentration of H atoms and can achieve higher phase purity and faster film
growth. The plasma temperature generated by HF-CVD is limited by the melting point of the
filaments and typically lower than the MW-CVD, but can grow over the larger surface area[168].
The gaseous form carbon source such as methane and dopant, such as B2H6 is fed into the CVD
chamber with hydrogen (H2) atmosphere. The presence of H2 is essential to react with sp and sp2
carbon on the surface and convert them to sp3 carbon[171]. Diamond grown by CVD can be
divided into two classes — polycrystalline and single crystalline diamond. Polycrystalline
diamond is composed by tightly-bonded small single crystals (grains). The grain size of
46

Figure 3-5 (a) schematic (b) scanning electron microscopy (SEM) image of the cross section of
diamond film of MC BDD. SEM imagines of (c) diamond nucleation side and (d) diamond growth
side. (a) and (b) are adopted and reprinted from [168]. (c) and (d) are taken using the films and
devices reported in Chapter 6.
polycrystalline diamond is determined by the process parameters such as temperature, pressure,
hydrogen to carbon and boron ratios and additional argon gas etc. [172] and can be categorized
into ultra-nanocrystal line (UNC) < nanocrystalline < microcrystalline based on grain size. For
fixed growth conditions, as grown time increases, a high density of nucleation sites first formed
on the substrate and then the sites with favored facets and orientations grow preferentially to those
with less favored ones, which results in an increasing size of crystalline facets as demonstrated in
47

Figure 3-5 (a) and (b)[168]. This will further lead to a smoother surface on the diamond nucleation
side than the diamond growth side as demonstrated in Figure 3-5 (c) and (d). The SEM imaging
of diamond nucleation and growth side are taken from the boron-doped polycrystalline diamond
(BDD) thin film described in Chapter 6. Non-diamond carbon (NDC) can be found predominately
at grain boundaries. UNC and NC BDD have more NDC than MC BDD due to more dense grain
boundaries[173]. The impurity caused by NDC of polycrystalline BDD film can cause the
electrode less electrocatalytically inert. Electrochemically, this can lead to a reduced potential
windows, increased background current and a surface more susceptible to fouling. Besides, in the
electron transfer process, a higher impurity of NDC results in slower electron transfer kinetics for
certain inner sphere redox couples. A comparative study of the potential window and background
current among diamond nucleation side, diamond growth side and a commercially available gold
electrode in 1M KCL solution with Pt as CE and Ag/Agcl as a reference at a scan rate of 0.1V/s is
shown in Figure 3-6. Those data are taken from the BDD film reported in Chapter 6. The
voltammograms are taken before and after the substrate transfer, which gives both diamond growth
side and nucleation side.
The easiest way to measure the impurity of diamond (primarily sp2 carbon) is to use Raman
spectroscopy. A pure sp3 diamond has only σ bonds, which results in a characteristic peak at 1332
cm-1. The line width of the peak is a qualitative measure of defects in the diamond film. The wider
the line width, the more defects. NDC, primarily sp2 carbon, which contains π bonds gives another
two peaks at 1355 cm-1 (D peak) and 1575 cm-1(G peak)[174]. The ratio of the relative intensity
of the sp3 peak to sp2 G peak is used to assess the quality of the diamond film in terms of nondiamond carbon impurity. Since π bonds are more polarizable than σ bonds, the excitation
wavelength of Raman spectroscopy is very important to accentuate the features of interest. Popular

48

laser wavelengths for Raman spectroscopy are 514.5nm green Ar ion laser, 632 nm HeNe laser,
and 785 nm IR diode laser. The longer wavelength of the excitation laser, the more sensitive to sp2
carbon[168]. Figure 3-7 shows Raman spectroscopy of heavy boron doped polycrystalline
diamond with different B/C ratio in the gas phase[175]. For boron-doped diamond, the ‘1332 cm1

’ peak decreases and down – shifts as B/C ratio (Boron to Carbon ratio) increases. Two new peaks

at 500 cm-1 and 1220 cm-1 appear. The relative intensities of both new peaks increase as boron

Figure 3-6 Voltammograms of Au, BDD growth side and nucleation side in 1M KCL solution
(WE: BDD / Au, CE: Pt, RE: Ag/Agcl), Scan rate: 0.1V/s. This experiment is done using the films
and devices reported in Chapter 6.
49

Figure 3-7 Raman spectroscope of heavily boron doped polycrystalline diamond with different
B/C ration in the gas phase. Reprinted from [175]
content increases in the gas phase. Two new broad peaks at 1350 cm-1 and 1580 cm-1 appear for
very high B/C ratio (14000ppm) ,which originates from micro crystalline graphite[176]. Figure 38 shows the Raman spectroscopy of both diamond nucleation side and growth side. The SEM
image of the films used in this measurement has been plotted in Figure 3-5 (c) and (d). The detail
50

fabrication process of the devices will be shown in Chapter 6. From the plot, we can see that the
‘1332 cm-1’peak shifts to 1305 cm-1 and two new peaks appear around 500 cm-1 and 1220 cm-1 for
boron contents. The intensity of both new peaks of diamond nucleation side is lower than that of
the diamond growth side, which indicates that diamond growth side has a higher boron doping
concentration. Besides, a new broad peak for diamond nucleation side appears at around 1470 cm1

, which indicates impurity of graphite (sp2 carbon).

Figure 3-8 Raman spectroscopy of BDD nucleation side and growth side. This experiment is done
using the films and devices reported in Chapter 6.

3.4 BDD devices for chemical sensing
Unlike the development of ÂľLED light delivery strategies, where extensive efforts
having been made, the engineering tools for in vivo / in vitro neurotransmitter detection is
still under developed, especially for BDD electrochemical sensor. Most of the BDD
neurotransmitter sensors fall into the category of BDD coated metal thin wires and BDD
microelectrodes. Most of BDD coated metal thin wires share a similar topology as shown
51

in Figure 3-9(a) [177]. The diamond coated metal wire (Pt in this case) was attached to a
longer copper wire using conductive Ag epoxy and inserted into a polypropylene pipet tip
for insulation. Then the tip was carefully heated and softened, which allows polypropylene
conformally flows over the BDD. Although those devices are convenient to fabricate, the
capabilities of multi-site recording and integration with other functionalities are limited.
BDD microelectrodes utilize the emerging microfabrication technology and offer
several advantages over BDD coated thin wires such as high density, high spatial
resolutions, and easy mass production. However, most diamond growth substrates, such as
silicon (with Young’s modules ~1011Pa[178]), are very rigid and do not match up with soft
tissue (with Young’s modules ~103~105 pa[179],[180]). Considering that diamond cannot
grow on flexible polymer substrate due to high synthesis temperature, the flexibility of
BDD microelectrodes is the barrier of applying such devices into implanted devices. To
overcome this barrier, possibilities of making flexible BDD microelectrodes have been
explored such as etching / thinning down diamond growth substrate and transferring BDD
onto a foreign flexible substrate.
A thinned down diamond neural probe was developed by Chan et al.[181] as shown in
Figure 3-9 (b). In this process, the undoped diamond was grown on silicon dioxide of a
silicon wafer, on top of which, doped diamond (BDD) was grown and patterned as
microelectrodes. The entire probe was released from silicon by etching down silicon wafer
in hydrofluoric-nitric acid. In this case, BDD is sandwiched between top silicon dioxide
passivation and bottom undoped diamond. While flexibility was shown due to the diamond
substrate being extremely thin (3Âľm), a thicker substrate with greater lateral dimension is

52

Figure 3-9 (a) BDD coated metal thin wires with Polypropylene pipet tip as insulation layer.
Reprinted from [177]. (b) A thinned down BDD neural probe with thickness of 3Âľm. Reprinted
from [181]. (c) A diamond polymer interface fabricated by transfer diamond from growth rigid
substrate onto spin-cast epoxy (only pads are made of diamond). Reprinted from [182]. (d) A BDD
microelectrode array fabricated using wafer transfer technology. Reprinted from [183].
desired to have stronger mechanical strength and application versatility. In viewing such
drawback, a wafer transfer process was proposed by Hess et al.[182] and also reported by
Bergonzo et al.[183], as shown in Figure 3-9 (c) and (d), respectively. In this process, silicon
diode was used as a sacrificial layer and spin-cast epoxy, such as polyimide, was used as a
flexible substrate. In detail, BDD was grown on silicon dioxide of a silicon wafer followed
by patterning as microelectrodes and contact pads. Then polyimide was spun on and cured.
The BDD devices were released from the silicon wafer by etching away silicon dioxide
layer in hydrofluoric acid (HF). The thickness of these devices is determined mainly by the
53

polyimide, which can be controlled by spin speed. However, such process can only transfer
small features such as electrodes and contact pads due to insufficient bonding strength
between BDD and polyimide. Besides, it requires multiple steps of HF etching since the
etching rate of silicon dioxide under larger patterns are faster than smaller patterns. In
addition, the bonding strength between diamond and polymer is unknown. To the best of
my knowledge, there are no published results on transferring wafer/large scale all diamond
patterns from rigid diamond growth substrate on to a flexible substrate, which will be the
focus of my work presented in Chapter 6.

54

Chapter 4. An implantable, miniaturized SU-8 optical
probe for Optogenetics-based deep brain stimulation
4.1 Motivation
Based on the review on the light delivery methods of Optogenetics, this chapter proposed a
method of making optical probes for Optogenetics-based deep brain optical stimulation using SU8. SU-8 is a negative photoresist, which has been used in many biomedical applications. SU-8 has
~95% optical transmittance around 460nm wavelength, which is the maximum activation
wavelength of the opsin used in this work. Therefore, by ÂľLEDs at the tip of an SU-8 probe and
directly inserting the light source into deep brain regions, attenuation caused by light transmission
n wave-guided structures such as optical fibers or optrodes can be minimized. Besides, compared
to silicon neural probes, SU-8 has a smaller Young’ s modulus[185], [186], which makes it more
flexible and thus causes less damage to the brain tissue. In addition, Additional Parylene-C

Figure 4-1 Conceptual diagram of the optical probe: (a) an overall view (b) a cross-section view.

55

encapsulation can potentially improve the long-term biocompatibility and reliability of the device
for chronic implantation. In this chapter, design considerations of the SU-8 optical probe are
described first, followed by the fabrication process. The optical, electrical properties of the asfabricated device are presented. Finally, the functionality has been proven in vivo by clearly lightinduced neural activity.

4.2 Optical probe device design fabrication
4.2.1 Device design
In the design, SU-8 2000 negative photoresist is used, which is an improved formulation of
SU-8. It can achieve over 200 Âľm thick coating with a single spinning process. Besides, it is
capable of making high aspect ratio structure and has very high optical transmission over 360 nm,
which makes SU-8 2000 a perfect candidate for this design purpose [186]. As shown in the concept
diagram (Figure 4-1), two layers of SU-8 photoresist were used in the fabrication, in which the top
layer was 10% wider than the bottom one to encapsulate ÂľLED chip. Several SU-8 anchors along
the central line of the shank were designed to improve the bonding strength between the two SU8 layers. The ÂľLED chip end of the SU-8 shank is 10% smaller than the contact end to make it
easier for penetrating into the brain. The overall device is encapsulated with Parylene coating to
further protect the device from exposure to body fluid. The Parylene package also reduces the risk
Table 4-1 Dimensions of the SU-8 probe and ÂľLED chips (L: length, W: width, H: height)
Probe (L×W)

4.2mm×0.86mm

Top SU-8 layer (W×H)

0.49mm×0.28mm

Bottom SU-8 layer (W×H)

0.42mm×0.12mm

Anchor diameter

0.03mm

Anchor interval

0.4mm

µLED (L×W×H)

0.550mm × 0.290mm × 0.100mm
56

of wiring being broken due to the dangling and stretching during device implantation. The
dimensions of the probe and the ÂľLED chip are listed in Table 4-1.

4.2.2 Fabrication process
The fabrication process is shown in Figure 4-2. In detail, a 3 inch silicon carrier wafer was
cleaned by isopropanol, acetone and deionized water. (a) 5Âľm Parylene C was deposited (PDS
2010, Specialty Coating System, Inc), followed by spinning the first layer of ~120 Âľm SU-8 2075.
(b) The first layer of SU-8 was patterned using standard lithography. (c) A 0.5 Âľm layer of Cu was
deposited on the first SU-8 layer using a thermal evaporator (Auto 306, Edward, Inc). (d) The Cu
layer was patterned and wet etched to form a connection between pads and leads. While copper
was used in the prototype as a proof-of-concept, devices for chronic implantation will use a noble
metal such as gold and platinum. During the copper patterning step, because the optical probe
stuck out of the wafer, thick photoresist residue was observed at the edge of the probe after
photolithography. As a result, a short circuit was formed between the anode and cathode of ÂľLED,
as shown in Figure 4-3(a). This is solved by increasing photoresist spinning speed by 50% and
reducing 20% exposure time, which can significantly reduce the residue without undercutting too
much copper. (e) S1813 photoresist was spun on top of the copper leads and exposed contact areas
with ÂľLED chips. The surface was then treated with O2 plasma (PX-250 plasma system, Nordson,
Inc) with 500 mTorr process pressure and 100 W power to remove photoresist residue on top of
the anode and cathode openings. Then low melting point (LMP) solder (melting point at ~62 ÂşC,
144 ALLOY Field’s Metal, Rotometals, Inc) was applied to the interconnects of the copper leads.
(f) The wafer was immersed in the acid solvent and suspended ÂľLED chips (Samsung, Inc) were
self-

57

Figure 4-3 Fabrication process

Figure 4-2: (a) A optical probe with thick photoresist residue at the edge. (b) A fabricated optical
probe with a ÂľLED mounted on the tip.
assembled onto the corresponding anode and cathode contacts. (g) S1813 was removed using
acetone followed by IPA and DI water rinse. (h) A ~280 Âľm layer of SU-8 was spun and patterned
to form the probe shape with a slanted tip for easy penetration of rat’s brain without removing the
dura. A dual spin coating process was used in order to achieve a better uniformity of the SU-8

58

coating and prevent wrinkles appearing by thermal stress during the post baking of SU-8. This was
done by spinning the first layer of ~120 Âľm SU-8 and then another layer of~ 160 Âľm SU-8. (i) The
LMP solder was applied onto pads in the solvent bath. Copper wires were bond to pads by melting
LMP. Epoxy was applied to strengthen the bonding between pads and copper wires. Then SU-8
optical probes were released from the silicon wafer. Then, the whole device was encapsulated with
~10 Âľm conformal Parylene C coating. A fabricated optical probe with a ÂľLED mounted on the
tip is shown in Figure 4-3 (b).

4.3 Results and discussions
4.3.1 Electrical properties
The current-voltage curve of the ÂľLED was characterized using a 4145B semiconductor
parameter analyzer, (Hewlett Packard, Inc), as plotted in Figure 4-4. The forward threshold of this
ÂľLED is around 2.6V. The typical voltages applied to ÂľLED and their corresponding input powers
re-listed in Table 4-2.

Figure 4-4 I-V curve of the ÂľLEDs

59

Table 4-2 Typical parameters using for optical stimulation

Applied voltage

Input current

Power consumption

3.0 V

11.2mA

33.6mW

3.2 V

22.6mA

72.4mW

3.4 V

38.2mA

130mW

4.3.2 Optical properties
The spectrum of light intensity coming through a ~280 Âľm thick top layer of SU-8 is shown
in Figure 4-5 (a). The peak intensity was located between 450 nm and 460 nm, which is within the
maximum activation wavelength region [187]. To study the light attenuation due to the SU-8
coating, the out-coupling light intensities of probes was measured with a 280 Âľm SU-8 coating
and without SU-8 coating, respectively. The data was recorded from a digital power meter (Model
815 Series, Newport, Inc) through an RHA 2000 evaluation board (Intan Technologies, Inc). As
plotted in Figure 4-5 (b), the light intensities of the probes with the SU-8 coating are 23 % and 7.1
% higher than those without SU-8, when the applied voltages are 3.2 V and 3.4 V, respectively.
This could be mainly attributed to the curvature of the top SU-8 layer formed due to the thermallyinduced polymer reflow, which effectively collimates the light of the embedded ÂľLED chip. The
SEM image, (HitachiS-4700, Hitachi, Inc) and profilometry, (NanoMap-500LS, Nanomap, Inc) of
the probe surface were taken as shown in Figure 4-5(c), where the clear curvature is shown.
However, it is also observed that, when the applied voltage increased to 3.6 V, the intensity of SU8 coated probe was slightly lower (7.6 %) than that of the uncoated probe. This inconsistency may
be due to the overheating effect of the ÂľLED at a high input voltage, which deteriorates SU-8 and
results in the reduced optical transparency. Besides, the high noise level of measurement
instrument might contribute to this as well. The variance of the data recorded from evaluation

60

Figure 4-5 Electrical and optical characteristics of SU-8 probe. (a) Spectrum of light emitting from
ÂľLED chip with a ~280Âľm SU-8 coating. (b) A comparison of light intensity of optical probe
coated with and without SU-8. (c)SEM image and profilometry of the SU-8 probe (d) Light
scattering property measured in gelatin.
board is quite noisy as shown from the error bars. The scattering property of ÂľLED light was
studied in gelatin to mimic the brain environment, where the ÂľLED was driven with 2.74 V input
voltage and 7 mW total power consumption as shown in Figure 3(d). The light propagation profile
was plotted in MatlabÂŽ (Figure 4-5 (d)) where blue spectra (450 ~ 495 nm) were extracted. The 90
% of the light intensity was within an elliptical-shaped scattering boundary (~ 10 mm in width and
~5mm in length), which enables optical excitation of a large population of neurons.

4.3.3 In-vivo LFP signal recordings
In vivo local field potential (LFP) upon optical stimulation was recorded in the unilateral
visual cortex of an anesthetized, channelrhodopsin-2 (ChR2) expressed rat as shown in Figure 46 (a). The rat was anesthetized with a mixture of ketamine and xylazine according to its weight.
After the rat was fully down, it was placed on a towel with heating pads underneath, which prevents
61

Figure 4-6 Demonstration of the efficacy of the deep brain optical stimulation using the fabricated
probe in the rat’s brain: (a) the experiment setup and (b) – (d) the recorded LFP with the µLED
input voltage of 3.0V, 3.2 V and 3.4V, respectively.

62

the rat from losing too much its body temperature during the surgery. Then the head of rat was
stabilized in a stereotaxic apparatus through ear bars. After shaving hairs, 70 % alcohol and
betadine were used to clean and disinfect the skin of the skull three times in turn. Then a rostralcaudal incision was made to expose the bony skull. Drilling was used to remove skullcap to prevent
bleeding and minimize the damage to the cortical area. A tungsten probe was used to record LFP
that was attached to a hydraulic microdrive to precisely control the penetration depth. The lightevoked signal was amplified and recorded using an RHA2000 amplifier and evaluation board
(Intan Technologies). Significant changes in neural activity were recorded when the ÂľLED was
driven by 3.2 Vpeak pulses with a 100 ms pulse width. A clear light-induced neural activity was
observed in time-domain LFP (1~100Hz) as shown in Figure 4-6(c), which demonstrates the
efficacy of deep brain stimulation using the fabricated SU-8 optical probe. As a comparison, LFP
oscillations excited with 3.0 Vpeak (Figure 4-6 (b)) and 3.4 Vpeak (Figure 4-6 (d)) pulses were
recorded. The results show that the average amplitude of LFP oscillations increased from 400ÂľV
to 750ÂľV when the input voltage of LED was changed from 3.0 Vpeak to 3.4 Vpeak. This is mainly
due to the increased light intensity, which can activate more neurons in a larger range.

4.4 Conclusion
This chapter reports a method of making SU-8 optical probes for Optogenetics-based deep
brain optical stimulation. The electrical and optical analyses show that the SU-8 layer not only
remains the central irradiation wavelength of the ÂľLED but also effectively enhance the light outcoupling intensity. Although the successfully demonstration of the efficacy of SU-8 probe, there
is one disadvantage of using SU-8 to drive ÂľLEDs due to the low thermal conductivity of SU-8.
The thermal property of the SU-8 probes will be discussed compared with diamond probes in the
next chapter.

63

Chapter 5. A hybrid neural interface Optrode with a
polycrystalline diamond heat spreader for Optogenetics
5.1 Motivation
In this section, to address the localized Joule heating issue, attempts was made by
mounting ÂľLEDs on a polycrystalline diamond (PCD) heat spreader, which can dissipate
electrically-induced heat rapidly and uniformly, and thus minimize tissue damage during
optical stimulation[146]. Compared to polymeric materials, such as SU-8 that has a thermal
conductivity of 0.3 Wm-1K-1, PCD has a better thermal conductivity up to 1800 Wm-1K-1
at 300K [73].
In the following sessions, device design and microfabrication method are first
introduced, where two different devices are presented, including single-shank and dualshank optrodes. In particular, the thermal properties of the single-shank PCD optrode are
analyzed both experimentally and computationally and compared with probes made of
other materials. Then, the optical and electrical properties of the optrodes are characterized.
Finally, the functionality of the dual-shank optrode is evaluated by recording light-induced
action potentials from primary visual cortex (V1) of a rat expressing channelrhodopsin-2
(ChR2).
A concept diagram of the heat spreading behavioral of SU-8 and PCD probes is shown
on the bottom left in Figure 5-1, where the outstanding thermal conductivity of PCD over
SU-8 inspired our proposed PCD-based optrodes on the bottom right. The optrodes are
designed to have either single or dual probe shanks (a dual probe shank optrode is shown

64

Figure 5-1 Concept diagram of the proposed neural interface optrode with a PCD heat spreader.
Table 5-1 Single/dual-shank probe dimensions
Single-shank probe

Dual-shank probe

Shank Dimension

0.9mm×5mm

0.9mm×5mm (each)

Base Dimension

2.38mm×6.5mm

2.38mm×6.5mm

# of Recording sites

3

4 (2 on each shank)

Recording site separation

1mm or 0.5mm

1mm

Shank separation

N.A.

1mm

# of stimulating sites

1

2 (1 on each shank)

Recording electrode
Dimension

90~100Âľm in diameter

90~100Âľm in diameter

amond DiamondDia

here). Each shank contains a blue ÂľLED chip on the tip for optical excitation of neural
activity and two microelectrodes for the recording of light-evoked signals. For the first

65

generation of the device, commercially available ÂľLED chips are utilized (Samsung, Inc),
with dimensions of 0.55mm (length) × 0.29mm (width) × 0.1mm (thickness). The
specifications of our design are listed in Table 1. The overall dimensions of the optrode are
determined mainly by the resolutions of the photolithography masks and laser cutting. The
probe is packaged with a Parylene C coating, which has been demonstrated to have the
longevity of protecting microelectronics for chronic implantation[188].

5.2 Methodology
5.2.1 Device fabrication
The fabrication process of PCD probes is shown in Figure 5-2, (a) PCD was grown on
a molybdenum substrate using a 2.45GHz microwave plasma assisted chemical vapor
deposition (MW-PACVD) reactor with 2-3kW microwave power in a methane and
hydrogen mixture atmosphere (4% CH4, 160-240Torr) and released by thermal stress
during cooling from growth temperature to room temperature. Then diamond was cut into
10mm by 10mm size chips to avoid large nucleation on the growing side. The side attached to
molybdenum (nucleation side) was used for fabrication. (b) The diamond substrate was
cleaned by sonication in isopropanol (IPA) and deionized (DI) water for 30min each and
then in nitric acid at 80ÂşC for 30min to create an oxygen-terminated diamond surface for
metal deposition. (c) A 3nm layer of Ti and a 0.5Âľm layer of Cu were deposited using a
thermal evaporator (Auto 306, Edward, Inc). Ti was used as an adhesion layer to improve
the bonding strength between Cu and PCD. A photoresist mask was patterned using a mask
aligner (ABM, Inc) for metal patterning. Then the Cu/Ti layer was wet etched using ferric
chloride (to remove Cu) and buffered oxide etchant (to remove Ti) to form microelectrodes,
contact pads, and interconnect wires. While Cu was used for prototyping, noble metal, such

66

Figure 5-2 Fabrication process for making the proposed PCD probe. (a) PCD growth on
molybdenum (Mo). (b) PCD release from Mo substrate. (c) Metal deposition and patterning. (d)(f) ÂľLEDs assembly. (g)-(h) Parylene C coating and patterning. (i) Probe shaping.
as gold or platinum, will be used for chronically implantable devices. (d) Photoresist
(S1813, Microchem) was spun on the substrate and selectively patterned to expose the
ÂľLED pad areas. Oxygen plasma (PX-250 plasma system, Nordson March, Inc) at a power
of 100W and pressure of 0.5Torr was applied for 5 min to remove photoresist residue on
the pad areas. (e) Low melting point (LMP) solder (62 ºC, 144 ALLOY Field’s Metal) was
applied in an acid bath[189] and the die-form LEDs were self-assembled onto the contact
pads wetted with LMP solder. (f) The substrate was rinsed with acetone, IPA, and then DI

67

water to remove the photoresist layer. (g) A ~5 Âľm Parylene C layer was deposited on top
of the probe as an encapsulation using a chemical vapor deposition (CVD) system (PDS
2010, Specialty Coating System, Inc). (h) Parylene C was then patterned using oxygen
plasma (RIE-1701, Nordson March, Inc) and a photoresist mask at the power of 300W and
pressure of 0.25Torr, in order to open the recording sites and contact pads for electrical
interconnects. (i) Finally, the probe was shaped using a Nd:YAG laser with a power of 814W (UltraShape 5xs, BettonvilleInc). To connect the probe to external powering and
recording electronics, the pad areas were covered with the LMP solder, and flexible thin
wires or Parylene ribbon cables were assembled to the probe. The Parylene ribbon cables
were made out of patterned metal sandwiched between two layers of Parylene C with LMP
solder applied at the contact pads and flipped-bonded onto the contact pads of the PCD
probe. Finally, epoxy was applied to the contacts to strengthen the bonding between the
pads and wires.

5.2.2 FEM Simulation of Device Thermal Properties
Finite element method (FEM) simulation was performed in COMSOLÂŽ Multiphysics
4.3 (COMSOLÂŽ, Inc.) with heat transfer in solids module. A time-domain solver was used
to solve equation (5-1), where ρ, Cp, k, and Q are density, heat capacity, thermal
conductivity, and heat source (or sink), respectively.
𝝆𝑪𝒑

𝝏𝑻
𝝏𝒕

− 𝛁 ∙ (𝒌𝛁𝑻) = 𝑸

(5-1)

A single-shank PCD optrode was simulated and compared to an SU-8 based optrode
fabricated using the method reported in [190]. The thicknesses of the PCD and SU-8 probe
were measured at 0.24-0.3mm and 0.22mm, respectively and were modeled as 0.27mm for

68

(a)

(b)

Figure 5-3 (a) Simulation model in COMSOLÂŽ. (b) Mesh generated for FEM simulation.
Table 5-2 Simulation Parameters in COMSOLÂŽ

mond

Density
(Kg/m3)

Heat Capacity
J/(kg·K)

Thermal Conductivity
W/(m·K)

Diamond [35], [126]

3000

0.512

1800

SU-8 [43], [127]

1236

1200

0.3

GaN [128], [129]

6100

500

130

Copper*

8700

385

400

*From material library in COMSOLÂŽ

the PCD probe and 0.22mm for the SU-8 probe in the simulation. The copper interconnects
were modeled as high conductive layers with a thickness of 500nm, as shown in Figure. 53(a). Two 4.5cm copper wires (~0.2mm in diameter) attached to the contact pads were
modeled as heat sinks of the system. From the simulation, the volumes of these 4.5cm long
copper wires were sufficient to dissipate the heat generated by the ÂľLED, given the fact
that increasing the length of copper wires will not change much of the maximum
temperature variation along the probe shank. The Parylene C coating (~5Âľm) was not
modeled because it was much thinner than the SU-8 and PCD (~200Âľm). The parameters
of the materials used in the simulation are listed in Table 2 [73], [191], [80], [192], [193],[194].

69

The geometry of the model was discretized using tetrahedral mesh with a maximum
element size of 1.73mm, minimum element size of 0.0743mm, and maximum element
growth rate of 1.35, as shown in Figure 5-3(b). The initial temperature was set at 24°C for
the SU-8 probe and 25°C for the PCD probe, which are the same as in the experiment. The
boundary of the model was set to thermal insulation as the probes were coated with Parylene
and epoxy at the base, which has relatively low thermal conductivity (Parylene C: ~0.084
(W/m·K) [195]). A heat source boundary of 18.6mW calculated from the experiment was
applied at the interface between ÂľLED and SU-8/PCD shanks with a100ms pulse. It is of
note that the above simulation does not include tissue, in order to compare with the
experimental studies where the device temperature was monitored in air to resemble the
worst scenario of the device application. In the brain tissue, the temperature variation of
the implanted device is expected to be lower than that measured in air, as the high thermal
capacity of brain tissue can counteract the temperature variation.

5.2.3 Device Characterization
Several experiments were conducted to fully characterize the thermal, electrical, and
optical properties of the optrode. In particular, the thermal images of both the PCD and SU8 devices were obtained in air using a high-resolution infrared camera (Delta Therm
HS1570 and DT v2.19 software) with a LED applied voltage of 3.4V at 1Hz and 100mS
pulse duration. The thermal images were processed using MATLABÂŽ (R2011a, The
MathWorks) and compared with the simulated results. The electrochemical impedances
were measured in saline (0.9% NaCl solution) at room temperature using an embedded
electrode-impedance-testing module in an Intan RHA 2000 evaluation board (Intan Tech.
LLC). The light intensity of the ÂľLED chip was measured using a digital power meter

70

(Model 815 Series, Newport, Inc) and read through the RHA 2000 evaluation board. The
light-scattering property of the optrode was also studied in a scattering media (20% gelatin).
The blue light was extracted and plotted in a three-dimensional (3-D) format using
MATLABÂŽ.

5.2.3 In-vivo Animal Experiments
In-vivo acute experiments were conducted in V1 of a rat (Long-Evans) to demonstrate
the functionality of the as-fabricated PCD probe. All procedures were approved by the
Institutional Animal Care and Use Committee (IACUC) at Michigan State University. Prior
to the in-vivo testing, the rat was transfected with ChR2 to enable the functionality of light
sensitivity upon blue light illumination. For viral transfection, the rat was anesthetized with
ketamine and xylazine and then placed in a stereotaxic apparatus with a heat pad underneath
to maintain its body temperature. A rostral-caudal incision was made in the skin to expose
the skull. Two holes, positioned rostral-caudal, were made through the skull overlying each
hemisphere of visual cortex. Each site was injected with 1.0ÂľL AAV-hSynhCHR2(H134R)-mCherry solution (1011-1012 vector genome (vg/ml), UNC Vector
Core), with an injection rate of 0.1ÂľL/min, using a micro syringe (5ÂľL, Model 75 RN SYR
and 100ÂľL, Neuros Adapter Kit, Hamilton, Inc). After each injection, the syringe was
maintained in place for an additional 5mins to allow the viral vector to diffuse away from
the injection site. Once the injections were completed, the cortical openings were plugged
with Bonewax and then covered with Gelfoam, after which the overlying skin was sutured.
Then the rat was given 5cc of sterile saline subcutaneously to prevent dehydration during
recovery, as well as an injection of buprenorphine for pain relief. Post-injection, and before
device implantation, the animal was housed in the animal care facility. A period of three

71

weeks was allowed for the animal to recovery fully from the injection surgery, and for
cortical expression of the ChR2 gene.
The functionality of the PCD optrode for simultaneous optical stimulation and
electrical recording was tested 3-4 weeks post-surgery. Following the same stereotaxic
procedure as used for the viral vector injection, the rat was anesthetized and the skull
opened over the region where the two injection sites. The PCD probe was inserted into V1
of the left hemisphere of the rat with the right hemisphere serving as a vector-injected
control. During the in-vivo experiments, the ÂľLED chips were driven by a continuous pulse
train with various input voltages for optical stimulation. The light-evoked neural activity
was recorded through the integrated microelectrodes (Channels 1-4) and amplified through
an RHD 2132 32-channels headstage and RHD 2000 evaluation system.

5.3 Results and discussions
5.3.1 Fabricated devices
Figure. 5-4(a) and (b) show the diamond growth side (rough side) and the nucleation
side (smooth side) of the PCD substrate used in our fabrication. The diamond growth side
of the PCD substrate has large diamond grain sizes up to 700Âľm and is not suitable for
microfabrication. The nucleation side attached to the molybdenum during diamond growth
has a relatively smoother surface and was used to construct microelectrodes, contacts, and
interconnect wires. To analyze the surface roughness of the nucleation side, a 3-D
quantitative atomic force microscope (AFM) image of a 5×5µm2 area was obtained (AFM,
Dimension 3100, Bruker Nano), as shown in Figure 5-4(c). The root-mean-square (RMS)
surface roughness was measured to be ~ 201.98nm. Figure 5-4(d) shows a fabricated PCD

72

Figure 5-4 Diamond growth side (rough side) of a PCD substrate. (b) Diamond nucleation (smooth
side) of the same PCD substrate. (c) An AFM imagine shows the surface morphology of the
nucleation side of the PCD substrate. (d) Fabricated PCD and SU-8 probe for heat dissipation
measurement. (e) A fabricated single-shank PCD probe with Parylene ribbon cable for
interconnection. (f) Fabricated single/dual-shank probes. Inset shows the ÂľLED, recording
electrode, and metal interconnects.
73

probe and an SU-8 probe used for heat distribution measurement. Both probes had the same
geometry, metal interconnects layout, and a similar average thickness (0.27mm for PCD
and 0.22mm for SU-8). The average thickness of the PCD chip was tapered from around
0.15 to 0.45mm and laser thinned down to ~0.27mm. Figure 5-4(e) shows a fabricated
single-shank probe with a Parylene ribbon cable attached to an interface with the recording
board. Figure 5-4 (f) shows fabricated single and dual-shank probes with a close-up view
of a ÂľLED, recording electrodes, and metal interconnects. Approximately 20% scaling
down of the width of the probe can be achieved in the future by cutting off the redundant
PCD substrate.

5.3.2 Optical and Electrical Properties
The electrochemical impedances of the electrodes at 1 KHz (Channels 1-4 in Figure
5-8) in saline were 89.0, 20.6, 38.5, and 54.8 Kohm, respectively. Channel 1 was found to
have high impedance due to incomplete Parylene removal and therefore was not used in
the in vivo studies. The variation of the impedances among different electrodes is caused
by the uneven surface morphology of the PCD, especially the one with large diamond
grains of 100Âľm. This affects the leverage of the substrate during photoresist (PR) spinning
and leads to uneven PR mask thickness and Parylene etching across the substrate. Hence,
the Parylene coating on the microelectrodes of the left side (Ch1 and Ch4) of the probe is
under-etched compared with Parylene on the right side (Ch2 and Ch3).The light spectrum
of the ÂľLED in Figure 5-5(a) shows that the peak intensity was located between 450 nm
and 460 nm, which is within the maximum activation wavelength region [196]. The light
intensity of the ÂľLED chip is plotted in Figure 5-5(b) (blue), as a function of the applied

74

Figure 5-5Figure 5-5 The electrical and optical properties of a fabricated PCD probe. (a) Light
spectrum of the ÂľLED. (b) Light intensity and power consumption of the ÂľLED. (c) Light
scattering property in gelatin. (d) Normalized blue light intensity (both amplitude and color
indicate the light intensity).
voltage. The light intensity with the applied voltage of 3.2V, 3.4V and 3.6V at 1Hz and
10% duty cycle corresponds to 0.79mW/mm2, 1.16mW/mm2, and 1.55mW/mm2,
respectively. The power consumption was calculated by multiplying the pulsating voltage
and current during the onset of the LED and then normalized to 1 sec. These three voltages
will be applied to the ÂľLEDs used later in acute animal studies, where the minimum light
75

intensity required for activating a ChR2 transfected ion channel is reported to be ~
1mW/mm2 [197]. To estimate the power consumption of the optrode, the instantaneous input
current of the ÂľLED was obtained by measuring the voltage drop across a 10ohm resistor
in series with the ÂľLED while being driven by 1Hz, 10% duty cycle pulses at different
voltages. The voltage across the ÂľLED was measured directly and the power consumption
was as plotted in Figure 5-5(b) (green). The corresponding normalized power consumptions
were 3.47mW, 6.38mW, and 26.82mW at the applied voltage of 3.2V, 3.4V, and 3.6V with
1Hz and 100ms duration, respectively. Figure 5-5(c) shows the light scattering profile of
the ÂľLED when the optrode was inserted into gelatin. In this study, the ÂľLED was driven
by square pulses with 3.4V amplitude, 1 Hz frequency, and 10% duty cycle. Images were
captured using a camera (EOS Rebel T2i, Canon) and analyzed using MATLABÂŽ to extract
blue light spectra (with a wavelength of 450-495nm), as shown in Figure 5-5 (d). An
effective area of optical stimulation, where the majority of the optical energy was delivered,
was estimated to be around ~ 2mm by 2.5mm

5.3.3 Thermal Properties
Figure 5-6(a)-(d) show the measured and simulated heat distribution of the SU-8 and
PCD optrodes, when the LED was powered by 3.4V, 1Hz, 100ms pulses. It can be seen that
the SU-8 probe accumulated heat at the tip due to the poor thermal conductivity of SU-8,
while the PCD probe dissipated heat throughout the entire shank in less than 0.5sec without
creating localized hot spots on the probe. Similar phenomena were observed in FEM
simulation. During the onset of LED stimulation, the maximum variations of the device
temperature relative to the ambient temperature were measured at 3.3°C and 2.5°C from

76

Figure 5-6 Heat distribution of the SU-8 and PCD probes with 3.4V, 1Hz, 100ms duration pulses:
(a)-(b) from the high-resolution infrared camera (c)-(d) from the FEM simulation in COMSOLÂŽ
(Enlarged tip areas are shown in dash squares.)
the experimental data and 1°C and 0.5°C from the FEM simulation for SU-8 and PCD
probe, respectively. Figure 5-7(a) and (b) show the measured cooling curves of the PCD
(solid line) and SU-8 (dash line) optrodes, after activating the ÂľLEDs for 60sec with
different input voltages. The maximum temperature variations of an SU-8 probe were
around 2.2ÂşC, 4.2ÂşC, and 10ÂşC for the applied voltages of 3.0V, 3.2V, and 3.4V,
respectively, whereas the PCD probe has maximum temperature variations of around 0.5ÂşC,
0.7ÂşC and 0.9 ÂşC at the same applied voltages. Furthermore, it can be seen that the SU-8

77

Figure 5-7 Cooling curves of the probes after activating ÂľLED for 60sec with different inputs of
(a) an SU-8 probe and (b) a PCD probe. (c) Instantaneous and steady state temperature variations
of the SU-8 probe and the PCD probe. (d) and (e) shows close-up view of instantaneous
temperature variations of the diamond probe in (c).
probe required more than 2 seconds in order to return to the baseline temperature. Hence,
sequential stimulation pulses come before the probe can be cooled down, which results in
significant heat accumulated on the probe, especially when the probe is driven by fast pulse
trains with high applied voltages. In contrast, the PCD probe has smaller temperature rises
and faster cooling rates than the SU-8 probe, thereby effectively reducing the localized heat
78

accumulation. This is demonstrated in the following study, where the LEDs were powered
by continuous pulse trains. We monitored device temperature versus time after the LED was
switched on, and defined the steady-state temperature as the temperature at t=1min when the
maximum temperature along the LED was no longer cranking up. Figure 5-7(c) shows the
instantaneous changes in the tip temperatures and the steady-state temperature variations of
the SU-8 and PCD optrodes for six repetitive heating on-off cycles. Close-up views of the
data for PCD were plotted in Figure 5-7 (d) and (e). The maximum temperature variations
of the SU-8 probe with input voltages of 3.0V, 3.2V, and 3.4V were 2.4ÂşC, 4.0ÂşC, and 5.5ÂşC
relative to environment temperature (~22.25 ÂşC) during the first duty cycle (100mS),
continued to increase to 3.7ÂşC, 6.8ÂşC, and 8.7ÂşC within the first 7sec, and then stabilized at 3.6ÂşC,
7.3ÂşC, and 11.8ÂşC. On the contrary, the temperature variation of the PCD probe with input voltages
of 3.0V, 3.2V, and 3.4V were 0.3ÂşC, 0.4ÂşC, and 0.5ÂşC for during the first duty cycle, continued
ramping up to 0.5ÂşC, 0.6ÂşC, and 0.9ÂşC within the first 7sec and then stabilized at 0.5ÂşC, 1.0ÂşC, and
1.3ÂşC. The maximum increase of transient temperature is around 1.0ÂşC for the PCD optrode, which
satisfies the safety requirement for implantable electronics[21]. The settled temperature variations
for PCD probe were lower than the initial variations in all three applied voltages. For the SU8
probe, the settled temperature variation with an applied voltage of 3.0V and 3.2V were similar to
the initial readings. However, the settled temperature variation with the applied voltage of 3.4V
was nearly 3.8ÂşC higher than the initial variation.

79

5.3.4 In-vivo Optical Neuromodulation and Recording
Figure 5-8 shows the experimental setup of the acute animal studies and the layout of
a dual-shank optrode. During the optical stimulation, the ÂľLED on the left side of the probe
was driven by repetitive pulses with a frequency of 1Hz, the pulse width of 10ms, and
amplitude of 2.9V, 3.2V, 3.4V, and 3.6V. Light-evoked neural responses were recorded
from three functioning electrodes (Channels 2-4) and post-processed using MATLABÂŽ.
Figure 5-9(a) shows the neural signals recorded from Channel 2 with different applied
voltages after applying a band-pass filter (60-3000Hz). Light-evoked action potentials were
observed when the optical stimuli were switched from on to off with the high input voltage
of 3.6 V, whereas no action potential was observed with 3.2 V and 3.4 V inputs. This is
because the input voltages of 3.2 V and 3.4 V of the ÂľLED may not be sufficient to evoke
any action potentials, considering the tissue scattering and adsorption of LED light. Figure
5-9 (b) shows the action potentials recorded from different channels with 40 trails aligned
and stacked when the applied voltage of LED was 3.6V, which demonstrated the
consistency of the recording. A higher-level photocurrent artifact signals were observed at
Ch4 because Ch4 is closer to the ÂľLED than Ch2 and Ch3.

Figure 5-8 (a) In-vivo testing setup.(b) Probe schematic shows the location of 4 recording channels.
80

Figure 5-9 (a) Recorded signals with the applied voltage of 3.2V, 3.4V, and 3.6V, respectively. (b)
Action potentials recorded from different channels with 40 trials stacking at the input voltage of
3.6 V.

5.4. Conclusion
In this chapter, a PCD-based, hybrid optoelectronic neural interfacing probe was designed,
fabricated, and characterized, which is capable of optical stimulation and simultaneous recording
of neural activity. The experimental and simulation results show that the application of the PCD
heat spreader can keep the temperature variations of the probe around 1.0ÂşC during optical
stimulation, which not only reduces the risk of thermally-induced tissue damage but also improves
the accuracy of Optogenetic experiments by minimizing biological interferences due to thermal
effects. In-vivo testing was performed in V1 of an Optogenetic rat, which demonstrated the
81

functionality of optical stimulating and electrical recording of action potentials of ChR2
transfected neurons. Light-evoked action potentials were observed when the LED was driven by
3.6V input, 10ms duration, 1Hz repetitive pulses. No action potentials were observed at 3.2V and
3.4V LED inputs. Studies verified that only 3.6V input can result in sufficient light intensity
exceeding the minimum requirement (1mW/mm2) for opening opsin ion channels.

82

Chapter 6. Large-scale, all polycrystalline diamond
structures transferred on flexible Parylene-C films for
electrochemical sensing
6.1 Motivation
Despite its many advantages of diamond as demonstrated in Chapter 5, diamond is a rigid
material with Young’s modules of ~1000 GPa[73], which is several orders of magnitude higher
than that of the brain tissues (~103 to 105 Pa[74]). The micromotion-induced strain between rigid
implants and surrounding soft tissues has been hypothesized to cause a harmful immune response
and even irreversible tissue damage[75]. Recently, mechanically flexible, polymer-based neural
implants have shown promises as the next generation of implanted devices[43],[76],[77],[78]. For
those devices, electrodes and interconnecting traces made of noble metals were constructed on soft
polymeric substrates with low Young’s moduli, such as polydimethylsiloxane (PDMS) (360-870
KPa[79]), polyimide (2.5 GPa[72]), SU-8 (SU-8 2000, 2.0 GPa[80]) and Parylene (2.8 GPa[81]).
Consequently, the overall effective Young’s modulus can be significantly reduced to minimize the
mechanical mismatches between rigid metal and soft tissues. Unfortunately, unlike noble metals,
BDD cannot be fabricated directly on a polymer substrate due to its high synthesis temperature
(500 – 900°C[82]) exceeding the glass transition temperature of polymers. To address this issue,
a wafer transfer process is required to transfer diamond patterns from diamond growth substrates,
such as silicon, onto flexible polymer substrates. Previously, Hess et al.[182] and Bergonzo et
al.[183] reported a method for making diamond-on-polymer electrodes. In their approaches, the
boron-doped polycrystalline diamond (BDD) was selectively grown and patterned on silicon
dioxide (SiO2) substrate, and then, transferred onto spin-casted polynorbornene and polyimide by
removing the SiO2 sacrificial layer in hydrofluoric acid (HF). Whereas such processes allow

83

transferring microscale BDD patterns such as electrodes, macroscale interconnects and contact
pads must be fabricated from metal separately. Additionally, a multi-step HF releasing process is
required since the release of polymers and large metal patterns are faster than the small BDD
electrodes[182].
In this chapter, an innovative method is introduced of wafer-scale transfer of all diamond
macro/micro patterns from a diamond growth silicon substrate onto a flexible Parylene-C
substrate[198],[199]. Parylene-C is a micromachinable, transparent, flexible, FDA-approved and
USP class VI biocompatible polymer, which has been widely used as an excellent structural and
packaging material in neural implants [200],[52]. As illustrated in Figure 6-1, this approach
involving three key steps: pre-transfer patterning, BDD-to-Parylene transfer, and post-transfer
fabrication. During the pre-transfer process (Figure 6-1-I), SiO2 was deposited onto a silicon wafer
as a sacrificial layer followed by BDD synthesis and patterning. After slight over-etching of SiO2
to create undercuts around the BDD patterns, Parylene-C was conformally coated to form
mechanical anchors that can improve the bonding strength between BDD and Parylene. A mesh
structure was employed on large patterns to introduce more anchors for further enhancement of
the bonding force. For BDD transfer (Figure 6-1-II), two methods for substrate removal were
explored and compared: potassium hydroxide (KOH) etching and deep reactive-ion etching
(DRIE). For KOH etching, a special wafer holder including a Teflon jig, KOH resistant O-ring,
and chromium-coated C-clamps was designed to expose only the backside of the silicon wafer
through the opening of the upper lid, while protecting the front side to prevent the delamination
between BDD and Parylene during etching. After BDD transfer, a post-transfer process (Figure 61-III) was employed where the free-standing BDD-polymer film was attached on a wafer followed

84

Figure 6-1. Illustration of the fabrication processes for I. pre-substrate transfer patterning; II.
Substrate transfer; and III. Post-substrate transfer processing.
by subsequent micromachining processes for electrical components assembly, interconnects and
wiring.

6.2 Methodology
Figure 6-1 shows a detailed process flow. In the pre-transfer step, (I-a) 1-Âľm-thick SiO2
was deposited on a 3-inch Si wafer using plasma enhanced chemical vapor deposition
(PECVD) (PlasmaLab 80plusÂŽ, Oxford Instruments). (I-b) Microcrystalline BDD film was
synthesized using a custom-designed 2.45GHz microwave plasma assisted chemical vapor
deposition reactor (MWPACVD) with a gas mixture of hydrogen-diborane and methane at
a temperature of 700°C. (I-c) Aluminum layer was sputtered as a hard mask for diamond
etching (Denton Desk Top Pro Sputtering System, Denton Vacuum, Inc.). (I-d) The

85

aluminum is patterned via photolithography (ABM, Inc.) and etched using an aluminum
etchant (Type A, TRANSENE, Inc.). Then BDD was plasma etched in an electron
cyclotron resonance reactive ion etcher (Lambda Technologies, Inc.) using SF6/Ar/O2 with
microwave power of 400 W and a radio-frequency (RF) bias of 100 W (150 V). Afterward,
the aluminum mask was stripped using the aluminum etchant. (I-e) SiO2 was over-etched
in buffered oxide etchant (BOE) to create undercuts for forming Parylene anchors. (I-f) The
wafer was treated with the Silane A174 adhesion promoter (Sigma Aldrich, Inc.), followed
by a conformal coating of ~15 Âľm Parylene-C (PDS 2010, Specialty Coating System, Inc.).
During the transfer step, Parylene coating on the backside of the Si wafer was first
removed in O2 plasma (PX-250 plasma system, Nordson March, Inc.). For wet etching of
silicon (II-a1), the wafer was attached to the jig and etched in the 35% KOH at 70 °C for
~9 hrs. For plasma dry etching of silicon (II-a2), the frontside of the wafer was bonded onto
a 4-inch carrier wafer using polyphenyl ether and the backside was completely etched with
repetitive SF6 and C4F8 cycles in a DRIE system (SPTS Pegasus 4, SPTS Technologies,
Ltd). (II-b) The BDD-Parylene layer was released from the carrier wafer by soaking the
wafer in acetone, followed by isopropanol alcohol (IPA) and deionized (DI) water rinses.
Afterward, (II-c) sacrificial SiO2 was etched completely in buffered oxide etchant (BOE).
During the post-transfer step, (III-a) the wafer-size BDD-Parylene film was bonded to a
carrier wafer using photoresist. (III-b) Ti/Cu was evaporated (Auto 306, Edward, Inc.) and
patterned (ABM, Inc.) to form contact pads onto the BDD film. (III-c) and (III-d) Low
melting point (LMP) solder (62 °C, 144 ALLOY Field’s Metal) was applied onto the
contact pads in an acidic solution. (III-e) Discrete active electronics, such as ÂľLEDs,
(Samsung, Inc) were self-assembled on the pads. (III-f) After releasing the device from the

86

Figure 6-2 Device design of ÂľLED probes and chemical sensors used to demonstrate the BDD
transfer process. The black and grey colors represent BDD and Parylene, respectively.
carrier wafer, flexible wires were soldered onto the pads using LMP as interconnects to
testing instruments. Epoxy was applied to strengthen the bonding between the wires and
pads. Finally, another 5 Âľm Parylene-C was deposited to encapsulate the device with the
electrodes being exposed for sensing.
As a proof-of-concept, two different BDD-polymer devices were designed as
illustrated in Figure 6-2. In particular, a ÂľLED probe was constructed in order to
demonstrate the post-transfer fabrication step and to test the mechanical robustness of
flexible BDD patterns during folding and bending. The ÂľLED probe consisted of two
electrodes of 60 Âľm in radius for both the anode and cathode. The wires were 2.4 mm long
with various widths of 200 Âľm and 50Âľm, resulting in different wire resistances (Figure 62(a)). Figure 6-2(b) shows the BDD electrochemical sensors designed in a three-electrode
configuration with: a reference electrode (RE), a working electrode (WE) and a counter
electrode (CE). The overall design dimensions of the sensor were 4×4.7 mm2. The
effectively exposed BDD areas of WE, CE and RE were around 0.8 mm2, 0.48 mm2, and
87

0.48 mm2, respectively, calculated by subtracting the areas covered with Parylene anchors.
To create more Parylene anchors, microholes with dimensions of 30× 50 µm2 and center to
center separation of 50 Âľm were uniformly distributed on large BDD patterns, such as
contact pads.

6.3 Results
6.3.1 Characteristics of the BDD films
A scanning electron microscope (SEM) image (6610V, JEOL Inc.) in Figure 6-3(a)
shows the surface morphology of the diamond nucleation side. The diamond nucleation
side, which was originally in contact with the SiO2/Si substrate, was exposed after the
transfer process and served as the sensing surface of electrochemical sensors. It shows a
smooth surface mainly due to the small crystal sizes of BDD. Figure 6-3(b) shows the
Raman spectrum (532 nm Laser, HORIBA Scientific Inc.) taken from the nucleation side
of BDD, where the characteristic diamond band at 1332 cm -1 (Sp3 carbon peak) decreases
and widens, two new boron bands appear at ~473 cm-1 and ~1209 cm-1 and Sp2 carbon peak
appears at 1470 cm-1. The Raman features are similar to heavily doped BDD thin films with
boron doping concentration estimated in the order of 1020 cm-3[175]. Table 6-1 summarized
the film properties including the average surface roughness of diamond nucleation side was
10.8 nm measured by an atomic force microscope (AFM) (D3100, VEECO Inc.). The
thickness of the BDD thin film was around 2.7Âľm for ÂľLED probes and 3.7Âľm for chemical
sensors estimated based on the deposition rate of the BDD synthesis system. Comparing
the two etching approaches, KOH etching results in weaker BDD-to-Parylene bond than
DRIE etching, mainly due to interface damage and delamination during 9 hours soaking in
corrosive

KOH.

While

DRIE

etching
88

eliminates

chemical

attack

on

Figure 6-3 (a) A SEM image shows the surface morphology of the BDD nucleation side. (b) A
Raman spectrum shows both boron and diamond bands from the BDD nucleation side.
Table 6-1 Summary of BDD film Characteristics

BDD-Parylene bond, the heat generated by aggressive plasma etching causes a
deterioration in mechanical flexibility and clarity of Parylene on some of the devices.

6.3.2 Mechanical properties of the transferred BDD-Parylene structure
Figure 6-4 (a) and 6-4(b) show a custom-designed KOH etching kit before and after
assembly, which includes a custom-made Teflon etching jig, a KOH resistant O-ring
(Mykin, Inc) and chromium-coated C-clamps (Graham field, Inc). Figure 6-4(c) shows
patterned BDD structures transferred on a 3-inch wafer-scale Parylene-C substrate, with a
close-up view of contact pads and ÂľLED electrodes. The SEM image in Figure 6-5(d)
shows Parylene anchors formed around the openings of the BDD mesh over large BDD
patterns. The flexibility of the BDD-Parylene structure was tested by wrapping the device
on the tip of a 2.5mm-diameter micro punch with the ÂľLED turned off (Figure 6-4 (e)) and
89

Figure 6-4 The custom designed KOH etching kit (a) before and (b) after assembly. (c) BDD
patterns on a wafer-scale flexible Parylene-C substrate after removing the Si substrate. Microscope
images shows contact pads and ÂľLED electrode probes with 200 Âľm and 50 Âľm trace width. (d)
A SEM image shows the mesh structure of the contact pads, where Parylene anchors, Parylene
substrate and BDD are highlighted in white, green, and blue, respectively. (e) and (f) Flexibility
of a BDD ÂľLED probe wrapped around the tip of a micro punch, with ÂľLED off and on. (g) and
(h) Scotch tapeÂŽ testing before(g) and after (h) peeling a BDD-Parylene sensor off the tape, with
the BDD side facing down.

90

Figure 6-4 (cont’d)

Figure 6-4 (cont’d) Scotch tape® testing before(g) and after (h) peeling a BDD-Parylene sensor
off the tape, with the BDD side facing down.
on (Figure 6-4 (f)), demonstrating the integrity of the conductive BDD wires during
bending. The enhanced adhesion between BDD and Parylene was demonstrated using
Scotch tapeÂŽ peeling test Figure 6-4(g) and (f). During the testing, a BDD-Parylene sensor
was pressed onto a piece of Scotch tapeÂŽ with the BDD side facing down and then peeled
off slowly by fingers. Afterward, the sensor and the tape were inspected under a
microscope. We did not observe BDD delamination or damage after five peelings,
suggesting a strong mechanical adhesion between the BDD patterns and the Parylene-C
substrate.

6.3.3 Characteristics of the BDD-Parylene electrochemical sensors
6.3.3.1 Potential window
Figure 6-5(a) shows a fabricated BDD-Parylene chemical sensor. The electrochemical
properties of the sensors were characterized using cyclic voltammetry (CV) (CHI604, CH
Instruments, Inc.). First, the potential window of an as-fabricated BDD chemical sensor
was quantified in 1.0 M potassium chloride (KCl) solutions, and compared with a
commercial standard gold (Au) electrode (CHI101, CH Instruments, Inc.), as shown Figure
6-5(b). During measurements, BDD or Au electrode was used as WE, commercially
91

available platinum (Pt) (CHI 102, CH Instruments, Inc.) and Ag/AgCl electrode (CHI 111,
CH Instruments, Inc.) were used as CE and RE, respectively, with a scan rate of 0.1 V/s.
The result shows that the BDD sensor exhibited featureless background current from -2.0
V to 1.5 V and a higher potential window of 3.5 V than the value of 2.2 V reported in [201]
under the same test conditions. The Au electrode shows reduction current at 0.5V and
oxidation current at a potential higher than 1.5 V, which defines a potential window of 1.0
V. The BDD electrode shows a much wider potential window than the Au electrode,
therefore providing a bigger degree of characterizing analytes.

Figure 6-5 (a) A fabricated BDD-Parylene sensor with three electrodes. (b) Voltammogram of Au
and BDD electrode in 1.0M KCL solution vs. Ag/AgCl at a scan rate of 0.1V/s (CE: Pt electrode,
RE: Ag/AgCl). (c) Voltammograms of BDD electrodes vs. BDD at various scan rates (CE: BDD,
RE: BDD). The voltammograms are offset for better visibility.

6.3.3.2 Double-layer capacitance
The double-layer capacitance (Cdl) resulting from an electrical double layer effect determines
92

the background current of a redox system using either a potential control or potential sweep
method. A lower double-layer capacitance can reduce the charging current and lower the
background noise, which contributes to a higher signal-to-noise ratio of chemical sensing. The Cdl
is determined by Eq. (6-1), where iav is the average current of the forward and reverse sweep of
CV, v is the scan rate of CV, and A is the electrode area[202]
Cdl = iav / vA

(6-1)

To quantify the Cdl of the BDD electrode, CV measurements were performed in 1.0 M KCl
solution with various scan rates. All the three electrodes on the same BDD sensor were used in the
measurement. Figure 6-5 (c) shows the voltammograms with scan rates of 0.1, 0.5, 1.0, 2.0 and
3.0 V/s. Figure 6-5 (d) plots the average current of forward and reverse sweeps at -0.9 V vs. BDD
with different scan rates. A linear regression equation (y=24x+23) with an R2 number of 0.979 is
generated and the slope of this equation defines the Cdl, which is 24 ÂľF/cm2.

6.3.3.3 Chemical redox characteristics
For a full evaluation of the electrochemical characteristics of the as-fabricated BDD sensor,
the sensors were tested with both outer and inner-sphere electron transfer processes by CV. For
outer-sphere electron, the one electron transfer, Ru(NH3)62+/3+ redox couple was studied. During
experiments, BDD was used as the WE, CE, and RE. Figure 6-6(a) shows the CV curves of the
sensor with various concentrations of Ru(NH3)62+/3+ (262005-250 MG, Sigma-Aldrich) in 1.0 M
KCl solution at a constant scan rate of 0.1 V/s. The peak potentials (Ep), half peak potentials (Ep/2)
and peak currents (ip) of oxidation process for different concentrations were extracted and
summarized in Table 2. In addition, the iR drop in the electrochemical cell caused by solution
resistance, contact resistance, BDD thin film resistance, etc. was measured, which were 0.94, 1.46,
1.69, and 2.65 mV for the Ru(NH3)62+/3+ concentration of 0.6, 0.8, 1, and 1.5 mM, respectively.

93

Those values are almost fifty times smaller compared with Ep and Ep/2 and can be ignored in the
data analysis. The peak current of oxidation process vs. different Ru(NH3)62+/3+concentrations is

Figure 6-6 Voltammograms of the BDD sensor with various concentrations of Ru(NH3)62+/3+ in
1.0M KCl solution vs. BDD at a scan rate of 0.1V/s. (CE: BDD, RE: BDD) (b) Fitting curve of
the oxidation peak current density versus different concentrations of Ruhex. (c) Voltammograms
of BDD electrodes vs. BDD at various scan rates (CE: BDD, RE: BDD) in 2 M Ru(NH3)62+/3+ with
1.0 M KCl solution. (d) Fitting curve of the oxidation peak current density versus square root of
scan rates.
Table 6-2 CV with different concentrations

plotted in Figure 6-6(b) and fitted to a linear regression curve (y=228.8x-14.6) with an R2 value of
0.991, indicating that this oxidation peak is assigned to Ruhex. The values of |Ep-Ep/2| were

94

calculated as shown in Table 6-2, which were 87, 86, 89, and 89 mV for the concentration of 0.6,
0.8, 1, and 1.5 mM, respectively. The experimental values were larger than the theoretical value
of 58.5 mV calculated from 58.5/n, where n is the number of electrons transferred in reaction
[202]. This implies a quasi-irreversible/irreversible electron transfer process with a small standard
rate constant (k0), during which an activationoverpotential Ρ beyond the Nernstian equilibrium
potential (Eeq) is required to drive the system to achieve the same exchange current level
determined by a reversible redox system, which leads to a relative larger Ep. The voltammograms
with different scan rates in 2.0 mM Ru(NH3)62+/3+ solution with 1.0 M KCl supporting electrolyte
are shown in Figure 6-6(c). The peak currents of oxidation process vs. the square root of the scan
rates are plotted in Figure 6-6(d). A linear regression (y=0.42x+0.52) was observed with an R2
value of 0.985, indicating that, although mesh structures were introduced at the electrodes, the
mass transfer process of such a system is still controlled by semi-infinite linear diffusion of
macroelectrodes rather than spherical diffusion of microelectrodes.

6.3.3.4 Dopamine sensing characteristics
For evaluation of the BDD sensors with an inner-sphere electron transfer system, DA was
studied by CV and chronoamperometry based on the chemical reaction in Figure 6-7(a), where
DA is oxidized to dopamine-o-quinone with two electrons transferred in the reaction[203]. Figure
6-7(b) shows the voltammograms of the BDD sensor vs. BDD at a scan rate of 1.0 V/s with various
concentrations of DA (H8502-10G, Sigma-Aldrich) in 0.1 M, pH=7.4 phosphate-buffered saline
(PBS) buffer solution. The inset of Figure 6-7(b) shows the voltammogram of 100 ÂľM DA after
subtracting the background charging current in the PBS solution, where an anodic peak potential
(Ep) was observed at 1.0 V vs. BDD with a peak current of 0.66 ÂľA/mm2. The corresponding halfpeak potential (Ep/2), which defines the potential at the half peak current, is 0.816 V. The iR drops

95

at Ep and Ep/2 were measured and calculated using the aforementioned method, which were 2.56
mV and 1.28 mV, respectively. These values are two orders of magnitude smaller than the values

Figure 6-7 (a) Dopamine / Dopamine-o-quinone redox. (b) Voltammograms of the BDD sensor in
various concentrations of DA with 0.1 M PBS vs. BDD. The scan rates are 1.0 V/s (c)
Chronoamperograms of various concentrations of DA in PBS buffer solution vs. BDD at an
applied potential of 1.0 V. (d) Fitting curve of the background corrected current versus the
concentrations of DA.
of Ep and Ep/2 and can be neglected. The calculated value of |Ep-Ep/2| was 184 mV, much larger
than the theoretical value of 28.25 mV [202], indicating an electrochemically irreversible/quasiirreversible process. Chronoamperometry was used to study the ability of BDD sensors to detect
low DA concentrations. In our study, a single step of 1.0 V vs. BDD was applied and current
decayed exponentially with time was recorded. Fig. 6-7(c) shows chronoamperograms with 0.5ÂľM
to 100 ÂľM DA in 0.1 M, pH=7.4 PBS buffer solution. With different DA concentrations, the
background corrected currents, which are calculated by subtracting the current of background PBS
96

buffer solution from the steady-state currents of chronoamperograms, are plotted in Figure 6-7(d).
A linear response of the BDD sensor to DA for the concentration of 0.5 ÂľM to 100 ÂľM is observed
(R2 = 0.999) with a detection limit of 0.5 ¾M and a sensitivity of 0.21 ¾A/cm2¡¾M

6.4 Conclusions
In summary, this section presented a novel fabrication process of transferring all diamond
structures from a diamond growth substrate, silicon, onto a flexible Parylene C substrate, which
enables the fabrication of all diamond on polymer devices without any metal traces in a wafer scale
fabrication fashion. The unique design of Parylene anchors greatly enhanced the bonding strength
between BDD and Parylene-C. The integrity of BDD thin films was demonstrated by bending a lit
ÂľLED BDD device on the tip of a micropunch. The comparative study against traditional gold
electrodes highlighted the advantages of BDD-polymer devices. Detailed experimental
investigation shows that the as-fabricated flexible BDD chemical sensor has a much wider
potential window and smaller double layer capacitance compared with traditional gold electrodes,
which gives wider freedom on detecting various chemicals. The electron transfer process and mass
transfer process were studied using outer-sphere redox couple Ru(NH3)62+/3+ and the capability of
sensing various concentrations of DA was demonstrated using chronoamperometry with the linear
response from 0.5 ÂľM to 100 ÂľM (R2 = 0.999). The fabrication methodology proposed in this
paper can be easily adapted into the fabrication of larger-scale flexible BDD devices for large area
or multiple targeted areas of interest. Future efforts include the development of multi-function
flexible BDD arrays for both neural electrical and chemical signal recording in vivo in a rat brain
model.

97

Chapter 7.

Conclusion

Recent advances in Optogenetics provide a unique neuromodulation technique,
allowing optical control of genetically targeted specific neurons that express light-sensitive
opsin proteins with sub-millisecond temporal precision. Applying Optogenetics into
studying DA release and uptake dynamics, researchers can have a more controlled
manipulation of the dopaminergic system and have an unbiased study on DA related
neurological diseases. However, the engineering tools which combines optical stimulation,
neuroelectrical recording and neurochemical recording in a compact, low power fashion
has not been reported. The ultimate goal of this thesis is to develop such tools from material
and device perspective, and gradually bring them from laboratory to the market.
The motivation of this thesis started from observing joule heating issue from an SU8 based
ÂľLED probes. To address such issue, polycrystalline diamond has been introduced as a heat
spreader for the first time for making ÂľLED neural probes. Targeting at dissipating joule heating
more efficiently, a PCD-based optoelectronic neural interfacing probe has been designed,
fabricated and characterized, which is capable of optical stimulation and simultaneous recording
of neural activity. The experimental and simulation results show that the application of the PCD
heat spreader can keep the temperature variations of the probe around 1.3°C during optical
stimulation. In-vivo testing was performed in V1 of an Optogenetic rat, which demonstrated the
functionality of optical stimulating and electrical recording of action potentials of ChR2
transfected neurons.
Besides, the very high thermal conductivity, polycrystalline diamond has many unique
features for chemical sensing, such as a large potential window, low background current and
resistance to surface fouling, and long – term implantation, such as biocompatibility and chemical

98

inertness. The combination of those superior properties makes polycrystalline diamond a very
promising candidate for a neural interface that has both functionalities of optical neuromodulation
and neurotransmitter detection.
The motivation of designing a wafer transfer process raised from the rigidity of diamond thin
film that the micro-motion between diamond and surrounding tissue can cause a harmful immune
response and the process intolerance that cannot synthesis diamond onto flexible polymer substrate
directly. The wafer transfer process introduced Parylene anchors that can increase the bonding
adhesion between transferred diamond and Parylene substrate, which has been demonstrated by
peeling test. To study the electrochemical properties of the transferred diamond-polymer
electrodes, an outer sphere redox couple Ru(NH3)6

2+/3+

has been used to study electron transfer

process. Then quantitative and qualitative studies of redox dopamine / dopamine-o-quinone have
been performed to demonstrate the efficacy of neurotransmitter detection.
The future work will involve the integration of individual modules such as ÂľLEDs,
neurotransmitter detection, and neuroelectrical signal recording. The interference of a
simultaneous recording neuroelectrical signal and neurotransmitter detection could be a challenge
since the applied voltage from chemical sensing may interference the recorded signals. This issue
requires careful design on the sensing circuitries and post signal processing. The initial success of
applying diamond-polymer devices into neural interface application has shone some light on a new
generation of neural interface strategy. It should be anticipated that diamond-Parylene technology
will become an emerging technique in the field of implanted devices in the near future.

99

BIBLIOGRAPHY

100

BIBLIOGRAPHY

[1]

Y. Lu, N. Driscoll, I. Ozden, Z. Yu, and A. V. Nurmikko, “Modulating dopamine release
by optogenetics in transgenic mice reveals terminal dopaminergic dynamics,” Neurophoton,
vol. 2, no. 3, pp. 031207–031207, 2015.

[2]

Z. Fekete, “Recent advances in silicon-based neural microelectrodes and microsystems: a
review,” Sensors and Actuators B: Chemical, vol. 215, pp. 300–315, Aug. 2015.

[3]

S. F. Cogan, “Neural Stimulation and Recording Electrodes,” Annual Review of Biomedical
Engineering, vol. 10, no. 1, pp. 275–309, 2008.

[4]

S. Butovas and C. Schwarz, “Spatiotemporal effects of microstimulation in rat neocortex: a
parametric study using multielectrode recordings,” J. Neurophysiol., vol. 90, no. 5, pp.
3024–3039, Nov. 2003.

[5]

G. Nagel et al., “Channelrhodopsin-2, a directly light-gated cation-selective membrane
channel,” Proc Natl Acad Sci U S A, vol. 100, no. 24, pp. 13940–13945, Nov. 2003.

[6]

B. Schobert and J. K. Lanyi, “Halorhodopsin is a light-driven chloride pump,” J. Biol. Chem.,
vol. 257, no. 17, pp. 10306–10313, Sep. 1982.

[7]

B. Y. Chow et al., “High-performance genetically targetable optical neural silencing by
light-driven proton pumps,” Nature, vol. 463, no. 7277, pp. 98–102, Jan. 2010.

[8]

F. Zhang et al., “Optogenetic interrogation of neural circuits: technology for probing
mammalian brain structures,” Nat. Protocols, vol. 5, no. 3, pp. 439–456, Mar. 2010.

[9]

F. Mora, G. Segovia, A. del Arco, M. de Blas, and P. Garrido, “Stress, neurotransmitters,
corticosterone and body–brain integration,” Brain Research, vol. 1476, pp. 71–85, Oct.
2012.

[10] V. Hefco, K. Yamada, A. Hefco, L. Hritcu, A. Tiron, and T. Nabeshima, “Role of the
mesotelencephalic dopamine system in learning and memory processes in the rat,”
European Journal of Pharmacology, vol. 475, no. 1–3, pp. 55–60, Aug. 2003.
[11] J. D. Berke and S. E. Hyman, “Addiction, Dopamine, and the Molecular Mechanisms of
Memory,” Neuron, vol. 25, no. 3, pp. 515–532, Mar. 2000.
[12] E. R. Kandel, J. H. Schwartz, and T. M. Jessell, Principles of Neural Science, Fourth Edition.
McGraw-Hill.
[13] R. H. McDonald and L. I. Goldberg, “Analysis of the Cardiovascular Effects of Dopamine
in the Dog,” J Pharmacol Exp Ther, vol. 140, no. 1, pp. 60–66, Apr. 1963.
101

[14] B. R, C. M, F. S, H. K, and M. J, “Low-dose dopamine in patients with early renal
dysfunction: a placebo-controlled randomised trial. Australian and New Zealand Intensive
Care Society (ANZICS) Clinical Trials Group.,” Lancet, vol. 356, no. 9248, pp. 2139–2143,
2000 1999.
[15] S. Threlfell, T. Lalic, N. J. Platt, K. A. Jennings, K. Deisseroth, and S. J. Cragg, “Striatal
Dopamine Release Is Triggered by Synchronized Activity in Cholinergic Interneurons,”
Neuron, vol. 75, no. 1, pp. 58–64, Jul. 2012.
[16] C. E. Bass, V. P. Grinevich, A. D. Kulikova, K. D. Bonin, and E. A. Budygin, “Terminal
effects of optogenetic stimulation on dopamine dynamics in rat striatum,” Journal of
Neuroscience Methods, vol. 214, no. 2, pp. 149–155, Apr. 2013.
[17] J. R. Melchior, M. J. Ferris, G. D. Stuber, D. R. Riddle, and S. R. Jones, “Optogenetic versus
electrical stimulation of dopamine terminals in the nucleus accumbens reveals local
modulation of presynaptic release,” J Neurochem, vol. 134, no. 5, pp. 833–844, Sep. 2015.
[18] Y. Iwai, S. Honda, H. Ozeki, M. Hashimoto, and H. Hirase, “A simple head-mountable LED
device for chronic stimulation of optogenetic molecules in freely moving mice,” Neurosci.
Res., vol. 70, no. 1, pp. 124–127, May 2011.
[19] P. of SPIE, “Optogenetics: Optical Methods for Cellular Control,” 2013, vol. 8586, pp.
858601-858601–12.
[20] J. G. McCall et al., “Fabrication and application of flexible, multimodal light-emitting
devices for wireless optogenetics,” Nat. Protocols, vol. 8, no. 12, pp. 2413–2428, Dec. 2013.
[21] P. Andersen and E. I. Moser, “Brain temperature and hippocampal function,” Hippocampus,
vol. 5, no. 6, pp. 491–498, Jan. 1995.
[22] C. Childs, “Human brain temperature: regulation, measurement and relationship with
cerebral trauma: Part 1,” Br J Neurosurg, vol. 22, no. 4, pp. 486–496, Jan. 2008.
[23] Y. Li et al., “Thermal analysis of injectable, cellular-scale optoelectronics with pulsed
power,” Proceedings of the Royal Society of London A: Mathematical, Physical and
Engineering Sciences, vol. 469, no. 2156, p. 20130142, Aug. 2013.
[24] Y. Li et al., “Thermal properties of microscale inorganic light-emitting diodes in a pulsed
operation,” Journal of Applied Physics, vol. 113, no. 14, p. 144505, Apr. 2013.
[25] T. Kim et al., “High-Efficiency, Microscale GaN Light-Emitting Diodes and Their Thermal
Properties on Unusual Substrates,” Small, vol. 8, no. 11, pp. 1643–1649, Jun. 2012.
[26] T. Kim et al., “Injectable, Cellular-Scale Optoelectronics with Applications for Wireless
Optogenetics,” Science, vol. 340, no. 6129, pp. 211–216, Apr. 2013.
[27] S. V. Kidalov and F. M. Shakhov, “Thermal Conductivity of Diamond Composites,”
Materials, vol. 2, no. 4, pp. 2467–2495, Dec. 2009.
102

[28] E. Fiedler, N. Haas, and T. Stieglitz, “Suitability of SU-8, EpoClad and EpoCore for flexible
waveguides on implantable neural probes,” Conf Proc IEEE Eng Med Biol Soc, vol. 2014,
pp. 438–441, Aug. 2014.
[29] L. Eldada and L. W. Shacklette, “Advances in polymer integrated optics,” IEEE Journal of
Selected Topics in Quantum Electronics, vol. 6, no. 1, pp. 54–68, Jan. 2000.
[30] M. Vogt and R. Hauptmann, “Plasma-deposited passivation layers for moisture and water
protection,” Surface and Coatings Technology, vol. 74–75, Part 2, pp. 676–681, Oct. 1995.
[31] J. Subbaroyan, D. C. Martin, and D. R. Kipke, “A finite-element model of the mechanical
effects of implantable microelectrodes in the cerebral cortex,” J Neural Eng, vol. 2, no. 4,
pp. 103–113, Dec. 2005.
[32] V. S. Polikov, P. A. Tresco, and W. M. Reichert, “Response of brain tissue to chronically
implanted neural electrodes,” J. Neurosci. Methods, vol. 148, no. 1, pp. 1–18, Oct. 2005.
[33] J. P. Harris et al., “Mechanically adaptive intracortical implants improve the proximity of
neuronal cell bodies,” J Neural Eng, vol. 8, no. 6, p. 66011, Dec. 2011.
[34] D.-H. Kim et al., “Dissolvable films of silk fibroin for ultrathin conformal bio-integrated
electronics,” Nat Mater, vol. 9, no. 6, pp. 511–517, Jun. 2010.
[35] F. Barz, P. Ruther, S. Takeuchi, and O. Paul, “Flexible silicon-polymer neural probe
rigidified by dissolvable insertion vehicle for high-resolution neural recording with
improved duration,” in 2015 28th IEEE International Conference on Micro Electro
Mechanical Systems (MEMS), 2015, pp. 636–639.
[36] http://www.dow.com/polyglycols/polyethylene/products/carbowaxp.htm, .
[37] G. Jiang and D. D. Zhou, “Technology Advances and Challenges in Hermetic Packaging
for Implantable Medical Devices,” in Implantable Neural Prostheses 2, D. Zhou and E.
Greenbaum, Eds. Springer New York, 2009, pp. 27–61.
[38] S. T, “Methods to Determine the Stability of Polymer Encapsulations.”
[39] M. Kazemi et al., “A test microchip for evaluation of hermetic packaging technology for
biomedical prosthetic implants,” in 26th Annual International Conference of the IEEE
Engineering in Medicine and Biology Society, 2004. IEMBS ’04, 2004, vol. 2, pp. 4093–
4095.
[40] X. Xie et al., “Long-term reliability of Al2O3 and Parylene C bilayer encapsulated Utah
electrode array based neural interfaces for chronic implantation,” J. Neural Eng., vol. 11,
no. 2, p. 26016, Apr. 2014.
[41] Y. Xie et al., “In vivo monitoring of glial scar proliferation on chronically implanted neural
electrodes by fiber optical coherence tomography,” Front Neuroeng, vol. 7, Aug. 2014.

103

[42] P. Fattahi, G. Yang, G. Kim, and M. R. Abidian, “A Review of Organic and Inorganic
Biomaterials for Neural Interfaces,” Adv. Mater., vol. 26, no. 12, pp. 1846–1885, Mar. 2014.
[43] C. Hassler, T. Boretius, and T. Stieglitz, “Polymers for neural implants,” J. Polym. Sci. B
Polym. Phys., vol. 49, no. 1, pp. 18–33, Jan. 2011.
[44] J. M. Morais, F. Papadimitrakopoulos, and D. J. Burgess, “Biomaterials/Tissue Interactions:
Possible Solutions to Overcome Foreign Body Response,” AAPS J, vol. 12, no. 2, pp. 188–
196, Feb. 2010.
[45] Y. Einaga, J. S. Foord, and G. M. Swain, “Diamond electrodes: Diversity and maturity,”
MRS Bulletin, vol. 39, no. 6, pp. 525–532, Jun. 2014.
[46] C. Bamann, R. Gueta, S. Kleinlogel, G. Nagel, and E. Bamberg, “Structural guidance of the
photocycle of channelrhodopsin-2 by an interhelical hydrogen bond,” Biochemistry, vol. 49,
no. 2, pp. 267–278, Jan. 2010.
[47] J. H. T. Luong, K. B. Male, and J. D. Glennon, “Boron-doped diamond electrode: synthesis,
characterization, functionalization and analytical applications,” Analyst, vol. 134, no. 10, pp.
1965–1979, Sep. 2009.
[48] F. Z. Cui and D. J. Li, “A review of investigations on biocompatibility of diamond-like
carbon and carbon nitride films,” Surface and Coatings Technology, vol. 131, no. 1–3, pp.
481–487, Sep. 2000.
[49] L. Tang, C. Tsai, W. W. Gerberich, L. Kruckeberg, and D. R. Kania, “Biocompatibility of
chemical-vapour-deposited diamond,” Biomaterials, vol. 16, no. 6, pp. 483–488, 1995.
[50] B. J. Kim et al., “3D Parylene sheath neural probe for chronic recordings,” J Neural Eng,
vol. 10, no. 4, p. 45002, Aug. 2013.
[51] S. R. Ali, R. R. Parajuli, Y. Balogun, Y. Ma, and H. He, “A Nonoxidative Electrochemical
Sensor Based on a Self-Doped Polyaniline/Carbon Nanotube Composite for Sensitive and
Selective Detection of the Neurotransmitter Dopamine: A Review,” Sensors, vol. 8, no. 12,
pp. 8423–8452, Dec. 2008.
[52] J. T. W. Kuo et al., “Novel flexible Parylene neural probe with 3D sheath structure for
enhancing tissue integration,” Lab Chip, vol. 13, no. 4, pp. 554–561, Jan. 2013.
[53] K. Jackowska and P. Krysinski, “New trends in the electrochemical sensing of dopamine,”
Anal Bioanal Chem, vol. 405, no. 11, pp. 3753–3771, Apr. 2013.
[54] A. Stoyanova and V. Tsakova, “Copper-modified poly(3,4-ethylenedioxythiophene) layers
for selective determination of dopamine in the presence of ascorbic acid: II Role of the
characteristics of the metal deposit,” J Solid State Electrochem, vol. 14, no. 11, pp. 1957–
1965, Nov. 2010.

104

[55] J. John, Y. Li, J. Zhang, J. A. Loeb, and Y. Xu, “Microfabrication of 3D neural probes with
combined electrical and chemical interfaces,” J. Micromech. Microeng., vol. 21, no. 10, p.
105011, Oct. 2011.
[56] P. Norlin, M. Kindlundh, A. Mouroux, K. Yoshida, and U. G. Hofmann, “A 32-site neural
recording probe fabricated by DRIE of SOI substrates,” J. Micromech. Microeng., vol. 12,
no. 4, p. 414, 2002.
[57] M. Jorfi, J. L. Skousen, C. Weder, and J. R. Capadona, “Progress towards biocompatible
intracortical microelectrodes for neural interfacing applications,” J Neural Eng, vol. 12, no.
1, p. 11001, Feb. 2015.
[58] R. L. McCreery, “Advanced Carbon Electrode Materials for Molecular Electrochemistry,”
Chem. Rev., vol. 108, no. 7, pp. 2646–2687, Jul. 2008.
[59] R. S. Balmer et al., “Chemical vapour deposition synthetic diamond: materials, technology
and applications,” J. Phys.: Condens. Matter, vol. 21, no. 36, p. 364221, 2009.
[60] “Carbon coatings for medical implants,” Journal of Achievements in Materials and
Manufacturing Engineering, no. 1–2, Feb. 2007.
[61] I. Dion, C. Baquey, and J. R. Monties, “Diamond: the biomaterial of the 21st century?,” Int
J Artif Organs, vol. 16, no. 9, pp. 623–627, Sep. 1993.
[62] M. Hupert et al., “Conductive diamond thin-films in electrochemistry,” Diamond and
Related Materials, vol. 12, no. 10–11, pp. 1940–1949, Oct. 2003.
[63] A. Kraft, “Doped diamond: a compact review on a new, versatile electrode material,” Int. J.
Electrochem. Sci, vol. 2, pp. 355–385, 2007.
[64] B. Fan, Rechenberg, Robert, F.Becker, Michael, and Li, Wen, “Fabrication of
polycrystalline diamond on a flexible parylene substrate,” presented at the The 18th
international conference on Solid-State Sensors, Actuators and Microsystems, 2015.
[65] “Silicon thermal conductivity.” [Online]. Available: https://en.wikipedia.org/wiki/Silicon.
[66] “Copper thermal conductivity.”
astr.gsu.edu/hbase/tables/thrcn.html.
[67] “Gold
thermal
conductivity.”
astr.gsu.edu/hbase/tables/thrcn.html.

[Online].
[Online].

Available:

http://hyperphysics.phy-

Available:

http://hyperphysics.phy-

[68] “Parylene
thermal
conductivity.”
[Online].
http://www.physics.rutgers.edu/~podzorov/parylene%20properties.pdf.

Available:

[69] “Iridium thermal conductivity.” [Online]. Available: https://en.wikipedia.org/wiki/Iridium.

105

[70] “SU8 thermal conductivity.” [Online]. Available: http://www.mit.edu/~6.777/matprops/su8.htm.
[71] “Platinum
thermal
conductivity.”
https://en.wikipedia.org/wiki/Platinum.
[72] “Material
Property
Database_Polyimide.”
http://www.mit.edu/~6.777/matprops/polyimide.htm.

[Online].
[Online].

[73] Diamond Materials, “The CVD diamond booklet.” [Online].
http://www.diamond-materials.com/downloads/cvd_diamond_booklet.pdf.

Available:
Available:
Available:

[74] P. Schiavone, F. Chassat, T. Boudou, E. Promayon, F. Valdivia, and Y. Payan, “In vivo
measurement of human brain elasticity using a light aspiration device,” Medical Image
Analysis, vol. 13, no. 4, pp. 673–678, Aug. 2009.
[75] K.-K. Lee et al., “Polyimide-based intracortical neural implant with improved structural
stiffness,” J. Micromech. Microeng., vol. 14, no. 1, p. 32, 2004.
[76] B. Fan, K. Y. Kwon, A. J. Weber, and W. Li, “An implantable, miniaturized SU-8 optical
probe for optogenetics-based deep brain stimulation,” in 2014 36th Annual International
Conference of the IEEE Engineering in Medicine and Biology Society, 2014, pp. 450–453.
[77] D. C. Rodger et al., “Flexible parylene-based multielectrode array technology for highdensity neural stimulation and recording,” Sensors and Actuators B: Chemical, vol. 132, no.
2, pp. 449–460, Jun. 2008.
[78] T. Stieglitz, H. Beutel, M. Schuettler, and J.-U. Meyer, “Micromachined, Polyimide-Based
Devices for Flexible Neural Interfaces,” Biomedical Microdevices, vol. 2, no. 4, pp. 283–
294.
[79] “Material
Property
Database_PDMS.”
http://www.mit.edu/~6.777/matprops/pdms.htm.

[Online].

Available:

[80] MicroChem, “SU-8 2000 Permanet Epoxy Negative Photoresist Datasheet.” [Online].
Available: http://www.microchem.com/pdf/SU-82000DataSheet2025thru2075Ver4.pdf.
[81] Specialty coating systems, “SCS Parylene Properties,” SCS Parylene Properties. [Online].
Available: http://www.physics.rutgers.edu/~podzorov/parylene%20properties.pdf.
[82] R. Ramamurti et al., “Boron doped diamond deposited by microwave plasma-assisted CVD
at low and high pressures,” Diamond and Related Materials, vol. 17, no. 4–5, pp. 481–485,
Apr. 2008.
[83] O. Yizhar, L. E. Fenno, T. J. Davidson, M. Mogri, and K. Deisseroth, “Optogenetics in
Neural Systems,” Neuron, vol. 71, no. 1, pp. 9–34, Jul. 2011.

106

[84] G. Nagel et al., “Channelrhodopsin-1: a light-gated proton channel in green algae,” Science,
vol. 296, no. 5577, pp. 2395–2398, Jun. 2002.
[85] E. S. Boyden, F. Zhang, E. Bamberg, G. Nagel, and K. Deisseroth, “Millisecond-timescale,
genetically targeted optical control of neural activity,” Nat Neurosci, vol. 8, no. 9, pp. 1263–
1268, Sep. 2005.
[86] B. J. Tromberg et al., “Non-Invasive In Vivo Characterization of Breast Tumors Using
Photon Migration Spectroscopy,” Neoplasia, vol. 2, no. 1–2, pp. 26–40, Jan. 2000.
[87] A. M. Aravanis et al., “An optical neural interface: in vivo control of rodent motor cortex
with integrated fiberoptic and optogenetic technology,” J Neural Eng, vol. 4, no. 3, pp.
S143-156, Sep. 2007.
[88] A. Berndt et al., “High-efficiency channelrhodopsins for fast neuronal stimulation at low
light levels,” Proc. Natl. Acad. Sci. U.S.A., vol. 108, no. 18, pp. 7595–7600, May 2011.
[89] G. Nagel, M. Brauner, J. F. Liewald, N. Adeishvili, E. Bamberg, and A. Gottschalk, “Light
activation of channelrhodopsin-2 in excitable cells of Caenorhabditis elegans triggers rapid
behavioral responses,” Curr. Biol., vol. 15, no. 24, pp. 2279–2284, Dec. 2005.
[90] F. Zhang et al., “Red-shifted optogenetic excitation: a tool for fast neural control derived
from Volvox carteri,” Nat Neurosci, vol. 11, no. 6, pp. 631–633, Jun. 2008.
[91] O. Yizhar et al., “Neocortical excitation/inhibition balance in information processing and
social dysfunction,” Nature, vol. 477, no. 7363, pp. 171–178, Sep. 2011.
[92] Photosensitive Molecules for Controlling Biological | James J. Chambers | Springer. .
[93] F. Zhang et al., “Multimodal fast optical interrogation of neural circuitry,” Nature, vol. 446,
no. 7136, pp. 633–639, Apr. 2007.
[94] V. Gradinaru et al., “Molecular and cellular approaches for diversifying and extending
optogenetics,” Cell, vol. 141, no. 1, pp. 154–165, Apr. 2010.
[95] A. Berndt, O. Yizhar, L. A. Gunaydin, P. Hegemann, and K. Deisseroth, “Bi-stable neural
state switches,” Nat Neurosci, vol. 12, no. 2, pp. 229–234, Feb. 2009.
[96] J.-M. Kim, J. Hwa, P. Garriga, P. J. Reeves, U. L. RajBhandary, and H. G. Khorana, “LightDriven Activation of β2-Adrenergic Receptor Signaling by a Chimeric Rhodopsin
Containing the β2-Adrenergic Receptor Cytoplasmic Loops,” Biochemistry, vol. 44, no. 7,
pp. 2284–2292, Feb. 2005.
[97] R. D. Airan, K. R. Thompson, L. E. Fenno, H. Bernstein, and K. Deisseroth, “Temporally
precise in vivo control of intracellular signalling,” Nature, vol. 458, no. 7241, pp. 1025–
1029, Apr. 2009.

107

[98] E. Oh, T. Maejima, C. Liu, E. Deneris, and S. Herlitze, “Substitution of 5-HT1A Receptor
Signaling by a Light-activated G Protein-coupled Receptor,” J. Biol. Chem., vol. 285, no.
40, pp. 30825–30836, Oct. 2010.
[99] K. M. Tye and K. Deisseroth, “Optogenetic investigation of neural circuits underlying brain
disease in animal models,” Nat Rev Neurosci, vol. 13, no. 4, pp. 251–266, Apr. 2012.
[100] M. Stierl et al., “Light modulation of cellular cAMP by a small bacterial photoactivated
adenylyl cyclase, bPAC, of the soil bacterium Beggiatoa,” J. Biol. Chem., vol. 286, no. 2,
pp. 1181–1188, Jan. 2011.
[101] M.-H. Ryu, O. V. Moskvin, J. Siltberg-Liberles, and M. Gomelsky, “Natural and engineered
photoactivated nucleotidyl cyclases for optogenetic applications,” J. Biol. Chem., vol. 285,
no. 53, pp. 41501–41508, Dec. 2010.
[102] E. S. Boyden, F. Zhang, E. Bamberg, G. Nagel, and K. Deisseroth, “Millisecond-timescale,
genetically targeted optical control of neural activity,” Nat Neurosci, vol. 8, no. 9, pp. 1263–
1268, Sep. 2005.
[103] T. Ishizuka, M. Kakuda, R. Araki, and H. Yawo, “Kinetic evaluation of photosensitivity in
genetically engineered neurons expressing green algae light-gated channels,” Neurosci. Res.,
vol. 54, no. 2, pp. 85–94, Feb. 2006.
[104] B. E. Losavio, V. Iyer, S. Patel, and P. Saggau, “Acousto-optic laser scanning for multi-site
photo-stimulation of single neurons in vitro,” J Neural Eng, vol. 7, no. 4, p. 45002, Aug.
2010.
[105] A. B. Arrenberg, D. Y. R. Stainier, H. Baier, and J. Huisken, “Optogenetic Control of
Cardiac Function,” Science, vol. 330, no. 6006, pp. 971–974, Nov. 2010.
[106] P. Zhu, O. Fajardo, J. Shum, Y.-P. Zhang Schärer, and R. W. Friedrich, “High-resolution
optical control of spatiotemporal neuronal activity patterns in zebrafish using a digital
micromirror device,” Nat Protoc, vol. 7, no. 7, pp. 1410–1425, Jul. 2012.
[107] A. M. Aravanis et al., “An optical neural interface: in vivo control of rodent motor cortex
with integrated fiberoptic and optogenetic technology,” J Neural Eng, vol. 4, no. 3, pp.
S143-156, Sep. 2007.
[108] R. D. Airan, K. R. Thompson, L. E. Fenno, H. Bernstein, and K. Deisseroth, “Temporally
precise in vivo control of intracellular signalling,” Nature, vol. 458, no. 7241, pp. 1025–
1029, Apr. 2009.
[109] P. Anikeeva et al., “Optetrode: a multichannel readout for optogenetic control in freely
moving mice,” Nat. Neurosci., vol. 15, no. 1, pp. 163–170, Jan. 2012.
[110] J. A. Cardin et al., “Targeted optogenetic stimulation and recording of neurons in vivo using
cell-type-specific expression of Channelrhodopsin-2,” Nat Protoc, vol. 5, no. 2, pp. 247–
254, Feb. 2010.
108

[111] D. F. English et al., “GABAergic circuits mediate the reinforcement-related signals of
striatal cholinergic interneurons,” Nat Neurosci, vol. 15, no. 1, pp. 123–130, Jan. 2012.
[112] K. R. Dhakal et al., “Non-scanning fiber-optic near-infrared beam led to two-photon
optogenetic stimulation in-vivo,” PLoS ONE, vol. 9, no. 11, p. e111488, 2014.
[113] X. Han et al., “Millisecond-timescale optical control of neural dynamics in the nonhuman
primate brain,” Neuron, vol. 62, no. 2, pp. 191–198, Apr. 2009.
[114] I. Ozden et al., “A coaxial optrode as multifunction write-read probe for optogenetic studies
in non-human primates,” Journal of Neuroscience Methods, vol. 219, no. 1, pp. 142–154,
Sep. 2013.
[115] I. Diester et al..., “An optogenetic toolbox designed for primates,” Nat Neurosci, vol. 14,
no. 3, pp. 387–397, Mar. 2011.
[116] O. Ruiz et al..., “Optogenetics through windows on the brain in the nonhuman primate,”
Journal of Neurophysiology, vol. 110, no. 6, pp. 1455–1467, Sep. 2013.
[117] E. Stark, T. Koos, and G. Buzsáki, “Diode probes for spatiotemporal optical control of
multiple neurons in freely moving animals,” Journal of Neurophysiology, vol. 108, no. 1,
pp. 349–363, Jul. 2012.
[118] Y. LeChasseur et al., “A microprobe for parallel optical and electrical recordings from
single neurons in vivo,” Nat Meth, vol. 8, no. 4, pp. 319–325, Apr. 2011.
[119] S. Dufour et al., “A Multimodal Micro-Optrode Combining Field and Single Unit Recording,
Multispectral Detection and Photolabeling Capabilities,” PLoS One, vol. 8, no. 2, Feb. 2013.
[120] M. Schwaerzle, K. Seidl, U. T. Schwarz, O. Paul, and P. Ruther, “Ultracompact optrode
with integrated laser diode chips and SU-8 waveguides for optogenetic applications,” in
2013 IEEE 26th International Conference on Micro Electro Mechanical Systems (MEMS),
2013, pp. 1029–1032.
[121] B. Rubehn, S. B. E. Wolff, P. Tovote, M. Schuettler, A. Lüthi, and T. Stieglitz, “Polymerbased shaft microelectrodes with optical and fluidic capabilities as a tool for optogenetics,”
Conf Proc IEEE Eng Med Biol Soc, vol. 2011, pp. 2969–2972, 2011.
[122] Y. Son et al., “A new monolithically integrated multi-functional MEMS neural probe for
optical stimulation and drug delivery,” in 2015 28th IEEE International Conference on
Micro Electro Mechanical Systems (MEMS), 2015, pp. 158–161.
[123] A. N. Zorzos, E. S. Boyden, and C. G. Fonstad, “Multiwaveguide implantable probe for
light delivery to sets of distributed brain targets,” Opt Lett, vol. 35, no. 24, pp. 4133–4135,
Dec. 2010.
[124] F. Pisanello et al., “Multipoint-emitting optical fibers for spatially addressable in vivo
optogenetics,” Neuron, vol. 82, no. 6, pp. 1245–1254, Jun. 2014.
109

[125] A. N. Zorzos, J. Scholvin, E. S. Boyden, and C. G. Fonstad, “Three-dimensional
multiwaveguide probe array for light delivery to distributed brain circuits,” Opt. Lett., OL,
vol. 37, no. 23, pp. 4841–4843, Dec. 2012.
[126] J. Zhang et al..., “Integrated device for optical stimulation and spatiotemporal electrical
recording of neural activity in light-sensitized brain tissue,” J Neural Eng, vol. 6, no. 5, p.
55007, Oct. 2009.
[127] T. V. F. Abaya, S. Blair, P. Tathireddy, L. Rieth, and F. Solzbacher, “A 3D glass optrode
array for optical neural stimulation,” Biomed Opt Express, vol. 3, no. 12, pp. 3087–3104,
Dec. 2012.
[128] F. Zhang et al., “Optogenetic interrogation of neural circuits: technology for probing
mammalian brain structures,” Nat. Protocols, vol. 5, no. 3, pp. 439–456, Mar. 2010.
[129] J. Zhang et al., “A microelectrode array incorporating an optical waveguide device for
stimulation and spatiotemporal electrical recording of neural activity,” Conf Proc IEEE Eng
Med Biol Soc, vol. 2009, pp. 2046–2049, 2009.
[130] J. Wang et al., “Approaches to optical neuromodulation from rodents to non-human
primates by integrated optoelectronic devices,” Conf Proc IEEE Eng Med Biol Soc, vol.
2011, pp. 7525–7528, 2011.
[131] J. Wang et al., “Integrated device for combined optical neuromodulation and electrical
recording for chronic in vivo applications,” J Neural Eng, vol. 9, no. 1, p. 16001, Feb. 2012.
[132] F. Wu et al., “An implantable neural probe with monolithically integrated dielectric
waveguide and recording electrodes for optogenetics applications,” J. Neural Eng., vol. 10,
no. 5, p. 56012, Oct. 2013.
[133] C.-W. Chang and J.-C. Chiou, “Development of a Three Dimensional Neural Sensing
Device by a Stacking Method,” Sensors, vol. 10, no. 5, pp. 4238–4252, Apr. 2010.
[134] Y. Yao, M. N. Gulari, J. A. Wiler, and K. D. Wise, “A Microassembled Low-Profile ThreeDimensional Microelectrode Array for Neural Prosthesis Applications,” Journal of
Microelectromechanical Systems, vol. 16, no. 4, pp. 977–988, Aug. 2007.
[135] A. N. Zorzos, J. Scholvin, E. S. Boyden, and C. G. Fonstad, “3-Dimensional
Multiwaveguide Probe Array for Light Delivery to Distributed Brain Circuits,” Opt Lett,
vol. 37, no. 23, pp. 4841–4843, Dec. 2012.
[136] C.-W. Jeon, K.-S. Kim, and M. d. Dawson, “Fabrication of Two-Dimensional InGaN-Based
Micro-LED Arrays,” phys. stat. sol. (a), vol. 192, no. 2, pp. 325–328, Aug. 2002.
[137] H. X. Zhang et al., “Individually-addressable flip-chip AlInGaN micropixelated light
emitting diode arrays with high continuous and nanosecond output power,” Opt Express,
vol. 16, no. 13, pp. 9918–9926, Jun. 2008.

110

[138] K. E. Jones, P. K. Campbell, and R. A. Normann, “A glass/silicon composite intracortical
electrode array,” Ann Biomed Eng, vol. 20, no. 4, pp. 423–437, Jul. 1992.
[139] K. D. Wise, J. B. Angell, and A. Starr, “An Integrated-Circuit Approach to Extracellular
Microelectrodes,” IEEE Transactions on Biomedical Engineering, vol. BME-17, no. 3, pp.
238–247, Jul. 1970.
[140] M. Schwaerzle, P. Elmlinger, O. Paul, and P. Ruther, “Miniaturized tool for optogenetics
based on an LED and an optical fiber interfaced by a silicon housing,” in 2014 36th Annual
International Conference of the IEEE Engineering in Medicine and Biology Society
(EMBC), 2014, pp. 5252–5255.
[141] K. Y. Kwon, H.-M. Lee, M. Ghovanloo, A. Weber, and W. Li, “Design, fabrication, and
packaging of an integrated, wirelessly-powered optrode array for optogenetics application,”
Frontiers in Systems Neuroscience, p. 69, 2015.
[142] N. Grossman et al., “Multi-site optical excitation using ChR2 and micro-LED array,” J.
Neural Eng., vol. 7, no. 1, p. 16004, Feb. 2010.
[143] K. Y. Kwon, B. Sirowatka, W. Li, and A. Weber, “Opto- #x03BC;ECoG array: Transparent
#x03BC;ECoG electrode array and integrated LEDs for optogenetics,” in 2012 IEEE
Biomedical Circuits and Systems Conference (BioCAS), 2012, pp. 164–167.
[144] N. McAlinden et al..., “Thermal and optical characterization of micro-LED probes for in
vivo optogenetic neural stimulation,” Opt Lett, vol. 38, no. 6, pp. 992–994, Mar. 2013.
[145] H. Cao, L. Gu, S. K. Mohanty, and J.-C. Chiao, “An integrated μLED optrode for
optogenetic stimulation and electrical recording,” IEEE Trans Biomed Eng, vol. 60, no. 1,
pp. 225–229, Jan. 2013.
[146] B. Fan et al., “A polycrystalline diamond-based, hybrid neural interfacing probe for
optogenetics,” in 2015 28th IEEE International Conference on Micro Electro Mechanical
Systems (MEMS), 2015, pp. 616–619.
[147] A. Branner, R. B. Stein, and R. A. Normann, “Selective Stimulation of Cat Sciatic Nerve
Using an Array of Varying-Length Microelectrodes,” Journal of Neurophysiology, vol. 85,
no. 4, pp. 1585–1594, Apr. 2001.
[148] A. Branner and R. A. Normann, “A multielectrode array for intrafascicular recording and
stimulation in sciatic nerve of cats,” Brain Research Bulletin, vol. 51, no. 4, pp. 293–306,
Mar. 2000.
[149] P. Degenaar et al., “Individually addressable optoelectronic arrays for optogenetic neural
stimulation,” in 2010 IEEE Biomedical Circuits and Systems Conference (BioCAS), 2010,
pp. 170–173.

111

[150] H. W. Choi, C. W. Jeon, and M. D. Dawson, “Fabrication of matrix-addressable micro-LED
arrays based on a novel etch technique,” Journal of Crystal Growth, vol. 268, no. 3–4, pp.
527–530, Aug. 2004.
[151] O. Yizhar, L. E. Fenno, T. J. Davidson, M. Mogri, and K. Deisseroth, “Optogenetics in
Neural Systems,” Neuron, vol. 71, no. 1, pp. 9–34, Jul. 2011.
[152] K. Y. Kwon, B. Sirowatka, W. Li, and A. Weber, “Opto-uECoG array: Transparent μECoG
electrode array and integrated LEDs for optogenetics,” 2012 IEEE Biomed. Circuits Syst.
Conf. BioCAS, pp. 164–167, 2012.
[153] K. Y. Kwon, B. Sirowatka, A. Weber, and W. Li, “Opto- Array: A Hybrid Neural Interface
With Transparent Electrode Array and Integrated LEDs for Optogenetics,” IEEE
Transactions on Biomedical Circuits and Systems, vol. 7, no. 5, pp. 593–600, Oct. 2013.
[154] M. Schwaerzle, P. Elmlinger, O. Paul, and P. Ruther, “Miniaturized 3 #x00D7;3 optical
fiber array for optogenetics with integrated 460 nm light sources and flexible electrical
interconnection,” in 2015 28th IEEE International Conference on Micro Electro
Mechanical Systems (MEMS), 2015, pp. 162–165.
[155] K. Y. Kwon, A. Weber, and W. Li, “Varying-Length Polymer Microneedle Arrays
Fabricated by Droplet Backside Exposure,” Journal of Microelectromechanical Systems,
vol. 23, no. 6, pp. 1272–1280, Dec. 2014.
[156] H.-M. Lee, K. Y. Kwon, W. Li, and M. Ghovanloo, “A Power-Efficient Switched-Capacitor
Stimulating System for Electrical/Optical Deep Brain Stimulation,” IEEE Journal of SolidState Circuits, vol. 50, no. 1, pp. 360–374, Jan. 2015.
[157] K. Kwon and W. Li, “Integrated multi-LED array with three-dimensional polymer
waveguide for optogenetics,” in 2013 IEEE 26th International Conference on Micro Electro
Mechanical Systems (MEMS), 2013, pp. 1017–1020.
[158] K. Y. Kwon, A. Khomenko, M. Haq, and W. Li, “Integrated slanted microneedle-LED array
for optogenetics,” in 2013 35th Annual International Conference of the IEEE Engineering
in Medicine and Biology Society (EMBC), 2013, pp. 249–252.
[159] K. Y. Kwon, H.-M. Lee, M. Ghovanloo, A. Weber, and W. Li, “A wireless slanted optrode
array with integrated micro leds for optogenetics,” in 2014 IEEE 27th International
Conference on Micro Electro Mechanical Systems (MEMS), 2014, pp. 813–816.
[160] K. Y. Kwon, X. Bi, and W. Li, “Droplet backside exposure for making slanted SU-8
microneedles,” in 2013 35th Annual International Conference of the IEEE Engineering in
Medicine and Biology Society (EMBC), 2013, pp. 767–770.
[161] Keiser, G., Optical Fiber Comm, 4th Edn. New York: McGraw-Hill, 2008.
[162] N. McAlinden et al.., “Thermal and optical characterization of micro-LED probes for in
vivo optogenetic neural stimulation,” Opt Lett, vol. 38, no. 6, pp. 992–994, Mar. 2013.
112

[163] N. McAlinden, E. Gu, M. D. Dawson, S. Sakata, and K. Mathieson, “Optogenetic activation
of neocortical neurons in vivo with a sapphire-based micro-scale LED probe,” Front Neural
Circuits, vol. 9, May 2015.
[164] H. Blatt and R. J. Tracy, Petrology.
[165] A. A. Finnerty and F. R. Boyd, “Evaluation of thermobarometers for garnet peridotites,”
Geochimica et Cosmochimica Acta, vol. 48, no. 1, pp. 15–27, Jan. 1984.
[166] F. P. Bundy, H. T. Hall, H. M. Strong, and R. H. Wentorf, “Man-Made Diamonds,” Nature,
vol. 176, no. 4471, pp. 51–55, Jul. 1955.
[167] P. M. Martineau, S. C. Lawson, A. J. Taylor, S. J. Quinn, D. J. F. Evans, and M. J. Crowder,
“Identification of Synthetic Diamond Grown using Chemical Vapor Deposition (CVD),”
Gems Gemol, vol. 40, pp. 2–25.
[168] J. V. Macpherson, “A practical guide to using boron doped diamond in electrochemical
research,” Phys Chem Chem Phys, vol. 17, no. 5, pp. 2935–2949, Feb. 2015.
[169] S. Ferro, “Synthesis of diamond,” J. Mater. Chem, vol. 12, pp. 2843–2855, 2002.
[170] “Observations on CVD-Grown Synthetic Diamonds: A Review.” [Online]. Available:
https://www.gia.edu/gems-gemology/fall-2016-observations-CVD-grown-syntheticdiamonds-review.
[171] J. E. Butler, Y. A. Mankelevich, A. Cheesman, J. Ma, and M. N. R. Ashfold, “Understanding
the chemical vapor deposition of diamond: recent progress,” J. Phys.: Condens. Matter, vol.
21, no. 36, p. 364201, 2009.
[172] P. W. May and Y. A. Mankelevich, “From Ultrananocrystalline Diamond to Single Crystal
Diamond Growth in Hot Filament and Microwave Plasma-Enhanced CVD Reactors: a
Unified Model for Growth Rates and Grain Sizes,” J. Phys. Chem. C, vol. 112, no. 32, pp.
12432–12441, Aug. 2008.
[173] O. A. Williams, “Nanocrystalline diamond,” Diamond and Related Materials, vol. 20, no.
5–6, pp. 621–640, May 2011.
[174] A. C. Ferrari and J. Robertson, “Interpretation of Raman spectra of disordered and
amorphous carbon,” Phys. Rev. B, vol. 61, no. 20, pp. 14095–14107, May 2000.
[175] M. Bernard, A. Deneuville, and P. Muret, “Non-destructive determination of the boron
concentration of heavily doped metallic diamond thin films from Raman spectroscopy,”
Diamond and Related Materials, vol. 13, no. 2, pp. 282–286, Feb. 2004.
[176]D. S. Knight and W. B. White, “Characterization of diamond films by Raman spectroscopy,”
Journal of Materials Research, vol. 4, no. 2, pp. 385–393, Apr. 1989.

113

[177] M. Hupert et al..., “Conductive diamond thin-films in electrochemistry,” Diamond and
Related Materials, vol. 12, no. 10–11, pp. 1940–1949, Oct. 2003.
[178] “Properties
of
Silicon.”
[Online].
http://www.design.caltech.edu/Research/MEMS/siliconprop.html.

Available:

[179] P. Schiavone, F. Chassat, T. Boudou, E. Promayon, F. Valdivia, and Y. Payan, “In vivo
measurement of human brain elasticity using a light aspiration device,” Medical Image
Analysis, vol. 13, no. 4, pp. 673–678, Aug. 2009.
[180] G. Soza, R. Grosso, C. Nimsky, P. Hastreiter, R. Fahlbusch, and G. Greiner, “Determination
of the elasticity parameters of brain tissue with combined simulation and registration,” Int
J Med Robot, vol. 1, no. 3, pp. 87–95, Sep. 2005.
[181] H.-Y. Chan, D. M. Aslam, J. A. Wiler, and B. Casey, “A novel diamond microprobe for
neuro-hemical and -electrical Recording in neural prosthesis,” Journal of
Microelectromechanical Systems, vol. 18, no. 3, pp. 511–521, Jun. 2009.
[182] A. E. Hess, D. M. Sabens, H. B. Martin, and C. A. Zorman, “Diamond-on-Polymer
Microelectrode Arrays Fabricated Using a Chemical Release Transfer Process,” Journal of
Microelectromechanical Systems, vol. 20, no. 4, pp. 867–875, Aug. 2011.
[183] P. Bergonzo et al., “3D shaped mechanically flexible diamond microelectrode arrays for
eye implant applications: The MEDINAS project,” IRBM, vol. 32, no. 2, pp. 91–94, Apr.
2011.
[184] K. V. Nemani, K. L. Moodie, J. B. Brennick, A. Su, and B. Gimi, “In vitro and in vivo
evaluation of SU-8 biocompatibility,” Materials Science and Engineering: C, vol. 33, no.
7, pp. 4453–4459, Oct. 2013.
[185] E. J. Boyd and D. Uttamchandani, “Measurement of the Anisotropy of Young’s Modulus in
Single-Crystal Silicon,” Journal of Microelectromechanical Systems, vol. 21, no. 1, pp.
243–249, Feb. 2012.
[186] “SU-8 2000 Permanent Epoxy Negative Photoresist Datasheet.”
[187] F. Zhang et al., “Optogenetic interrogation of neural circuits: technology for probing
mammalian brain structures,” Nat. Protocols, vol. 5, no. 3, pp. 439–456, Mar. 2010.
[188] E. M. Schmidt, J. S. Mcintosh, and M. J. Bak, “Long-term implants of Parylene-C coated
microelectrodes,” Med. Biol. Eng. Comput., vol. 26, no. 1, pp. 96–101, Jan. 1988.
[189] K. Y. Kwon, B. Sirowatka, W. Li, and A. Weber, “Opto-uECoG array: Transparent μECoG
electrode array and integrated LEDs for optogenetics,” 2012 IEEE Biomed. Circuits Syst.
Conf. BioCAS, pp. 164–167, 2012.
[190] B. Fan, K. Y. Kwon, A. J. Weber, and W. Li, “An implantable, miniaturized SU-8 optical
probe for optogenetics-based deep brain stimulation,” in 2014 36th Annual International
114

Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2014, pp.
450–453.
[191] Element Six, “ENABLE EXTREME PERFORMANCE PHOTONICS.” [Online].
Available:
http://www.e6.com/wps/wcm/connect/1d1f8101-c317-419f-a310c819ba00b0f9/Element+Six_Optical.pdf?MOD=AJPERES&CACHEID=1d1f8101-c317419f-a310-c819ba00b0f9.
[192] J.-L. Lai, C.-J. Liao, and G.-D. J. Su, “Using an SU-8 Photoresist Structure and Cytochrome
C Thin Film Sensing Material for a Microbolometer,” Sensors, vol. 12, no. 12, pp. 16390–
16403, Nov. 2012.
[193] A. Prejs, S. Wood, R. Pengelly, and W. Pribble, “Thermal analysis and its application to
high power GaN HEMT amplifiers,” in Microwave Symposium Digest, 2009. MTT ’09.
IEEE MTT-S International, 2009, pp. 917–920.
[194] S. Lee, S. Y. Kwon, and H. J. Ham, “Specific Heat Capacity of Gallium Nitride,” Jpn. J.
Appl. Phys., vol. 50, no. 11S, p. 11RG02, Nov. 2011.
[195] SCS
coating
systems,
“SCS
Parylene
property.”
[Online].
http://www.physics.rutgers.edu/~podzorov/parylene%20properties.pdf.

Available:

[196] F. Zhang et al., “Optogenetic interrogation of neural circuits: technology for probing
mammalian brain structures,” Nat Protoc, vol. 5, no. 3, pp. 439–456, Mar. 2010.
[197] E. Stark, T. Koos, and G. Buzsáki, “Diode probes for spatiotemporal optical control of
multiple neurons in freely moving animals,” Journal of Neurophysiology, vol. 108, no. 1,
pp. 349–363, Jul. 2012.
[198] B. Fan, R. Rechenberg, M. F. Becker, and W. Li, “Fabrication of polycrystalline diamond
on a flexible Parylene substrate,” in 2015 Transducers - 2015 18th International Conference
on Solid-State Sensors, Actuators and Microsystems (TRANSDUCERS), 2015, pp. 892–895.
[199] B. Fan, Y. Zhu, R. Rechenberg, M. F. Becker, and W. Li, “A Flexible, Large-Scale
Diamond-Polymer Chemical Sensor For Neurotransmitter Detection,” presented at the
Hilton Head Workshop 2016: A Solid-State Sensors, Actuators and Microsystems
Workshop, Hilton Head Island, SC.
[200] E. M. Schmidt, J. S. Mcintosh, and M. J. Bak, “Long-term implants of Parylene-C coated
microelectrodes,” Med. Biol. Eng. Comput., vol. 26, no. 1, pp. 96–101, Jan. 1988.
[201] H. Chan, “Polycrystalline CVD Diamond Probes for Use in in vivo and in vitro Neural
Studies,” Doctor of Philosophy, Michigan State University, 2008.
[202] Allen J. Bard and Larry R. Faulkner, Electrochemical Methods Fundamentals and
Applications (Second Edition). .

115

[203] B. J. Venton and R. M. M. Wightman, “Psychoanalytical electrochemistry,” Analytical
Chemistry, vol. 75, no. 19, Oct. 2003.

116