ufilh\un
3t i
I :-

11“ .9
3. . .5.
«1

mm

saiua.

 

‘ 331.:
Q: .4 ¢

. a.
at 3322...! in
..L r. .lvaholbxs.
\ A371...»
.I

or

3‘

.2.
lies“:

. , ‘ kn E.
A 2...: .: .. z
. ‘ ... ‘ . ‘ . 1.9.5:».
‘ 2. ..i.

[I

 

This is to certify that the
dissertation entitled

POLYCRYSTALLINE CVD DIAMOND PROBES FOR USE IN
IN VIVO AND IN VITRO NEURAL STUDIES

presented by
HO-YlN CHAN

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Electrical Engineering

9% flaw

Major Professor’s Signature
//- // — 0 8/

Date

MSU is an Affirmative Action/Equal Opportunity Employer

 

POLYCRYSTALLINE CVD DIAMOND PROBES FOR USE IN IN VI V0
AND IN VI T R0 NEURAL STUDIES

By

Ho-yin Chan

A DISSERTATION

Submitted to
Michigan State University
In partial fulfilhnent of the requirements
For the degree of

DOCTOR OF PHILOSOPHY
Electrical Engineering

2008

ABSTRACT

POLYCRYSTALLINE CVD DIAMOND PROBES FOR USE IN IN VI V0
AND IN VITRO NEURAL STUDIES

By

Ho-yin Chan

Microprobes are an effective analysis and stimulation tool for neural studies.
Analysis can take place in one of two modes; the microprobe may be used to detect or
measure chemical indicators within the brain, or, it may be used to detect electrical
signals emanating fi'om firing neurons.

A polycrystalline diamond based probe provides many potential advantages over

probes of other materials. Carbon sp3 bonded diamond exhibits transparency and is a

good insulator due to its large band gap (5.5 eV). Diamond is also mechanically versatile

due its Young’s modulus of ~1011 Pa. A boron doped diamond electrode is

advantageous when compared to other electrode materials as it is resistive to fouling and
chemically inert. Developing a sound design and fabrication process for the diamond
probe has proven to be problematic.

In this research, the design, microfabrication and testing of a novel polycrystalline
diamond (poly-C) based microprobe, for possible applications in neural prosthesis, are
developed and presented for the first time. The associated poly-C micromachining
technologies are also developed and discussed in depth. Two types of poly-C probes are

realized. One is designed for electrical and the other for electrochemical measurements.

The probes utilize undoped poly—C with a resistivity in the range of 105 Qcm as a

supporting material, which has a Young’s modulus in the range of 400 — 1,000 GPa and

is biocompatible. Boron doped poly-C with a resistivity in the range of 10-3 Gem is used

as an electrode material, which provides a chemically stable surface for both chemical

and electrical detections in neural studies. The poly-C electrode capacitance is

approximately 87 uF/cmz, which is small as compared to other metal electrodes. The

measured water potential window of the poly-C electrode spans across negative and
positive electrode potentials and typical has a total value of 2.2 V in 1M KCl. The
smallest detectable concentration of norepinephrine (a neurotransmitter) was on the order
of 10 nM. The poly-C probe has also been successfully implanted in the auditory cortex
area of guinea pig brain for in viva neural studies. The recorded signal amplitude was 30-

40 uV and had a duration of lms.

ACKNOWLEDGMENTS

I would like to thank my whole family for their patience, understanding and
support during this study. I owe my greatest thanks to Skye Chan for her hard work and

sacrifice.

I would like to thank my advisor, Dr. Dean Aslam, for his encouragement,
guidance and financial support throughout this research. I also thank my committee
members, Dr. Donnie Reinhard, Dr. Tim Hogan, Dr. George Abela and especially Dr
Kensall Wise for the advice and support. Thanks go to the members of our research group
for their help and discussion in particular, Zongliang Cao, Jing Lu and Aixia Shao.
Thanks also go to the faculties, stafi's and fiiends of the WIMS center at University of
Michigan who gave me great help for the probe fabrication, especially, Dr. Ying Yao,

James Wiler and Brendan Casey.

TABLE OF CONTENT

LIST OF TABLES ....................................................................................... viii
LIST OF FIGURES ........................................................................................ ix
Chapter 1 Research Motivation and Goals ..................................................... 1
1.1 Introduction ............................................................................ ' ......... 1
1.2 Microelectrodes/Microprobes ......................................................... 2
1.3 Objective of this Work ..................................................................... 5
1.4 Dissertation Outline ......................................................................... 7
Chapter 2 Microprobe/Microelectrode Technologies ..................................... 8
2.1 Introduction ..................................................................................... 8
2.2 Cochlear Probes ............................................................................... 9
2.3 , Neural Probes ................................................................................ 10
2.4 Probes for Drug and Chemical Delivery .......................................... 14
2.5 Probes used in Electrochemical Analysis/Detections ..................... 17
2.6 Summary ........................................................................................ 20
Chapter 3 Monitoring Neural Activities Using Microelectrodes ................. 21
3.1 Introduction ................................................................................... 21
3.2 Electrode Model ............................................................................ 22
3.2.1 Electrical Double Layer ................................................................... 23
3.2.2 Electrical Model for Electrode/ Electrolyte ..................................... 26
3.3 Techniques in Monitoring Neural Activities ................................... 31
3.3.1 Electrical Recording of Action Potential ......................................... 31

V

3.3.2
3.4

Neurochemical Detection of Neurotransmitters ............................ 35

Summary ........................................................................................ 38

Chapter 4 Poly-C Micromachining Technologies Relevant to Probe

Fabrication ..................................................................................................... 39
4.1 Introduction ................................................................................... 39
4.2 Properties of Polycrystalline Diamond ........................................... 40
4.2.1 Physical and Chemical Properties of Diamond ............................... 40
4.2.2 Biocompatibility of Diamond .......................................................... 42
4.2.3 Electrochemical Properties of Diamond ......................................... 43
4.3 Poly-C Technologies ....................................................................... 45
4.3.1 Diamond Nucleation ...................................................................... 45
4.3.2 Diamond Growth and Doping ........................................................ 47
4.3.3 Inadvertent SiOz Etching ................................................................ 49
4.3.4 Conductivity in Undoped Poly-C Films ............................................ 53
4.4 Diamond Etching ............................................................................ 57
4.5 Summary ........................................................................................ 61
Chapter 5 Poly-C Probe Technology ............................................................ 62
5.1 introduction ................................................................................... 62
5.2 Design ............................................................................................ 63
5.2.1 Electrical Neural Probe ................................................................... 65
5.2.2 Electrochemical Neural Probe ........................................................ 67
5.3 Fabrication ..................................................................................... 69
5.3.1 Electrical Neural Probe ................................................................... 69
5.3.2 Electrochemical Neural Probe ........................................................ 74

vi

5.4 Probe Mounting and Packaging ..................................................... 78

5.5 Summary ........................................................................................ 79
Chapter 6 Characterizations and Applications of Poly-C Probes ................. 80
6.1 introduction ................................................................................... 80
6.2 Poly-C Site Characterizations .......................................................... 81
6.2.1 Electrical Impedance Spectroscopy ................................................ 81
6.2.2 Cyclic Voltammetry ........................................................................ 88
6.2.3 EC Probes ....................................................................................... 94
6.3 Probe Applications ......................................................................... 97
6.3.1 Electrical Detection ........................................................................ 97
6.3.2 Electrochemical Detection ........................................................... 104
6.4 Summary ...................................................................................... 107
Chapter 7 Summary and Recommendations ............................................... 108
7.1 Summary of Contributions ........................................................... 108
7.2 Future Research Areas ................................................................. 109
APPENDIX ................................................................................................. 111
BIBLIOGRAPHY ....................................................................................... 120

vii

LIST OF TABLES

Table 2-1. Summaries of previous work on neural probes. ............................................... 12
Table 2-2. Summary of the previous work on microprobes for drug delivery. ................. 16
Table 3-1. Concentrations of ions for mammalian skeletal muscle [86]. .......................... 34
Table 4-1. Properties of silicon and diamond [95]. ........................................................... 41

Table 4-2. Comparison of different seeding methods on seeding density, substrate
material and surface morphology [128] ............................................................................. 46
Table 4-3 Conditions for diamond growth using MPCVD ............................................... 48

Table 4-4. Etch rates of different types of silicon dioxide (no annealing was done after
growth) ............................................................................................................................... 52

Table 4-5. Effect of boron out-diffusion from the neighborhood wafer on the diamond
resistivity. .......................................................................................................................... 56

Table 4-6. Effect of process history of the MPCVD chamber on the resistivity of undoped

diamond. ............................................................................................................................ 57
Table 4-7 Parameters for diamond etch ............................................................................. 58
Table 5-1. A list of the nine masks used in this project ..................................................... 63
Table 5-2. The six masks used in the fabrication of BL probes. ....................................... 66
Table 5-3 The eight masks used in the fabrication of electrochemical neural probes ....... 69
Table 5-4. AgCl formation on Ag electrodes using FeCl3. ............................................... 78
Table 6-1. Fitted parameters for electrode/electrolyte model ............................................ 88

viii

LIST OF FIGURES

Figure 1-1. An overview of poly-C neural probe technology for neural recording. ........... 6
Figure 2-1. The fabricated cochlear probes. ...................................................................... 10
Figure 2-2. Filtered neural data recorded from a single channel of the probe in a rat’s

cortex [42] .......................................................................................................................... 11
Figure 2-3. Different types of neural probes: (a) Planar (b) Three dimensional. .............. 12
Figure 2-4. Fabrication process and images of the microneedles ...................................... 15
Figure 2-5. Neural probes with stimulation sites and microchannels ................................ 16
Figure 2-6. Carbon fiber electrode for electrochemical detection [69]. ............................ 18

Figure 2-7. Cyclic voltammetry of the electrodes in 0.5M H2SO4 (a, b) polycrystalline

diamond, (c) platinum and ((1) highly oriented pyrographite [7 5] ..................................... 18
Figure 2-8. A diamond microprobe with tungsten connected wire [81]. .......................... 19
Figure 3-1. A typical electrochemical cell which has two electrodes. .............................. 23
Figure 3-2. A systematic representation of the double layer structure. ............................. 25
Figure 3-3. Randles’ equivalent circuit for the electrode/electrolyte interface. ................ 27
Figure 3-4. A typical Nyquist plot for the Randles model in both high and low

frequencies. ........................................................................................................................ 28
Figure 3-5. Low frequency model of an electrode/electrolyte interface. .......................... 29
Figure 3—6. High frequency model of an electrode/electrolyte interface ........................... 30
Figure 3-7. A typical example of an action potential [87] ................................................. 32
Figure 3-8. A triangular potential waveform used in cyclic voltammetry [69] ................. 37
Figure 3-9. A typical cyclic voltarnmogram [69]. ............................................................. 38

Figure 4-1. Cyclic voltammetries of diamond films with different C/I-I source gas

mixtures in 0.1M HClO4[114]. .......................................................................................... 44
Figure 4-2. Seed distribution on (a) a non-treated and (b) an RCA-treated Si surface. ....47
Figure 4-3. Schematic of the MPCVD system used in this project. .................................. 48
Figure 4-4. (a) An SEM image of a poly-C surface and (b) its Raman spectrum. ............ 49

Figure 4-5. The relationships between the seeding density and time to as a frmction of
seeding spin rate. ............................................................................................................... 52

Figure 4-6. Diamond seeding density on an RCA treated Si surface with different spin
rates (the samples have been undergone 40 mins diamond growth). ................................ 53

Figure 4-7. Schematics of an experimental setup for showing boron-out diffusion from Si
substrate in the MPCVD chamber. .................................................................................... 55

Figure 4-8. SEM images of the diamond surface at different stages of the etching process.
The bar shown on the lower right corner of each image is 2 pm. ..................................... 59

Figure 4-9. A rough Si surface which has been over etched caused by CF4 in the diamond
etching recipe ..................................................................................................................... 60

Figure 4-10. The diamond sample patterned using 2 step etching. Oxide layer is well

protected and its roughness is only 2-3 nm. ...................................................................... 60
Figure 5-1. Mask layout for different types of diamond probes ........................................ 64
Figure 5-2. Design of the EL probe. ............................................................ . ...................... 66
Figure 5-3. Design of the EC probe ................................................................................... 68
Figure 5-4. Fabrication process of all-diamond probes. ................................................... 71

Figure 5-5. Released all-diamond probe on a US penny. The highly doped Si back end
was removed in this probe. ................................................................................................ 72

Figure 5-6. An all—diamond probe was bent in the out-of-plane direction with a pair of

tweezers (bending angle was larger than 180°). ................................................................ 73
Figure 5-7. Fabrication process of the EC probe. .............................................................. 75
Figure 5-8. (a) A fabricated EC probe, which has an (b) Au counter electrodes, (c)

Ag/AgCl reference electrodes and a (d) poly-C working electrodes ................................. 76
Figure 5-9. Probe mounting. .............................................................................................. 79
Figure 6-1. Modified Randles model of the electrode/electrolyte interface [86]. ............. 83
Figure 6-2. Experimental setup for the EIS measurements. .............................................. 84

Figure 6-3. Impedance spectrum for a diamond electrode with diameter of 30 pm, (a)
Nyquist plot of the real and imaginary part of the measured impedance, (b) Bode plot. ..87

Figure 64. Experimental setup for cyclic voltammetry. ................................................... 89

Figure 6-5. Voltarnmograms of Au, Pt, IrOx and diamond electrodes in (a) 1MKC1, (b)
Krebs, and (0) PBS. ........................................................................................................... 91

Figure 6-6. (a) Cyclic voltammetry current-voltage curve for the diamond probe in 1M
KCl. (b) The CV of the diamond probe in 1 mM Fe(CN)r;'4 in 1M KC]. The scan rate was
100 mV/s. The reference and counter electrodes are Ag/AgCl and platinum, respectively.
........................................................................................................................................... 93

Figure 6-7. A comparison on the cyclic voltammograms in 1M KCl. .............................. 95
Figure 6-8. Cyclic voltammograms of the EC probe in 1mM Fe(CN)6'3/'4 + 1M KCl ...... 97
Figure 6-9. Acute recording of neural activity using the diamond probe .......................... 99

Figure 6-10. (a) Recorded neural activity in guinea pig’s auditory cortex using the
diamond probe with an electrode area of IOOumZ, (b) A recorded neural spike. ............ 100

Figure 6-11. A recorded in vivo signal from an all-diamond EL probe. Site separation is
~500 um ........................................................................................................................... 102

xi

Figure 6-12. Histograms of the recorded neural signals shown in previous figure. ........ 103

Figure 6-13. (a) Measured cyclic voltammograms of the poly-C probe for different
concentrations of NE in Krebs buffer solution using a scan rate of 100 mV/s. (b) It shows
the background current in Krebs solution ........................................................................ 105

Figure 6-14. Detection of NE and DA with same concentration ..................................... 106

xii

Chapter 1

Research Motivation and Goals

1.1 Introduction

The study of cellular function and response has been used for decades to better
understand the principles of human body function in order to prevent, treat or cure
diseases in an effective way. Among all the parts of a human body, the human brain is the
core and is also the most complex as it contains an incredibly large amount of neurons.
These neurons control body movement, human emotion, reaction, drinking and etc. Due
to its complexity, it contains many long-standing mysteries which are of interest to
neuroscientists [1]. For instance, it is not yet known how information is encoded in the
brain, what emotions really are, or how memories are stored and retrieved. The
exploration and understanding of the brain require the study of neural activities and how
these signals correlate to human reactions. Thus, it is highly desired to have a way to
interact with the human brain in order to extract information from it. As technology
continues to advance, the capabilities and sensitivity of cellular studies have greatly
improved. Current techniques allow for the examination of a single cell’s electrical
characteristics either in vivo (within the body) or in vitro (outside of the body) using glass
micropipettes or wire microelectrodes [2]. These tools are able to detect electrical signals

and detect chemicals electrochemically.

1.2 Microelectrodes/Microprobes

One of the key components in neural prosthetic systems is the
microelectrode/microprobe, which interfaces with neurons in order to record electrical
signals or detect the presence of a particular chemical. Microprobes can also be used for
stimulation. In early in vivo studies, metal wires coated with insulating materials [3] were
used as electrodes, which are associated with limited spatial resolution. In the late 19608,
microprobes with multiple recording or stimulation sites were fabricated with high
reproducibility utilizing lithographic thin-film techniques [3]. With the development of
Micro-Electro-Mechanical Systems (MEMS) technologies, sophisticated microprobes
with electrode arrays have been developed and are now widely used in neural activity
studies [4][5][6][7], drug delivery [8][9], and cochlear implants [10][11]. 3D microprobe
arrays have been fabricated, rnicroassembled and integrated with a hybrid application
specific integrated circuit (ASIC) chip. [12]. Most of today’s probes are made out of
silicon due to the fact that silicon based bulk micromachining technology is well
developed. The silicon microprobes are either fabricated on bulk silicon or silicon on
insulator (SOI) wafers. In the former case, highly boron doped silicon is used as the etch
stop in the ethylenediamene pyrocatecol (EDP) or Tetramethylammonium hydroxide
(TMAH) etching process to release the probe structure. The oxide layer of the silicon on
insulator (SOI) wafer is used as the etch stop in deep reactive ion etching (DRIE) for
probe release. Due to the flexibility issues associated with Si probes, recently, researchers
have been studying the use of other materials in probe fabrication. For example, parylene
C [l3][l4][15], polyimide [l6] and SU-8 [17] are often used as materials for the probe

substrate or for coating.

The most important component on the microprobes is the metal site, which is used for
signal recording and electrical stimulation. Commonly used site materials include
platinum [18], gold [19], titanium nitride [20] and iridium oxide [21]. The sites should be
inert and chemically stable in the working environment-especially in ion- and protein-
rich conditions. Otherwise, the corresponding background noise would be too high, or the
target signal could be suppressed due to site blocking caused by residues formed or
absorbed on the site surface. For stimulation, the electrode should have a large charge
storage capacity as well as a large charge delivery capacity in order to deliver a large
amount of charge to a small geometric area. Sites for recording, on the other hand, should

be associated with relatively low background noise in order to achieve a high signal to

noise ratio. The exposed site area should be large (< 400 umz) for collecting a greater

number of signals in the case of electrical recording. Also, thermal noise is reduced
because of the low impedance of larger sites [22], but the selectivity (i.e. the ability to
discriminate between signals from different neurons) would be low. Thus, the signal
received by the probe would be from multiple units but not from isolated unit, which is
undesirable. On the contrary, in electrochemical recording, the site area as well as the
double layer capacitance must be small in order to minimize non-Faradic currents. These
currents are usually caused by the charging of the double layer structure formed at the
electrode/electrolyte interface. Therefore, in order to have a good signal to noise ratio
(>3), the electrode should have low capacitance. The electrode should also exhibit a wide
water potential window (potential range in which there is no current caused by oxygen or
hydrogen evolution). Carbon fiber [23] is one of the most promising materials for

electrochemical detection as it has most of the characteristics mentioned above. However,

one of the main disadvantages of carbon fiber is its fouling which requires constant

surface polishing.

In this study, polycrystalline diamond (poly-C) is chosen as the core material in the

microprobe due to its unique properties. The presence of sp3 C-C bonds in the diamond

lattice leads to its unique mechanical properties (large Young’s modulus, ~1011 Pa).

Poly-C also has a large band gap (5.5 eV) which makes it a good insulating and optically
transparent material [24]. The optical transparency of the diamond substrate is important
for in vitro experiment because it allows the electrode position on the probe to be easily
located under a microscope. In addition, the comparatively wide potential window (the

reported values range from 1.4 to 4 V) in an aqueous environment [25][26][27][28], low

double layer capacitance (ranging from 5 to 40 uF/cmz) [29], chemical inertness and

stability, resistance to fouling of boron doped diamond make it an excellent site material
for microprobes [25]. The biocompatibility of diamond surfaces has been studied
intensively based on protein adsorption on chemical-vapor-deposited (CVD) diamond [30]
and diamond-liked carbon (DLC), and cellular responses to each [31]. It is found that
CVD diamond absorbed and denatured relatively small amounts of fibrinogen which is
commonly used as a biocompatibility indicator, and no cytotoxicity was seen, which is an
inflammatory reaction and has adverse effects on cells [32]. In electrochemistry,
researchers have extensively studied boron-doped diamond as an electrode material
[33][34]. The previously existing boron doped poly-C electrodes, which are used in

neural studies [35][36], were fabricated by depositing diamond on the tip of a wire. These

electrodes display limited spatial resolution for studying brain activities. Recently, poly-C

sensors and electrodes were integrated in Si cochlear microprobes for the first time [ l 1].

1.3 Objective of this Work

An overview of the work done in this project is shown in Figure 1-1. The focus of
the current work is to integrate polycrystalline diamond thin films into microprobe
fabrication to develop a novel all-diamond probe for neural and biomedical applications.
The fabricated all-diamond probes with diamond electrodes would be used for chemical
detection (as conductive diamond has a wide potential window) and electrical sensing in
biological environments (as diamond chemical inertness is advantageous). For such
measurements, a neural probe having precisely positioned arrays of electrodes with
micrometer and nanometer dimensions, are needed.

The work addressed the following issues:

1) Development of the process for integrating undoped and doped diamond as the
probe substrate, and interconnects and electrode/ site, respectively.

2) Design and fabrication of all-diamond neural probes using silicon existing surface
micromachining technologies.

3) Design and fabrication of diamond based probes which include sets of three
electrode configurations for electrochemical analysis

4) Application of the fabricated diamond probes in electrochemical and electrical
measurements in neural studies.

Successful accomplishment of these goals has to lead to the following unique

contributions:

1) First fabrication of diamond probes with diamond electrodes, substrate and
interconnects in a microprobe structure.

2) First fabrication of diamond probes with counter, working and reference
electrodes on a single probe

3) First application of a diamond probe in neural recording.

 

Polycrystalline CVD Diamond Based Neural Probe for Neural Prosthesis I
l

i

 

P l —C Fl
0 y I m Poly-C Neural Probe

Technology

 

U Unique contribution

   
       

 

 

 

 

 

 

   

 

 

 

   

 

_ ‘ I In vivo neuro-electrical In vitro neuro- U I
_i recording using poly-C probes electrochemical recording I
Enahllnatechnology - using poly-C probes .

 

 

, . _ . 7 . .l I
i First time reported I

Figure 1-1. An overview of poly-C neural probe technology for neural recording.

1.4 Dissertation Outline

This thesis presents the development and characterization of poly-C neural
microprobes for neural prosthesis. A literature review on the current
microprobe/microelectrode technologies is discussed in Chapter 2. The work on
microprobes, which include cochlear, neural and electrochemical probes, are presented
and summarized. The key component of these probes is the electrode/site, which is used
for stimulation and recording. Thus, an electrode model for an electrode/electrolyte
interface is presented in chapter 3, which aids in understanding reactions taking place at
the electrode surface. As the focus of this research is to utilize poly-C electrodes for in
vivo and in vitro recording, methods useful in neural recording are also discussed in
chapter 3.

All of the work performed in this study is presented in chapters 4, 5 and 6. In
chapter 4, poly-C micromachining technologies relevant to probe fabrication are
presented, which include diamond nucleation, growth, doping and patterning. In chapter
5, the design and fabrication of two types of novel poly-C probes are presented. The two
types include electrical (EL) and electrochemical (EC) probes. Then, the characterization
and applications of these two types of poly-C probes are presented in chapter 6. Finally, a
summary on the work done and also recommendations for future related research are

included in chapter 7.

Chapter 2

Microprobe/Microelectrode Technologies

2.1 Introduction

Microprobes, sometimes called microelectrodes, have been widely used in
chemical and biomedical applications. Depending on the particular applications,
microprobes can be involved in chemical or electrical sensing, electrical stimulation or
chemical delivery. Microprobes are crucial in any of these aspects because they interface
with biological entities (in biochemical applications) and target species (in
electrochemistry applications). They are most commonly used in the recording of
electrical signals which are associated with reduction-oxidation (redox) reactions
involving analytes or action potentials in the brain. Some of them even have the ability to
deliver chemicals in order to stimulate cells or aid in healing applications. Most of
today’s microprobes are made of silicon as silicon lends itself to on-probe CMOS
development. Also, processes for micromachining silicon are well-developed. The silicon
probes are fabricated using either an ethylenediamene pyrocatecol (EDP) silicon etching
process with a deep boron etch stop [37] or a DRIE etch of SOI wafers [3 8]. Such probes
are widely used in biomedical applications for neural simulation, neural activity
recording and drug delivery. Microprobes used in electrochemical aspects are simply
metal wires coated with a layer of insulator. In this chapter, a brief introduction to various

types of microprobes used in both biomedical applications and electrochemistry is

8

presented. It also highlights the structure of various probes and the materials used for the

probes.

2.2 Cochlear Probes

Researchers at the University of Michigan have been actively researching the
fabrication and design of many different types of microprobes. Prof. K. D. Wise’s group
has demonstrated the use of microprobes for cochlear prosthesis [37][39]. In cochlear
prosthesis, the probe is implanted in the ear in an effort to recover the hearing ability of
deaf people. The microprobe, shown in Figure 2-1, was fabricated on a silicon wafer and
includes stimulation sites and position sensors. The position sensors on the probe retrieve
the shape of the inserted microprobe and the electrodes are used for sending electrical
pulses to nerves inside the ear. The probe shaft is defined by boron doped silicon which
acted as an etch stop in an EPD process. The probe was around 1.6cm long and 6pm
thick. Dr. Yuxing Tang, a member of our group, has integrated polycrystalline diamond
(poly-C) strain sensors on the probes which have a higher gauge factor (GF) as compared
to those made of polysilicon [40]. However, these probes have some limitations. The

silicon substrate is conductive and has to be passive with oxide/nitride/oxide stack.
However, it is known that Si02 will be hydrated over time [41]. On the other hand, the
stiffness of the probe limits its flexibility which can cause problems during insertion.

Also in response to the problems outlined above, C. Pang et al have encapsulated probes

in parylene so as to increase their flexibility [42].

 

Figure 2-1. The fabricated cochlear probes.

2.3 Neural Probes

The design and fabrication processes of neural probes must focus on minimization
of insertion damage, reduction of tissue reactive responses and minimization of
perturbing cellular environment. Therefore, the probe size, material construction and the
technique used to package the probe are important. Current probes used in neural
prosthesis contain several metal sites for recording neural signals. An example is shown
in Figure 2-2. There are numerous groups that have been working on the design and
fabrication of neural probes [43-56]. G. T. A. Kovacs et al has briefly summarized the
early work on rnicroelectrode arrays for neural recording and stimulation [57].

Table 2-1 summarizes the recent work on neural probes. It is found that silicon is
often chosen probe substrate material over the past 10 years. Recently, researchers have
been looking for other probe materials and fabrication techniques which could overcome
some of the limitations associated with silicon probes. For example, probes made from

polyimide can be very flexible which is critical during insertion. On the other hand, using

10

electroplating technology, probes can be fabricated such that its thickness varies along its
length; this increases the stiffness of the probe so that the probe will not break easily. It is
seen that, depending on different applications, probes can be in different forms and make
use of different substrate materials. However, the materials used in fabricating electrodes
are similar in most all probes. Commonly used materials include platinum, gold and

iridium because of their associated chemical stability and inertness.

 

 

Amplified signal (V)

 

 

r r r r r
2.55 2.7 2.75 2.8 2.85 2.9 2.95

 

Time (s)

Figure 2-2. Filtered neural data recorded from a single channel of the probe in a rat’s
cortex [42].
A typical design of a neural probe array is shown in
Figure 2-3 (a). It consists of several shanks each having its own recording site
located at the tip portion of the probe. These arrays provide the ability to record neural
signals from various points.
Figure 2-3 (b) shows another design of a three dimensional probe array. This probe

array provides 3-D spatial resolution.

11

 

Figure 2-3. Different types of neural probes: (a) Planar (b) Three dimensional.

Table 2-1. Summaries of previous work on neural probes.

 

 

 

 

 

Researchers Substrate Electrode Probe Applications
material(s) material(s) dimensions

D. J. Anderson Silicon Ir =3mm, Electrical

et al (U. of stimulation
W=150-250

Mich, 1989)
um

[21]

P. K. Campbell et Silicon Au L=l.5mm Electrical

al (U. of Utah, . recording,

1991) [47] (multi probes)

J. H. M. Put et al Silicon Pt L=lmm Electrical

(Netherlands, stimulation

1991) [50] (multi probes) (10pmx505um) T=45um

G. T. A Kovacs Silicon Cr/Ir L=lmm Electrical

et al (Stanford, _ sensing of the

1996) [51] (multI-probes) W=30ilm cerebral cortex

of an
T—25um anesthetized
rat

12

Table 2-1 (continue)

 

 

 

 

 

 

 

 

 

L. P. Lee et al Silicon Pt T=25 um Electrical
simulation and
(U CB, 2000) [52] (multi-probe) recording
S. J. Kim et la Silicon Cr/Au L=1mm Recording
action
$931.63, 2000) (multi probes) W=100um potentials in rat
T=30um “me"
P. Norlin et a1 SOI Ti/Ir and Ti/Pt L=l.6mm Electrical
(Germay, 2003) ' sensing
[46] (mulit-prbes) (10 pm x 10 W=50um
rim)
H. Fujita et al Polyimide Ti/Al L=1.2mm Action
potentials
[43] PTObCS) rat’s visual
T=5um cortex
C. Pang et al Silicon/parylene Ti/Pt L=4mm Electrical
C Neural
(Caltech, 2005) W=45um recording
[41] (multi probes)
T=50um
D. R. Kipke et al Silicon Pt, IrOx L=2mm Rat brain
(U .of Mich. , ' cortical neural
2005) [44] (multi probes) (D=20 um) =60um signal
[58][59] recording and
neurochemical
sensing
(dopamine)
J. W. Judy et al Electroplated Ni Cr/Pt L=22mm Deep-brain
(UCLA, 2005) (single probe) simulation
[53] T=50 urn
W. C. Tang et al Silicon Ir L=750 um Electrical
(UCI, 2005) [54] neural
(millti probes) recording
S. H. Cho et a1 SU-8 Au L=5mm, Electrical
(U. of Texas at . recording
Dallas, 2006) (Single probe) (D=15 um) W=240 um

[60]

13

Table 2-1 (continue)

 

 

 

Y. Kato et a1 Parylene C Au L=3.7 mm, Electrical
recording and
(U. of Tokyo, (Single probe) (D=20 um) W= 554 um chemical
2006) [61] delivery
J. C. Sanchez et - Polyimide L=l.6mm Electrical
a1 (U. of Florida, coated recording
2006) [62] tungsten/Au
(D=50 urn)

P. Ruther et al Silicon Pt L=4mm or Electrical

8mm recording
(U. of Freiburg,
2007) [63]

 

2.4 Probes for Drug and Chemical Delivery

In neural activity studies, it is sometimes not just interacting with neurons
electrically but also chemically stimulating neurons in order to understand the complex
reactions happened inside the brain. Therefore, there are probes which are specially
outfitted with embedded microchannels for chemicals delivery. These probes are often
called microneedles or microcapillaries [64][65][66][67] and are summarized in Table
2-2. Figure 2-4 shows an example of a fabricated microneedle, which is made of boron
doped silicon (12pm) and silicon nitride (l um) [65]. Silicon dioxide (8pm) was used as
the sacrificial layer which was covered by silicon nitride. Numerous etch holes were
opened on the nitride surface for sacrificial layer etching. Then, the etch holes were
sealed by depositing silicon nitride (1 .Sum) again.

A sophisticated microprobe reported by J. Chen et a1 [64] contains both simulation

sites (made of Titanium/Iridium) and microchannels (made of silicon) as shown in Figure

14

2-5. As mentioned earlier in this section, the delivery capability of microprobes allows
scientists to study complex chemical reaction inside neural systems. This probe was used
to deliver both kainic acid (a nemal stimulant) and y-aminobutyric acid (a neural

depressant) into guinea pig brains in viva.

Inflation-dailies: lsrrrm
ISilimdiuido INrura.

 

 

 

 

Figure 2-4. Fabrication process and images of the microneedles.

15

 

Figure 2-5. Neural probes with stimulation sites and microchannels.

Table 2-2. Summary of the previous work on microprobes for drug delivery.

 

 

 

 

Researchers Substrate Channel dimensions Probe dimensions Applications
material(s) and materials
J. Chen et a1 (U. Silicon W=10-32 urn L=4mm deliver kainic and
of Mich, 1994) _ ' '
H=15 llm y ammobutyrrc
[64] acid
Silicon
A. B. Frazier ct - Palladium L=3mm Injection
a1 (Georgia
H=20 um
Tech, 1998) [67]
W=40 pm
A. P. Pisano et al Silicon Silicon and nitride L=l-6mm Injection

(UCB, 1999 and

1993) [66]

W=50um

H=9um

 

16

2.5 Probes used in Electrochemical Analysis/Detections

Probes or electrodes are used in an electrochemical manner for analytical purposes
such as environments monitoring, industrial quality control and biomedical analysis.
These applications require electron transfer between electrolytes and electrodes through
redox reaction processes [68][69]. Therefore, electrodes must be resistant to fouling. The
designs of probes or electrodes used in electrochemistry are simple as compared to
MEMS-based microprobes. They are generally thin metal wires coated with insulating
materials such as glass. As mentioned in [69], glassy carbon and carbon fiber are the
most common insulating materials used as shown in Figure 2-6. Carbon electrodes are
desirable because of their low cost, simplicity of preparation, the possibility of achieving
large surface area and a relatively wide potential window of water and their high degree
of stability as compared to metal materials [70]. However, electrode fouling limits their
long term stability and leads to a need to frequently polish or dispose of the electrode
after only a few uses. Thus, diamond, another form of carbon has drawn the attention of
researchers because of its extreme chemical stability [71] and resistance to fouling
[72][73]. In addition, diamond electrodes exhibit a comparatively wide potential window
([74][75]) for oxygen and hydrogen evolutions (-1.25V to +2.3V vs NHE) as shown in
Figure 2-7. This allows us to analyze species (e.g., persulfate ion [76]) which cannot be

analyzed using conventional metal electrodes.

17

 

200nm

Figure 2-6. Carbon fiber electrode for electrochemical detection [69].

 

 

 

 

Figure 2-7. Cyclic voltammetry of the electrodes in 0.5M H2SO4 (a, b) polycrystalline
diamond, (c) platinum and (d) highly oriented pyrographite [75].

18

Several groups have reported using boron-doped diamond as an electrode material
for electrochemical detection [77], waste treatment [69], in viva detection of serotonin
[78][79] and adenosine [80] and neurodynamic studies [81]. An example of a boron-
doped diamond probe used in electrochemistry is shown in Figure 2-8. A tungsten wire
was encapsulated by a quartz tube and boron-doped diamond was growth at the tube end
using microwave plasma assisted chemical vapor deposition (MPCVD). The diameter of
the tube end was around 35 um which is useful for precision in viva detection. Owing to
the small effective area of the exposed electrode, the iR drop in the electrolyte and the
changing capacitive current of the electrode-electrolyte interface, which is undesirable,

are small.

   

   

500x

\-

Figure 2-8. A diamond microprobe with tungsten connected wire [81].

19

2.6 Summary

This chapter reviews previous work done on various types of microprobes. Most of

the microprobes are made of silicon. The commonly used electrode materials are Pt, Au,

TiN and IrOx, which can be used either for stimulation or recording. In electrochemistry,

the commonly used electrodes are carbon fiber and diamond coated microelectrode. It is
found that diamond electrodes are advantageous in chemical detection and analysis
because of their chemical inertness and stability. However, to date, there is no integration
of diamond electrodes with microprobe fabrication with the exception diamond
piezoresistors. The integration of diamond electrodes on a probe will lead to promising

applications in both in viva and in vitro neural recording.

20

Chapter 3
Monitoring Neural Activities Using

Microelectrodes

3.1 Introduction

In the first part of this chapter, a review of the electrochemical processes taking
place at electrode surfaces when an electrode is immersed in an electrolyte is presented.
The behavior of the electrochemical system can be represented by an eqmvalent circuit
consisting of a combination of resistors and capacitors. The model parameters in the
circuit can be extracted indirectly by Electrical Impedance SpectroscOpy (BIS), which is
commonly used to characterize the electrode behavior for small perturbations and high
frequency excitations.

In the second part of this chapter, techniques used for monitoring neural activities
are discussed. These techniques can be classified into two categories: non-invasive and
invasive recording [82]. Non-invasive recording modalities, including functional
magnetic resonance imaging (fMRI), electroencephalograms (EEG) and
magnetoencephalograrns (MEG), can provide most of current data from the brain.
However, the data returned from these techniques lack the spatial resolution required to

understand physiological processes at the cellular level. On the contrary, the development

21

of invasive recording allows us to retrieve information down to the cellular level (i.e.
local field potential or even action potential generated by a single neuron). In addition,
recent advances in micromachining technology have made real-time multichannel
monitoring of cellular and molecular dynamics possible both in vitra and in viva.
Numerous research groups have been working on cell cultures in vitra using devices
which have single or multi planar electrodes [83][84][85]. These devices have the
capability to measure the action potential and impedance of the target cells.
Microprobes, on the other hand, allow us to retrieve data from the brain in viva. In
summary, in this chapter, two possible ways of neural recording are discussed which

include recording of action potential and neurochemical release.

3.2 Electrode Model

In an electrochemical system, or any measurement system involving the use of
electrodes, there must be at least two electrodes in order to complete the circuit as shown
in Figure 3-1. Charge can then be transferred by electrons (in the electrode and external
circuit) and by ions (in the electrolyte). At the electrode/electrolyte interface, there are
two types of processes which can occur. One of them involves charge transfer across the
electrode/electrolyte interface; this type of process is called a Faradaic process. The other
type, a non-Faradaic process, does not involve charge transfer across the interface. For
example, adsorption or desorption phenomena. As there are different processes taking

place at the electrode/electrolyte interface, an electrical model of the situation should be

22

developed in order to gain a thorough understanding of the interface and relevant

 

 

 

processes.
V
Indicator Counter
electrode
electrode

 

 

 

 

 

Figure 3-1. A typical electrochemical cell which has two electrodes.

3.2.1 Electrical Double Layer

When an electrode is immersed in an electrolyte, an electrical double layer must
be formed before any charge may move across the electrode/electrolyte interface. The
structure of the electrical double layer formed is shown in Figure 3-2. For a given applied

potential, and before any charge can move across the interface, the total charge on the

electrode (-qM) and in the solution (qS) must be equal but opposite in magnitude in order

to maintain charge neutrality. This region of charge separation is called space charge

region.

23

The region next to the electrode surface (moving away from the electrode) is
made of solvent molecules (e. g. water) or adsorbed species. The water dipoles orient
themselves according to an electric field developed across the interface caused by the
double layer formed. This layer of dipoles is called the inner Helmholtz plane (IHP). The
next layer towards the electrolyte is called the outer Helmholtz plane (OHP), or,
sometimes, the plane of closest approach (PCA). This plane defines the nearest locus of
centers of solvated ions. These ions, called nonspecifically adsorbed ions, are those
which undergo electron exchange with the electrode. The distance of this plane from the
electrode surface is ~10A. These nonspecifically adsorbed ions are distributed and extend
into the bulk solution. This layer is called the diffuse layer. The thickness of the electrical
double layer grows as the time. A typical thickness in the time scale (reaction time) of is
is ~30um. Since the cations are mostly concentrated at the OHP, the potential drop occurs
mostly across the electrode and OHP. Therefore, when this double layer structure is
formed, a large potential field develops at the interface which drives the electron transfer

across it.

24

Electrolyte

Metal __
electrode -

 

 

IHP OHP Diffuse space charge

@@

0 ® 0 8699

Speciallyadsorbed ”mam?” Unsolvatedt—i ion (D
(arrow pants to +

charge an dipole) Solvated H '00

anion

Figure 3-2. A systematic representation of the double layer structure.

As discussed, a space charge region is formed at the electrode/electrolyte interface.
As the water dipoles act as a dielectric, the interface should, in theory, behave like a
capacitor. However, unlike a simple capacitor, the capacitance of the double layer

structure is voltage dependent because ions move in response to an applied potential. The

total interfacial capacitance C] of the double layer structure is given by

25

1 1 1
_=_+__

(1)
Cr CH CGCV)

where CH is the Hehnholtz capacitance and CG is the Gouy-Chapman capacitance [68].

The Gouy-Chapman capacitance C0 is a voltage dependent term. The idea is that the ions

in the diffuse layer are pulled toward the electrode surface when the applied voltage
increases. Therefore, the separation of the charge on the electrode and in the electrolyte

decreases which leads to an increase in the capacitance.

3.2.2 Electrical Model for Electrode/Electrolyte

The electrochemical cell behaves as an impedance. It is therefore possible to
represent its behavior by an equivalent circuit. The most commonly used
electrode/electrolyte model was developed by Randles in 1973 [86] and is shown in
Figure 3-3. All current passes through a resistance called the spreading resistance Rs. It
includes any resistance in the external circuit. For instance, cable resistance, solution
resistance and electrode resistance. Then, the parallel elements are introduced because the
total current through the interface is the sum of both Faradaic (i.e. current from redox

reactions) and non-faradaic processes (i.e. double layer charging). The double layer

structure can be modeled by a capacitor Cd], as discussed in the last section, which has an

impedance Zc. The F aradaic process is represented by a resistance R, (called charge

transfer resistance). Another impedance term W, called the Warburg impedance, is related

to the mass transfer resistance. This is due to the fact that solvated ions have to be

26

brought from the bulk to the interface before they can undergo a redox reaction. The

impedance Wis defined by

 

(2)

  
 

Electrode

 

Figure 3-3. Randles’ equivalent circuit for the electrode/electrolyte interface.

The total impedance Z for the Randles model is

zc(Rt+W)
W+Zc+Rt'

 

Z=RS+ (3)

The variation of total impedance Z with frequency is often of interest. This
relationship can be represented in two ways: Bode plots and Nyquist plots. In a Bode
plot, long] and phase 6 are plotted against log[ca]. In the Nyquist plot, the imaginary part
Z ” of the total impedance Z is plotted against the real part Z’ of Z for different values of

ca. The total impedance Z can be fitted to the Nyquist plot. In this way, the parameters of

27

the Randles model can be obtained. A typical Nyquist plot of an electrode/electrolyte

interface is shown in Figure 3—4.

Z» A

I
Kinetic control Mass transfer control

Decreasingw

  

O
l
O

l
h
I
c

l
O
l
a

I
.

l
o
l
O

l
u
I
I

l
O
l
I

     

 

 

Rs I (R5 + R; IZIIRS +IRt) Z: I

Figure 3-4. A typical Nyquist plot for the Randles model in both high and low
frequencies.

The curve can be divided into two regions: low (mass transfer limted) and high

frequencies (kinetic limited). At low frequencies, the impedance 20 is large enough that it

can be approximated as an open circuit. Therefore, the circuit model can be simplified to
the form shown in Figure 3-5 for low frequencies.

The low fiequency impedance is then:

Z=RS+Rt+ 1 1 (4)

(joon)§

 

28

The real and imaginary parts are

.1.
2

Z' = R, + R, + (2mcw)’ (5)

Z" = —(2(r)Cw)-% (6)

Eliminating a), we get,

u I
—Z = Z — (Rs + Rt) (7)
This corresponds to a straight line with a slope of l (with —Z " plotted against Z’)

and an x-intercept of R S + R, in the Nyquist plot.

 

Electrolyte Electrode
r—’VV\I ‘VVV—I—
Rs Rt

Figure 3-5. Low frequency model of an electrode/electrolyte interface.

At high frequencies, the Warburg impedance is relatively low as compared to Rt so

that it can be considered as a short circuit. The simplified circuit for this case is shown in
Figure 3-6.

The high frequency impedance is then:

29

Rt

 

 

 

Z = R3 + 1+ijthl
The real and imaginary parts are
I Rt
Z = R5 + 1+w2C§le
n _ mCdrRi
— 1+m2C(2nR%

Eliminating m, we get,

(2’ — (R, + iRtDZ + Z"2 = (i Rt)2

(8)

(9)

(10)

(11)

This corresponds to a half circle centered at Z ’= RS+Rt/2 with radius Rt/2 in the

Nyquist plot.

 

 

Electrolyte * ’\/\/\r Electrode
"MAE—I R |_.
R nt ,

5 II
Cdl

Figure 3-6. High frequency model of an electrode/electrolyte interface.

30

3.3 Techniques in Monitoring Neural Activities

Microprobes are one of the most commonly used devices in studying real-time
neural activities. Because information is transmitted and propagates both electrically and
chemically between neurons, the microprobes either record extracellular neural activities
or electrochemically monitor neurotransmitter synaptic overflow. The microprobes are
often brought in close proximity to the cell for extracellular recordings.

The neural signal pathway starts at the stimulated neuron’s axon hillock, where an
action potential is generated. Then, this electrical action potential is transmitted along the
axon until it reaches a junction called the synapse. This junction is where neurons
communicate with each other. When the action potential reaches the end of axon, the
axon will release certain types of neurotransmitters (e. g. dopamine and/or serotonin).
Then, the dendrite of another neuron will collect the released neurotransmitters and fire
an action potential. Repeating this process, information can propagate through the neural
network. Because of the nature of the information transmission in the brain, the study of
neural activities can include both electrical recording, where action potentials are

measured and chemical sensing, where neurotransmitters are detected.

3.3.1 Electrical Recording of Action Potential

The action potential is the electrical signal of interest in the electrical recording.
The action potential is a spike of electrical energy that travels along the membrane of a
cell, and carries and transmits information along the cell and between different cells.
Thus, studying brain activity can be achieved by monitoring and processing the action

potentials occurring within cells.

31

 
    

Membrane potential (mV)

 

 

 

 

beginstoenter
can 13 5 KWhamatsciou.
{' Natchamelsreset
-70 \f—
Threahotdot 6 ExtraK+outside
excitation dittuaesamy

Figure 3-7. A typical example of an action potential [87].

An action potential can be created by many types of cells, but are used most
extensively by the nervous system for communication between neurons. They are also
used to transmit information from neurons to other body tissues. Action potentials
differ based on what type of cell they emanate from. However, a typical action
potential is shown in Figure 3-7. An action potential is caused by an imbalance of

ions moving into and out of a cell across its membrane. This membrane is semi-

32

permeable to different ions existing inside and outside of the cell and has many

different channels to control the in-flow and out-flow of certain ions. Typically, the

+ + + -
common ions found inside and outside of the cell are Na , K , Ca2 and Cl . An

example of the concentrations of these ions is shown in Table 3-1.

The sequences of events that underlie the action potential are shown in Figure 3-7.
If there is no excitatory stimulus, the cell is at rest with cell potential ~ -70 mV
(measured as the intracellular potential with respect to extracellular potential). This
potential is created by the ion concentration difference inside and outside the cell. At

this point, the cell is said to be polarized. Upon receiving a stimulus, the cell becomes

+ +
excited and Na ion channels on the membrane will be opened to allow Na ions to
. + . +
enter the cell. At the same trrne, the voltage-gated K ron channel opens and K

+
moves out of the cell due to its concentration gradient. The influx of Na ion causes

the increase in the cell potential to ~ +20 mV. Now, the cell is said to be depolarized.

+
By the time the cell potential reaches +50 mV, the Na channel will start to close and

+
reduce the inflow of sodium ions. The K ions continue to diffuse out of the cell and

lower the cell potential below the threshold. Then, the voltage-gated channel will

+
close and prevent K ions from leaving the cell. The cell is now re-polarized. The

cell enters a period called the refractory period, at which time it cannot receive any

+
stimulus. The rest potential will then be restored by the active transport of K into the

33

+
cell and Na out of the cell. The actual mechanism is more complex than that

discussed here and can be found in most biology textbooks. However, the basic
concept is shown to give a brief understanding of action potentials. After the process
outlined here, this action potential propagates along the cell membrane and will be
transferred to the other cells.

Microelectrodes/microprobes, with low input impedances, can be used to
capacitively record the action potential, which, as explained earlier, is due to charge
movement. The probe must be connected to a high input impedance amplifier and
necessary noise filters [88]. Most often, local field potential, which is the sum of
action potentials from a number of neurons, is recorded--rather than an individual
action potential. As there are numerous electrodes on a given probe, a reliable data
acquisition system is desired which can be found in the market [89]. Researchers, on
the other hand, have also developed implantable rnicrosystems used to record signals
received on a probe. These systems can also transmit data wirelessly and provide

more flexibility in neural studies [90].

Table 3-1. Concentrations of ions for mammalian skeletal muscle [86].

 

 

 

 

 

Ion Extracellular Intracellular
concentration concentration (mM)
(M

+
Na 145 12
+
K 4 155
+ -7
Ca2 1'5 10 M
Cl- 123 4.2

 

 

 

 

 

34

3.3.2 Neurochemical Detection of Neurotransmitters

Neurochemical detection is a method used to measure F aradic currents resulting
fiom the oxidation or reduction of neurochemical species released from neurons. These
neurotransmitters are highly localized inside the synapse and their concentrations
decrease with increasing distance from the synapse. Therefore, the microprobe has to be
placed as closely as possible to the synapse. (<50um) [91]. The current is caused by
electron transfer to and from an electrode across the electrode/electrolyte interface. This
measurement can be performed using electrochemical techniques such as cyclic
voltarmnetry and amperometry. These techniques require three electrodes: a working
electrode (WE), reference electrode (RE) and counter (CE) electrode. The oxidation or
reduction reaction of the redox couple is forced to occur at the WE by applying sufficient
voltage between itself and the CE. The RE is used as the reference for measuring the
potential of the WE and does not take any current from the cell. It should have low
impedance and be capable of maintaining a constant potential throughout the
measurement process. The most commonly used RE is Ag/AgCl. It is stable and able to
maintain a constant potential independent of which electrolyte is under examination. By
controlling the potential of the WE (as referenced to the RE) and simultaneously
measuring the amount of current passing through it, information concerning the charge
transfer processes that occur at the electrode-electrolyte interface is obtained.

Considering the following redox system,

0+ne‘ —->R

35

where O and R are the oxidized and reduced forms of the redox couple,
respectively.

This reaction will occur in a potential region where the electron transfer is
thermodynamically or kinetically favorable. For systems controlled by the laws of
thermodynamics, the potential of the electroactive species at the surface Co(0,t) and

CR(0,t) vary according to the Nemst equation

 

 

+ 2.3RT C0(0,t)

bg
nF CR (0, t) (12>

E=E°

o . . . . .
where E rs the standard potential for the redox reaction, R rs the umversal gas

-1 -l
constant (8.314JK mol ), T is the temperature in Kelvin, n is the number of electrons

transferred in the reaction, and F is the Faraday constant (96487 coulombs). If the

potential E is smaller than E0, the oxidized form tends to be reduced. But if the potential

E is larger than E0, the reduced form tends to be oxidized. Therefore, current flows due to

the electron transfer between the electrode and redox couple. Every redox couple has its
own standard potential which can be found in electrochemistry textbooks.

Cyclic voltammetry is the most widely used technique for acquiring qualitative and
quantitative information about electrochemical reactions. A triangular potential
waveform as shown in Figure 3-8 is applied to the working electrode. The amplitude of
the final and inital potentials and the scan rate can be adjusted depending on the

particular application. The voltage-current plot, called a cyclic voltammogram, is shown

36

Figure 3-9. For the forward scan, there is no significant current measured until the

potential reaches E0. As the applied voltage reaches E0, the oxidized form of the redox

couple starts to be reduced and thus, current flow fi'om the WE to the CE increases as the
voltage increases. The peak current represents the voltage where the rate of oxidation
equals the rate of redox couple reaching the surface. For the reverse scan, it is simply the
reverse of the process--the reduced form of the redox couple is reduced to its oxidized

form. As mentioned earlier, different redox couples have different standard potentials,

E0. Since these standard potentials can be obtained from a cyclic voltammogram, cyclic

voltammetry is a powerful tool in indentifying redox couples. For example, it can be used

in detecting neurotransmitters in viva, which are indicative of neural activity.

- Cycrer ..l

Efinal

 

3} Reverse
“E scan
a *1 4
0 Forward
0' Scan
E SWITCHING

. initial POTENTI A] .
Time

Figure 3-8. A triangular potential waveform used in cyclic voltammetry [69].

37

 

cathodic
O
i

  
 

forward scan

 

Current

anodic

 

 

04—}!
Potential

 

Figure 3-9. A typical cyclic voltammograrn [69].

3.4 Summary

This chapter reviews the basic electrode/electrolyte model which aids in the
understanding of reactions taking place on an electrode surface. It is found that a double
layer is formed at the electrode/electrolyte interface. The interface can be modeled by an
equivalent circuit, which is a combination of resistors and capacitors. In addition, two
techniques used for monitoring neural activities are discussed. These include the

monitoring of action potentials and the detection of the release of neurotransmitters.

38

Chapter 4

Poly-C Micromachining Technologies

Relevant to Probe Fabrication

4.1 Introduction

CVD poly-C is becoming a more and more attractive material for MEMS
applications due to its unique properties. MEMS applications, for example sensors [92],

packaging [93] and electronics [94], have already been fabricated using diamond.

However, the difficulty in growing sp3 bonded carbon, which usually requires high

temperature as a growth condition, is currently causing a delay in the development of a
reliable and economical diamond micro-fabrication technology that is compatible with
conventional microsystems/MEMS technologies. An alternative to diamond is diamond-
like carbon (DLC) which can be deposited at room temperature. Unfortunately, the
quality of the film (i.e., its mechanical strength) is not as strong as that of diamond as
there is no long-range order of the position of atoms.

In this study, CVD poly-C is used as the core material for probe fabrication. In
order to integrate CVD poly-C into Si MEMS, several important issues must first be
addressed. These issues include seeding, grth and etching of CVD diamond. In this

chapter, physical, chemical and biomedical properties of CVD poly-C films and its

39

associated micromachining technologies are presented. Combined, these properties and

technologies have made the development of a diamond neural probe possible.

4.2 Properties of Polycrystalline Diamond

4.2.1 Physical and Chemical Properties of Diamond

Some of the mechanical and electrical properties of poly-C diamond are listed in
Table 4-1. It is well known that diamond is the hardest and strongest material found in

nature. Its mechanical strength comes as a result of the diamond lattice; each carbon atom
is covalently bonded with its four neighboring atoms by hybridized sp3 atomic orbitals.
Thus, it is a good candidate for high temperature (due to its large band gap), high speed
(owing to its high carrier mobility) and high power (because of its high thermal
conductivity) applications. Compared with silicon, diamond has a much higher Young’s
modulus which indicates its stiffness. In addition to Young’s modulus, fracture toughness
is also a good indicator of structural integrity. Fracture toughness is a property which

describes the ability of a material containing a crack to resist fracture. If a material has a

large value of fracture toughness, it will likely undergo ductile fracture. Typical values

2
for facture toughness are ~150 MPa ml/ for ductile metals, 25 MPa 1111/2 for brittle

1/2
ones and 1-10 MPa m for glasses and brittle polymers. Diamond exhibits a high

fracture toughness as compared to silicon (Table 4-1).

40

Diamond also exhibits good chemical properties. It is found that diamond films are
stable during anodic polarization in acidic fluoride, alkaline, acidic chloride and neutral
chloride media [71][114]. Also, diamond is resistant to fouling compared to graphite as
mentioned in the previous section. For all of these reasons, diamond film is an excellent

material for use in harsh environments.

Table 4-1. Properties of silicon and diamond [95].

 

 

 

 

 

Properties Silicon Poly-C Diamond
Density (g/cml) 2.329 3.52
Melting point (K) 1412 --
Young’s Modulus (GPa) 130-180 800-1 180
1/2 12.5 [95][97] 3-6 [98][99]

Fracture toughness (MPa rn )

 

 

 

 

Poisson’s ratio 0.22-0.24 0.07-0.148
Band gap (eV) 1.12 5.45
Carrier Mobility

2
Electron (cm N s) 1450 23

500 10-1000
2

Hole (cm /Vs)
Dielectric constant 11.7 6.5

41

Table 4-1 (Continue)

 

 

 

 

Breakdown voltage (MV/cm) 0.3 7-0.5 0.1-1

Intrinsic resistivity (0cm) 1x103 ”1016

Thermal conductivity (W/ch) 1.5 4-20
2.6 2.6

Thermal expansion coefficient (10-6

/°C)

 

4.2.2 Biocompatibility of Diamond

Diamond has been called the “biomaterial of the 21St Century” due to its excellent
biocompatibility [100]. With recent advances in industrial synthesis of diamond and
diamond-like carbon film bringing prices down significantly, researchers are increasingly
willing to experiment with diamond coatings for medical implants, for example.
Examples of diamond use in implants includes artificial heart valves [101], prosthetic
devices [102] and joint replacements [103]. Diamond is advantageous in these situations
because of its excellent wear resistance and high resistance to blood platelet adhesion.
Diamond has also been used as the material for biosensors [78][104][105]. Diamond-like
carbon (DLC) coated orthopedic pins implanted in sheep also demonstrate low diamond

bioactivity [106]. Even more, CVD diamond coatings for artificial joints are said to have

“low immunoreactivitYT 1 07].

42

The first study of protein adsorption on CVD diamond was performed in 1995, and
focused on the absorption of fibrinogen [108]. Fibrinogen is the major protein involved in
the initiation of coagulation and inflammation including fibrosis around implanted
biomaterials. It has been found that the CVD diamond absorbed and denatured relatively
small amounts of fibrinogen. Also, in vitra and in viva experiments showed less cell
adhesion to the diamond surface and activation on the surface of CVD diamond when
compared to both titaninum and 316 stainless steel. In addition, CVD diamond has also
shown that it has no cytotoxic effect, no hemolytic effect or any complementary
activation [100][109]. More detailed studies on diamond biocompatibility can be found in

[110][111][112][113].

Although diamond is reported to be a biocompatible material, it is important to point
out that biocompatibility is highly application-specific. Some applications require a non-
adhesive interface, while other applications will require complete tissue integration with

the implanted devices.

4.2.3 Electrochemical Properties of Diamond

Boron-doped diamond is found to have a very wide potential window (the anodic
evolution potential is especially high) as compared to other materials, which is
advantageous in the electroanalysis of substances which have a similar or higher
oxidation potential than oxygen. A high potential for oxygen evolution permits
electrogeneration of oxidants such as hydroxyl radicals from water. The potential window
of the diamond depends heavily on its growth conditions. The relationship between

potential window and the C/H ratio during diamond growth is shown in Figure 4-1

43

[1 l4][1 15]. The potential window decreases and the background current magnitude

increases with increasing C/I-I ratio.

 

 

 

 

 

 

 

 

 

 

:/
3
E [I
a !’
I10 11A
“4600‘ i A i: ‘ ‘ ‘1o‘oo‘ A nzo‘oo‘
Potential (mV vs. AgrAgcr)

Figure 4-1. Cyclic voltammetries of diamond fihns with different C/H source gas
mixtures in 0.1M HClO4[114].

In addition, when immersed in an electrolyte, diamond also displays low
background current, low double layer capacitance, chemical and mechanical stability and
resistance to fouling. It also shows the lack of a surface oxide, and a controllable surface
termination [100][116]. These characteristics have led to the application of diamond in
electrochemical sensors, for electroanalysis, for electrochemical synthesis, and for anodic

destruction of organic wastes. Other properties such as low iR drop and high sensitivity

are discussed in detail in [117].

44

4.3 Poly-C Technologies

4.3.1 Diamond Nucleation

The CVD growth of poly-C on non-diamond substrates such as Si requires a
pretreatment step to generate seeds (or nuclei) on the substrate. One of the following are
typically used for the pretreatment: surface abrasion [118][119][120] followed by the
sonication of diamond powder loaded solution [121], bias enhanced nucleation (BEN)
[122][123][124], spinning of diamond-powder-loaded photoresist (DPR) [125], spraying
of diamond-loaded fluids [126][127] or diamond-loaded water (DW) [128]. Table 4-2
summarizes these pretreatments and their effects on seeding.

The DW pretreatment has been used in previous studies covering the integration of

poly-C piezoresistors into Si cochlear probes [128]. During the probe fabrication process,

poly-C diamond piezoresistors are fabricated on top of SiOz. This surface is hydrophilic.

In contrast to growth on SiOz, in the development of poly-C probes, poly-C films are

sometimes grown on directly on an Si substrate--which is a hydrophobic surface. A
surface treatment is required for more uniform seeding on the Si substrate. In this study,
RCA-clean has been used to treat the Si surface. This cleaning process increases the

contact angle between a water droplet and the Si surface. In fact, this process gives the

smallest contact angle 50 (i.e. more hydrophilic surface) among all other surface

treatments listed in [129]. Therefore, RCA-clean was performed on the Si sample before

nucleation throughout this study.

45

Table 4-2. Comparison of different seeding methods on seeding density, substrate

material and surface morphology [128].

 

 

 

 

 

 

. Sonication
sea”? BEN DPR Spray/paint DW
e o /Abrasion
Seeding 10 11 8 11 8 11
Density Up to 10 Up to 10 ~10 Up to 10 10 ~10
Hydrophilic
Substrate . MOSI Conductlve . MOSI . MOSI surface Ilkc
S l . . drelectrrc & dielectric & dielectrrc & (SiO &
e ectrvrty metal Si 01' Metal, metal metal 2
Si3N4)
scratch
Surface surface, not No l l
Affection good for thin NO damage damage NO ge NO ge
frlm
Uniform
Uniformity Not Uniform & & Not Uniform &
& uniform repeatable on repeatable uniform repeatable on
Controllability whole wafer on whole whole wafer
wafer

 

 

 

 

 

 

A comparison of nucleation on a non-treated and RCA-treated Si surface is shown
in Figure 4-2. The DW is supplied by Advanced Abrasives Corporation and was diluted
to contain 0.5% diamond powder in water. The average seed size was lum. Then, the
DW was applied to the substrate with a 500 rpm pre-spin for 10 sec followed by a 2000

rpm phase for 30 sec. As shown in the Figure 4-2, the RCA-treated surface shows a better

46

 

and more uniform seeding density as compared to the non-treated surface. This is due to

the fact that the Si surface becomes hydrophilic after RCA cleaning. The resulting

0 -2
seeding density is in the range of 109 — 101 seeds/cm .

 

(a) (b)
Figure 4-2. Seed distribution on (a) a non-treated and (b) an RCA-treated Si surface.

4.3.2 Diamond Growth and Doping

Microwave Plasma Chemical Vapor Deposition (MPCVD) was used for poly-C
growth in this project. A schematic of the MPCVD system is shown in Figure 4-3. This
MPCVD (Wavernat TM MPDR 313EHP) system has a chamber (D = 9”) and quartz bell
jar (D = 5”). A 2.45 GHz, 5 kW microwave power supply (Sairem TM, GMP60KSM) and
the large chamber size mentioned above ensured the uniformity of the plasma in the
vicinity of wafers up to 2”x 2”. To ensure this, a sample wafer was heated by the plasma
and its temperature was monitored by a pyrometer outside the chamber. Typical

deposition conditions used in this study are listed in Table 4-3.

47

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Probe W gurde I Microwave
I Generator
/’ -------------- \
1’ \‘l l
r
I Microwave ' '
. I Cavity E 0: Power
I Pyrometer : “NW
1
r
r
r

 

Gas inlet

 

g

o 3

110'
r;

0'

§

i-

" I
;{

——-------~-
-~-

g...

Roughing Punp

 

 

 

 

 

 

 

 

Figure 4-3. Schematic of the MPCVD system used in this project.

Table 4-3 Conditions for diamond growth using MPCVD

 

 

 

 

Parameter(s) Value(s)
Growth Temperature 750 0C

Growth Pressure 40 Torr

Gases Used for Growth Methane: 1 seem

Hydrogen: 100 sccm

 

Dopant Used for B—doped Poly-C Trimethylboron diluted in Hz

(0.1% of B(CH3)3 by volume in H2): 5-10

 

 

sccm
Microwave Power Used for Growth 2.3 kW, 2.4 GHz
Growth Rate 0.10 - 0.15 nm/h

 

 

Poly-C Film Thickness; Undoped/Doped 3 um/ 0.5 pm

 

 

48

 

Diamond film was in-situ boron-doped by introducing trimethylboron (TMB),
(B(CH3)3), gas diluted in hydrogen. The boron doping level, according to the data

presented in [128], is in the range of 3000 - 7000 ppm (B/C ratio in gas phase) and
resistivities are in the range of 10-1 - 10.3 Item. The average grain size on the surface of
poly-C film is approximately 2 pm for a growth time of 20 h. The film quality, as

indicated by the sp3/sp2 C-C bond ratio, is very good as indicated by a sharp Raman peak

at 1332 cm.1 as shown in Figure 4—4.

 

 

 

 

3 1332cm-1
'5 r
- C
3
S
700 1100 1500 1900
Raman Shift (cm-1) V

 

 

 

(a) (b)

Figure 4-4. (a) An SEM image of a poly-C surface and (b) its Raman spectrum.

4.3.3 Inadvertent SiOz Etching

As shown in Table 4-3, the growth temperature is high enough that it can melt
most of the metals which have low melting temperature (e.g. Al). In addition, the high

power of the hydrogen plasma is another factor that limits the use of metals in the

49

MPCVD system. Metals (e.g. Au and Cu) can be etched or deformed mechanically
during the growth of diamond. Therefore, the harsh environment in the MPCVD chamber
during diamond growth limits the integration of poly-C into conventional Si or
semiconductor micromachining technology. Also, it is found in this work that SiOz,

which is commonly used in semiconductor fabrication, can be etched significantly during

diamond growth. The hydrogen plasma continues to etch Si02 through the holes between

the diamond seeds until the SiOz is fully covered by poly-C. This can cause problems if

the SiOz layer is used as an insulating layer. Since the etching of SiOz happens only

when there are pin-holes in the poly-C film, increasing seeding density can help in
minimizing the effect of this etching issue. Increased seeding density can be achieved by
either lowering the spin rate for nucleation or by increasing the concentration of DW.

However, with this method, pin-holes still exist before a continuous diamond film is

formed, and the SiOz can still be etched. Thus, it is important to characterize the etch rate

of SiOz in the MPCVD chamber so that the amount of S102 etched becomes a known

factor and can be taken into consideration in any process design. The etch rate of

different types of commonly used SiOz layers during diamond growth at 750°C are listed

in Table 4-4.

Besides the etch rate of SiOz, the time tc for obtaining a pin-hole free poly-C film

should be known in order to calculate the total thickness of oxide etched. This time tc is

50

plotted as a function of spin speed during the DW seeding as shown in Figure 4-5. The
resulting density is also plotted. It is noted that as the spin speed increases, the seeding

density decreases leading to a longer time necessary for obtaining a pinhole free poly-C

film. The total thickness t of SiOz is then calculated by the following equation,

t = tC x etch rate.

51

Table 4-4. Etch rates of different types of silicon dioxide (no annealing was done after

growth)

 

Types of silicon
oxide lgrth
temperature

Growth temperature
0
( C)

Etch rate
750°C

 

Thermal /
1000°C

1100

1.0-3.0

 

LPCVD /
450°C

450

1.4-3.1

 

PECVD /
380°C

 

 

380

 

2.8-4.4

 

 

H N N
U1 0 U1
1 l 1

Seeding Density (x109 lcmz)
S

 

 

 

 

5 — I Seeding density
50
0 9 Time required for having a pinhole free poly-C film 0
0 1000 2000 3000 4000 5000
Spin Rate (rpm)

Figure 4-5. The relationships between the seeding density and time tc as a

firnction of seeding spin rate.

52

, 100 ‘ '1 ‘200‘0‘rpmif

u“

4000rpm

 

Figure 4-6. Diamond seeding density on an RCA treated Si surface with different spin
rates (the samples have been undergone 40 mins diamond growth).

4.3.4 Conductivity in Undoped Poly-C Films

Ideally, undoped diamond is a large band gap material with a value of 5.5 eV. It
should not be conducting in any case. However, it is also a fact that polycrystalline
diamond can show conductivities (low resistivities) due to many factors. For example,
growth conditions, grain boundaries, impurities and etc. In this project, undoped poly-C
diamond is used as the probe substrate and the conducting interconnects are fabricated on

top of the poly-C. In this study, as-grown diamond exhibited finite conductive, which,

53

given its purpose in the probe, is undesirable. Therefore, it is important to find the cause
of the conductivity of the as-grown undoped poly-C.

There are several possible causes for the undesired conductivity. First, impurities
in the chamber may incorporate into the diamond film during its growth. Obviously,
hydrogen is one of the main sources which contributes to the conductivity of diamond.
The surface of an as-grown poly-C film is hydrogen-terminated. Because of the

hydrogen termination, the surface becomes conductive after exposure to the atmosphere.

This conductive surface layer can be removed by annealing in oxygen at 600°C or by

oxygen plasma etching [l30][13l]. In addition to its role in increasing surface
conductivity, hydrogen may incorporate deep into the poly-C film and cause bulk

conduction. The hydrogen trapped in the poly-C film can be released by annealing under

nitrogen at 8000C for an hour [132]. Also, the intrinsic defects in poly-C films can be a

. . . . . 2
factor Wthh contributes to the conductrvrty of undoped diamond, for example, the sp C-

C bond. These defects appear mostly on the grain boundaries [133][134] and
unfortunately, cannot be avoided in poly-C film as grain boundaries must always exist.

In this project, it is found that there are other factors affecting the conductivity of
undoped diamond. They are in the form of impurities (i.e., boron) which originate from
Si substrates and the MPCVD chamber and can be incorporated into the diamond fihn
during growth. The effect of boron impurities from Si substrates can be verified by the
following experiment. Two samples (a & b) were placed next to each other in the

MPCVD chamber as shown in Figure 4-7. Sample a is the control sample, which has a

layer of SiOz (2pm) on top of Si in each experiment. The layer on top of sample b is

54

varied in different experiments as listed in Table 4-5. The resistivity (after removing the
conductive surface layer) of sample a was monitored for each experiment. It is found that
the resistivity of sample a is related to the previous treatment on sample b. If sample b is

highly boron doped and not covered with $02, the resistivity of sample a is lower. But,

sample a’s resistivity seems unaffected b sample b is n—type Si and even without SiOz

layer on top. It is likely that the boron in sample b got diffused out and incorporate into

sample a. Also, the SiOz can help in blocking the boron out diffusion.

Quartz dome \

Sample a Sample b

   

Sample holder"

Figure 4-7. Schematics of an experimental setup for showing boron-out diffusion from Si
substrate in the MPCVD chamber.

55

Table 4-5. Effect of boron out-diffusion from the neighborhood wafer on the diamond
resistivity.

 

 

 

 

 

 

Resistivity of sample a (plasma Types of wafers put in the
etched for 10mins and annealed same growth (sample b)
at 7500C for 10mins)
+
~105 flcm Sioqp Si
.H.
~104 0cm SiOZ/p Si
+
~103 0cm p Si
++
~103 chn p Si
5 - .
~10 Dem n S1

 

 

 

 

Another factor is the impurities coming from the chamber. In this experiment,
sample a and b were put in the MPCVD chamber for diamond growth separately. Poly-C
film was grown on sample a first. It is followed by undoped diamond on sample b
immediately. It is observed that if doped diamond is grown on sample a, the resistivity of
sample b (undoped diamond) will be lower. This is due to the fact that boron atoms are
either stick on the dome or the sample holder and ruin the next sample. Therefore, it is
not recommended to have a single system doing both doped and undoped diamond
growth. Or, the chamber has to be cleaned using hydrogen or more effectively oxygen

plasma before each run.

56

Table 4-6. Effect of process history of the MPCVD chamber on the resistivity of undoped
diamond.

 

Sample a’s history Sample b’s resistivity

 

Undoped diamond 20hrs ~105 Dem

 

Doped diamond 10hrs (1.5% TMB) ~101 (2cm

 

Doped diamond 10hrs (10% TMB) 401 mm

 

 

 

 

4.4 Diamond Etching

Although the reactive ion etching of diamond has been studied intensively
[135][l36][l37], poly-C dry etching used in the neural probe fabrication in the current

work has led to some problems. The recipe used for the diamond etch is shown in Table

4-7. The problem comes as a result of the CF4 in the etching gases--Si or SiOz will be

etched in any etching gases which contain fluorine. In the poly-C probe fabrication, poly-

C film is either grown on top of the Si or SiOz surface. Due to these facts, Si or SiOz will

be inadvertently etched during poly-C etching.

An example is shown in Figure 4-8. This figure is a group of SEM micrographs
taken at different stages of poly-C dry etching. As the etching of poly-C continues to
reach the seeding depth of the poly-C film, holes appear in the poly-C fihn as seen in

Figure 4-8 (c). At this stage of the etching process, if the etching gases contain fluorine,

etching of the underlying substrate (Si or SiOz) takes place leaving the substrate surface

very rough. Also, it is well known that micro columns (due to micromasking) would be

57

formed in the absence of fluorine in the etching gases [138], which is due to the

sputtering of Al by 02 plasma. The showering of sputtered Al particles on the surface

being etched leads to micromasking.
A two step etching process was developed to eliminate the problems of
micromasking and the substrate etching [139]. Two step etching process has been

proposed in [140]. The etching process produces a very smooth etched diamond surface.

However, it does not take into the consideration of protecting Si or SiOz underneath

when the diamond is fully etched. In this project, diamond layers will be fully etched and

the Si or SiOz underneath have to be protected. Therefore, a new two step etching

process is developed. First, poly-C is etched in the presence of only CF4 until a very thin

poly-C film is left on the surface. Then, the final etching stage is performed in pure

oxygen. This process leads to a smooth Si or SiOz surface.

Table 4-7 Parameters for diamond etch

 

 

 

 

Parameter(s) Value(s)
Power 300 W
Pressure 50 Torr
Gases 02: 40 sccm

Tetrafluoromethane (CF4): 1 sccm

 

Bias 293V

 

Etch rate 1.2 [rm/h

 

 

 

 

58

   

(c) Poly-C surface after 2b of etchmg when 0.6 pm of poly-Cremams; pinholes can be
observed and the S102 layer starts to be etched.

Perspective \ 1c

 

- g, ., 24‘“ '..

a. .9 ‘ y ._ .
(d) Poly-C surface after 3b of etching when the poly-C is fully etched and a rough
Si02 surface is left due to the presence of fluoride in the etching gases.

Figure 4-8. SEM images of the diamond surface at different stages of the etching process.

The bar shown on the lower right corner of each image is 2 pm.

59

Perspective View

 

Figure 4-9. A rough Si surface which has been over etched caused by CF4 in the diamond
etching recipe.

I

.7,
at,

1)"

- ‘. Diamond J . '

a .a.
\ .\

 

Figure 4-10. The diamond sample patterned using 2 step etching. Oxide layer is well
protected and its roughness is only 2-3 11m.

60

4.5 Summary

In this chapter, poly-C micromachining technologies are developed and discussed
in detail. These technologies are specially developed in order to address the problems

associated with poly-C probe fabrication and also diamond MEMS. Examples include

diamond nucleation on Si and SiOz surfaces, and inadvertent Si02 etching during

diamond growth and etching. These issues are important in poly-C probe fabrication.

61

Chapter 5

Poly-C Probe Technology

5.1 Introduction

A Poly-C probe is a novel idea in both MEMS and neuroscience. It utilizes, for
the first time, polycrystalline diamond as the probe substrate, interconnect material and
electrode material. As discussed in the last section, the harsh environment in the MPCVD
chamber during growth limits the integration of diamond into MEMS technologies.
Therefore, the design and fabrication processes of poly-C probes must be specifically
developed as presented in this chapter.

In this chapter, the design and fabrication processes of two types of poly-C probes
are presented. The first type is a probe which is to be used for electrical recording of
neural activity, called the EL probe. This probe has eight poly-C electrodes. The other
type of probe is to be used for electrochemical recording and is called the E C probe. This
probe has Au counter electrodes, Ag/AgCl reference electrodes and poly-C working

electrodes.

62

5.2 Design

In total, nine masks have been designed for the fabrication of the EL and EC
probes as listed in Table 5-1. Depending on the type of probe (e.g. EL or EC) and the
materials used for the interconnects/leads (e. g. poly-C or Au), 3 different number of
masks are needed. Specifics are discussed in later sections. An example of the mask
layout is shown in Figure 5-1. In the layout, there are different types of poly-C probes;
there are EL, EC, cochlear and cholesterol probes. The EL probe has eight poly-C sites;
the EC probe has 2 sets of thee-electrodes (poly-C, Ag/AgCl and Au) for a total of six
electrodes. The cochlear and cholesterol probes have poly-C piezoresistors along their
probe shank. In this research, EL and EC probes are the major focus. In the mask design,

there are also regions for resistivity testing and contact resistance testing.

Table 5-1. A list of the nine masks used in this project.

 

 

 

 

 

 

 

 

 

 

Mask Layer Description
_gds # name
1 BDD Deep boron diffusion for the definition of probe backend
2 BDP Define highly doped diamond electrodes/interconnects
3 PP Define diamond piezoresistors
4 UDP Define undoped diamond probe patterns
5 REP Define reference electrodes
6 CEP Define counter electrodes
7 IP Define metal or diamond interconnects
8 TIP Define top layer insulators
9 CPP Define contact pads for the probe

 

 

 

 

 

63

mmmcm
mczmmu

8305 H. 3333mm

 

 

8303 um.

meshed.
>323me

$305
623866

3205 c8268..

 

 

Figure 5-1. Mask layout for different types of diamond probes.

64

5.2.1 Electrical Neural Probe

A schematic of the probe (depicted in Figure 5-2) reveals details of the probe’s
dimensions used in the design, which requires a six-mask fabrication process. The front
end of the fabricated probes is semitransparent and flexible. Several designs of the
diamond probes have been studied ranging from all-diamond (in which the fi'ont end is
made out of poly-C only) to diamond-based (in which the front end is not completely
made out of poly-C).

Usually, the structural material of the probe, as shown in Figure 5-2, is poly-C
(thickness =3 pm) which provides strong mechanical support which is important during
in viva insertion. The probe has a pointed tip with an tip angle of 30° which enhances the
penetration abilities of the probe in terms of biological entities. It has eight highly boron
doped poly-C electrodes with diameters ranging from 2 to 150 pm [141]. These
electrodes, which are made of poly-C, act as the sites for physiological signal recording.
Leads connected between the sites and metal (i.e. Ti/Au) bonding pads communicate with

outside circuits and systems. The resistivity of the poly-C films used as electrode and

lead material is in the range of 5 x 10.2 — 10.3 Gem [141][142]. The undoped poly-C

. . . . . 5 8 ,
resrstrvrty rs 1n the range of 10 - 10 Item. The backend of the probe is also made of

undoped poly-C supported by thick (thickness = 15 um) highly boron doped silicon.

Shown in the Table 5-2 are the selected masks needed for the fabrication of BL probes.

65

   

a
.3

1.2 mm

TiIAu

(bonding pad)

p"+ -Si

" (backend)
ndoped diamond

7 (substrate)
Si02( pinsuiator)

         
     

(electrode an -
interconnect)

Figure 5-2. Design of the EL probe.

Table 5-2. The six masks used in the fabrication of BL probes.

 

 

 

 

 

 

 

 

 

Mask Layer Description
ids # name
1 BDD Define backend of a probe
2 BDP Define interconnects and boron doped poly-C electrodes
4 UDP Define undoped poly-C probe shank/ sroz layer
8 TIP Define top Si02 insulator which exposes poly-C electrodes
9 CPP Define bonding pads on the probe backend

 

66

 

5.2.2 Electrochemical Neural Probe

The design of the EC probe is similar to that of the EL probe. This probe is used
for electrochemical detection and thus, it contains three different types of electrodes: the
counter electrodes (Au), reference electrodes (Ag/AgCl) and working (poly-C)
electrodes. Each probe has two sets of these three electrode setups. The interconnects are
made of Au. The schematic of the electrochemical probe is shown in Figure 5-3, and the
masks required are listed in Table 5-3.

Gold is chosen as the material for the counter electrode because of its inertness.
The Au electrode is used to complete the circuit in an electrochemical setup and
therefore, any reaction taking place at its surface is not of interest. However, the size of
the counter electrode should be big enough in order to prevent any electron transfer
limitations in the electrochemical circuit. The magnitude of current, and thus the rate of
electron transfer, depends on the electrode area. If it is too small as compared to the
working electrode, it will impose a limitation on the current across the working electrode.
Therefore, as compared to the poly-C electrode (D = 4 pm), the Au electrode is larger (D
= 80 pm) in the electrochemical probe design. Its position on the probe is not a crucial
issue as compared to the reference electrode.

Ag/AgCl is used as the material for the reference electrode. One of the most
important reasons for this choice of material is the ease of fabrication as compared to
other reference electrodes such as standard hydrogen reference electrode. It can provide a
constant potential reference (0.222 vs. NHE) for the poly-C electrode regardless of the

electrolyte. The reaction that takes place is:

67

AgCl(s)+e' e Ag(s)+Cl'

Since the AyAgCl electrode draws no current from circuit, it acts as a potential
reference. It must be positioned near the poly-C electrode on the probe. Otherwise, the
measured potential between the working and reference electrodes is the sum of the
potential of poly-C electrode (as referenced to the reference electrode) and the voltage
(iR) drop in the electrolyte between the poly-C and Ag/AgCl electrodes. In this probe
design, the Ag/AgCl electrode is placed close to the working electrode as compared to

the Au electrode.

 

Doped diamond AgIAgQureference
(site, D=4pm) electrode, D=800pm)

Figure 5-3. Design of the EC probe.

68

Table 5-3 The eight masks used in the fabrication of electrochemical neural probes

 

 

 

 

 

 

 

 

 

Mask Layer Description
_gds # name
1 BDD Define backend of a probe
2 BDP Define poly-C working electrodes
4 UDP Define undoped poly-C probe shank/ SiOz layer
5 REP Define Ag/AgCl reference electrodes
6 CEP Define Au counter electrodes
7 IP Define Ti/Au interconnects
8 TIP Define top SiOz insulator which exposes Ag/AgCl, Au, poly-C
electrodes
9 CPP Define bonding pads on the probe backend

 

 

 

 

 

5.3 Fabrication

5.3.1 Electrical Neural Probe

Shown in Figure 5-4 is the fabrication process of EL probes (i.e. all-diamond
probes and poly-C probes with metal leads). A more detailed procedure can be found in
Appendix A.’The process starts with a bare n-type silicon wafer. A layer of 1.2 pm thick
thermal oxide was grown on top of the wafer as an intermediate step for patterning the
highly boron doped Si areas, which served as the backends of the diamond probes. As
undoped poly-C grth directly on boron doped Si increases the conductivity of the
poly-C due to boron out diffusion (during diamond growth) from neighborhood Si and

the corresponding boron incorporation into poly-C as discussed in the last section, a layer

of silicon dioxide (S102) was grown on top of the Si wafer to prevent boron out diffusion.
It is found that the difference in the resistivities of poly-C films grown on Si substrates

69

with and without S102 protection can be 2 orders in magnitude. Therefore, SiOz is an

essential layer for diamond growth on a boron doped silicon substrate (Figure 5-4a).

Next, undoped (3 pm) and doped (1 pm) poly-C films were grown using the method

mentioned in the previous section with resistivities of ~10 Gem and ~10 chn,

respectively (Figure 5-4b). Aluminum (700nm) was used as the mask for etching as
mentioned above (Figure 5-4c). The Al mask was washed away by Aluminum Etch Type
A solution. The undoped poly-C film served as the substrate for the probe, while the
doped poly-C film was used as the electrodes and leads.

In the case of using metal as the lead material, an additional step is added to
pattern a Ti/Au (50 nrn/300 nm) layer (Figure 5-4c’). Then, a 1.4 pm SiOz layer is
deposited using GSI plasma enhanced chemical vapor deposition (PECVD). This layer
was patterned and etched using buffered oxide solution in order to expose the contact
pads located on the backend and the poly-C electrodes (Figure 5-4d). A layer of
titanium/gold (50 nm/300 nm) was then deposited and patterned to form the contact pads
(Figure 5-4e). Finally, the Si substrate was thinned by HF-nitric (HNA) down to 200 um
and the probes were released by the ethylene-diamine—pyrocatechol (EDP) process
(Figure 5-4f) with boron etch stop. A fabricated all-diamond probe is shown in Figure 5-5.

. The all-diamond front end of the probe is flexible in the out-of-plane direction
(Figure 5-6a) and also it can easily be twisted without breaking (Figure 5-6b). The
flexibility is advantageous in terms of brain implants since the probe must not break

during insertion.

7O

 

 

Initial steps which
are common to
both processes

 

 

 

Undoped
diamond

 

Process flow

for all-diamond E
probe

 

 

 

 

 

Process flow of
diamond probe
with metal
interconnect

 

 

 

 

 

Figure 5-4. Fabrication process of all-diamond probes.

71

 

Figure 5-5. Released all-diamond probe on a US penny. The highly doped Si back end
was removed in this probe.

72

 

(bi

Figure 5-6. An all-diamond probe was bent in the out-of-plane direction with a pair of
tweezers (bending angle was larger than 1800).

73

5.3.2 Electrochemical Neural Probe

The fabrication process used for the EC probe is shown in Figure 5-7. The first .

four steps of the process are exactly the same as that of the EL probe. In surmnary, a

group of undoped/doped diamond layers is grown on top of a SiOz substrate. Then, these

layers are etched using RIE with aluminum (1 pm) as the masking layer. For the EL
probes, highly boron doped diamond is used as the working electrode (Figure 5-7d). The
metal stack for the reference electrode is Ti/Au/Ag which have thicknesses of 40 nm/ 10
urn/300 nm and are patterned using sputtering and a lift off process. The Ti/Au layer
underneath the Ag is used to provide adhesion between doped diamond and Ag.
Formation of Ag/AgCl is performed after the probe release. The materials for counter
electrode and leads are Ti/Au using lift off process. Then, a layer of PECVD oxide was
deposited on top with the three electrodes and bonding sites exposed. Ti/Au bonding pads
were patterned on the very top layer. Finally, the probe was released by HF-nitric and
EDP process. Shown in Figure 5-8 is a fabricated EC probe which has two sets of

electrodes (i.e. working, counter and reference).

74

flirt-2.2 mums:~21!.!.~1-3.-<m'«r:-.‘-t-'=:02:-‘firtirrt131‘s: ‘v’s :‘JT‘. 1‘." m
..-

--;.u- J'.LD '
.«u.,..( 1 E. a

 

doped silicon

— Doped diamond
'1! Undoped diamond

n I. \‘o‘r', a
u

 

__}1

 

 

 

a»
2" a
z .-
.- ’ 3",
3;
IQ” -' 1“th

..... (c) Etch doped diamond by RIE

g
l

 

.
‘..._.—. r. Mv—wfiquemv—r‘wq raw-un-

(d) Etch undoped diamond by RIF.

TllAuiA

- --.<.~r~,.-1mcnrr-.-.~.—a-:~zr 1...

 

 

 

rm”; ‘ .
@PaueihT1(50myAu (300nm) by lift-off process
.3122 .........r.-.... .11.. n...

-.qanmm¢rn—1—l-§Or‘1

 

--I-7h-a-—-:'.T-1-.r Wu.“
. r

r“

' «no.3

3..'-- .35. 'Iinter‘t' 27—:- H...Vf .fi'flu 14‘4" .’“M‘ .31.

(11) Deposit and pattern PECVD oxide (1 .6 pm)
Au

E. :3 U .. I ' ' firm-erg . ,
HG) Pattern Ti (30nm) /Au (300nm) by lift-off” proces;
flvwwrifiiflaah

 

fringwrrr-rvnrm Hf": .. r.-- ---m* W11} fig“

 

 

Er: . ~ - 'r -.-' --~ r ‘- ".'"1'45'7‘31"."r‘f‘fi’"ltfllm‘fft‘m-‘J'I‘I‘fi
11' - . ‘ - ,

 

Figure 5-7. Fabrication process of the EC probe.

75

 

Figure 5-8. (a) A fabricated EC probe, which has an (b) Au counter electrodes, (c)

Ag/AgCl reference electrodes and a (d) poly-C working electrodes.

Afier the EC probe is released, the last step of the fabrication process is the
formation of AgCl on top of Ag electrode surface. There are different ways to form AgCl
from Ag [143]. For example, anodization of Ag in chloride-containing solutions,

chemical oxidation, and thermal or plasma treatment in chloride-containing atmospheres.

The chemical oxidation of Ag in aqueous FeCl3 is chosen as the formation method in this
project because this process is compatible with the probe fabrication process and is easily
performed. The EC probe was immersed in ferric chloride FeC13. SEM images of AgCl
formation on Ag electrodes under different FeCl3 concentrations and immersion times
are shown Table 5-4. It is observed that the average grain size of AgCl increases with

increasing FeCl3 concentration and reaction time.

76

One of the problems associated with the use of an Ag/AgCl electrode is that it
requires a constant source of chloride in order to maintain its constant potential. In other
words, it must always be immersed in a chloride rich solution (e. g. KCl) to maintain its
potential. Otherwise, the AgCl may dissolve completely if the probe is immersed for a
long period of time. These facts suggest a stability problem relating to the Ag/AgCl
electrode in an aqueous enviromnent. If the stability of the electrode cannot be
maintained, its potential will change with respect to time and therefore, there will no
longer be a a fixed reference for electrochemical measurements. Typically, it is observed
that the Ag/AgCl electrode is stable for a time period in the range of a few minutes to a
few hours [144]. One way to lengthen the electrode’s maximum immersion time is to

coat the Ag/AgCl electrode with a gel or polymer material (e. g. Nafion or polyurethane)

[l45][146][l47]. This creates a diffiision barrier between the C1- ions dissolved from the

electrode to the electrolyte, and therefore, lowers the rate at which the AgCl dissolves.

77

Table 5-4. AgCl formation on Ag electrodes using FeCl3.

 

 

 

 

Dipping Time 1M FeCl3 50mM FeCl3
(S)

30

60

120

 

 

 

 

 

 

5.4 Probe Mounting and Packaging

To increase the ease of accessing sites on the poly-C probe, the probe is mounted
on a printed circuit board (PCB) which extends the leads and connects easily with outside
circuits or systems as shown in Figure 5-9. The probe is mounted on the PCB using
Crystalbond 200 which is a low temperature adhesive gel. Then, the wire bonding is

performed using an aluminum wedge bonder which connects the pads on the probe to that

78

of the PCB. Finally, for the purpose of in vivo testing, the wires and the backend of the
probe are encapsulated by an optically transparent silicone elastomer from Silicone
Technology. This encapsulation provides electrical and chemical isolation of the wires

and the backend fi'om the working environment.

Aluminum wires POW-C probe

Encapsulatinggel

 

Figure 5-9. Probe mounting.

5.5 Summary

Two types of poly-C probes, EL and EC probes, are presented for the first time.

. . . . . . . . -3 .
The EL probe cons1sts of eight poly-C Sites (With res1st1v1t1es of ~10 0cm) which are

used for electrical recording. The lead material can be either doped poly-C or Au. The
EC probe, on the other hand, has three types of electrodes (i.e. Ag/AgCl reference

electrodes, Au counter electrodes and poly-C working electrodes) used for

5
electrochemical recording. The probe shank is undoped poly-C with a resistivity of ~10

Qcm in both cases. The fabricated poly-C probes are then mounted on a PCB in order to

interface with external systems.

79

Chapter 6
Characterizations and Applications of Poly-C

Probes

6.1 Introduction

The common techniques for characterizations of an electrode include Electrical
Impedance Spectroscopy (EIS) and Cycle Voltammetry (CV). Electrical Impedance
Spectroscopy is used for electrode electrical modeling which characterizes the electrode’s
behavior across a wide range of fiequencies [86][l48]. Cyclic Voltammetry, on the other
hand, is used for studying electrode reactions, which yield information not only on the
thermodynamics of reactions but also on kinetic parameters (e.g. heterogeneous electron-
transfer rate constant). In this chapter, EIS and CV will be discussed in detail in the first
two sections. In the later two sections, two applications of the poly-C neural probes will
be discussed: electrochemical recording of neurotransmitters and electrical recording in

vivo.

80

6.2 Poly-C Site Characterizations

6.2.1 Electrical Impedance Spectroscopy

Electrical Impedance spectroscopy (EIS) is a technique used to characterize an
electrode, in an electrical sense, for small signal perturbations (V pp < 0.05 V). Together
with an appropriate equivalent circuit model for the electrode/electrolyte interface,
parameters such as interfacial capacitance and charge transfer resistance can be obtained.
Since a small perturbation is applied to the electrode, the electrode is working close to
equilibrium and thus, a linear current-voltage response is obtained. The measurement is
performed over a wide range of frequencies (l'Hz to 100 kHz). The magnitude of the
electrical impedance Z and phase angle 6 are recorded at each frequency. The impedance,

Z, can be represented as,
I 0
Z, = Z — )Z" (13)

where Z’ (>0) and Z” (>0) are the real and imaginary parts of the total impedance
Z, respectively. It is realized that the definition of Z always has a negative imaginary part.
This is because the electrode impedance is almost always capacitive as discussed in

chapter 3. The phase 6 is represented by,
_ -1 Z_"
0 — tan (2,) (14)

As discussed in chapter 3, an electrical double layer is formed when an electrode is
placed in an electrolyte. The electrons in the metal and the charged ions close to the
electrode surface are attracted to each other and separated by solvent molecules (usually

water). This interface acts like a capacitor which charges or discharges depending on the

81

imposed voltage. The capacitance of this double layer structure governs the electrode’s
ability to deliver charge (in electrical stimulation) and its background current (in
electrical and electrochemical detection).

For the model used in this project, the Warburg impedance and the double layer
capacitance in the Randles model are replaced by two constant phase elements (CPE, W
and Q) which results in a better curve fitting, and therefore a more accurate
representation of the electrode in this study. The constant phase element has a constant
phase different from an ideal capacitor and its impedance is defined by the following
equation.

1

Zcpra = W (15)

The modified equivalent circuit for the electrode/electrolyte model is shown in

Figure 6-1. The total impedance is given by,

_ : Q(Rt+W)
Z _ RS ' W+Q+Rt (16)

 

1 1

= (ide1)“’ W _ (iwa)B

 

 

where Q

andO 5. (LB S 1.

After Q and W are substituted in, the total impedance and the phase of Z can be

computed as shown in Appendix C.

82

At low frequencies, the total impedance Z is dominated by the Warburg impedance,
which represents the mass transfer of ions to and fi'om the electrode, and becomes

approximately,

1
(l (”Cwfls

 

Z] = RS + Rt 'l'" (17)

. . . . n3 . .
This corresponds to a straight line With a slope of tan (7)1n the Nyquist plot.

On the other hand, at high frequencies, the effect of the Warburg impedance is no
longer significant as compared to that of the double layer capacitance. Therefore, the total
impedance becomes approximately,

Rt
1+RtCIde1)“

 

Zh = Rs + (18)

This corresponds to the semi-circle in the high frequency range of the Nyquist plot

as discussed in chapter 3.

 

Figure 6-1. Modified Randles model of the electrode/electrolyte interface [86].

83

To measure the impedance spectrum, the poly-C probe is mounted as explained the
in previous chapter and placed in 0.9% saline solution (9000 mg NaCl: 1000 ml DI H20);
the poly-C site acts as the working electrode. The reference and counter electrodes are
commercially available Ag/AgCl and platinum electrodes, respectively. The
measurement is performed inside a faraday cage in order to minimize noise from external
sources as shown in Figure 6-2. The electrochemical station CH 750C from CH

Instruments which has the ability to do EIS was used.

 

 

Faradaypage

 

 

 

 

 

 

   

Computer

I
Platinum A9! AQCI Electrolyte
electrode

Figure 6-2. Experimental setup for the EIS measurements.

In this measurement, a small signal of (5 mVpp) was applied to the poly-C
electrode, and divided by the induced current to find Z ' and Z The Nyquist and Bode

plots in the frequency, co, range of 0.1 — 100 kHz are shown in Figure 6-3. The predicted

84

total impedance is described by (16). Least squares curve fitting was preformed to obtain

the model parameters (R S, Rt, Wand Q). The error is defined as:

2
N (Zpredicted,i‘zmeasuredj)
e = i=1

 

Zz . (19)
measured,1

where N is the total number of data points.

In this method, the fimction e is minimized by adjusting the model parameters; the
optimal parameters obtained are listed in Table 6-1. At low frequencies (< 4 kHz), both
the real and imaginary parts of the total impedance Z decrease with increasing frequency,
and Z has a negative phase 6 throughout the spectrum as shown in Figure 6-1(a). This
indicates that the electrode behaves capactively. This low frequency behavior is described
by the Warburg impedance term, in (4), which is defined as a constant phase element

(fl=0.43). This element describes the ions movement in the bulk solution to and from the

electrode interface. The impedance becomes more resistive (i.e. 6 changes from -230 to -

150) as the frequency increases as shown in Figure 6-1 (b). The cutoff frequency is around

4 kHz at which point the impedance enters the high frequency regime which can be

described by (5); 6 reaches its minimum. The impedance becomes more capacitive again

(i.e. 6 changes from -150 to -500) at high frequencies (> 4 kHz) as shown in Figure 6-1

(b). In this region, the double layer capacitance Q and charge transfer resistance R,

dominate in the total impedance Z. In other words, the current is mainly due to the

charging of the double layer at the electrode/electrolyte interface at high frequencies. It is

85

noted that there is no further decrease in 6 within the measurable frequency range of the
equipment. Thus, the poly-C electrode behaves similar to a capacitor within the

frequency range of 0.1 — 100 kHz for small signal perturbations. The double layer

capacitance Cd] as obtained from fitting the measured data to the electrical model, is ~87

uF/cm2 which is within the range of reported values (3 - 100 uF/cmz) [149][150]. This

normalized capacitance is calculated using the geometric area of the electrode. In the case
of a poly—C site, the effect of its surface roughness on the double layer capacitance can be
significant--especially in the high frequency regime. This is because the thickness of the
interfacial double layer structure can be less than 5 nm [8]. Thus, the true double layer
capacitance can be lower by a factor of 3 to 10 if the roughness of diamond is considered
[149]. Polishing the diamond surface is one way to address this issue. Despite their
roughness, diamond electrodes are known to have low double layer capacitance as
compared to most metal electrodes [151]. This helps keep the RC time constant low.
Therefore, the time needed to charge and discharge the double layer structure is small
which is important in terms of in viva electrochemical detection as fast scan voltammetry
is often used. Also, the background current, as discussed in the next section, can be

lower due to the small double layer capacitance.

86

 

45.0 .

 

40.0 j
35.05
30.0]

IIIIIIIIIIIIIIIIII

 

25.0 ;
20.0-f

- 2” (k0)

150-}

 

 

10 0: - gm h 42.)- . - Wm--.” .......... . ...... 5 ______ :
g g Low to t
1

r i 1 1 ; I ii I I T

50

   

I I
ID..........Q‘.;9.......I.Q.OOD.0......OIOICC9.;0I‘ll-IO...”.O...O...‘..‘O..'~.U....O'.....OWOQOOI........‘.” .........

  

A! ALI LL14 I ninmnnln ngnnnnl
' . Measured value
.............. ’ooaonoouooooou—OOODI-DOD:I... ..o. . no... no

" Predicted value

 

 

 

 

 

 

100 20.0 ”000' 40.0 50

0000' 100'

80 0 90.0

 

 

 

 

Z’{kl’2)
(a)
5.2 r _ '60
= 5 4 Phase -50
N.
3 4.8 a _ 40¢
1 .-g an
5 a
.N 4.6 ~ - ’ . so:
a
on '08 2 log 2 g
e
'1 4.4 ~ 409*
N
en
.2 4.2 ~ - -10
log -z"
4 r r 0
1.8 2.8 3.8 4.8
logo)
(b)

Figure 6-3. Impedance spectrum for a diamond electrode with diameter of 30 mm, (a)
Nyquist plot of the real and imaginary part of the measured impedance, (b) Bode plot.

87

Table 6-1. Fitted parameters for electrode/electrolyte model.

 

 

 

 

 

 

Parameters Values

. . fiZ
Spreading resrstance RS (0) 4953(=0.03 Qcm )

4 2

Charge transfer resistance Rt(Q) 4.02x10 (=0.284 9cm )
Warburg impedance w (s secB) 2.22x10fl6 (0:043)
Constant phase impedance Q (S seca) 1.12x10-9 (a=0.95)
Double layer capacitance Cd1(uF/cmf) 87'45

 

 

 

 

6.2.2 Cyclic Voltammetry

As discussed in earlier chapters, cyclic voltammeU'y is a useful tool in studying
electrode kinetics and characterizing the electrode’s performance. In this section, CV is
used to compare the potential window of different materials: diamond, gold, platinum and
iridium oxide when they are used as the working electrode. The experimental setup is
shown in Figure 6—4. As shown in the figure, a specially designed 25 ml glass flask which
has 4 openings is used. The CE and RE electrodes are placed at the top and are immersed
in the electrolyte. The third opening at the top can be used to purge oxygen which has

dissolved in the electrolyte. The WE is placed such that is covers the bottom opening of

the flask. The opening on the bottom of the flask has an area of 0.2 cm2. The metal or

diamond films are prepared on top of an Si substrate. These three electrodes are then
connected to the potentiostat as mentioned in the last section. The experiment is

performed inside the faraday cage.

88

/CE

RE

 
 
 
  
   
 

Bottom
opening
(O—ring)

/WE

Metal piece

 

Figure 6-4. Experimental setup for cyclic voltammetry. .

In order to compare the performance of diamond electrodes with other electrodes

commonly used in neural studies such as Au, Pt and IrOx electrodes, a 300 nm thick film

of Au, Pt or Ir was deposited on a highly doped Si substrate. IrOx was formed by
activating Ir in phosphate buffered saline (PBS) solution [152].the complete procedure
for IrOx activation is listed in Appendix B. The background current of the electrodes in

potassium chloride (KCl), Krebs and PBS solutions are shown in Figure 6-5. It is found
that IrOx has the largest background current in each of the solvents and exhibits several

peaks due to reduction-oxidation reactions that involve the transfer of electrons across the

interface [153]. As evidenced by Figure 6-5, IrOx not only has a high charge storage

89

capacity and also a large charge injection ability. Therefore, IrOx is a material of choice

for cell stimulation applications. The reason of the high Qin associated with IrOx is that

the mechanism of its charge injection is the F aradaic reduction and oxidation between the

3+ 4+ . .
Ir and Ir states of the oxrde at the electrode surface whrch correspond to the peaks

seen in the CV in Figure 6-5 [27]. Poly-C, on the other hand, shows no peaks and also
shows the lowest background current among the materials. This is an advantage-
especially for electrochemical detection because the low background current and lack of

peaks implies that its surface is chemically inert and stable.

90

 

asses

 

 

 

.eeeeeé

0:3 ' 1:2" 106

Current Densitv (uA/cmzl

 

é

       

$$e$$g

as

 

 

 

Current Densitv (uA/cmzl

.a

“aogeeeaeae

 

 

vvvvvvvvv

0.8 0.6 0.4 0,2 0 0.2 0.4 0.6 0.0 1.0 0102
Voltage

Current Densitv (uA/cmzl

Figure 6-5. Voltarnmograms of Au, Pt, IrOx and diamond electrodes in (a) lMKCl, (b)
Krebs, and (c) PBS.

91

In addition to low background current, poly-C electrodes also exhibit a wide water
potential window. The water potential window is defined as the range between those
potentials where oxygen and hydrogen evolution take place. There is no charge transfer
across the electrode/electrolyte interface within the potential window. In other words, it is
the range of potential where an electrode behaves as an ideal polarized electrode. The

potential window can be determined fiom CV curves recorded using an electrode in an

aqueous solution. In an aqueous solution, water (H20) acts as the source for the oxygen

and hydrogen that will evolve. The two chemical reactions are,

ZHZO -) 02 + 4H+ '1' 4e (20)
2H+ + 23 —) H2 (21)

A species can only be detected if it undergoes redox reactions at potentials less
than that of hydrogen evolution, or greater than oxygen evolution. Therefore, the wider
the water potential window an electrode has, the more chemicals it can detect. The
potential window of the poly-C site was recorded and is shown in Figure 6-6. The poly-C
site on the probe reveals a potential window from -0.8 to 1.4 V in 1M KCl; these over-
potentials for both oxygen and hydrogen evolution are relatively high. In other words,
there is no current caused by oxygen and hydrogen evolution which would interfere with
current from a target species undergoing redox reaction within this range. As a result, the

poly-C site is useful in a wide range of potentials, and more species can be detected.

92

 

ty (mA/cmz)
N DJ '5 UI
O O O O

y—n
O
I

 

 

 

———
A
V

Current Densi
O

 

 

 

 

 

-10 — =
Water Potential Window
.20 j 1
-1 0 l 2
Voltage (V)
(a)
18
0.377 V (half-wave
12 _ potential)

 

 

 

 

Current Densitry (mA/cmz)
Ox

 

 

-6 1 1 1
-0.5 0 0.5 l 1.5

Voltage (V)
(b)

Figure 6-6. (3) Cyclic voltammetry current-voltage curve for the diamond probe in 1M
.4
KCl. (b) The CV of the diamond probe in 1 mM Fe(CN)6 in 1M KCl. The scan rate

was 100 mV/s. The reference and counter electrodes are Ag/AgCl and platinum,
respectively.

93

In diamond related electrochemistry, the Fe(CN)6.3/.4 redox system has been

widely used to evaluate the reactivity of electrodes as they undergo electron transfer via a

simple outer-sphere mechanism. In addition, physical and chemical properties of the

- /-4
diamond electrode make the surface highly sensitive to this Fe(CN)6 3 redox system

[25]. Thus, this redox system can be used to compare the quality of different diamond

electrodes. The cyclic voltammogram of a poly-C site shown in Figure 6-6 (inset) reveals

the half-wave potential E1/2 which is 377 mV (vs. Ag/AgCl). This is the potential at

which the value of the current is one half of the difference between the peak current and
the residual current. This figure falls within the range (~ 250 mV to 450 mV) that a

typical diamond electrode should exhibit.

6.2.3 EC Probes

Having the success in the fabrication of poly-C electrodes on poly-C probes, two
additional electrodes with different materials are integrated to the probe for advance
electrochemical testing applications. The two additional electrodes are Ag/AgCl and Au,
as discussed in chapter 5, which replace the two commercial metal electrodes used in
electrochemical measurement. The main advantage of the EC probe is that the sizes of
the three types of electrodes can be minimized, which allows electrochemical detection in
small areas. It is especially important in the applications for neural studies.

Preliminary tests have been performed on the EC probe including the CV of the

- -4
probe in KCl and Fe(CN)6 3/ . Shown in the Figure 6-7 is a comparison of the CV in

94

1M KCl using (1) three electrodes on EC probe and (2) poly-C site on the probe and
commercial RE (Ag/AgCl electrode) and CE (Platinum electrode). It is shown that the
shapes of the two curves are almost the same. This shows that the CE and RE on probe
work the same as commercial electrodes. The lower in the current for the EC probe may

due to the increase in lead resistances in the CE and the difference in the current path in

the electrolyte side between the two cases.

 

1 2 __.L A A. J. A L .L J A L .1 l A .1 1 I A A A l n L A l A I
O a

1"“7

1,0{ A WE on EC probe; Commercial CE and RE ;

1 ——- WE, CE and RE on EC probe 3: :
0.8 1 ‘: _~
0.6-j. ;. z'
0.43 .. :-

1

4 b
02* ~
0 d p
I I

J
0- -
«r b
4 h
p

.b
m

Current Density (Alcmz)

b
.5

 

 

 

-1.6 -1.2 08 -0.4 0 04 0.8 1.2

Voltage (V)

Figure 6-7. A comparison on the cyclic voltammograrns in 1M KC].

- -4
Shown in Figure 6-8 is the CVs of an EC probe on Fe(CN)6 3/ (in 1M KCl) with

different scan rate v. The diameter of the poly-C WE is 20 pm. It can be extracted from

the curves that the E 1 /2 is ~230 mV, which confirms the diamond quality as discussed in

95

the previous section. At scan rate of 0.05V/s, the shape of the curve is sigrnoidal which a
typical property of a microelectrode is. Its limiting current can be calculated by the

following equation,

it = 4nFDCr (22)

where n is the number of electrons involved in n electrode reaction (n=1), F is the

- 2 -
Faraday constant (F=95600 C), D is the diffusion coefficient (D=6.5x10 6 cm s 1), C is

the analyte concentration (C=1 mM) and r is the radius of the electrode (r=10 pm). The

calculated 1'] is 1.25 nA. This value matched with the measured value 1.1 nA (=1.6 nA —

0.5 nA), which is background subtracted. As the scan rate v increases, the shape of the
curves begins to show a peak. There is not enough time for the analytes to diffuse up to
the electrode surface. In other words, there is not enough analyte for taking the reaction,

therefore shape begins to show a peak.

96

J.

l A l A J A l l A l I A l j l I l I L .l

 

N 9°
C :5

Current (nA)
3

i h
b

0 o". l-. ‘ *-
ll...-...-.'II-II.III.II.--- ‘ h

d ‘ "

 

   

I A I ‘ l A ;‘;.
A“ t
4‘ v=O.ZVIS ‘Abntaoaaaaaaat“‘ .
— A
o v-O.1V/s ‘ _ mt
. v=0.05V/s . f f.
A .IIIII-
A 8 .II" a? ’
‘ .. '. ".. h
L b
‘ i-
A
A‘ .
A “ °'.
At“ a“ °
5.6““. . g I 0°;
‘ . ultnau""""'”. '- 0‘
..IIH“” . --- -
-”“”3IIIIIIIII"" I 3"

A L
‘

 

 

nhaoannaona“‘ A

“““ ‘

-1 0 I r I ?A¢A¢A‘v T
C

I V T I I t'

-100

 

0 100 200000400000
Voltage (mV)

Figure 6-8. Cyclic voltammograms of the EC probe in 1mM Fe(CN)6-3/-4 + 1M KCl.

6.3 Probe Applications

6.3.1 Electrical Detection

Poly-C EL probes are used for the first time to record in vivo neural activity by
surgically inserting the probe into the auditory cortex area of a guinea pig brain as shown
in Figure 6-9. The in viva electrical detection was done in the Kresge Hearing Research
Institute at the University of Michigan and performed on pigmented guinea pigs using
approved animal use and care procedures. The poly-C probe was mounted on a probe
connector which is precisely controlled by a sub-micron positioner. The poly-C probe

was then'placed in the auditory cortex of the guinea pig which is located in the right

97

hemisphere of its brain. The probe was connected to a Plexon Neural Data Acquisition
System to record the stimulated neural signals.

During the electrical data recording, a broadband sound signal (stimulus) was
applied twice per second to stimulate the brain of the guinea pig. Each stimulus lasts for
200 ms. The all-diamond EL probe monitored the effect of the sound signal on the neural
activity inside the cortex; an example of the recorded signal from one electrode of the EL
probe is shown in Figure 6-10. The recorded electrical signal was amplified 20,000 times
and filtered using a bandpass filter designed to pass frequencies in the 100 Hz - 10kHz
range. It can be seen in Figure 6-10 (b) that the probe received a relatively large signal of
>25 11V in response to the sound signal. It is demonstrated that the diamond probe
received the stimulated neural signal at a signal to noise (S/N) ratio of ~2.

There are several possible reasons for the low SfN ratio. One of the possible
reasons is that the probe position may not be close enough to the body of the firing
neurons. The strength of the neuron signal attenuates with increasing distance between
the recording site and target neurons [154]. Increasing the SNR will be a priority in future

work.

98

 

Figure 6-9. Acute recording of neural activity using the diamond probe.

99

Stimulu
i | I I l

Voltage

 

 

 

 

o .
DD 1 I i ‘i I, i
g i
O
> .
_ 10 ms
1 1 I L I l I r— l
Time
0’)

Figure 6-10. (a) Recorded neural activity in guinea pig’s auditory cortex using the
diamond probe with an electrode area of 100nm , (b) A recorded neural spike.

Another recorded signal from an in vivo experiment using an all—diamond EL
probe is shown in Figure 6-11. Again, the probe was placed in the auditory cortex of a
guinea pig’s brain. The signals recorded from each of the eight sites are shown along with

their corresponding positions on the probe. It can be easily inferred fiom the signals that

100

the most intense neural activity was taking place around sites 1, 2 and 3. The sites
recorded a peak signal every half second which is consistent with the fiequency of the
stimulus signal. On the rest of the sites, it seems that there is no neural activity recorded
or the induced signal amplitude is too low to be observed. Another way of representing
the recorded signals is shown in Figure 6-12. This acts as further evidence that the probe
can successfully record neural activities. These histograms, shown in the figure, represent
the recorded signal which has amplitude larger than a preset threshold at a particular time
internal called bin (=1 ms). In the plot, the onset of the stimulus always takes place at t=0
and lasts for 200 ms. Therefore, the recorded signals beyond t = 200 ms are noise from
external sources or related to background brain activity. It is shown in the plots that site 1,
2, 3 and 4 were receiving neural activity caused by the stimulus. This is proven by the
fact that most of the recorded neural signals were received just as the stimulus began (i.e.
higher counts/bin). Some signals were still recorded between the onset of the stimulus
and 200 ms. The stimulus is still present at this time; it is seen that neural activity is
more sensitive to the onset of a stimulus rather than its presence. Beyond 200 ms, little or
no activity is recorded; these signals are considered noise. Sites 5, 6, 7 and 8 show little
or no neural activity because the amplitude of all the recorded signals are around noise

level.

101

 

 

 

 

 

 

 

 

 

62.3
_
.
i—

 

 

 

 

 

 

 

 

 

 

 

 

 

1 2 3 4 5 6 7 B 9
50 r
43W.
0 1 2 3 4 5 6 7 B 0
so I I . I 1 I T I
S 0W
4'” I l l l I I l
=— 0 1 2 3 4 5 6 7 8 9
v50 1 l r r r I l
0
Q -50 l l J. l _l l l 1
CD 0 1 2 3 4 5 o 7 B 11
.5 fiWfin—W
l ' i
3600 r 2 3 4 5 o 1 a ll
50
> 0W
-50

 

O
a.
u
to
a-
01
O
V
a
O

 

2.36203
to
—4 D
@—

O
N
u
5
0|
9
a
O
9

Site 1

 

Figure 6-11. A recorded in vivo signal from an all-diamond EL probe. Site separation is

~500 pm.

102

 

 

 

 

2m . . . r a . .
.=
5 m -
C
3 0 ”W... - . . . . .

0 5] [[1 611 an 251] (ll) :50 41]] 45] am
Time(ms)

= 2m . . . .

   

Coun

 

 

 

 

 

Countefoin

 

Countsfoin

 

 

 

Countst

5

0 9] [I] 60 21]] 23] 300 38] 4:0 45] 5]]

Time rams)?

V

8 8

.

Countslbill

 

O

0 50 m 50 200 250 300 350 am 450 an
'rl’imor(ms)T

 

I
I I

 

 

 
 

40 .
.s
E
8
0 50 [11 1'50 21]] 250 300 350 400 450 500
I) Timo(ms)
g.
c 5 ‘ . ix . q
3 ; lL . 5 I i I i 10 :1[
0 0

 

0150:050200250330350400450500
Timo(ms)

Figure 6-12. Histograms of the recorded neural signals shown in previous figure.

103

6.3.2 Electrochemical Detection

Norepinephrine (N E) is a neurotransmitter which can be found in the sympathetic

nervous system and acts as a stress hormone. Norepinephrine, detected electrochemically

according to the following 2e-/2H- redox reaction, is oxidized to form NE-o-quinone.:

OH OH
HO NH2 0 W2
___. + 2H+ + 2e"
HO 0
NE NE-o-quinone (23)

In a preliminary study focusing on the detection of NE in Krebs buffer solution,
current density was recorded as a function of voltage applied to the poly—C electrode with
respect to an Ag/AgCl reference electrode. The recorded data shown in Figure 6-13 are
presented after subtracting the background current (as shown in the inset of Figure 6-13)
as measured in Krebs buffer solution. The curves shown in Figure 6-13 are the currents
recorded at different NE concentrations. The recorded current starts to increase when the
applied voltage reached ~0.2 V. The lowest detectable concentration of NE is roughly 10
nM. The results are similar to those reported in [36] for diamond coated electrodes but
there is no peak observed in the recorded current using the poly-C electrode. One

possible reason for this could be the variation in diamond quality which leads to a small

heterogeneous rate constant k0 [27].

104

w

t" S" .
UIHMNMUJM-h

P
o

 

Current Density (mA/cmz)

.5
U!

 

 

 

 

 

 

 

 

4
a; 2 —-
:5; o
-4 r r r
O 0.2 Vogt'gge (OV'IG 0.8

Figure 6-13. (a) Measured cyclic voltammograms of the poly-C probe for different
concentrations of NE in Krebs buffer solution using a scan rate of 100 mV/s. (b) It shows
the background current in Krebs solution.

105

In addition to NE detection, the detection of dopamine (DA), which is another
important neurotransmitter, and is believed to be related to Parkinson’s disease, is
demonstrated. Shown in Figure 6-14 is a comparison of the detection of NE and DA with

the same concentration. Clearly, there is a difference in the magnitude of the current

density between the two neurotransmitters. However, it is difficult to identify E 1/2 for the

two curves as there is no clear plateau (i.e. limiting or saturated current). In fact, the

reported E 1/2 (vs. SCE) of DA and NE are 200 and 240 mV, respectively [155].

 

3.8 A 1 L ‘ ‘ 1 L 4 ‘ L 1 1 ’2’.
T ,3: :
j .e‘ :
3 0: t 1mM Norepinephrine I!!!“ 1
. I 1mM Dopamine .e‘“ r“ .

 

d
0'1
.-.AJ.-..

.0
co

 

Current Density (mA/cmz)

 

0 0.20

i r

0.00 ' 1.00

0.40 ' 0.00
Voltage (V)

Figure 6-14. Detection of NE and DA with same concentration.

106

6.4 Summary

Poly-C probes are characterized using EIS and CV. It is found that the poly-C

2
electrode acts similar to a capacitor and has a double layer capacitance of ~87 uF/cm . In

addition, the poly-C electrode has the lowest background current (on the order of nA for a

centimeter scale electrode) and a wider potential window (2.2V in total) as compared to

Au, Pt and IrOx. The EC probe is also characterized and shown to perform as well as

commercial electrodes.

Poly-C probes are applied to in viva electrical and in vitra electrochemical neural
recording. The probes are successfully implanted in a guinea pig’s brain and used to
recorded in viva stimulated neural signal. It is also demonstrated that the poly-C probes

are able to detect both NE and DA.

107

Chapter 7

Summary and Recommendations

7.1 Summary of Contributions

1. Fundamental research on poly-C technology

Poly-C technologies used in this project, including seed nucleation, growth and
patterning, have been studied and addressed. These poly-C technologies are developed
specifically for later poly-C probe fabrication. Two issues, which are crucial in terms of

the fabrication of poly-C probes, have been reported and discussed. They are

unintentional SiOz or Si etching during poly-C growth and diamond etching. The
corresponding solutions are discussed.

2. Enabling technology development in poly-C probe fabrication

The integration of poly-C into neural probe technology has been successful. All-
diamond probes, which use poly-C as the material for the probe shank, leads and
electrodes, have been realized for the first time. With the development of the poly-C
technologies in this study, two types of poly-C probes (i.e. EL and EC probes) are
fabricated and reported for the first time. EL probes consist of a number of poly-C sites

for electrical (physiological) detection. EC probes, on the other hand, consist of three

108

different types of electrodes (i.e. RE, CE and WE) which are necessary for performing

electrochemical detection.

3. Poly-C neural probe applications

Implementations of the poly-C probes in neuro-electrical and neuro-chemical
recording have been studied. EL probes have been successfully implemented in a guinea
pig’s brain and acute electrical signals have been recorded in viva. EC probes, on the
other hand, have demonstrated the ability to detect neurotransmitters in vitra. This
pioneer work on poly-C probes will provide a strong basis and guidance for future

developments in poly-C neural probe technologies.

7.2 Future Research Areas

As for the long term development of the poly-C neural probe project, the following
research areas must be addressed and discussed intensively in order to provide more
promising performance of the poly-C neural probe.

1. Further improvements on the conductivity of undoped poly-C must be
made. One possible way of accomplishing this is the introduction of
oxygen during diamond growth. This may lead to having undoped poly-C
packaging for the all-diamond probe.

2. Further study on the electrical and electrochemical properties of poly-C
electrodes and the integration of carbon nanotubes with poly-C electrodes.
With further development of the poly-C electrodes, it may be possible to

have a poly-C electrode serve as a stimulating electrode.

109

. Integration of ultra nanocrystalline diamond (UN CD) into the probe
fabrication. This can provide a smoother surface that is easily built upon.

. Improvement on the design of the EL and EC probes, for instance,
minimization of the probe size, optimization of the layout of the three
electrode sets on the EC probes and the integration of channels in the
probes.

. Further study of the chronic implantation of the poly-C probes.
Investigation of the long term performance of the poly-C probes in viva.

. System integration of the poly-C probes with a microcontroller chip for
signal recording, processing and transmission.

. Exploration of the possible applications of poly-C probes. For instance,

surface modification of poly-C electrodes for biosensing applications.

110

APPENDIX

A. Process F low of the Fabrication of Diamond Probes

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

S Process Operat- M Descript Equip- Parameters ln-line
ion a -ion ment
t Measure-
S
e ment
k
P
1 Pattern Pro-furnace
clean
probe
2 Grow oxide B2 tube Recipe: Oxide thickness
“01‘6““ DWD/I‘CA ~12 um
Parameter
table:
DWDSKIN
4 h, 1000 °C
3 Spin 1813 1.4um Spinner 4000 rpm, 30
sec
4 Soft bake 95 °C Hotplate l min
5 Expose B Define MA6 5 sec
D backend
F
6 Develop MIF 318 l min
7 Oxide etch BHF 17 mins (until
patterned areas
are
hydrophilic)
8 Pre-furnace
clean

 

111

 

 

 

 

 

 

 

 

 

 

 

9 Deep boron Define A2 tube Recipe:
diffusion backend BEDEP99
Parameter
table:
BORON99
5 h, ll75°C
10 Drive-in A4 tube Recipe:
N2ANIJOX
Parameter
table:
N2ANUOX
5 h, ll75°C
11 Strip deep HFzHZO (1:1) 7mins
boron
masking
oxide
12 Pre—furnace
clean
—— Dielectric _ , _
13 d osition LPCVD Dielectric D4 tube Recrpe: HTO ~l.2 um
ep oxide deposition
3 h 17mins,
450°C
14 HF—dip Refiesh ’HEHZO (1:7) 5 sec
LPCVD
oxide
surface
15 Un dope d Nucleation Diamond Spinner 500rpm for 5
diamond powder sec, 2000mm
loaded for 30 sec
growth
water
(DW.
0.5%)
16 Diamond MPCVD 20h, 750°C ~3 um
growth
17 Dielectric PECVD PECVD 3 times (1 pm
deposition oxide @deposition)
18 Annealing 650°C RIP 10 mins

 

 

 

 

 

 

 

 

 

112

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

19 Annealing 750°C RTP 10 mins
20 Annealing 970°C RTP 3 mins
21 HF -dip Refiesh HFzHZO (1:7) 5 sec
PECVD
oxide
22 Doped Nucleation DW, Spinner 500rpm for 5
diamond
owth 3.75% sec, 2000rpm
gr for 30 sec
23 Diamond MPCVD 13h, 750°C, ~l um
growth TMB (2%)
23 Diamond PT790 10mins
etch
a
24 Spin 1827 3 pm Spinner 3000rpm, 30
sec
25 Soft bake 95 °C Hotplate 1 min
26 Expose B Define MA6 17 sec
D doped
P diamond
electrode
window
27 Develop MIF 318 1 min
28 Pattern Masking Al Enerjet
doped layer . . (1 m) Evaporator
diamond deposrtron 1‘
29 13"” Lift-off Acetone 1 h
30 RIE Plasma- Recipe:
CAO_NEW
Therm 790 (mainly 02
etch)
2 h
31 HF-dip Removing HFszO (l :7) 10 sec
diamond
whiskers
32 Al removal Al etchant

 

 

 

 

 

 

 

 

113

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

33 Etch Spin 9260 6 pm Spinner 4000mm, 30
backside sec
oxide
34 Etch BHF 15mins (until
hydrophobic
surface)
35 PR remove Acetone
36 Masking Al (1 pm) Enerjet
layer Evaporator
deposition
37 Spin 5214 1.6 pm Spinner 3000rpm, 30
sec
38 Soft bake 95 °C Hotplate l min
39 Expose MA6 4 sec
40 Hard bake 115 °C Hotplate 1:30 mins
41 Pattern Flood MA6 l min
undOped Expose
diamond
42 Develop A2300 40 sec
44 Mask etch Al etchant
Type A
45 Diamond Plasma-Therm Recipe: It also etches
etch 790 diamond (with the LPCVD
F), 3 h oxide
underneath.
Check and etch
until the expose
surface is
conducting.
46 Al removal Al etchant
47 Dielectric PECVD 1 11m PECVD 500nm @
deposition deposition
and
48 patterning Spin 9260 6 um Spinner 4000rpm, 30
sec
49 Soft bake 95 °C Hotplate 2 V2 min

 

 

 

 

 

 

 

 

114

 

 

50

 

 

51

 

 

52

 

 

53

 

 

54

 

 

55

 

 

56

 

 

57

 

 

58

 

 

59

 

60

 

 

 

61

 

62

 

 

 

63

 

64

 

 

 

65

 

 

66

 

 

67

 

 

 

Expose B doped MA6 35 sec
D diamond
P pattern
Develop AZ400: H20 2 min
(4:1)
RIE Plasma-Therm Recipe: P.S. if 65mins
790 n_Si02, 50 for 1.2 pm
mins PECVD
PR remove Acteone
Ashing March asher 100W, 100 sec
Spin 5214 1.6 pm Spinner 3000rpm, 30
sec
Soft bake 95 °C Hotplate 1 min
Expose U MA6 4 sec
D
P
Hard bake 115 0C Hotplate 1:30 mins
Flood MA6 1 min
Expose
Develop A2300 40 sec
RIE Plasma- Recipe:
Therm 790 n_SiOZ, 25
mins
PR remove Acteone
Ashing March asher 100W, 100 sec
Spin 5214 1.6 pm Spinner 3000rpm, 30
sec
Soft bake 95 °C Hotplate l min
Expose U MA6 4 sec
D
P
Hard bake 115 °C Hotplate 1:30 mins

 

 

 

 

 

 

115

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

68 Flood MA6 1 min
Expose
69 Develop A2300 40 sec
70 RIE Plasma-Therm Recipe:
790 n_Si02, 25
mins
71 PR remove Acetone
72 Ashing March asher 100W, 100 sec
73 Spin 9260 6 pm Spinner 4000rpm, 30
sec
74 Soft bake 95 °C Hotplate 2 ‘/2 min
75 Expose C Counter MA6 35 sec
E electrode
Pattern P
counter .
76 Develop AZ400: H20 2 mm
electrode (4. 1)
77 Metal Ti/Au SI 20
deposition
(50 nm/
400 nm)
78 Lift-off Acetone 2 h
79 Spin 9260 6 pm Spinner 4000rpm, 30
sec
80 Soft bake 95 °C Hotplate 2 1/2 min
81 Expose R Reference MA6 35 sec
E electrode
Pattern P
32 “fem“ Develop A2400: H20 2 min
electrode (4.1)
83 Metal Ti/Au/Ag Enerjet
deposition (10 nm/ 8 Sputter
nm/ 400
nm)
84 Lift-off Acetone Overnight

 

 

 

 

 

 

 

116

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

85 Spin 9260 6 pm Spinner 4000rpm, 30
sec
86 Soft bake 95 °C Hotplate 2 '/2 min
87 Expose I Lead MA6 35 sec
Pattern P pattern
88 metal Develop A2400: H20 2 min
intercon- (4:1)
nect
89 Metal Ti/Au S120
deposition
(50 nm/
400 nm)
90 Lift-off Acetone
91 PECVD 2 pm PECVD 500nm @
deposition
92 Spin 9260 6 pm Spinner 4000rpm, 30
sec
93 Soft bake 95 °C Hotplate 2 '/2 min
94 Expose T Oxide MA6 35 see
I insulator
Top oxide P
insulator
95 Develop AZ400: H20 2 min
(4:1)
96 RIE Plasma- Recipe:
Therm 790 n_SiOZ, 90
mins
97 PR remove Acetone
98 Ashing March asher 100W, 100 sec
99 Probe Wafer HF -Nitric 30 mins
1 . . g
l
100 re ease Release 115 °c EDP 4 h

 

 

 

 

 

 

 

117

 

 

B. Activation of Ir

1)

2)
3)
4)
5)

6)

Put 1 pack of P-5368 Phosphate Buffered Saline pH7.4 in 25mL of DI water. The
solution is ~0.3M.

Use “amperometric i-t curve” to set the -3.0V for 5 mins.

Clean the bubbles on the surface of Ir

Use “arnperometric i-t curve” to set the 2.5V for 5 mins.

Use “CV” to scan though -0.6 to 0.72 vs. Ag/AgCI. With 100mV/s for 100 cycles
(=200 segments)

Use “Chronocoulometry” to set pulse width = lsec, initE= -0.8 and finalE=0.7.

118

C. Equations

@eeAélheve f?

N

3+3

Va.

Aflwliam galls ..evaemf a...

 

Tee. ..r. r..- e.

 

AW vm8m<N+A§va8 .mN+..N +Nev5mm<N+m¢+~<+~m

PPSN + 3v 5». J l 552 1 3am I 05% 1 8V an £3.

E

 

 

 

 

A

@aaiinwmagfiA eN Iveem<N+¥+~<+~m

at.

 

@meeff waveaegeamllfleveaé a .meeveieaweveeetea

 

7:8 0

GD

(oon)B.

I.

ta 0

N

((DCdl)“ and B =

where A -'=

119

[1]

[2]

[3]

[4]

[5]

[6]

[7]

[8]

[9]

[10]

BIBLIOGRAPHY

David Eagleman, “10 unsolved mysteries of the brain”, Discover, August 2007.

N. B. Standen, P. T. A. Gray, and M. J. Whittaker, Microelectrode techniques,
The Plymouth workshop handbook, Cambridge, 1987.

K. D. Wise, J. B. Angel] and A Starr, “An integrated-circuit approach to
extracellular microelectrodes”, IEEE Trans. Biomed Eng, v. 17, no. 3, pp. 238-
247, 1970.

P. Norlin, M. Kindlundh, A. Mouroux, K. Yoshida and U. G. Hofmann, “A 32-
site neural recording probe fabricated by DRIE of SOI substrates,” 1 of
Micromech. & Microeng., pp. 414-419, 2002.

D. R. Kipke, “Implantable neural probe systems for cortical neuroprosthesis”,
Proc. of Int. Com of the IEEE Eng. in Med and Bio. Soc., pp. 5344-5347, 2004.

Q. Bai and K. D. Wise, “Single-unit neural recording with active microelectrode
arrays”, IEEE Trans. Biomed Eng, v. 48, pp. 911-920, 2001.

C. Pang, J. G. Cham, Z. Nenadic, S. Musallam, Y. C. Tai, J. W. Burdick and R A.
Andersen, “ A new multi-site probe array with monolithically integrated parylene
flexible cable for neural prostheses,” Proc. of Int. Conf of the IEEE Eng. in Med
andBio. Soc., pp. 7114- 7117, 2005.

J. Chen and K. D. Wise, “A multichannel neural probe for selective chemical
delivery at the cellular level”, IEEE Trans. Biomed Eng, v. 44, no. 8, pp. 760-769,
1997.

L. Lin and A. Pisano, “Silicon processed microneedles”, IEEE J. Micromech.
Syst, v. 8, no. 1, pp. 78-84, 1999.

J. Wang, M. Gulari, P. T. Bhatti, B. Y. Arcand, K. Beach, C. R. Fredrich and K. D.
Wise,, “A cochlear electrode array with built-in position sensing,” Proc. of Int.
Conf on MEMS, pp. 786-789, 2005.

120

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

Y. Tang, D. M. Aslam, J. Wang and K. D. Wise, “Poly-Crystalline Diamond
Piezoresistive Position Sensors for Cochlear Implant Probe”, Proc. of Solid-State
Sensors, Actuators and Microsystems, v.1, pp. 542-546, 2005.

Y. Yao, M. N. Gulari, J. A. Wiler and K. D. Wise, “A microassembled low-
profile three-dimensional microelectrode array for neural prosthesis applications”,
J. ofMicro. Mech. Syst., v. 16, no. 4, pp. 977-988, 2007.

C. Pang, J. G. Cham, Z. Nenadic, Y. C. Tai, J. W. Burdick and R. A. Andersen,
“A New Neural Recording Electrode Array with Parylene Insulating Layer”, Proc.
of Int. Conf on Miniaturized Syst. for Chemistry and Life Sciences, pp. 675-677,
2005.

E. Patrick, M. Ordonez, N. Alba, J. C. Sanchez and T. Nishida, “Design and
fabrication of a flexible substrate microelectrode array for brain machine
interfaces”, Proc. of IEEE EMBS, pp. 2966-2969, 2006.

Y. Kato, M. Nishino, I. Saito, T. Suzuki and K. Mabuchi, “Flexible intracortical
neural probe with biodegradable polymer for delivering bioactive components”,
Proc. of IEEE MS, pp. 143-146, 2006.

S. Takeuchi, T. Suzuki, K. Mabuchi and H. Fujita, “3D flexible multichannel
neural probe array”, .1. Micomech. Microeng, v. 14, pp. 104-107, 2004.

H. Lu, S. H. Cho, J. B. Lee, L. Cauller, M. R. Ortega and G. Hughes, “SU8-based
micro neural probe for enhanced chronic in-vivo recording of spike signals from
regenerated axons”, IEEE Int. Conf on Sensors, pp. 66-69, 2006.

M. D. Johnson, R. K. Franklin, K. A. Scott, R B. Brown and D. R. Kipke,
“Neural probes for concurrent detection of neurochemical and
electrophysiological signals in vivo”, Proc. of IEEE EMBS, pp. 7325-7328, 2005.

K. L. Drake, K. D. Wise, J. Farraye, D. J. Anderson and S. P. Bement,
“Performance of planar multisite microprobes in recording extracelluar single-unit
intracortical activity”, IEEE Trans. on Bio.Eng, v. 35, no. 9, pp. 719-732, 1988.

T. L. Rose, E. M. Kelliher and L. S. Robbless, “Assessment of capacitor
electrodes: preliminary evaluation,” J. of Neurosci. Methods, no. 12, pp. 181-193,
1985

121

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[23]

[29]

D. J. Anderson, K. Najafr, S. J. Tanghe, D. A Evans, K. L. Levy, J. F. Hetke, X.
Xue, J. J. Zappia and K. D. Wise, “Batch-fabricated thin-film electrodes for
stimulation of the central auditory system”, IEEE Trans. of Bio. Eng, v. 36, no. 7,
1989.

M. P. Hughes, K. Bustamante, D. J. Banks and D. J. Ewims, “Effects of electrode
size on the performance of neural recording microelectrodes”, IEEE Cory? on
Microtech. in Med And Bio., pp. 220-223, 2000.

J. A. Stamford, P. P alij, C. Davidson, C. M. Jorm and J. Millar, “Simultaneous
“real time” electrochemical and electrophysiological recording in brain slices with
a single carbon-fibre microelectrode”, J. of Neuroscience Methods, v. 50, pp. 279-
290, 1993.

L. S. Pan, D. R Kania, Diamond: Electronic Properties and Applications,
Springer, 1994.

M. Hupert, A. Muck, J. Wang, J. Stotter, Z. Cvackova, S. Haymond, Y. Show
and G. M. Swain, “Conductive diamond thin-films in electrochemistry,” Dia. and
Rel. Mat, v. 12, pp. 1940-1949, 2003.

AN. Ndao, F. Zenia, A Deneuville, M. Bernard, C. L. 1e'ment, “Effect of boron
concentration on the electrochemical reduction of nitrates on polycrystalline
diamond electrodes”, Dia. and Rel. Mat, v. 9, pp. 1175-1180, 2000.

A. Bennett, J. Wang, Y. Show and G. M. Swain, “Effect of sp2-bonded
nondiamond carbon impurity on the response of boron-doped polycrystalline
diamond thin-film electrodes”, J. of Electrochem. Soc., v. 151, no. 9, pp. 306-313,
2004.

C. E. Nebel, H. Kato, B. Rezek, D. Shin, D. Takeuchi, H. Watanabe and T.
Yamamoto, “Electrochemical properties of undoped hydrogen terminated CVD
diamond”, Dia. and Rel. Mat, v. 15, pp. 264-268, 2006.

G. M. Swain and R. Ramesham, “The electrochemical activity of boron—doped
polycrystalline diamond thin film electrode”, Anal. Chem, v. 65, pp. 345-351,
1993.

122

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

'L. Tang, C. Tsai, W. W. Gerberich, L. Kruckeberg and D. R Kania,

“Biocompatibility of chemical-vapour—deposited diamond”, Biomaterials, v. 16,
pp. 483-488, 1995.

R. A Freitas, Nanomedicine, Vol. IIA: Biocompatibility, Landes Bioscience, 2003.

E. W. Salzman, J. Linden, G. McManama and J. A. Ware, “Role of fibrinogen in
activation of platelets by artificial surfaces,” Annals N. Y. Acad Sci, v. 516, pp.
184-195, 1987.

A. Kraft, “Doped diamond: A Compact Review on a new, versatile, electro
material”, Int. J. Electrochem. Sci., v. 2, pp. 355-385, 2007.

J. Cvacka, V. Quaiserova, J. W. Park, Y. Show, A. Muck and G. M. Swain,
“Boron-doped Diamond Microelectrodes for Use in Capillary Electrophoresis
with Electrochemical Detection”, Anal. Chem, v.75, pp. 2678-2687, 2005.

J. M. Halpern, S. Xie, G. P. Sutton, B. T. Higashikubo, C. A. Chestek, H. Liu H.
J. Chiel, and H. B. Martin, “Diamond electrodes for neurodynamic studies in
Aplysia californica”, Dia. and Rel. Mat, v. 15, pp. 183-187, 2006.

J. Park, V. Q. Mocko, K. Peckova, J. J. Galligan, G. D. Fink and G. M. Swain,
“ Fabrication, characterization, and application of a diamond microelectrode for
electrochemical measurement of norepinephrine release from the sympathetic
nervous system”, Dia. and Rel. Mat, v. 15, pp. 761-772, 2006.

P. T. Bhatti, B. Y. Arcand, J. Wang, N. V. Butala, C. R. Friedrich and K. D.
Wise,”A high-density electrode array for cochlear prosthesis”, Proc. of Int. Conf
on Solid State Sensors, Actuators and Microsystems, pp. 1750-1753, 2003.

P. Norlin, M. Kindlundh, A. Mouroux, K. Yoshida, W. Jensen, U. G. Hofmann,
“A 32-site neural recording probe fabricated by double-sided deep reactive ion
etching of silicon-on-insulator substrates”, Proc. of the Micromechanics Europe
Workshop, 2001.

J. Wang, M. Gulari, P. T. Bhatti, B. Y. Arcand, K. Beach, C. R. Friedrich and K.
D. Wise, “A cochlear electrode array with built-in position sensing,” Proc. of Int.
Conf on Micro Electro Mechanical Systems, pp. 786-789, 2005.

123

 

[40]

[41]

[42]

[43]

I44]

[45]

[46]

[47]

[43]

[49]

Y. Tang, D. M. Aslam, J. Wang and K. D. Wise, “Poly-Crystalline Diamond
Piezoresistive Position Sensors for Cochlear Implant Probe”, Proc. of Solid-State
Sensors, Actuators and Microsystems, v.1, pp. 542-546, 2005.

C. Pang, “Parylene technology for neural probes applications”, PhD thesis,
California Institute of Technology, 2008.

C. Pang, J. G. Cham, Z. Nenadic, S. Musallam, Y. C. Tai, J. W. Burdick and R. A.
Andersen, “ A new multi-site probe array with monolithically integrated parylene
flexible cable for neural prostheses,” Proc. of Int. Conf of the IEEE Eng. in Med
andBio. Soc., pp. 7114- 7117, 2005.

S. Takeuchi, T. Suzuki, K. Mabuchi and H. Fujita, “3D flexible multichannel
neural probe array,” J. of Micromech and Microeng, v.14, pp. 104-107, 2004.

D. R. Kipke, “Implantable neural probe systems for cortical neuroprosthesis,”
Proc. of Int. Cont: of the IEEE Eng in Med and Bio. Soc, 5344-5347, 2004.

Q. Bai and K. D. Wise, “Single-unit neural recording with active microelectrode
arrays,” IEEE Trans Biomed Eng, v. 48, pp. 911-20, 2001.

P. Norlin, M. Kindlundh, A. Mouroux, K. Yoshida, W. Jensen, U. G. Hofmann,
“A 32-site neural recording probe fabricated by double-sided deep reactive ion
etching of silicon-on-insulator substrates,” Proc. of the Micromechanics Europe
Workshop, 2001.

P. K. Campbell, K. E. Jones, R. J. Huber, K. W. Horch and R A. Normann, “A
silicon-based three-dimensional neural interface: manufacturing process for an
intracortical electrode array,” IEEE Trans. Biomed Eng, v. 38, no. 8, pp.758-768,
1991.

M. K. Kindlundh, P. Norlin, U. G. Hofinann, “A neural probe process enabling
variable electrode configurations,” Sensors and Actuators B: Chemical, v. 102, pp.
51-58, 2004.

T. H. Yoon, E. J. Hwang, D. Y. Shin, S. J. Parl, S. J. Oh, S. C. Jung, H. C. Shin
and S. J. Kim, “A micromachined silicon depth probe for mulitchannel neural
recording,” IEEE Trans. Biomed Eng, v. 47, no. 8, pp. 1082-1087, 2000.

124

[50]

[51]

[52]

[53]

[54]

[55]

[56]

[57]

[53]

[59]

W. L. C. Rutten, H. J. vanWier and J. H. M. Put, ”Sensitivity and selectivity of
intranueral stimulation using a silicon electrode array,” IEEE Trans. Biomed. Eng,
v. 38, no. 2, pp.192-198, 1991.

D. T. Kewley, M. D. Hills, D. A. Borkholder, I. E. Opris, N. I. Maluf, C. W.
Storrnent and G. T. A Kovacs “Plasma-etched neural probes,” Sens. Actuators A,
v. 58, pp. 27-35, 1997.

K. Chueng, L. Gun, K. Djupsund, D. Yang and L. P. Lee, “A new neural probe
using SOI wafers with topological interlocking mechanism,” Proc. of IEEE Int.
Conf on Microtechnologies in Medicine & Biology, pp. 507-511, 2000.

P. S. Motta and J. W. Judy, “Multielectrode microprobes for deep-brain
simulation fabricated with a customizable 3-D electroplating process,” IEEE Tans.
0n Biomed Eng, v. 52, no. 5, 2005.

J. Wu, L. Yan, W. C. Tang and F. G. Zeng, “Micromachined electrode arrays with
forrn-fitting profile for auditory nerve prostheses,” Proc. of IEEE Eng in
Medicine and Biology, 2005.

K. D. Wise, A. M. Sodagar, Y. Yao, M. N. Gulari, G. E. Perlin and K. Najafi,
“Microelectrode, microelectronics, and implantable neural microsystems”, Proc.
ofthe IEEE, v. 96, no. 7, pp. 1184-1202, 2008.

P. S. Motta and J. W. Judy, “Multielectrode microprobes for deep-brain
simulation fabricated with a customizable 3-D electroplating process,” IEEE Tans.
on Biomed Eng, v. 52, no. 5, 2005.

G. T. A. Kovacs, C. W. Storrnent and J. M. Rosen, “Regeneration microelectrode
array for peripheral nerve recording and stimulation, ” IEEE Trans. on
BiomedEng, v. 39, no. 9, pp. 893-902, 1992.

M. D. Johnson, R. K. Franklin, K. A Scott, R B. Brown, D. R Kipke, “Neural
probes for concurrent detection of neurochemical and electrophysiological signals
in vivo”, Prof of IEEE Eng In Med And Bio, pp. 7325-7328, 2005.

R. K. Franklin, M. D. Johnson, K. A. Scott, J. H. Shim, H. Nam, D. R Kipke and
RB. Brown, “Iridium oxide reference electrodes for neurochemical sensing with
MEMS microelectrode arrays”, IEEE Sensors, pp. 1400-1403, 2005.

125

[60]

[61]

[62]

[63]

[64]

[65]

[66]

[67]

[63]

[69]

H. Lu, S. H. Cho, J. B. Lee, L. Cauller, M. R Ortega and G. Hughes, “SU8-based
micro neural probe for enhanced chronic in-vivo recording of spike signals fiom
regenerated axons”, Proc. of IEEE SENSORS, pp. 65-69, 2006.

Y. Kato, M. Nishino, I. Saito, T. Suzuki and K. Mabuchi, “Flexible intracortical
neural probe with biodegradable polymer for delivering bioactive components”,
Proc. of IEEE Microtech in Med And Bio., pp. 143-146, 2005.

J. C. Sanchez, N. Alba, T. Nishida, C. Batich and P. R Carney, “Structural
modifications in chronic microwire electrodes for cortical neuroprosthetics: A
case study”, IEEE Trans. on Neural systems and rehabilitation Eng, pp. 217-221,
2006.

H. P. Neves and P. Ruther, “The neuroprobes project”, Proc. of the IEEE Ell/BS,
pp. 6442-6444, 2007.

J. Chen and K. D. Wise, “A multichannel neural probe for selective chemical
delivery at the cellular level,” IEEE Trans. Biomed Eng, v. 44, no. 8, pp. 760-769,
1997.

L. Lin and A P. Pisano, “Silicon processed microneedles,” IEEE J. Micromech.
Syst, v.8, no.1, pp. 78-84, 1999.

N. H. Talbot and A. P. Pisano, “Polymolding: two wafer polysilicon
micromolding of closed-flow passages for microneedles and microfluidic
devices,” T ech. Dig. IEEE Solid-State Sens. Actuator Workshop, pp. 265-268,
1998.

J. D. Brazzle, I. Papautsky, A. B. Frazier, “Fluid-coupled hollow metallic
micromachined needle arrays, ” Proc. SPIE Conf Microfluidic Dev. Syst, pp. 1 16-
124, 1998.

AJ. Bard and L. R. Faulkner, Electrochemical methods: Fundamentals and
applications, John Wiley & Sons. Inc., 2001.

J. Wang, Analytical Electrochemistry, Wiley- VCH, 2000.

126

[70]

[71]

[72]

[73]

[74]

[75]

[76]

[77]

[73]

[79]

D. T. Fagan, I. F. Hu and T. Kuwana, “Vaccum heat treatment for activation of
glassy carbon electrodes,” Anal. Chem, v. 57, pp. 2759-2763, 1985.

G. M. Swain, “The susceptibility of surface corrosion in acidic fluoride media: a
comparison of diamond, HOPG, and glassy carbon electrodes,” J. Electrochem.
Soc., v. 141, no. 12, pp. 3382-3393, 1994.

T. N. Rao, I. Yagi, T. Miwa, D. A Tryk and A. Fujishinma, “Electrochemical
oxidation of NADH at highly boron-doped diamond electrodes,” Anal. Chem, v.
71, pp. 2506-2511, 1999.

R. Declements and G. M. Swain, “The formation and electrochemical activity of
microporous diamond thin film electrodes in concentrated KOH,” J. Electrochem.
Soc., v. 144, pp. 856-866, 1997.

Y. V. Pleskov, “Synthetic diamond in electrochemistry,” Russian Chemical
Reviews, v. 68, no. 5, pp. 381-392, 1999.

H. B, Martn, A. Argoitia, U. Landau, A. B. Anderson and J. C. Angus, “Hydrogen
and oxygen evolution on boron-doped diamond electrodes,” J. Electrochem. Soc,
v. 143, no. 6, pp 133-136, 1996.

P. A. Michaud, E. Mahe, W. Haenni, A Perret and C. Cominellis, “Preparation of
peroxodisulfirric acid using boron-doped diamond thin film electrodes”,
Electrochem. Solid-State Lett., v. 3, pp.77-79, 2000.

J. Wang, G. Chen and M. P. Chatrathi, “Microchip capillary electrophoresis
coupled with a boron-doped diamond electrode-based electrochemical detector,”
Anal. Chem, v. 75, pp. 935-939, 2003.

B. V. Sarada, T. N. Rao, D. A Tryk and A Fuhishima, “Electrochemical
oxidation of histamine and serotonin at highly boron-doped diamond electrodes,”
Anal Chem, v. 72, pp. 1632-1638, 2000.

O. Herlambang, B. V. Sarada, T. N. Rao and A. Fujishima, “Miniaturization of
diamond microsensor system for in vivo detection”, Proc. of the Chemical Sensor,
v. 18, pp. 121-123, 2002.

127

[30]

[81]

[82]

[83]

[84]

[35]

[36]

[37]

[83]

[89]

S. Xie, G. Shafer, C. G. Wilson and H. B. Martin, “ In vitro adenosine detection
with a diamond-based sensor,” Diamond & related material, v. 15, pp. 225-228,
2005.

J. M. Halpern, S. Xie, G. P. Sutton, B. T. Higashikubo, C. A. Chestek, H. Lu, H. J.
Chiel and H. B. Martin, “Diamond electrode for neurodyanmice studies in
Aplysia calofornica,” Diamond & related material, v. 15, pp. 183-187, 2006.

A. K. Engel, C. K. E. Moll, I. Fried and G. A. Ojemann, “Invasive recordings
from the human brain: clinical insights and beyond”, Nature review: neuroscience,
v.6,pp.35-47,2005.

P.Mitra, C. R. Keese and I. Giaever, “Electrical measurements can be used to

monitor the attachment and spreading of cells in tissue culture”, BioTechniques, v.
11, pp. 504-510, 1991.

Y. Jimbo and A. Kawanna, “Electrical stimulation of cultured neural cells by
planar electrode array’ Proc. of IEEE Eng. in Med And Bio. Soc., v. 12, pp.
1741-1742, 1990.

G. W. Gross, E. Rieske, G. W. Kreutzberg and A. Meyer, “A new fixed-array
multi-electrode system designed for long-term monitoring of extracellular single
unit neuronal activity in vitro”, Neuroscience Lett. V. 6, pp. 101-105, 1977.

J. E. B. Randles, “Kinetics of rapid electrode reactions. Part 3. — electron
exchange reactions,” KW Somerton Trans. of the Faraday Society, v. 48, pp. 937-
950, 1952.

http://www.lhsc.on.ca/critcare/ucicu/educate/orient/neuroap.htrn

D. T. Kewley, M. D. Hills, D. A Borkholder, I. E. Opris, N. I. Maluf, C. W.
Storrnent, J. M. Bower and G. T. A. Kovacs, “Plasma-etched neural probe”,
Sensors and Actuators A, v. 58, pp. 27-35, 1997.

www.plexoninc.com

128

[90]

[91]

[92]

[93]

[94]

[95]

[96]

[97]

[93]

[99]

K. D. Wise, A. M. Sodagar, Y. Yao, M. N. Gulari, G. E. Perlin and K. Najafi,
“Microelectrode, microelectronics, and implantable neural microsystems”, Proc.
of the IEEE, v. 96, no. 7, pp. 1184-1202, 2008.

C. Nicholson, “Diffusion and related transport mechanisms in brain tissue”, Rep
Prog Phys, v. 64, pp. 815-884, 2001.

D. R. Wur, J. L. Davidson, W. P. Kang and D. L. Kinser, “Polycrystalline
diamond pressure sensor”, J. ofllflZMS, v. 4, no. 1, pp. 34-41, 1995.

X. Zhu, D. M. Aslam, Y. Tang, B. H. Stark and K. Najafi, “The Fabrication of all-
diamond packaging panels with built-in interconnects for wireless integrated
microsystems”, J. of MS, v. 13, no. 3, 2004.

A. Aleksov, M. Kubovic, M. Kasu, P. Schmid, D. Grobe, S. Ertl, M. Schreck, B.
Stritzker and E. Kohn, “Diamond-based electronic for RF applications”, Diamond
& Related Materials, v. 13, pp. 233-240, 2004.

S. Sahli, Electronic characterization and fabrication of CVD diamond

piezoresistive pressure sensors, Ph. D Dissertation, Michigan State University,
1997.

T. Ando, X. Li, S. Nakao, T. Kasai, M. Shikida and K. Sato, “Effect of crystal
orientation fracture strength and fi'acture toughness of single crystal silicon,” Proc.
of Int. Cord on W48, pp. 177-180, 2004.

T. Ando, X. Li, S. Nakao, H. Tanska, M. Shikida, K. Sato, ”Fracture toughness
measurement of thin-film silicon,” Fatigue & fracture of engineering materials
and structure, v. 28, no. 8, pp. 687-694, 2005.

S. Kamiya, M. Sato, M. Saka and H. Abe, “Fracture toughness estimation of thin
chemical vapor deposition diamond films based on the spontaneous fracture
behavior on quartz glass substrates,” J. Appl. Phys, v. 81, no. 12, pp. 6056—6061,
1997.

M. D. Drory. R. H. Dauskardt, A. Kant and R. O. Ritchie, “Fracture of synthetic
diamond”, J. Appl. Phys, v. 78, pp. 3083-3088, 1995.

129

[100]

[101]

[102]

[103]

[104]

[105]

[106]

[107]

[108]

[109]

[110]

[111]

I. Dion, C. Baquey, J .R. Monties, “Diamond: the biomaterial of the let century?”
Int. J. Artrf Organs, v.16, 623-627, 1993.

R. F. Davis, Diamond films and coatings: Development, properties, and
applications, Noyes Publications, pp. 1-30, 1993.

D. P. Dowling, P. V. Kola, K Domelly and T. C. Kelly, “Evaluation of diamond-
like carbon coated orthopaedic implants”, Dia. Rel. Mat, no. 6, pp. 380-393, 1997 .

S. Afi’atato, M. Frigo, A Toni, “An in vitro investigation of diamond-like carbon
as a femoral head coating”, J. Biomed Mat. Res, no. 53, pp. 221-226, 2000.

T. Tatsuma, H. Mori and A. Fujisjhima, “Electron transfer from diamond
electrode to heme peptide and peroxidase”, Ana. Chem, v. 72, pp. 2919-2924,
2000.

W. Yang, J. E. Butler, J. N. Russell and R J. Hamers, “Direct electrical detection
of antigen-anitbody binding on diamond and silicon substrates using electrical
impedance spectroscopy”, The Analst, v. 132, pp. 296-306, 2007.

A. Lettington and J. W Steeds, Thin film diamond, Chapman and Hall, pp. 117-
125, 1994.

M. Pinneo, “Diamond growth: Today and tomorrow”, Proc. of the Ist general
conf on Nanotechnology, pp. 147-172, 1995.

L. Tang, C. Tsai, W.W. Gerberich, L. Kruckeberg, D.R Kania, “Biocompatibility
of chemical-vapor-deposited diamond,” Biomaterials, v. 16, pp. 483-488, 1995.

J. Grabarczk, D. Batory, P. Louda, P. Couvrat, I. Kotela and K. Bakowicz-Mitura,
“Carbon coatings for medical implants”, J. of Achievements in Mat. and Manuf
Eng, v. 20, pp. 107-110, 2007.

R. A. Freitas, Nanomedicine, Vol 11A: Biocompatibiltiy, Landes Bioscience, 2003.

R. Hauert, “A review of modified DLC coatings for biological applications”, Dia.
and Rel. Mat, v. 12, pp. 583-589, 2003.

130

[112]

[113]

[114]

[115]

[116]

[117]

[118]

[119]

[120]

[121]

A. Grill, “Diamond-like carbon coatings as biocompatible materials-an overview”,
Dia. and Rel. Mat, v.12 , pp. 166-170, 2003.

M. Amaral, P. S. Gomes, M. A. Lopes, J. D. Santos, RF. Silva and M. H.
F ermandes, “Nanocrystaline diamond as a coating for joint implants: cytotoxicity
and biocompatibility assessment”, J. of Nanomaterials, accepted, 2008.

G. M. Swain, A. B. Anderson, J. C. Augus, “Application of diamond thin films in
electrochemistry,” MS bulletin, v.23, no. 9, pp. 56-60, 1998.

J. A. Bennett, J. Wang, Y. Show and G. M. Swain, “Effect of sp2-bonded
nondiamond carbon impurity on the response of boron-doped polycrystalline
diamond thin-film electrodes”, J. of the Electrochem. Soc., v. 151, no.9, pp. 306-
313, 2004.

J. Xu, Q. Chen and G. M. Swain, “Anthraquinedisulfonate electrochemistry: A
comparison of glassy carbon, hydrogenated glassy carbon, highly oriented
pyrolytic graphite, and diamond electrodes,” Anal. Chem, v. 70, pp. 3146-3154,
1998.

A Fujishima, Y. Einaga, T. N. Rao and D. A. Tryk, Diamond electrochemistry,
Elasevier, 2005.

A. A Morrish and P. E. Pehrsson, "Effects of Surface Pretreatments on
Nucleation and Growth of Diamond Films on a Variety of Substrates," Appl. Phy.
Lett., vol. 59, pp. 417-419, 1991.

L. Demuynck, J. C. Amault, C. Speisser, R. Polini and F. LeNormand,
"Mechanisms of CVD diamond nucleation and growth on mechanically scratched
and virgin Si(100) surfaces," Diamond and Related Materials, vol. 6, pp. 23 5-239,
1997.

P. A Dennig, H. Shiomi, D. A. Stevenson and N. M. Johnson, "Influence of
Substrate Treatments on Diamond Thin-Film Nucleation," Thin Solid Films, vol.
212, pp. 63-67, 1992.

E. J. Bienk and S. S. Eskildsen, "The Effect of Surface Preparation on the
Nucleation of Diamond on Silicon," Diam. and Rel. Mat, vol. 2, pp. 432-437,
1993.

131

[122]

[123]

[124]

[125]

[126]

[127]

[128]

[129]

[130]

[131]

F. S. Lauten, Y. Shigesato and B. W. Sheldon, "Diamond Nucleation on
Unscratched Si02 Substrates," App]. Phy. Lett., vol. 65, pp. 210-212, 1994.

M. J. Chiang and M. H. Hon, "Positive dc bias-enhanced diamond nucleation with
high CH4 concentration," Diam. andRe]. Mat, vol. 10, pp. 1470-1476, 2001.

M. D. Irwin, C. G. Pantano, P. Gluche and E. Kohn, "Bias-enhanced nucleation of
diamond on silicon dioxide," App]. Phy. Lett.,, vol. 71, pp. 716-718, 1997.

A Masood, M. Aslam, M. A Tamor and T. J. Potter, "Techniques for Patterning
of CVD Diamond Films on Non-Diamond Substrates," J. of the Electrochemical
Society, vol. 138, pp. L67-L68, 1991.

G. S. Yang, M. Aslam, K. P. Kuo, D. K. Reinhard and J. Asmussen, "Effect of
Ultrahigh Nucleation Density on Diamond Growth at Different Growth-Rates and
Temperatures," J. of Vacuum Science & Technology B, vol. 13, pp. 103 0-103 6,
1995.

A. P. Malshe, R. A. Beera, A. A. Khanolkar, W. D. Brown and H. A. Naseem,
"Initial results of a novel pre-deposition seeding technique for achieving an ultra-
high nucleation density for CVD diamond growth," Diam. and Re]. Mat vol. 6, pp.
430-434, 1997.

Y. X. Tang and D. M. Aslam, Polycrystalline diamond technology and
piezoresistive sensor application for cochlear prosthesis, Ph D. Thesis, Michigan
State University, 2006.

K. Herrnansson, Ulf Lindberg, Bertill Hok and Goran Plamskog, “Wetting
properties of silicon surfaces”, IEEE Int. Conf on Solid-state Sensors and
Actuators, pp. 193-196, 1991.

E. Snidero, D. Tromson, C. Mer, P. Bergonzo, J. S. Foord, C. Nebel, et al.,
"Influence of the postplasma process conditions on the surface conductivity of
hydrogenated diamond surfaces," J. of Applied Physics, vol. 93, pp. 2700-2704,
2003.

J. Ristein, M. Riedel, F. Maier, B. F. Mantel, M. Starnmler and L. Ley, "Surface
doping: a special feature of diamond," J. of Physics-Condensed Matter, vol. 13,
pp. 8979-8987, 2001.

132

[132]

[133]

[134]

[135]

[136]

[137]

[138]

[139]

[140]

[141]

M. I. Lanstrass and K. V. Ravi, “Resistivity of chemical vapor deposited diamond
films”, Appl. Phys. Lett., v. 55, no. 10, pp. 975-977, 1989.

B. J. Lee, B. T. Ahn, J. K. Lee, Y. J. Baik, “A study on the conduction path in
undoped polycrystalline diamond films”, Diamond and Related Materials, v. 10,
App]. Phys. Lett., pp. 2174-2177, 2001.

M. C. Rossi, “Micro-raman fingerprint of grain boundry in [100] oriented
diamond films stress distribution and diamond phases”, App]. Phy. Lett., v. 73, no.
9, 1998.

Y. Ando, Y. Niishibayashi, K. Kobashi, T. Hirao and K. Oura, “Smooth and high-
rate reactive ion etching of diamond”, Diam. and Re]. Mat, v. 11, pp. 824-827,
2002.

1. Bell, M. K. Fung, W.J Zhang, K. H. Lai, Y. M. Wang, Z. F. Zhou, R. K. W. Yu,
C. S. Lee and S. T. Lee, “Effects at reactive ion etching of CVD diamond”, Thin
Solid Films, v. 368, pp. 222-226, 2000.

G. F. Ding, H. P. Mao, Y. L. Cao, Y. H. Zhang, X. Yao and X. L. Zhao,
“Micromachining of CVD diamond by RIE for MEMS applications”, Dia. and
Re]. Mat, v. 14, pp. 1543-1548, 2005.

O. Dorsch, M. Werner and E. Oberrneier, “Dry etching of undoped and boron
doped polycrystalline diamond films”, Diam. and Re]. Mat, v. 4, pp. 456-459,
1995.

Z. L. Cao, H. Y. Chan and D. M. Aslam, “Diamond technology”, in preparation.

T. Yamada, H. Yoshikawa, H. Uetsuka, S. Kumaragurubaran, N. Tokuda and S.
Shikata, “Cycle of two-step etching process using ICP for diamond MEMS
applications”, Diam. and Re]. Mat, v. 16, pp. 996-999, 2007.

H. —Y. Chan, D. M. Aslam, S. Wang, G. M. Swain and K. D. Wise, “Fabrication
and testing of a novel all-diamond neural probe for chemical detection and
electrical sensing applications,” Proc. of Int. Conf on AfliMS, pp. 244-247, 2008.

133

[142]

[143]

[144]

[145]

[146]

[147]

[148]

[149]

[150]

[151]

Y. Tang and D. M. Aslam, “Technology of polycrystalline diamond thin films for
microsystems applications”, J. Vac. Sci. T ech., V. 3, pp. 1088-1095, 2005.

B. J. Polk, A. Stelzenmuller, G. Mijares, W. MacCrehan and M. Gaitan,
“Ag/AgCl microelectrodes with improved stability for micrfoluidics”, Sensors
and Actuators B, v. 114, pp. 239-247, 2006.

T. Matsumoto, A. Ohashi and N. Ito, “Development of a micro-planar Ag/AgCl

quasi-reference electrode with long-term stability for an amperometric glucose
sensor”, Anal. Chim. Acta, v. 462, pp. 253-259, 2002.

J. S. Lee, S. D. Lee, G. Cui, J. J. Lee, J. H. Shin, G. S. Cha and H. Nam,
“Hydrophilic polyurenthane coated silver/silver chloride electrode for the
determination of chloride in blood”, Electroanalysis, v. 11, pp. 260-267, 1999.

L. Y. Huang, R. S. Huang and L. H. Lo, “Improvement of integrated Ag/AgCl
thin-film electrodes by KCl-gel coating for ISFET applications”, Sens. Actuators
B, v. 94, pp 53-64, 2003.

M. A. Nolan, S. H. Tan and S. P. Kounaves, “Fabrication and characterization of
a solid state reference electrode for electroanalysis of natural waters with
ultrarnicroelectrode”, Anal. Chem. v. 69, pp. 1244—1247, 1997.

A J. Bard and L. R. Faulkner, Electrochemical methods: fimdamentals and
applications, Wiley Publishers, 2000.

J. Xu, Q. Chen and G. M. Swain, “Anthraquinonedisulfonate electrochemistry: A
comparison of glassy carbon, hydrogenated glassy carbon, highly oriented
pyrolytic graphite, and diamond electrodes,” Anal. Chem, v. 70, pp. 3146-3154,
1998.

R. Ramesham and M. F. Rose, “Cyclic voltammetric, a. c. and d. c. polarization
behavior of boron-doped CVD diamond,” Thin Solids Films, v. 300, pp. 144-153,
1997 .

A Fujishma, Y. Einaga, T. N. Rao and D. A Tryk, Diamond electrochemistry,
Elsevier, 2005.

134

[152] J. D. Weiland, “Electrochemical properties of iridium oxide stimulating
electrodes,” Ph. D. dissertation, University of Michigan, 1997.

[153] L. S. Robbless, J. L. Leflco and S. B. Brummer, “Activated Ir: An electrode
suitable for reversible charge injection in saline solution,” J. Electrochem. Soc., v.
130, no. 3, pp. 731-732, 1983.

[154] V. S. Polikov, P. A. Tresco, W. M. Reichert, “Response of brain tissue to
chronically implanted neural electrodes,” J. of Neuroscience Method, v. 148, pp.
1-18, 2005.

[155] G. M. Swain, Lecture note, Michigan State University.

135

 

 

    

 
 

AAAAAAAAAA

lll]llllll llllllllnl‘ll llllllllll