University of Southern California Dissertations and Theses

The electrochemical evaluation of Parylene-based electrodes for neural applications

(USC Thesis Other)

The electrochemical evaluation of Parylene-based electrodes for neural applications

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

The Electrochemical Evaluation of
Parylene-based Electrodes for Neural
Applications

Dissertation by

Seth Aogu Hara

In Partial Fulfillment of the Requirements
for the Degree of

Doctor of Philosophy

University of Southern California

Los Angeles, California

2015

PRESENTED TO THE FACULTY OF THE USC VITERBI SCHOOL OF ENGINEERING
AUGUST 2015

© 2015

Seth A. Hara

All Rights Reserved
-II-

To Nicole
-III-

ACKNOWLEDGEMENTS

I am truly blessed to have had the opportunity to do this work and to have learned
all that I have over the course of my Ph.D. tenure. I thank God for providing me with this
opportunity and for the support system that has sustained me through it all. This has been
a long road and I did not always know where I was going or how I was going to get there,
but with the help and guidance of many amazing individuals, I somehow arrived. I would
like to take the time to express my gratitude to those people here.
First and foremost, I thank Dr. Ellis Meng for her mentorship over these past six
years. Ellis took me as a completely novice researcher and patiently guided me along the
Ph.D. path, allowing me to stumble and learn on my own, but giving me a stronger nudge
in the right direction when necessary. Ellis facilitated a lab culture in which the work
truly was a group effort and each of us was asked to step up and claim ownership of a
part of the project. When someone encountered an obstacle, however, everyone else
chipped in to help. Knowledge and skills were openly shared and likewise the
accomplishments and victories of one were shared by all. Perhaps most importantly, Ellis
served as an ideal role model with her incredible work ethic, critical analysis of issues,
and her penchant for open collaboration with other researchers in order to bring the best
minds to the table to tackle a project.
I am very grateful to the other faculty members who have helped me along the
way. To Dr. James Weiland, Dr. Malancha Gupta, and Dr. Krishna Nayak, who have
advised and mentored me over the years, thank you. I especially thank Dr. Victor Pikov
from the Huntington Medical Research Institutes for his mentorship and for sharing his
extensive experience in neural engineering with me.
The members of the Biomedical Microsystems Laboratory have been invaluable
to me and I would still be stuck in my first year struggling to run my first successful CV
if it were not for them. They are a talented and dedicated group of colleagues to work
alongside, but over the years we have also become something of a family, celebrating
each other’s achievements, supporting each other through difficulties, and laughing
together every step of the way. For their friendship and for their support, I thank them. Dr.
-IV-

Christian Gutierrez has always been a steadfast mentor, fielding my frantic questions and
pleas of desperation when things were not working and working through the problem
with me so that we could discover the answer together. My senior labmates, Dr. Heidi Tu
and Dr. Jonathan Kuo, were valuable founts of knowledge offering their expertise and
experience to me when I needed it. I could always count on Dr. Curtis Lee to provide a
unique perspective on things, both in lab and at the lunch table, and I have always valued
the many conversations and discussions Curtis and I had. Dr. Brian Kim’s confidence in
the success of our projects and in each of us individually was a constant source of
encouragement and inspiration as was his ingenuity and, of course, his surgeon-steady
hands. I thank Lawrence Yu, my co-computer administrator, for all the help keeping the
technology in the lab up and running, in spite of the printer’s ceaseless quest to never
print another page, and for geeking out with me about our planned Arduino projects.
Angelica Cobo’s dedicated work ethic never gets in the way of her generous heart and
helping hand. I will miss our weekly racquetball sessions and Game of Thrones/The
Walking Dead debriefing sessions. As I leave the lab, I am encouraged by the promise
shown by the newbies: Alex Baldwin, Jessica Lizbeth Ortigoza-Díaz, and Ahuva
Weltman. They each show a desire to learn, a tenacity to overcome the obstacles they
will surely face, and most importantly, some great personalities to keep the work fun.
Welcome to the family. Lastly, I have to especially acknowledge Dr. Roya Sheybani. As
a fellow researcher, lab social coordinator, roomie, and friend, there is no equal. She is a
brilliant engineer and a treasured friend. Mengsters, thank you for everything.
I would like to thank the many administrators in the Department of Biomedical
Engineering who have supported me both as a student and as a researcher. To
Mischalgrace Diasanta, Sandra Johns, Karen Johnson, and Daisy Rusli, thank you. I must
also acknowledge Dr. Tuan Hoang, who managed so much of the lab behind-the-scenes
and was always so willing to help troubleshoot problems, and Dr. Donghai Zhu, who
tirelessly managed and maintained the cleanroom facilities.
To my family, whose faith in me has never faltered, I am forever indebted. The
support of Rachel and Nathan has always pushed me forward, even in the most difficult
times. When I was unsure, they gave me the confidence to keep my eyes on the prize. My
parents, trailblazers in their own right, have given me such very high shoulders upon
-V-

which to stand. My father, who has always given me the perfect example of what it
means to be strong and to tenaciously strive for my goals, has always been a touchstone
for me and, when I need it, a wise counselor. My mother, whose gentle spirit gave me
comfort when I needed it and whose inner strength showed me how to be true to myself,
had unwavering confidence in my success. Her pride in me pushed me to be my best and
I pray that I will always give her reason to look upon me with pride.
To Nicole, my partner, my wife, and my love, I dedicate this dissertation.
Whether right by my side or on the other side of the world, she has been my greatest
supporter and loudest cheerleader. She doled out tough love when I needed it and always
helped be back to my feet when I got knocked down. I owe so much of who I am and
what I have done to her. Thank you.

-VI-

TABLE OF CONTENTS

Acknowledgements ......................................................................................................... IV
Table of contents ........................................................................................................... VII
List of Tables .................................................................................................................... X
List of Figures .................................................................................................................. XI
Abstract ....................................................................................................................... XVII
Chapter 1 Introduction..................................................................................................... 1
1.1 The study of electrical activity in the brain .......................................................... 1
1.2 Intracortical neural probes .................................................................................... 3
1.2.1 Microwires ......................................................................................................... 4
1.2.2 MEMS neural probes ......................................................................................... 4
1.2.3 Limitations of current technology ...................................................................... 5
1.3 Strategies to improve intracortical recording reliability .................................... 5
1.3.1 Strategies for non-biological failure modes ....................................................... 5
1.3.2 Strategies for biological failure modes .............................................................. 7
1.4 Parylene C.............................................................................................................. 12
1.4.1 Development of Parylene ................................................................................. 12
1.4.2 Usage in medical implants and neural probes.................................................. 13
1.4.3 Parylene C adhesion issues .............................................................................. 14
1.4.4 Techniques to analyze Parylene-based electrodes ........................................... 16
1.5 Electrochemical Background and Theory .......................................................... 16
1.5.1 Understanding the electrode-electrolyte interface ........................................... 16
1.5.2 Electrochemical techniques ............................................................................. 22
1.6 Objectives............................................................................................................... 24
References .................................................................................................................... 25
Chapter 2 Parylene Sheath Electrode: Design, Fabrication, and Testing ................. 33
2.1 Design of the Parylene Sheath Electrode ............................................................ 33
2.2 Fabrication of the PSE.......................................................................................... 34
2.2.1 Microfabrication .............................................................................................. 34
2.2.2 3D sheath formation ......................................................................................... 35
2.3 Electrochemical techniques for neural electrode testing ................................... 36
2.3.1 Basic electrochemistry for neural electrodes ................................................... 36
2.3.2 Electrochemical impedance spectroscopy ....................................................... 37
2.3.3 Cyclic voltammetry .......................................................................................... 45
2.3.4 Fully-formed PSE ............................................................................................ 49
-VII-

2.4 Biofunctional coatings .......................................................................................... 50
2.4.1 EC effects of biofunctional coatings ................................................................ 50
2.4.2 Body temperature soak of coated probes ......................................................... 52
2.4.3 Biofunctional coating: analysis and conclusions ............................................. 53
2.5 Electrochemical effects of ethylene oxide sterilization ...................................... 53
2.6 Discussion and conclusions .................................................................................. 54
References .................................................................................................................... 55
Chapter 3 Reliability Testing of the Parylene Sheath Electrode ................................ 58
3.1 Background and aims ........................................................................................... 58
3.1.1 Non-biological failure ...................................................................................... 58
3.1.2 Biological failure ............................................................................................. 58
3.2 Reliability testing of Parylene-based electrodes ................................................. 59
3.2.1 Accelerated lifetime testing ............................................................................. 59
3.2.2 Insulation testing .............................................................................................. 64
3.3 28-day in vivo testing ............................................................................................ 73
3.4 Discussion and conclusion .................................................................................... 75
References .................................................................................................................... 78
Chapter 4 Parylene Adhesion Reliability ..................................................................... 79
4.1 Background and aims ........................................................................................... 79
4.1.1 Transverse and lateral impedance paths .......................................................... 79
4.2 Test device design, fabrication, and packaging .................................................. 80
4.2.1 Device design ................................................................................................... 80
4.2.2 Fabrication ....................................................................................................... 83
4.2.3 Packaging ......................................................................................................... 85
4.3 Testing protocol ..................................................................................................... 87
4.4 Results .................................................................................................................... 88
4.4.1 Scanning electron microscopy ......................................................................... 89
4.4.2 Optical microscopy .......................................................................................... 91
4.4.3 Transverse device substrate treatments ............................................................ 92
4.4.4 Lateral device substrate treatments .................................................................. 95
4.4.5 Transverse device electrode design ................................................................. 97
4.4.6 Lateral device electrode design ...................................................................... 100
4.5 Discussion and conclusions ................................................................................ 102
References .................................................................................................................. 106
Chapter 5 Conclusion ................................................................................................... 108
Appendix A Parylene Sheath Electrode Testing Procedures .................................... 110
1.1 Electrode numbering .......................................................................................... 110
1.2 Electrochemical (EC) cleaning........................................................................... 111
-VIII-

1.3 Electrochemical impedance spectroscopy (EIS) .............................................. 112
1.4 Electrochemical effects of thermoforming with cyclic voltammetry in
ferrocyanide ............................................................................................................... 112
1.5 Effect of sheath opening and thermoforming on the Parylene sheath electrode
..................................................................................................................................... 113
1.6 Electrochemical impedance spectroscopy of biofunctional coatings.............. 113
1.7 Leakage current testing on PSE-style devices .................................................. 114
1.8 Leakage current on glass-substrate interdigitated electrodes ........................ 115
1.9 Accelerated life testing (ALT) ............................................................................ 116
Appendix B A-174 Silane Treatment .......................................................................... 117
Appendix C Fabrication Process Flow for Reliability Test Devices......................... 118
1.10 Process recipe for glass-substrate devices....................................................... 118
1.11 Process recipe for Parylene-substrate devices ................................................ 119
1.12 Masks for reliability test devices ..................................................................... 120

-IX-

LIST OF TABLES

Table 2-1. Dimensions for the three different PSE sheath geometries. For all designs, sheaths were 800 µm
in length. ____________________________________________________________________ 34
Table 2-2. Calculated component values from modeled EIS data. From [2]. _______________________ 44
Table 2-3. Comparison of 1 kHz impedance magnitude values before and after electrochemical cleaning.
Probe design C, 2
nd
generation, pre-thermoform. _____________________________________ 48
Table 4-1. Parylene-substrate sample groups for each parameter studied on transverse and lateral
impedance paths. The names of the sample groups are given in quotations. ________________ 83

-X-

LIST OF FIGURES

Figure 1-1. Intracellular recording of an action potential by Hodgkin and Huxley in 1939 from the squid
giant axon. Vertical axis is in mV. Time marker is 500 Hz. (Adapted from [2]). ______________ 1
Figure 1-2. A participant of the Braingate project, who was paralyzed from the neck down by a stroke, uses
a BMI to control a prosthetic arm and drink her morning coffee. Photo credit: Brown University
News [12]. ____________________________________________________________________ 3
Figure 1-3. Current commercial neural probe technologies: (a) Microwire array [18], (b) Michigan
array[19] and (c) Utah electrode array [20]. _________________________________________ 5
Figure 1-4. Cross-sectional schematic of a possible cause of delamination of polymers. Inadequate
adhesion may result in the formation of voids (a), which then allow water vapor to condense and
ions to penetrate under the polymer, breaking adhesive bonds between the polymer layers (b). __ 7
Figure 1-5. Neurotrophic cone electrode [43]. ______________________________________________ 10
Figure 1-6. Polymer-based probes: (a) Parylene Sheath Electrode, (b) polyimide-based probe [77], and (c)
Parylene probe with microchannels [67] ___________________________________________ 12
Figure 1-7. Schematic of the Gorham deposition process. _____________________________________ 13
Figure 1-8. Schematic of a simple Parylene neural electrode with a top view (a) and cross-section (b). _ 14
Figure 1-9. Schematic of the double layer formed at the electrode-electrolyte interface in the case of
adsorbed anions. Modified from [91]. _____________________________________________ 18
Figure 1-10. Equivalent electrical circuit model proposed by Randles. R r and C r stand for the resistance
and capacitive portions, respectively, of the Faradaic reactions. C l represents the double-layer
capacitance and R C represents the solution resistance. Adapted from [96]. ________________ 20
Figure 1-11. Simplified Randles circuit. R ct = Charge-transfer resistance, C dl = double-layer capacitance,
R s = solution resistance. ________________________________________________________ 21
Figure 1-12. Coating model with constant phase elements. CPE F = constant phase element of Faradaic
(double-layer) capacitance, R f = Faradaic or charge-transfer resistance, R pore = pore resistance,
CPE C = constant phase element of coating capacitance, and R u = uncompensated or solution
resistance. ___________________________________________________________________ 22
Figure 1-13. Representative Bode plot of an EIS measurement for a simplified Randles circuit model. __ 23
Figure 1-14. Example CV of Pt in sulfuric acid. The integration of the hydrogen desorption peaks to find
Q H for ESA calculation is highlighted. _____________________________________________ 24
Figure 2-1. Optical micrograph of the Parylene Sheath Electrode, Design B (a) 1
st
generation and (b) 2
nd

generation. __________________________________________________________________ 34
Figure 2-2. Cross-sectional schematic of the major steps in the microfabrication process of the PSE (From
[2]). ________________________________________________________________________ 35
Figure 2-3. Schematic of the thermoforming process. (From [2]) _______________________________ 36
Figure 2-4. Three-electrode cell setup. WE = working electrode, CE = counter electrode, and RE =
reference electrode. Electrolyte is represented by the circle. ____________________________ 36
-XI-

Figure 2-5. EC test setup showing (a) ambient temperature Faraday cage setup and (b) temperature-
controlled, semi-caged setup. ____________________________________________________ 37
Figure 2-6. EIS impedance (a) magnitude and (b) phase curves of 1
st
generation PSE electrodes, separated
by inner electrodes (black) and outer electrodes (white). Probe design C. Mean ± SE, n = 4
electrodes across 1 probe for each data set. _________________________________________ 39
Figure 2-7. (a) Optical microscopy and (b) SEM depicting cracking out outer electrodes on 1
st
generation
PSE electrodes. _______________________________________________________________ 39
Figure 2-8. EIS curves of the PSE both before and after mechanical opening of the sheath structure with
the microwire. Probe design C, 2
nd
generation. Mean ± SE, n = 8 electrodes over 2 probes. (a)
Inner electrodes show a slight decrease in impedance following sheath opening, while (b) outer
electrodes remain unchanged. ____________________________________________________ 40
Figure 2-9. EIS curves of the PSE both before and after heat treatment. Probe design A, 2
nd
generation.
Mean ± SE, n = 16 electrodes across 2 probes. ______________________________________ 42
Figure 2-10. Comparison of the cyclic voltammograms taken before (black solid) and after (red dash) heat
treatment. Probe design A, 2
nd
generation. Mean ± SE, n = 16 electrodes across 2 probes. ____ 43
Figure 2-11. Coating model with constant phase elements. CPE F = constant phase element of Faradaic
(double-layer) capacitance, R f = Faradaic or charge-transfer resistance, R pore = pore resistance,
CPE C = constant phase element of coating capacitance, and R u = uncompensated or solution
resistance. ___________________________________________________________________ 44
Figure 2-12. Pre-heat treatment (a) and post-heat treatment (b) measured impedance data (magnitude,
square markers; phase, circle markers) and model fit (dashed red line). ___________________ 44
Figure 2-13. Cyclic voltammogram of PSE following microfabrication, but prior to thermoforming. Probe
design C, 2
nd
generation. Mean ± SE, n = 8 electrodes on 1 probe. _______________________ 47
Figure 2-14. Effects of EC clean on EIS curves. Impedance magnitude (a) and phase (b) of electrodes on
the PSE before and after EC clean. Data taken before thermoforming, Probe design C, 2
nd

generation. Mean ±SE, n = 40 electrodes across 5 probes. _____________________________ 47
Figure 2-15. Calculation of the ESA of the PSE. Probe design C, 2
nd
generation. Mean ± SE, n = 8
electrodes on 1 probe. Q H = hydrogen desorption charge. Inset: SEM of the native surface
roughness of the electrode. ______________________________________________________ 49
Figure 2-16. EIS curves of the fully-formed PSE. Probe design A, 2
nd
generation. Mean ± SE, n = 32
electrodes across 4 probes. ______________________________________________________ 49
Figure 2-17. EIS curves taken before (black) and after (white) coating. Probe design A, 2
nd
generation. (a)
Matrigel-only coating, Mean ± SE, n = 7 electrodes on 2 probes, (b) NGF/NT-3-loaded coating,
Mean ± SE, n = 4 electrodes on 1 probe, and (c) DEX-loaded coating, Mean ± SE, n = 3
electrodes on 1 probe. __________________________________________________________ 51
Figure 2-18. Normalized impedance magnitude of coated probes over 3 days at 37° C. Probe design A, 2
nd

generation. (a) Matrigel-only coating, (b) NGF/NT-3-loaded coating, and (c) DEX-loaded coating.
____________________________________________________________________________ 52
Figure 2-19. Heat map of individual electrode 1 kHz impedance magnitude measurements over the course
of the soak. Probe design A, 2
nd
generation. (a) Matrigel-only coating, (b) NGF/NT-3-loaded
coating, and (c) DEX-loaded coating. ______________________________________________ 52
-XII-

Figure 2-20. Impedance magnitude (a) and phase (b) of the PSE before (black squares) and after (white
circles) ethylene oxide sterilization. Probe design A, 2
nd
generation. Mean ± SE, n = 36 electrodes
across 5 probes. ______________________________________________________________ 54
Figure 3-1. Test setup for ALT showing connections in vial cap. PSE probes were epoxied into vial caps for
soaking. Caps were removed from soaking vials, reference and counter electrodes were positioned
through the cap, and caps were placed on vials with fresh PBS for testing. _________________ 59
Figure 3-2. EIS (a) magnitude and (b) phase of the PSE undergoing ALT at 80° C. Probe designs A and C,
2
nd
generation. Mean ± SE, n = 16 electrodes across 2 probes. Day 6 data not included due to a
connection issue. ______________________________________________________________ 60
Figure 3-3. (a) SEM and (b) optical micrograph of a probe following ALT at 80° C. SEM shows deposits of
salts from PBS, but no delamination. Optical microscopy reveals wrinkling of Pt traces that
indicates delamination from the Parylene substrate. __________________________________ 61
Figure 3-4. EIS (a) magnitude and (b) phase of the PSE soaking at 37° C. Probe design A, 2
nd
generation.
Mean ± SE, n = 16 electrodes across 2 probes. ______________________________________ 62
Figure 3-5. Optical micrograph of electrodes on the PSE after 2 weeks of soaking at 37° C. Some
electrodes maintained their smooth surface (a), but others were already delaminating from the
Parylene (b). _________________________________________________________________ 62
Figure 3-6. Optical micrograph of electrodes on the PSE after 10 weeks of soaking at 37° C. At this time
point, most electrodes were partially delaminated from the Parylene (both a and b). _________ 63
Figure 3-7. Delamination of the Parylene insulation layer from the underlying Pt and Parylene substrate.
____________________________________________________________________________ 63
Figure 3-8. Cross-sectional schematics depicting (a) transverse and (b) lateral impedance paths that were
evaluated in the leakage current studies. ___________________________________________ 65
Figure 3-9. Electrode naming on the PSE. Blue arrows indicate electrodes inside of the sheath structure,
red indicate electrodes outside. ___________________________________________________ 65
Figure 3-10. Exploded cross-sectional view of the Parylene (blue) and Pt (gray) layers present for the
inner and outer electrodes. ______________________________________________________ 66
Figure 3-11. EIS (a) magnitude and (b) phase taken prior to soaking at 37° C. Other than electrode 1
(black), the curves indicate a well-insulated electrode. Probe design B, 2
nd
generation. E1 and E3
insulated with 1 µm, E6 and E8 insulated with 5 µm. __________________________________ 67
Figure 3-12. EIS (a) magnitude and (b) phase after 1 day soaking at 37° C. Insulation integrity is
compromised for all electrodes. Probe design B, 2
nd
generation. E1 and E3 insulated with 1 µm,
E6 and E8 insulated with 5 µm. __________________________________________________ 67
Figure 3-13. Leakage current measurement indicated insulation failure on all tested electrodes after one
day of soaking. Probe design B, 2
nd
generation. E1 and E3 insulated with 1 µm, E6 and E8
insulated with 5 µm. ___________________________________________________________ 68
Figure 3-14. EIS (a) magnitude and (b) phase taken prior to soaking at 37° C. All curves indicate a
compromised insulation. Probe design B, 2
nd
generation. E1 and E3 insulated with 1 µm, E6 and
E8 insulated with 5 µm. _________________________________________________________ 68
Figure 3-15. Leakage current measurement indicated insulation failure on all tested electrodes within the
first day of soaking. Probe design B, 2
nd
generation. E1 and E3 insulated with 1 µm, E6 and E8
insulated with 5 µm. ___________________________________________________________ 69
-XIII-

Figure 3-16. Glass-substrate IDE coated with Parylene for insulation testing. _____________________ 70
Figure 3-17. EIS (a) magnitude and (b) phase measured across the transverse impedance path on heat
treated device #1. Insulation integrity is clearly compromised on Day 0, prior to soaking._____ 70
Figure 3-18. EIS (a) magnitude and (b) phase measured across the lateral impedance path on heat treated
device #1. Insulation integrity is clearly compromised on Day 0. ________________________ 70
Figure 3-19. EIS (a) magnitude and (b) phase measured across the transverse impedance path on heat
treated device #2. Insulation integrity is clearly compromised on Day 0. __________________ 71
Figure 3-20. EIS (a) magnitude and (b) phase measured across the lateral impedance path on heat treated
device #2. Insulation integrity is clearly compromised on Day 0. ________________________ 71
Figure 3-21. EIS (a) magnitude and (b) phase measured across the transverse impedance path on non-heat
treated device #1. Curves indicate that the insulation integrity is compromised on Day 14. ____ 72
Figure 3-22. EIS (a) magnitude and (b) phase measured across the lateral impedance path on non-heat
treated device #1. Curves indicate that the insulation integrity is compromised on Day 14. ____ 72
Figure 3-23. EIS (a) magnitude and (b) phase measured across the transverse impedance path on non-heat
treated device #2. Curves indicate that the insulation integrity is compromised on Day 7. _____ 73
Figure 3-24. EIS (a) magnitude and (b) phase measured across the lateral impedance path on non-heat
treated device #2. Curves indicate that the insulation integrity is compromised on Day 1. _____ 73
Figure 3-25. Representative electrophysiological traces at (a) day 0 and (b) day 14. Note a lack of
resolvable neuronal activity at day 0 and emergence of well-resolved neuronal activity at day 14.
The data are from the outer electrode sites on probe design B coated with Matrigel only. The key
electrochemical and electrophysiological parameters for this recording site are noted above the
traces [9]. ___________________________________________________________________ 74
Figure 3-26. Changes in (a) 1kHz impedance, (b) SNR, (c) noise, and (d) event rate over time after the
probe implantation (mean ± SD, n = 37 recording sites in 5 probes). The selected 5 probes are
from 3 animals (1 with probes A and 2 with probes B, 1 with DEX-loaded coating and 2 with
Matrigel only coating), for which all weekly data were available [9]. _____________________ 75
Figure 4-1. Cross-sectional schematics depicting (a) transverse and (b) lateral impedance paths on
Parylene-based electrodes. ______________________________________________________ 80
Figure 4-2. Schematic showing top view (a) and cross-sectional view (b) of a single transverse channel. 81
Figure 4-3. Electrode design parameters considered for Parylene adhesion reliability study. Side width
path and electrode trace width (a), pitch (b), and perforated electrode trace (c) were investigated.
____________________________________________________________________________ 82
Figure 4-4. Cross-section view of the fabrication process for the glass devices. ____________________ 84
Figure 4-5. Optical image of a fabricated glass device. _______________________________________ 84
Figure 4-6. Cross-section view of the fabrication process for the Parylene devices. _________________ 85
Figure 4-7. Optical image of a fabricated Parylene device. ____________________________________ 85
Figure 4-8. Parylene device with PEEK backing for insertion into ZIF connector (a) and glass device with
ZIF connector directly soldered to contact pads (b). Scale bar applies to both (a) and (b). ____ 86
Figure 4-9. ZIF connector soldered to FFC for connection to test setup. __________________________ 86
-XIV-

Figure 4-10. Transverse Parylene device secured in sample vial with marine epoxy. ________________ 87
Figure 4-11. Schematics of the test setup for transverse (a) and lateral (b) devices. _________________ 88
Figure 4-12. Impedance magnitude (a) and phase (b) of a single electrode on a PSE in 1× PBS. EIS
measurements conducted with the ADG1206 mux (red) showed a minor addition of noise, mostly
evident in the phase, as compared to measurements taken without the mux (black). __________ 88
Figure 4-13. SEM images of glass (a,b) and Parylene (c,d) devices. The Parylene-glass interface of a
single channel (a) shows isotropic etching of the Parylene layer. The diced edge of the devices (b)
shows good adhesion between the Parylene and the glass. Isotropic etching is clearly evident on
the Parylene devices (c,d), but the two layers of Parylene are well-adhered as no interface
between the layers is visible. _____________________________________________________ 90
Figure 4-14. Representative SEM images of glass (a) and Parylene (b) devices after soaking. _________ 90
Figure 4-15. Optical micrographs of glass and Parylene transverse devices before (a, glass; b, Parylene)
and after soaking (c, glass; d, Parylene), showing the appearance of diffraction patterns around
the electrode traces. ___________________________________________________________ 91
Figure 4-16. Optical micrograph of a Parylene lateral device before (a) and after (b) soaking showing
diffraction patterns along the edges of the device channel and to a lesser degree along the leftmost
electrode. ____________________________________________________________________ 92
Figure 4-17. Normalized impedance magnitude comparing tested substrate treatments on transverse
devices. Data shown for measurements taken after 1 day (a), 1 week (b), and 1 month (c) of
soaking. Note different scale for (a) due to annealed glass data. Mean ± SE, n = 4-24 electrodes
across 2-4 devices _____________________________________________________________ 93
Figure 4-18. Normalized 1 Hz impedance comparing different substrate treatments used for transverse
devices. Mean ± SE, n = 4-24 electrodes across 2-4 devices. ____________________________ 94
Figure 4-19. Normalized 1 Hz impedance substrate treatment subsets that provide pairwise comparisons:
the effect of AdPro Poly (a), annealing of glass devices (b), annealing of AdPro Poly devices (c),
and annealing of Parylene devices (d). Mean ± SE, n = 4-24 electrodes across 2-4 devices. ___ 94
Figure 4-20. Normalized impedance magnitude comparing tested substrate treatments on lateral devices.
Data shown for measurements taken after 1 day (a), 1 week (b), and 1 month (c) of soaking. Mean
± SE, n = 2-24 electrodes across 2-4 devices. _______________________________________ 95
Figure 4-21. Normalized 1 Hz impedance comparing different substrate treatments used for lateral devices.
Mean ± SE, n = 2-24 electrodes across 2-4 devices. __________________________________ 96
Figure 4-22. Normalized 1 Hz impedance substrate treatment subsets that provide pairwise comparisons:
the effect of AdPro Poly (a), annealing of glass devices (b), annealing of AdPro Poly devices (c),
and annealing of Parylene devices (d). Mean ± SE, n = 2-24 electrodes across 2-4 devices. ___ 97
Figure 4-23. Normalized impedance magnitude comparing tested electrode designs on transverse devices.
Data shown for measurements taken after 1 day (a), 1 week (b), and 1 month (c) of soaking. Mean
± SE, n = 6-8 electrodes across 4 devices. __________________________________________ 98
Figure 4-24. Normalized 1 Hz impedance comparing electrode designs used for transverse devices. Mean ±
SE, n = 6-8 electrodes across 4 devices. ____________________________________________ 99
Figure 4-25. Normalized 1 Hz impedance electrode design subsets that provide pairwise comparisons: the
effect of a narrower side width (a), wider side width (b), narrower electrode trace (c), and a
perforated electrode trace (d). Mean ± SE, n = 6-8 electrodes across 4 devices. ____________ 99
-XV-

Figure 4-26. Normalized impedance magnitude comparing tested electrode designs on lateral devices.
Data shown for measurements taken after 1 day (a), 1 week (b), and 1 month (c) of soaking. Mean
± SE, n = 1-6 electrodes across 3 devices. _________________________________________ 100
Figure 4-27. Normalized 1 Hz impedance comparing electrode designs used for lateral devices. Mean ± SE,
n = 1-6 electrodes across 3 devices. ______________________________________________ 101
Figure 4-28. Normalized 1 Hz impedance electrode design subsets that provide pairwise comparisons: the
effect of a narrower pitch (a), wider pitch (b), narrower electrode trace (c), and a perforated
electrode trace (d). Mean ± SE, n = 1-6 electrodes across 3 devices. ____________________ 101
Figure 4-29. Optical image of a glass-substrate lateral device showing delamination centered between
electrodes. __________________________________________________________________ 102
Figure 4-30. Schematic of polymer failure due to the presence of voids at the interface. _____________ 103

-XVI-

ABSTRACT

A comprehensive understanding of the human brain and the ability to tap into the
power it possesses is a daunting task, but one that physicians, scientists, and engineers are
tackling together. A crucial step toward attaining that goal is the development of
intracortical electrodes capable of reliably recording neural activity chronically. Neural
probes created with MEMS technology continue to push the limits of neural technology
and indeed may be the key to unlock the enormous mystery that is the human brain.
As new technologies are developed, it is paramount to evaluate and analyze them
to fully understand not only how they function, but also how they may fail. As engineer
and author, Henry Petroski wrote in his book Design Paradigms: Case Histories of Error
and Judgment in Engineering, “It is imperative in the design process to have a full and
complete understanding of how failure is being obviated in order to achieve success.
Without fully appreciating how close to failing a new design is, its own designer may not
fully understand how and why a design works.” In this dissertation, electrochemical
techniques are implemented to evaluate the Parylene Sheath Electrode and understand its
principal failure modes. An introduction of the problems encountered at the neural
interface and a survey of the strategies being explored to overcome current limitations to
neural probes is presented in Chapter 1. Additionally, an introduction to Parylene as a
medical device material and a primer on the electrochemical techniques used with neural
electrodes is provided. Chapter 2 gives the systematic analysis of the Parylene Sheath
Electrode to understand the electrochemical impact of the novel fabrication processes
used. Failure analysis of the probe is presented in Chapter 3 with along with the
fabrication and testing of devices designed to isolate the failure mechanisms of Parylene
as an insulator in a saline environment. These results prompted a comprehensive study to
investigate the impact of several strategies and design parameters to improve the
performance of Parylene-based electrodes, which is detailed in Chapter 4.
-XVII-

1.1 The study of electrical activity in the brain
As the center of activity, the brain controls all of the processes that we humans
carry out, both consciously and unconsciously. From fascinating processes such as
perception, learning, or playing a game of chess to the seemingly mundane processes
such as digestion or breathing, the brain controls them all. In spite of its centrality to our
everyday functioning, we know remarkably little about how the brain operates. Through
the study of the nervous systems of other organisms and the isolation of single neurons,
science has deduced the basic functionality of how neurons generate and carry neural
signals in the form of electrical action potentials [1-3]. Nonetheless, as a network of over
86 billion neurons [4], the brain is an incredibly complex organ, and we are still in the
process of piecing together the link between brain physiology and behavior.

Figure 1-1. Intracellular recording of an action potential by Hodgkin and Huxley in 1939
from the squid giant axon. Vertical axis is in mV. Time marker is 500 Hz. (Adapted from
[2]).
As an example, one need only to look to the success seen in deep brain
stimulation (DBS). DBS is a neurological technique for the treatment of various diseases,
ranging from Parkinson’s disease to depression and obsessive-compulsive disorder that
Chapter 1
INTRODUCTION
-1-

has been in use for several decades [5]. Despite the efficacy of the technique, the
mechanisms that produce the therapeutic effect are not very well understood. Given the
dramatic benefits obtained with DBS using our limited understanding of the brain, the
potential value of an improved grasp of neural physiology is undeniable. In addition to
the fundamental science of understanding the link between neural physiology and
behavior, we can use that information to develop technologies to treat illness and injury
or even extend the current capabilities of humankind.
Launched in 2013, both the U.S. BRAIN Initiative and the EU’s Human Brain
Project have highlighted the relevance of developing a complete understanding of the
human brain. Until recently, the human brain has been too daunting and too complex to
understand. Slowly the barriers to understanding the brain have begun to fall and
researchers are developing tools, techniques, and models to study the brain. Even so,
initially, the research priorities described in recent NSF and NIH reports about the
BRAIN Initiative emphasize tool development that can lead to mapping projects in the
future [6].
Intracellular recording involves the placement of an electrode through the cell
membrane to the inside of the neuron. This technique was utilized by Hodgkin and
Huxley to form our understanding of the ionic nature of the electrical signaling
mechanisms of neurons [2]. This sort of recording, however, is not possible to study the
brain in its entirety, due to the pulsatile nature of the brain. Normal physiological sources
such as cardiac rhythm and fluctuations in respiratory pressure result in micromotion of
the brain that can make intracellular recording of neural activity difficult [7].
Extracellular neural interfaces allow one to record the electrical activity of individual
neurons in the intact brain. The action potentials that take place in the neurons result in
local potential differences in the extracellular space. By placing conductive materials (i.e.
electrodes) in the extracellular space, the electrical activity of the adjacent neurons can be
recorded. Vernon Mountcastle and his colleagues developed extracellular recording
techniques to study the physiology of the cerebral cortex at the cellular level in the 1950s
[8] and by the 1960s single neuron recordings were successfully obtained in humans and
behaving primates [9, 10]. It was not a far jump for researchers to realize that these sorts
of neural interfaces also held the potential to link the brain to external devices,
-2-

positioning neuroprosthetics and the establishment of a brain machine interface (BMI)
from unattainable fantasies, to very real possibilities. In the fifty years since researchers
first obtained single neuron recordings from humans, the BMI field has seen remarkable
achievements in the degree of control that current BMI systems offer. Through programs
like the Braingate project, individuals have been able to control not only computer
cursors on a 2D screen, but also robotic arms in a 3D space [11].

Figure 1-2. A participant of the Braingate project, who was paralyzed from the neck
down by a stroke, uses a BMI to control a prosthetic arm and drink her morning coffee.
Photo credit: Brown University News [12].
1.2 Intracortical neural probes
In order to establish a link between the brain and an external device, intracortical
neural probes provide a means to electrically communicate with neurons in the brain. The
development of intracortical neural probes for BMI applications is an ongoing challenge.
Extensive research over the past few decades has yielded several technologies for BMIs.
Using these technologies, acute recordings have been successful, but reliable chronic
recordings have only been elusive.
-3-

1.2.1 Microwires
Perhaps the most basic form of creating an electrical interface with neurons,
microwires are simply insulated metal wires that are exposed at the tip. These microwires
are typically made of tungsten, stainless steel, or iridium and insulated with a
nonconducting polymer [13]. Microwires are implanted into the cortex either individually
or bundled together in arrays for access to more neurons and improved spatial
resolution[14]. Their simplicity makes them some of the most widely used intracortical
neural electrodes, but is also their greatest limitation. Each microwire provides only one
recording site, making high electrode density and depth control difficult to achieve.
1.2.2 MEMS neural probes
Microfabrication technologies overcome the limitations of microwires by
allowing precise control over electrode placement and geometries. An additional benefit
to these MEMS (micro-electro-mechanical systems) neural electrodes is the fact that they
can be fabricated in batches and in a highly repeatable fashion. Borrowing techniques
from the semiconductor industry, MEMS neural electrodes can be created from materials
that are compatible with the integration of CMOS (complementary metal oxide
semiconductor) electronics that can add functionality to the electrode [15]. In this
category, two silicon (Si) neural electrode designs are dominant. The Michigan probes
were first introduced in 1985 and consist of a planar Si shank with several electrodes
positioned along the shank, providing recording capabilities along the length of the shank,
instead of only at the tip [16]. The Utah electrode array (UEA), introduced in 1991, is
unique in that it is fabricated as an array, not as individual electrodes that are then
assembled into an array [17].
Although MEMS neural electrodes are typically fabricated on Si, MEMS
techniques are also compatible with polymers. This opens the door to the possibility of
flexible neural electrodes, which may yield a better interface with neural tissue.
-4-

Figure 1-3. Current commercial neural probe technologies: (a) Microwire array [18], (b)
Michigan array[19] and (c) Utah electrode array [20].
1.2.3 Limitations of current technology
Studies attribute the inability of current neural probe technology to attain reliable
chronic recordings to both non-biological and biological factors. Non-biological failures
occur due to connector breakage, insulation delamination, and corrosion. In a failure
mode analysis of 78 UEA implants in non-human primates, it was discovered that the
most common cause of acute (< 7 days) failure was mechanical failure of connectors,
highlighting the importance of design and material choice to ensure electrical connection
to the recording site. In the same study, chronic (> 7 days to years) failure was most
commonly attributed to biological failure, specifically meningeal encapsulation of the
UEA [21]. Biological failures result from the brain’s immune response to the neural
probe. The trauma of insertion, the blocking of chemical signaling in the brain, and the
chronic tissue agitation arising from the mechanical mismatch between the neural probe
and tissue, are all possible contributors to biological failure [7, 22-25]. In the effort to
produce a reliable chronic intracortical interface, it is vital to understand and overcome
both non-biological and biological failure modes.
1.3 Strategies to improve intracortical recording reliability
1.3.1 Strategies for non-biological failure modes
Living cortical tissue is a harsh environment for neural probes that compromises
the integrity of the materials used, thereby demanding robust design to withstand
degradation of the implant. An aggressive chemical environment is created by activated
-5-

immune cells, which produce reactive oxygen species [26] that can cause corrosion of
electrode metals, such as commonly used tungsten [27], prompting the use of noble
metals such as platinum that are resistant to this sort of corrosion.
Even with chemically inert materials, the mechanical strains experienced in vivo
can lead to breakage and electrical discontinuity between the electrode site and the
recording system. Kozai et al. demonstrated through FEM and in vivo work that probe
geometry and mechanical mismatch between implant materials can exacerbate strain
experienced by silicon probes and cause breakage [28]. Novel materials, such as carbon
fiber electrodes or compliant polymer-based probes consisting of a polymer-metal-
polymer sandwich, may better accommodate strain and resist breakage. Furthermore, the
compliance of polymer probes may reduce tissue inflammation, as will be discussed later.
While polymer probes have many desirable properties, they face several non-
biological modes of failure. Cracking and delamination of the thin film metal traces used
to form the electrode sites and traces are known complications that must be accounted for
in probe design and fabrication [29, 30]. Additionally, it is well-known that polymers are
permeable, to varying degrees, to water vapor and ions [31, 32]. If surface contamination
or poor adhesion results in the formation of a void between polymer layers, the water
vapor is able to condense. Once water vapor has condensed in the void, ions are free to
diffuse into the void, providing additional pressure to delaminate the polymer further as
well as potentially producing an electrolytic pathway between metal traces, effectively
shorting them together. If water vapor and ions remain within the polymer and do not
condense into liquid form between the polymer layers, the electrodes remain properly
insulated [33, 34]. Thus, mitigation of this phenomenon hinges on adequate adhesion and
lack of voids in the encapsulation.
-6-

Figure 1-4. Cross-sectional schematic of a possible cause of delamination of polymers.
Inadequate adhesion may result in the formation of voids (a), which then allow water
vapor to condense and ions to penetrate under the polymer, breaking adhesive bonds
between the polymer layers (b).
1.3.2 Strategies for biological failure modes
Although generally less of an issue for acute studies, the biological failure that
results from neuronal retraction, cell death, and glial encapsulation [22, 35, 36] becomes
a challenge for chronic intracortical recording. To overcome this challenge, various
approaches have been, and continue to be, explored. These strategies focus on mitigation
of trauma to neural tissue and the blood-brain barrier (BBB) upon insertion, alteration of
the brain’s immune response through bioactive molecules or novel probe geometries, and
management of chronic irritation caused by mechanical mismatch and brain micromotion.
1.3.2.1 Insertion is a major factor for biological failure
The initial insertion of a probe into the cortex inherently causes damage to the
neural tissue and neurovasculature that triggers a cascade of events that can be
detrimental to reliable neural recordings. The effects of probe cross-sectional area,
insertion speed, and imaging techniques to avoid key vasculature have been investigated.
Szarowski et al. investigated silicon probes with varying cross-sectional areas and found
that, as expected, the smallest cross-section devices damaged less tissue upon insertion.
Interestingly, the sustained response (> 2 weeks) of the neural tissue was similar across
all of the tested devices, resulting in an encapsulating sheath composed of reactive
-7-

astrocytes and microglia. The authors suggested that although a smaller penetration
profile is likely to be beneficial, the reactive tissue response and damage to vasculature is
a more dominant factor in establishing a reliable neural interface [37]. In a similar study,
it was found that a faster insertion speed positively correlated with tissue health.
Differing tip geometries (sharp, medium, and blunt) with the same cross-sectional area
were also compared, but no correlation with tissue health was observed [38]. Of note,
both of these studies highlighted the paramount importance of avoiding penetration of
neurovasculature and maintaining the integrity of the BBB as much as possible for a
favorable tissue response.
The role of a compromised BBB was further studied by Saxena et al. with non-
invasive fluorescence molecular tomography of implant sites over 16 weeks. Both
microwire arrays and Michigan arrays were implanted. The study concluded that,
although the microwire arrays displace a larger volume of neural tissue, they provided
stable recordings for longer than the Michigan arrays. This improved performance was
inversely related to BBB permeability as the microwire arrays tended to breach the BBB
less than the Michigan arrays. The dominance of BBB permeability over other variables
in the study was demonstrated as varying recording performance amongst identical
microwire arrays was also linked to BBB permeability: the poorly performing array had
enhanced BBB permeability [39]. In an effort to avoid disruption of the BBB, two-photon
microscopy has been demonstrated to provide a map of the neurovasculature to a depth of
500 µm below the cortical surface. With this technique, insertion location and trajectory
was modified and reduced neurovasculature damage was accomplished through
avoidance of major vessels [40]. Combining the conclusions of these two studies suggests
that in order to mitigate insertion trauma, chronic intracortical neural probes may need to
be flexible enough to navigate around the neurovasculature and coupled with an insertion
mechanism to properly map and direct placement of the probes.
1.3.2.2 Foreign body response results in neurodegradation and probe
encapsulation
The brain’s response to an intracortical probe that is initiated by the trauma of
insertion, is sustained by a foreign body response to the indwelling probe. This sustained
foreign body response is characterized by the activation and proliferation of astrocytes,
-8-

microglia, and macrophages. This cellular response results in the secretion of factors that
recruit microglia and macrophages to the site of the probe as well as factors that lead to
neuronal apoptosis. Furthermore, as immune cells are unable to degrade or phagocytose
the probe, glial tissue surrounds and forms a “glial scar” or “glial sheath” around the
probe, effectively isolating it from the rest of the neural tissue, presumably to maintain
the BBB [22, 35]. The end result for chronic implants is encapsulation by the glial sheath
and a zone of inflammation and neurodegradation within the vicinity of the probe [41,
42]. As theoretical models and previous studies have indicated that recording sites need
to be within 50-130 µm of a neuron in order to measure an action potential [22], this
isolation is a serious impediment to stable chronic recording.
Bioactive molecules have been shown to effectively draw neuronal processes closer
to the probe’s recording sites to produce the strongest signal. This has been demonstrated
both by directly encouraging neuronal attachment in or on to the probe [43, 44] as well as
controlling the growth of astrocytes along the probe to indirectly stimulate the extension
of neuronal processes along the layer of astrocytes toward recording sites [45]. Other
approaches seek to curtail the immune response and prevent the activation of astrocytes
and microglia, thereby inhibiting the formation of a glial sheath [46] [47-50].
Neural probe geometry also has an impact on glial encapsulation and activation of
the immune response. Correlating with previously mentioned work that suggested a
smaller cross-section area portends less neural damage, Karumbaiah et al. concluded that
a decreased probe size could reduce glial scarring and neurodegradation [51]. Skousen et
al. demonstrated that reduced surface area through an open probe architecture can reduce
microglial activation and the impact of the resulting released chemical factors that
promote inflammation, thereby reducing neuronal apoptosis [52]. These results
corroborate the proposals of Seymour and Kipke that open-architecture probes may
provide and improved interface as they increase tissue integration, reduce adhesion of
reactive cells, and improve cell-to-cell communication through increased diffusion of
chemical factors in local tissue [24].
Neurotrophic cone electrode
The neural probe technology that has had the most success in obtaining reliable
chronic neural recordings in humans is the neurotrophic cone electrode, first introduced
-9-

in 1989 [53]. A novel approach, the neurotrophic cone electrode addresses biotic failure
by incorporating the use of neurotrophins that encourage dendritic growth into the cone,
thereby mitigating many of the adverse biotic effects of the implant. With this technology,
reliable chronic neural recordings have been demonstrated in humans for nearly 5 years
[54-56]. Despite the success, the neurotrophic cone electrode is fabricated manually,
thereby limiting repeatability, recording site density, and scalability [43]. Furthermore,
the body of the probe is made of glass, resulting in a mechanical mismatch with neural
tissue.

Figure 1-5. Neurotrophic cone electrode [43].
1.3.2.3 Brain micromotion and mechanical mismatch aggravates the
foreign body response
Brain micromotion and its implications for chronic neural implants has been
investigated in order to understand its effects on extracellular recording. This
micromotion arises as a result of physiological origins such as cardiac rhythm and
fluctuations in respiratory pressure as well as mechanical origins such as movement of
the head or external disturbances to the probe and can cause additional irritation to the
foreign body response or simply result in micro-migration of the probe away from target
tissue. In the rat cortex, this motion has been quantified to be 2-4 µm due to vascular
pulsatility and 10-30 µm due to respiratory pressure [7]. The micromotion-induce stresses
in the rat model between the cortex and a stainless steel microelectrode were measured to
be 0.1-2.6 kPa over the course of an 8-week implantation [57]. Further work has
-10-

suggested that tethering probes to the skull further aggravates the situation, suggesting
that a better mechanical match to cortical tissue will help alleviate the damage [25, 41].
One alternative approach to manage the effects of micromotion between the probe
and cortical tissue is to incorporate CMOS multiplexers into the probe itself. In addition
to affording high electrode densities, this creation of “active” electrodes allows for
adaptive selection of recording channels to compensate for probe micro-migration with
respect to tissue [58, 59].
1.3.2.4 Flexible probes improve recordings
A popular approach to mitigate damage due to micromotion and to lessen
neurodegradation and the foreign body response has been to create intracortical probes
that more closely mimic the mechanical properties of cortical tissue. Studies have shown
that unphysiologically stiff materials, such as microwires and silicon, may bear some
responsibility in the activation of astrocytes and microglia [60], potentially due to the
additional strain they put on tissue than more flexible materials [61]. Taking it a step
further, a direct comparison between chemically-matched stiff and flexible chronic
implants showed that mechanically compliant probes resulted in less neurodegeneration
and a more robust BBB, foreshadowing improved chronic recording [62].
To this end, a large body of work has been dedicated to fabricating neural probes
from materials that more closely mimic the mechanical properties of brain tissue, as
mentioned earlier. Considering the negative results associated with the large mechanical
mismatch between conventional neural probe materials (metal microwires, silicon, and
ceramic) with a Young’s modulus of 100-400 GPa and brain tissue with a Young’s
modulus of 0.4-6 kPa [63], flexible substrate neural probes are under development to
better match the soft mechanical properties of the brain. Although thin film silicon has
been utilized as a flexible substrate [64], most efforts employ polymer substrates that are
less susceptible to breakage such as polyimide [23, 65, 66], Parylene C [67-70], or SU-8
[71]. Polyimide, Parylene, and SU-8 have Young’s moduli of ~3 GPa [61, 72], 2.7-3.7
GPa [73], and 2 GPa [74], respectively, which are over an order of magnitude closer to
the mechanical properties of brain tissue than conventional substrate materials (e.g.
silicon). Of the polymers that have been used, Parylene C, stands unique in that it is a
-11-

United States Pharmacopeia (USP) class VI material (the highest biocompatibility rating
in the U.S.), provides a conformal, pinhole-free coating at room temperature, and is
widely used in FDA-approved implants [75, 76].

Figure 1-6. Polymer-based probes: (a) Parylene Sheath Electrode, (b) polyimide-based
probe [77], and (c) Parylene probe with microchannels [67]

1.4 Parylene C
1.4.1 Development of Parylene
The term “Parylene” refers to a family of polymers made from the para-xylylene
compound. The molecule consists of a benzene ring with two methylene groups attached
through double carbon bonds. Various molecules can be attached to points of the benzene
ring to create Parylenes with different material properties.
Although first synthesized and characterized in 1947 by Michael Mojzesz Swarc,
the process of synthesis was impractical for commercialization at first. However, the
physical and chemical inertness of Parylene caught the attention of both academia and
industry and through the work of Donald Cram at UCLA and William Gorham at Union
Carbide, a commercially-viable synthesis process, dubbed the “Gorham process” was
developed [78, 79].
In the Gorham process, the dimer for the para-xylylene, di-p-xylylene, is used to
vapor deposit Parylenes more efficiently and without the gaseous byproducts of Swarc’s
original synthesis process. The dimer is first vaporized by heating to 130-150° C at a
pressure of 25-35 mTorr and then pyrolyzed into monomer form at 690-750° C. The
gaseous monomer then passes into the deposition chamber at room temperature (~25° C)
-12-

where it polymerizes into poly(p-xylylene) in a conformal, pin-hole free film. Any excess
monomer that does not polymerize is collected by a cold trap, in order to protect the
vacuum pump from damage [80].

Figure 1-7. Schematic of the Gorham deposition process.

Several Parylenes have been commercialized based upon their properties.
Parylene N was the initial Parylene that was synthesized, with no additional molecules
bonded to the benzene ring. Parylene C, with one chlorine atom on the monomer
molecule, has good chemical and electrical barrier properties, a low dielectric constant,
and biological stability. Parylene D, with two chlorine atoms, has a higher density and
better barrier properties than Parylene C, but film uniformity is inconsistent, limiting its
practical application. The most recently commercialized form, Parylene HT, features
improved resistance to temperatures as high as 450° C, but its biocompatibility has yet to
be confirmed [81]. As biocompatibility is paramount for neural applications, the term
“Parylene” will refer to Parylene C for the remainder of this work.
1.4.2 Usage in medical implants and neural probes
Parylene C has been classified as a USP class VI material, designating it for
implant applications, and it has been successfully tested according to the ISO 10993
standard. As a result, it has been used to encapsulate many medical devices, including
-13-

pacemakers, coronary and cerebral stents, and drug-eluting stents [82]. For Parylene
neural electrodes, Parylene is used both as substrate and insulator. The electrodes,
typically made from thin film metal, are patterned on a layer of Parylene (the substrate).
A second layer of Parylene is then deposited on top of the electrodes as an insulation
layer, sandwiching the conductive thin film metal. Finally, electrode sites are exposed by
selectively etching or ablating the insulating Parylene layer. This results in devices that
are composed entirely of just Parylene and metal.

Figure 1-8. Schematic of a simple Parylene neural electrode with a top view (a) and
cross-section (b).
1.4.3 Parylene C adhesion issues
Perhaps the biggest limiting factor for the application of Parylene C in medical
devices is its poor adhesion to substrates in saline environments. This poor adhesion has
been attributed to many possible causes and several techniques have been investigated to
improve adhesion. The adhesion between two materials can loosely be broken into two
categories: chemical, in which molecules on the surfaces of the materials form a chemical
bond that adheres the materials together, and physical, in which the two surfaces
interleave and increase van der Waals forces to adhere. In the case of the Parylene-metal
and Parylene-Parylene interfaces present in Parylene-based electrodes, both modalities
have been explored to improve adhesion.
-14-

1.4.3.1 Chemical interventions
Parylene is a non-polar film and does not form any chemical bonds with the
substrate as it is deposited [83] and so one strategy to improve adhesion is through
chemical adhesion promoters. Organosilanes, such as A-174 (gamma-
methacryloxypropyl trimethoxy silane), are commonly used to improve adhesion to
silicon and its derivatives, but is less effective on noble metal substrates [83]. Other
intermediaries, such as ceramics or plasma-polymerized films have been shown to
improve adhesion to metals [84, 85]. When intermediaries are introduced to medical
devices, however, their effect on biocompatibility is an essential consideration. A-174,
for example, is known to be toxic. Specialty Coating Systems (SCS) has developed
proprietary adhesion promoters specifically for the adhesion of Parylene to metals and
polymers, Ad Pro Plus and Ad Pro Poly, respectively that are undergoing
biocompatibility testing and may offer a promising solution for medical applications.
1.4.3.2 Physical interventions
Adhesion promotion through physical means aims to increase contact area or
encourage the interleaving of polymer chains between the different layers. Reactive ion
etching (RIE) with oxygen plasma is commonly used in MEMS fabrication and is known
to roughen Parylene surfaces. This technique has been implemented prior to deposition of
the second Parylene layer with some success [84], but in some cases it has impaired
adhesion, possibly due to the fact that the plasma alters the surface chemistry of the
Parylene making it more hydrophilic, which impairs adhesion to the second deposition of
Parylene [86].
Annealing of Parylene layers allows the polymer chains of the each layer to
interleave with each other at the interface. This process involves heating the Parylene
above its glass transition point [87] to allow for inter-diffusion of the polymer chains and
is conducted in either a vacuum or nitrogen environment in order to avoid oxidative
degradation [88]. In addition to improving adhesion, this technique has also been shown
to reduce permeability to water vapor by increasing the crystallinity of the polymer [89].
-15-

1.4.4 Techniques to analyze Parylene-based electrodes
As Parylene-based electrodes are developed, it is essential to test and analyze
them to ensure they fulfill their intended purpose. This purpose, at the most basic level, is
to interact with the body and more specifically, the brain, through the transfer of charge.
This transfer of charge occurs through electrochemical reactions at the interface between
the electrode and the ionic fluid found in the extracellular space. As such, analysis of
Parylene-based electrodes may be conducted through microscopy or elemental analysis
techniques such as x-ray photoelectron spectroscopy (XPS) to a limited extent, but
electrochemical methods are necessary to acquire a full understanding of electrode
functionality for neural applications.
1.5 Electrochemical Background and Theory
Electrochemistry is the branch of chemistry that deals with the transfer of charge
between electrons in an electrode and ions in an electrolyte. Much of the body’s
physiology is governed by the flow of ions and so an understanding of electrochemistry is
vital for any device that interfaces with that physiology through the use of electrodes.
The basis of this work is the utilization of electrochemical (EC) techniques to
evaluate Parylene-based electrodes and so this section details some fundamental
electrochemical principles and techniques used to study electrodes.
1.5.1 Understanding the electrode-electrolyte interface
When an electrode, most commonly a metal, is placed into an ionic conductor, or
electrolyte, two types of processes can occur at the interface. One kind is the transfer of
charge across the electrode-electrolyte interface, causing an oxidative or reductive
reaction to occur and converting the charge from ionic to electronic, or vice versa. This
process follows Faraday’s law, meaning that the extent of the chemical reaction is
directly proportional to the amount of electricity that is passed, and is referred to as a
faradaic process. Faradaic reactions, by their very nature, behave similar to and are often
modeled as a resistance. The other type of process, a nonfaradaic process, involves
changes to the interface without the transfer of charge between the electrode and the
-16-

electrolyte. The most common nonfaradaic process involves the adsorption and
desorption of ions to the electrode. This process can result in transient current flow when
the electrode potential or solution composition is altered, but charge does not cross the
interface, giving the interface a capacitive component.
Experimentally, one cannot isolate a single electrode-electrolyte interface. Instead,
one must study a collection of interfaces that together form an electrochemical cell. The
EC cell typically consists of the electrode of interest, referred to as the working electrode,
as well as a counter electrode to complete the circuit through the electrolyte. A third
reference electrode can help isolate the EC reactions being examined to those of the
working electrode through the use of an electrode material that has very stable, well-
understood reactions that standardize the rest of the EC cell. Although the standard
hydrogen electrode (SHE) is accepted as the universal primary reference electrode,
experiment practicalities limit its use. Instead, other standard reference electrodes such as
the saturated calomel electrode (SCE) have been developed and potentials that are
measured or applied to working electrodes are stated with respect to the reference
electrode used. Commonly used in biomedical applications, the silver/silver chloride
(Ag/AgCl) electrode was selected for this work. With the proper selection and usage of
counter and reference electrodes, the reactions and properties of the working electrode
can be highlighted and studied. For the remainder of this work, the properties and
processes of the working electrode will be the main focus unless otherwise specified.
The electrode-electrolyte interface has been studied extensively and equivalent
circuit models have been created to describe the system behavior. These models are
composed of electrical components that represent the different processes that occur at the
interface. This section will detail the different processes that contribute to the EC
behavior of the interface and how electrical components are used to model that behavior.
1.5.1.1 Interface capacitance
The simple act of introducing an electrode into an electrolyte initiates
spontaneous oxidation and reduction reactions at the interface due to charges at the
surface of the electrode and ions in the electrolyte. Without the application of an external
potential to the electrode, these reactions reach a dynamic equilibrium with balanced
-17-

redox currents known as the equilibrium exchange current density, J0, resulting in zero
net current flow and establishing an electric field between the electrode and the
electrolyte. These charges, both in the electrode and in the electrolyte, distribute
themselves into layers, creating what is known as the electrical double layer. This double
layer results in a capacitance that is a function of the potential applied across it.
The double layer was first described by Helmholtz in the late 1800’s as two sheets
of opposite charge in the electrolyte separated by a finite distance [90]. The innermost
layer consists of specifically adsorbed molecules, which is referred to as the inner
Helmholtz layer (IHP). The second layer, the outer Helmholtz layer (OHP), is separated
from the electrode by a distance, dOHP, and is composed of solvated ions that are
nonspecifically adsorbed.

Figure 1-9. Schematic of the double layer formed at the electrode-electrolyte interface in
the case of adsorbed anions. Modified from [91].

These Helmholtz layers form a parallel plate capacitor with a capacitance that is
given by:
𝐶𝐶 𝐻𝐻 =
𝜀𝜀 𝜀𝜀 0
𝑑𝑑 𝑂𝑂 𝐻𝐻 𝑂𝑂
where CH is capacitance normalized to area, ε is the permittivity of the medium between
the layers of charge, ε0 is the permittivity of free space, and dOHP is the distance between
the layers of charge. This description is limited, however, in that the capacitance is
-18-

defined as independent of the applied voltage, which is not what has been observed
experimentally.
In order to more accurately describe the double-layer capacitance phenomenon,
the Gouy [92] and Chapman [93] independently proposed an alternate model in early
1900’s. Later modified by Stern [94], the Gouy-Chapman-Stern model accounts for a
diffuse layer of charge in the solution that more fully captures the double layer behavior
in real systems (CGCS) . In this model, the double-layer capacitance is defined as [91, 95]:
1
𝐶𝐶 𝑑𝑑𝑑𝑑
=
1
𝐶𝐶 𝐻𝐻 +
1
𝐶𝐶 𝐺𝐺𝐺𝐺𝐺𝐺 =
𝜀𝜀 𝜀𝜀 0
𝑑𝑑 𝑂𝑂 𝐻𝐻 𝑂𝑂 +
𝐿𝐿 𝐷𝐷 𝜀𝜀 𝜀𝜀 0
cosh (
𝑧𝑧 𝜑𝜑 0
2 𝑘𝑘𝑘𝑘
)

where z is the charge on the ion in the electrolyte, φ0 is the applied potential, k is
Boltzmann’s constant, and T is absolute temperature. The Debye length, LD, is given by:
𝐿𝐿 𝐷𝐷 = �
𝜀𝜀 𝜀𝜀 0
𝑘𝑘𝑘𝑘
2 𝑛𝑛 𝑧𝑧 2
𝑞𝑞
where n is the bulk concentration of ions in the electrolyte and q is the elementary charge.
1.5.1.2 Charge transfer and solution resistances
The equilibrium exchange current density mentioned earlier, J0, depends on the
electrode material and the electrolyte composition. These parameters also control the
charge transfer process across the electrode-electrolyte interface and so the charge
transfer resistance, Rct, is a function of J0. When a potential is applied to the electrode
that drives the electrode potential away from the equilibrium potential, the charge transfer
resistance impedes the flow of current. For a small amplitude perturbation, the charge
transfer resistance is linear and is given by:
𝑅𝑅 𝑐𝑐𝑐𝑐
=
𝑅𝑅 𝑘𝑘 𝐹𝐹 𝐽𝐽 0

where R is the gas constant and F is Faraday’s constant. J0 is not readily estimated, but
typically determined experimentally and so likewise, Rct is typically inferred from
empirical data.
The resistance to current flow through the electrolyte between the working
electrode and counter electrode is referred to as the solution or spreading resistance.
Assuming a large area counter electrode and full contact between the working electrode
-19-

and the electrolyte, this resistance is dependent on the working electrode’s two
dimensional geometry and the electrolyte resistivity. For a circular electrode with radius,
r, in an electrolyte with resistivity, ρ, the solution resistance is:
𝑅𝑅 𝑠𝑠 =
𝜌𝜌 4 𝑟𝑟
1.5.1.3 Randles model
The most commonly used equivalent circuit model for a metal electrode in
solution is the Randles model, which was first proposed by John E. B. Randles in 1947
[96].

Figure 1-10. Equivalent electrical circuit model proposed by Randles. Rr and Cr stand for
the resistance and capacitive portions, respectively, of the Faradaic reactions. Cl
represents the double-layer capacitance and RC represents the solution resistance.
Adapted from [96].
This model is often employed in a simplified form for the analysis of neural
electrodes under the assumption that mass transfer is negligible. In this circuit, the
Faradaic or charge-transfer, reactions are represented by a resistor, which is in parallel
with the Helmholtz double-layer, characterized as a capacitor. Finally, the solution
impedance is given by an additional resistor which is in series with the parallel
combination (Figure 1-10).
-20-

Figure 1-11. Simplified Randles circuit. Rct = Charge-transfer resistance, Cdl = double-
layer capacitance, Rs = solution resistance.
1.5.1.4 More complex models
EC cells are not always well represented by the Randles model and so many
models and model components are created to address the various system conditions and
better model empirical data. Constant phase elements (CPE) are often used in place of an
ideal capacitor to represent the double-layer capacitance. The impedance of these
elements is defined by:
𝑍𝑍 𝐺𝐺𝑂𝑂 𝐶𝐶 ( 𝜔𝜔 ) =
1
( 𝑗𝑗 𝜔𝜔 𝑗𝑗 )
𝛼𝛼
where ω is frequency in radians, Q is the magnitude of the CPE with units of s
α
/Ω, and α
is a constant between 0 and 1. As α approaches 1, the behavior of the constant phase
element approaches that of an ideal capacitor. This component is used to account for
frequency dispersions at the electrode surface, most likely due to inhomogeneities such as
surface roughness [97] and so are valuable for the analysis of Parylene-based electrodes,
which have an inherent surface roughness due to the native roughness of Parylene.
Although typically used for the analysis of protective layers for corrosion studies,
a more complicated coating equivalent circuit model can be used for the study of
Parylene-based electrodes. This coating model incorporates an additional capacitance in
parallel with a simplified Randles circuit to model the effects of a non-conductive film on
the electrode. In spite of its name, the pore resistance may not actually represent physical
pores in the coating, but it does represent the resistance of the coating. This model is
-21-

typically used to represent an electrode that is completely encapsulated by a coating
material, but it can also be used to account for other non-conductive coatings, such as the
insulation along the electrode traces. Constant phase elements can be used in place of
ideal capacitors in this coating model to create an equivalent circuit that is well-suited to
model Parylene-based electrodes.

Figure 1-12. Coating model with constant phase elements. CPEF = constant phase
element of Faradaic (double-layer) capacitance, Rf = Faradaic or charge-transfer
resistance, Rpore = pore resistance, CPEC = constant phase element of coating capacitance,
and Ru = uncompensated or solution resistance.
1.5.2 Electrochemical techniques
There are a multitude of EC techniques that are employed in the field of
electrochemistry. In this section, the techniques most commonly used for neural
electrodes, and which are utilized in this work, are briefly introduced.
1.5.2.1 Electrochemical impedance spectroscopy
Electrochemical impedance spectroscopy, or EIS, is a technique that characterizes
the EC properties of the electrode interface that are described in the models discussed
above. EIS perturbs the electrochemical cell with a small-amplitude voltage signal and
monitors the resulting current to calculate the electrode impedance across a range of
frequencies. The EIS measurements are often presented in Bode plots, such as the one in
Figure 1-12. Through the use of a small-amplitude signal, the system can be
approximated to be linear despite the fact that EC cells are inherently non-linear. A broad
-22-

frequency spectrum is studied to acquire a comprehensive understanding of the electrode
impedance, and thus, the electrode’s EC properties.

Figure 1-13. Representative Bode plot of an EIS measurement for a simplified Randles
circuit model.
1.5.2.2 Cyclic voltammetry
Cyclic voltammetry, or CV, involves the application of a large-amplitude triangle
voltage waveform to the electrode, sweeping the voltage of the electrode between two set
potentials cyclically and recording the current that is produced by the resulting redox
reactions. To avoid the initiation of irreversible reactions, the potential endpoints are set
to be within the “water window”, that is, the potential limits at which electrolysis occurs.
As CV captures all of the possible reactions in a given EC cell, it is a useful tool to
characterize the electrode surface. Unique to platinum (Pt) electrodes, CV conducted in a
solution containing hydrogen atoms, such as sulfuric acid, can be used to measure the
electroactive surface area (ESA) of the electrode, which may differ from the geometric
surface area (GSA) due to surface topology. In this process, hydrogen atoms adsorb to
and desorb from the electrode surface, producing distinct peaks of current in the cyclic
voltammogram. As shown in Figure 1-13, these peaks can be integrated with respect to
the scan rate used to calculate the amount of charge that was passed during the
adsorption/desorption processes.
-23-

Figure 1-14. Example CV of Pt in sulfuric acid. The integration of the hydrogen
desorption peaks to find QH for ESA calculation is highlighted.

The characteristic charge density associated with the monolayer of hydrogen
atoms adsorbed to polycrystalline Pt is 210 µC/cm
2
[98]. Using this value with the
hydrogen desorption charge measured via CV, the ESA can be calculated:
𝐸𝐸 𝐸𝐸𝐸𝐸 =
𝑗𝑗 𝐻𝐻 210 𝜇𝜇 𝐶𝐶 /𝑐𝑐𝑐𝑐 2

1.6 Objectives
The aim of this work is to further the development of Parylene-based electrodes
for neural applications through the electrochemical evaluation of the processes and
techniques utilized to create these devices. As polymer intracortical neural probe
technology matures, these types of probes must be studied electrochemically to develop
an understanding of how these electrodes perform and identify obstacles that may affect
their use in neural interfaces. The Biomedical Microsystems Laboratory specializes in the
use of Parylene for the fabrication of medical microdevices. The lab has developed the
Parylene Sheath Electrode (PSE) technology for intracortical neural recording
applications. Using this technology as a platform, EC techniques commonly used for the
-24-

study of neural electrodes were utilized and modified for the specific study of Parylene-
based neural probes. The work evaluates the fabrication process of the PSE, studies the
effect of biofunctional coatings to EC properties, and uses EC testing to assess the
lifetime reliability of the PSE. Additionally, test structures were created to further
interrogate electrode design and processing techniques with the aim of identifying
variables that impact the robustness and reliability of the Parylene-based electrodes.
References
[1] S. M. Schuetze, “The discovery of the action potential,” Trends in Neurosciences,
vol. 6, no. 0, pp. 164-168, 1983.
[2] A. L. Hodgkin, and A. F. Huxley, “Action potentials recorded from inside a nerve
fibre,” Nature, vol. 144, no. 3651, pp. 710-711, 1939.
[3] A. L. Hodgkin, and A. F. Huxley, “A Quantitative Description of Membrane
Current and its Application to Conduction and Excitation in Nerve,” Journal of
Physiology-London, vol. 117, no. 4, pp. 500-544, 1952.
[4] F. A. C. Azevedo, L. R. B. Carvalho, L. T. Grinberg, J. M. Farfel, R. E. L. Ferretti,
R. E. P. Leite, W. Jacob, R. Lent, and S. Herculano-Houzel, “Equal Numbers of
Neuronal and Nonneuronal Cells Make the Human Brain an Isometrically Scaled-
Up Primate Brain,” Journal of Comparative Neurology, vol. 513, no. 5, pp. 532-
541, Apr, 2009.
[5] M. I. Hariz, P. Blomstedt, and L. Zrinzo, “Deep brain stimulation between 1947
and 1987: the untold story,” Neurosurgical Focus, vol. 29, no. 2, Aug, 2010.
[6] B. Keim, “The U.S. BRAIN Initiative Boldly Begins,” IEEE Spectrum, 31
December 2013, 2013.
[7] A. Gilletti, and J. Muthuswamy, “Brain micromotion around implants in the
rodent somatosensory cortex,” Journal of Neural Engineering, vol. 3, no. 3, pp.
189-195, Sep, 2006.
[8] V. B. Mountcastle, and T. P. S. Powell, “NEURAL MECHANISMS
SUBSERVING CUTANEOUS SENSIBILITY, WITH SPECIAL REFERENCE
TO THE ROLE OF AFFERENT INHIBITION IN SENSORY PERCEPTION
AND DISCRIMINATION,” Bulletin of the Johns Hopkins Hospital, vol. 105, no.
4, pp. 201-232, 1959.
[9] E. Marg, and J. E. Adams, “Indwelling multiple micro-electrodes in the brain,”
Electroencephalography and clinical neurophysiology, vol. 23, no. 3, pp. 277-280,
1967.
[10] E. V. Evarts, “Pyramidal tract activity associated with a conditioned hand
movement in the monkey,” J Neurophysiol, vol. 29, no. 6, pp. 1011-1027, 1966.
[11] J. L. Collinger, B. Wodlinger, J. E. Downey, W. Wang, E. C. Tyler-Kabara, D. J.
Weber, A. J. C. McMorland, M. Velliste, M. L. Boninger, and A. B. Schwartz,
“High-performance neuroprosthetic control by an individual with tetraplegia,”
The Lancet, no. 0.
-25-

[12] D. Orenstein, "People with paralysis control robot using brain-computer
interface," 2012.
[13] D. A. Robinson, “The electrical properties of metal micro-electrodes,”
Proceedings of the IEEE, vol. 56, no. 6, pp. 1065-1071, 1968.
[14] M. A. L. Nicolelis, “Brain-machine interfaces to restore motor function and probe
neural circuits,” Nature Reviews Neuroscience, vol. 4, no. 5, pp. 417-422, May,
2003.
[15] K. Seidl, S. Herwikl, Y. Nurcahyo, T. Torfs, M. Keller, M. Schuttler, H. Neves, T.
Stieglitz, O. Paul, and P. Ruther, "CMOS-based high-density silicon microprobe
array for electronic depth control in neural recording," Proceedings of the IEEE
International Conference on Micro Electro Mechanical Systems (MEMS). pp.
232-235.
[16] K. Najafi, K. D. Wise, and T. Mochizuki, “A high-yield IC-compatible
multichannel recording array [neural activity probe],” IEEE Transactions on
Electron Devices, vol. ED-32, no. 7, pp. 1206-11, 1985.
[17] P. K. Campbell, K. E. Jones, R. J. Huber, K. W. Horch, and R. A. Normann, “A
silicon-based, three-dimensional neural interface: manufacturing processes for an
intracortical electrode array,” Biomedical Engineering, IEEE Transactions on, vol.
38, no. 8, pp. 758-768, 1991.
[18] M. A. L. Nicolelis, D. Dimitrov, J. M. Carmena, R. Crist, G. Lehew, J. D. Kralik,
and S. P. Wise, “Chronic, multisite, multielectrode recordings in macaque
monkeys,” Proceedings of the National Academy of Sciences of the United States
of America, vol. 100, no. 19, pp. 11041-11046, Sep, 2003.
[19] K. D. Wise, A. M. Sodagar, Y. Ying, M. N. Gulari, G. E. Perlin, and K. Najafi,
“Microelectrodes, microelectronics, and implantable neural microsystems,”
Proceedings of the IEEE, vol. 96, no. 7, pp. 1184-202, 07/, 2008.
[20] P. J. Rousche, and R. A. Normann, “Chronic recording capability of the Utah
Intracortical Electrode Array in cat sensory cortex,” Journal of Neuroscience
Methods, vol. 82, no. 1, pp. 1-15, 7/1/, 1998.
[21] J. C. Barrese, N. Rao, K. Paroo, C. Triebwasser, C. Vargas-Irwin, L. Franquemont,
and J. P. Donoghue, “Failure mode analysis of silicon-based intracortical
microelectrode arrays in non-human primates,” Journal of Neural Engineering,
vol. 10, no. 6, pp. 23, Dec, 2013.
[22] V. S. Polikov, P. A. Tresco, and W. M. Reichert, “Response of brain tissue to
chronically implanted neural electrodes,” Journal of Neuroscience Methods, vol.
148, no. 1, pp. 1-18, Oct, 2005.
[23] P. J. Rousche, D. S. Pellinen, D. P. Pivin Jr, J. C. Williams, and R. J. Vetter,
“Flexible polyimide-based intracortical electrode arrays with bioactive capability,”
Biomedical Engineering, IEEE Transactions on, vol. 48, no. 3, pp. 361-371, 2001.
[24] J. P. Seymour, and D. R. Kipke, “Neural probe design for reduced tissue
encapsulation in CNS,” Biomaterials, vol. 28, no. 25, pp. 3594-3607, 2007.
[25] Y.-T. Kim, R. W. Hitchcock, M. J. Bridge, and P. A. Tresco, “Chronic response
of adult rat brain tissue to implants anchored to the skull,” Biomaterials, vol. 25,
no. 12, pp. 2229-2237, 2004.
[26] B. Halliwell, “Reactive Oxygen Species and the Central Nervous System,”
Journal of neurochemistry, vol. 59, no. 5, pp. 1609-1623, 1992.
-26-

[27] E. Patrick, M. E. Orazem, J. C. Sanchez, and T. Nishida, “Corrosion of tungsten
microelectrodes used in neural recording applications,” Journal of neuroscience
methods, vol. 198, no. 2, pp. 158-171, 2011.
[28] T. D. Kozai, K. Catt, X. Li, Z. V. Gugel, V. T. Olafsson, A. L. Vazquez, and X. T.
Cui, “Mechanical failure modes of chronically implanted planar silicon-based
neural probes for laminar recording,” Biomaterials, vol. 37, pp. 25-39, 2015.
[29] R. Metzen, and T. Stieglitz, “The effects of annealing on mechanical, chemical,
and physical properties and structural stability of Parylene C,” Biomedical
Microdevices, pp. 1-9, 2013/03/15, 2013.
[30] T. Ware, D. Simon, C. Liu, T. Musa, S. Vasudevan, A. Sloan, E. W. Keefer, R. L.
Rennaker, and W. Voit, “Thiol‐ ene/acrylate substrates for softening intracortical
electrodes,” Journal of Biomedical Materials Research Part B: Applied
Biomaterials, vol. 102, no. 1, pp. 1-11, 2014.
[31] S. Kirsten, M. Schubert, J. Uhlemann, and K.-J. Wolter, "Characterization of
ionic permeability and water vapor transmission rate of polymers used for
implantable electronics." pp. 6561-6564.
[32] P. R. Menon, W. Li, A. Tooker, and Y. C. Tai, "Characterization of water vapor
permeation through thin film parylene C," TRANSDUCERS 2009 - 15th
International Conference on Solid-State Sensors, Actuators and Microsystems. pp.
1892-1895.
[33] P. Donaldson, “Aspects of silicone rubber as an encapsulant for neurological
prostheses,” Medical and Biological Engineering and Computing, vol. 29, no. 1,
pp. 34-39, 1991.
[34] A. Vanhoestenberghe, and N. Donaldson, “Corrosion of silicon integrated circuits
and lifetime predictions in implantable electronic devices,” Journal of neural
engineering, vol. 10, no. 3, pp. 031002, 2013.
[35] W. He, and R. V. Bellamkonda, "A Molecular Perspective on Understanding and
Modulating the Performance of Chronic Central Nervous System (CNS)
Recording Electrodes," Indwelling Neural Implants: Strategies for Contending
with the In Vivo Environment, W. M. Reichert, ed., Boca Raton (FL), 2008.
[36] T. D. Y. Kozai, “Brain Tissue Responses to Neural Implants Impact Signal
Sensitivity and Intervention Strategies,” ACS chemical neuroscience, 2015.
[37] D. H. Szarowski, M. D. Andersen, S. Retterer, A. J. Spence, M. Isaacson, H. G.
Craighead, J. N. Turner, and W. Shain, “Brain responses to micro-machined
silicon devices,” Brain Research, vol. 983, no. 1-2, pp. 23-35, Sep 5, 2003.
[38] C. S. Bjornsson, S. J. Oh, Y. A. Al-Kofahi, Y. J. Lim, K. L. Smith, J. N. Turner, S.
De, B. Roysam, W. Shain, and S. J. Kim, “Effects of insertion conditions on
tissue strain and vascular damage during neuroprosthetic device insertion,”
Journal of Neural Engineering, vol. 3, no. 3, pp. 196-207, Sep, 2006.
[39] T. Saxena, L. Karumbaiah, E. A. Gaupp, R. Patkar, K. Patil, M. Betancur, G. B.
Stanley, and R. V. Bellamkonda, “The impact of chronic blood–brain barrier
breach on intracortical electrode function,” Biomaterials, vol. 34, no. 20, pp.
4703-4713, 7, 2013.
[40] T. D. Y. Kozai, T. C. Marzullo, F. Hooi, N. B. Langhals, A. K. Majewska, E. B.
Brown, and D. R. Kipke, “Reduction of neurovascular damage resulting from
-27-

microelectrode insertion into the cerebral cortex using in vivo two-photon
mapping,” Journal of Neural Engineering, vol. 7, no. 4, pp. 12, Aug, 2010.
[41] R. Biran, D. C. Martin, and P. A. Tresco, “The brain tissue response to implanted
silicon microelectrode arrays is increased when the device is tethered to the skull,”
Journal of Biomedical Materials Research - Part A, vol. 82, no. 1, pp. 169-178,
2007.
[42] G. C. McConnell, H. D. Rees, A. I. Levey, C. A. Gutekunst, R. E. Gross, and R.
V. Bellamkonda, “Implanted neural electrodes cause chronic, local inflammation
that is correlated with local neurodegeneration,” Journal of Neural Engineering,
vol. 6, no. 5, pp. 056003 (12 pp.), 2009.
[43] J. Bartels, D. Andreasen, P. Ehirim, H. Mao, S. Seibert, E. J. Wright, and P.
Kennedy, “Neurotrophic electrode: Method of assembly and implantation into
human motor speech cortex,” Journal of Neuroscience Methods, vol. 174, no. 2,
pp. 168-176, Sep, 2008.
[44] C. L. Kolarcik, D. Bourbeau, E. Azemi, E. Rost, L. Zhang, C. F. Lagenaur, D. J.
Weber, and X. T. Cui, “In vivo effects of L1 coating on inflammation and
neuronal health at the electrode–tissue interface in rat spinal cord and dorsal root
ganglion,” Acta biomaterialia, vol. 8, no. 10, pp. 3561-3575, 2012.
[45] R. Biran, M. D. Noble, and P. A. Tresco, “Directed nerve outgrowth is enhanced
by engineered glial substrates,” Experimental neurology, vol. 184, no. 1, pp. 141-
152, 2003.
[46] Y. Zhong, and R. V. Bellamkonda, “Dexamethasone Coated Neural Probes Elicit
Attenuated Inflammatory Response and Neuronal Loss Compared to Uncoated
Neural Probes,” Brain research, vol. 1148, pp. 15-27, 02/22, 2007.
[47] X. Luo, C. Matranga, S. Tan, N. Alba, and X. T. Cui, “Carbon nanotube
nanoreservior for controlled release of anti-inflammatory dexamethasone,”
Biomaterials, vol. 32, no. 26, pp. 6316-6323, 2011.
[48] S. M. Gutowski, K. L. Templeman, A. B. South, J. C. Gaulding, J. T. Shoemaker,
M. C. LaPlaca, R. V. Bellamkonda, L. A. Lyon, and A. J. Garcia, “Host response
to microgel coatings on neural electrodes implanted in the brain,” Journal of
Biomedical Materials Research - Part A, vol. 102, no. 5, pp. 1486-1499, 2014.
[49] A. J. Sawyer, W. Tian, J. K. Saucier-Sawyer, P. J. Rizk, W. M. Saltzman, R. V.
Bellamkonda, and T. R. Kyriakides, “The effect of inflammatory cell-derived
MCP-1 loss on neuronal survival during chronic neuroinflammation,”
Biomaterials, vol. 35, no. 25, pp. 6698-6706, 8//, 2014.
[50] T. D. Kozai, X. Li, L. M. Bodily, E. M. Caparosa, G. A. Zenonos, D. L. Carlisle,
R. M. Friedlander, and X. T. Cui, “Effects of caspase-1 knockout on chronic
neural recording quality and longevity: Insight into cellular and molecular
mechanisms of the reactive tissue response,” Biomaterials, vol. 35, no. 36, pp.
9620-9634, 2014.
[51] L. Karumbaiah, T. Saxena, D. Carlson, K. Patil, R. Patkar, E. A. Gaupp, M.
Betancur, G. B. Stanley, L. Carin, and R. V. Bellamkonda, “Relationship between
intracortical electrode design and chronic recording function,” Biomaterials, vol.
34, no. 33, pp. 8061-8074, 2013.
[52] J. L. Skousen, M. E. Merriam, O. Srivannavit, G. Perlin, K. D. Wise, and P. A.
Tresco, “Reducing surface area while maintaining implant penetrating profile
-28-

lowers the brain foreign body response to chronically implanted planar silicon
microelectrode arrays,” Brain Machine Interfaces: Implications for Science,
Clinical Practice and Society, vol. 194, pp. 167-180, 2011.
[53] P. R. Kennedy, “The cone electrode: a long-term electrode that records from
neurites grown onto its recording surface,” Journal of neuroscience methods, vol.
29, no. 3, pp. 181-193, 1989.
[54] P. R. Kennedy, and R. A. Bakay, “Restoration of neural output from a paralyzed
patient by a direct brain connection,” Neuroreport, vol. 9, no. 8, pp. 1707-11, Jun
1, 1998.
[55] P. R. Kennedy, R. A. Bakay, M. M. Moore, K. Adams, and J. Goldwaithe, “Direct
control of a computer from the human central nervous system,” IEEE Trans
Rehabil Eng, vol. 8, no. 2, pp. 198-202, Jun, 2000.
[56] L. R. Hochberg, M. D. Serruya, G. M. Friehs, J. A. Mukand, M. Saleh, A. H.
Caplan, A. Branner, D. Chen, R. D. Penn, and J. P. Donoghue, “Neuronal
ensemble control of prosthetic devices by a human with tetraplegia,” Nature, vol.
442, no. 7099, pp. 164-171, Jul, 2006.
[57] A. Sridharan, S. D. Rajan, and J. Muthuswamy, “Long-term changes in the
material properties of brain tissue at the implant-tissue interface,” Journal of
Neural Engineering, vol. 10, no. 6, pp. 16, Dec, 2013.
[58] K. Seidl, M. Schwaerzle, I. Ulbert, H. P. Neves, O. Paul, and P. Ruther, “CMOS-
based high-density silicon microprobe arrays for electronic depth control in
intracortical neural recording–characterization and application,”
Microelectromechanical Systems, Journal of, vol. 21, no. 6, pp. 1426-1435, 2012.
[59] C. M. Lopez, A. Andrei, S. Mitra, M. Welkenhuysen, W. Eberle, C. Bartic, R.
Puers, R. F. Yazicioglu, and G. G. Gielen, “An implantable 455-active-electrode
52-channel CMOS neural probe,” Solid-State Circuits, IEEE Journal of, vol. 49,
no. 1, pp. 248-261, 2014.
[60] P. Moshayedi, G. Ng, J. C. F. Kwok, G. S. H. Yeo, C. E. Bryant, J. W. Fawcett, K.
Franze, and J. Guck, “The relationship between glial cell mechanosensitivity and
foreign body reactions in the central nervous system,” Biomaterials, vol. 35, no.
13, pp. 3919-3925, 4, 2014.
[61] J. Subbaroyan, D. C. Martin, and D. R. Kipke, “A finite-element model of the
mechanical effects of implantable microelectrodes in the cerebral cortex,” Journal
of Neural Engineering, vol. 2, no. 4, pp. 103, 2005.
[62] J. K. Nguyen, D. J. Park, J. L. Skousen, A. E. Hess-Dunning, D. J. Tyler, S. J.
Rowan, C. Weder, and J. R. Capadona, “Mechanically-compliant intracortical
implants reduce the neuroinflammatory response,” Journal of neural engineering,
vol. 11, no. 5, pp. 056014, 2014.
[63] J. R. Capadona, D. J. Tyler, C. A. Zorman, S. J. Rowan, and C. Weder,
“Mechanically adaptive nanocomposites for neural interfacing,” MRS Bulletin,
vol. 37, no. 6, pp. 581-589, 2012.
[64] J. Du, M. L. Roukes, and S. C. Masmanidis, “Dual-side and three-dimensional
microelectrode arrays fabricated from ultra-thin silicon substrates,” Journal of
Micromechanics and Microengineering, vol. 19, no. 7, pp. 075008, 2009.
[65] T. Stieglitz, H. Beutel, C. Blau, and J.-U. Meyer, "Micromachined devices for
interfacing neurons." pp. 174-185.
-29-

[66] Z. Xiang, S.-C. Yen, N. Xue, T. Sun, W. M. Tsang, S. Zhang, L.-D. Liao, N. V.
Thakor, and C. Lee, “Ultra-thin flexible polyimide neural probe embedded in a
dissolvable maltose-coated microneedle,” Journal of Micromechanics and
Microengineering, vol. 24, no. 6, pp. 065015, 2014.
[67] S. Takeuchi, D. Ziegler, Y. Yoshida, K. Mabuchi, and T. Suzuki, “Parylene
flexible neural probes integrated with microfluidic channels,” Lab on a Chip, vol.
5, no. 5, pp. 519-523, 2005.
[68] P. J. Gilgunn, R. Khilwani, T. D. Y. Kozai, D. J. Weber, X. T. Cui, G. Erdos, O.
B. Ozdoganlar, and G. K. Fedder, "An ultra-compliant, scalable neural probe with
molded biodissolvable delivery vehicle," 2012 IEEE 25th International
Conference on Micro Electro Mechanical Systems (MEMS). pp. 56-9.
[69] J. T. W. Kuo, B. J. Kim, S. A. Hara, C. D. Lee, C. A. Gutierrez, T. Q. Hoang, and
E. Meng, “Novel flexible Parylene neural probe with 3D sheath structure for
enhancing tissue integration,” Lab on a Chip, vol. 13, no. 4, pp. 554-561, 2013.
[70] V. Castagnola, E. Descamps, A. Lecestre, L. Dahan, J. Remaud, L. Nowak, and C.
Bergaud, “Parylene-based flexible neural probes with PEDOT coated surface for
brain stimulation and recording,” Biosensors and Bioelectronics, 2014.
[71] A. Altuna, L. Menendez de la Prida, E. Bellistri, G. Gabriel, A. Guimerá, J.
Berganzo, R. Villa, and L. J. Fernández, “SU-8 based microprobes with integrated
planar electrodes for enhanced neural depth recording,” Biosensors and
Bioelectronics, vol. 37, no. 1, pp. 1-5, 8//, 2012.
[72] J. C. Coburn, M. T. Pottiger, S. C. Noe, and S. D. Senturia, “Stress in polyimide
coatings,” Journal of Polymer Science Part B: Polymer Physics, vol. 32, no. 7, pp.
1271-1283, 1994.
[73] B. Kim, B. Chen, M. Gupta, and E. Meng, “Formation of three-dimensional
Parylene C structures via thermoforming,” Journal of Micromechanics and
Microengineering, vol. 24, no. 6, pp. 065003, 2014.
[74] K. Wouters, and R. Puers, “Determining the Young's modulus and creep effects in
three different photo definable epoxies for MEMS applications,” Sensors and
Actuators, A: Physical, vol. 156, no. 1, pp. 196-200, 2009.
[75] D. C. Rodger, W. Li, J. D. Weiland, M. S. Humayun, and Y.-C. Tai, Flexible
Circuit Technologies for Biomedical Applications, 2013.
[76] J. H. C. Chang, Y. Liu, D. Y. Kang, M. Monge, Y. Zhao, C. C. Yu, A. Emami-
Neyestanak, J. Weiland, M. Humayun, Y. C. Tai, and Ieee, "PACKAGING
STUDY FOR A 512-CHANNEL INTRAOCULAR EPIRETINAL IMPLANT,"
26th Ieee International Conference on Micro Electro Mechanical Systems,
Proceedings IEEE Micro Electro Mechanical Systems, pp. 1045-1048, 2013.
[77] A. Mercanzini, K. Cheung, D. L. Buhl, M. Boers, A. Maillard, P. Colin, J. C.
Bensadoun, A. Bertsch, and P. Renaud, “Demonstration of cortical recording
using novel flexible polymer neural probes,” Sensors and Actuators a-Physical,
vol. 143, no. 1, pp. 90-96, May, 2008.
[78] M. Szwarc, “Some remarks on the CH2[graphic omitted]CH2 molecule,”
Discussions of the Faraday Society, vol. 2, no. 0, pp. 46-49, 1947.
[79] "History of Parylene," http://scscoatings.com/parylene_knowledge/parylene.
-30-

[80] W. F. Gorham, “A New, General Synthetic Method for the Preparation of Linear
Poly‐ p‐ xylylenes,” Journal of Polymer Science Part A‐ 1: Polymer Chemistry, vol.
4, no. 12, pp. 3027-3039, 1966.
[81] Y. Qin, M. M. R. Howlader, M. J. Deen, Y. M. Haddara, and P. R.
Selvaganapathy, “Polymer Integration for Packaging of Implantable Sensors,”
Sensors and Actuators B: Chemical, no. 0, 2014.
[82] "Parylene Application Evolution,"
http://scscoatings.com/what_is_parylene/application_evolution.aspx.
[83] P. K. Lightfoot, and J. Stejny, “The effect of water desorption and organosilane
coupling agents on the adhesion of poly(p-xylylene) films to a silicon wafer
surface,” Journal of Materials Science-Materials in Electronics, vol. 12, no. 10,
pp. 581-586, Oct, 2001.
[84] F. G. Yamagishi, “Investigation of plasma-polymerized films as primers for
parylene-c coatings on neural prosthesis materials,” Thin Solid Films, vol. 202, no.
1, pp. 39-50, 07/15, 1991.
[85] X. Xianzong, L. Rieth, S. Merugu, P. Tathireddy, and F. Solzbacher, “Plasma-
assisted atomic layer deposition of Al2O3 and parylene C bi-layer encapsulation
for chronic implantable electronics,” Applied Physics Letters, vol. 101, no. 9, pp.
093702 (5 pp.), 2012.
[86] C. Hassler, R. P. Von Metzen, P. Ruther, and T. Stieglitz, “Characterization of
parylene C as an encapsulation material for implanted neural prostheses,” Journal
of Biomedical Materials Research - Part B Applied Biomaterials, vol. 93, no. 1,
pp. 266-274, 2010.
[87] H. S. Noh, Y. Huang, and P. J. Hesketh, “Parylene micromolding, a rapid and
low-cost fabrication method for parylene microchannel,” Sensors and Actuators
B-Chemical, vol. 102, no. 1, pp. 78-85, Sep, 2004.
[88] T. E. Nowlin, D. Fosssmith, and G. S. Cieloszyk, “THERMAL OXIDATIVE
STABILITY OF POLY-P-XYLYLENES,” Journal of Polymer Science Part a-
Polymer Chemistry, vol. 18, no. 7, pp. 2103-2119, 1980.
[89] E. M. Davis, N. M. Benetatos, W. F. Regnault, K. I. Winey, and Y. A. Elabd,
“The influence of thermal history on structure and water transport in Parylene C
coatings,” Polymer, vol. 52, no. 23, pp. 5378-5386, Oct, 2011.
[90] H. Helmholtz, “Studien uber electricshe grenzschichten,” Annalen der Physic und
Chemie, vol. 7, 1879.
[91] A. J. Bard, and L. R. Faulkner, Electrochemical methods : fundamentals and
applications, 2nd ed., New York: John Wiley, 2001.
[92] M. Gouy, “Sur la constitution de la charge electrique a la surface d'un electrolyte,”
J. Phys. Theor. Appl., vol. 9, no. 1, pp. 457-468, 1910.
[93] D. L. Chapman, “LI. A contribution to the theory of electrocapillarity,” The
London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science,
vol. 25, no. 148, pp. 475-481, 1913.
[94] O. Stern, “Zur theorie der elektrolytischen doppelschicht,” Zeitschrift für
Elektrochemie und angewandte physikalische Chemie, vol. 30, no. 21‐ 22, pp. 508-
516, 1924.
-31-

[95] W. Franks, I. Schenker, P. Schmutz, and A. Hierlemann, “Impedance
characterization and modeling of electrodes for biomedical applications,” IEEE
Transactions on Biomedical Engineering, vol. 52, no. 7, pp. 1295-1302, 2005.
[96] J. E. B. Randles, “Kinetics of rapid electrode reactions [with discussion],”
Discussions of the Faraday Society, vol. 1, pp. 11-50, 1947.
[97] E. T. McAdams, A. Lackermeier, J. A. McLaughlin, D. Macken, and J. Jossinet,
“The linear and non-linear electrical properties of the electrode-electrolyte
interface,” Biosensors and Bioelectronics, vol. 10, no. 1–2, pp. 67-74, //, 1995.
[98] A. N. Frumkin, “Hydrogen overvoltage and adsorption phenomena -- 2,”
Advances in Electrochemistry and Electrochemical Engineering, vol. 3, pp. 287-
391, 1962.

-32-

2.1 Design of the Parylene Sheath Electrode
Of the existing intracortical neural probes, the neurotrophic cone electrode has
had the most success with chronic implementation in the human cortex. The potential of
the neurotrophic cone electrode, however, is limited by its fabrication process and the
materials that it is made of. As a manually fabricated neural probe, repeatability between
devices is difficult to maintain. Furthermore, recording site density is limited as bulk
microwire is used. Lastly, the majority of the probe consists of a glass micropipette tip,
with a Young’s modulus of ~100 GPa, that presents a large mechanical mismatch to
neural tissue [1]. The Parylene Sheath Electrode (PSE) was designed to overcome these
limitations while also capitalizing on the beneficial attributes of the neurotrophic cone
electrode. The design of the PSE borrows the form factor of the neurotrophic cone
electrode, creating a 3D sheath structure entirely out of Parylene. Thin-film platinum was
used to create the electrode sites, traces, and contact pads. Eight recording sites were
positioned on PSE, four inside of the sheath structure, and four on the outside. The 1
st

generation of the PSE had the four outer electrodes on the outer surface of the sheath
(Figure 2-1a), but they were then moved to the periphery of the sheath in the 2
nd

generation (Figure 2-1b). Recording sites were designed to be circular in shape with a 45
µm diameter.
Chapter 2
PARYLENE SHEATH ELECTRODE: DESIGN,
FABRICATION, AND TESTING
-33-

Figure 2-1. Optical micrograph of the Parylene Sheath Electrode, Design B (a) 1
st

generation and (b) 2
nd
generation.

Three different sheath geometries were created to explore how sheath geometry
may effect performance in vivo (Table 2-1). Design A uses the same dimensions as the
neurotrophic cone electrode. For points of comparison, Design B features a conical shape
with a wider taper and Design C is a straight cylinder with no taper.
Table 2-1. Dimensions for the three different PSE sheath geometries. For all designs,
sheaths were 800 µm in length.
Design Base diameter Tip diameter
A 300 50
B 450 50
C 300 300
2.2 Fabrication of the PSE
2.2.1 Microfabrication
The PSE is manufactured entirely from Parylene and platinum with conventional
MEMS fabrication techniques. The surface micromachining process is depicted
graphically in Figure 2-2. To form the substrate, 5 µm of Parylene C was first deposited
on a silicon wafer with its native oxide intact through a chemical vapor deposition (CVD)
process (Figure 2-2a). Platinum electrode sites, traces, and contact pads were then e-beam
evaporated onto the substrate and patterned with a lift-off process (Figure 2-2b). A
second layer of Parylene, 2 µm thick, was deposited and the electrode sites and contact
pads were exposed with an oxygen plasma reactive ion etch (Figure 2-2c). Photoresist
was then patterned to create a sacrificial layer and a third layer of Parylene, 5 µm thick,
was deposited. Again, the electrode sites and contact pads were exposed with an oxygen
-34-

plasma reactive ion etch (Figure 2-2d). The outlines of the individual PSEs and access
ports to the sacrificial layer were etched in a final oxygen plasma etch. The individual
PSEs were manually removed from the Si wafer, using water to aid in the release. The
photoresist sacrificial layer was then removed with an acetone soak, revealing a Parylene
microchannel (Figure 2-2e).

Figure 2-2. Cross-sectional schematic of the major steps in the microfabrication process
of the PSE (From [2]).
2.2.2 3D sheath formation
Following microfabrication, a novel post-fabrication thermoforming process was
completed to form the 3D sheath structure that emulates the glass micropipette form
factor in the neurotrophic cone electrode. This thermoforming process capitalizes on the
thermoplastic nature of Parylene to create a 3D sheath structure that retains its shape even
after deformation. The application of heat increases the crystallinity of the Parylene so
that it can hold its shape [3], but the process has also been shown to improve adhesion
between the multiple layers of Parylene [4] , as discussed in Chapter 1.
To obtain the 3D sheath structure, a custom-tapered stainless steel microwire was
manually inserted into the initially flat Parylene microchannel to serve as a mold for the
Parylene. The assembly was then placed into a vacuum oven and heated to 200° C for 48
hours under vacuum so that the Parylene does not oxidize. The oven temperature was
then slowly ramped down to room temperature and the chamber was vented. Finally, the
microwire mold was removed and the sheath structure maintained its shape.

-35-

Figure 2-3. Schematic of the thermoforming process. (From [2])
2.3 Electrochemical techniques for neural electrode testing
Following fabrication, neural probes are typically electrochemically tested to
verify that the electrode possess appropriate the EC properties for the intended
application as well as to screen devices for malfunction.
2.3.1 Basic electrochemistry for neural electrodes
As the functionality of neural electrodes is, at the heart, electrochemical, EC
testing of the electrochemical cell is an essential tool to evaluate and understand neural
electrodes. For in vitro testing, a three-electrode cell setup, with a working, a counter, and
a reference electrode, is commonly used (Figure 2-4). For this work, a 1 cm
2
platinum
square or 10 mm
2
platinum wire was used as the counter electrode, and an Ag/AgCl
electrode filled with 3M NaCl was used as a reference.

Figure 2-4. Three-electrode cell setup. WE = working electrode, CE = counter electrode,
and RE = reference electrode. Electrolyte is represented by the circle.
In vivo EC testing often utilizes the use of a two-electrode cell in which the
counter electrode and the reference electrode are combined into one electrode. This is due
-36-

to the fact that standard reference electrode materials are difficult to incorporate in vivo
and can be toxic [5].
All EC testing in this work was conducted with a Gamry Reference 600
potentiostat (Gamry Instruments, Warminster, PA) and a Faraday cage was used to
minimize ambient electrochemical noise. A waterbath, which also served to partially
shield the EC cell from noise, was used when temperature controlled tests were
conducted (Figure 2-5).

Figure 2-5. EC test setup showing (a) ambient temperature Faraday cage setup and (b)
temperature-controlled, semi-caged setup.
2.3.2 Electrochemical impedance spectroscopy
Neural action potentials result in membrane potential changes on the order of
millivolts, but that potential change does not manifest itself as strongly in the
extracellular space. As a result, recording electrodes must be able to consistently measure
extracellular potential changes on the order of 100-500 µV [6-9] and so electrode
impedance is a critical EC property for neural electrodes. The electrical signals generated
by action potentials reside in the frequency range of 100 Hz to 10 kHz [10-12]. For that
reason, the electrical impedance of neural electrodes is often tested toward the center of
that range using the single frequency of 1 kHz. While that measurement provides a small
window into the EC properties of the electrode, it does not tell the whole story. Studies
have shown that electrode changes that may dramatically affect the performance of the
neural probe may not be evident in changes of the 1 kHz impedance [9, 13]. For that
reason, it is advisable to measure the EC impedance of neural electrodes across a range of
-37-

frequencies in order to obtain a more comprehensive understanding of their EC
properties. Electrochemical impedance spectroscopy (EIS) is a commonly used technique
that does just that. EIS applies a small-amplitude sinusoidal voltage signal to an electrode
in solution and measures the resulting current that flows between it and another electrode
that is also placed in the solution.
EIS of the PSE was conducted in vitro in 1× phosphate buffered saline (PBS; all
concentrations used were 1× and so this modifier will be omitted from now on) at 37° C.
A 10 mVrms perturbation signal was used over the range of 1-100,000 Hz.
2.3.2.1 Modification of electrode placement – 2
nd
generation PSE
The 1
st
generation of the PSE had outer electrodes positioned on the top of the
sheath, as opposed to on the periphery (Figure 2-1. Optical micrograph of the Parylene
Sheath Electrode, Design B (a) 1
st
generation and (b) 2
nd
generation.). Mechanically
opening the sheath exerted a tensile stress on these top electrodes, adding to the tensile
stress that the electrodes already experience as a result of the deposition process [14, 15]
and therefore leading to cracking of the electrode sites. This cracking was evident in the
EIS curves as the impedance magnitude of the outer electrodes was greater than the inner
electrodes and the phase was shifted, or even exhibited multiple time constants (Figure
2-6). Optical microscopy and SEM confirmed the presence of cracks across the top
electrodes (Figure 2-7).
-38-

Figure 2-6. EIS impedance (a) magnitude and (b) phase curves of 1
st
generation PSE
electrodes, separated by inner electrodes (black) and outer electrodes (white). Probe
design C. Mean ± SE, n = 4 electrodes across 1 probe for each data set.

Figure 2-7. (a) Optical microscopy and (b) SEM depicting cracking out outer electrodes
on 1
st
generation PSE electrodes.
Although EIS alone is not sufficient to identify electrode cracking on its own, it is
a useful tool that may be used in conjunction with microscopy to diagnose this issue
unique to 3D polymer-based electrodes. Also, EIS is more amenable to high throughput
evaluation of electrode quality and may prove useful as a screening technique for
selecting devices suitable for implantation.
Interestingly, no cracking was observed on the inner electrodes. Due to their
placement, the inner electrodes experienced compressive stress as opposed to the outer
electrodes which were under tensile stress. As such, the inner electrodes were not
vulnerable to cracking. Due to the reduction in yield of useable devices resulting from
-39-

electrode cracking, the PSE was redesigned with the outer electrode relocated to the
probe periphery for the 2
nd
generation. Additionally, this placement could result in
improved recordings as the electrodes could benefit from reduced encapsulation and
increased proximal neuronal density compared to other locations on the PSE [16].
2.3.2.2 Mechanical opening of sheath
The formation of the 3D sheath structure required mechanical opening of the
initially flat microchannel with a custom-tapered microwire. This process was studied
with EIS to confirm that the expansion of the microchannel did not have negative effects
on the EC performance of the electrodes. As the placement of the electrodes either inside
of the sheath or on the periphery can also alter their EC performance, the effects of
mechanically opening the sheath were studied on each group separately. As can be seen
in Figure 2-8, the inner electrodes exhibit a slight decrease in impedance, likely due to
the widening of the conductive path to the counter electrode following expansion of the
sheath structure. The outer electrodes, however, were unaffected by opening of the sheath.
The radius of curvature for this probe was measured to be 76 to 77 μm. The data
indicated that thin-film Pt electrodes subjected to compressive strain following
thermoforming retain desirable EC characteristics.

Figure 2-8. EIS curves of the PSE both before and after mechanical opening of the sheath
structure with the microwire. Probe design C, 2
nd
generation. Mean ± SE, n = 8 electrodes
over 2 probes. (a) Inner electrodes show a slight decrease in impedance following sheath
opening, while (b) outer electrodes remain unchanged.
-40-

2.3.2.3 Thermoforming: heat treatment
Upon EC testing of the PSE following the heat treatment step of the
thermoforming process, it was discovered that the process significantly changed the EC
properties of the electrodes (no significant difference under optical or electron
microscopy observation). As seen in Figure 2-9, the impedance magnitude of the
electrodes increased at lower frequencies and changed slightly in slope after the heat
treatment process. The impedance phase shifted towards lower frequencies and became
slightly more negative. Although the exact causes of these changes cannot be deduced
from these results, a few possibilities are suggested. Assuming the system behaves
according to the simplified version of the Randles model [17], the increase in electrode
impedance at lower frequencies may indicate a decrease in electrode surface area. This
could be caused by the deposition or formation of an occluding substance onto the
electrode surface during the heat treatment process that was not detected with standard
SEM imaging.
Another possible explanation that could contribute to the observed increase in
impedance could be the sealing of Parylene layers at exposed interfaces. Prior to heat
treatment, it is possible that gaps exist between the Parylene layers due to inadequate
adhesion between the layers. As a result, once the devices are immersed, the solution
would be in contact with more of the metal surface than intended, producing a lower
measured impedance. As the annealing improves the adhesion between the two Parylene
layers, these gaps would lessen or disappear, resulting in less metal surface exposed to
the solution, and thus, a higher measured impedance. In an attempt to identify the
changes occurring at the electrode surface, energy-dispersive X-ray spectroscopy (EDX)
and X-ray photoelectron spectroscopy (XPS) measurements were conducted, but proved
inconclusive; the penetration depth of the EDX was too great and the spot size of the XPS
was too wide to produce a reliable signal just from the electrode surface.
-41-

Figure 2-9. EIS curves of the PSE both before and after heat treatment. Probe design A,
2
nd
generation. Mean ± SE, n = 16 electrodes across 2 probes.
To further investigate the observed effects of heat treatment, CV was performed
with the PSE in a ferrocyanide solution both before and after the heat treatment process.
The well-understood redox behavior of the ferricyanide/ferrocyanide couple was utilized
to investigate the possibility that thermoforming obscures the electrode surface which
may account for the observed electrochemical changes in the EIS results. This redox
couple is often used with CV as it is a relatively uncomplicated model of a highly
reversible reaction [18]. To this end, a CV in freshly-prepared 6 mM ferrocyanide
(K4[Fe(CN)6]) was conducted with a scan rate of 50 mV/s from -0.1 to 0.5 V vs.
Ag/AgCl (3M NaCl).
-42-

Figure 2-10. Comparison of the cyclic voltammograms taken before (black solid) and
after (red dash) heat treatment. Probe design A, 2
nd
generation. Mean ± SE, n = 16
electrodes across 2 probes.
CVs conducted in ferrocyanide revealed a distinct difference in electrode
performance between electrodes tested prior to and those tested after undergoing the heat
treatment process. The dramatic drop in current observed correlates with the EIS
observations in that it is also indicative of a decrease of electrode surface area.
Unfortunately, this experiment did not shed any additional light as to the cause of the
decreased surface area.
Modeling of the EIS data may help interpret the observed changes. The data sets
exhibited two time constants, prompting the use of an equivalent electrical circuit model
that includes a coating capacitance in parallel with a simplified Randles circuit, which
was described in Chapter 1 and included here in Figure 2-11 for ease of reference. This
capacitance models the effects of a non-conductive film on the electrode, such as the
insulating Parylene layer along the traces of the probe, which may introduce a second
time constant to the system. Constant phase elements were used in the model in place of
capacitors to better represent the surface inhomogeneities [19]. Modeling software
(Echem Analyst, Gamry Instruments, Warminster, PA) was used to calculate the values
of the model components, which are detailed in Table 2-2, and the resulting fits are
compared to the data sets in Figure 2-12.
-43-

Figure 2-11. Coating model with constant phase elements. CPEF = constant phase
element of Faradaic (double-layer) capacitance, Rf = Faradaic or charge-transfer
resistance, Rpore = pore resistance, CPEC = constant phase element of coating capacitance,
and Ru = uncompensated or solution resistance.

Table 2-2. Calculated component values from modeled EIS data. From [2].
Component Pre-heat treatment Post-heat treatment
Uncompensated resistance (Ru) 7.5 kΩ 5.8 kΩ
Faradaic resistance (Rf) 111.8 MΩ 698.1 MΩ
Pore resistance (Rpore) 1.6 MΩ 46.2 kΩ
Faradaic CPE (CPEf) 19.9× 10
-9
s×s
α
5.2 × 10
-9
s×s
α

α 0.640 0.930
Equivalent Faradaic capacitance 30.4 nF 5.7 nF
Coating CPE (CPEC) 9.0 × 10
-9
s×s
β
4.5 × 10
-9
s×s
β

β 0.799 0.867
Equivalent coating capacitance 3.1 nF 1.2 nF

Figure 2-12. Pre-heat treatment (a) and post-heat treatment (b) measured impedance data
(magnitude, square markers; phase, circle markers) and model fit (dashed red line).
-44-

Based on the model, the changes observed after the heat treatment process are
consistent with a sealing of the Parylene layers at the exposed interfaces. The dominant
changes occur in the Faradaic resistance, Faradaic capacitance, and the pore resistance
components. The increase in Faradaic resistance and decrease in Faradaic capacitance are
consistent with a decrease in exposed electrode surface area, as indicated by previous
observations, but the value of the Faradaic capacitance may yield an explanation as to
what caused the decrease in electrode area. The theoretical capacitance of thin film
platinum microelectrodes in physiological saline has been calculated to be 0.545 pF/µm
2

[20]. For our designed electrode area of 1590 µm
2
, this equates to a capacitance of 0.9 nF.
Although the theoretical calculation does not account for surface roughness, it is expected
that the observed electrode roughness would result in a larger capacitance. The Faradaic
capacitance modeled from the data approaches the designed theoretical value following
heat treatment, suggesting that the exposed electrode surface approaches the desired
electrode area. The reduction of Rpore further supports this conclusion. Complete removal
of this parameter from the model would result in a modified Randles model, a commonly
accepted model for a metal electrode in solution, indicating that the sealing of the
Parylene layers effectively removes the “pores” that are responsible for the pore
resistance element.
2.3.3 Cyclic voltammetry
Following fabrication, cyclic voltammetry (CV) was conducted on the PSE. CV
was performed either in 1× PBS or 0.05M sulfuric acid (H2SO4), depending on the
intended purpose of the CV. Sulfuric acid is often used in traditional electrochemistry as
a standard solution for the EC cleaning and characterization of a Pt surface. This
electrode-electrolyte combination produces a characteristic cyclic voltammogram that
indicates a clean electrode surface [21]. PBS, on the other hand, provides a good model
for the EC properties of physiological solution, such as that which may be found in the
extracellular space [22].
-45-

2.3.3.1 EC cleaning procedure
Following microfabrication, standard cleanroom cleaning is conducted on the PSE
that involves repeated rinses with acetone, isopropyl alcohol, and deionized water. This
cleaning process, however, does not always leave a clean electrode surface, despite the
fact that they appear clean by visual inspection and under optical microscopy.
Electrochemically, this has been observed as the introduction of multiple time constants
to the EIS curves of electrodes on the PSE. Various other subgenres of electrochemistry
regularly employ EC methods to produce a clean electrode surface [23, 24] and yet, this
practice is not commonly practiced in the neural interface field. Otto et al. made the case
for the voltage pulsing of neural electrodes to rejuvenate electrode surfaces plagued by
biofouling in vivo [25] and others have suggested its use to remove debris that may have
settled onto the electrodes during storage or sterilization and showed a 50% impedance
reduction as a result of the pulsing [26, 27]. These groups, however, do not address the
potential contamination of the electrode surface prior to implantation as a result of the
microfabrication process. Considering that solvents and other toxic materials, such as
photoresist, are used in the fabrication process of MEMS neural probes, it is essential to
remove these substances from the probe prior to implantation. The voltage pulsing that
was utilized by Otto et al. was observed to be too aggressive of a process for the PSE,
destroying the thin-film electrodes on Parylene. For this reason, CV was employed to
anodize and oxidize the electrode surface to remove any debris or residue that may have
adsorbed to the electrode surface, as suggested by Petrossians, et al. [28].
To conduct the EC clean, each electrode of the PSE was cycled at 250 mV/s from
-0.2 to 1.2 V vs. Ag/AgCl in 0.05 M H2SO4 for 30 cycles. This produced the
voltammogram seen in Figure 2-13. As is consistent with a clean Pt surface, hydrogen
adsorption and desorption peaks are visible in the CV and the formation and subsequent
removal of an oxide layer resulted in two other distinct peaks.
-46-

Figure 2-13. Cyclic voltammogram of PSE following microfabrication, but prior to
thermoforming. Probe design C, 2
nd
generation. Mean ± SE, n = 8 electrodes on 1 probe.
As seen in Figure 2-14, this process removed discontinuities observed prior to the
EC cleaning process in the EIS phase curve. Additionally, the 1 kHz impedance of the
electrodes decreased by 80.2% and the overall impedance curve shift was consistent with
the removal of debris from the electrode surface.

Figure 2-14. Effects of EC clean on EIS curves. Impedance magnitude (a) and phase (b)
of electrodes on the PSE before and after EC clean. Data taken before thermoforming,
Probe design C, 2
nd
generation. Mean ±SE, n = 40 electrodes across 5 probes.

-47-

Table 2-3. Comparison of 1 kHz impedance magnitude values before and after
electrochemical cleaning. Probe design C, 2
nd
generation, pre-thermoform.
Condition
Impedance
Magnitude (kΩ)
Standard Error
(kΩ)
Impedance
Reduction
Pre-clean 144.3 405.7 ---
Post-clean 28.6 18.2 80.2%

2.3.3.2 CV for surface area measurements
CV is also a useful tool for determining true electroactive surface area (ESA), as
introduced in Chapter 1. As most electrodes are not perfectly smooth, the surface
roughness can manifest itself electrochemically as more electrode surface area exposed to
the solution than that expected from the geometric surface area (GSA). This roughness is
important in stimulation electrodes as a tool to increase charge-injection capacity. For
recording applications, however, it is useful as it usually results in a decrease in electrode
impedance. To calculate the ESA of the PSE, the phenomenon of hydrogen adsorption on
Pt was used and the roughness factor, defined as the ratio of ESA to GSA, was estimated.
The ESA was measured by conducting a CV in 0.05M H2SO4 as part of the EC
cleaning process. The hydrogen desorption charge was measured to be approximately 41
nC, which correlates to an ESA of 19,567 µm
2
. Given that the GSA was 1590 µm
2
, this
suggests a roughness factor of 12.
-48-

Figure 2-15. Calculation of the ESA of the PSE. Probe design C, 2
nd
generation. Mean ±
SE, n = 8 electrodes on 1 probe. QH = hydrogen desorption charge. Inset: SEM of the
native surface roughness of the electrode.
2.3.4 Fully-formed PSE
As is commonly done with neural electrodes, the fully-formed PSE was tested
with EIS following all microfabrication, thermoforming, and EC cleaning. The EIS
curves presented the expected curve for an electrode in an electrolyte, indicating that the
electrodes on the PSE were clean and intact.

Figure 2-16. EIS curves of the fully-formed PSE. Probe design A, 2
nd
generation. Mean ±
SE, n = 32 electrodes across 4 probes.
-49-

2.4 Biofunctional coatings
Much of the success of the neurotrophic cone electrode is attributed to the
integration of neurotrophins in the cone to encourage dendritic growth towards recording
sites. Similarly, the PSE was designed to accommodate biofunctional factors to mitigate
the immune response. To accomplish this, Matrigel was selected to functionalize the PSE
surface as an extracellular matrix to support neuronal growth that can also be loaded with
neurotrophins and anti-inflammatory factors. Matrigel is a commercially-available
substance that was isolated from the Engelbreth-Holm-Swarm mouse sarcoma and
consists of many matrix proteins (adhesion molecules) such as laminin, collagen, and
entactin. It also contains growth factors such as epidermal, nerve, and fibroblast growth
factors, and has been successfully implemented to support cell cultures [29]. Matrigel
was coated onto the PSE in three variations: in its original form, loaded with the
neurotrophins, nerve growth factor (NGF) and neurotrophin-3 (NT-3), and loaded with an
anti-inflammatory, dexamethasone (DEX).
2.4.1 EC effects of biofunctional coatings
The PSE was coated with these biofunctional coatings after the thermoforming
process. Each of the coating variations was tested to determine the EC effect that they
had on the PSE. To do this, baseline EIS measurements were taken prior to coating the
PSE probes. The probes were coated and then immersed in PBS at 37° C. A post-coat
EIS was taken and then the probe was soaked in PBS at 37° C and daily EIS
measurements were taken for three days. EIS was conducted with a 10 mVrms
perturbation signal in the frequency range of 1-100,000 Hz.
Several devices and individual electrodes failed over the course of testing and so
were excluded from the data analyses. “Failure” in this case was determined as an
impedance value greater than 1MΩ for an individual electrode, indicating an open circuit.
This was usually due to a tear across leads caused by handling or detachment of the
sheath. Over the course of testing, 7 electrodes across 2 devices from the Matrigel set, 4
electrodes across 1 device from the NGF/NT-3 set, and 3 electrodes across 1 device from
the DEX set survived.
-50-

EIS conducted immediately after coating indicated that the coatings did slightly
impede charge transfer, as expected. No clear distinction was observed between the
different coating variations (Figure 2-17).

Figure 2-17. EIS curves taken before (black) and after (white) coating. Probe design A,
2
nd
generation. (a) Matrigel-only coating, Mean ± SE, n = 7 electrodes on 2 probes, (b)
-51-

NGF/NT-3-loaded coating, Mean ± SE, n = 4 electrodes on 1 probe, and (c) DEX-loaded
coating, Mean ± SE, n = 3 electrodes on 1 probe.
2.4.2 Body temperature soak of coated probes
For a better understanding of how the coatings will affect the probes in vivo over
time, the coated probes were soaked in PBS at 37° C over the course of three days, the
expected time for the complete elution of bioactive factors in the coating. EIS was
conducted daily and the data was compared, normalizing to the pre-coat data.

Figure 2-18. Normalized impedance magnitude of coated probes over 3 days at 37° C.
Probe design A, 2
nd
generation. (a) Matrigel-only coating, (b) NGF/NT-3-loaded coating,
and (c) DEX-loaded coating.

A heat map of 1 kHz impedance magnitude was used to provide a complete
picture of the changes observed at each electrode in each sample set over the course of
the three day soak.

Figure 2-19. Heat map of individual electrode 1 kHz impedance magnitude
measurements over the course of the soak. Probe design A, 2
nd
generation. (a) Matrigel-
only coating, (b) NGF/NT-3-loaded coating, and (c) DEX-loaded coating.
-52-

2.4.3 Biofunctional coating: analysis and conclusions
Following coating of the probes, the electrode impedance increased, indicating
that the coating did, indeed, impede charge transfer. The impedance increase observed
was consistent across the three coating variations and no distinction was seen. Following
body temperature soaking, however, a difference between the coatings became apparent.
As was expected, the increase in impedance that was observed due to the coating
gradually decreases back to baseline or even below as the coating dissolves and elutes
from the surface of the electrodes for the Matrigel-only and the NGF/NT-3-loaded
coatings. The impedance drop below baseline is not alarming as it is a fairly common
occurrence due to the gradual penetration of solution into the insulation, which will be
discussed in the next chapter. Interestingly, the DEX-loaded coating caused a continued
increase in impedance over the course of the soaking. This is likely due to the fact that
DEX often comes with a carrier, cyclodextrin, that swells in the presence of water [30].
Thus, as the DEX-loaded coating soaked in the PBS, it would swell and thicken, further
impeding charge transfer at the electrode surface.
2.5 Electrochemical effects of ethylene oxide sterilization
For concurrent in vivo work, the PSE was sterilized with ethylene oxide (EtO)
prior to implantation. To determine the EC effects of this process, the PSE was EC tested
before and after sterilization. Sterilization was performed at room temperature with the
Anprolene AN74i sterilizer system (Andersen Products, Inc., Haw River, NC) using a 24
hour cycle. Testing before and after the EtO sterilization process showed no impact on
the performance of the electrodes. The minor differences seen in the EIS curves (Figure
2-20) manifest themselves only at high frequencies, where the solution impedance
dominates the system response. This change is clearly the result of differences in solution
conductivity, due to slight variation in the measurement solution ionic concentration, and
not changes to the electrodes.
-53-

Figure 2-20. Impedance magnitude (a) and phase (b) of the PSE before (black squares)
and after (white circles) ethylene oxide sterilization. Probe design A, 2
nd
generation.
Mean ± SE, n = 36 electrodes across 5 probes.
2.6 Discussion and conclusions
Microfabrication techniques of flexible neural probes were studied through the
use of EC methods. The use of EIS aided in the identification of electrode cracking on the
1
st
generation PSE, motivating the relocation of the outer electrodes to the periphery of
the sheath structure in the 2
nd
generation probes. EC testing of the PSE indicated that the
mechanical opening of the Parylene microchannel to form the sheath structure has no
detrimental effect on the EC performance of the electrodes, provided that the electrodes
are not subjected to tensile strain. The heat treatment portion of the thermoforming
process, however, resulted in performance variation as the impedance spectra shifts and
the current is attenuated in the ferrocyanide CV. Modeling of the data suggests that this
impedance change is consistent with a sealing of the Parylene layers at the exposed
interface, indicating that the heat treatment process is effectively improving adhesion
between the distinct Parylene layers.
An EC cleaning procedure conducted with CV following standard cleanroom
cleaning procedures was shown to reduce electrode impedance and correct abnormal
impedance performance without causing damage to the electrodes. These results suggest
that potentially toxic residue may remain on neural probes from the microfabrication
-54-

process and that an EC cleaning procedure may be useful in removing such residues prior
to implant.
Qualitatively, electrode surface roughness was confirmed observed through the
use of SEM. CV was used to measure the electroactive surface area of the electrodes on
the PSE indicating that the electrodes have a roughness factor of 12.
One distinct strategy that the PSE borrowed from the neurotrophic cone electrode
is the use of biofunctional factors to encourage neuronal growth and mitigate the immune
response. The EC effects of the coating of the PSE with Matrigel loaded with
biofunctional factors was evaluated. The coating was shown to increase the electrode
impedance, but soaking in PBS at 37° C to mimic in vivo conditions indicated that the
electrode impedance would drop back down as the coating dissolved and the factors
eluted from the probe. The exception to this phenomenon was the DEX-loaded coating,
which actually increased in impedance over time. It is hypothesized that this is a
consequence of the carrier molecule, cyclodextrin, which is known to swell in solution.
While this swelling may not be a serious concern, other forms of dexamethasone that do
not use cyclodextrin as well as other methods of applying the dexamethasone to the PSE
were explored.
References

[1] J. T. W. Kuo, B. J. Kim, S. A. Hara, C. D. Lee, C. A. Gutierrez, T. Q. Hoang, and
E. Meng, “Novel flexible Parylene neural probe with 3D sheath structure for
enhancing tissue integration,” Lab on a Chip, vol. 13, no. 4, pp. 554-561, 2013.
[2] S. A. Hara, B. J. Kim, J. T. W. Kuo, and E. Meng, “An Electrochemical
Investigation of the Impact of Microfabrication Techniques on Polymer-based
Microelectrode Neural Interfaces,” Journal of Microelectromechanical Systems,
Submitted 3/16/2015, 2015.
[3] H. Jui-Mei, L. Rieth, S. Kammer, M. Orthner, and F. Solzbacher, “Effect of
thermal and deposition processes on surface morphology, crystallinity, and
adhesion of parylene-C,” Sensors and Materials, vol. 20, no. 2, pp. 87-102, 2008.
[4] D. C. Rodger, L. Wen, A. J. Fong, H. Ameri, E. Meng, J. W. Burdick, R. R. Roy,
V. R. Edgerton, J. D. Weiland, M. S. Humayun, and T. Yu-Chong, "Flexible
microfabricated parylene multielectrode arrays for retinal stimulation and spinal
cord field modulation," 2006 International Conference on Microtechnologies in
Medicine and Biology (IEEE Cat. No.06EX1357). p. 4 pp.
-55-

[5] K. R. Cote, and R. C. Gill, “DEVELOPMENT OF A PLATINIZED
PLATINUM/IRIDIUM ELECTRODE FOR USE IN VITRO,” Annals of
Biomedical Engineering, vol. 15, no. 5, pp. 419-426, 1987.
[6] P. J. Rousche, and R. A. Normann, “Chronic recording capability of the Utah
Intracortical Electrode Array in cat sensory cortex,” Journal of Neuroscience
Methods, vol. 82, no. 1, pp. 1-15, 7/1/, 1998.
[7] M. A. L. Nicolelis, D. Dimitrov, J. M. Carmena, R. Crist, G. Lehew, J. D. Kralik,
and S. P. Wise, “Chronic, multisite, multielectrode recordings in macaque
monkeys,” Proceedings of the National Academy of Sciences of the United States
of America, vol. 100, no. 19, pp. 11041-11046, Sep, 2003.
[8] R. J. Vetter, J. C. Williams, J. F. Hetke, E. A. Nunamaker, and D. R. Kipke,
“Chronic neural recording using silicon-substrate microelectrode arrays implanted
in cerebral cortex,” IEEE Transactions on Biomedical Engineering, vol. 51, no. 6,
pp. 896-904, 06/, 2004.
[9] S. Suner, M. R. Fellows, C. Vargas-Irwin, G. K. Nakata, and J. P. Donoghue,
“Reliability of signals from a chronically implanted, silicon-based electrode array
in non-human primate primary motor cortex,” IEEE Transactions on Neural
Systems and Rehabilitation Engineering, vol. 13, no. 4, pp. 524-41, 12/, 2005.
[10] E. R. Kandel, J. H. Schwartz, and T. M. Jessell, Principles of neural science:
McGraw-Hill New York, 2000.
[11] K. J. Otto, K. A. Ludwig, and D. R. Kipke, "Acquiring Brain Signals from within
the Brain," Brain-computer interfaces: principles and practice, J. Wolpaw and E.
W. Wolpaw, eds., pp. 80-110: Oxford University Press, 2012.
[12] D. A. Robinson, “The electrical properties of metal micro-electrodes,”
Proceedings of the IEEE, vol. 56, no. 6, pp. 1065-1071, 1968.
[13] P. Takmakov, K. Ruda, K. S. Phillips, I. S. Isayeva, V. Krauthamer, and C. G.
Welle, “Rapid evaluation of the durability of cortical neural implants using
accelerated aging with reactive oxygen species,” Journal of Neural Engineering,
vol. 12, no. 2, pp. 026003, 2015.
[14] D. F. Bahr, D. A. Crowson, J. S. Robach, and W. W. Gerberich, "The effects of
residual stress on modulus measurements by indentation," Thin-Films - Stresses
and Mechanical Properties VII. Symposium. pp. 85-90.
[15] V. Branger, V. Pelosin, K. F. Badawi, and P. Goudeau, “Study of the mechanical
and microstructural state of platinum thin films,” Thin Solid Films, vol. 275, no.
1-2, pp. 22-24, 1996.
[16] J. P. Seymour, and D. R. Kipke, “Neural probe design for reduced tissue
encapsulation in CNS,” Biomaterials, vol. 28, no. 25, pp. 3594-3607, 2007.
[17] J. E. B. Randles, “Kinetics of rapid electrode reactions [with discussion],”
Discussions of the Faraday Society, vol. 1, pp. 11-50, 1947.
[18] A. J. Bard, and L. R. Faulkner, Electrochemical methods : fundamentals and
applications, 2nd ed., New York: John Wiley, 2001.
[19] E. T. McAdams, A. Lackermeier, J. A. McLaughlin, D. Macken, and J. Jossinet,
“The linear and non-linear electrical properties of the electrode-electrolyte
interface,” Biosensors and Bioelectronics, vol. 10, no. 1–2, pp. 67-74, //, 1995.
-56-

[20] W. Franks, I. Schenker, P. Schmutz, and A. Hierlemann, “Impedance
characterization and modeling of electrodes for biomedical applications,” IEEE
Transactions on Biomedical Engineering, vol. 52, no. 7, pp. 1295-1302, 2005.
[21] D. T. Sawyer, "Voltammetric Indicator Electrodes," Experimental
electrochemistry for chemists, pp. 60-100, New York: John Wiley and Sons, 1974.
[22] E. Hudak, J. Mortimer, and H. Martin, “Platinum for neural stimulation:
voltammetry considerations,” Journal of Neural Engineering, vol. 7, no. 2, pp.
026005, 2010.
[23] R. A. Illsley, S. G. Roscoe, E. D. Jackson, and T. J. Hughes, “An electrochemical
investigation of the fouling of a model surface by a coliform bacterium,” 1997.
[24] M. Okochi, N. Nakamura, and T. Matsunaga, “Electrochemical control of
bacterial cell accumulation on submerged glass surfaces,” Clean Products and
Processes, vol. 1, no. 1, pp. 53-59, 1998.
[25] K. J. Otto, M. D. Johnson, and D. R. Kipke, “Voltage pulses change neural
interface properties and improve unit recordings with chronically implanted
microelectrodes,” IEEE Transactions on Biomedical Engineering, vol. 53, no. 2,
pp. 333-40, 2006.
[26] I. Ulbert, E. Halgren, G. Heit, and G. Karmos, “Multiple microelectrode-
recording system for human intracortical applications,” Journal of Neuroscience
Methods, vol. 106, no. 1, pp. 69-79, 2001.
[27] L. Grand, A. Pongrácz, É. Vázsonyi, G. Márton, D. Gubán, R. Fiáth, B. P.
Kerekes, G. Karmos, I. Ulbert, and G. Battistig, “A novel multisite silicon probe
for high quality laminar neural recordings,” Sensors and Actuators A: Physical,
vol. 166, no. 1, pp. 14-21, 2011.
[28] A. Petrossians, J. J. Whalen III, J. D. Weiland, and F. Mansfeld,
“Electrodeposition and characterization of thin-film platinum-iridium alloys for
biological interfaces,” Journal of The Electrochemical Society, vol. 158, no. 5, pp.
D269-D276, 2011.
[29] C. Lee, L. Yu, J. T. W. Kuo, B. Kim, T. Hoang, and E. Meng, "Drug eluting
coating for 3D Parylene sheath electrode." pp. 839-842.
[30] T. R. Thatiparti, A. J. Shoffstall, and H. A. von Recum, “Cyclodextrin-based
device coatings for affinity-based release of antibiotics,” Biomaterials, vol. 31, no.
8, pp. 2335-2347, Mar, 2010.
-57-

3.1 Background and aims
3.1.1 Non-biological failure
The purpose of the PSE is to establish a neural interface that can reliably record
neural activity chronically. As mentioned earlier, chronic neural interfaces can fail due to
both non-biological and biological factors. Non-biological failure can be tested on the
bench top through the use of several EC techniques combined with models of the cortical
environment. In in vivo studies, failure mechanisms are often difficult to determine due to
the combination of various factors that can contribute to neural probe failure. Through the
use of the Arrhenius relationship, the aim of the bench top testing was to determine the
lifetime of the PSE in a purely abiotic environment through acceleration of the aging
process. To determine the non-biological failure modes of the PSE, accelerated lifetime
testing, EIS, and leakage current testing were implemented.
3.1.2 Biological failure
The design of the PSE implements several strategies to mitigate failure due to
biological factors. Those strategies can only be tested in vivo where the PSE can be fully
exposed to the biotic environment. In collaboration with the Huntington Medical
Research Institutes (HMRI), PSE neural probes were implanted into the rat motor cortex
for a preliminary one month study. The aim of this testing was to evaluate the PSE’s
ability to record neural activity and track performance through electrophysiology,
electrode impedance, and histology.
Chapter 3
RELIABILITY TESTING OF THE PARYLENE
SHEATH ELECTRODE
- 58 -
3.2 Reliability testing of Parylene-based electrodes
3.2.1 Accelerated lifetime testing
Accelerated Lifetime Testing (ALT) is a commonly used technique that leverages
the Arrhenius relationship between temperature and reaction rate to accelerate lifetime
testing of devices. According to this relationship, increasing the temperature by 10° C
will roughly double the rate of many polymer reactions, thereby accelerating the rate of
aging [1]. Using the equation below, the relationship between testing time at the
accelerated temperature (tT) and the simulated time at body temperature (t37) can be
calculated.
𝑡𝑡 3 7
= 𝑡𝑡 𝑇𝑇 × 2
( 𝑇𝑇 − 3 7)
1 0

Samples were immersed in PBS to mimic physiological conditions in sealed vials.
To accelerate aging of the PSE, a circulating water bath was used to maintain the samples
at the elevated temperature. For testing, the samples were removed from the elevated
temperature and EIS was conducted in fresh PBS at 37° C.

Figure 3-1. Test setup for ALT showing connections in vial cap. PSE probes were
epoxied into vial caps for soaking. Caps were removed from soaking vials, reference and
counter electrodes were positioned through the cap, and caps were placed on vials with
fresh PBS for testing.
3.2.1.1 80° C acceleration
At 80° C, aging is accelerated approximately 20 times, so 6 months can be
simulated in approximately 9 days.
- 59 -

Figure 3-2. EIS (a) magnitude and (b) phase of the PSE undergoing ALT at 80° C. Probe
designs A and C, 2
nd
generation. Mean ± SE, n = 16 electrodes across 2 probes. Day 6
data not included due to a connection issue.
Once the PSE is soaked in PBS, the impedance magnitude drops at high
frequencies. This is a common occurrence in aging probes [2]. This may be due to water
absorption in the insulating Parylene C around the electrode site, thereby altering the EC
properties. It is worth noting that the rate of change slows and begins to stabilize around
Day 5 (correlates to approximately 100 days at 37° C), indicating that the changes to the
electrodes are reaching equilibrium.
Following ALT, both SEM optical microscopy was used to examine the electrode
surface. SEM indicated no delamination of the Parylene. Under optical microscopy,
however, it was observed that the Pt in the PSE had begun to detach from the Parylene at
larger area interfaces.

- 60 -

Figure 3-3. (a) SEM and (b) optical micrograph of a probe following ALT at 80° C. SEM
shows deposits of salts from PBS, but no delamination. Optical microscopy reveals
wrinkling of Pt traces that indicates delamination from the Parylene substrate.

It was suspected that 80°C was too high of a temperature for accelerated testing of
Parylene-Pt devices and actually initiated other artifacts that would never be observed at
body temperature, such as buckling of the metals due to mismatched thermal coefficients
of expansion (TCE) of Parylene and Pt. Additionally, the glass transition temperature of
Parylene is thought to lay between 35° C to 80° C [3]. For this reason, it may be best to
avoid soaking Parylene-based devices at 80° C. Personal communication with other
groups that work with Parylene-substrate electrodes also suggested this may be the case.
3.2.1.2 37° C soak
To avoid the complications experienced at elevated temperature, the longevity of
the probes was tested as they soaked in PBS at 37° C. Unfortunately, as no elevated
temperature was used, no acceleration of aging was accomplished. Probes were soaked
and EIS measurements were taken over the course of 3 months.
- 61 -

Figure 3-4. EIS (a) magnitude and (b) phase of the PSE soaking at 37° C. Probe design
A, 2
nd
generation. Mean ± SE, n = 16 electrodes across 2 probes.
As was observed at 80° C, the impedance magnitude decreased and the phase
curve shifted. Interestingly, despite the fact that soaking took place at a lower
temperature, the impedance drop was more drastic than at 80° C. Optical microscopy
taken during the soak revealed that the Pt detached from the Parylene. Additionally, the
Parylene insulation layer covering the Pt delaminated from the Pt and Parylene substrate.

Figure 3-5. Optical micrograph of electrodes on the PSE after 2 weeks of soaking at 37°
C. Some electrodes maintained their smooth surface (a), but others were already
delaminating from the Parylene (b).
- 62 -

Figure 3-6. Optical micrograph of electrodes on the PSE after 10 weeks of soaking at 37°
C. At this time point, most electrodes were partially delaminated from the Parylene (both
a and b).

Figure 3-7. Delamination of the Parylene insulation layer from the underlying Pt and
Parylene substrate.
3.2.1.3 ALT discussion and conclusions
The discrepancies seen between the 80° C ALT data and the 37° C soak data
suggests that, despite the fact that an elevated temperature may indeed introduce
confounding failure mechanisms that would not be seen at 37° C, the elevated
temperature was not the sole reason for detachment of the Pt from the Parylene and the
drop in impedance. Both sets of data indicate that adhesion between Pt and Parylene is
not robust and delamination of the Pt from the Parylene substrate is an issue that warrants
attention.
Adhesion issues between Parylene and metal layers has been extensively
documented and various possible remedies have been suggested. As detailed in Chapter
- 63 -
1, the suggested solutions typically involve the application of an additional material
during the fabrication process [4-6], which may compromise biocompatibility. For this
reason, it was decided to avoid the use of materials other than Parylene and Pt, which
both have a history of being successfully used in implanted devices.
Techniques to improve the adhesion between two Parylene layers have been
developed and may prove useful for improved longevity of the PSE. One such technique,
the annealing of Parylene layers [7], was implemented in the PSE through the
thermoforming process. Current data has not shown this technique to be effective for the
PSE, but work is underway to optimize the parameters of the process to achieve the
desired results as shown in literature. Additionally, redesign of the Pt layout on the PSE
may limit delamination of the Pt through the inclusion of perforations in the metal traces
to leverage the improved Parylene-Parylene adhesion and reduce the amount of large area
Pt that may be more prone to delamination [8]. This strategy was evaluated in test
structures, the results of which will be elaborated in Chapter 4.
3.2.2 Insulation testing
A fundamental part of probe reliability is the performance of the insulation
material. It is critical for the insulating layer to sufficiently isolate electrode traces both
from other traces as well as from the surrounding environment in order to reliably record
neural signals. Failed insulation between an electrode and the surrounding environment
compromises the electrode’s ability to isolate targeted neural signals from other electrical
activity in the tissue. Similarly, if electrode traces are insufficiently insulated from each
other, signals being carried on one electrode can interfere with those on a neighboring
electrode, a phenomenon known as cross-talk. To avoid these failure modes, a high
impedance is desired along these electrical paths. Transverse impedance (between each
electrode and the surrounding environment, Ztrans), as well as lateral impedance (between
adjacent electrodes, Zlat) were evaluated in Parylene-based electrodes. To evaluate these
impedance paths, EIS and leakage current testing were conducted.
- 64 -

Figure 3-8. Cross-sectional schematics depicting (a) transverse and (b) lateral impedance
paths that were evaluated in the leakage current studies.
EIS was conducted across each impedance path in PBS at 37° C with a 25 mVrms
perturbation signal from 1-100,000 Hz. Leakage current testing was conducted by
applying 5 VDC across the impedance path and monitoring the resulting current. For the
purpose of this study, any current over 5 µA was considered a failure as this correlated to
a DC impedance of 1 MΩ. Test samples were soaked in 37° C PBS and tested daily.
3.2.2.1 PSE-style test devices
To imitate the conditions that the PSE experiences as closely as possible,
insulation testing was conducted on PSE probes with electrode sites that had not been
exposed and were still covered with 1 or 5 µm of Parylene depending if the electrode is
inside or outside of the sheath structure, respectively. The same thermoforming process
that was used for the PSE (heating to 200° C for 48 h in vacuum) was also used for these
devices. For each of the 2 test devices, 2 electrodes inside and 2 electrodes outside of the
sheath were tested. Electrodes were numbered according to Figure 3-9.

Figure 3-9. Electrode naming on the PSE. Blue arrows indicate electrodes inside of the
sheath structure, red indicate electrodes outside.
For measurements of the transverse impedance path, the impedance was measured
across Parylene 1 or 5 µm thick, depending on whether the electrodes being tested were
- 65 -
inside or outside of the sheath structure (Figure 3-10). This measurement could also take
a path between the Parylene layers to the solution. The distance to solution between the
Parylene layers varied depending on the electrode being tested, with the shortest distance
being approximately 10 µm. For the lateral impedance path, the shortest distance between
leads was 30 µm.

Figure 3-10. Exploded cross-sectional view of the Parylene (blue) and Pt (gray) layers
present for the inner and outer electrodes.
EIS measurements taken across the transverse impedance path prior to soaking
indicated that three of the four electrodes on a single probe were well-insulated (Figure
3-11), with the fourth electrodes showing signs of insulation beginning to fail. After one
day of soaking in 37° C PBS, the insulation was compromised on all electrodes, as can be
seen by the drop in impedance magnitude and the introduction of a phase shift other than
-90°, the expected phase shift for a dielectric such as Parylene (Figure 3-12). This was
confirmed by leakage current testing, which showed a current signal that saturated the
settings at over 600 µA (Figure 3-13).

- 66 -

Figure 3-11. EIS (a) magnitude and (b) phase taken prior to soaking at 37° C. Other than
electrode 1 (black), the curves indicate a well-insulated electrode. Probe design B, 2
nd

generation. E1 and E3 insulated with 1 µm, E6 and E8 insulated with 5 µm.

Figure 3-12. EIS (a) magnitude and (b) phase after 1 day soaking at 37° C. Insulation
integrity is compromised for all electrodes. Probe design B, 2
nd
generation. E1 and E3
insulated with 1 µm, E6 and E8 insulated with 5 µm.
- 67 -

Figure 3-13. Leakage current measurement indicated insulation failure on all tested
electrodes after one day of soaking. Probe design B, 2
nd
generation. E1 and E3 insulated
with 1 µm, E6 and E8 insulated with 5 µm.
This dramatic failure was observed in both devices tested. The second device
failed within the first day, prior to soaking overnight, as seen in the EIS (Figure 3-14) and
leakage current (Figure 3-15) data below. No imperfections to the insulation were
observed under optical microscopy, though it is possible that the insulation was
compromised with pinholes prior to testing.

Figure 3-14. EIS (a) magnitude and (b) phase taken prior to soaking at 37° C. All curves
indicate a compromised insulation. Probe design B, 2
nd
generation. E1 and E3 insulated
with 1 µm, E6 and E8 insulated with 5 µm.
- 68 -

Figure 3-15. Leakage current measurement indicated insulation failure on all tested
electrodes within the first day of soaking. Probe design B, 2
nd
generation. E1 and E3
insulated with 1 µm, E6 and E8 insulated with 5 µm.
The PSE-style test devices failed across the transverse impedance path before the
lateral impedance path could be tested. As each electrode had an electrolytic connection
to a common ground, there was no need to conduct lateral impedance testing.
3.2.2.2 Glass-substrate interdigitated electrode devices
In an effort to isolate the Parylene insulation layer and remove any effects the
Parylene-substrate may have had on testing, the insulation testing was repeated on glass-
substrate interdigitated (IDE) Pt electrodes under the same conditions as for the PSE-style
devices. Each digit was 100 µm wide and 5 µm of Parylene was deposited onto the
electrodes. No adhesion promoter, such as A-174, was used in an effort to minimize any
confounding factors and so that the test results maintain applicability for in vivo
applications. This resulted in a transverse impedance path of 5 µm through the Parylene
and a lateral impedance path of 100 µm between digits. Two devices were heat treated at
200° C for 48 hours, while two devices were tested without any heat treatment. In this
way, the effect of the heat treatment process that is used during thermoforming on the
insulation integrity could be studied.
- 69 -

Figure 3-16. Glass-substrate IDE coated with Parylene for insulation testing.
Both devices that were heat treated displayed indications of compromised
insulation both along the transverse (Figure 3-17 and Figure 3-19) and the lateral (Figure
3-18 and Figure 3-20) impedance paths within the first day of testing (Day 0).

Figure 3-17. EIS (a) magnitude and (b) phase measured across the transverse impedance
path on heat treated device #1. Insulation integrity is clearly compromised on Day 0,
prior to soaking.

Figure 3-18. EIS (a) magnitude and (b) phase measured across the lateral impedance path
on heat treated device #1. Insulation integrity is clearly compromised on Day 0.
- 70 -

Figure 3-19. EIS (a) magnitude and (b) phase measured across the transverse impedance
path on heat treated device #2. Insulation integrity is clearly compromised on Day 0.

Figure 3-20. EIS (a) magnitude and (b) phase measured across the lateral impedance path
on heat treated device #2. Insulation integrity is clearly compromised on Day 0.
The devices that were not heat treated maintained their integrity for longer, but
not satisfactorily. After one to two weeks of soaking at body temperature, the EIS curves
showed a breach of the insulation, both across the transverse as well as the lateral
impedance paths.

- 71 -

Figure 3-21. EIS (a) magnitude and (b) phase measured across the transverse impedance
path on non-heat treated device #1. Curves indicate that the insulation integrity is
compromised on Day 14.

Figure 3-22. EIS (a) magnitude and (b) phase measured across the lateral impedance path
on non-heat treated device #1. Curves indicate that the insulation integrity is
compromised on Day 14.
- 72 -

Figure 3-23. EIS (a) magnitude and (b) phase measured across the transverse impedance
path on non-heat treated device #2. Curves indicate that the insulation integrity is
compromised on Day 7.

Figure 3-24. EIS (a) magnitude and (b) phase measured across the lateral impedance path
on non-heat treated device #2. Curves indicate that the insulation integrity is
compromised on Day 1.
3.3 28-day in vivo testing
The ultimate reliability test, of course, is implementation in vivo. PSE probes
were implanted into the motor cortex of 19 young, male Sprague Dawley rats. All
procedures for the animal experiments were in accordance with the animal protocol
approved by the Huntington Medical Research Institutes Institutional Animal Care and
Use Committee (HMRI IACUC) and in compliance with the Animal Welfare Act. To
assess the functionality of the probes, in vivo EIS and electrophysiological neural
recordings were conducted weekly. EIS data was collected using the same parameters as
- 73 -
used in vitro. Electrophysiological data was acquired at 16 bit and 40 kHz per channel
using a 64-channel data acquisition system (OmniPlex; Plexon Inc., Dallas, TX) and
high-pass filtered at 300 Hz to remove the low-frequency fluctuations from the baseline.
Spike detection of action potentials was performed using the Nonlinear Energy Operator
to provide more accurate spike detection, albeit at lower signal-to-noise ratios (SNR) than
might be obtained using a simple amplitude thresholding spike detection method.
Neural activity was successfully recorded with the PSE within 2 weeks of
implantation. As was expected, electrode impedance increased after implantation, but
began to show signs of leveling off. Over the course of the 28-day period, the event rate
and SNR continued to gradually increase while the noise level began to plateau,
suggesting stabilization of the interface [9].

Figure 3-25. Representative electrophysiological traces at (a) day 0 and (b) day 14. Note
a lack of resolvable neuronal activity at day 0 and emergence of well-resolved neuronal
activity at day 14. The data are from the outer electrode sites on probe design B coated
with Matrigel only. The key electrochemical and electrophysiological parameters for this
recording site are noted above the traces [9].
- 74 -

Figure 3-26. Changes in (a) 1kHz impedance, (b) SNR, (c) noise, and (d) event rate over
time after the probe implantation (mean ± SD, n = 37 recording sites in 5 probes). The
selected 5 probes are from 3 animals (1 with probes A and 2 with probes B, 1 with DEX-
loaded coating and 2 with Matrigel only coating), for which all weekly data were
available [9].

3.4 Discussion and conclusion
Accelerated lifetime testing was conducted to assess the time-to-failure of the
PSE and identify failure mechanisms. An elevated temperature was initially chosen to
accelerate the processes that lead to failure and therefore minimize the experiment
timeframe. Under these conditions, the PSE was found to exhibit signs of insulation
failure within the first day of soaking, raising suspicion around the use of an elevated
temperature. Although the glass transition temperature is not agreed upon, the proposed
values of 35° to 80° C suggest that the chosen soak temperature of 80° C is too high and
may initiate changes to the material properties of the PSE. Given these results, it was
determined that testing of the PSE at temperatures above that which would be experience
in vivo (37° C) would not provide an accurate measurement of time-to-failure. To avoid
any extraneous effects that an elevated temperature may have on the PSE that would not
- 75 -
be observed at body temperature, subsequent soak testing was conducted at 37° C.
Nonetheless, in both experiments EIS measurements indicated insulation failure. This
was confirmed with optical microscopy that showed visible delamination of the Pt
electrodes and traces from the Parylene substrate, indicating that adhesion of these two
materials may cause problems for longevity of the PSE.
Additional testing was conducted with test devices fabricated to better isolate
insulation failure. PSE-style test devices were fabricated identically to the PSE with the
one exception of keeping the electrodes completely encapsulated. EIS measurements and
leakage current testing demonstrated insulation failure within one day of soaking in PBS
at body temperature. Isolation of the Parylene insulation was taken one step further
through the use of glass-substrate electrodes for soak testing. Results were consistent
with previous findings and suggested that insulation integrity, perhaps as a byproduct of
poor adhesion, is a point of concern and an issue that must be addressed.
Adhesion issues between Parylene and metals have been documented in literature,
raising caution over its use in wet environments. In spite of these adhesion issues,
Parylene continues to be used in medical implants in favor of its other desirable
properties and has performed admirably. For neural probe applications such as the PSE,
however, solving the issue of adhesion is critical. Several solutions have been suggested,
typically in the form of an adhesion layer or promoter. While these suggested materials
may prove to be the solution to adhesion, biocompatibility must be properly addressed.
Parylene and Pt were selected for the PSE as they have long track records of being used
in implanted devices. As such, techniques to improve adhesion that do not require the
addition of materials that may compromise the vetted biocompatibility of Parylene and Pt
are desirable. The annealing process utilized by Rodger et al. offers such a technique and
therefore was incorporated into the fabrication process for the PSE [7]. The process,
however, did not yield the necessary adhesion to improve the longevity of the PSE. The
heat treatment step of the thermoforming process is meant to create a mechanical bond by
providing energy for the polymer chains of each Parylene layer to interweave with the
other. In the procedure currently being implemented, this mechanical bond is not being
fully accomplished and solution during soaking is penetrating either through or between
the layers, leading to delamination of the Parylene. Further work is being conducted to
- 76 -
optimize the parameters of the process, such as the temperature and duration, specifically
for the PSE to yield a robust mechanical bond. Additionally, as delamination of the Pt
from the Parylene occurred first and most evidently at areas of large, uninterrupted Pt,
redesign of the Pt layout on the PSE may include design considerations to limit wide Pt
traces and electrodes.
Reliability testing conducted with the glass-substrate IDE devices indicated an
additional failure mode of the Parylene insulation. A single layer of Parylene was coated
around the entire device after the individual dies were diced, thereby leaving no layer
interface for solution to delaminate. Even so, the insulation of the devices failed, thereby
suggesting that water penetrated through the Parylene. The earlier failure of the heat
treated devices as compared to non-heat treated devices was unexpected given the
evidence in literature that the heat treatment process increases the crystallinity of
Parylene and thus decreases the water vapor permeability of the polymer [10, 11].
However, it is possible that the TCE mismatch between the glass substrate (~3 × 10
-6
/K
at 20° C) and the Parylene insulation (3.5 × 10
-5
/K at 25° C) resulted in stress at the
interface between the two materials when the devices were heat treated, thereby breaking
the bonds at the interface and compromising the integrity of the Parylene insulation and
accelerating failure [12].
While in vitro testing revealed several areas that could be addressed to improve
the PSE, in vivo work proved encouraging. Preliminary in vivo data confirmed that the
PSE technology is capable of successfully obtain neural recordings over several weeks. It
is unclear if the same delamination issues occur in vivo as were observed in vitro,
however the data suggests that whatever non-biological processes that may be occurring
in vivo are not preventing the PSE from recording neural activity. This may speak to the
inaccuracy of the in vitro models used for reliability testing. For example, the brain is not
a bath of fluid, as we have modeled it to be. Instead, the tissue in the brain provides a
measure of mechanical support to the probes. The ability of the PSE to successfully
record neural activity over several weeks despite the failure mechanisms seen in in vitro
testing highlights the need to develop better tests and models to predict the performance
of neural electrodes in vivo.
- 77 -
References
[1] D. W. L. Hukins, A. Mahomed, and S. N. Kukureka, “Accelerated aging for
testing polymeric biomaterials and medical devices,” Medical Engineering and
Physics, vol. 30, no. 10, pp. 1270-1274, 2008.
[2] P. Takmakov, ““Artificial Brain” for Aging of Neural Implants,” in Neural
Interfaces, Salt Lake City, UT, 2012.
[3] J.-M. Hsu, L. Rieth, S. Kammer, M. Orthner, and F. Solzbacher, “Effect of
thermal and deposition processes on surface morphology, crystallinity, and
adhesion of Parylene-C,” Sensors and Materials, vol. 20, no. 2, pp. 87-102, 2008.
[4] F. G. Yamagishi, “Investigation of plasma-polymerized films as primers for
parylene-c coatings on neural prosthesis materials,” Thin Solid Films, vol. 202, no.
1, pp. 39-50, 07/15, 1991.
[5] J. P. Seymour, and D. R. Kipke, “Neural probe design for reduced tissue
encapsulation in CNS,” Biomaterials, vol. 28, no. 25, pp. 3594-3607, 2007.
[6] J. Ordonez, M. Schuettler, C. Boehler, T. Boretius, and T. Stieglitz, “Thin films
and microelectrode arrays for neuroprosthetics,” MRS Bulletin, vol. 37, no. 6, pp.
590-598, 2012.
[7] D. C. Rodger, L. Wen, A. J. Fong, H. Ameri, E. Meng, J. W. Burdick, R. R. Roy,
V. R. Edgerton, J. D. Weiland, M. S. Humayun, and T. Yu-Chong, "Flexible
microfabricated parylene multielectrode arrays for retinal stimulation and spinal
cord field modulation," 2006 International Conference on Microtechnologies in
Medicine and Biology (IEEE Cat. No.06EX1357). p. 4 pp.
[8] R. Metzen, and T. Stieglitz, “The effects of annealing on mechanical, chemical,
and physical properties and structural stability of Parylene C,” Biomedical
Microdevices, pp. 1-9, 2013/03/15, 2013.
[9] B. J. Kim, J. T. W. Kuo, S. A. Hara, C. D. Lee, L. Yu, C. A. Gutierrez, T. Q.
Hoang, V. Pikov, and E. Meng, “3D Parylene sheath neural probe for chronic
recordings,” Journal of Neural Engineering, vol. 10, no. 4, pp. 045002, 2013.
[10] E. M. Davis, N. M. Benetatos, W. F. Regnault, K. I. Winey, and Y. A. Elabd,
“The influence of thermal history on structure and water transport in Parylene C
coatings,” Polymer, vol. 52, no. 23, pp. 5378-5386, Oct, 2011.
[11] P. R. Menon, W. Li, A. Tooker, and Y. C. Tai, "Characterization of water vapor
permeation through thin film parylene C," TRANSDUCERS 2009 - 15th
International Conference on Solid-State Sensors, Actuators and Microsystems. pp.
1892-1895.
[12] H. Jui-Mei, L. Rieth, S. Kammer, M. Orthner, and F. Solzbacher, “Effect of
thermal and deposition processes on surface morphology, crystallinity, and
adhesion of parylene-C,” Sensors and Materials, vol. 20, no. 2, pp. 87-102, 2008.
- 78 -

4.1 Background and aims
In light of the insulation failure that was observed with the in vitro testing
conducted on the PSE and subsequent test structures, a comprehensive study was
designed to evaluate and compare various parameters that may improve Parylene
adhesion reliability for Parylene-based electrodes. In addition to substrate treatments such
as adhesion promoters and annealing, electrode design parameters were examined. Test
devices were designed and fabricated to isolate these parameters in an effort to directly
compare the effects each may have in improving the insulation performance of Parylene
in a saline environment at body temperature.
4.1.1 Transverse and lateral impedance paths
Failure of Parylene-based electrodes can occur as a result of insulation failure
between an electrode and the surrounding environment (transverse impedance path), as
well as due to insulation failure between different electrodes within the same device,
(lateral impedance path). These impedance paths (schematics from Chapter 3 included
here in Figure 4-1 for reference) reflect slightly different failure modes of the Parylene
insulation. Failure at the lateral impedance path indicates the intrusion of electrolyte into
the device, either through the Parylene-Parylene interface or through the Parylene itself,
thereby resulting in cross-talk between electrodes, but not necessarily causing a
conductive path to the surrounding environment. Failure of the transverse impedance also
arises as a result of solution intrusion into the device. Unlike failure at the lateral
impedance path, however, transverse impedance failure indicates the creation of an
undesired fluidic connection between the electrodes and the surrounding environment.
Chapter 4
PARYLENE ADHESION RELIABILITY
- 79 -
Devices were designed and fabricated to assess both of these impedance paths. The
devices were immersed in 1× PBS and EC impedance was monitored to determine device
time-to-failure.

Figure 4-1. Cross-sectional schematics depicting (a) transverse and (b) lateral impedance
paths on Parylene-based electrodes.
4.2 Test device design, fabrication, and packaging
4.2.1 Device design
The test devices were simple Parylene-metal-Parylene sandwiches with an array
of different electrode designs on each device. The electrodes were designed to be fully
encapsulated by Parylene with no area exposed to solution. Each layer of Parylene was
designed to be 12 µm thick as a result of previous observations and conversations with
researchers at Specialty Coating Systems that indicated the need for Parylene thicknesses
>10 µm for soaking applications. Transverse devices, designed to study the transverse
impedance path, had 12 electrodes per device and lateral devices, created for the study of
the lateral impedance path, had six electrode pairs. The devices were designed so that
each channel (a single electrode for transverse devices or electrode pair for lateral
devices) was isolated from the other channels. Various techniques centered on substrate
treatments and different electrode designs were investigated for their impact on adhesion
reliability.
- 80 -

Figure 4-2. Schematic showing top view (a) and cross-sectional view (b) of a single
transverse channel.
4.2.1.1 Substrate treatments
As discussed in Chapter 1, the annealing Parylene devices has been shown to
improve soak performance [1], presumably through the enhancement of mechanical
interlocking between polymer chains in the abutting Parylene layers [2]. Additionally,
annealing has been shown to increase the crystallinity of Parylene, thereby reducing
water vapor permeability [3]. The effectiveness of this technique has been inconsistent as
one group saw that it had little effect on the adhesion between two Parylene layers [4]. It
should be noted, however, in that study, annealing was combined with the use of A-174,
which may have impeded the polymer chain entanglement expected to occur during
annealing. Since the utility of annealing is yet unproven, this technique was evaluated in
this study.
Specialty Coating Systems (SCS; Indianapolis, IN) offers two proprietary
adhesion promoters to address adhesion issues where A-174 cannot be used. AdPro Plus
was designed to improve adhesion of Parylene to metals, while AdPro Poly was designed
specifically for the adhesion of Parylene to another polymer [5]. For Parylene-based
electrodes, both Parylene-Parylene and a Parylene-metal interfaces are present. The
majority of the surface area, however, consists of the Parylene-Parylene interface.
Therefore AdPro Poly was investigated in this work.
- 81 -
4.2.1.2 Electrode design
When considering the design of Parylene-based electrodes to improve Parylene
adhesion, the key concern is the distinction between the Parylene-Parylene interface and
the Parylene-metal interface as each presents different mechanisms for adhesion. Thus,
the ratio of Parylene to metal is of interest. To this end, the side width path, that is the
lateral distance between an electrode and the exposed Parylene-Parylene interface, and
the electrode trace width were varied (Figure 4-3a). For the lateral impedance path,
electrode trace pitch was also considered (Figure 4-3b). Additionally, a unique electrode
design that aims to improve adhesion by disrupting the amount of contiguous Parylene-
metal interface present was studied (Figure 4-3c). This design has been shown in
literature to improve adhesion in a saline environment as it provides anchor points for
bonding between the two Parylene layers [6].

Figure 4-3. Electrode design parameters considered for Parylene adhesion reliability
study. Side width path and electrode trace width (a), pitch (b), and perforated electrode
trace (c) were investigated.
A full breakdown of the parameters and sample groups examined is given in
Table 4-1. Trace width, side width, and annealing sample groups were fabricated on a
glass substrate for comparison. As AdPro Poly was designed for polymer adhesion, it was
not applied to glass devices, but A-174 (Specialty Coating Systems, Indianapolis, IN)
was used for all glass devices for the best possible Parylene adhesion. Different electrode
designs were compared only under the untreated Parylene substrate condition to avoid
any confounding factors from other substrate conditions.
- 82 -
Table 4-1. Parylene-substrate sample groups for each parameter studied on transverse and
lateral impedance paths. The names of the sample groups are given in quotations.
Parameter Sample groups
Transverse Trace width
(50 µm side width)
10 µm
“tracewidth10”
20 µm
“tracewidth20”
Perforated 30 µm
“perforated”
Side width
(20 µm trace width)
10 µm
“sidewidth10”
50 µm
“tracewidth20”
100 µm
“sidewidth100”
Anneal Yes No
AdPro Poly Yes No
Lateral Trace width
(50 µm pitch)
10 µm
“tracewidth10”
20 µm
“tracewidth20”
Perforated 30 µm
“perforated”
Pitch
(20 µm tracewidth)
10 µm
“pitch10”
50 µm
“tracewidth20”
100 µm
“pitch100”
Anneal Yes No
AdPro Poly Yes No
4.2.2 Fabrication
4.2.2.1 Glass-substrate devices
Glass devices were fabricated in a two mask process on three inch borosilicate
glass wafers (Figure 4-4a). First, wafers were dehydrated and the electrode pattern was
defined through contact lithography. Platinum electrodes were deposited onto wafers
with electron beam (e-beam) evaporation. To aid adhesion of platinum to the glass
substrate, a 20 nm titanium (Ti) layer was first deposited followed by 200 nm of platinum
(Pt) without breaking vacuum. The electrodes were then patterned through a lift-off
process (Figure 4-4b). Following lift-off, wafers were cleaned through a descum process
in O2 RIE and A-174 was applied by a liquid immersion bath. 12 µm of Parylene was
deposited on top of the electrodes in the room temperature chemical vapor deposition
(CVD) process described in Chapter 1 (Figure 4-4c). The contact pads and outlines of
each channel were etched out of the Parylene with a second O2 RIE (Figure 4-4d). Finally,
individual devices were cut from the wafer with a dicing saw.
- 83 -

Figure 4-4. Cross-section view of the fabrication process for the glass devices.

Figure 4-5. Optical image of a fabricated glass device.
4.2.2.2 Parylene-substrate devices
Parylene devices were fabricated on three inch silicon (Si) carrier wafers with the
native oxide layer intact (Figure 4-6a). First, 12 µm of Parylene was deposited (Figure
4-6b). Onto this substrate layer, the electrode pattern was defined through contact
lithography and the substrate was prepared for metal deposition with an O2 RIE descum.
200 nm of Pt was then e-beam evaporated onto the substrate and lift-off patterned (Figure
4-6c). An O2 RIE descum was performed and then a second layer of Parylene, also 12
µm thick, was deposited (Figure 4-6d). For devices that used the AdPro Poly adhesion
promoter, this step and the subsequent deposition of the second Parylene layer were
carried out by SCS. Contact pads and the outlines of the devices were etched with a final
O2 RIE (Figure 4-6e). Finally, devices were immersed in deionized water and released
from the Si carrier wafer (Figure 4-6f).
- 84 -

Figure 4-6. Cross-section view of the fabrication process for the Parylene devices.

Figure 4-7. Optical image of a fabricated Parylene device.
4.2.3 Packaging
Zero insertion force (ZIF) connectors (Hirose Electric Co., Ltd., Tokyo, Japan)
were used to electrically connect devices to the test setup. These connectors are
commonly used to interface with Parylene devices since conventional techniques such as
wire bonding and soldering damage contact pads on Parylene substrates. Parylene devices
required the addition of a polyether ether ketone (PEEK) backing (McMaster-Carr
Supply Company, Elmhurst, IL) to provide the necessary thickness and stiffness for a
good connection with the ZIF connector (Figure 4-8a). Glass devices, however, could be
directly soldered to the ZIF connector (Figure 4-8b). Flat flexible cables (FFC; Johnson
- 85 -
Electric, Hong Kong, China) were then used to interface the ZIF connector with the test
setup.

Figure 4-8. Parylene device with PEEK backing for insertion into ZIF connector (a) and
glass device with ZIF connector directly soldered to contact pads (b). Scale bar applies to
both (a) and (b).

Figure 4-9. ZIF connector soldered to FFC for connection to test setup.
Each of the devices was potted with marine epoxy (Loctite Corporation,
Dusseldorf, Germany) such that only a 2 mm length of each channel would be exposed to
1× PBS when immersed. Marine epoxy was also used to secure the devices into the caps
of sample vials, providing a water tight seal to prevent solution evaporation. For
transverse devices, an additional hole was drilled into the vial caps and sealed with a
rubber stopper to allow access for the counter electrode.
- 86 -

Figure 4-10. Transverse Parylene device secured in sample vial with marine epoxy.
4.3 Testing protocol
Test devices were immersed in 1× PBS in sample vials and soaked in a waterbath
at 37° C. Although the sample vials were designed to limit evaporation, the PBS was
replaced every two weeks to maintain a constant conductivity. Two-point EIS was
conducted with a 25 mVrms perturbation signal in the frequency range of 1-100,000 Hz
with a Reference 600 potentiostat (Gamry Instruments, Warminster, PA). A Faraday cage
was used to minimize ambient electromagnetic noise. For transverse devices, a Pt wire
with an exposed surface area of 15.9 mm
2
was used as the counter electrode. For lateral
devices, the electrode pair of each channel served as working and counter electrodes.
Devices were tested daily for the first week and then weekly after that. To
facilitate the testing of such a large sample set, a solid-state multiplexer (ADG1206,
Analog Devices, Norwood, MA) was used to automate the testing of individual channels
for each device. Testing was conducted to evaluate the effect of the multiplexer (mux) on
the EIS measurement, using an electrode on the PSE as the test EC cell (Figure 4-10).
Although it was evident that the mux contributed to noise susceptibility, the changes were
minor and the benefit of automation was determined to outweigh the risk of added noise.

- 87 -

Figure 4-11. Schematics of the test setup for transverse (a) and lateral (b) devices.

Figure 4-12. Impedance magnitude (a) and phase (b) of a single electrode on a PSE in 1×
PBS. EIS measurements conducted with the ADG1206 mux (red) showed a minor
addition of noise, mostly evident in the phase, as compared to measurements taken
without the mux (black).
4.4 Results
Scanning electron microscopy (SEM) and optical microscopy images were taken
before and after soak testing to evaluate device integrity. EIS measurements at each time
- 88 -
point were normalized to the initial EIS measurement (Day 0) of that specific channel so
that the progressive failure of the channels can be compared to each other and differences
in initial impedance would not confound observations. Over the course of the study, it
was discovered that the connection to the potentiostat was not robust and the connector
had to be replaced occasionally due to wear. As a result, some measurements were found
to be unreliable and so were discarded. Additionally, some channels were excluded due
to breakage caused by handling and poor connection to the potentiostat due to corrosion
of the FFC contact pads.
4.4.1 Scanning electron microscopy
SEM revealed that the O2 RIE process used to etch out the outline of the devices
and the channels on each device produced a poorly defined edge due to the isotropic
nature of the RIE. For devices previously fabricated out of thinner layers of Parylene, the
isotropy of RIE was not an issue since the etch time was sufficiently low that lateral
etching was negligible. However, as these devices were fabricated from much thicker
layers of Parylene (each 12 µm thick), the required etch time was long enough for lateral
etching to erode the photoresist mask and begin to etch the top layer of Parylene. While
undesirable, the lateral etching was not severe enough to compromise any of the channels
and so testing was continued.
- 89 -

Figure 4-13. SEM images of glass (a,b) and Parylene (c,d) devices. The Parylene-glass
interface of a single channel (a) shows isotropic etching of the Parylene layer. The diced
edge of the devices (b) shows good adhesion between the Parylene and the glass.
Isotropic etching is clearly evident on the Parylene devices (c,d), but the two layers of
Parylene are well-adhered as no interface between the layers is visible.

Following soaking, no changes were evident under SEM. The Parylene-glass and
Parylene-glass interfaces appeared well-adhered and no defects were observed.

Figure 4-14. Representative SEM images of glass (a) and Parylene (b) devices after
soaking.
- 90 -
4.4.2 Optical microscopy
Although no changes were observed under SEM, optical microscopy revealed
delamination of the Parylene layers. On both transverse and lateral devices, delamination
was evident by the appearance of diffraction patterns around the electrodes or on the
edges of the channel. No buckling or delamination of the metal from the substrate was
observed. This delamination was evident across all substrate treatments and electrode
designs with representative images shown in Figures 4-15 and 4-16. As SEM was
performed at high vacuum and only inspects the surface of the sample, delamination was
not visible in the SEM images.

Figure 4-15. Optical micrographs of glass and Parylene transverse devices before (a,
glass; b, Parylene) and after soaking (c, glass; d, Parylene), showing the appearance of
diffraction patterns around the electrode traces.

- 91 -

Figure 4-16. Optical micrograph of a Parylene lateral device before (a) and after (b)
soaking showing diffraction patterns along the edges of the device channel and to a lesser
degree along the leftmost electrode.
4.4.3 Transverse device substrate treatments
As seen in Figure 4-17, the impedance of all substrate treatments dropped
significantly after a single day of soaking. The addition of AdPro Poly delayed the high
frequency impedance drop of electrodes, but not at low frequencies. Annealing does not
consistently improved the performance of all substrates, but it did seem to slightly
improve the glass-substrate devices. It is suspected that the initial EIS measurement for
the annealed glass data may have been inaccurate and therefore resulted in an unusually
high normalized impedance although no issues were found in the measurement setup and
test files.
- 92 -
4.4.3.1 Normalized EIS comparison

Figure 4-17. Normalized impedance magnitude comparing tested substrate treatments on
transverse devices. Data shown for measurements taken after 1 day (a), 1 week (b), and 1
month (c) of soaking. Note different scale for (a) due to annealed glass data. Mean ± SE,
n = 4-24 electrodes across 2-4 devices
4.4.3.2 Normalized 1 Hz impedance
As the most pronounced impedance changes observed in this study occurred at
low frequencies, which has also been shown in literature [7], the normalized impedance
at 1 Hz was chosen to monitor device failure over time. Across all substrate treatments,
the 1 Hz impedance dropped and to the same baseline within a week of soaking.
- 93 -

Figure 4-18. Normalized 1 Hz impedance comparing different substrate treatments used
for transverse devices. Mean ± SE, n = 4-24 electrodes across 2-4 devices.

Figure 4-19. Normalized 1 Hz impedance substrate treatment subsets that provide
pairwise comparisons: the effect of AdPro Poly (a), annealing of glass devices (b),
- 94 -
annealing of AdPro Poly devices (c), and annealing of Parylene devices (d). Mean ± SE,
n = 4-24 electrodes across 2-4 devices.
4.4.4 Lateral device substrate treatments
4.4.4.1 Normalized EIS comparison
As observed with data from the transverse devices, AdPro Poly and annealing
marginally delayed the impedance drop of the electrodes, but all substrates reached the
same baseline within a month of soaking across all frequencies, indicating complete
failure of the insulation. Among lateral devices, failure was more consistent across all
substrates than with transverse devices, indicating that this impedance path is more
vulnerable to the dominant failure mechanism of these devices.

Figure 4-20. Normalized impedance magnitude comparing tested substrate treatments on
lateral devices. Data shown for measurements taken after 1 day (a), 1 week (b), and 1
month (c) of soaking. Mean ± SE, n = 2-24 electrodes across 2-4 devices.

- 95 -
4.4.4.2 Normalized 1 Hz impedance
At 1 Hz, the annealed AdPro Poly and Parylene lateral devices exhibited some
increased variability, but again, no substrate treatments prevented the electrode
impedance from falling to less than half of its initial value within a week.

Figure 4-21. Normalized 1 Hz impedance comparing different substrate treatments used
for lateral devices. Mean ± SE, n = 2-24 electrodes across 2-4 devices.
- 96 -

Figure 4-22. Normalized 1 Hz impedance substrate treatment subsets that provide
pairwise comparisons: the effect of AdPro Poly (a), annealing of glass devices (b),
annealing of AdPro Poly devices (c), and annealing of Parylene devices (d). Mean ± SE,
n = 2-24 electrodes across 2-4 devices.
4.4.5 Transverse device electrode design
Although EIS measurements comparing electrode designs exhibited a high degree
of variability, some trends were evident. Devices with a higher metal-to-Parylene ratio
displayed slightly better performance with a smaller impedance drop than designs with a
greater proportion of Parylene. As such, although the “sidewidth10” design presents a
shorter path along the Parylene-Parylene interface between the electrode and the
surrounding solution, the impedance drop was less pronounced than the “sidewidth100”
design. Nonetheless, it is important to note that this distinction is less evident at 1 Hz and
that despite the slight difference, impedance dropped across all electrode designs to less
than 50% of the initial impedance.
- 97 -
4.4.5.1 Normalized EIS comparison

Figure 4-23. Normalized impedance magnitude comparing tested electrode designs on
transverse devices. Data shown for measurements taken after 1 day (a), 1 week (b), and 1
month (c) of soaking. Mean ± SE, n = 6-8 electrodes across 4 devices.

- 98 -
4.4.5.2 Normalized 1 Hz impedance

Figure 4-24. Normalized 1 Hz impedance comparing electrode designs used for
transverse devices. Mean ± SE, n = 6-8 electrodes across 4 devices.

Figure 4-25. Normalized 1 Hz impedance electrode design subsets that provide pairwise
comparisons: the effect of a narrower side width (a), wider side width (b), narrower
- 99 -
electrode trace (c), and a perforated electrode trace (d). Mean ± SE, n = 6-8 electrodes
across 4 devices.
4.4.6 Lateral device electrode design
For lateral devices, variation of electrode design showed very little effect on
insulation performance. Within a week of soaking, the impedance of all designs across all
frequencies dropped below 30% of their initial impedances (Figure 4-26). As with the
substrate treatment data, it is evident that the lateral impedance path is the most
susceptible to failure in saline environments.
4.4.6.1 Normalized EIS comparison

Figure 4-26. Normalized impedance magnitude comparing tested electrode designs on
lateral devices. Data shown for measurements taken after 1 day (a), 1 week (b), and 1
month (c) of soaking. Mean ± SE, n = 1-6 electrodes across 3 devices.

- 100 -
4.4.6.2 Normalized 1 Hz impedance

Figure 4-27. Normalized 1 Hz impedance comparing electrode designs used for lateral
devices. Mean ± SE, n = 1-6 electrodes across 3 devices.

Figure 4-28. Normalized 1 Hz impedance electrode design subsets that provide pairwise
comparisons: the effect of a narrower pitch (a), wider pitch (b), narrower electrode trace
(c), and a perforated electrode trace (d). Mean ± SE, n = 1-6 electrodes across 3 devices.
- 101 -
4.5 Discussion and conclusions
In this study, various substrate treatments and electrode designs were evaluated
with the intent of identifying the fabrication and design parameters most relevant for
improving the adhesion reliability, and thus the insulation integrity, of Parylene-based
electrodes. Unfortunately, all of the devices tested, no matter what substrate treatment or
electrode design was used, displayed drastic insulation failure and electrochemical
impedance drop. It is evident that the parameters investigated are not sufficient on their
own to guarantee a stable and reliable interface in a saline environment at body
temperature. Although no definitive solution to the problem of Parylene adhesion was
revealed, the results of this study do provide evidence as to the dominant mechanism of
failure for Parylene-metal-Parylene devices.
Across all substrate treatments and electrode designs, failure occurred more
quickly and uniformly amongst lateral devices than transverse devices. As the lateral
devices measure the impedance between adjacent electrodes within the same Parylene-
metal-Parylene sandwich, this distinction indicates that the mode of Parylene device
failure is solution permeation through the bulk Parylene film, not propagation between
layers. This correlates to microscopy images that showed the formation of delaminated
areas throughout the devices and not solely propagating along the etched interface in
contact with solution. Several lateral devices displayed delamination almost entirely
between the electrodes with little to no delamination evident along the periphery. This
occurred even with glass-substrate devices, in spite of the use of A-174 adhesion
promoter, as seen in Figure 4-30.

Figure 4-29. Optical image of a glass-substrate lateral device showing delamination
centered between electrodes.
- 102 -
Furthermore, SEM images did not indicate failure along the etched interface, but
instead showed the two layers well-adhered. These results may indicate that the failure
mechanism at play with these devices is that which was introduced in Chapter 1 in which
voids between the Parylene layers allow for the condensation of water vapor and
transmission of ions (Figure 4-29). Once this process is initiated, it begins a positive
feedback loop in which the condensed water vapor and transmitted ions further degrade
the adhesion at the Parylene-Parylene and Parylene-metal interfaces, resulting in
additional delamination.

Figure 4-30. Schematic of polymer failure due to the presence of voids at the interface.
For neural probes such as the Utah Electrode Array or microwires, this failure
mode may be less of an issue as each electrode channel is isolated from the others and the
main function of the insulating layer is to insulate the transverse impedance path. For
microfabricated probes, however, this failure mode is of great concern as adjacent
electrode channels are subjected to increased cross-talk if the lateral impedance path is
compromised.
With regards to different electrode designs, there was little to no distinguishing
behavior observed. A minor trend that was seen with the transverse devices was a
correlation between a larger metal-to-Parylene ratio and improved performance,
suggesting that adhesion between Parylene and metal is more robust than Parylene-to-
Parylene adhesion. This is consistent with what was found in a concurrent peel-strength
study [8] that showed an increased adhesion strength at the metal/Parylene interface when
- 103 -
compared to a Parylene-Parylene interface. It is important to note, however, that the peel-
strength study was conducted on dry samples. A follow-up peel-strength study to
investigate samples that have been soaked is planned and should shed some light on the
results observed in the work presented here.
In an effort to improve the integrity of the Parylene-based electrodes, 12 µm thick
Parylene layers were used both for the substrate and insulation layers. This was the first
time that the Biomedical Microsystems Laboratory used Parylene of this thickness for
microfabrication and the results of the fabrication process suggested a modification of the
typical etch protocol. O2 RIE was used to define the device outlines and expose the
contact pads, which had been successfully performed for similar devices with thinner
Parylene layers. Nonetheless, the isotropic nature of the RIE process proved to be ill-
suited for Parylene layers thicker than 5 µm, resulting in lateral etching of the top
Parylene layer. The lateral etch did not compromise the devices, but the results indicate
that an alternate etch be used for subsequent fabrication processes. The optimization of
deep reactive ion etching (DRIE) for thicker layers of Parylene is currently under
investigation to replace RIE for Parylene etching. Interestingly, the use of thicker
Parylene layers (12 µm, compared to 5 µm used for the PSE) did not appreciably improve
soak performance. The expectation was that a thicker Parylene layer would impede
solution intrusion by presenting a longer path across which water vapor would have to
diffuse. As detailed in Chapter 3, soak testing of the PSE showed loss of insulation after a
day of soaking and very similar performance was observed in this study. Although a
thicker Parylene layer has previously been shown to improve insulation integrity under
soaking conditions [9], the results presented here suggest that a thicker layer alone does
not prevent solution intrusion and insulation failure.
Although non-ideal for etching through thick layers of Parylene, plasma treatment
of a Parylene substrate prior to deposition of the insulation layer roughens the surface,
and is another technique that has been demonstrated to improve adhesion between the
two layers. This has been performed with argon, methane, and oxygen plasmas in situ in
the Parylene deposition chamber [10]. The use of plasmas during the fabrication process
is standard practice with Parylene-based electrodes; an oxygen reactive ion etching (RIE)
“descum” is performed to the substrate prior to the deposition of another material in order
- 104 -
to establish a clean surface prior to the deposition. This descum, however, is performed in
a separate tool from the deposition tool and so vacuum must be broken between the
descum and the subsequent deposition. It is possible that the plasma treatment is less
effective when performed with a separate tool, as the chemical changes that occur at the
Parylene surface due to the plasma may be negated once vacuum is broken.
Improving the adhesion of Parylene to different substrates is a topic of interest
across various fields with as many different techniques suggested to produce a robust
interface. Recent work by other groups has shown that a bilayer of Parylene and Al2O3
improves the insulation performance of Parylene on fused silica and silicon substrates
under soaking conditions [11, 12]. Al2O3 serves as an improved moisture barrier with a
water vapor transmission rate (WVTR) of ~10
-10
(g × mm)/(m
2
× day
1
) [13, 14], as
compared to Parylene C with a WVTR of 0.08 (g × mm)/(m
2
× day
1
) [15]. Al2O3 is
readily etched by water, but when Parylene is used to block contact with liquid water, the
improved moisture barrier properties of Al2O3 can prevent the condensation of water
vapor that can lead to delamination and insulation failure. This result has not been
confirmed on Parylene substrates, however, and results from peel-strength and
mechanical bend testing conducted in a concurrent study suggest that the addition of
Al2O3 may actually inhibit adhesion of Parylene-metal-Parylene devices [8], which may
limit its effectiveness for these types of devices.
The techniques investigated in this study, however, proved to be inadequate to
provide the complete adhesion between layers of Parylene-based electrodes in a warm
saline environment. The results did suggest, however, that the predominant mode of
failure is through the transmission of water and ions through the bulk Parylene film. This
is consistent with previously proposed theories of polymer encapsulation failure due to
the presence of voids or contaminants at the interface between the polymer and a
substrate (polymer or otherwise). As such, it is likely that a more thorough cleaning or
surface treatment process, such as immersion in a dilute bath of hydrofluoric acid [16],
can reduce or remove any voids or contaminants prior to Parylene deposition and should
be investigated. That being said, the PSE, which was fabricated with techniques similar to
those used in this study, has shown success recording neural signals in vivo for over 15
months (data in preparation). The results presented here indicate that it is very likely the
- 105 -
insulation of the PSE probes was compromised over the course of those 15 months,
suggesting that the insulation failure observed on the benchtop with sensitive techniques
such as EIS may not fully capture electrode performance in vivo. Nonetheless, reliability
studies such as the present work provide insight into the mechanisms at play in order to
inform neural probe design and develop techniques to improve chronic performance.
References
[1] W. Li, D. C. Rodger, E. Meng, J. D. Weiland, M. S. Humayun, and Y.-C. Tai,
"Flexible parylene packaged intraocular coil for retinal prostheses." pp. 105-108.
[2] H. S. Noh, Y. Huang, and P. J. Hesketh, “Parylene micromolding, a rapid and
low-cost fabrication method for parylene microchannel,” Sensors and Actuators
B-Chemical, vol. 102, no. 1, pp. 78-85, Sep, 2004.
[3] P. R. Menon, W. Li, A. Tooker, and Y. C. Tai, "Characterization of water vapor
permeation through thin film parylene C," TRANSDUCERS 2009 - 15th
International Conference on Solid-State Sensors, Actuators and Microsystems. pp.
1892-1895.
[4] C. Hassler, R. P. Von Metzen, P. Ruther, and T. Stieglitz, “Characterization of
parylene C as an encapsulation material for implanted neural prostheses,” Journal
of Biomedical Materials Research - Part B Applied Biomaterials, vol. 93, no. 1,
pp. 266-274, 2010.
[5] "Advances in Parylene Adhesion Technologies," http://scscoatings.com/adhesion-
coating-technology.
[6] R. Metzen, and T. Stieglitz, “The effects of annealing on mechanical, chemical,
and physical properties and structural stability of Parylene C,” Biomedical
Microdevices, pp. 1-9, 2013/03/15, 2013.
[7] P. Takmakov, K. Ruda, K. S. Phillips, I. S. Isayeva, V. Krauthamer, and C. G.
Welle, “Rapid evaluation of the durability of cortical neural implants using
accelerated aging with reactive oxygen species,” Journal of Neural Engineering,
vol. 12, no. 2, pp. 026003, 2015.
[8] C. D. Lee, “Strategies for Improving Mechanical and Biochemical Interfaces
Between Medical Implants and Tissue,” Biomedical Engineering, University of
Southern California, 2015.
[9] H.-w. Lo, and Y.-C. Tai, "Characterization of Parylene as a Water Barrier via
Buried-in Pentacene Moisture Sensors for Soaking Tests." pp. 872-875.
[10] A. K. Sharma, and H. Yasuda, “Effect of glow discharge treatment of substrates
on parylene‐ substrate adhesion,” Journal of Vacuum Science and Technology, vol.
21, no. 4, pp. 994-998, 1982.
[11] X. Z. Xie, L. Rieth, L. Williams, S. Negi, R. Bhandari, R. Caldwell, R. Sharma, P.
Tathireddy, and F. Solzbacher, “Long-term reliability of Al2O3 and Parylene C
bilayer encapsulated Utah electrode array based neural interfaces for chronic
implantation,” Journal of Neural Engineering, vol. 11, no. 2, pp. 9, Apr, 2014.
[12] S. Minnikanti, G. Q. Diao, J. J. Pancrazio, X. Z. Xie, L. Rieth, F. Solzbacher, and
N. Peixoto, “Lifetime assessment of atomic-layer-deposited Al2O3-Parylene C
- 106 -
bilayer coating for neural interfaces using accelerated age testing and
electrochemical characterization,” Acta Biomaterialia, vol. 10, no. 2, pp. 960-967,
Feb, 2014.
[13] E. Langereis, M. Creatore, S. Heil, M. Van de Sanden, and W. Kessels, “Plasma-
assisted atomic layer deposition of Al 2 O 3 moisture permeation barriers on
polymers,” Applied physics letters, vol. 89, no. 8, pp. 081915-081915-3, 2006.
[14] P. Carcia, R. McLean, M. Reilly, M. Groner, and S. George, “Ca test of Al 2 O 3
gas diffusion barriers grown by atomic layer deposition on polymers,” Applied
physics letters, vol. 89, no. 3, pp. 031915-031915-3, 2006.
[15] L. W. McKeen, Permeability properties of plastics and elastomers: William
Andrew, 2011.
[16] J. H. Chang, L. Bo, and T. Yu-Chong, "Adhesion-enhancing surface treatments
for parylene deposition." pp. 390-393.

- 107 -

Recent advances in medicine and technology have enabled incredible steps
toward decoding the human brain. The development and implementation of neural
technology has already improved the quality of life for many and holds tremendous
potential to further our understanding of cognitive function and behavior, to create new
therapies or neuroprosthetics, and even to change how we experience and interact with
the world. Neural probes created with MEMS technology continue to push the limits of
neural technology and indeed may be the key to unlock the enormous mystery that is the
human brain.
However, as detailed in Chapter 1, a plethora of challenges stand in the path of
realizing a stable intracortical interface. In addition to the need for a robust design to
maintain device integrity in the brain, intracortical neural probes must also manage the
brain’s immune response in order to reliably record neural activity. This is an area of
active research with several strategies being developed to address the failure modes faced
by intracortical probes.
The Parylene Sheath Electrode (PSE) integrates several of these strategies by
implementing the form factor and bioactive modulation approach of the neurotrophic
cone electrode on a flexible Parylene C substrate. In Chapter 2, this new technology was
characterized and the novel fabrication steps used to create it were systematically
evaluated through the use of electrochemical techniques to determine their impact on
electrode functionality. It was determined on the benchtop that the PSE was fully
functional, which was confirmed with the successful recording of neural activity in vivo.
In Chapter 3, additional testing was conducted to assess the reliability of the PSE,
isolating and analyzing its failure modes. Elevated temperature soak testing was initially
used for accelerated lifetime testing, but rapid failure of the devices suggested that the
Chapter 5
CONCLUSION
- 108 -
low glass transition temperature of Parylene was obfuscating the failure mechanisms that
would actually occur in vivo at body temperature. As such, subsequent soak testing was
conducted at 37° C. Nonetheless, probe failure in the form of insulation and electrode
delamination was observed. Given these results, test devices were fabricated to further
investigate Parylene insulation failure on both PSE-style devices and on a glass substrate.
Soak testing of the PSE-style devices revealed device failure within a day of soaking
preventing further testing. Glass-substrate devices failed in the lateral and transverse
impedance paths within a day and two weeks, respectively, of soaking. In spite of success
recording neural activity in vivo, the failure modes observed at the benchtop necessitated
further investigation to understand the mechanisms at play and evaluate strategies to
improve the reliability of Parylene-based electrodes.
Chapter 4 details a comprehensive study designed to do just that. Various
substrate treatments and electrode designs were utilized in the fabrication of test devices
to evaluate the effect each parameter may have in improving the performance of
Parylene-based electrodes in a warm, saline environment such as would be found in the
brain. Unfortunately, none of the parameters evaluated provided the necessary adhesion
to reliably insulate the electrodes. The measured impedance dropped drastically within a
week of soaking and delamination of the Parylene layers was observed. The results of the
study indicated that the predominant mode of failure was the transmission of solution
through the bulk Parylene film and not along the seam between Parylene layers, as was
suspected. This mechanism resulted in the earlier and more consistent failure of the
lateral impedance path than the transverse impedance path, which portends cross-talk
between adjacent channels to be a concern for Parylene-based electrodes, but also helps
focus further efforts to improve Parylene adhesion reliability.
It is my hope that the work presented in this dissertation will help direct the
design and fabrication of future neural electrodes. Failure analysis and reliability testing
is a frustrating business wrought with negative results and lacks the glamour and
excitement of proof-of-concept work. Nonetheless, it is a vital part of the engineering
process. To make something work once is the spark of innovation, but to make something
that performs consistently and reliably is when the true value of a technology is realized.
- 109 -

All EC testing in this work was conducted with a Gamry Reference 600
potentiostat (Gamry Instruments, Warminster, PA) and a Faraday cage was used to
minimize ambient electrochemical noise. A waterbath was used when temperature
controlled tests were conducted.
1.1 Electrode numbering
To clarify the nomenclature used in this document, some explanation is necessary.
At this point, two major versions of probes were developed with the difference being the
location of the electrodes on the probe exterior (top of sheath or wing). Both versions
used 3 different layout designs, designated “A,” “B,” and “C.” Each probe has 8
electrodes, 4 located inside of the sheath and 4 outside of the sheath. A few early devices
only had 4 electrodes on the inside of the sheath. To distinguish between different probes,
a naming convention was devised for the probe with electrodes on the sheath:

Probes named by:
• # of electrodes (4E or 8E)
• Probe design (Probe A = “PA”, Probe B = “PB”, Probe C (cylinder) = “PC”)
• Number of device

So, an 8 electrode probe A is named “8EPA” and to distinguish between different
8EPAs, a number is added, e.g., “8EPA_1” or “8EPA_3”

Electrode designation on a single probe:
• Inside of cone, from base to tip = 1, 2, 3, 4
• Outside of cone, from base to tip = 5, 6, 7, 8

The wing probe naming convention is as follows:
APPENDIX A
PARYLENE SHEATH ELECTRODE TESTING
PROCEDURES
- 110 -

Probes named by:
• Probe design (Probe A = “A”, Probe B = “B”, Probe C (cylinder) = “C”)
• Designated as “wing” probes (“W”)
• Number of device (2 digits)

So, a probe is designated “AW,” “BW,” or “CW” and to distinguish between different
probes of the same design, a number is added, e.g., “AW_03” or “BW_05”

Electrode designation on a single probe:
• Inside of sheath, from base to tip = 1, 2, 3, 4
• Outside of sheath, from base to tip = 5, 6, 7, 8
1.2 Electrochemical (EC) cleaning
Using a three-electrode cell, electrodes were immersed in 0.05 M H2SO4 with
constant N2 purging. The working electrode was cycled between -0.2 to 1.2 V with
respect to an Ag/AgCl (3M NaCl) reference. A 1 cm
2
Pt plate served as the counter
electrode. Using a scan rate of 250 mV/s, each electrode was cycled for 30 cycles, as it
was observed that the curves reached a steady-state at this point, indicating that no further
cleaning was occurring. The electroactive surface area (ESA) was measured by
conducting a CV in 0.05M H2SO4 as part of the EC cleaning process.

Protocol:
1. Prepare the 0.05 M H2SO4 solution and pour into a 50 mL beaker.
2. Purge the solution with N2 both before and during the CV. N2 flow should be
adjusted so that there is a constant stream of bubbles in the solution, but not so
much that the electrodes are disturbed or bubbles attach to the electrodes.
3. Rinse the PSE and counter electrode with acetone, IPA, and DI water. Gently dry
the PSE with a Kimwipe and blow dry the counter with N2.
4. Rinse the reference electrode with ddH2O and dry with a Kimwipe.
5. Insert the PSE, counter, and reference electrodes and conduct the CV:
a. -0.2 to 1.2 V vs Ag/AgCl (3M NaCl)
b. 250 mV/s
c. 30 cycles
6. IMPORTANT: Set the “max current” setting of the software to the maximum
expected current, but as close as possible. This is not a safety setting, it is to tell
- 111 -
the potentiostat what current range to measure with. Overestimating the maximum
current will set the potentiostat into the wrong current range, leading to inaccurate
and noisy CVs.
1.3 Electrochemical impedance spectroscopy (EIS)
EIS measurements of the PSE were conducted in vitro in 1× phosphate buffered
saline (PBS; all concentrations used were 1× and so this modifier will be omitted from
now on) at 37°. A 10 mVrms perturbation signal was used over the range of 1-100,000 Hz.

Protocol:
1. Immerse PSE in IPA prior to placing device in PBS to wet the sheath.
2. Place device, counter electrode, and Ag/AgCl reference electrode (if being used)
in sample vial or beaker with sufficient PBS to completely cover electrodes.
3. Using Gamry Framework software, click on “Experiment” ”Physical
Electrochemistry” ”Potentiostatic EIS”
4. Name “output file”
5. “Notes”: Device, Solution, Reference electrode used, Any other necessary info
6. “Initial Freq.” = 100,000 Hz
7. “Final Freq.” = 1 Hz
8. “AC Voltage” = 10 mV
9. “Init. Delay” = 300 s
10. “Stability” = 0.01 mV/s
1.4 Electrochemical effects of thermoforming with cyclic voltammetry in
ferrocyanide
1. Once in solution, ferrocyanide is sensitive to light and will dissociate to
ferricyanide. Prepare the solution immediately before testing.
2. Standard clean probes with acetone, IPA, and DI rinse followed by drying with a
Kimwipe.
3. First wet the surface with IPA, immerse array in 1×PBS at 37°C, and conduct EIS.
a. 1-100,000 Hz
b. 10 mVrms
c. RE = Ag/AgCl (3M NaCl), CE =1 cm
2
Pt plate
4. Standard clean probes with acetone, IPA, and DI rinse followed by drying with a
Kimwipe.
5. First wet the surface with IPA, immerse array in 6 mM K4[Fe(CN)6] (in 1× PBS)
with a 1 cm
2
Pt plate counter and Ag/AgCl (3M NaCl) reference.
6. Allow electrodes to equilibrate for 10 min.
7. Scan all electrodes of the array as follows:
- 112 -
a. Scan the working electrode -0.1 to 0.5 V at 50 mV/s and 100 mV/s, 10
cycles each, letting the solution settle for 30 min. between scans.
8. Rinse arrays in DI and dry with a Kimwipe.
1.5 Effect of sheath opening and thermoforming on the Parylene sheath
electrode
Take 2 sets of probes through post-fabrication processing via different paths. Set 1
will be EIS tested “as fabricated,” then thermoformed and EIS, and finally mechanically
opened and thermoformed and EIS. Set 2 will be EIS tested as fabricated, mechanically
opened and EIS, and finally thermoformed and EIS. Cylindrical probes with electrodes
on the wing (“CW”) were used.

1. Set 1 schedule:
a. EIS
b. EC clean
c. EIS
d. Anneal without opening and EIS
e. Mechanically open, anneal a second time, and EIS
f. EC clean
g. Final EIS
2. Set 2 schedule:
a. EIS
b. EC clean
c. EIS
d. Mechanically open, EIS
e. Thermoform (open and anneal) and EIS
3. Note:
a. Annealing without opening was seen to collapse the channel permanently
on one probe.
1.6 Electrochemical impedance spectroscopy of biofunctional coatings
1. Following thermoforming, each device is EC cleaned as described above.
2. Baseline EIS curves are taken with parameters listed below.
3. Probes are coated with their respective coating.
4. Post-coat EIS is measured.
5. Probes are placed in 37°C waterbath and EIS is conducted daily for 3 days.
EIS parameters:
• 1×PBS @37°C
• 10 mV rms perturbation signal
• 1-100,000 Hz frequency range
• CE = 1 cm
2
Pt plate
- 113 -
• RE = Ag/AgCl (3M NaCl)
Sample set CV clean Pre-coat
EIS
Post-coat
EIS
Day 1 EIS Day 2 EIS Day 3 EIS
Matrigel 9/20/12 9/21/12 9/21/12 9/23/12* 9/24/12 9/25/12
MG+NGF 10/5/12 10/5/12 10/5/12 10/7/12* 10/8/12 10/9/12
MG+Dex 10/5/12 10/5/12 10/5/12 10/7/12* 10/8/12 10/9/12
*Had to epoxy probes to vial cap and leave to cure overnight. Soaking began once epoxy
cured.
1.7 Leakage current testing on PSE-style devices
Protocol-Ztrans:
1. Take dry EIS and leakage current measurements (1 hour each electrode) for all
electrodes for the first day.
2. Soak probes in 1×PBS at 37°C, IPA-wetting is used to fill sheaths.
3. EIS curves are taken between each selected electrode and an external counter
electrode (Ztrans).
4. Leakage current measurements are measured for one electrode on each probe over
the first hour of soaking.
5. Continue leakage current measurements until next EIS measurements.
EIS parameters:
• 1×PBS @37°C
• 25 mVrms perturbation signal
• 1-100,000 Hz frequency range
• RE = CE = 1 cm
2
Pt plate
Leakage current parameters:
• 1×PBS @ 37°C
• 5 VDC
• Ztrans: RE = CE = 1 cm
2
Pt plate
Probes were designated “LC” for “leakage current” testing. All probes were
design A. 2 probes were used, with 2 electrodes from inside of sheath (electrode numbers
1 and 3) and 2 from outside of sheath (electrode numbers 6 and 8).

During first hour of soak:
LC2: measure leakage current of E1 (electrode 1)
LC4: measure leakage current of E8 (electrode 8)

Ztrans-Leakage Current/EIS Soak
Schedule
Day LC2 LC4
- 114 -
0 Ztrans-E1,EIS EIS
1 Ztrans-E3, EIS EIS
2 Ztrans-E1
3 EIS Ztrans-E3, EIS
4 Ztrans-E6
5 Ztrans-E8, EIS EIS
6 Ztrans-E6
7 EIS Ztrans-E8, EIS

Protocol-Zlat:
Same protocol as for Ztrans, except measurements taken between electrode pairs
(E1 to E2, E3 to E4, E5 to E6, and E7 to E8) instead of with an external counter electrode.

During first hour of soak:
LC5: measure leakage current of E3
LC6: measure leakage current of E6

Zlat-Leakage Current/EIS Soak
Schedule
Day LC5 LC6
0 Zlat -E1,EIS EIS
1 Zlat -E3, EIS EIS
2 Zlat -E1
3 EIS Zlat -E3, EIS
4 Zlat -E6
5 Zlat -E8, EIS EIS
6 Zlat -E6
7 EIS Zlat -E8, EIS

1.8 Leakage current on glass-substrate interdigitated electrodes
1. Four interdigitated electrode (IDE) devices were coated with Parylene.
2. Two devices are annealed in the vacuum oven at 200ºC for 48 h.
3. One side of the IDE was designated for Ztrans measurements (left side).
4. Take dry EIS and leakage current measurements (10 min. each electrode) for all
electrodes for the first day.
5. Soak IDEs in 1×PBS at 37°C.
6. Measure leakage current measurements across Ztrans for one electrode on each IDE
over the first hour of soaking.
7. Each day:
a. Remove IDEs from the soaking solution and placed in fresh PBS for
measurements.
- 115 -
b. Take EIS curves between each selected electrode and an external counter
electrode (Ztrans) as well as between the IDEs (Zlat).
c. Measure 10 min. of leakage current for both Ztrans (vs. external Pt
electrode) and Zlat (vs. the two IDE).
8. Every two weeks, replace the soaking PBS with fresh PBS.
EIS parameters:
• 1×PBS @37°C
• 25 mV rms perturbation signal
• 1-100,000 Hz frequency range
• Ztrans: RE = CE = 1 cm
2
Pt plate, Zlat: RE = CE = opposite electrode (right side)
Leakage current parameters:
• 1×PBS @ 37°C
• 5 VDC
• Ztrans: RE = CE = 1 cm
2
Pt plate, Zlat: RE = CE = opposite electrode (right side)
1.9 Accelerated life testing (ALT)
1. EC clean each electrode using the protocol detailed above.
2. Immerse probes in IPA prior to immersing probes in PBS for soaking.
3. Place probes in vials with fresh 1× PBS to begin soak.
4. When testing, remove vials from higher temperature waterbath, place in 37º C
waterbath to run EIS.
5. Allow vials to come to 37°C (~10 min.).
6. Begin EIS testing.
7. When testing is complete, return to original waterbath for soaking at higher
temperature.
8. Replace PBS in soaking vial every 2 weeks to maintain proper ionic
concentration.

- 116 -

CAUTION: To be performed under the solvents Fume Hood only.

Mixing steps:
1. Always check the shelf-life tag on the A-174 bottle. If > 6 months, solution needs
to be replaced.
2. Make 0.5% of A-174 in IPA and DI water. Mixing recipe:
a. IPA: DI H20: A-174 = 500:500:5 ml. Stir for 30 seconds with a dedicated
glass stirring rod. Make sure the solution is mixed well
b. Let it stand for 2 hours prior to use.
c. Lifetime of mixed solution is 24 hours. Must use within that time or you
need to remake the solution.
3. After 2 hours, submerge the cleaned wafer(s) in the solution for 15-30 min.
4. Air dry for 15-30 min or carefully blow dry or spin dry in spin dryer (wafer must
be dry before proceeding to next step).
a. Do not let surfaces to be coated to come into contact with dirty surfaces
(prop up wafers at an angle).
5. Immerse in IPA bath for 15-30 sec. Gently agitate wafers/parts in bath.
6. Air dry for 30-60 seconds or spin dry in spin dryer.
7. Bake dry to remove moisture in an oven for 15 min @ 100°C (wafer must be
completely dry before parylene coating).

Clean up:
1. Treat used Silane A-174 as combustible/toxic waste
a. Collect the waste in a properly labeled waste glass container and store in
flammable cabinet.
b. Contact safety office (213-740-7215) for pick up when full.
2. Store A-174 back in the flammable cabinet.
Store A-174 dedicated glassware/labware back in the designated drawer in lab.

APPENDIX B
A-174 SILANE TREATMENT
- 117 -

1.10 Process recipe for glass-substrate devices
Obtain prime 3" borosilicate wafer out of box
Dehydration bake on hotplate @ 120 °C for ~ 30 min.
Prime wafer w/ HMDS (place wafer under glass beaker along with plastic cap of HMDS)
AZ 5214 E-IR lithography for metal deposition
5 sec @ 500 rpm; 45 sec @ 2 krpm (~2 µm)
Bake 1:10 min @ 90 °C
Expose 50 mJ = 10 mW/cm
2
* 5 sec
IR bake 45 sec @ 120 °C
Global exposure 300 mJ = 10 mW /cm
2
* 30 sec
Develop: 30-35 sec
RIE descum 100 W : 100 mT : 30 sec
Ebeam Ti (20 nm)/Pt (3 runs of 666 Å for total of ~2000 Å)
Liftoff
Measure out 28 g of PxC dimer and take to cleanroom
Immerse wafers in A-174 and allow to dry in oven (see A-174 protocol, Appendix A)
Deposit Parylene (~12 µm thick, 28 g dimer)
AZ4620 double-spin lithography for etch mask for contact pads
5 sec @ 500 rpm; 45 sec @ 2 krpm (~9.6 µm)
Bake 5 min @ 90 °C
5 sec @ 500 rpm; 45 sec @ 2 krpm (~9.6 µm)
Bake 6 min @ 90 °C
Expose 540 mJ = 10 mW/cm
2
* 54 sec
Develop 50-60 sec
RIE etch 100 W : 100 mT : 5 min (12 times rotating every 3 times)
Remove PR mask with acetone/IPA/DI baths
Dice wafer
• Z-ind = 0.5
• Spd = 1 mm/s
• Cut strk = 80 mm
• Water = 1 L/min
Solder ZIFs onto contact pads, insert FFCs, and pot devices with marine epoxy down to
the first mark (2 mm from tip of electrodes)
APPENDIX C
FABRICATION PROCESS FLOW FOR
RELIABILITY TEST DEVICES
- 118 -
1.11 Process recipe for Parylene-substrate devices
This recipe will be interrupted at certain points for the application of AdPro Poly by SCS.
As it is a proprietary adhesion promoter, wafers must be sent to SCS for them to do the
coating and subsequent Parylene deposition. It should be noted that since they are
depositing the Parylene, the quality of the Parylene may vary from that which is
deposited in our lab.

Obtain prime 3" Si wafers out of box
Dehydrate wafers for at least 15 min. at 140º C
Deposit Parylene (12 µm, 28 g dimer)
AZ 5214 E-IR lithography for metal deposition
5 sec @ 500 rpm; 45 sec @ 2 krpm (~2 µm)
Bake 1:10 min @ 90 °C
Expose 50 mJ = 10 mW/cm
2
* 5 sec
IR bake 45 sec @ 120 °C
Global exposure 300 mJ = 10 mW /cm
2
* 30 sec
Develop: 30-35 sec
RIE descum 100 W : 100 mT : 30 sec
Ebeam Pt (3 runs of 666 Å for total of ~2000 Å)
Liftoff
Descum 100 W : 100 mT : 30 sec
Deposit Parylene (12 µm, 28 g dimer)
AZ 4620 double-spin lithography for etch mask for contact pads
5 sec @ 500 rpm; 45 sec @ 2 krpm for (~9.6 µm)
Bake 5 min @ 90 °C
5 sec @ 500 rpm; 45 sec @ 2 krpm for (~9.6 µm)
Bake 6 min @ 90 °C
Expose 540 mJ = 10 mW/cm
2
* 54 sec
Develop 50-60 sec
RIE contact pad etch 100 W : 100 mT : 5 min (12 times rotating every 3 times)
Remove PR mask with acetone/IPA/DI baths
Etch mask for cutout
AZ 4620
5 sec @ 500 rpm; 45 sec @ 2 krpm for (~9.6 µm)
Bake 5 min @ 90 °C
5 sec @ 500 rpm; 45 sec @ 2 krpm for (~9.6 µm)
Bake 6 min @ 90 °C
Expose 540 mJ = 10 mW/cm
2
* 54 sec
Develop 50-60 sec
RIE 100W : 100 mT : 5 min (12 times, rotating ever 3 times)

- 119 -
1.12 Masks for reliability test devices

- 120 -

- 121 -

- 122 -

Abstract (if available)

Abstract A comprehensive understanding of the human brain and the ability to tap into the power it possesses is a daunting task, but one that physicians, scientists, and engineers are tackling together. A crucial step toward attaining that goal is the development of intracortical electrodes capable of reliably recording neural activity chronically. Neural probes created with MEMS technology continue to push the limits of neural technology and indeed may be the key to unlock the enormous mystery that is the human brain. ❧ As new technologies are developed, it is paramount to evaluate and analyze them to fully understand not only how they function, but also how they may fail. As engineer and author, Henry Petroski wrote in his book Design Paradigms: Case Histories of Error and Judgment in Engineering, “It is imperative in the design process to have a full and complete understanding of how failure is being obviated in order to achieve success. Without fully appreciating how close to failing a new design is, its own designer may not fully understand how and why a design works.” In this dissertation, electrochemical techniques are implemented to evaluate the Parylene Sheath Electrode and understand its principal failure modes. An introduction of the problems encountered at the neural interface and a survey of the strategies being explored to overcome current limitations to neural probes is presented in Chapter 1. Additionally, an introduction to Parylene as a medical device material and a primer on the electrochemical techniques used with neural electrodes is provided. Chapter 2 gives the systematic analysis of the Parylene Sheath Electrode to understand the electrochemical impact of the novel fabrication processes used. Failure analysis of the probe is presented in Chapter 3 with along with the fabrication and testing of devices designed to isolate the failure mechanisms of Parylene as an insulator in a saline environment. These results prompted a comprehensive study to investigate the impact of several strategies and design parameters to improve the performance of Parylene-based electrodes, which is detailed in Chapter 4.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Development of implantable Parylene-based MEMS technologies for cortical applications

PDF

Parylene C bioMEMS for implantable devices with electrochemical interfaces

PDF

Development of micromachined technologies for neural interfaces

PDF

Strategies for improving mechanical and biochemical interfaces between medical implants and tissue

PDF

Parylene-based implantable interfaces for biomedical applications

PDF

Penetrating parylene neural probe array for dense, in vivo, chronic recordings

PDF

Modular bio microelectromechanical systems (bioMEMS): intraocular drug delivery device and microfluidic interconnects

PDF

Fabrication and packaging of three-dimensional Parylene C neural interfaces

PDF

Integrated wireless piezoelectric ultrasonic transducer system for biomedical applications

PDF

A wireless implantable MEMS micropump system for site-specific anti-cancer drug delivery

PDF

Multi-region recordings from the hippocampus with conformal multi-electrode arrays

PDF

Development of fabrication technologies for robust Parylene medical implants

PDF

Next generation neural interfaces for neural modulation and prosthesis

PDF

Parylene-based biomems sensors for multiple physiological systems

PDF

High-resolution data acquisition with neural and dermal interfaces

PDF

Thin-film impedimetric sensors for chronic in vivo use: design and application to hydrocephalus treatment

PDF

Wireless electrochemical drug delivery micropump with fully integrated electrochemical dose tracking feedback system

PDF

Liquid flow rate sensing via solution resistance for in vivo applications

PDF

Prosthetic vision in blind human patients: Predicting the percepts of epiretinal stimulation

PDF

Investigation of the electrode-tissue interface of retinal prostheses

Asset Metadata

Creator Hara, Seth Aogu (author)

Core Title The electrochemical evaluation of Parylene-based electrodes for neural applications

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Biomedical Engineering

Publication Date 07/07/2016

Defense Date 04/30/2015

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag bioMEMS,electrochemistry,microelectrodes,neural engineering,OAI-PMH Harvest,Parylene

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Meng, Ellis (committee chair), Nayak, Krishna S. (committee member), Weiland, James D. (committee member)

Creator Email sethhara@gmail.com,sethhara@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-587732

Unique identifier UC11301309

Identifier etd-HaraSethAo-3559.pdf (filename),usctheses-c3-587732 (legacy record id)

Legacy Identifier etd-HaraSethAo-3559.pdf

Dmrecord 587732

Document Type Dissertation

Format application/pdf (imt)

Rights Hara, Seth Aogu

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA