Citation
Fabrication and Characterization of Zinc Oxide and Gallium Nitride Based Sensors

Material Information

Title:
Fabrication and Characterization of Zinc Oxide and Gallium Nitride Based Sensors
Creator:
Wang, Hung Ta
Place of Publication:
[Gainesville, Fla.]
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (130 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Chemical Engineering
Committee Chair:
Ren, Fan
Committee Members:
Norton, David P.
Tseng, Yiider
Pearton, Stephen J.
Graduation Date:
5/1/2008

Subjects

Subjects / Keywords:
Annealing ( jstor )
Diodes ( jstor )
Electric current ( jstor )
Electric potential ( jstor )
Hydrogen ( jstor )
Ions ( jstor )
Low temperature ( jstor )
Nanorods ( jstor )
Room temperature ( jstor )
Sensors ( jstor )
Chemical Engineering -- Dissertations, Academic -- UF
Genre:
Electronic Thesis or Dissertation
bibliography ( marcgt )
theses ( marcgt )
Chemical Engineering thesis, Ph.D.

Notes

Abstract:
Pt-coated ZnO nanorods show a decrease of 8 % resistance upon exposure to 500 ppm hydrogen in room temperature. This is a factor of two larger than that obtained with Pd; approximately 95 % of the initial ZnO conductance was recovered within 20 s by exposing the nanorods to O2. This rapid and easy recoverability makes the ZnO nanorods suitable for ppm-level sensing at room temperature with low power consumption. Pt-gated AlGaN/GaN based high electron mobility transistors (HEMTs) showed that Schottky diode operation provides large relative sensitivity over a narrow range around turn-on voltage; the differential designed Schottky diodes with AlGaN/GaN hetero-structure was shown to provide robust detection of 1 % H2 in air at 25 ?C, which remove false alarms from ambient temperature variations; moreover, the use of TiB2-based Ohmic contacts on Pt-Schottky contacted AlGaN/GaN based hydrogen sensing diodes was shown to provide more stable operation. Thioglycolic acid functionalized Au-gated AlGaN/GaN based HEMTs were used to detect mercury (II) ions. A fast detection ( > 5 seconds) was achieved. This is the shortest response ever reported. The sensors were able to detect mercury (II) ion concentration as low as 10?7 M. The sensors showed an excellent sensing selectivity of more than 100 of detecting mercury ions over sodium, magnesium, and lead ions, but not copper. AlGaN/GaN based HEMTs were used to detect kidney injury molecule-1 (KIM-1), an important biomarker for early kidney injury detection. The HEMT gate region was coated with KIM-1 antibodies and the HEMT source-drain current showed a clear dependence on the KIM-1 concentration in phosphate-buffered saline (PBS) solution. The limit of detection (LOD) was 1ng/ml using a 20 ?m ?50 ?m gate sensing area. This approach shows a potential for both preclinical and clinical disease diagnosis with accurate, rapid, noninvasive, and high throughput capabilities. The rest of this dissertation includes ZnO band edge electroluminescence from N+-implanted ZnO bulk, and the investigation of cryogenic gold Schottky contact on GaAs for enhancing device thermal stability. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (Ph.D.)--University of Florida, 2008.
Local:
Adviser: Ren, Fan.
Statement of Responsibility:
by Hung-Ta Wang

Record Information

Source Institution:
UFRGP
Rights Management:
Copyright by Hung-Ta Wang. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
7/11/2008
Classification:
LD1780 2008 ( lcc )

Downloads

This item has the following downloads:


Full Text







10
1 20x50 gm gate
S8-

-4

I I
S4-



0

1 10 100
KIM-1 Conc. (ng/ml)

Figure 5-4 Current change in HEMT as a function of KIM-1 concentration.









0.020


,0.015-

r-
( 0.010

00.005


0.000
0


1x10 -
1xi0',
1x10-2
1x103

1x10"
lx10-W
l x10i

1x10-

1x10-10
0


1 2
Voltage(V)


(b)
-o--air
-- 1% H


i-'t.K)
I


1 2
Voltage(V)


log (b) form of Pt-gated diode measured in air or 1


Figure 3-10 I-V characteristics in linear (a) or
% hydrogen ambient at 25 C.









[149] K. S. Hwang, J. H. Lee, J. Park, D. S. Yoon, J. H. Park, and T. S. Kim, Lab Chip. 4, 547
(2004).

[150] K. W. Wee, G. Y. Kang, J. Park, J. Y. Kang, D. S. Yoon, J. H. Park, and T. S. Kim,
Biosens. Bioelectron. 20, 1932 (2005).

[151] R. J. Chen, S. Bangsaruntip, K. A. Drouvalakis, N. W. S. Kam, M. Shim, Y. Li, W. Kim,
P. J. Utz, and H. Dai, Proc. Natl. Acad. Sci. USA 100, 4984 (2003).

[152] C. Li, M. Curreli, H. Lin, B. Lei, F. N. Ishikawa, R. Datar, R. J. Cote, M. E. Thompson,
and C. Zhou, J. Am. Chem. Soc. 127, 12484 (2005).

[153] G. Zheng, F. Patolsky, Y. Cui, W. U. Wang, and C. M. Lieber, Nature Biotechnology 23,
1294 (2005).

[154] F. Patolsky, G. Zheng, and C. M. Lieber, Nanomedicine 1, 51 (2006).

[155] F. Patolsky, G. Zheng, and C. M. Lieber, Nature Protocals 1, 1711 (2006).

[156] F. Patolsky, B. P. Timko, G. Zheng, and C. M. Lieber, MRS Bulletin 32, 142 (2007).

[157] H.-T. Wang, B. S. Kang, T. F. Chancellor, Jr., T. P. Lele, Y. Tseng, and F. Ren, S. J.
Pearton, J. W. Johnson, P. Rajagopal, J. C. Roberts, E. L. Piner, and K. J. Linthicum, Appl.
Phys. Lett. 91, 042114 (2007).

[158] R. Thadhani, M. Pascual, and J. V. Bonventre, N. Engl. J Med. 334, 1448 (1996).

[159] G. M. Chertow, E. M. Levy, K. E. Hammermeister, F. Grover, and J. Daley, Amer. J Med.
104, 343 (1998).

[160] J. V. Bonventre and J. M. Weinberg, J. Am. Soc. Nephrol. 14, 2199 (2003).

[161] T. Ichimura, J. V. Bonventre, V. Bailly, H. Wei, C. A. Hession, R. L. Cate, and M.
Sanicola, J. Biol. Chem. 273, 4135 (1998).

[162] S. J. Pearton, D. P. Norton, K. Ip, Y. W. Heo, and T. Steiner, Prog. Mater. Sci. 50, 293
(2005).

[163] D. C. Look, Mater. Sci. Eng. B 80, 383 (2001).

[164] G.-C. Yi, C. Wang, and W. I. Park, Semicond. Sci. Technol. 20, S22 (2005).

[165] Ya. I. Alivov, E. V. Kalinina, A. E. Cherenkov, D. C. Look, B. M. Ataev, A. K. Omaev,
M. V. Chukichev, and D. M. Bagnall, Appl. Phys. Lett. 83, 4719 (2003).

[166] A. Osinsky, J. W. Dong, M. Z. Kauser, B. Hertog, A. M. Dabiran, P. P. Chow, S. J.
Pearton, O. Lopatiuk, and L. Chernyak, Appl. Phys. Lett. 85, 4272 (2004).

[167] B. W. Thomas and D. Walsh, Electron. Lett. 9,362 (1973).









or in a 500 ppm H2 in N2 ambient. The FET mode of operation (monitoring of change in drain-

source current) shows a much larger signal over a broad range of gate voltage. By sharp contrast,

the diode mode of detection shows a large relative change in current only at high forward gate

biases. This shows the advantage of using the 3-terminal device structure, with its attendant

current gain. The percentage change in both drain-source and gate currents when the hydrogen is

introduced into the measurement ambient are shown at the bottom of Figure 3-3. The relative

change can be much larger in the diode mode of operation at small forward bias (-1 V) due to

the lower baseline current. At higher forward bias the effects of series resistance dominate the

current both in N2 and H2-containing ambients.

Figure 3-4 shows some of the recovery characteristics of the HEMTs upon multiple

cycling of the ambient from N2 to 500 ppm H2 in N2 and back again The sensors show good

recyclability and recovery in both modes of operation. Once again the change in drain-source

current is much larger in the FET mode. The initial response to hydrogen in both cases is rapid

(<5 sec), while the recovery back to the N2 ambient value takes much longer (100-200 secs)

because of the mass transport characteristics of gas in our test chamber. From the fast initial

response of the sensors in both modes, the effective diffusivity of the atomic hydrogen through

the Pt is >4x10-13 cm2/V's at 25 C. This is only an estimate, since the response time includes the

gas flow dynamics of the gas into the test chamber, the dissociation of the molecular hydrogen

and the diffusion to the Pt/AlGaN interface of the atomic hydrogen.

In conclusion, Pt/AlGaN/GaN HEMTs operated in either a diode mode or in an FET mode

show the ability to detect 500 ppm H2 in N2 at room temperature. The FET mode provides a

larger total current change with introduction of hydrogen into the ambient, but the diode mode

shows a higher relative sensitivity over a limited range of forward biases.









CHAPTER 6
ZNO BASED LIGHT EMITTING DIODE

6.1 Background

ZnO is attracting renewed interest for use in blue/UV light-emitting diodes (LEDs) and

photodectors with potential advantages over the III-nitride system due to the higher exciton

binding energy, availability of high quality bulk substrates and ease of wet etching [162-166].

The reports of ZnO metal-insulator-semiconductor (MIS) electroluminescent diodes go back to

the 1970"s, with most of the emission being due to defect bands in the blue/green and infra-red

(IR) [167-171]. However, in some cases, small band-edge emission was observed at low

temperatures [171], with little understanding of the origin of the holes in these n-type ZnO

structures. Generally, no electroluminescence was observed in these devices in the reverse bias

or without the i-layer. More recently, a number of groups have reported hybrid heterojunction

LEDs using n-type ZnO deposited on top of p-type layers of GaN, AlGaN or conducting oxides

[165, 166, 172-175]. Homojunction ZnO LEDs have been reported by Tsukazaki [176, 177] who

used temperature modulation epitaxy for p-type doping of ZnO using N as dopant and fabricated

a p-ZnO/i-ZnO/n-ZnO LED on a ScAlMgO4 substrate. Most of the emission consisted of bands

at 420 and 500 nm, with a small shoulder at 395 nm assigned to radiative recombination in the p-

ZnO through donor-acceptor pair transitions. Another homojunction ZnO LEDs have also been

recently reported by Jae-Hong [178, 179] who used rf sputtering technique for P dopant p-type

ZnO and fabricated p-ZnO/n-ZnO LED. The emission consisted of a near band edge emission at

380nm and broad deep level emission at approximately 640 nm. With p-ZnO/barrier-MgZnO/n-

ZnO/b-MgZnO/n-ZnO structure, carrier combination process is confined in high quality n-type

ZnO thin film and the defect related emission at 640 nm is removed. In addition, it has been

suggested that semiconducting nanowires may offer additional advantages for light emission due









3-2 Gate I-V characteristics at 0 V IDS measured at 25 C under pure N2 ambient or in a
500 ppm H 2 in N 2 am bient. ....................................................................... ....................48

3-3 Change in drain-source or gate currents as a function of gate voltage (top), and
percentage changes in these currents (bottom) for measurement under pure N2
ambient or in a 500 ppm H2 in N2 ambient .................................... ......... ............... 49

3-4 Time dependence of drain-source (top) or gate current (bottom) when switching from
pure N2 ambient to a 500 ppm H2 in N2 ambient and back again ....................................50

3-5 Microscopic images of differential sensing diodes. The opening of the active diode
was deposited with 10 nm of Pt, and the reference diode was deposited with Ti/Au. ......52

3-6 Absolute (a) and differential (b) current of HEMT diode measured at 25 C ..................53

3-7 Absolute (a) and differential (b) current of HEMT diode measured at 50 C...................54

3-8 Time dependent test of differential HEMT diodes at 25 and 50 C. ................................55

3-9 Schematic of HEMT diode hydrogen sensor using either conventional or TiB2-based
O h m ic co n tacts...................................................... ................ 5 7

3-10 I-V characteristics in linear (a) or log (b) form of Pt-gated diode measured in air or 1
% hydrogen ambient at 25 C...... ........................... .........................................58

3-11 Time-dependence of current test biased by 1.5 V of Pt-gated diode as the ambient is
switched from air to 1 % hydrogen and back to air. ................. ............................... 59

3-12 Variation in forward current at fixed bias for diodes with boride-based Ohmic
contacts (top) or conventional Ohmic contacts (bottom) as a function of time under
field conditions where the temperature increases during the day and decreases at
night .......................................................... ....................................59

4-1 (a) A schematic of AlGaN/GaN HEMT. The Au-coated gate area was functionalized
with thioglycolic acid. (b) Plan view photomicrograph of a completed device with a
5 nm Au film in the gate region. ............................................... .............................. 65

4-2 Photographs of contact angle of water drop on the surface of bare Au (left) and
thioglycolic acid functionalized Au (right) ........................................ ........ ............... 65

4-3 (a) Time dependent response of the drain current for bare Au-gate AlGAN/GaN
HEMT sensor and thioglycolic acid functionalized Au-gate HEMT sensor. (b) Drain
current of a thioglycolic acid functionalized Au-gate HEMT sensor as a function of
the H g2 ion concentration. ...................................................................... ....................66

4-4 Time dependent response of the drain current for detecting Na+, Mg2+ or Hg2+ with a
thioglycolic acid functionalized Au-gate HEM T sensor. ..................................................67










101
100 ,
E 10"1 -A
AA a,

Au 1000 A A "O
S10- A
S10 200 pm circle A'[
10- 300K
O 10o7 77 K
10 8
10 -
-6 -5 -4 -3 -2 -1 0
Voltage(V)


101
100

E 10"'
0-3,
10
~- 10"-

0 10
6io
-6


-5 -4 -3 -2 -1 0 1
Voltage(V)


0.2 0.4 0.6 0.8
Voltage(V)


0.0 0.2 0.4 0.6 0.8
Voltage(V)


Figure 7-1 I-V characteristics of Au/GaAs Schottky diodes deposited at either 77 K(A) or 300
K(o) for both 200 km dia.(left) and 800 km dia.(right) contact. An expanded view of
the forward voltage part of the curves is shown at bottom.


1x1 01 300K(RT) 100nm Au
--- 200 pm dia.
-,- 400 mpr dia. C
1 x10 r 800 pm dia.,0

1x10-3 r


7 t77

1xio10 c' -
1x10- '

0.0 0.2 0.4 0


"A
a
*I


K(LT) 100nm Au
-200 pm dia.
- 400 pm dia.
800 pm dia.


.6


Voltage(V)


0.8 1.0


Figure 7-2 Forward current densities as a function of bias for diodes of different diameter
deposited at either 77 or 300 K.


6A A


A Q AI

AuOO 1000 A A% A
800 pm circle 1
--o- 300 K
a 77 K 1
. I 1


101
10
10-1
E 10



10-
o 10
-106


10
10


o
A

o/ /
p A,
D A- Au 1000 A
P AA 200 pm circle
----300 K
A-77K


0


CN
E



I-
L

O









CHAPTER 2
HYDROGEN SENSOR USING MULTIPLE ZNO NANORODS

2.1 Background

There is a strong need to develop hydrogen sensors for use with proton-exchange

membrane and solid oxide fuel cells for space craft and other long-term applications. A key

requirement for these sensors is the ability to selectively detect hydrogen at room temperature

with minimal power use and weight. Nanorods and nanotubes are potential candidates for this

type of sensing. In the case of hydrogen sensing with carbon nanotubes (CNTs), several groups

have reported that use of Pd doping or films or loading with Pd nanoparticles can functionalize

the surface of nanotubes for catalytic dissociation of H2 to atomic hydrogen [48, 49]. Of course,

thin-film sensors of Si, GaAs, InP, SiC, and GaN with Pd contacts have been used previously as

hydrogen sensors [64]. ZnO nanowires and nanorods have shown potential for use in gas,

humidity, and chemical sensing [65, 66, 67]. The ability to make arrays of nanorods with large

total surface area has been demonstrated with a number of different growth methods [5, 68-89]

and a large variety of ZnO one-dimensional structures has been demonstrated [87]. The large

surface area of the nanorods and bio-safe characteristics of ZnO makes them attractive for both

chemical sensing and biomedical applications.

In this chapter, we demonstrate hydrogen detection use catalytic metal coated ZnO

nanorod field effect based sensor. Section 2.2 presents a comparison of different metal coating

layers on multiple ZnO nanorods for enhancing the sensitivity to detection of hydrogen at room

temperature. Pt is found to be the most effective catalyst, followed by Pd. The resulting sensors

are shown to be capable of detecting hydrogen in the range of ppm at room temperature using

very small current and voltage requirements, and recovering quickly after the source of hydrogen

is removed. In section 2.3, issues like quantifying the sensitivity, limit of detection (LOD) at









The electrical properties of the devices, source and drain current (IDS) vs. voltage (VDs),

were measured in PBS buffer and 100 ng/ml KIM-1 in PBS buffer, as shown in Figure 5-2.

There is a clear conductance decrease with KIM-1 exposure and this suggests that through the

selective binding of KIM-1 with antibody, there are charges accumulated at the surface and these

surface charges are transduced into a change in the carrier concentration of AlGaN/GaN 2DEG,

leading to the obvious decrease in the conductance of the device after KIM-1 exposure.

Figure 5-3 shows the time dependent source-drain current signal with constant bias of 500

mV for KIM-1 detection in PBS buffer solution. No current change can be seen with the addition

of buffer solution around 50 sec. This stability is important to exclude possible noise from the

mechanical change of the buffer solution. By sharp contrast, the current change showed a rapid

response in less than 20 seconds when target 1 ng/ml KIM-1 was switched to the surface at 150

sec. The abrupt current change due to the exposure of KIM-1 in a buffer solution stabilized after

the KIM-1 thoroughly diffused into buffer to reach a steady state. 10 ng/ml KIM-1 was then

applied at 350 sec and it was accompanied with a larger signal correlated to the higher KIM-1

concentration. Further real time tests were carried out to explore the limit of detection of KIM-1

antibody (Figure 5-4). The device was exposed to 10 pg/ml, 100 pg/ml, Ing/ml, 10ng/ml, and

100ng/ml individually and each concentration was repeated five times to obtain the standard

deviation of source-drain current response for each concentration. The limit of detection of this

device was Ing/ml KIM-1 in PBS buffer solution and the source-drain current change is

nonlinearly proportional to KIM-1 concentration. Between each test, the device was rinsed with

a wash buffer of 10 tM phosphate buffer solution containing 10 gM KC1 with pH 6 to strip the

antibody from the antigen. These results suggest that our HEMTs are compatible with AKI

biomarker, KIM-1, are very sensitive compared to nano-devices [151-156] and are useful for



















Figure 7-5 Optical microscope images of Au contacts deposited at 77 K (left) or 300 K (right).


108
107
10
105
104
103
102
101
100


1 2 3 4
theta-2theta


5 6


Figure 7-6 XRR of thin (-90 A) Au layers of GaAs for the two different deposition temperatures
and the associated Au surface roughness and Au/GaAs interfacial roughness derived
from the XRR.


100 A Au/GaAs
-r 77 K (LT) deposition
- 300 K (RT) deposition




, ,, ,









[185] V. A. Coleman, H. H. Tan, C. Jagadish, S. O. Kucheyev, and J. Zou, Appl. Phys. Lett. 87,
231912 (2005)

[186] E. Alves, E. Rita, U. Wahl, J. G. Correia, T. Monteiro, J. Soares, and C. Boemare, Nucl.
Instrum. Meth. Phys. Res. B 206, 1047 (2003).

[187] T. S. Jeong, M. S. Han, C. J. Youn, and Y. S. Park, J. Appl. Phys. 96, 175 (2004).

[188] F. Reus, C. Kirchner, Th. Gruber, R. Kling, S. Maschek, W. Limmer, A. Waag, and P.
Ziemann, J Appl. Phys. 95, 3385 (2004).

[189] K. Lorenz, E. Alves, E. Wendler, O. Bilani, W. Wesch, and M.Hayes, Appl. Phys. Lett. 87,
191904 (2005).

[190] G. Braunstein, A. Muraviev, H. Saxena, N. Dhere,V.Richter, and R. Kalish, Appl. Phys.
Lett. 87, 192103 (2005).

[191] H. S. Yang, D. P. Norton, S. J. Pearton, and F. Ren, Appl. Phys. Lett. 87, 212106 (2005).

[192] G. D. Mahan, L. M. Levinson, and H. R. Philipp, J. Appl. Phys. 50,2799 (1979).

[193] G. E. Pike, S. R. Kurtz, P. L. Gourley, H. R. Philipp, and L. M. Levinson, J Appl. Phys.
57, 5512 (1985).

[194] T. M. Barnes, K. Olson, and C. A. Wolden, Appl. Phys. Lett. 86, 112112 (2005).

[195] See, for example, R. E. Williams, GaAs Processing Techniques, Artech House, Dedham,
MA (1984).

[196] See, for example, L. F. Eastman and U. K. Mishra, IEEE Spectrum Magazine, 39(5), 28
(2002).

[197] H. Xing, P. Chavarkar, S. Keller, S. P. DenBaars, and U. K. Mishra, IEEE Electron. Dev.
Lett. 24, 141 (2003).

[198] S. Rajan, H. Xing, D. Jena, S. P. DenBaars, and U. K. Mishra, Appl. Phys. Lett. 84, 1591
(2004).

[199] J. Sun, H. Fatima, A. Koudymov, A. Chitnis, X. Hu, H.-M. Wang, J. Zhang, G. Simin, J.
Yang, and M. A. Khan, IEEE Electron Dev. Lett. 24, 375 (2003).

[200] A. P .Zhang, L. B. Rowland, E. B. Kaminsky, J. W. Kretchmer, R. A. Beaupre, J. L.
Garrett, J. B. Tucker, B. J. Edward, J. Foppes, and A. F. Allen, Solid-State Electron 47,
821 (2003).

[201] A. P. Zhang, L. B. Rowland, E. B. Kaminsky, V. Tilak, J. C. Grande, J. Teetsov, A.
Vertiatchikh, and L. F. Eastman, J. Electron. Mater. 32, 388 (2003).

[202] A. Z. Wang and W. A. Anderson, Appl. Phys. Lett. 66, 1963 (1995).


128









[76] Y. W. Heo, V. Varadarjan, M. Kaufman, K. Kim, D. P. Norton, F. Ren, and P. H. Fleming,
Appl. Phys. Lett. 81, 3046 (2002).

[77] D. P. Norton, Y. W. Heo, M. P. Ivill, K. Ip, S. J. Pearton, M. F. Chisholm, and T. Steiner,
Mater. Today 7, 34 (2004).

[78] P. J. Poole, J. Lefebvre, and J. Fraser, Appl. Phys. Lett. 83, 2055 (2003).

[79] M. He, M. M. E. Fahmi, S. Noor Mohammad, R. N. Jacobs, L. Salamanca-Riba, F. Felt, M.
Jah, A. Sharma, and D. Lakins, Appl. Phys. Lett. 82, 3749 (2003).

[80] X. C. Wu, W. H. Song, W. D. Huang, M. H. Pu, B. Zhao, Y. P. Sun, and J. J. Du, Chem.
Phys. Lett. 328, 5 (2000).

[81] M. J. Zheng, L. D. Zhang, G. H. Li, X. Y. Zhang, and X. F. Wang, Appl. Phys. Lett. 79, 839
(2001).

[82] S. C. Lyu, Y. Zhang, H. Ruh, H. J. Lee, H. W. Shim, E. K. Suh, and C. J. Lee, Chem. Phys.
327, 104 (1999); B. P. Zhang, N. T. Binh, Y. Segawa, K. Wakatsuki, and N. Usami, Appl.
Phys. Lett. 83, 1635 (2003).

[83] W. I. Park, Y. H. Jun, S. W. Jung, and G. Yi, Appl. Phys. Lett. 82, 964 (2003).

[84] B. D. Yao, Y. F. Chan, and N. Wang, Appl. Phys. Lett. 81, 757 (2002).

[85] J. Y. Lao, J. Y. Huang, D. Z. Wang, and Z. F. Ren, Nano Lett. 3, 235 (2003).

[86] X. W. Sun, S. F. Yu, C. X. Xu, C. Yuen, B. J. Chen, and S. Li, J. Appl. Phys. 42, L1229
(2003).

[87] Z. L. Wang, Mater. Today 7, 26 (2004).

[88] A. A. Tomchenko, G. P. Harmer, B. T. Marquis, and J. W. Allen, Sens. Actuators B 93, 126
(2003).

[89] H. L. Hartnagel, A. L. Dawar, A. K. Jain, and C. Jagadish, Semiconducting Transparent
Thin Films (IOP, Bristol, 1995).

[90] Y. W. Heo, D. P. Norton, L. C. Tien, Y. Kwon, B. S. Kang, F. Ren, S. J. Pearton, and J. R.
LaRoche, Mater. Sci. Eng. R 47, 1 (2004).

[91] See the databases at http://www.rebresearch.com/H2perm2.htm and
http://www.rebresearch.com/H2sol2.htm

[92] W. Eberhardt, F. Greunter, and E.W. Plummer, Phys. Rev. Lett. 46, 1085 (1981).

[93] K. D. Mitzner, J. Sternhagen, and D. W. Galipeau, Sens. Actuators B 93, 92 (2003).

[94] P. Mitra, A. P. Chatterjee, and H. S. Maiti, Mater. Lett. 35, 33 (1998).









electrons will fill the triangular well and form 2DEG. From another viewpoint, as described by

Ambacher et al., 2DEG is the compensation to a fixed sheet charge induced by both spontaneous

polarization (Psp) and piezoelectric or strain-induced polarization (PpE) [38, 39, 40]. The

spontaneous polarization for AlxGal-xN is,

P, (x) = (-0.052x- 0.029) C/m2, (1-1)

where the x is the Al concentration in AlxGal-xN. In both A1N and GaN system, the

spontaneous polarization is negative meaning that the spontaneous polarization is pointing

toward substrate for Ga(Al)-face, and toward surface for N-face. The piezoelectric polarization

can be calculated by,

PPE =e33z +e31Yx + Yy) (1-2)

where e33 and e31 are piezoelectric coefficients, yZ = (c c, /Co) is the strain in c-axis,

yx = y, = (a /a0) is the in-plane strain that is assumed isotropic, and a, c are the lattice

constants of the strained layer. Negative piezoelectric polarization represents that a GaN layer

under compressive strain, and/or a AlGaN layer under tensile strain for Ga(Al)-face crystal.

For Ga-face AlGaN/GaN grown on c-A1203 substrate, tensile strain AlGaN layer contact

with relaxed GaN layer and the spontaneous polarization and piezoelectric polarization align in

parallel (Figure 1-3). The fixed charge induced by total polarization can be derived by,

a = P(bottom) P(top) = Pp (GaN) [Pp (AlGaj _N) + PE (AlGal _N)]
= [P,(GaN) P(AlxGa l N)]+ [0 Pp(AlGaj-xN] (1-3)
= a(P ) + a(PP)

The polarization induced sheet charge in this case is positive (+ a ), and free electrons,

therefore, tend to compensate it to form 2DEG at GaN interface with a sheet carrier

concentration (ns), which is expected to be,









However, others have disputed this interpretation and suggested the results were due to an

inhomogeneous Schottky barrier height in the diodes due to a dependence of the local interface

dipole on the local interface structure [205].

This chapter mainly examines the effect of cryogenic deposition temperatures on the

properties of Au Schottky contacts on n-type GaAs. Section 7.2 presents the role of deposition

temperature on the electrical properties of Au/GaAs diodes. We find the barrier height is

increased by cryogenic deposition and the interfacial roughness is decreased. Section 7.3

examines the effect of post-deposition annealing temperature on the barrier height and reverse

breakdown voltage of Au/n-GaAs diodes deposited at either 77 or 300K. The barrier height is

increased by cryogenic deposition and remains higher throughout the annealing temperature

range up to 300 C. The reverse breakdown voltage is also increased by the low temperature Au

deposition. Finally, section 7.4 reports on X-ray reflectivity (XRR) studies of the interface

between Au deposited on n-GaAs at either 77 K or 300 K, followed by post-deposition annealing

at temperatures up to 300 C, for comparing with the results in section 7.2 and 7.3. The barrier

height is increased by -0.09 eV by cryogenic deposition relative to the room temperature

deposition value of 0.73 eV. This is accompanied by a smoother metal surface, while the

metal/GaAs interfacial roughness is similar. As the diodes are annealed to 300 OC, the barrier

height enhancement disappears; the Au/GaAs interfaces continue to show the same degree of

roughness while the metal surface becomes rougher. Other metals such as Pt, Ti, Pd and Ni were

also examined for barrier height enhancement.









BIOGRAPHICAL SKETCH

Hung-Ta Wang was born in Nantou, Taiwan, in 1977. He received both the bachelor and

the master degrees in the Department of Chemical Engineering of National Cheng Kung

University, Tainan, Taiwan, in 1999 and 2001 respectively. He attended Taiwan Army for the

mandatory training and service, and served as vice company commander with the position of 1st

lieutenant officer from 2001 to 2003. After military service, he worked for Taiwan

Semiconductor Manufacturing Company (TSMC) as a process integration engineer monitoring

front-end process until 2004.

In 2004 Fall, he was enrolled in the Ph.D. program of the Department of Chemical

Engineering of the University of Florida. He was under the guidance of Professor Fan Ren

studying wide bandgap semiconductor chemical/bio sensors, light emitting diodes, as well as

high speed devices. He earned his Doctor of Philosophy degree from the Chemical Engineering

Department of the University of Florida in May 2008 with 2 filed patents, 32 SCI journal

publications in highly recognized journals (Applied Physics Letters, Nanotechnology, Journal of

Electronic Materials, Applied Surface Science, Electrochemical and Solid-State Letters, etc.),

and 8 international conference oral presentations.












1 0x10"
0.0

$ -1.0x104
0,4
C -2.0x104

3 -3.0x104

-4.0x 104

-5.0x104


0.02 0.04 0.06 0.08
Dia. (cm)


1.0x10"4
o ..
_,O -4
-1.0x10 a

-2.0x10'4 0

-3.0x104 Leakage current @ -4V`
1000 A Au
-4.0x10 --o- 300 K o
77K
-50x1.000 0.001 0.002 0.003 004 0.005 0.006 0007
Dia.2 (cm2)


-5
1.0x10

5.0x10"6

< 0.0

) -5.0x10

-1.0x105

-1.5x105

-2.0x105


Leakage current @ -4V
1000 A Au
-*-77K
S-
0.
S' -


0.02 0.04 0.06
Dia. (cm)


0.08


1.0x10"5 .

5.0x10"6 Leakage current @ -4V
1000 A Au
0.0 --77 K

-5 0x10-"

-1 0x10" -- -

-1.5x10"5

-2 0x 10' .
-2 O :..00, 0.001 0.002 0.003 0.004 0.005 0.006 0.007

Dia.2 (cm2)


Figure 7-3 Reverse current at -4 V for diodes deposited at either 77 or 300 K, as a function of
either contact diameter (top) or area (bottom).


1.1-i


0.8


0.6


0.4


0.2


0.0-
0


Au 1000 A
400 um diode
77 K(LT)
-300 K(RT)


20 40 60 80 100120140 160180
Time (s)


Figure 7-4 Time dependence of forward bias at a current of 10 mA for diodes deposited at either
77 or 300 K.


-* : .-* .- -..
0o

"0

Leakage current @ -4V
1000 A Au
o 300 K o
- 77 K


t


-

-









low power requirements and minimal weight. Combined with the native advantageous

characteristics of ZnO and GaN, ZnO and GaN nanorods/wires are natural candidates for this

type of sensing application.

GaN/AlGaN high electron mobility transistors (HEMTs) have been extremely useful for

gas and liquid sensor for primarily two reasons: 1) they consist of a high electron sheet carrier

concentration channel induced by piezoelectric polarization of the strained AlGaN layer and 2)

the carrier concentration strongly depends on the ambient [7-9]. In addition, sensors fabricated

from these wide band-gap semiconductors could be readily integrated with solar blind UV

detectors or high temperature, high power electronics on the same chip. For these reasons, nitride

HEMTs are versatile devices that may be used for a variety of sensing applications.

On the other hand, ZnO is an attractive candidate for Ultra Violent (UV) Light Emitting

Diodes (LEDs) since it is an environmentally friendly material which is grown at low

temperatures on cheap transparent substrates and has both a direct wide band gap of 3.3 eV and a

very large exciton binding energy of 60 meV, important for robust light emission [10, 11].

Finally, it is important to increase Schottky barrier height in order to solve the reliability issues

to compound semiconductor based HMETs. In particular, the gate reliability has been

problematic. Increasing the Schottky barrier heights can improve the gate leakage current, gate-

drain breakdown voltage, output resistance and power gain, and noise performance. Promising

results engineering Schottky barrier heights have been demonstrated by cryogenic metal

deposition at 77 K for GaAs, InP, and InGaAs [12, 13, 14].









NO2:

02:

03:

Pb:

PBS:

Pd:

PEI:

PEM:

PIN diode:

PL:

PMMA:

ppb:

ppm:

Pt:

RTA:

SiC:

SnO2:

TE:

TFE:

Ti:

TiB2:

UV:

XPS:

XRR:


nitrogen dioxide

Oxygen

Ozone

Lead

Phosphate Buffered Saline

Palladium

polyethyleneimine

proton-exchange membrane

Positive-Intrinsic-Negative diode

photoluminescence

polymethyl methacrylate

parts per billion

parts-per-million

Platinum

Rapid Thermal Annealing

Silicon Carbide

Tin(IV) oxide

Thermionic Emission

Thermionic Field Emission

Titanium

Titanium Boride

Ultra Violent

X-ray Photoelectron Spectroscopy

X-ray reflectivity










0.08

0.06

<" 0.04

0.02

0.00


0 2 4 6 8
Vds(V)


"4

E
E 3
E 2
-1


100
80
E
60
40oE
20 o

0


-4 -2 0 2
V (V)
g\


Figure 3-1 IDs-VDs characteristics (top) and transfer characteristics (bottom) of Pt-gated HEMT
measured at 25 C under pure N2 ambient or in a 500 ppm H2 in N2 ambient.









nanorods suitable for practical applications in hydrogen selective sensing at ppm levels at room

temperature with -0.4 mW power consumption.

In conclusion, both Pt-coated and Pd-coated ZnO nanorods appear well suited to detection

of ppm concentrations of hydrogen at room temperature. Pd coated nanorods showed better

recovery characteristics. The ZnO nanorods can be placed on cheap transparent substrates such

as glass, making them attractive for low-cost sensing applications, and can also operate at very

low power conditions. Of course, there are many issues still to be addressed, in particular

regarding the reliability and long-term reproducibility of the sensor response before it can be

considered for space-flight applications. In addition, the slow response of the Pt coated sensors at

room temperature is a major issue in some applications.



8.2 Hydrogen Sensor Using AIGaN/GaN Schottky Diode and High Electron Mobility
Transistor

Pt-gated AlGaN/GaN high electron mobility transistors can be used as room temperature

hydrogen gas sensors at hydrogen concentrations as low as 100 ppm. A comparison of the

changes in drain and gate current-voltage (I-V) characteristics with introduction of 500 ppm H2

into the measurement ambient shows that monitoring the change in drain-source current (FET

mode) provides a wider gate voltage operation range for maximum detection sensitivity and

higher total current change than measuring the change in gate current (Schottky diode mode).

However, over a narrow gate voltage range, the relative sensitivity of detection by monitoring

gate current changes (Schottky diode mode) is up to an order of magnitude larger than that of

drain-source current changes (FET mode). In both cases, the changes are fully reversible in -2-3

mins at 250 C upon removal of the hydrogen from the ambient. These Pt/AlGaN/GaN HEMTs

operated in either a diode mode or in an FET mode show the ability to detect 500 ppm H2 in N2









The origin of the reverse current can be examined by looking at the dependence of current

on perimeter/area ratio. Figure 7-3 shows reverse current at -4 V for diodes deposited at either 77

or 300 K, as a function of either contact diameter (top) or area (bottom).The reverse current was

proportional to both the perimeter and area of the rectifying contact, suggesting that both surface

and bulk contributions are present in this voltage range. Therefore, low temperature deposition

does not seem to reduce Fermi level pinning by surface states in GaAs. Table 7-2 gives a

summary of the electrical properties of the diodes. The main differences are an effective increase

in barrier height of 10-13 % for cryogenic deposition, from a mean value of 0.73 eV for room

temperature deposition to a mean value of 0.82 eV for 77 K deposition. However, it should also

be noted that the ideality factors were larger for the low temperature diodes, perhaps suggesting

the presence of interfacial contamination gettered to the cold surface during the pump-down and

initial stages of deposition. This could be improved by including gettering sources such as

tungsten filaments within the chamber during evaporation. It has been proposed that these low

temperature contacts could be governed by the lack of a thermally diffused interfacial layer,

geometry differences due to reduced metal clustering, or strain imposed on the interface caused

by thermal expansion differences between the metal and substrate, or some combination of the

three. TEM and XPS studies by D. S. Cammack et. al support the assertion of increased barrier

heights due to an interfacial layer differences [207]. Their TEM study of low temperature

deposited contacts showed a more abrupt metal-semiconductor interface, than room temperature

contacts.

All of the diodes showed excellent stability under forward bias aging. Figure 7-4 shows the

time dependence of forward bias at a current of 10 mA for diodes deposited at either 77 or 300

K, with no evidence of drift due to trapping effects. There was also no apparent difference in









1.5 Dissertation Outline

The main objective of this work focuses on the design and fabrication of GaN and ZnO

based sensing devices, and analysis of their sensitivity to hydrogen, heavy metal ions, and

disease biomarkers. Chapter 1 reviews the material properties and recent sensor technologies

combining field effect based devices with surface chemical functionalization. Chapter 2

illustrates details of fabrication and sensitivity measurement to hydrogen using catalytic metal

coated multiple ZnO nanowires. Chapter 3 presents hydrogen detection using AlGaN/GaN

HEMTs and Schottky diodes with the design of differential diodes and the use of TiB2 based

Ohmic contact. This is the extension for the previous work done by Dr. Byoung Sam Kang [63],

toward practical applications. Chapter 4 focuses on the mercury ion detection using AlGaN/GaN

HEMTs through a surface coating of carboxyl groups on gate region. Chapter 5 presents the

detection of disease biomarker, kidney injury molecule-1 (KIM-1), using AlGaN/GaN HEMTs

which are gate-coated with KIM-1 antibody.

The rest of this dissertation also includes zno band edge electroluminescence n -implanted

zno bulk (chapter 6), and investigation of cryogenic schottky contact on gaas regarding thermal

stability and interfacial differences (chapter 7). Chapter 8 briefly summarizes a conclusion of

above works and suggests future studies.









[19] D. C. Look, J. W. Hemsky, and J. R. Sizelove, Phys. Rev. Lett. 82, 2552 (1999).

[20] B. J. Jin, S. H. Bae, S. Y. Lee, and S. Im, Mater. Sci. Eng. B 71, 301 (2000).

[21] D. M. Hofmann, A. Hofstaetter, F. Leiter, H. Zhou, F. Henecker, B. K. Meyer, S. B.
Orlinskii, J. Schmidt, and P. G. Baranov, Phys. Rev. Lett. 88, 045504-1 (2002).

[22] C. G. Van de Walle, Phys. Status Solidi B 229, 221 (2002).

[23] S. F. J. Cox, E. A. Davis, P. J. C. King, J. M. Gil, H. V. Alberto, R. C. Vilao, J. Piroto
Duarte, N. A. De Campos, and R. L. Lichti, J. Phys.: Condens. Matter 13, 9001 (2001).

[24] C. G. Van de Walle, Phys. Rev. Lett. 85, 1012 (2000).

[25] D. M. Bagnall, Y. F. Chen, M. Y. Shen, Z. Zhu, T. Goto, and T. Yao, J. Cryst. Gi 1, n tl 184,
605 (1998).

[26] B. J. Jin, S. Im, and S. Y. Lee, Thin SolidFilms 366, 107 (2000).

[27] E. G. Bylander, J. Appl. Phys. 49, 1188 (1978).

[28] N. Riehl, O. Ortman, Z. Elektrochem. 60, 149 (1952).

[29] F. A. Kroger and H. J. Vink, J. Chem. Phys. 22, 250 (1954).

[30] I. Y. Prosanov and A. A. Politov, Inorg. Mater. 31, 663 (1955).

[31] N. Y. Garces, L. Wang, L. Bai, N. C. Giles, L. E. Halliburton, and G. Cantwell, Appl. Phys.
Lett. 81, 622 (2002).

[32] H.-J. Egelhaaf and D. Oelkrug, J. Cryst. Gi/,n th 161, 190 (1996).

[33] D. Hahn, R. Nink, and Physik, Cond. Mater. 3, 311 (1965).

[34] M. Liu, A. H. Kitai, P. Mascher, and J. Lumin. 54, 35 (1992).

[35] H. Xing, S. Keller, Y.-F. Wu, L. McCarthy, I. P. Smorchkova, D. Buttari, R. Coffie, D. S.
Green, G. Parish, S. Heikman, L. Shen, N. Zhang, J. J. Xu, B. P. Keller, S. P. DenBaars, and
U. K. Mishra, J. Phys.: Condens. Matter 13, 7139 (2001).

[36] R. J. Trew, Proceedings of the IEEE 90, 1032 (2002).

[37]D. Pavlidis, Technical Digest International Electron Devices Meeting, 795 (2004).

[38] E. T. Yu, G. J. Sullivan, P. M. Asbeck, C. D. Wang, D. Qiao, and. S. Lau, Appl. Phys. Lett.
71, 2794 (1997).









Chemical Engineering; Dr. Ivan Kravchenko and Mr. Bill Lewis in UF NanoFab. Without their

professional technical support, I could not have a smooth research life in Gainesville.

Finally, I thank my family and friends in Taiwan for their endless love and support.

Especially my parents, their enormous love and spiritual encouragements helped me to realize

my dream. This dissertation and my Ph.D. degree could not be accomplished without their

invaluable supports.









(200 A)/A1 (800 A)/Pt (400 A)/Au (800 A). The contacts were annealed at 850 C for 45 sec

under a flowing N2 ambient in a Heatpulse 610T system. A 200 A thick circular Pt Schottky

contact was deposited for the gate metal. The final step was deposition of e-beam evaporated

Ti/Au (200 A/2000 A) interconnection contacts. The gate dimension of the device was 1 x50

lm2. The devices were bonded to electrical feed-through and exposed to either pure N2 or 500

ppm H2 in N2 ambient in an environmental chamber in which the gases were introduced through

electronic mass flow controllers.

Figure 3-1 shows the HEMT drain-source current-voltage (IDs-VDs) characteristics (top)

and the transfer characteristics (bottom) at 25 C measured in both the pure N2 and the 500 ppm

H2 in N2 ambients. The increase in drain current at each applied gate voltage current is consistent

with the hydrogen molecules dissociating into atoms through the catalytic action of the Pt gate

contact and diffusing to the Pt/AlGaN interface where it screens some of the piezo-induced

channel charge [118]. Previous measurements have shown an effective decrease in the effective

barrier height of Pt on GaN by 30-60 meV by introduction of hydrogen into a N2 ambient [103].

This data represents the FET mode of operation for hydrogen gas sensing. The transconductance

of the HEMT also increases slightly when measured in the hydrogen-containing ambient due to

the increase in effective channel charge, as shown at the bottom of Figure 3-1.

Figure 3-2 shows the gate I-V characteristics at 0 V IDS measured at 25 C under pure N2

ambient or in a 500 ppm H2 in N2 ambient. Both the forward and reverse currents increase due to

the reduction in barrier height. This data represents the Schottky diode mode of operation for

hydrogen gas sensing.

Figure 3-3 shows the change in drain-source or gate currents as a function of gate voltage

(top) and percentage change in these currents (bottom) for measurement under pure N2 ambient









contact morphology at low resolution. Figure 7-5 shows optical microscope images of Au

contacts deposited at 77 K (left) or 300 K (right), with both exhibiting excellent morphology.

Cryogenic metal deposition also reduces the resistivity, of very thin films, four or five orders of

magnitude compared to contacts deposited at room temperature [208]. By limiting surface

diffusion, atoms deposited on cryogenically cooled substrates tend to stay close to where they

impinge on the substrate surface. Conversely, room temperature deposited atoms are more likely

to re-evaporate or diffuse along the surface. Diffusion can lead to clustering, and eventually

coalescence leaving voids on the substrate [12]. The lack of temperature related diffusion at 77 K

causes the cryogenic metal film to become continuous at lower film thicknesses than films

deposited at room temperature. As metal thickness increases, room temperature films experience

secondary nucleation, and the voids fill. Void filling makes the room temperature films more

continuous, while the larger kinetic energy of room temperature deposited makes these atoms

more likely to form regular crystal lattices than low temperature films. The increased continuity

of room temperature films coupled with better crystal quality lead to a cross over, where the

room temperature films then have lower resistances than comparable thickness cryogenic films.

The lack of temperature related lateral surface diffusion of metal deposited at 77 K implies that

cryogenic metal deposition may also be used to enhance the adhesion of refractory metals (such

as molybdenum). The adhesion of such metals limits their application at short gate lengths.

Successful use of refractory metals at short gate lengths may enhance the reliability of devices at

elevated temperatures.

Figure 7-6 shows XRR of thin (-100 A) Au layers of GaAs for the two different deposition

temperatures, and the associated Au surface roughness and Au/GaAs interfacial roughness

derived from the XRR. The Au surface is clearly smoother for low temperature deposition when




































2008 Hung-Ta Wang











0.85
0.80
0.75
0.70
0.65
0.60
0.55
0.50
0.45
0.40


550 600


Figure 7-9 Schottky barrier height as a function of annealing temperature for diodes deposited at
either 77 or 300 K.


Ak A 4




1000 A Au/GaAs
leakage current @ -4V
c-- 300 K diode
S77 K diode

300 350 400 450 500 5!
Temperature(K)


-1x10-2


<-2x10 2
C -3x10-2
I -4x104
S-5x10-2
-6x10 2


300 350 400 450 500
Temperature(K)


Figure 7-10 Reverse leakage current (@ -4 V) on two different scales as a function of annealing
temperature for diodes deposited at either 77 or 300 K.


0 400 450 500 '
emperature(K)


4)





-c
0)
U,

cc


A-A -- -







1000 A AulGaAs
-o-- 300 K diode
S77 K diode


300 35(
T


0

S-2x103
' -4x10
CD
. -6x103
-x
-8x10'
-1x10-2


A
', ,



1000 A AuiGaAs
leakage current @ -4V
o 300 K diode
--- 77 K diode
0









sensors showed an excellent sensing selectivity of more than 100 of detecting mercury ions over

sodium, magnesium, and lead ions, but not copper.

AlGaN/GaN based HEMTs were used to detect kidney injury molecule-1 (KIM-1), an

important biomarker for early kidney injury detection. The HEMT gate region was coated with

KIM-1 antibodies and the HEMT source-drain current showed a clear dependence on the KIM-1

concentration in phosphate-buffered saline (PBS) solution. The limit of detection (LOD) was

Ing/ml using a 20 gm x50 tm gate sensing area. This approach shows a potential for both

preclinical and clinical disease diagnosis with accurate, rapid, noninvasive, and high throughput

capabilities.

The rest of this dissertation includes ZnO band edge electroluminescence from N -

implanted ZnO bulk, and the investigation of cryogenic gold Schottky contact on GaAs for

enhancing device thermal stability.









CHAPTER 4
MERCURY ION SENSOR USING ALGAN/GAN HIGH ELECTRON MOBILITY
TRANSISTOR

4.1 Background

The toxicity of heavy metal ions, including mercury(II) (Hg2) lead(II) (Pb2+), copper(II)

(Cu2+), and zinc(II) (Zn2+) has long been recognized as a chronic environmental problem [121-

125]. In particular, mercury is released into the environment through a variety of courses

including the combustion of fossil fuels, mining, volcanic emissions and solid waste incineration.

Mercury has attracted a great deal of attention around the world for its impact on wild life

ecology and human health. Certain bacteria convert inorganic mercury Hg2 into neuro-toxic

organic-mercury compounds, which bio-accumulate through the plant, animals, and can food

chain and affect the entire eco-system [126, 127].

It is highly desirable to develop sensitive and selective analytical methods for the

quantitative detection of Hg2+, which are applicable in a wide range of different sites and

environments. Traditionally, there are several methods for heavy metal detection including

spectroscopic (atomic absorption spectroscopy (AAS), Auger electron spectroscopy (AES), or

inductively coupled plasma-Mass Spectrometry (ICP-MS)), or electrochemical (ion selective

electrodes (ISE) or polarography), however, these methods are either expensive or not useful for

detection on-site, where hand-held portable devices could be invaluable for metal detections at

low concentrations [128-130]. To date, a number of selective Hg2 ion sensors have been devised

utilizing redox, chromogenic or fluorogenic changes. Most of these systems display

shortcomings in practical use, such as interference from other metal ions, delayed response to

Hg2+, and/or lack of water solubility [131-134]. Therefore, development of fast response and

inexpensive methods for detection of bioavailable heavy metal concentrations is highly desirable.








contact morphology [120].We expect that the AlGaN/Ti interface is therefore more uniform with

the TiB2 overlayer and this translates to less noise in the current at fixed voltage. Note that this

leads to a lower threshold for hydrogen detection.

In conclusion, Pt-gated AlGaN/GaN HEMT diodes show greatly improved current stability

under field conditions with use of Ti/Al/TiB2/Ti/Au contacts replacing the more conventional

Ti/Al/Pt/Au. Combined with the superior thermal stability of these boride-based contacts, this

metallization system appears attractive for sensors for long-term monitoring applications.



Ohmic Metal:
1. TilAIIPt/Au
2. TilAI/TiB2/Ti/Au


Ti/Au TilAu TilAu
SiNx I Pt SiN,
AI,, :,,Ga,, ;,N

GaN 2D G


Sapphire


Figure 3-9 Schematic of HEMT diode hydrogen sensor using either conventional or TiB2-based
Ohmic contacts.









samples. Similar measurements were performed for the diodes deposited with Ni, Pt, Pd or Ti at

77 K or 300 K. Table 7-3 shows a summary of the extracted barrier heights for all the metals

examined. Only Au, Pd and Ni showed any significant change in barrier height for cryogenic

metal deposition. Note also that we observed cracking and peeling of the metal in the case of Pt,

Pd and Ni, leaving Au as the only metal that showed an enhancement in barrier height and good

adhesion to the GaAs. Optical micrographs of the Ti and Au diodes deposited at different

temperatures are shown in Figure 7-12 as an example of a situation where the metal (Ti in this

case) peels off when deposited at low temperature.

Figure 7-13 shows the XRR spectra from 77 K Au/GaAs diodes as a function of post-

deposition anneal temperature. From this data, it is possible to deconvolute the interfacial

roughness between the Au and GaAs and also the metal roughness at the air interface. To make

the differences in the spectra more obvious, Figure 7-14 shows the XRR spectra from Au/GaAs

diodes deposited at either 77 K or 300 K before (left) or after (right) annealing at 300 C.Two

things are obvious from this data, firstly, there is a clear difference between the samples

deposited at different temperatures and secondly, the effect of the 300 C annealing is to wash

out these differences.

Figure 7-15 shows the interfacial Au/GaAs roughness and metal/air roughness data derived

from the XRR spectra for samples deposited at either 300 K(left) or 77 K(right) as a function of

post-deposition annealing temperature. The interfacial roughness in the room temperature

deposited diodes is basically constant with anneal temperature, whereas that for the cryogenic

diodes is initially smoother but roughens with annealing, reaching a similar value to that in the

room temperature diodes (Figure 7-16). By contrast, the metal/air roughness improves above 200









shown promise for such applications due to a high electron sheet carrier concentration channel

induced by both piezoelectric polarization and spontaneous polarization [7, 8, 135-139, 157].

Acute Kidney Injury (AKI) or Acute Renal Failure (ARF) is one of the most common and

serious medical complications that is closely associated with high mortality [158-160]. Despite

the improvements in dialysis and kidney transplantation techniques over the past two decades,

the high mortality rate has remained. The AKI diagnosis by detecting the urinary biomarker,

kidney injury molecule-1 or KIM-1 (a specific sensitive AKI biomarker) [161], have been

proved with enzyme-linked immunsorbent assay (ELISA) technology [145]. Most state-of-art

testing methods for kidney injury disease biomarkers have limitations due to the laboratory-

oriented nature of the measurements requiring sample transportation, time consuming analysis

and high cost of detection. In this chapter, we report the detection of KIM-1 with KIM-1

antibody functionalized Au-gated GaN/AlGaN HEMTs (Section 5.2). We quantified the

sensitivity of the HEMT sensor and the temporal resolution, along with the limit of detection

(LOD) and selectivity.



5.2 Kidney Injury Molecule-1 Detection Using AIGaN/GaN High Electron Mobility
Transistors

The HEMT structures consisted of a 2 [tm thick undoped GaN buffer and 250 A thick

undoped A10.25Gao.75N cap layer. The epi-layers were grown by metal-organic chemical vapor

deposition on 100 mm (111) Si substrates. Mesa isolation was performed with Inductively

Coupled Plasma (ICP) etching with C12/Ar based discharges at -90 V dc self-bias, ICP power of

300 W at 2 MHz and a process pressure of 5 mTorr. 50 x 50 .im2 Ohmic contacts separated with

gaps of 20 rm consisted of e-beam deposited Ti/Al/Pt/Au patterned by lift-off and annealed at

850 C for 45 sec under flowing N2. 5 nm thin gold film was deposited as gate metal to









thioglycolic acid functionalized Au surface is affected by several factors, like oxygen level, light,

initial packing quality, chain length, and terminal functional group [142, 143]. Our devices has

been stored in nitrogen ambient and repeatedly used over a couple of weeks. The long term

stability of the thioglycolic acid functionalized Au surface is under investigation.

The current sensor operates at 0.5 V of drain voltage and 2 mA of drain current. However,

the operation voltage and device size can be further reduced to minimize the power consumption

to .iW range. The sensor can be integrated with a commercial available hand-held wireless

transmitter to realize a portable, fast response and high sensitivity Hg2+ and Cu2+ ion detector.

In summary, we have demonstrated bared Au-gate and thioglycolic acid functionalized

AlGaN/GaN HEMT sensors to heavy ion detections. The bare Au-gate sensor was sensitive to

Hg2 and thioglycolic acid functionalized sensors could detect both Hg2 and Cu2+ ions. By

fabricating an array of the sensors on a single chip and selectively functionalizing some sensors

with thioglycolic acid, a multi-functional specific detector can be fabricated. Such a sensor array

can be used to detect quantitatively Hg2+ ions in Cu2+ ion solution or Cu2+ ions in Hg2+ ion

solution. Both bare Au-gate and thioglycolic acid functionalized sensor can be repeatedly used

after a simple DI water rinse.









[39] O. Ambacher, J. Smart, J. R. Shealy, N. G. Weimann, K. Chu, M. Murphy, W. J. Schaff, L.
F. Eastman, R. Dimitrov, L. Wittmer, M. Stutzmann, W. Rieger, and J. Hilsenbeck, J. Appl.
Phys. 85, 3222 (1999).

[40] O. Ambacher, B. Foutz, J. Smart, J. R. Shealy, N. G. Weimann, K. Chu, M. Murphy, A. J
Sierakowski, W. J. Schaff, L. F. Eastman, R. Dimitrov, A. Mitchell, and M. Stutzmann, J.
Appl. Phys. 87, 334 (2000).

[41] P. Bergveld, IEEE Trans. Biomed. Eng. 17, 70 (1970).

[42] P. Bergveld, IEEE Trans. Biomed. Eng. 19, 342 (1972).

[43] G. F. Blackburn, Biosensors: Fundamentals and Applications, A. P. F. Turner, G. Karube,
G. S. Wilson, Oxford University Press, Oxford, 481 (1987).

[44] D. G. Hafeman, J. W. Parce, and H. M. McConnell, Science 240, 1182 (1988).

[45] J. Kong, N. R. Franklin, C. Zhou, M. G. Chapline, S. Peng, K. Cho, and H. Dai, Science
287, 622 (2000).

[46] P. Qi, O. Vermesh, M. Grecu, A. Javery, Q. Wang, H. Dai, S. Peng, and K. J. Cho, Nano
Lett. 3, 347 (2003).

[47] A. Star, T. R. Han, V. Joshi, J.-C. P. Gabriel, and G. Gruner, Adv. Mater. 16, 2049 (2004).

[48] I. Sayago, E. Terrado, E. Lafuente, M. C. Horillo, W. K. Maser, A. M. Benito, R. Navarro,
E. P. Urriolabeita, M. T. Martinez, and J. Gutierrez, Syn. Metals 148,15 (2005).

[49] Y. Lu, J. Li, H. T. Ng, C. Binder, C. Partridge, and M. Meyyapan, Chem. Phys. Lett. 391,
344 (2004).

[50] A. Kolmakov, D. O. Klenov, Y. Lilach, S. Stemmer, and M. Moskovits, Nano Lett. 5, 667
(2005).

[51] E. Katz and I. Willner, Chem. Phy. Chem. 5, 1084 (2004).

[52] R. J. Chen, S. Bangsaruntip, K. A. Drouvalakis, N. Wong Shi Kam, M. Shim, Y. Li, W.
Kim, P. J. Utz, and H. Dai, Proc. Natl. Acad. Sci. USA 100, 4984 (2003).

[53] A. Star, J.-C. P. Gabriel, K. Bradley, and G. Griner, Nano Lett. 3, 459-463 (2003).

[54] C. Li, M. Curreli, H. Lin, B. Lei, F. N. Ishikawa, R. Datar, R. J. Cote, M. E. Thompson, C.
Zhou, J. Am. Chem. Soc. 127, 12484 (2005).

[55] B. R. Azamian, J. J. Davis, K. S. Coleman, C. B. Bagshaw, M. L. H. Green, J. Am. Chem.
Soc. 124, 12664 (2002).

[56] J. J. Davis, K. S. Coleman, B. R. Azamian, C. B. Bagshaw, and M. L. H. Green, Chem. Eur.
J. 9, 3732 (2003).









measured by this higher resolution technique. There is also a slight decrease in metal/GaAs

interfacial roughness with cryogenic deposition, suggesting less diffusion of the initially

deposited Au atoms, as discussed above. Details of XRR results are discussed in section 7.4.

In conclusion, the results of this study are summarized as follows:

* The use of low temperature deposition of Au on n-GaAs produces an increase in Schottky
barrier height of 10-13 % relative to conventional room temperature deposition.

* The improved barrier height is accompanied by a smoother Au surface and more abrupt
interface between the Au and the underlying GaAs.

* Additional work is needed to determine the origin of the increased ideality factors in low
temperature diodes.


Table 7-1 Metals, melting temperature, and recrystallization temperature.
Melting Temp. (C) Recrystallization Temp.
0. 4Tmek (oC)
Aluminum 660 264
Gold 1064 425
Indium 156 62
Nickel 1453 581
Palladium 1552 620
Platinum 1769 707
Titanium 1668 667


Table 7-2 Summary of Au/GaAs diode characteristics for deposition of the Au at either 77 K or
300 K
J(A, Diameter barrier
slope J(A, Diaeter R(cm) Js(A/cm2) n e d d%
intercept) (Uim) height
14.21 7.0x10-11 200 0.01 2.23x10-7 1.17 0.73
300 K 14.24 2.8x10-10 400 0.02 2.27x10-7 1.17 0.73
14.23 9.6x10-10 800 0.04 1.91x10-7 1.17 0.73
11.66 1.6x10-12 200 0.01 5.27x10-9 1.43 0.83 0.097 13.3
77 K 11.12 1.0x10-11 400 0.02 8.18x10-9 1.50 0.82 0.086 11.8
11.06 5.1x10-11 800 0.04 1.02x 10-8 1.51 0.81 0.076 10.3









removal of hydrogen from the ambient. The ZnO nanorods can be placed on cheap transparent

substrates such as glass, making them attractive for low-cost sensing applications and operate at

very low power conditions.


0 30 60 90
Time(min)


120 150


Figure 2-6 Time dependence of resistance of either Pd-coated or uncoated multiple ZnO
nanorods as the gas ambient is switched from N2 to various concentrations of H2 in
air (10-500 ppm) as time proceeds. There was no response to 02.









Figure 7-9 shows the Schottky barrier height as a function of annealing temperature for

diodes deposited at either 77 or 300 K. The enhancement in barrier height of -0.09 eV between

the two types of diodes is retained over the entire annealing temperature range investigated.

After annealing at 200 OC, the low temperature diodes exhibit a barrier height of 0.53 eV,

compared to 0.44 eV for the comparable room temperature deposited diode. Note the ideality

factor for the low temperature diodes was always larger than those for the room temperature

devices. For example, the as-deposited diodes showed a value of 1.17 for the room temperature

devices and 1.43 for the cryogenic diodes. This would be consistent with the presence of an

interfacial layer that produces more of an MIS behavior than a true Schottky contact.

In the case of breakdown being initiated in the bulk, the reverse breakdown voltage of a

diode VB can be expressed as [209],

eND W,2
VB eNDW (7-3)
2E

where ND is the doping on the epilayer, WB is the depletion depth at breakdown and is the

dielectric constant of GaAs. Figure 7-10 shows the reverse leakage current (@ -4 V) on two

different scales as a function of annealing temperature for diodes deposited at either 77 or 300 K.

We defined the breakdown voltage as the reverse bias needed to reach a current of 100 tA-cm2.

These values are shown as a function of anneal temperature in Figure 7-11. There is roughly a 50

% increase in VB for the diodes deposited at 77 K as a result of the increased barrier height.

Previous electron microscopy and chemical bonding studies are supportive of interfacial layer

differences being the cause of the differences in barrier height, with a more abrupt metal-

semiconductor interface for low temperature deposited contacts [207].

Au rectifying contacts deposited at 77 K on n-GaAs show an enhancement in barrier height

of -12 % over their value for room temperature deposition (0.73 eV). This enhancement persists












I MgO|
7



violet
C3 5 6- \
_0
o GnN]
0.... o violet,
Z3 e
2N CdO
c 2 red -
I -I NJ .. .

2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0
Lattice constant, a (A)

Figure 1-1 Energy band-gap of GaN and ZnO based compound semiconductors as a function of
lattice constant.



1.3 Two Dimensional Electron Gas of AIGaN/GaN High Electron Mobility Transistor

AlGaN/GaN heterojunction has been shown to form a potential well and a two dimensional

electron gas (2DEG) at the lower heterointerface, as show in Figure 1-2. The application of

AlGaN/GaN 2DEG had accelerated the development of high voltage, high power operated

microwave devices for use in broad band power amplifiers in wireless base station applications

[35, 36, 37]. These devices are so called AlGaN/GaN high electron mobility transistor, or

AlGaN/GaN HEMTs, well known for high electron mobility in the 2DEG channel, highest sheet

carrier concentration among III-V material system, high saturation velocity, high breakdown

voltage, and thermal stability.

When wide band AlxGal-xN and narrow band GaN are brought into contact, thermal

equilibrium align their respective Fermi levels (EF) that both conduction (Ec) and valence (Ev)

band are bent and cause the GaN conduction band at the interface to drop below EF. Free









(a) 40 (b) 440
Au-gate HSCH COOH-gate
3 30 --- Hg2+ 30 -- Hg"
2 --Cu Cu2
S20 (50x10 pm gate) 20 (50 10 pm gate)

-10 10-

0 -* ----- ---- 1 --A r I -------

10-8 10-7 10-6 10 10-8 107 106 105
Ion cone. (Ni) Ton cone. (M)

Figure 4-7 Drain current changes in response to Hg2 and Cu2+ ions as a function of the ion
concentration for (a) the bare Au-gate and (b) the thioglycolic acid functionalized Au-
gate AlGaN/GaN HEMT sensor.


Figure 4-8 Plan view photograph of a multiple cell AlGaN/GaN HEMT sensors.









MIS diodes. Future work on acceptor implantation should focus on achieving p-type

conductivity in the ZnO so that true injection LEDs may be realized.



8.6 Increasing Schottky Barrier Height with Cryogenic Metal Deposition

The use of low temperatures (-77 K) during Au Schottky contact deposition onto n-GaAs

produces an increase in barrier height from 0.73 eV for room temperature diodes to 0.82 eV. The

increase in barrier height translates to a decrease in reverse current density of several orders of

magnitude for Au/GaAs diodes deposited at 77 K. The reverse breakdown voltage of low

temperature deposited diodes was -50 % larger than conventional Au/GaAs diodes. There is no

evidence of drift in the forward current in either type of diode, and the low temperature deposited

samples show smoother Au layers and more abrupt Au/GaAs interfaces as determined by X-Ray

Reflectivity measurements. Both types of diodes show surface and bulk contributions to the

reverse bias current. The ideality factor of the cryogenically processed devices (-1.43) was

higher than for room temperature diodes (-1.17) and it may result from contaminants gettered to

the cold GaAs surface. Not all Schottky metals show this enhancement; for example Pt and Ti do

not show any significant change in barrier height whereas Au, Pd and Ni show increases between

7-18%.

The enhancement of -0.09 eV (a 12 % increase) in Schottky barrier height for Au

deposited at cryogenic temperatures on n-type GaAs relative to conventional deposition at 300 K

is shown to persist for annealing temperatures up to 200 OC. At higher anneal temperatures (300

C), both types of diodes show a severe deterioration in rectifying behavior. We used X-Ray

Reflectivity to show that the main difference between Au deposited at 77 K and room

temperature is a decreased interfacial roughness between the Au and GaAs. As the diodes are

annealed to 300 C both the difference in barrier height and interfacial roughness is lost. This is a









2.3 Hydrogen Sensing Using Pd-Coated ZnO Multiple Nanowires

The device fabrication details are as described in section 2.2. Figure 2-6 shows the time

dependence of resistance of either Pd-coated or uncoated multiple ZnO nanorods as the gas

ambient is switched from N2 to various concentrations of H2 in air (10-500 ppm) as time

proceeds. There are several aspects of the data. First, there is a strong increase (approximately a

factor of 5) in the response of the Pd-coated nanorods to hydrogen relative to the uncoated

devices. The addition of the Pd appears to be effective in catalytic dissociation of the H2 to

atomic hydrogen. Second, there was no response of either type of nanorod to the presence of 02

in the ambient at room temperature. Third, the effective conductivity of the Pd-coated nanorods

is higher due to the presence of the metal. Fourth, the recovery of the initial resistance is rapid

(<20 s) upon removal of the hydrogen from the ambient, while the nanorod resistance is still

changing at least 15 min after the introduction of the hydrogen. The reversible chemisorption of

reactive gases at the surface of metal oxides such as ZnO can produce a large and reversible

variation in the conductance of the material [93]. Fifth, the relative response of Pd-coated

nanorods is a function of H2 concentration in N2. The Pd-coated nanrods detected hydrogen

down to <10 ppm, with relative responses of >2.6 % at 10 ppm and >4.2 % at 500 ppm H2 in N2

after a 10 min exposure, as shown in Figure 2-7. By comparison, the uncoated devices showed

relative resistance changes of -0.25 % for 500 ppm H2 in N2 after a 10 min exposure, and the

results were not consistent for lower concentrations. The gas-sensing mechanisms suggested in

the past include the desorption of adsorbed surface hydrogen and grain boundaries in poly-ZnO

[94], exchange of charges between adsorbed gas species and the ZnO surface leading to changes

in depletion depth [89] and changes in surface or grain boundary conduction by gas

adsorption/desorption [95]. The detection mechanism is still not firmly established in these









potentiometric chemical signal can be measured by monitoring the conductance of the

semiconductor channel that is field-effect sensitively modulated in a FET structure [43, 44]. The

dependence of the conductance on gate voltage hence makes FETs natural candidates for

electrical sensing of any binding of charged species. Thin film FET sensor had, however,

constrained by limited sensitivity until the advent of one dimensional nanotubes, wires, and rods.

Field-effect based semiconductor sensors made with those nanostructures have made a great

impact, due to a dramatically increased surface to volume ratio with nanoscale diameter channel,

that single molecule sensitivity might be possible.

The progress of nanostructure FET sensor recently has been accelerated by the successful

surface functionalizations on semiconductor surface for selectivity detection purpose. For gas

sensing, bare Carbon Nanotubes (CNTs) was reported that nitrogen dioxide (NO2) and ammonia

(NH3) adsorptions result in reverse conductance responses because of the electron

withdraw/donating mechanism[45]; The coating of amine group rich polymer, polyethyleneimine

(PEI), on CNT causes an improved sensitivity and selectivity to NO2 and other acidic gas, while

the sulfonic acid group (-SO3H) rich polymer, Nafion, obtain NH3 selective detection [46, 47];

Catalytic metal, Pt or Pd, coating made CNTs [48, 49] and Tin(IV) Oxide (SnO2) nanowires [50]

highly sensitive to hydrogen. For biosensing in aqueous solution, CNT have successfully

detected specific proteins [51, 52, 53, 54] and glucose [51, 55, 56]; Si nanowires are able to

detect protein [57, 58], deoxyribonucleic acid (DNA) [57, 59], drug [57, 60], and virus [57, 61];

Indium(III) oxide (In203) nanowires are also sensitive to protein [54] and DNA [62] though

appropriate surface functionalizations. Table 1-2 is a summery of diverse chemical

functionalization approaches applied to different nanostructure FET sensor for specific bio-

























(b)
*0
:HS* 0H


rr r 1r1 : KIM-1 antibody
Au
-a P: KIM-1


S: mismatched antigen



Figure 5-1 (a) Plan view photomicrograph of a completed device with a 5 nm Au film on the gate
region. (b) schematic device cross section. The Au-coated gate area was
functionalized with KIM-1 antibody on thioglycolic acid.










> 0
S-1 1000 A Au/GaAs
S voltage @ -100 A
-2 --- 300 K diode
0 A 77 K diode
> -3


CO -5
s0 -5-
S -6 A ...
m .7I- I I I '
300 350 400 450 500 550 600
Temperature(K)

Figure 7-11 Reverse breakdown voltage (@ -100 [tA) as a function of annealing temperature for
diodes deposited at either 77 or 300 K.



7.4 Interfacial Differences in Enhanced Schottky Barrier Height Au/n-GaAs Diodes
Deposited at 77 K

100 A thick Au films deposited at 77 K or 300 K, followed by post-deposition annealing at

temperatures up to 300 C, were used in this X-Ray Reflectivity (XRR) study. We interest in

interfacial differences of 77 K and 330 K deposited Au on n-GaAs substrate before and after

thermal treatment, which correlate with Au/n-GaAs diode characteristics and thermal behaviors

as discussed in section 7.2 and 7.3. In addition, front-side contacts of 1000 A thick Au, Pt, Ni, Pd

or Ni were deposited at 77 K or 300 K onto n-GaAs (n ~1017 cm-3) substrates to examine their

individual barrier height enhancements.

As shown in Figure 7-1 (top), the diodes deposited at low temperature have reverse current

densities approximately two orders of magnitude lower than those deposited at room

temperature. The respective barrier heights extracted from the forward I-V characteristics (Figure

7-1, bottom) were 0.73 eV for the room temperature diodes and 0.82 eV for the low temperature









GaN/AlGaN high electron mobility transistors (HEMTs) have also shown promise for gas

and liquid sensor applications due to primarily two reasons: 1) they consist of a high electron

sheet carrier concentration channel induced by both piezoelectric polarization of the strained

AlGaN layer and the difference in spontaneous polarization between AlGaN and GaN. Unlike

conventional semiconductor field effect transistors, there is no intentional dopant in the

AlGaN/GaN HEMT structure. 2) the electrons in the two-dimensional electron gas (2DEG)

channel are located at the lower interface between the AlGaN layer and GaN layer. The electron

carrier concentration in 2DEG strongly depends on the ambient [7-9, 135-139]. We have recently

exploited these properties to detect a variety of species in gases and liquids using appropriately

functionalized AlGaN/GaN HEMTs [135-139]. For these reasons, nitride HEMTs are versatile

devices that may be used for a variety of sensing applications.

Section 4.2 presents the detection of Hg2 with sensors fabricated with Au-gated and

thioglycolic acid functionalized Au-gated GaN/AlGaN HEMTs. We investigated a wide range of

concentration from 10 iM to 10 nM. The temporal resolution of the device was quantified, along

with limit of detection selectivity over sodium as well as magnesium and precision of

measurements. Section 4.3 illustrates the detection of Hg2+ and Cu2+ ions with sensors fabricated

with Au-gated and thioglycolic acid functionalized Au-gated GaN/AlGaN HEMTs. We

investigated a wide range of concentration from 10 iM to 10 nM. The temporal resolution of the

device was quantified, along with limit of detection and selectivity over sodium, magnesium and

lead ions. The recyclability of the sensors between measurements was also explored.









LIST OF REFERENCES

[1] Sensors: A Global Strategic Business Report, released by Global Industry Analysts, Inc.
http://www.strategyr.com/MCP-2740.asp

[2] S. J. Pearton, D. P. Norton, K. Ip, Y. W. Heo, and T. Steiner, J Vac. Sci. Technol. B 22, 932
(2004).

[3] S. J. Pearton, J. C. Zolper, R. J.Shul, and F. Ren, J. Appl. Phys. 86, 1 (1999).

[4] M. H. Huang, Y. Wu, H. Feick, N. Tran, E. Weber, and P. Yang, Adv. Mater. 13, 113 (2001).

[5] Z. W. Pan, Z. R. Dai, and Z. L. Wang, Science 291, 1947 (2001).

[6] W. Park, G. Yi, M. Kim, and S. Pennycook, Adv. Mater. 15, 526 (2003).

[7] Y. Alifragis, A. Volosirakis, N. A. Chaniotakis, G. Konstantinidis, E. Iliopoulos, and A.
Georgakilas, Phy. Stat. Sol. (a) 204, 2059 (2007).

[8] M. Eickhoff, J. Schalwig, G. Steinhoff, O. Weidemann, L. Gorgens, R. Neuberger, M.
Hermann, B. Baur, G. Miller, O. Ambacher, and M. Stutzmann, Phys. Stat. Sol. (c) 0, 1908
(2003).

[9] O. Ambacher, M. Eickhoff, A. Link, M. Hermann, M. Stutzmann, F. Bernardini,V.
Fiorentini, Y. Smorchkova, J. Speck, U. Mishra, W. Schaff, V. Tilak, and L. F. Eastman,
Phys. Stat. Sol. (c) 0, 1878 (2003).

[10] B. Claflin, D. C. Look, S. J. Park, and G. Cantwell, J. Cryst. GI/,n ii 287, 16 (2006).

[11] S. Limpijumnong, S. B. Zhang, S.-H. Wei, and C. H. Park, Phys. Rev. Lett. 92, 155504
(2004).

[12] H. J. Lee, W.A. Anderson, H. Hardtdegen, and H. Luth, Appl. Phys. Lett. 63, 1939 (1993).

[13] S. A. Clark, S. P. Wilks, A. Kestle, D. I. Westwood, and M. Elliot, Surf Sci. 850, 352-354,
(1996).

[14] D. S. Cammack, S. M. McGregor, J. J. McChesney, S. A. Clark, P. R. Dunstan, S. R.
Burgess, S. P. Wilks, F. Peiro, J. C. Ferrer, A. Cornet, J. R.Morante, A. Kestle, D. I.
Westwood, and M. Elliott, Appl. Surf Sci. 501, 123-124 (1998).

[15] S. L. Rumyantsev, M. S. Shur, and M. E. Levinshtein, Selected Topics in Electronics and
System 33, 1-19 (2004).

[16] D. W. Palmer, http://www.semiconductors.co.uk, 2002. 06.

[17] J.-Y. Duboz, Phy. Stat. Sol. (a) 176, 5 (1999).

[18] A. R. Hutson, Phy. Rev. 108, 222 (1957).









8.4 Disease Biomarker Sensor Using AIGaN/GaN High Electron Mobility Transistor

AlGaN/GaN high electron mobility transistors (HEMTs) were used to detect kidney injury

molecule-1 (KIM-1), an important biomarker for early kidney injury detection. The gate region

consisted of 5 nm gold deposited onto the AlGaN surface. The gold was conjugated to highly

specific KIM-1 antibodies through a self-assembled monolayer of thioglycolic acid. The HEMT

source-drain current showed a clear dependence on the KIM-1 concentration in phosphate

buffered saline solution. The limit of detection was 1 ng/ml using a 20 .m x 50 tm gate sensing

area. This electronic detection of disease biomarker is a significant step towards a compact

sensor chip, which can be integrated with a commercial available hand-held wireless transmitter

to realize a portable, fast and high sensitive device for multiple disease diagnosis. Our approach

shows potential for both preclinical and clinical disease diagnosis with accurate, rapid,

noninvasive, and high throughput capabilities.



8.5 ZnO Based Light Emitting Diode

N+ ion implantation at moderate doses (1013-1014 cm-2) into nominally undoped (n -1017

cm-3) bulk single crystal ZnO substrates followed by annealing in the range 600-950 C was used

to fabricate diodes that show band-edge electroluminescence at 120 K (-390 nm) under forward

bias conditions. The current-voltage (I-V) behaviors of the diodes are characteristics of metal-

insulator-semiconductor (MIS) devices but not p-n junctions, and suggest the implantation

creates a more resistive region in the n-ZnO in which holes are created by impact ionization

during biasing, similar to the case of electroluminescence in ZnO varistors. The series resistance

is only 25 Q due to the use of the conducting ZnO substrate. We demonstrated that band-edge

and yellow EL could be obtained from N -implanted bulk ZnO diodes similar to that observed in









3.3 Robust Detection of Hydrogen Using Differential AIGaN/GaN High Electron Mobility
Sensing Diode

A maskset was designed for fabricating differential diodes to eliminate the temperature

effect on the diode characteristics. Schottky contacts of 100 A Pt for the active diode and Ti (200

A)/Au (1200 A) for the reference diodes were deposited by e-beam evaporation. Details of

device fabrication are as described in Section 3.2. Figure 3-5 shows an optical microscope image

of the completed devices. The devices were bonded to an electrical feed-through and exposed to

a 1 % H2 ambient in an environmental chamber.

Figure 3-6 shows the absolute and differential forward current-voltage (I-V) characteristics

at 25 C of the HEMT active (top) and reference (bottom) diodes, both in air and in a 1 % H2 in

air atmosphere. For the active diode, the current increases upon introduction of the H2, through a

lowering of the effective barrier height. The H2 catalytically decomposes on the Pt metallization

and diffuses rapidly to the interface where it forms a dipole layer [117]. The differential change

in forward current upon introduction of the hydrogen into the ambient is -1-4 mA over the

voltage range examined and peaks at low bias. This is roughly double the detection sensitivity of

comparable GaN Schottky gas sensors tested under the same conditions, confirming that the

HEMT based diode has advantages for applications requiring the ability to detect hydrogen even

at room temperature.

As the detection temperature is increased to 50 C, the differential current response of the

HEMT diode pair was almost constant over a wide range of voltages due to more efficient

cracking of the hydrogen on the metal contact, as shown in Figure 3-7. The maximum

differential current is similar to that at 25 C, but the voltage control to achieve maximum

detection sensitivity for hydrogen is not as important at 50 OC.









4.2 Fast Electrical Detection of Hg(II) Ions with AIGaN/GaN High Electron Mobility
Transistors

The HEMT structures consisted of a 2 |tm thick undoped GaN buffer and 250 A thick

undoped Al0.25Gao.75N cap layer. The epi-layers were grown by metal-organic chemical vapor

deposition on 100 mm (111) Si substrates at Nitronex Corporation. Mesa isolation was

performed with an Inductively Coupled Plasma (ICP) etching with C12/Ar based discharges at -

90 V dc self-bias, ICP power of 300 W at 2 MHz and a process pressure of 5 mTorr. 50x50 im2

Ohmic contacts separated with gaps of 10, 20, and 50 [im consisted of e-beam deposited

Ti/Al/Pt/Au patterned by lift-off and annealed at 850 C, 45 sec under flowing N2 for source and

drain metal contacts and 5-nm thin gold film was deposited as gate metal to functionalize a self-

assembled monolayer of thioglycolic acid. 500 nm-thick polymethyl methacrylate (PMMA) was

used to encapsulate the source/drain regions, with only the gate region open to allow the liquid

solutions to cross the surface by e-beam lithography. The source-drain current-voltage

characteristics were measured at 25 C using an Agilent 4156C parameter analyzer with the Au-

gated region exposed to different concentrations of Hg2+, Mg2+ or Na+ solutions. Ac

measurements were performed to prevent side electrochemical reactions with modulated 500-mV

bias at 11 Hz.

A schematic cross-section of the device with Hg2 ions bound to thioglycolic acid

functionalized on the gold gate region and plan view photomicrograph of a completed device is

shown in Figure 4-1. The thioglycolic acid, HSCH2COOH, is an organic compound and contains

both a thiol mercaptann) and a carboxylic acid functional group. A self assembled monolayer of

thioglycolic acid molecule was adsorbed onto the gold gate due to strong interaction between

gold and the thiol-group. The extra thioglycolic acid molecules were rinsed off with de-ionized

(DI) water. An increase in the hydrophilicity of the treated surface by thioglycolic acid
























000 0.02


5.0x10.8


a TU
4.0x108 O

3.0x108 '
2.OxlO08
1.0X108
2.0x10.8 .

1.0x10'8 .


0.04 0.06 0.08 0.10
Current (A)


Figure 6-4 I-V characteristics and forward bias current dependence of integrated EL intensity
from an MIS diode annealed at 800 OC. The EL intensity was measured by a Si
photodiode.


Figure 6-5 Optical microscope image of the emission from the diode in the dark (top) and photos
of the diode under bias from the probe contact taken both in the light and dark
(bottom).









functionalize a self-assembled monolayer of thioglycolic acid. 500-nm-thick polymethyl

methacrylate (PMMA) was used to encapsulate the source/drain regions, with only the gate

region opened using e-beam lithography. A plan view photomicrograph of a completed device is

shown in Figure 5-1(a).

Before thioglycolic acid coating, the sample was exposed to UV ozone for 5 mins to clean

surface contamination. The thioglycolic acid, HSCH2COOH, is an organic compound and

contains functional groups of a thiol mercaptann) and a carboxylic acid functional group. A self-

assembled monolayer of thioglycolic acid molecule was adsorbed onto the Au-gate due to strong

interaction between gold and the thiol-group. The extra thioglycolic acid molecules were rinsed

off with de-ionized water. An increase in the hydrophilicity of the treated surface by thioglycolic

acid functionalization was confirmed by contact angle measurements which showed a change in

contact angle from 58.40 to 16.20 after the surface treatment. The sample was treated with

monoclonal anti rat KIM-1 antibody in a solution of 10 mM phosphate-buffered saline (PBS)

buffer solution containing 4 mM sodium cyano-borohydride, pH 8.8 at room temperature for 2

hours. This antibody immobilization is based on a strong reaction between carboxyl group on

thioglycolic acid and amine group on KIM-1 antibody. Excess KIM-1 antibodies were washed

off by PBS buffer and the unreacted surface carboxyl groups were passivated by a blocking

solution of 100 mM ethanolamine in 10 mM phosphate buffer pH 8.8. Figure 5-1(b) shows a

schematic device cross section with thioglycolic acid followed by KIM-1 antibody coating. The

source-drain current-voltage characteristics were measured at 25 C using an Agilent 4156C

parameter analyzer with the KIM-1 antibody functionalized Au-gated region exposed to different

concentrations of KIM-1/PBS buffer. AC measurements were performed to prevent side

electrochemical reactions with modulated 500 mV bias at 11 Hz.









lowering of the effective barrier height. The H2 catalytically decomposes on the Pt metallization

and diffuses rapidly to the interface where it forms a dipole layer [117]. The differential change

in forward current upon introduction of the hydrogen into the ambient is ~1 mA over the voltage

range examined. By sharp contrast, the passive diodes with Ti/Au gates showed no difference in

current when measured in H2-containing ambients.

Figure 3-11 shows the time dependence of forward current for the active diodes at a fixed

bias of 1.5 V as the 1 % H2 ambient was switched into the chamber through a mass flow

controller for 200 seconds and then switched back to air. The response of the sensor is rapid (<1

sec based on a series of switching tests). The decay time of the forward current back to its initial

value is a function of the volume of the test chamber and the flow rate of the input gases and is

not limited by the response of the HEMT diode itself. As we have previously noted, the use of

the differential pair geometry removes false alarms due to changes in ambient temperature or

voltage drifts [119].

The TiB2-based Ohmic contact devices showed much more stable forward currents at fixed

bias than their conventional counterparts. Figure 3-12 shows the time dependence of forward

current at 1.5 V gate bias for devices with both types of Ohmic contacts. These tests were carried

out in the field, where temperature and humidity were not controlled. There are several features

of note. First, the current is much higher in the diodes with TiB2-based contacts because of their

lower contact resistance (1.6x10-6 Q'cm2 vs 7.5x10-6 Q'cm2 for the conventional Ti/Al/Pt/Au).

Second, there is much better stability of the devices with TiB2-based contacts. There is much

less temperature dependence to the contact resistance of the boride contacts and this translates to

less variation in gate current as the temperature cycles from day to night. We know from

previous results that the TiB2 is an effective diffusion barrier and prevents degradation of the






























Figure 7-12 Optical micrograph images of Ti deposited at either 77 K (top left) or 300 K (top
right) and Au at 77 K (bottom left) or 300 K (bottom right) on GaAs.


1012


101

108


4-104

102


100
C


1 2 3 4 5 6 7
theta-2theta


Figure 7-13 XRR spectra from 77 K Au/GaAs diodes as a function of post-deposition annealing
temperature.


77 K deposition
r *
100 A Au on GaAs
r 1

r
r
r- <" -JY, 2000C

r f .
... .. as-grown,
i I I I I


)












0.24
0.20
0.16
0.12
0.08
0.04
0.00
0


- Thioglycolic acid gate
(50 x50 pm gate)








. ... I . .. i . .. .


107 10 4
Hg(II) cone.


600 800


105
(M)


Figure 4-3 (a) Time dependent response of the drain current for bare Au-gate AlGAN/GaN
HEMT sensor and thioglycolic acid functionalized Au-gate HEMT sensor. (b) Drain
current of a thioglycolic acid functionalized Au-gate HEMT sensor as a function of
the Hg2 ion concentration.


(a)
0.28


200 400
Time (s)


--Au gate
---- Thioglycolic acid gate
- %- -

10 M 10-6M 105IM
S HgCI HgCI2 HgCI


50x50

(50x50 pm)


(b)








on
<-


40

30


-









LIST OF FIGURES


Figure page

1-1 Energy band-gap of GaN and ZnO based compound semiconductors as a function of
lattice constant. ............................................................................24

1-2 Schematic diagram of normal AlGaN/GaN heterostructure with band diagram in the
equilibrium state. 2DEG is located at the lower AlGaN/GaN interface..........................26

1-3 Polarization induced sheet charge in Ga(Al)-face strained/relaxed AlGaN/GaN
heterostructure ............................................................... .... ..... ......... 27

1-4 Cross section of a p-channel FET under positive VG and negative VG. S, D, and G
represent source, drain and gate electrodes respectively. ...............................................29

2-1 Scanning electron micrograph of ZnO multiple nanorods ............... ...... .............35

2-2 Schematic of contact geometry for multiple nanorod gas sensor (left) and a picture of
packaged sensor (right). .......................... ........................................... .. ..... 36

2-3 Time dependence of relative resistance response of metal coated multiple ZnO
nanorods as the gas ambient is switched from N2 to 500 ppm of H2 in air as time
proceeds. There w as no response to 02........ ....................................... .....................36

2-4 I-V characteristic of Pt-coated nanowires in air and after 15 min in 500 ppm of H2 in
a ir .. .......................................................... ..................................... 3 7

2-5 Time dependence of resistance change of Pt-coated multiple ZnO nanorods as the
gas ambient is switched from N2 to various concentrations of H2 in air (10-500 ppm)
and then back to N2. ..... ....................... ........................ ........... 37

2-6 Time dependence of resistance of either Pd-coated or uncoated multiple ZnO
nanorods as the gas ambient is switched from N2 to various concentrations of H2 in
air (10-500 ppm) as time proceeds. There was no response to 02............... ...............40

2-7 Relative response of Pd-coated nanorods as a function of H2 concentration in N2..........41

2-8 Time dependence of relative resistance of Pd-coated multiple ZnO nanorods as the
gas ambient is switched from N2 to oxygen or various concentrations of H2 in air
(10-500 ppm) and then back to N2. ...................................................... ..................41

2-9 Arrhenius plot of rate of resistance change after exposure to 500 ppm H2 in N2..............42

3-1 IDs-VDs characteristics (top) and transfer characteristics (bottom) of Pt-gated HEMT
measured at 25 C under pure N2 ambient or in a 500 ppm H2 in N2 ambient ................47









7-1 I-V characteristics of Au/GaAs Schottky diodes deposited at either 77 K(A) or 300
K(o) for both 200 [pm dia.(left) and 800 [pm dia.(right) contact. An expanded view of
the forward voltage part of the curves is shown at bottom....................... ................97

7-2 Forward current densities as a function of bias for diodes of different diameter
deposited at either 77 or 300 K ................................................ .............................. 97

7-3 Reverse current at -4 V for diodes deposited at either 77 or 300 K, as a function of
either contact diameter (top) or area (bottom). ...................................... ............... 98

7-4 Time dependence of forward bias at a current of 10 mA for diodes deposited at either
7 7 or 3 0 0 K ............................................................................ 9 8

7-5 Optical microscope images of Au contacts deposited at 77 K (left) or 300 K (right). ......99

7-6 XRR of thin (-90 A) Au layers of GaAs for the two different deposition
temperatures and the associated Au surface roughness and Au/GaAs interfacial
roughness derived from the XRR ............................................ ............................. 99

7-7 I-V characteristics of 400 ptm diameter diodes deposited at either 300 K (left) or 77
K (right), as a function of post-deposition annealing temperature. ............................102

7-8 Forward I-V characteristics of 400 pm diameter diodes deposited at either 300 K
(left) or 77 K (right), as a function of post-deposition annealing temperature..............102

7-9 Schottky barrier height as a function of annealing temperature for diodes deposited at
either 77 or 300 K .........................................................................103

7-10 Reverse leakage current (@ -4 V) on two different scales as a function of annealing
temperature for diodes deposited at either 77 or 300 K................................................ 103

7-11 Reverse breakdown voltage (@ -100 pA) as a function of annealing temperature for
diodes deposited at either 77 or 300 K ..................................... ..................... ............. 104

7-12 Optical micrograph images of Ti deposited at either 77 K (top left) or 300 K (top
right) and Au at 77 K (bottom left) or 300 K (bottom right) on GaAs..........................108

7-13 XRR spectra from 77 K Au/GaAs diodes as a function of post-deposition annealing
tem perature. .............................................................................108

7-14 XRR spectra from Au/GaAs diodes deposited at either 77 K or 300 K before (left) or
after (right) annealing at 300 C ............................................. .............................. 109

7-15 Interfacial Au/GaAs roughness and metal/air roughness data derived from the XRR
spectra for samples deposited at either 300 K (left) or 77 K (right), as a function of
annealing tem perature .................. ................................ ........ .. ........ .... 109









[168] T. Minami, M. Tanigawa, M. Yamanishi, and T. Kawamura, Jpn. J. Appl. Phys. 13, 1475
(1974).

[169] T. Minami, S. Takata, M. Yamanishi, and T. Kawamura, Jpn. J Appl. Phys. 18, 1617
(1979).

[170] A. Shimizu, M. Kanbara, M. Hada, and M. Kasuga, Jpn. J. Appl. Phys. 17, 1435 (1978).

[171] S. Takata, T. Minami, H. Nanto, and T. Kawamura, Phys. Stat. Solidi A 65, K83 (1981).

[172] Y. I. Alivov, J. E. Van Nostrand, D. C. Look, M. V. Chukichev, and B. M. Ataev, Appl.
Phys. Lett. 83, 2943 (2003).

[173] H. Hosono, H. Ohta, K. Hayashi, M. Orita, and M. Hirano, J. Cryst. G/n t1,1 469, 237
(2001).

[174] D. K. Hwang, S.H. Kang, J. H. Lim, E. J. Yang, J. Y. Oh, J. H.Yang, and S. J. Park, Appl.
Phys. Lett. 86, 222101 (2005).

[175] H. S. Yang, S. Y. Han, Y. W. Heo, K. H. Baik, D. P. Norton, S. J. Pearton, F. Ren, A.
Osinsky, J. W. Dong, B. Hertog, A. M. Dabiran, P. P. Chow, L. Chemyak, T. Steiner, C. J.
Kao, and G. C. Chi, Jpn. J. Appl. Phys 44, 7296 (2005).

[176] A. Tsukazaki, A Ohtomo, T. Onuma, M. Ohtani, T. Makino, M. Sumiya, K. Ohtani, S. F.
Chichibu, S. Fuke, Y. Segawa, H. Ohno, H. Koinuma, and M. Kawasaki, Nature Mater. 4,
42 (2005).

[177] A. Tsukazaki, M. Kubota, A. Ohtomo, T. Onuma, K. Ohtani, H. Ohno, S. F. Chichibu, and
M. Kawasaki, Jpn. J. Appl. Phys. 21, L643 (2005).

[178] K. K. Kim, H. S. Kim, D. K. Hwang, J. H. Lim, and S. J. Park, Appl. Phys. Lett. 83, 63
(2003).

[179] J. H. Lim, C. K. Kang, K. K. Kim, I. K. Park, D. K. Hwang, and S. J. Park, Adv. Mater. 18,
2720 (2006).

[180] R. Konenkamp, R. C. Word, and C. Schlegel, Appl. Phys. Lett. 85, 6004 (2006).

[181] Y. Du, S. Han, W. Jin, C. Zhou, and A. F. J. Levi, Appl. Phys. Lett. 83, 996 (2003).

[182] S. O. Kucheyev, J. S. Williams, C. Jagadish, J. Zou, C. Evans, A. J. Nelson, and A. V.
Hamza, Phys. Rev. B 67, 094115 (2003).

[183] S. O. Kucheyev, P. N. K. Deenapanray, C. Jagadish, J. S. Williams, Mitsuaki Yano,
Kazuto Koike, Shigehiko Sasa, Masataka Inoue, and Ken-ichi Ogata, Appl. Phys. Lett. 81,
3350 (2002).

[184] S. O. Kucheyev, C. Jagadish, J. S. Williams, P. N. K. Deenapanray, M. Yano, K. Koike, S.
Sasa, M. Inoue, and K.-I. Ogata, J. Appl. Phys. 93, 2972 (2003).


127















Z 400000 T= 298 K
3 350000
S Un-implanted ZnO
300000 -
-- Implanted ZnO
v 250000
t 200000
150000
(,
C 100000
J 50000


12000-
~1 0 1= 30 mA
C 10000
T= 120 K
-2 8000- T= 298 K

6000

c 4000

.- 2000-
-j
LU 0-

400 450 500 550 600
wavelength (nm)





Figure 6-3 Room temperature PL from ZnO before and after N+ implantation and annealing at
800 oC for 2 mins (top) and EL from MIS diode at room temperature and 120 K
(bottom).









devices and needs further study. It should be remembered that hydrogen introduces a shallow

donor state in ZnO and this change in near-surface conductivity may also play a role.

Figure 2-8 shows the time dependence of relative resistance change of Pd-coated multiple

ZnO nanorods as the gas ambient is switched from vacuum to N2, oxygen or various

concentrations of H2 in air (10-500 ppm) and then back to air. These data confirm the absence of

sensitivity to 02. The resistance change during the exposure to hydrogen was slower in the

beginning and the rate resistance change reached maximum at 1.5 min of the exposure time. This

could be due to some of the Pd becoming covered with native oxide, which is removed by

exposure to hydrogen. Since the available surface Pd for catalytic chemical absorption of

hydrogen increased after the removal of oxide, the rate of resistance change increased. However,

the Pd surface gradually saturated with the hydrogen and the rate of resistance change decreased.

When the gas ambient switched from hydrogen to air, the oxygen reacted with hydrogen right

away, with the resistance of the nanorods changed back to the original value instantly. Moreover,

the data were recorded at a power level of- 0.4 mW, which is low even in comparison with

CNTs [48, 49]. This is attractive for long-term hydrogen sensing applications.

Figure 2-9 shows the Arrhenius plot of nanorod resistance change rate. The rate of

resistance change for the nanorods exposed to the 500 ppm H2 in N2 was measured at different

temperatures. An activation energy of 12 KJ/mole was calculated from the slope of the Arrhenius

plot. This value is larger than that of a typical diffusion process. Therefore, the dominant

mechanism for this sensing process is more likely to be the chemisorption of hydrogen on the Pd

surface.

In conclusion, Pd-coated ZnO nanorods appear well suited to detection of ppm

concentrations of hydrogen at room temperature. The recovery characteristics are fast upon









TABLE OF CONTENTS

page

A CK N O W LED G M EN T S ................................................................. ........... ............. .....

L IST O F T A B L E S ............. ..... ............ ................................................................... . 8

LIST OF FIGURES .................................. .. ..... ..... ................. .9

LIST OF ABBREVIATION S ........................... ........................... .... .................. 14

A B S T R A C T ............ ................... ............................................................ 18

CHAPTER

1 INTRODUCTION ............... ................. ........... .............................. 20

1.1 M otiv action ........................................ ............. 20
1.2 Properties of G aN and ZnO ................................................ ........ ...................... 22
1.3 Two Dimensional Electron Gas of AlGaN/GaN High Electron Mobility Transistor ......24
1.4 Field-Effect Based Semiconductor Sensors................................. ........................ 27
1.5 Dissertation Outline ..................................................... .......... ............... 30

2 HYDROGEN SENSOR USING MULTIPLE ZNO NANORODS .................................31

2 .1 B a c k g ro u n d ................................................ .... .......................... .. .... ...................... 3 1
2.2 Detection of Hydrogen at Room Temperature with Catalyst-Coated ZnO Multiple
N anow ires .............................................. ....... ...... ............ 32
2.3 Hydrogen Sensing Using Pd-Coated ZnO Multiple Nanowires.................. ............38

3 HYDROGEN SENSOR USING ALGAN/GAN SCHOTTKY DIODE AND HIGH
ELECTRON M OBILITY TRANSISTOR ........................................ ......................... 43

3 .1 B ackgrou n d ................. ... ..................... ...... ........... ... .. ............... ............. 4 3
3.2 Comparison of Gate and Drain Current Detection of Hydrogen at Room
Temperature with AlGaN/GaN High Electron Mobility Transistor..............................44
3.3 Robust Detection of Hydrogen Using Differential AlGaN/GaN High Electron
M ability Sensing D iode ................................... .... ... .. ..... ............. ............ .....51
3.4 Stable Hydrogen Sensors from AlGaN/GaN Heterostructure Diodes with TiB2-
Based Ohm ic Contacts ......... .... .............. .............. .. ...... 55

4 MERCURY ION SENSOR USING ALGAN/GAN HIGH ELECTRON MOBILITY
T R A N S IS T O R .................................................................................................................. 6 0

4 .1 B background ............... .... ................................................................... .........60
4.2 Fast Electrical Detection of Hg(II) Ions with AlGaN/GaN High Electron Mobility
T ran sisto rs ...................................... .................................... ............... 6 2









[57] F. Patolsky and C. M. Lieber, Materialstoday, April, 20 (2005).

[58] Y. Cui, Q. Wei, H. Park, and C. M. Lieber, Science 293, 1289 (2001).

[59] J. Hahm and C. M. Lieber, Nano Lett. 4, 51 (2004).

[60] W. Wang, C. Chen, K. Lin, Y. Fang, and C. M. Lieber, Proc. Natl. Acad. Sci. USA 102,
3208 (2005).

[61] F. Patolsky, G. Zheng, O. Hayden, M. Lakadamyali, and C. M. Lieber, Proc. Natl. Acad.
Sci. USA 101, 14017 (2004).

[62] M. Curreli, C. Li, Y. Sun, B. Lei, M. A. Gundersen, M. E. Thompson, and C. Zhou, J. Am.
Chem. Soc. 127, 6922 (2005).

[63] Byoung Sam Kang, University of Florida dissertation, 2005.

[64] See, for example, the review by G. W. Hunter, P. G. Neudeck, R. S. Okojie, G. M. Beheim,
V. Thomas, L. Chen, D. Lukco, C. C. Liu, B. Ward, and D. Makel, Proc.-Electrochem. Soc.
114, 93 (2002).

[65] Q. Wan, Q. H. Li, Y. J. Chen, T. H. Wang, X. L. He, X. G. Gao, and J. P. Li, Appl. Phys.
Lett. 84, 3654 (2004).

[66] K. Keem, H. Kim, G. T. Kim, J. S. Lee, B. Min, K. Cho, M. Y. Sung, and S. Kim, Appl.
Phys. Lett. 84, 4376 (2004).

[67] M. H. Huang, S. Mao, H. Feick, H. Yan, Y. Wu, H. Kind, E. Weber, R. Russo, and P. Yang,
Science 292, 1897 (2001).

[68] Q. H. Li, Q. Wan, Y. X. Liang, and T. H. Wang, Appl. Phys. Lett. 84, 4556 (2004).

[69] B. H. Kind, H. Yan, B. Messer, M. Law, and P. Yang, Adv. Mater. (Weinheim, Ger.) 14,
158(2002).

[70] C. H. Liu, W. C. Liu, F. C. K. Au, J. X. Ding, C. S. Lee, and S. T. Lee, Appl. Phys. Lett. 83,
3168 (2003).

[71] W. I. Park, G. C. Yi, J. W. Kim, and S. M. Park, Appl. Phys. Lett. 82, 4358 (2003).

[72] J. J. Wu and S. C. Liu, Adv. Mater. (Weinheim, Ger.) 14, 215 (2002).

[73] H. T. Ng, J. Li, M. K. Smith, P. Nguygen, A. Cassell, J. Han, and M. Meyyappan, Science
300, 1249 (2003).

[74] J. Q. Hu and Y. Bando, Appl. Phys. Lett. 82, 14019 (2003).

[75] W. I. Park, G. C. Yi, M. Y. Kim, and S. J. Pennycook, Adv. Mater. (Weinheim, Ger.) 15,
526 (2003).









LIST OF TABLES


Table page

1-1 Physical properties of GaN and ZnO................................... ................................. 23

1-2 Summery of diverse biosensors with CNT, Si nanowire, and In203 nanowire FETs........29

7-1 Metals, melting temperature, and recrystallization temperature .............. ...............96

7-2 Summary of Au/GaAs diode characteristics for deposition of the Au at either 77 K or
3 0 0 K .......................................................... .................................... 9 6

7-3 Barrier height enhancement observed for different metals on n-GaAs .........................107









7.3 Thermal Stability of Au Schottky Diodes on GaAs Deposited at Either 77 K or 300 K

The devices used for this study are those reported in section 7.2. The I-V's were obtained

as a function of post-deposition annealing temperature (up to 300 OC, 30 minutes anneals under

air ambient). The Schottky barrier height, Ob, and diode ideality factor, n, were extracted from

the relation for the thermionic emission (eq. 7-1) as well. Figure 7-7 shows I-V characteristics of

400 [tm diameter diodes deposited at either 300 K (left) or 77 K (right), as a function of post-

deposition annealing temperature. The diodes deposited at low temperature have reverse current

densities approximately two orders of magnitude lower than those deposited at room

temperature. The respective barrier heights extracted from the forward I-V characteristics were

0.73 eV for the room temperature diodes and 0.82 eV for the low temperature samples. Both

types of diodes show increases in reverse current density after annealing at 200 OC or higher,

with very significant increases after 300 C anneals. The forward I-V characteristics as a

function of annealing temperature are shown in more detail in Figure 7-8. Note that while both

diode types show a deterioration in rectifying behavior, the samples with low temperature

deposited contacts still retain lower current densities at all annealing conditions. The forward

turn-on voltage for rectifiers is given by [206],

nkT J
VF n( iF )+ n +Ro JF (7-2)
e A T2

where k is Boltzmann's constant, Tis the absolute temperature, e is the electronic charge, A** is

Richardson's constant and RON the on-state resistance. Thus the observed increase in turn-on

voltage is also consistent as resulting from a larger barrier height. The on/off ratio of the diodes

was ~104 at 1 V/-4 V for the low temperature deposited devices and ~102 for the room

temperature diodes.









Zn: Zinc

ZnO: Zinc Oxide









10-2 ,
102 Vds= 0V

10-4




iI
10-8 "

-4 -2 0 2 4
v (V)


Figure 3-2 Gate I-V characteristics at 0 V IDS measured at 25 C under pure N2 ambient or in a
500 ppm H2 in N2 ambient.









gated region exposed to different concentrations of Hg2+, C2+, Pb2+, Mg2+ or Na+ solutions. AC

measurements were performed to prevent side electrochemical reactions with modulated 500 mV

bias at 11 Hz.

Figure 4-5 shows the change in drain current of a bare Au-gated AlGaN/GaN HEMT

sensor and a thioglycolic acid functionalized AlGaN/GaN HEMT sensor exposed to 10-5 M Hg2+

ion solutions as compared to exposed to DI water. The drain current of both sensors decreased

after exposure to Hg2 ion solutions. The drain current reduction of the thioglycolic acid

functinalized AlGaN/GaN HEMT sensors was almost 80 % more than that of the bare Au-gate

sensor. The mechanism of the drain current reduction for bare Au-gate and thioglycolic acid

functionalized AlGaN/GaN HEMT sensors is probably quite different. For the thioglycolic acid

fictionalized AlGaN/GaN HEMT, the thioglycolic acid molecules on the Au surface align

vertically with carboxylic acid functional group toward the solution [140]. The carboxylic acid

functional group of the adjacent thioglycolic acid molecules probably forms chelates

(R-COO-(Hg2+)-OOC-R) with the Hg2+ ions. If the chelates are indeed forming, one would

expect the charges of trapped Hg2+ ion in the R-COO-(Hg2+ -OOC-R to change the polarity of the

thioglycolic acid molecules. This is probably why the drain current changes in response to

mercury ions. A similar type of surface functionalization was used by Chang et. al. and the

detection performed with gold-nanoparticle-based fluorescence[141], but the detection time is

longer than the nitride HEMT based sensor. Because Hg2 ions were used in our experiments, we

do not expect an Au-mercury amalgam to form on the bare Au-surface. The detailed mechanism

for mercury ion induced reduction in drain current of the Au-gate device is not clear and

currently under further investigation.









ACKNOWLEDGMENTS


I deeply appreciate my dissertation advisor, Professor Fan Ren, for his mentoring,

guidance, and encouragement throughout my whole Ph.D. studies. He has created a world-class

research environment in his group, giving me the opportunity to conduct many intriguing and

important projects, to collaborate with many experts of different background levels, to broaden

my fields and build my multidisciplinary knowledge, to operate many state-of-the-art

instruments and utilize many research resources, and to accomplish so many research results that

I never thought I could achieve within a few years. I respect his professional attitudes to research

and development, and his tremendous experience and acknowledge that helped me not only to

complete my degree but also to develop myself as a top-notch researcher.

I am also truly indebted to my committee members, Professor Stephen Pearton, Professor

David Norton, and Professor Yiider Tseng, for their significant contributions to my research and

my dissertation. Besides, I sincerely appreciate Professor Jenshan Lin and Professor Tanmay

Lele for their important advice on to my research.

I thank my group members: Dr. Byoung Sam Kang, Dr. Soohwan Jang, Dr. Jau-Jiun Chen,

Travis Anderson, Yu-Lin Wang, Ke-Hung Chen, and Barrett Hicks. I especially thank Byoung

Sam for his great help with my research life in this group. I also thank other colleagues, Dr.

Brent Gila, Dr. Luc Stafford, Dr. Mark Hlad, Dr. Chih-Yang Chang, Dr. Rohit Khanna, Lars

Voss, Jonathan Wright, Andrew Herrero, Wantae Lim, Thomas Chancellor, Jr., Changzhi Li,

Zhen-Ning Low, and Sheng-Chung Hung. I will always relish my collaborations with these

experts as well as the friendships built during these years. Also, I have to extend my

acknowledgements to Mr. Dennis Vince, Mr. James Hinnant, and Dr. Santiago Alves Tavares in









exchange of charges between adsorbed gas species and the ZnO surface leading to changes in

depletion depth [94] and changes in surface or grain-boundary conduction by gas

adsorption/desorption [95]. Finally, Figure 2-3 shows an incubation time for response of the

sensors to hydrogen. This could be due to some of the Pd (or Pt) becoming covered with native

oxide, which is removed by exposure to hydrogen. A potential solution is to use a bi-layer

deposition of the Pt/Pd followed by a very thin Au layer to protect the Pd from oxidation.

However, this adds to the complexity and cost of the process and, since the Pd is not a

continuous film, the optimum coverage of Au would need to be determined. We should also

point out that the I-V characteristics were the same when measured in vacuum as in air,

indicating that the sensors are not sensitive to humidity.

The power requirements for the sensors were very low. Figure 2-4 shows the I-V

characteristics measured at 25 C in both a pure N2 ambient and after 15 min in a 500 ppm H2 in

N2 ambient. Under these conditions, the resistance response is 8 % and is achieved for a power

requirement of only 0.4 mW. This compares well with competing nanotechnologies for hydrogen

detection such as Pd-loaded carbon nanotubes [48, 49]. Moreover, the 8 % response compares

very well to the existing SiC-based sensors, which operate at temperatures >100 C through an

on-chip heater in order to enhance the hydrogen dissociation efficiency [64]. Figure 2-5 shows

the sensors can detect 100 ppm H2.

In conclusion, Pt-coated ZnO nanorods appear well suited to detection of ppm

concentrations of hydrogen at room temperature. The recovery characteristics are fast upon

removal of hydrogen from the ambient. The ZnO nanorods can be placed on cheap transparent

substrates such as glass, making them attractive for low-cost sensing applications, and can

operate at very low power conditions. Of course, there are many issues still to be addressed, in









CHAPTER 3
HYDROGEN SENSOR USING ALGAN/GAN SCHOTTKY DIODE AND HIGH ELECTRON
MOBILITY TRANSISTOR

3.1 Background

There is great current interest in detection of hydrogen sensors for use in hydrogen-fueled

automobiles and with proton-exchange membrane (PEM) and solid oxide fuel cells for space

craft and other long-term sensing applications. These sensors are required to selectively detect

hydrogen near room temperature with minimal power consumption and weight and with a low

rate of false alarms. Due to their low intrinsic carrier concentrations, wide bandgap

semiconductor sensors based on GaN or SiC can be operated at lower current levels than

conventional Si-based devices and offer the capability of detection to -600 OC [8, 96-117]. The

ability of electronic devices fabricated in these materials to function in high temperature, high

power and high flux/energy radiation conditions enable performance enhancements in a wide

variety of spacecraft, satellite, homeland defense, mining, automobile, nuclear power, and radar

applications.

AlGaN/GaN high electron mobility transistors (HEMTs) show promising performance for

use in broad-band power amplifiers in base station applications due to the high sheet carrier

concentration, electron mobility in the two dimensional electron gas (2DEG) channel and high

saturation velocity. The high electron sheet carrier concentration of nitride HEMTs is induced by

piezoelectric polarization of the strained AlGaN layer and spontaneous polarization is very large

in wurtzite III-nitrides. This provides an increased sensitivity relative to simple Schottky diodes

fabricated on GaN layers [8, 99-117]. An additional attractive attribute of AlGaN/ GaN diodes is

the fact that gas sensors based on this material could be integrated with high-temperature

electronic devices on the same chip. The advantages of GaN over SiC for sensing include the

presence of the polarization-induced charge, the availability of a heterostructure and the more

























1 2 3
theta-2theta


4 5


1 2 3 4 5
theta-2theta


Figure 7-14 XRR spectra from Au/GaAs diodes deposited at either 77 K or 300 K before (left) or
after (right) annealing at 300 C.


300K deposition
A metal/air
SGaAs/metal


--
00 o-- 0 0



300 400 500 60
Temperature (K)


77 K deposition A
A metal/air
o GaAs/metal A
/
/
/
A meau A re(K)


300 400 500 60
Temperature (K)


Figure 7-15 Interfacial Au/GaAs roughness and metal/air roughness data derived from the XRR
spectra for samples deposited at either 300 K (left) or 77 K (right), as a function of
annealing temperature.


As Grown
S77 K
300 K
11


-L'^ *fc .
4-M


30

-25

cn 20
U,
| 15
0-

0
Z 5

0











1.0


Pt coated ZnO nanowires
0.5 -o- 15 min exposed to
< 500ppm H2
E -air
0.0


0-0.5



-1.0
-0.4 -0.2 0.0 0.2 0.4
Voltage(V)


Figure 2-4 I-V characteristic of Pt-coated nanowires in air and after 15 min in 500 ppm of H2 in


0 5 10 15 20
Time(min)


25 30


Figure 2-5 Time dependence of resistance change of Pt-coated multiple ZnO nanorods as the gas
ambient is switched from N2 to various concentrations of H2 in air (10-500 ppm) and
then back to N2.









The results of our study may be summarized as follows:


* The deposition of Au at low temperatures produces an increase in barrier height from 0.73
eV to 0.82 eV. Ni shows an even larger enhancement, but the metal contact in that case
shows cracking and peeling.

* No significant enhancement in barrier height was observed for low temperature deposited
Pt and Ti.

* The improved barrier height in the case of Au is accompanied by a sharper metal/GaAs
interface. As the samples are annealed to 300 C, this interface roughens to the same value
as in room temperature deposited diodes and the enhancement in barrier height disappears.


Table 7-3 Barrier height enhancement observed for different metals on n-GaAs
Metal Cracks at 77K 0(300K)(eV) 0(77K)(eV) d (%)
Au No 0.73 0.82 12
Pt Yes 0.79 0.79 0
Ti No 0.69 0.71 3
Pd Yes 0.76 0.81 7
Ni Yes 0.67 0.79 18










detections using selective bio-interaction, e.g. antigen-antibody reaction and DNA-DNA


interaction.


Table 1-2 Summery of diverse biosensors with CNT, Si nanowire, and In203 nanowire FETs.
1-D 2nd 3rd 4th 5th
1st layer sensing mechanism
nanomaterial layer layer layer layer
CNT hydrophobic-hydrophobic
(ref. 51, 52) interaction
CNT N -aliphatic ::'-bootin protein biotin-protein interaction
(ref. 51, 52) molecules
CNT N -aliphatic
(ref. 51, 52) mlecules ',-protein protein antobody-antigen interaction
(ref. 51, 52) molecules
CNT N -@-
NT(ref. ) pol) *: otin protein biotin-protein interaction
(ref. 51, 53) polymer(PEI)
CNT -phosphatic
CNref.T @-phosphic --ani:oo, antigen antobody-antigen interaction
(ref. 54) molecules
CNT N -glucose guce redox process
(ref. 51, 55, 56) oxidase glucose rx
Si-NW @-amine
re. 5 5 ) g p -bjiotin protein biotin-protein interaction
Si-NW -aine @-protein @-biotin @-DNA c-Dr A DNA-DNA interaction
(ref. 57, 59) group
Si-NW @-amine
@-protein drug inhibitor reaction
(ref. 57, 60) group
Si-NW @-amine
Si-NW -amine @-antibody ..rus antibody-virus interaction
(ref. 57, 61) group
In203-NW @-phosphatic @-DNA c-Di JA DNA-DNA interaction
(ref. 54) molecules
fn203-NW -phosphatic @-antibody antigen antobody-antigen interaction
(ref. 62) molecules
Note: @ represents covalent functionalization; N representscovalent functionalization; highlighted final layer is the target
analyte.


VG>O VG



p-type p-type
Space charge
region



Figure 1-4 Cross section of a p-channel FET under positive VG and negative VG. S, D, and G
represent source, drain and gate electrodes respectively.









to the increased junction area, reduced temperature sensitivity, enhanced polarization

dependence of reflectivity and improved carrier confinement in 1-D nanostructures [180-181].

In this chapter, we demonstrate that N+ implantation into bulk single-crystal ZnO

substrates can be used to achieve bandedge electroluminescence (EL) in simple diode structures.

The mechanism for bandedge EL is most likely hole creation by impact ionization in the MIS

structure.



6.2 Band-Edge Electroluminescence from N+-Implanted Bulk ZnO

There have also been recent breakthroughs in the understanding of damage creation and

annealing in ion implanted ZnO [182-189] and reports of p-type doping using As implantation at

low temperatures, followed by multiple step annealing [190]. Ion implantation is an attractive

process for low-cost, high throughput device manufacturing and in this section we show that N

implantation into bulk single-crystal ZnO substrates can be used to achieve bandedge

electroluminescence (EL) in simple diode structures. The mechanism for bandedge EL is most

likely hole creation by impact ionization in the MIS structure.

The ZnO samples were (0001) undoped grade I quality bulk, single- crystal ZnO crystals

from Cermet. They were epiready with one-side-Zn-face-polished by the manufacturer. The

room temperature electron concentration and mobility established by van der Pauw

measurements were 1017 cm-3 and 190 cm2/Vs, respectively. Ion implantation was performed at

300K with N ions of energy 5 keV (dose of 1.5 x1013 cm-2), 20 keV (dose of 5 x 1013 cm-2) plus

50 keV (dose of 1.3 x1014 cm-2) and 130 keV (dose of 3.5 x 1014 cm-2), followed by rapid thermal

annealing (RTA)for 2 mins under a flowing 02 ambient. We also annealed some of the samples

in either a conventional tube furnace or a pulsed laser deposition chamber under 02 ambients for

45 mins, with the same basic trends observed in diode behavior as for the RTA processed









C in the room temperature diodes, but worsens with annealing in the 77 K diodes. Note that the

metal is initially smoother on the low temperature diodes (Figure 7-16).

The correlation of electrical and x-ray data clearly shows that the enhanced barrier height

is associated with a smoother Au/GaAs interface and that post-deposition annealing of the

cryogenic diodes roughens this interface to a value similar to that of diodes deposited at room

temperature while at the same time reducing the Schottky barrier height back to the values

obtained on the room temperature diodes annealed under the same conditions. Previous electron

microscopy and chemical bonding studies are also supportive of interfacial layer differences

being the cause of the differences in barrier height, with a more abrupt metal-semiconductor

interface for low temperature deposited contacts [207].There are still many issues to be resolved,

for example, why only certain metals exhibit the increased barrier height when deposited at low

temperatures. This may be related to the crystal structure and grain size of the particular metal

layers on GaAs obtained at different deposition temperatures. The melting temperature of Au

(1064 C) is well below that of the other metals studied here(1453 C for Ni, 1552 C for Pd,

1769 C for Pt, and 1668 C for Ti) and thus the recrystallization (onset of grain growth)

temperature will also be lower, since this is typically about 40 % of the melting temperature. It is

desirable that that interface diffusion and grain growth be minimized during encapsulation of the

GaAs device by a SiNx dielectric for packaging. This dielectric is deposited by PECVD and the

temperature that the substrate is exposed to is 200-300 C. For significant interfacial diffusion to

take place, the constituents of either the semiconductor or the metal contact must become mobile

and form a solid solution. The contact metals studied here should have little diffusion below

0.4Tmelt.










0.020 ---
on a 50C
ni .- non-active
0.018 -




0.014

0 0.012 -1% H test in air non-activ
& active
Biased at 2 5V
0.010 -
0 300 600 900 1200
Time (s)

Figure 3-8 Time dependent test of differential HEMT diodes at 25 and 50 C.



3.4 Stable Hydrogen Sensors from AIGaN/GaN Heterostructure Diodes with TiB2-Based
Ohmic Contacts

We compared two types of Ohmic contacts formed by sputter deposition and lift-off, i.e. Ti

(200 A) /Al (1000 A) /Pt (600 A) /Au (800 A) or Ti (200 A) /Al (1000 A) /TiB2 (400 A) /Ti (200

A) /Au (800 A). All of the metals were deposited by Ar plasma-assisted rf sputtering at pressures

of 15-40 mTorr and rf (13.56 MHz) powers of 200-250 W. Other details of device fabrication are

as described in Section 3.2. Figure 3-9 shows a schematic of a completed device. The device was

bonded to an electrical feed-through and exposed to a 1 % H2 ambient in an environmental

chamber.

Figure 3-10 shows the linear (top) and log scale (bottom) forward current-voltage (I-V)

characteristics at 25 C of the HEMT active (ie. Pt-gate) diode, both in air and in a 1 % H2 in air

atmosphere. For these diodes, the current increases upon introduction of the H2, through a










0 08


0.06


S0.04
i-
i-

O 0.02


0 00



0.08


-0.06


0.04


S0.02
o 0.02


(b)
Reference diodl
@ 50o C
* air
-o- 1% H2


H- -I,


* ~P1
3-L


/


P


2 4 6 8 10


Voltage (V)


Figure 3-7 Absolute (a) and differential (b) current of HEMT diode measured at 50 C.


0.006

0.004 N
_X


L
0


Voltage (V)


0.00
0


f nfl^


0.008

0.006


0.004 -
I
0.002


-.lr .- n I. 2 1 .. ... ....... ULJ LL


/IL #W>










e de 2
njx) u- () ) AIx)( ) (1-4)


where o(x) is piezoelectric polarization, e(x) is the dielectric constant, dd is the AlGaN layer

thickness, eb is the Schottky barrier of the gate contact on AlGaN, EF is the Fermi level and

AEc is the conduction band discontinuity between AlGaN and GaN. Because the spontaneous

and piezoelectric polarization increase with Al concentration of AlGaN layer, the typical sheet

carrier concentration could reach 1.6 x 1013 cm-2 for x = 0.3, in excess of other available III-V

material systems.


AIGaN

2DEG

GaN


mmm
< -n n

Figure 1-2 Schematic diagram of normal AlGaN/GaN heterostructure with band diagram in the
equilibrium state. 2DEG is located at the lower AlGaN/GaN interface.









CHAPTER 1
INTRODUCTION

1.1 Motivation

According to 2007 published "Sensors: A Global Strategic Business Report", the global

sensor market grows averagely at an annual rate of 4.5 % between 2000 and 2008 and is

expected to reach US $61.4 billion by 2010 [1]. The US $11.5 billion worth of chemical sensor

market represents the largest segment of this global sensor market. This includes chemical

detection in gas, chemical detection in liquid, flue gas and fire detection, liquid quality sensor,

and biosensor. Semiconductor based sensor fabricated using the mature micro-fabrication

techniques and/or novel nanotechnology is one of the major contestants in this market. Silicon

based devices remain dominating due to their low cost, reproducible and controllable electronic

behaviors, and abundant data of chemical treatment on silicon oxide or glass. However, they are

unable to be operated at harsh environment, for instance, high temperature, high pressure, and

corrosive ambient, so the application area is still limited. The two wide band-gap compound

semiconductors, Gallium Nitrite (GaN) and Zinc Oxide (ZnO), are very good alternative options

to replace silicon because of many advantages, for example, highly chemical resistance, potential

for high power operation, and blue and ultraviolet optoelectronic behaviors [2, 3]. A variety of

sensors have been reported using GaN or ZnO materials, such as nanorod/wire, homo-structured

thin film, and hetero-structured thin film based devices (diodes, transistors, surface acoustic

wave devices, or electrochemical electrodes).

ZnO and GaN nanorods/wires are extremely attractive for sensing applications. In nature,

1-D nanostructures could dramatically enhance the sensitivity due to their high surface to volume

ratio, Debye length comparable diameter, better crystallinity than 2-D thin film, and quantum

effect [4, 5, 6]. In addition, for most of these applications, the nanorod/wire sensors have very


































To my parents, my family.










0.15
0.1 Na ion test
0.12
09 Mg ion test
0.09
0.06 -
107 M 106 M 105M V ~
S003 NaCI NaC1 NaCI
0.00 HO 10" M1
S0.15" 10 M 10 M 10-M HgCI,
0.12 lgCI, NlgCI, MAnCI.


0.069
Thioglycolic acid gate

0.00
0 200 400 600 800 1000
Time (s)


Figure 4-4 Time dependent response of the drain current for detecting Na+, Mg2+ or Hg2+ with a
thioglycolic acid functionalized Au-gate HEMT sensor.



4.3 Selective Detection of Hg(II) Ions from Cu(II) and Pb(II) Using AIGaN/GaN High
Electron Mobility Transistors

The HEMT structures consisted of a 2 ptm thick undoped GaN buffer and 250 A thick

undoped Al0.25Gao.75N cap layer. The epi-layers were grown by molecular beam epitaxy system

on 2" sapphire substrates at SVT Associates. Details of the device fabrication are as described in

Section 4.2. 5-nm thin gold film was deposited as gate metal for two set of samples. One was for

the bare Au-gate sensor and the other was for functionalizing a self-assembled monolayer of

thioglycolic acid on the Au-gate. An increase in the hydrophilicity of the treated surface by

thioglycolic acid functionalization was confirmed by contact angle measurement as well. A

schematic cross-section of the device with Hg2+ ions bound to thioglycolic acid functionalized on

the gold gate region is as shown in Figure 4-1(a). The source-drain current-voltage

characteristics were measured at 25 C using an Agilent 4156C parameter analyzer with the Au-









sensor took less than 15 seconds for the drain to reach steady state. However, the drain current

required 30-55 seconds to reach steady state, when the sensor was exposed to the less

concentrated Hg2 ion solutions.

A less than 5 second response time was obtained for the thioglycolic acid functionalized

AlGaN/GaN HEMT sensors, when the sensor was exposed to the 105 M of the Hg2 ion solution.

This is the shortest response time of Hg2 ion detection ever reported. For the thioglycolic acid

functionalized AlGaN/GaN HEMT, the thioglycolic acid molecules on the Au surface align

vertically with carboxylic acid functional group toward the solution [140]. The carboxylic acid

functional group of the adjacent thioglycolic acid molecules form chelates of

R-COO-(Hg2+)-OOC-R with Hg2+ ion, when the sensors are exposed to the Hg2+ ion solution.

The charges of trapped Hg2+ ion in the R-COO-(Hg2+)-OOC-R chelates changed the polarity of

the thioglycolic acid molecules, which were bonded to the Au-gate through -S-Au bonds. This is

why the drain current changes in response to mercury ions. Similar surface functionalization was

used by Chang et. al. and the fluorescence was use for the detections [141]. The difference of

drain current for the device exposed to different Hg2 ion concentration to the DI water is

illustrated in Figure 4-3(b). The Hg2 ion concentration detection limit for the thioglycolic acid

functionalized sensor is 107 M, which is approximately equivalent to 27 ppb (parts per billion).

The thioglycolic acid functionalized sensor also showed excellent sensing selectivity (over 100

times higher selectivity) over Na+ and Mg2+ ions, as illustrated in Figure 4-4.

Since our sensor chip is very compact (1 mm x 5 mm) and operates at extremely low

power (8 iW based on 0.5 V of drain voltage and 80 |A of drain current operated at 11 Hz), it

can be integrated with a commercial available hand-held wireless transmitter to realize a

portable, fast response and high sensitivity Hg2 ion detector.











-0.01
-- ZnO nanorod with Pd
S-0.00
C) N2 012 Air Air Air Air
C -0.01
U)
-0.02

l-0.03- 1 om
< 2
-0.04 H100pppm
2 2 I 2
0 30 60 90 120 150

Time(min)


Figure 2-7 Relative response of Pd-coated nanorods as a function of H2 concentration in N2.


5

04

o 3

<2

1

0


0 5 10 15 20
Time(min)


25 30


Figure 2-8 Time dependence of relative resistance of Pd-coated multiple ZnO nanorods as the
gas ambient is switched from N2 to oxygen or various concentrations of H2 in air (10-
500 ppm) and then back to N2.









[95] J. F. Chang, H. H. Kuo, I. C. Leu, and M. H. Hon, Sens. Actuators B 84, 258 (1994).

[96] G. W. Hunter, P. G. Neudeck, R. S. Okojie, G. M. Beheim, V. Thomas, L. Chen, D. Lukco,
C. C. Liu, B. Ward, and D. Makel, Proc. ECS Vol 01-02, 212 (2002).

[97] A. Baranzahi, A. Lloyd Spetz, and I. Lundstrom, Appl. Phys. Lett. 67, 3203 (1995).

[98] A. Lloyd Spetz, L. Uneus, H. Svenningstorp, P. Tobias, L.-G. Ekedahl, O. Larsson, A.
Goras, S. Savage, C. Harris, P. Martensson, R. Wigren, P. Salomonsson, B. Haggendahl, P.
Ljung, M. Mattsson, and I. Lundstrom, Physica Status Solidi A 185,15 (2001).

[99] B. P. Luther, S. D. Wolter, and S. E. Mohney, Sensors Actuators B 56, 164 (1999).

[100] J. Schalwig, G. Muller, O. Ambacher, and M. Stutzmann, Phys Status Solidi A 185, 39
(2001).

[101] J. Schalwig, G. Miller, M. Eickhoff, O. Ambacher, and M. Stutzmann Sens Actuat B 87,
425 (2002).

[102] J. Schalwig, G. Miller, U. Karrer, M. Eickhoff, O. Ambacher, M. Stutzmann, L. Gorgens,
and G. Dollinger, Appl. Phys. Lett. 80, 1222 (2002).

[103] J. Kim, F. Ren, B. Gila, C. R. Abernathy, and S. J. Pearton, Appl. Phys. Lett. 82, 739
(2003).

[104] H. T. Wang, B. S. Kang, F. Ren, R. C. Fitch, J. Gillespie, N. Moser, G. Jessen, R. Dettmer,
B.P. Gila, C. R. Abernathy, and S. J. Pearton, Appl. Phys. Lett. 87, 172105 (2005).

[105] A. El. Kouche, J. Lin, M.E. Law, S. Kim, B. S. Kim, F. Ren, and S.J. Pearton, Sensors and
Actuators B: Chemical 105, 329 (2005).

[106] B. S. Kang, R. Mehandru, S. Kim, F. Ren, R. Fitch, J. Gillespie, N. Moser, G. Jessen, T.
Jenkins, R. Dettmer, D. Via, A. Crespo, B.P. Gila, C. R. Abernathy, and S.J. Pearton, Appl.
Phys. Lett. 84, 4635 (2004).

[107] B. S. Kang, F. Ren, B.P. Gila, C. R. Abernathy, and S. J. Pearton, Appl. Phys. Lett. 84,
1123 (2004).

[108] J. Kim, B. Gila, G.Y. Chung, C. R. Abernathy, S. J. Pearton, and F. Ren, Solid-State
Electron. 47, 1069 (2003).

[109] J.-R. Huang, W.-C. Hsu, Y.-J. Chen, T.-B. Wang, K.-W. Lin, H.-I. Chen, and W.-C. Liu,
Sensors Actuators B 117, 151 (2006).

[110] B. S. Kang, S. Kim, F. Ren, B. P. Gila, C. R. Abernathy, and S. J. Pearton, Sensors
Actuators B 104, 232 (2005).

[111] K. Matsuo, N. Negoro, J. Kotani, T. Hashizume, and H. Hasegawa, Appl. Surf Sci. 244,
273 (2005).









devices. The backside of the substrates was deposited with full area contacts of e-beam deposited

Ti (20 nm)/Au (200 nm) annealed at 400 OC [191]. Circular front-side contacts of Ni (20 nm)/Au

(80 nm) with diameter 200 pm were deposited by e-beam evaporation and patterned by

lithography and lift-off. A schematic of the completed diodes is shown in Figure 6-1. The

current-voltage (I-V) characteristics were measured at 300 K using a probe station and Agilent

4145B parameter analyzer. The EL spectrum and output power from the structures were

measured using a spectrometer and Si photodiode, respectively while the photoluminescence

(PL) was excited with a He-Cd laser.

Figure 6-2 shows the I-V characteristics from the implanted structures as a function of

post-implant RTA temperature under an 02 ambient for 2 mins. The I-Vs are characteristic of

back-to-back diodes for low anneal temperatures and transition to Schottky-diode like-behavior

at the highest anneal temperature. Note that for anneals at 800 OC the behavior might be

misinterpreted as that from a pn junction because the forward turn-on voltage is that expected

from a material with bandgap around 3 eV but this is misleading if not considered in the context

of all the data. Thus we do not believe that we create a p-type region by activation of the

implanted N acceptors. This is consistent with our relatively low dose, the large ionization

energy of the N and the residual n-type background of the substrate. With all of these

considerations, it is not likely we have converted the implanted region to p-type conductivity.

Figure 6-3 shows room temperature PL from the bulk ZnO before and after N+

implantation and annealing at 800 OC for 2 mins (top) and EL from MIS diode at room

temperature and 120 K (bottom). The unimplanted ZnO shows strong band-edge (-380 nm) PL,

whereas after implantation and annealing the intensity of this transition is decreased and deep

level-related emission peaked at >600 nm is introduced. This is expected, since the annealing of









CHAPTER 8
SUMMARY AND FUTURE WORK

8.1 Hydrogen Sensor Using Multiple ZnO Nanorods

A variety of different metal catalyst cluster coatings (Pt, Pd, Au, Ag, Ti, and Ni) deposited

on multiple ZnO nanorods were compared for their effectiveness in enhancing sensitivity for

detecting hydrogen at room temperature. The metal cluster coated nanorods were biased at 0.5 V

and power levels for these diode sensors were -0.4 mW. Pt-coated nanorods showed an increase

of conductance up to 8 % in room temperature upon exposure to 500 ppm hydrogen in N2. This

is a factor of two larger than that obtained with Pd, and more than an order of magnitude larger

than that achieved with the remaining metals. Pt-coated ZnO nanorods easily detected hydrogen

down to 100 ppm, with 4 % increase of conductance at this concentration after 10 min exposure.

It took a few minutes for the nanorods to return to their original conductance after switching

hydrogen off and back to air. The slow response at room temperature is a drawback in some

applications, but the sensors do offer low power operation and very good detection sensitivity.

The sensitivity for detecting hydrogen using multiple ZnO nanorods with cluster coating of

Pd on the surface is further investigated with different hydrogen concentrations. The nanorods

show changes of conductance upon exposure to hydrogen concentrations of 10-500 ppm

balanced with N2 approximately a factor of five larger than without Pd. Pd-coated ZnO nanorods

detected hydrogen down to -10 ppm, with a increase of conductance >2.6 % at 10 ppm and >4.2

% at 500 ppm H2 in N2 after 10 min exposure. The nanorods had no response to 02 at room

temperature. As opposed to the slow recovery for Pt coated nanorods, Pd coated nanorods

showed a much quicker recovery time upon switching the ambient from hydrogen to either air or

pure 02, for which approximately 95 % of the initial ZnO conductance after exposure to

hydrogen was recovered within 20 s. This rapid and easy recoverability make the Pd-coated









4-5 Changes in HEMT drain-source current for bare Au-gate and Au-gate with
thioglycolic acid functionalization exposed to 10-5 M Hg2+ ion solutions. .....................71

4-6 (a) Time dependent response of the drain current as a function of Hg2+, Cu2+, Pb2
ion concentrations for bare Au-gate AlGaN/GaN HEMT sensor. (b) Time dependent
response of the drain current as a function of Hg2+, Cu2+, Pb2 ion concentrations for
thioglycolic acid functionalized Au-gate AlGaN/GaN HEMT sensor. ..........................71

4-7 Drain current changes in response to Hg2+ and Cu2+ ions as a function of the ion
concentration for (a) the bare Au-gate and (b) the thioglycolic acid functionalized
Au-gate AlGaN/GaN HEM T sensor ........................................ ............................ 72

4-8 Plan view photograph of a multiple cell AlGaN/GaN HEMT sensors.............................72

4-9 Time dependent change in the drain current in response to Na+ and Mg2+ with bare
Au-gate and thioglycolic acid functionalized Au-gate HEMT sensor.............................73

4-10 Recyclability for (a) the bare Au-gate, and (b) the thioglycolic acid functionalized
A u-gate surface. ......................................................... ................. 73

5-1 (a) Plan view photomicrograph of a completed device with a 5 nm Au film on the
gate region. (b) schematic device cross section. The Au-coated gate area was
functionalized with KIM -1 antibody on thioglycolic acid...............................................79

5-2 IDs-VDs characteristics of HEMT in both PBS buffer and 100 ng/ml KIM-1 ...................80

5-3 Time dependent current signal when exposing the HEMT to 1 ng/ml and 10 ng/ml
KIM-1 in PBS buffer .............................................. ......... 80

5-4 Current change in HEMT as a function of KIM-1 concentration............................... 81

6-1 Schematic of ZnO MIS diode formed by N+ implantation into a bulk single crystal
sub state ......................................................... .................................86

6-2 I-V characteristics as a function of post-implant annealing temperature under an 02
am bient for 2 m ins. .........................................................................87

6-3 Room temperature PL from ZnO before and after N+ implantation and annealing at
800 C for 2 mins (top) and EL from MIS diode at room temperature and 120 K
(b bottom ) ................... .......................................................... ................. 88

6-4 I-V characteristics and forward bias current dependence of integrated EL intensity
from an MIS diode annealed at 800 OC. The EL intensity was measured by a Si
p h o to d io d e ............................. ................................................................. ............... 8 9

6-5 Optical microscope image of the emission from the diode in the dark (top) and
photos of the diode under bias from the probe contact taken both in the light and dark
(b bottom ) ................... .......................................................... ................. 89











0.042
0.041
< 0.040
- 0.039
0.038
0.037





6.0x10
5.0x10
4.0x10


S3.0xl
2.0xl
1.0xl


0 100 200 300 400 500
Time(sec)


0 200 400 600
Time(sec)


800


Figure 3-4 Time dependence of drain-source (top) or gate current (bottom) when switching from
pure N2 ambient to a 500 ppm H2 in N2 ambient and back again.









8.3 Mercury Ion Sensor Using AIGaN/GaN High Electron Mobility Transistor

Bare Au-gated and thioglycolic acid functionalized Au-gated AlGaN/GaN high electron

mobility transistors (HEMTs) were used to detect mercury(II) ions. Fast detection of less than 5

seconds was achieved for thioglycolic acid functionalized sensors. This is the shortest response

time ever reported for mercury detection. Thioglycolic acid functionalized Au-gated

AlGaN/GaN HEMT based sensors showed 2.5 times larger response than bare Au-gated based

sensors. The sensors were able to detect mercury (II) ion concentration as low as 10-7 M. The

sensors showed an excellent sensing selectivity of more than 100 for detecting mercury ions over

sodium or magnesium ions.

Bare Au-gated and thioglycolic acid functionalized Au-gated AlGaN/GaN HEMTs were

further used to detect both mercury(II) and copper(II) ions. The bare Au-gate sensor was only

sensitive to Hg2+, and thioglycolic acid functionalized sensors could detect both Hg2 and Cu2

ions. Both surfaces had a selectivity of approximately a hundred-fold over other contaminating

ions of sodium, magnesium and lead. Both bare Au-gate and thioglycolic acid functionalized

sensor can also be repeatedly used after a simple DI water rinse. By fabricating an array of the

sensors on a single chip and selectively functionalizing some sensors with thioglycolic acid, a

multi-functional specific detector can be fabricated. Such a sensor array can be used to detect

quantitatively Hg2+ ions in Cu2+ ion solution or Cu2+ ions in Hg2+ ion solution. The dimensions of

the active area of the sensor and the entire sensor chip are 50 .im x 50 .im and 1 mm x 5 mm,

respectively. Our results show that portable, selective, and fast Cu2+ and Hg2+ sensors can be

realized by combining bare Au-gated and thioglycolic acid-functionalized surface in one sensor.
















T 3 100 C
e
500C :
e 2 adsorption curve
S-- Arrhenius fitting

e slope= -1420.00457 room T
activation energy (E)= 11.805 kJ/mol
e0
0.0020 0.0025 0.0030 0.0035
1/T (K1)



Figure 2-9 Arrhenius plot of rate of resistance change after exposure to 500 ppm H2 in N2.


I I I I I I I













0.8


0.6



Qn.


0.0 ""
0.0


0.5 1.0 1.5 2.0 2.5 3.0
Voltage (V)


Figure 5-2 IDs-VDs characteristics of HEMT in both PBS buffer and 100 ng/ml KIM-1.


0 100 200


300 400


Time (s)

Figure 5-3 Time dependent current signal when exposing the HEMT to 1 ng/ml and 10 ng/ml
KIM-1 in PBS buffer.


150



145



U
5-
u 4


135


20x50 igm gate

KIM-l KIM-i
PBS I ngml 10 ng/ml







PBS biuffer-+* lg nil KIM-1-lOng/ml KIM-I
k I I


500









[131] A. Caballero, R. Martinez, V. Lloveras, I. Ratera, J. Vidal-Gancedo, K. Wurst, A. Tarraga,
P. Molina, and J. Veciana, J. Am. Chem. Soc. 127, 15666 (2005).

[132] E. Coronado, J. R. Galan-Mascar6s, C. Marti-Gastaldo, E. Palomares, J. R. Durrant, R.
Vilar, M. Gratzel, and Md. K. Nazeeruddin, J. Am. Chem. Soc. 127, 12351 (2005).

[133] Y. K. Yang, K. J. Yook, and J. Tae, J. Am. Chem. Soc. 127, 16760 (2005).

[134] M. Matsushita, M. M. Meijler, P. Wirsching, R. A. Lerner, and K. D. Janda, Org. Lett. 7,
4943 (2005).

[135] H. T. Wang, B. S. Kang, F. Ren, R. C. Fitch, J. K. Gillespie, N. Moser, G. Jessen, T.
Jenkins, R. Dettmer, D. Via, A. Crespo, B. P. Gila, C. R. Abernathy, and S. J. Pearton,
Appl. Phys. Lett. 87, 172105 (2005).

[136] B. S. Kang, S. Kim, F. Ren, J. W. Johnson, R. Therrien, P. Rajagopal, J. Roberts, E. Piner,
K. J. Linthicum, S. N .G. Chu, K. Baik, B. P. Gila, C. R. Abernathy, and S. J. Pearton,
Appl. Phys. Lett. 85, 2962 (2004).

[137] B. S. Kang, F. Ren, L. Wang, C. Lofton, Weihong Tan, S. J. Pearton, A. Dabiran, A.
Osinsky, and P. P. Chow, Appl. Phys. Lett. 87, 023508 (2005).

[138] B. S. Kang, F. Ren, M. C. Kang, C. Lofton, Weihong Tan, S. J. Pearton, A. Dabiran, A.
Osinsky, and P. P. Chow, Appl. Phys. Lett. 86, 173502 (2005).

[139] B. S. Kang, J. J. Chen, F. Ren, S. J. Pearton, J. W. Johnson, P. Rajagopal, J. C. Roberts, E.
L. Piner, and K. J. Linthicum, Appl. Phys. Lett. 89, 122102 (2006).

[140] F. Malem and D. Mandler, Anal. Chem. 65, 37 (1993).

[141] C. C. Huang and H. T. Chang, Anal. Chem. 78, 8332 (2006).

[142] M. H. Schoenfisch and J. E. Pemberton, J. Am. Chem. Soc. 120, 4502 (1998).

[143] N. J. Brewer, T. T. Foster, G. J. Leggett, M. R. Alexander, and E. McAlpine, J. Phys.
Chem. B 108, 4723 (2004).

[144] R. Lequin, Clin. Chem. 51, 2415 (2005).

[145] V. S. Vaidya, V. Ramirez, T. Ichimura, N. A. Bobadilla, and J. V. Bonventre, Am. J
Physiol. Renal. Physiol. 290, F517 (2006).

[146] Dario A. A. Vignali, Journal oflmmunological Methods 243, 243 (2000).

[147] J. Wang, Biosens. Bioelectron. 21, 1887 (2006)

[148] C. F. Sanchez, C. J. McNeil, K. Rawson, and 0. Nilsson, Anal. Chem. 76, 5649 (2004).











Ga(AI)-face


Al AIxGal.xN

SN ^i -



1 GaN


0-

Figure 1-3 Polarization induced sheet charge in Ga(Al)-face strained/relaxed AlGaN/GaN
heterostructure.



1.4 Field-Effect Based Semiconductor Sensors

The idea for semiconductor sensor based on field-effect transistors (FETs) operation

principle was first addressed by Bergveld who reported the use of metal-oxide-semiconductor

field effect transistor (MOSFET), without incorporation of the regular gate electrode, for sodium

ion and hydrogen ion detections in aqueous solution [41, 42]. In principle, for FET on p-type

semiconductor (Figure 1-4), holes are injected from source electrode into the channel and

collected at drain electrode. The conductance of this p-type channel can be tuned by a third gate

electrode capacitively coupled through a thin dielectric layer: a positive gate voltage depletes

carriers to cause a increase of space charge region underneath the gate electrode and reduces the

conductance, while a negative gate voltage attracts carriers to compensate space charge region

and leads an increase in conductance. Since the electric field resulting from binding of a charged

species to the gate dielectric is analogous to applying a voltage using a gate electrode, the









preclinical and clinical applications. Similar surface modifications can be applied for detecting

other important disease biomarkers and a compact disease diagnosis array can be realized for

multiplex disease analysis.

In summary, we have shown that the Au-gated region of an AlGaN/GaN HEMT structure

can be functionalized with KIM-1 (a kidney injury disease biomarker) antibody for the detection

of KIM-1 with a limit of detection of 1 ng/ml in PBS buffer. This electronic detection of disease

biomarker is a significant step towards a compact sensor chip, which can be integrated with a

commercial available hand-held wireless transmitter to realize a portable, fast and high sensitive

device for multiple disease diagnosis.









Figure 7-1 shows I-V characteristics of Au/GaAs Schottky diodes deposited at either 77 K

(A) or 300 K (o) with 200 [tm (left) or 800 [m (right) in contact diameter. An expanded view of

the forward voltage part of the curves is shown at the bottom of the figure. There is a clear

decrease in both reverse and forward bias current for the diodes with Au deposited at 77 K,

consistent with an increase in the effective Schottky barrier height. It has been reported that the

crystal structure and grain size of the low temperature deposited metal are different from metals

deposited at room temperature.

The changes in electrical behavior were consistent both spatially within a 2" wafer and for

different contact diameters. Figure 7-2 shows the forward current densities as a function of bias

for diodes of different diameter, deposited at either 77 or 300 K. The Schottky barrier height

enhancement by deposition at cryogenic temperatures is thought in part to result from the

minimized interaction between the metal and semiconductor during the metal deposition. The

stability of the metal-semiconductor interface and the stability of the small grain size metal

contact are important for device reliability. Typically after the gate metallization step, the device

is encapsulated by a SiNx dielectric for packaging. This dielectric is deposited by PECVD and

the temperature that the substrate is exposed to is 200-300 oC. At these temperatures, it is

desirable that interface diffusion and grain growth be minimized. Table 7-1 lists the common

metals used for Schottky contacts on GaAs and their melting temperatures. A recognized

measure for the onset of grain growth is 0.4 Tmelt. As indicated in the table, contacts of Ti, Pt,

and Pd should not recrystallize at PECVD deposition temperatures and therefore these may be

more effective choices for actual GaAs devices. For significant interfacial diffusion to take place,

the constituents of either the semiconductor or the metal contact must become mobile and form a

solid solution. The contact metals of Ti, Pt, and Pd have little diffusion below 0.4 Tmelt.









[112] J. Song, W. Lu, J. S. Flynn, and G. R. Brandes, Solid-State Electronics 49, 1330 (2005).

[113] M. Ali, V. Cimalla, V. Lebedev, H. Romanus, V. Tilak, D. Merfeld, P. Sandvik, and 0.
Ambacher, Sensors Actuators B 113, 797 (2006).

[114] 0. Weidemann, M. Hermann, G. Steinhoff, H. Wingbrant, A. Lloyd Spetz, M. Stutzmann,
and M. Eickhoff, Appl. Phys. Lett. 83, 773 (2003)

[115] L. Voss, B. P. Gila, S. J. Pearton, H. T. Wang, and F. Ren J. Vac. Sci. Technol. B 23,
2373 (2005).

[116] J. Song, W. Lu, J.S. Flynn, and G. R. Brandes, Appl. Phys. Lett. 87, 133501 (2005).

[117] F. Yun, S. Chevtchenko, Y.-T. Moon, H. Morkoc, T. J. Fawcett, and J. T. Wolan, Appl.
Phys. Lett. 87, 073507 (2005)

[118] M. Eickhoff, R. Neuberger, G. Steinhoff, O. Ambacher, G. Muller, and M. Stutzmann,
Phys. Status Solidi B 228, 519 (2001).

[119] H. T. Wang, T. J. Anderson, F. Ren, C. Li, Z. N. Now, J. Lin, B. P. Gila, S. J. Pearton, A.
Osinsky, and A. Dabiran, Appl. Phys. Lett. 89, 242111 (2006).

[120] R. Khanna, S. J. Pearton, F. Ren, and I. I. Kravchenko, J. Vac. Sci. Technol. B 24, 4
(2006).

[121] A. Renzoni, F. Zino, and E. Franchi, Environ. Res. 77, 68 (1998).

[122] J. M. Llobet, G. Falco, C. Casas, A. Teixido, and J. L. Domingo, J Agric. Food Chem. 51,
838 (2003).

[123] A. H. Stern, Environ. Res. 98, 133 (2005).

[124] F. Zahir, S. J. Rizwi, S. K. Haq, and R. H. Khan, Environ. Toxicol. Pharmacol. 20, 351
(2005).

[125] J. Mutter, J. Naumann, R. Schneider, H. Walach, and B. Haley, Neuroendocrinol. Lett. 26,
439 (2005).

[126] W. Zheng, M. Aschner, and J.-F. Ghersi-Egea, Toxicol. Appl. Pharmacol. 192, 1 (2003).

[127] I. Hoyle and R. D. Handy, Aquat. Toxicol. 72, 147 (2005).

[128] A. L. Burlingame, R. K. Boyd, and S. J. Gaskell, Anal. Chem. 68, 599 (1996).

[129] K. W. Jackson and G. Chen, Anal. Chem. 68, 231 (1996).

[130] J. L. Anderson, E. F. Bowden, and P. G. Pickup, Anal. Chem. 68, 379 (1996).





















Figure 2-2 Schematic of contact geometry for multiple nanorod gas sensor (left) and a picture of
packaged sensor (right).


0 5 10 15 20
Time(min)


25 30


Figure 2-3 Time dependence of relative resistance response of metal coated multiple ZnO
nanorods as the gas ambient is switched from N2 to 500 ppm of H2 in air as time
proceeds. There was no response to 02.


I 'A









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

FABRICATION AND CHARACTERIZATION OF ZINC OXIDE AND GALLIUM NITRIDE
BASED SENSORS

By

Hung-Ta Wang

May, 2008
Chair: Fan Ren
Major: Chemical Engineering

Pt-coated ZnO nanorods show a decrease of 8 % resistance upon exposure to 500 ppm

hydrogen in room temperature. This is a factor of two larger than that obtained with Pd;

approximately 95 % of the initial ZnO conductance was recovered within 20 s by exposing the

nanorods to 02. This rapid and easy recoverability makes the ZnO nanorods suitable for ppm-

level sensing at room temperature with low power consumption.

Pt-gated AlGaN/GaN based high electron mobility transistors (HEMTs) showed that

Schottky diode operation provides large relative sensitivity over a narrow range around turn-on

voltage; the differential designed Schottky diodes with AlGaN/GaN hetero-structure was shown

to provide robust detection of 1 % H2 in air at 25 C, which remove false alarms from ambient

temperature variations; moreover, the use of TiB2-based Ohmic contacts on Pt-Schottky

contacted AlGaN/GaN based hydrogen sensing diodes was shown to provide more stable

operation.

Thioglycolic acid functionalized Au-gated AlGaN/GaN based HEMTs were used to detect

mercury (II) ions. A fast detection (> 5 seconds) was achieved. This is the shortest response ever

reported. The sensors were able to detect mercury (II) ion concentration as low as 10 7 M. The




Full Text

PAGE 1

1 FABRICATION AND CHARACTERIZATION OF ZINC OXIDE AND GALLIUM NITRIDE BASED SENSORS By HUNG-TA WANG A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Hung-Ta Wang

PAGE 3

3 To my parents, my family.

PAGE 4

4 ACKNOWLEDGMENTS I deeply appreciate my dissertation adviso r, Prof essor Fan Ren, for his mentoring, guidance, and encouragement throughout my whole Ph.D. studies. He has created a world-class research environment in his group, giving me the opportunity to conduct many intriguing and important projects, to collaborate with many expe rts of different backgr ound levels, to broaden my fields and build my multidisciplinary knowledge, to operate many state-of-the-art instruments and utilize many research resources, and to accomplish so many research results that I never thought I could achieve within a few years. I respect his professional attitudes to research and development, and his tremendous experience and acknowledge that helped me not only to complete my degree but also to devel op myself as a top-notch researcher. I am also truly indebted to my committee members, Professor Stephen Pearton, Professor David Norton, and Professor Yiider Tseng, for thei r significant contributions to my research and my dissertation. Besides, I sinc erely appreciate Professor Jens han Lin and Professor Tanmay Lele for their important a dvice on to my research. I thank my group members: Dr. Byoung Sam Kang, Dr. Soohwan Jang, Dr. Jau-Jiun Chen, Travis Anderson, Yu-Lin Wang, Ke-Hung Chen, a nd Barrett Hicks. I es pecially thank Byoung Sam for his great help with my research life in this group. I also thank other colleagues, Dr. Brent Gila, Dr. Luc Stafford, Dr. Mark Hlad, Dr. Chih-Yang Chang, Dr. Rohit Khanna, Lars Voss, Jonathan Wright, Andrew Herrero, Wantae Lim, Thomas Chancellor, Jr., Changzhi Li, Zhen-Ning Low, and Sheng-Chung Hung. I will al ways relish my collaborations with these experts as well as the friendships built duri ng these years. Also, I have to extend my acknowledgements to Mr. Dennis Vince, Mr. James Hinnant, and Dr. Santiago Alves Tavares in

PAGE 5

5 Chemical Engineering; Dr. Ivan Kravchenko an d Mr. Bill Lewis in UF NanoFab. Without their professional technical support, I could not ha ve a smooth research life in Gainesville. Finally, I thank my family and friends in Taiwan for their endless love and support. Especially my parents, their e normous love and spiritual encour agements helped me to realize my dream. This dissertation and my Ph.D. de gree could not be accomplished without their invaluable supports.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES.........................................................................................................................9 LIST OF ABBREVIATIONS........................................................................................................ 14 ABSTRACT...................................................................................................................................18 CHAPTER 1 INTRODUCTION..................................................................................................................20 1.1 Motivation.........................................................................................................................20 1.2 Properties of GaN and ZnO.............................................................................................. 22 1.3 Two Dimensional Electron Gas of AlGa N/GaN High Electron Mo bility Transistor...... 24 1.4 Field-Effect Based Semiconductor Sensors...................................................................... 27 1.5 Dissertation Outline....................................................................................................... ...30 2 HYDROGEN SENSOR USING MULTIPLE ZNO NANORODS....................................... 31 2.1 Background.......................................................................................................................31 2.2 Detection of Hydrogen at Room Temperature with Catalyst-Coated ZnO Multip le Nanowires.......................................................................................................................32 2.3 Hydrogen Sensing Using Pd-Coated ZnO Multiple Nanowires....................................... 38 3 HYDROGEN SENSOR USING ALGAN/G AN SCHOTTKY DIODE AND HIGH ELECTRON MOBILITY TRANSISTOR............................................................................. 43 3.1 Background.......................................................................................................................43 3.2 Comparison of Gate and Drain Current Detection of Hydrogen at Room Temp erature with AlGaN/GaN High Electron Mobility Transistor............................... 44 3.3 Robust Detection of Hydrogen Using Differential AlGaN/GaN High Electron Mobility Sensing Diode.................................................................................................. 51 3.4 Stable Hydrogen Sensors from AlGaN/ GaN Heterostructure Diodes with TiB2Based Ohmic Contacts.................................................................................................... 55 4 MERCURY ION SENSOR USING ALGA N/GAN HIGH ELECTRON MOBILITY TRANSISTOR ........................................................................................................................60 4.1 Background.......................................................................................................................60 4.2 Fast Electrical Detection of Hg(II) I ons with AlGaN/GaN High Electron Mobility Transistors.................................................................................................................... ...62

PAGE 7

7 4.3 Selective Detection of Hg(II) Ions from Cu(II) and Pb(II) Using AlGaN/GaN High Electron Mobility Transistors .........................................................................................67 5 DISEASE BIOMARKER SENSOR USING ALGAN/GAN HIGH ELECTRON MOBILITY TRANSISTOR ...................................................................................................74 5.1 Background.......................................................................................................................74 5.2 Kidney Injury Molecule-1 Detection Using AlGaN/GaN High Electron Mobility Transistors.......................................................................................................................75 6 ZNO BASED LIGHT EMITTING DIODE........................................................................... 82 6.1 Background.......................................................................................................................82 6.2 Band-Edge Electroluminescence from N+-Implanted Bulk ZnO.....................................83 7 INCREASING SCHOTTKY BARRIER HEIGHT WITH CRYOGENIC METAL DEPOSITION..................................................................................................................... ....90 7.1 Background.......................................................................................................................90 7.2 Improved Au Schottky Contacts on GaAs Using Cryogenic Metal Deposition .............. 92 7.3 Thermal Stability of Au Schottky Diode s on GaAs Deposited at Either 77 K or 300K..............................................................................................................................100 7.4 Interfacial Differences in Enhanced Schottky Barrier Height Au/n-GaAs Diodes Deposited at 77 K..........................................................................................................104 8 SUMMARY AND FUTURE WORK...................................................................................111 8.1 Hydrogen Sensor Using Multiple ZnO Nanorods.......................................................... 111 8.2 Hydrogen Sensor Using AlGaN/GaN Scho ttky Diode and High Electron Mobility Transistor......................................................................................................................112 8.3 Mercury Ion Sensor Usi ng AlGaN/GaN High Electr on Mobility Transistor................. 114 8.4 Disease Biom arker Sensor Using AlGaN/ GaN High Electron Mobility Transistor...... 115 8.5 ZnO Bas ed Light Emitting Diode...................................................................................115 8.6 Increasing Schottky Barrier Height with Cryogenic Metal Deposition.......................... 116 LIST OF REFERENCES .............................................................................................................118 BIOGRAPHICAL SKETCH.......................................................................................................130

PAGE 8

8 LIST OF TABLES Table page 1-1 Physical properties of GaN and ZnO.................................................................................23 1-2 Summery of diverse biosensors with CNT, Si nanowire, and In2O3 nanowire FETs........ 29 7-1 Metals, melting tem perature, and recrystallization temperature........................................96 7-2 Summary of Au/GaAs diode characteristics fo r deposition of the Au at either 77 K or 300 K ..................................................................................................................................96 7-3 Barrier height enhancement obser ved for different me tals on n-GaAs........................... 107

PAGE 9

9 LIST OF FIGURES Figure page 1-1 Energy band-gap of GaN and ZnO based compound sem iconductors as a function of lattice constant............................................................................................................... ....24 1-2 Schematic diagram of normal AlGaN/GaN heterostructure with band diagram in the equilibrium state. 2DEG is located at the lower AlGaN/GaN interface............................ 26 1-3 Polarization induced sheet charge in Ga(Al)-face strained/relaxed AlGaN/GaN heterostructure....................................................................................................................27 1-4 Cross section of a pchannel FET under positive VG and negative VG. S, D, and G represent source, drain and ga te electrodes respectively................................................... 29 2-1 Scanning electron micrograph of ZnO multiple nanorods................................................. 35 2-2 Schematic of contact geometry for multiple nanorod gas sensor (left) and a picture of packaged sensor (right)...................................................................................................... 36 2-3 Time dependence of relative resistance response of metal coated multiple ZnO nanorods as the gas ambient is switched from N2 to 500 ppm of H2 in air as time proceeds. There was no response to O2..............................................................................36 2-4 IV characteristic of Pt-coated nanowi res in air and after 15 mi n in 500 ppm of H2 in air............................................................................................................................ ...........37 2-5 Time dependence of resistance change of Pt-coated m ultiple ZnO nanorods as the gas ambient is switched from N2 to various concentrations of H2 in air (10 ppm) and then back to N2............................................................................................................37 2-6 Time dependence of resistance of either Pd-coated or uncoated multiple ZnO nanorods as the gas ambient is switched from N2 to various concentrations of H2 in air (10 ppm) as time proceeds. There was no response to O2....................................40 2-7 Relative response of Pd-coate d nanorods as a function of H2 concentration in N2...........41 2-8 Time dependence of relative resistance of Pd-coated multiple ZnO nanorods as the gas ambient is switched from N2 to oxygen or various concentrations of H2 in air (10 ppm) and then back to N2....................................................................................41 2-9 Arrhenius plot of rate of resist ance change after exposure to 500 ppm H2 in N2..............42 3-1 IDS-VDS characteristics (top) and transfer characteristics (bottom) of Pt-gated HEMT measured at 25 C under pure N2 ambient or in a 500 ppm H2 in N2 ambient.................. 47

PAGE 10

10 3-2 Gate I-V characteristics at 0 V IDS measured at 25 C under pure N2 ambient or in a 500 ppm H2 in N2 ambient................................................................................................. 48 3-3 Change in drain-source or gate currents as a function of gate voltage (top), and percentage changes in these currents (bottom) for measurement under pure N2 ambient or in a 500 ppm H2 in N2 ambient........................................................................ 49 3-4 Time dependence of drain-source (top) or gate current (bottom) when switching from pure N2 ambient to a 500 ppm H2 in N2 ambient and back again...................................... 50 3-5 Microscopic images of differential sens ing diodes. The opening of the active diode was deposited with 10 nm of Pt, and the re ference diode was deposited with Ti/Au....... 52 3-6 Absolute (a) and differential (b) curr ent of HEMT diode me asured at 25 C................... 53 3-7 Absolute (a) and differential (b) cu rrent of HEMT diode me asured at 50 C................... 54 3-8 Time dependent test of differential HEMT diodes at 25 and 50 C.................................. 55 3-9 Schematic of HEMT diode hydrogen sens or using either conventional or TiB2-based Ohmic contacts................................................................................................................. ..57 3-10 I-V characteristics in linear (a) or log (b) form of Pt-gat ed diode measured in air or 1 % hydrogen ambient at 25 C............................................................................................ 58 3-11 Time-dependence of current test biased by 1.5 V of Pt-gated diode as the ambient is switched from air to 1 % hydrogen and back to air. .......................................................... 59 3-12 Variation in forward current at fixed bias for diodes with boride-based Ohmi c contacts (top) or conventional Ohmic contacts (bottom) as a function of time under field conditions where the temperature in creases during the day and decreases at night...................................................................................................................................59 4-1 (a) A schematic of AlGaN/GaN HEMT. Th e Au-coated gate area was functionalized with thioglycolic acid. (b) Plan view phot omicrograph of a com pleted device with a 5 nm Au film in the gate region......................................................................................... 65 4-2 Photographs of contact angle of water drop on the surface of bare Au (left) and thioglycolic acid functio nalized Au (right). .......................................................................65 4-3 (a) Time dependent response of the dr ain curren t for bare Au-gate AlGAN/GaN HEMT sensor and thioglycolic acid functi onalized Au-gate HEMT sensor. (b) Drain current of a thioglycolic acid functionalized Au-gate HEMT sensor as a function of the Hg2+ ion concentration................................................................................................. 66 4-4 Time dependent response of the drain current for detecting Na+, Mg2+ or Hg2+ with a thioglycolic acid functionaliz ed Au-gate HEMT sensor................................................... 67

PAGE 11

11 4-5 Changes in HEMT drain-source curren t for bare Au-gate and Au-gate with thioglycolic acid functio nalization exposed to 10-5 M Hg2+ ion solutions........................ 71 4-6 (a) Time dependent response of th e drain current as a function of Hg2+, Cu2+, Pb2+ ion concentrations for bare Au-gate AlGa N/GaN HEMT sensor. (b) Time dependent response of the drain curr ent as a function of Hg2+, Cu2+, Pb2+ ion concentrations for thioglycolic acid functionalized Au -gate AlGaN/GaN HEMT sensor..............................71 4-7 Drain current changes in response to Hg2+ and Cu2+ ions as a function of the ion concentration for (a) the bare Au-gate a nd (b) the thioglycolic acid functionalized Au-gate AlGaN/GaN HEMT sensor..................................................................................72 4-8 Plan view photograph of a multip le cell AlGaN/GaN HEMT sensors.............................. 72 4-9 Time dependent change in th e drain current in response to Na+ and Mg2+ with bare Au-gate and thioglycolic acid func tionalized Au-gate HEMT sensor............................... 73 4-10 Recyclability for (a) the bare Au-gate, and (b) the thioglycolic acid functionalized Au-gate surface. ............................................................................................................... ..73 5-1 (a) Plan view photomicrograph of a comp leted device with a 5 nm Au film on the gate region. (b) schem atic device cross section. The Au-coated gate area was functionalized with KIM-1 an tibody on thioglycolic acid................................................. 79 5-2 IDS-VDS characteristics of HEMT in both PBS buffer and 100 ng/ml KIM-1................... 80 5-3 Time dependent current signal when exposing the HEMT to 1 ng/ml and 10 ng/ml KIM-1 in PBS buffer......................................................................................................... 80 5-4 Current change in HEMT as a function of KIM-1 concentration...................................... 81 6-1 Schematic of ZnO MIS diode formed by N+ implantation into a bulk single crystal substrate...................................................................................................................... .......86 6-2 I-V characteristics as a function of post-implant ann ealing te mperature under an O2 ambient for 2 mins............................................................................................................. 87 6-3 Room temperature PL from ZnO before and after N+ implantation and annealing at 800 C for 2 mins (top) and EL from MI S diode at room temperature and 120 K (bottom)....................................................................................................................... .......88 6-4 I-V characteristics and forward bias cu rrent dependence of in tegrat ed EL intensity from an MIS diode annealed at 800 C. The EL intensity was measured by a Si photodiode..........................................................................................................................89 6-5 Optical microscope image of the emissi on from t he diode in the dark (top) and photos of the diode under bias from the probe contact taken both in the light and dark (bottom)....................................................................................................................... .......89

PAGE 12

12 7-1 I-V characteristics of Au/GaAs Schottky diodes deposited at either 77 K( ) or 300 K( ) for both 200 m dia.(left) and 800 m dia.(right) contact. An expanded view of the forward voltage part of the curves is shown at bottom................................................ 97 7-2 Forward current densities as a function of bias for di odes of different diameter deposited at either 77 or 300 K. ......................................................................................... 97 7-3 Reverse current at -4 V for diodes deposited at either 77 or 300 K, as a function of either contact diameter (top) or area (bottom )...................................................................98 7-4 Time dependence of forward bias at a curr ent of 10 mA for diodes deposited at either 77 or 300 K........................................................................................................................98 7-5 Optical microscope images of Au cont acts deposited at 77 K (left) or 300 K (right)....... 99 7-6 XRR of thin (~90 ) Au layers of GaAs for the two different deposition temperatures and the associated Au surface rough ness and Au/GaAs interfacial roughness derived from the XRR......................................................................................99 7-7 I-V characteristics of 400 m dia meter diodes deposited at either 300 K (left) or 77 K (right), as a function of post-de position annealing temperature.................................. 102 7-8 Forward I-V characteristics of 400 m diameter diodes deposited at either 300 K (left) or 77 K (right), as a function of post-deposition annealing tem perature................ 102 7-9 Schottky barrier height as a function of annealing temperature for diodes deposited at either 77 or 300 K. ........................................................................................................... 103 7-10 Reverse leakage current (@ -4 V) on two different scales as a function of annealing temp erature for diodes deposited at either 77 or 300 K................................................... 103 7-11 Reverse breakdown voltage (@ -100 A) as a function of annealing temperature for diodes deposited at either 77 or 300 K............................................................................. 104 7-12 Optical micrograph images of Ti deposited at either 77 K (top left) or 300 K (top right) and Au at 77 K (bottom left) or 300 K (bottom right) on GaAs............................ 108 7-13 XRR spectra from 77 K Au/GaAs diodes as a function of post-d eposition annealing temperature. ................................................................................................................... ..108 7-14 XRR spectra from Au/GaAs diodes deposited at either 77 K or 300 K before (left) or after (right) annealing at 300 C.......................................................................................109 7-15 Interfacial Au/GaAs roughness and metal/a ir roughness data derived from the XRR spectra for samples deposited at either 300 K (left) or 77 K (right), as a function of annealing temperature. ..................................................................................................... 109

PAGE 13

13 7-16 Comparison of metal ro ughness (left) and me tal-semiconductor interfacial roughness (right) for the two types of diodes as a function.............................................................. 110

PAGE 14

14 LIST OF ABBREVIATIONS 2DEG: Two Di mensional Electron Gas AAS: Atomic Absorption Spectroscopy AES: Auger Electron Spectroscopy Ag: Silver AKI: Acute Kidney Injury Al: Aluminium Al2O3: Aluminium oxide AlGaN: Aluminum Gallium Nitride Ar: Argon ARF: Acute Renal Failure As: Arsenic Au: Gold Cd: Cadmium Cl2: Chlorine CNT: Carbon Nanotube Cu: Copper DNA: deoxyribonucleic acid ELISA: enzyme-linked immunsorbent assay FET: Field Effect Transistor EL: electroluminescence GaAs: Gallium Arsenide GaN: Gallium Nitride H2: Hydrogen He: Helium

PAGE 15

15 HEMT: High Electron Mobility Transistor Hg: Mercury ICP: Inductively Coupled Plasma ICP-MS: Inductively Coupled Plasma-Mass Spectrometry In: Indium In2O3: Indium(III) oxide InGaAs: Indium Gallium Arsenide InP: Indium Phosphide ISE: Ion Selective Electrodes I-V: Current-Voltage IR: Infra Red KIM-1: Kidney Injury Molecules-1 LD: Laser Diode LED: Light Emitting Diode LOD: Limit of Detection MBE: Molecular Beam Epitaxy Mg: Magnesium MIS: Metal Insulator Semiconductor MOCVD: Metal Organic Chemical Vapor Deposition MOSFET: Metal Oxide Semiconductor Field Effect Transistor MSFET/MESFET: Metal Semiconducto r Field Effect Transistors N2: Nitrogen Na: Sodium NH3: ammonia Ni: Nickel

PAGE 16

16 NO2: nitrogen dioxide O2: Oxygen O3: Ozone Pb: Lead PBS: Phosphate Buffered Saline Pd: Palladium PEI: polyethyleneimine PEM: proton-exchange membrane PIN diode: Positive-Intrinsic-Negative diode PL: photoluminescence PMMA: polymethyl methacrylate ppb: parts per billion ppm: parts-per-million Pt: Platinum RTA: Rapid Thermal Annealing SiC: Silicon Carbide SnO2: Tin(IV) oxide TE: Thermionic Emission TFE: Thermionic Field Emission Ti: Titanium TiB2: Titanium Boride UV: Ultra Violent XPS: X-ray Photoelectron Spectroscopy XRR: X-ray reflectivity

PAGE 17

17 Zn: Zinc ZnO: Zinc Oxide

PAGE 18

18 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy FABRICATION AND CHARACTERIZATION OF ZINC OXIDE AND GALLIUM NITRIDE BASED SENSORS By Hung-Ta Wang May, 2008 Chair: Fan Ren Major: Chemical Engineering Pt-coated ZnO nanorods show a decrease of 8 % resistance upon exposure to 500 ppm hydrogen in room temperature. This is a factor of two larger than that obtained with Pd; approximately 95 % of the initial ZnO conductance was recovered within 20 s by exposing the nanorods to O2. This rapid and easy recoverability makes the ZnO nanorods suitable for ppmlevel sensing at room temperature with low power consumption. Pt-gated AlGaN/GaN based high electron mobi lity transistors (HEMTs) showed that Schottky diode operation provides large relative sensitivity over a narrow range around turn-on voltage; the differential designed Schottky diodes with AlGaN/Ga N hetero-structure was shown to provide robust detection of 1 % H2 in air at 25 C, which rem ove false alarms from ambient temperature variations; mo reover, the use of TiB2-based Ohmic contacts on Pt-Schottky contacted AlGaN/GaN based hydrogen sensing diodes was shown to provide more stable operation. Thioglycolic acid functionalized Au-gated Al GaN/GaN based HEMTs were used to detect mercury (II) ions. A fast detection (> 5 seconds) was achieved. This is the shortest response ever reported. The sensors were able to detect mercury (II) ion concentration as low as 107 M. The

PAGE 19

19 sensors showed an excellent sensing selectivity of more than 100 of detecting mercury ions over sodium, magnesium, and lead ions, but not copper. AlGaN/GaN based HEMTs were used to detect kidney injury molecule-1 (KIM-1), an important biomarker for early kidney injury dete ction. The HEMT gate region was coated with KIM-1 antibodies and the HEMT source-drain current showed a clear dependence on the KIM-1 concentration in phosphate-buffered saline (PBS ) solution. The limit of detection (LOD) was 1ng/ml using a 20 m m gate sensing area. This appr oach shows a potential for both preclinical and clinical disease diagnosis with accurate, rapid, noninva sive, and high throughput capabilities. The rest of this dissertation includes Zn O band edge electroluminescence from N+implanted ZnO bulk, and the investigation of cryogenic gold Schottky contact on GaAs for enhancing device thermal stability.

PAGE 20

20 CHAPTER 1 INTRODUCTION 1.1 Motivation According to 2007 published Sensors: A Global Strategic Business Report, the global sensor m arket grows averagely at an a nnual rate of 4.5 % between 2000 and 2008 and is expected to reach US $61.4 billion by 2010 [1]. The US $11.5 billion worth of chemical sensor m arket represents the largest segment of this global sensor market. This includes chemical detection in gas, chemical det ection in liquid, flue gas and fire detection, liquid quality sensor, and biosensor. Semiconductor based sensor fabr icated using the matu re micro-fabrication techniques and/or novel nanotechnology is one of the major contestants in this market. Silicon based devices remain dominating due to their lo w cost, reproducible and controllable electronic behaviors, and abundant data of chemical treatment on silicon oxide or glass. However, they are unable to be operated at harsh environment, for instance, high temperature, high pressure, and corrosive ambient, so the application area is still limited. The two wide band-gap compound semiconductors, Gallium Nitrite (GaN) and Zinc Ox ide (ZnO), are very good alternative options to replace silicon because of many advantages, fo r example, highly chemical resistance, potential for high power operation, and blue and u ltraviolet optoelect ronic behaviors [ 2, 3]. A variety of sensors have been reported usi ng GaN or ZnO materials, such as nanorod/wire, homo-structured thin film and hetero-structured thin film base d devices (diodes, transistors, surface acoustic wave devices, or elec trochemical electrodes). ZnO and GaN nanorods/wires are extremely attr active for sensing applications. In nature, 1-D nanostructures could dramatically enhance the sensitivity due to thei r high surface to volume ratio, Debye length comparable diameter, better cr ystallinity than 2-D thin film, and quantum effect [ 4, 5, 6]. In addition, for most of these applicat ions, the nanorod/wire sensors have very

PAGE 21

21 low power requirements and minimal weight. Combined with the native advantageous characteristics of ZnO and GaN, ZnO and GaN nanorods/wires are natural candidates for this type of sensing application. GaN/AlGaN high electron mobility transistors (HEMTs) have been extremely useful for gas and liquid sensor for primarily two reasons: 1) they consist of a hi gh electron sheet carrier concentration channel induced by piezoelectric po larization of the strained AlGaN layer and 2) the carrier concentration strongly depends on the ambient [ 79]. In addition, sensors fa bricated from these wide band-gap semiconductors could be readily integrated with solar blind UV detectors or high temperat ure, high power electronics on the sa me chip. For these reasons, nitride HEMTs are versatile devices that may be us ed for a variety of sensing applications. On the other hand, ZnO is an attractive candidate for Ultra Violent (UV) Light Emitting Diodes (LEDs) since it is an environmentally friendly material which is grown at low temperatures on cheap transparent substrates and has both a direct wide band gap of 3.3 eV and a very large exciton binding energy of 60 me V, important for robust light emission [ 10, 11]. Finally, it is important to increase Schottky barrier height in order to solve the reliability issues to compound semiconductor based HMETs. In pa rticular, the gate reliability has been problematic. Increasing the Schottky barrier heights can improve th e gate leakage current, gatedrain breakdown voltage, output resistance and power gain, and noise performance. Promising results engineering Schottky barrier heights have been demonstrated by cryogenic metal deposition at 77 K for GaAs, InP, and InGaAs [ 12 13, 14].

PAGE 22

22 1.2 Properties of GaN and ZnO The properties of GaN and ZnO [ 2 15, 16] are summarized and lis ted in Table 1-1 GaN can fo rm either Wurtzite crystal structure with a= 3.19 and c= 5.19 or Zinc Blende crystal structure with a=4.52 and c= 4.5 Because of its large direct band-gap (Eg=3.5 eV), high thermal stability, high el ectron mobility (1000 cm2/V s), and other physical properties, GaN and its alloys with Al and In have been the basic materials for short-wavelength optoelectronics, and high-power, high-temperature elec tronic devices and sensors. The energy gaps in these considered compounds (6.2, 3.4 and 1.9 eV for AlN, GaN and InN respectively) cover the whole visible spectrum and a large part of the UV range, as shown in Figure 1-1 At present, GaN based high-brightn ess blue and green light emitting diodes (LEDs) and low-power blue laser diodes (LDs) are commercially available. On the othe r hand, however, the development of GaN-based technology was, and still is, strongly limited by difficulties in obtaining large, high-quality crystals which could be used as substrates for epitaxial deposition of multilayer quantum structures necessary for devices [ 3, 17]. ZnO norm ally forms in the hexagonal wurtzite crystal structure with a=3.25 and c= 5.20 The Zn atoms are tetrahedrally coordi nated to four O atoms, where the Zn d electrons hybridize with the O p electrons. ZnO is also a direct band-gap semiconductor with Eg=3.4 eV. The band gap of ZnO, similar to GaN, can be tuned either up via Mg substitution or down via Cd substitution on the catio n site, as shown in Figure 1-1 Substituting Mg on the Zn site in epitaxial films can increase the band gap to approxim ately 4.0 eV while still main taining the wurtzite structure. The electron Hall mobility in Zn O single crystals is on the order of 200 cm2/V s at room temperature [ 18]. While the electron mobility is lowe r than that for GaN, ZnO has a higher theoretical saturation velocity. Electron doping in nom inally undoped ZnO has been attributed to Zn interstitials, oxygen vacancies, or hydrogen [ 1924]. The intrinsic defect levels that lead to n -

PAGE 23

23 type doping lie approximately 0.01.05 eV below the conduction band. A strong room temperature near-band-edge UV photoluminescence p eak at ~3.2 eV is attributed to an exciton state, as the exciton binding en ergy is on the order of 60meV [ 25]. In addition, visible emission is also observed due to defect states. A blue-green emission, cent ered at around 500nm in wavelength, has been explained within the context of transitions involving self-activated centers formed by a doubly ionized zinc vaca ncy and an ionized interstitial Zn+ [ 26], oxygen vacancies [ 2730], donoracceptor pair recom binati on involving an impurity acceptor [ 31], and/or interstitial O [ 3234]. Table 1-1 Physical proper ties of GaN and ZnO. Propert y GaN ZnO Crystal structure Wurtzite Zinc Blende Wurtzite Lattice constant(nm) a0 0.3189 0.452 0.3249 c0 0.5185 0.45 0.5207 a0/c0 1.6259 1.602 Density(g/cm3) 6.15 5.606 Thermal conductivity(W cm-1C-1) >2.1 0.6, 1-1.2 Linear expansion coefficient(C-1) a0 5.59-6 6.50-6 c0 3.17-6 3.00-6 Energy bandgap, Eg (eV) 3.51, direct 3.3, direct 3.4, direct Exciton binding energy(meV) 28 60 for n-type Electron effective mass 0.2 0.24 Electron Hall mobility at 300K(cm2V-1s-1) ~1000 ~1000 200 for p-type Hole effective mass 0.8 0.59 Hole Hall mobility 200 350 5~50 Electron saturation velocity(107cm s-1) 2-2.5 2.0 3.2

PAGE 24

24 Figure 1-1 Energy band-gap of GaN and ZnO base d compound semiconductors as a function of lattice constant. 1.3 Two Dimensional Electron Gas of AlGaN /GaN High Electron Mobility Transistor AlGaN/GaN heterojunction has been shown to form a potential well and a two dimensional electron gas (2DEG) at the lowe r heteroin terface, as show in Figure 1-2 The application of AlGaN/GaN 2DEG had accelerated the developm ent of high voltage, h igh power operated microwave devices for use in broad band power amplifiers in wireless base station applications [ 35, 36, 37]. These devices are so cal led AlGaN/GaN high electron mobility transistor, or AlGaN/GaN HEMTs, well known for high electron m obility in the 2DEG channel, h ighest sheet carrier concentration among III-V material system, high satu ration velocity, high breakdown voltage, and thermal stability. When wide band AlxGa1-xN and narrow band GaN are broug ht into contact, thermal equilibrium align their respective Fermi levels (EF) that both conduction (Ec) and valence (Ev) band are bent and cause the GaN conduction band at the interface to drop below EF. Free

PAGE 25

25 electrons will fill the triangular well and form 2DEG. From a nother viewpoint, as described by Ambacher et al., 2DEG is the compensation to a fixed sheet charge i nduced by both spontaneous polarization (SPP ) and piezoelectric or st rain-induced polarization (PEP) [ 38, 39, 40]. The spontaneous polarization for AlxGa1-xN is, )029 .0052.0()( x xPSP C/m2, (1-1) where the x is the Al concentration in AlxGa1-xN. In both AlN and GaN system, the spontaneous polarization is ne gative meaning that the sponta neous polarization is pointing toward substrate for Ga(Al)-face, and toward surface for N-face. The piezoelectric polarization can be calculated by, )(31 33 yx z PEeeP (1-2) where 33e and 31e are piezoelectric coefficients, ) /(00cccz is the strain in c-axis, )/(00aaayx is the in-plane strain that is assumed isotropic, and a, c are the lattice constants of the strained layer. Negative piezoelectric polariza tion represents that a GaN layer under compressive strain, and/or a AlGaN layer under tensile strain fo r Ga(Al)-face crystal. For Ga-face AlGaN/GaN grown on c -Al2O3 substrate, tensile stra in AlGaN layer contact with relaxed GaN layer and the spontaneous polar ization and piezoelectric polarization align in parallel ( Figure 1-3 ). The fixed charge induced by to tal polarization can be derived by, )()( (0) ()( ) () ()()()(1 1 1 1PE SP xxPE xxSP SP xxPE xxSP SPPP NGaAlPNGaAlPGaNP NGaAlPNGaAlPGaNPtopPbottomP (1-3) The polarization induced sheet charge in this case is positive ( ), and free electrons, therefore, tend to compensate it to form 2D EG at GaN interface with a sheet carrier concentration (sn ), which is expected to be,

PAGE 26

26 )()()( )( )( )(2 0xExExe ed x e x xnC F b d s (1-4) where )( x is piezoelectric polarization, ) ( x is the dielectric constant, dd is the AlGaN layer thickness, be is the Schottky barrier of the gate contact on AlGaN, FE is the Fermi level and CE is the conduction band discontinuity between AlGaN and GaN. Because the spontaneous and piezoelectric polarization increase with Al concentration of AlGaN la yer, the typical sheet carrier concentration could reach 13106.1 cm-2 for3.0 x, in excess of other available III-V material systems. Figure 1-2 Schematic diagram of normal AlGaN/Ga N heterostructure with band diagram in the equilibrium state. 2DEG is located at the lower AlGaN/GaN interface.

PAGE 27

27 Figure 1-3 Polarization induced sheet charge in Ga(Al)-face strained/relaxed AlGaN/GaN heterostructure. 1.4 Field-Effect Based Semiconductor Sensors The idea for semiconductor sensor based on field-effect transistors (FETs) operation principle was first addressed by Bergveld w ho reported the use of metal-oxide-semiconductor field effect transistor (MOSFET), without incorpor ation of the regular gate electrode, for sodium ion and hydrogen ion detect ions in aqueous solution [ 41, 42]. In principle, for FET on p-type sem iconductor ( Figure 1-4 ) holes are injected from sour ce electrode into the channel and collected at drain electro de. The conductance of this p-type channel can be tuned by a third gate electrode capacitively coupled through a thin dielectric layer: a positive gate voltage depletes carriers to cause a increase of space charge regi on underneath the gate electrode and reduces the conductance, while a negative gate voltage attrac ts carriers to compensate space charge region and leads an increase in conductance. Since the electric field resulting fr om binding of a charged species to the gate dielectric is analogous to applying a voltage using a gate electrode, the

PAGE 28

28 potentiometric chemical signal can be meas ured by monitoring the conductance of the semiconductor channel that is field-effect sensitively modulated in a FET structure [ 43, 44]. The dependence of the conductance on gate voltage hence m a kes FETs natural candidates for electrical sensing of any binding of charged sp ecies. Thin film FET sensor had, however, constrained by limited sensitivity until the advent of one dimensional nanotubes, wires, and rods. Field-effect based semiconductor sensors made with those nanostructures have made a great impact, due to a dramatically in creased surface to volume ratio w ith nanoscale diameter channel, that single molecule sensitivity might be possible. The progress of nanostructure FET sensor r ecently has been accelerated by the successful surface functionalizations on semiconductor surface for selectivity detection purpose. For gas sensing, bare Carbon Nanotubes (CNTs) wa s reported that nitrogen dioxide (NO2) and ammonia (NH3) adsorptions result in reverse conducta nce responses because of the electron withdraw/donating mechanism[ 45 ]; The coating of a m ine group ri ch polymer, polyethyleneimine (PEI), on CNT causes an improved sensitivity and selectivity to NO2 and other acidic gas, while the sulfonic acid group (-SO3H) rich polymer, Nafion, obtaine NH3 selective detection [ 46, 47]; Catalytic m etal, Pt or Pd, coating made CNTs [ 48 49] and Tin(IV) Oxid e (SnO2) nanowires [ 50] highly sensitive to hydrogen. For biosensing in aqueous solution, CNT have successfully detect ed specific proteins [ 51, 52, 53, 54] and glucose [ 51, 55, 56]; Si nanowires are able to detect protein [ 57, 58], d eoxyribonucleic acid (DNA) [ 57 59 ], drug [ 57, 60], and virus [ 57, 61]; Indium (III) oxide (In2O3) nanowires are also sensitive to protein [ 54] and DNA [62] though appropriate surface functionalizations. Table 1-2 is a summ ery of diverse chem ical functionalization approaches applied to differe nt nanostructure FET se nsor for specific bio-

PAGE 29

29 detections using selective bi o-interaction, e.g. antigen-antibody reaction and DNA-DNA interaction. Table 1-2 Summery of diverse biosenso rs with CNT, Si nanowire, and In2O3 nanowire FETs. 1-D nanomaterial 1st layer 2nd layer 3rd layer 4th layer 5th layer sensing mechanism CNT (ref. 51, 52) protein hydrophobic-hydrophobic interaction CNT (ref. 51, 52) N -alip hatic molec ules -biotin protein biotin-protein interaction CNT (ref. 51, 52) N -alip hatic molec ules -protein protein antobody-antigen interaction CNT (ref. 51, 53) N -pol ymer(PEI) -biotin protein biotin-protein interaction CNT (ref. 54) -phosphatic molecules -antibody antigen antobody-antigen interaction CNT (ref. 51, 55, 56) N -gluc o se oxidase glucose redox process Si-NW (ref. 57, 58) amine group -biotin protein biotin-protein interaction Si-NW (ref. 57, 59) amine group -protein -biotin -DNA c-DNA DNA-DNA interaction Si-NW (ref. 57, 60) amine group -protein drug inhibitor reaction Si-NW (ref. 57, 61) amine group -antibody virus antibody-virus interaction In2O3-NW (ref. 54) -phosphatic molecules -DNA c-DNA DNA-DNA interaction In2O3-NW (ref. 62) -phosphatic molecules -antibody antigen antobody-antigen interaction Note: represents covalent functionalization; N representscovalent functionalization; highlighed final layer is the target analyte. Figure 1-4 Cross section of a p-channel FET under positive VG and negative VG. S, D, and G represent source, drain and ga te electrodes respectively.

PAGE 30

30 1.5 Dissertation Outline The main objective of this work focuses on the design and fabrication of GaN and ZnO based sensing devices, and analysis of their sensitivity to hydrogen, heavy metal ions, and disease biomarkers. Chapter 1 reviews the materi al properties and recent sensor technologies combining field effect based devices with su rface chemical functi onalization. Chapter 2 illustrates details of fabrication and sensitivit y measurement to hydrogen using catalytic metal coated multiple ZnO nanowires. Chapter 3 pr esents hydrogen detec tion using AlGaN/GaN HEMTs and Schottky diodes with the design of differential diodes and the use of TiB2 based Ohmic contact. This is the extension for the previous work done by Dr. Byoung Sam Kang [ 63 ], toward practical applications. Chapter 4 focuse s on the m e rcury ion detection using AlGaN/GaN HEMTs through a surface coating of carboxyl gro ups on gate region. Chapter 5 presents the detection of disease biomarker, kidney injury molecule-1 (KIM-1), using AlGaN/GaN HEMTs which are gate-coated with KIM-1 antibody. The rest of this dissertation also includes zno band edge electroluminescence n+-implanted zno bulk (chapter 6), and investigation of cryoge nic schottky contact on gaas regarding thermal stability and interfacial differences (chapter 7) Chapter 8 briefly summarizes a conclusion of above works and suggests future studies.

PAGE 31

31 CHAPTER 2 HYDROGEN SENSOR USING MULTIPLE ZNO NANOR ODS 2.1 Background There is a strong need to develop hydroge n sensors for use with proton-exchange membrane and solid oxide fuel cells for space craft and other long-term applications. A key requirement for these sensors is the ability to selectively detect hydrogen at room temperature with minimal power use and weight. Nanorods an d nanotubes are potential candidates for this type of sensing. In the case of hydrogen sens ing with carbon nanotubes (CNTs), several groups have reported that use of Pd doping or films or loading with Pd nanoparticles can functionalize the surface of nanotubes for cat alytic dissociation of H2 to atomic hydrogen [ 48, 49]. Of course, thin-film se nsors of Si, GaAs, InP, SiC, and GaN with Pd contacts have been used previously as hydrogen sensors [ 64]. ZnO nanowires and nanorods have shown potential for use in gas, hum i dity, and chemical sensing [ 65, 66, 67]. The ability to make arrays of nanorods with large total surface area has be en dem onstrated with a number of different growth methods [ 5 68 89] and a large variety of ZnO one-dimensiona l stru ctures h as been demonstrated [ 87]. The larg e surface area of the nanorods and bi o-safe characteris tics of ZnO makes them attractive for both chemical sensing and biomedical applications. In this chapter, we demonstrate hydrogen detection use catalytic metal coated ZnO nanorod field effect based sensor. Section 2.2 pr esents a comparison of different metal coating layers on multiple ZnO nanorods for enhancing the sensitivity to detection of hydrogen at room temperature. Pt is found to be the most effec tive catalyst, followed by Pd The resulting sensors are shown to be capable of detecting hydrogen in the range of ppm at room temperature using very small current and voltage requirements, a nd recovering quickly after the source of hydrogen is removed. In section 2.3, i ssues like quantifying the sensitivity, limit of detection (LOD) at

PAGE 32

32 room temperature, power consumption of the sensors, and time response upon switching away from the H2-containing ambient are discussed in a Pd-coated multiple ZnO nanorod sensor. The sensors are also shown to detect ppm hydrogen at room temperature using <0.4 mW of power. 2.2 Detection of Hydrogen at Room Tempera ture with Catalyst-C oated ZnO Multiple Nanowires ZnO nanorods were grow n by nucleating on an Al2O3 substrate coated with Au islands [ 90]. For nom i nal Au film thicknesses of 20 discontinuous Au island s are realized after annealing. The nanorods were deposited by molecu lar beam epitaxy (MBE) w ith a base pressure of 58 mbar using high-purity (99. 9999 %) Zn metal and an O3/O2 plasma discharge as the source chemicals. The Zn pressure was varied between 46 and 27 mbar, while the beam pressure of the O3/O2 mixture was varied between 56 and 54 mbar. The growth time was ~2 h at 600 C. The typical length of the resu lting nanorods was 2m, with typical diameters in the range of 30 nm. Figure 2-1 s hows a scanning electronmicrograph of the asgrown rods. Selected area diffrac tion patterns showed the nanorods to be single crystalline. In som e cases, the nanorods were coated with Pd, Pt, Au, Ni, Ag or Ti thin films (~100 thick) deposited by sputtering. Pd and Pt are known to be the most effective catalysts for dissociation of molecular hydrogen; Au was chosen to see if it could provide any enhancement in hydrogen sensitivity since it might potentially be used as an over-layer to prevent oxidation of the other metals, which are all significantly cheaper than Pd and Pt and were explored from the viewpoint of keeping the overall cost of the sens or fabrication as low as possible. Contacts to the multiple nanorods were formed using a shadow mask and e-beam evaporation of Al/Ti/Au el ectrodes. The separation of the electrodes was ~300 m. A schematic of the resulting sensor (left) and a pictur e of packaged sensor (right) are shown in Figure 2-2 Au

PAGE 33

33 wires were bonded to the contact pad for current voltage (IV) measurements performed at 25 C in a range of different ambients (N2, O2 or 10 ppm H2 in N2). Note that no currents were measured through the discontinuous Au islands and no thin film of ZnO on the sapphire substrate was observed with the growth condition for the na norods. Therefore, the m easured currents are due to transport through the nanorods themselv es. The IV characteristics of the multiple nanorods were linear with typical currents of 0.8 mA at an applied bias of 0.5 V. Figure 2-3 shows the tim e dependence of relative resistance change of either metal-coated or uncoated multiple ZnO nanorods as the gas ambient is switched from N2 to 500 ppm of H2 in air and then back to N2 as time proceeds. These were measured at a bias voltage of 0.5 V. The first point of note is that there is a strong increase (by approximately a factor of five) in the response of the Pt-coated nanorods to hydrogen re lative to the uncoated devices. The maximum response was ~8 %. There is also a strong enha ncement in response with Pd coatings, but the other metals produce little or no change. This is consistent with the known catalytic properties of these metals for hydrogen dissociation. Pd has a highe r permeability than Pt but the solubility of H2 is larger in the former [ 91]. Moreover, studies of the bonding of H to Ni, Pd and Pt surfaces have shown that the adsorption energ y is lowest on Pt [ 92]. There was no response of either type of nanorod to the presence of O2 in the ambient at room temperature. Once the hydrogen is removed from the ambient, the recovery of th e initial resistance is rapid (<20 s). By sharp contrast, upon introduction of th e hydrogen, the effective nanorod resistance continues to change for periods of >15 min. This suggests that th e kinetics of the chemis orption of molecular hydrogen onto the metal and its dissociation to atomic hydrogen are the ra te-limiting steps in the resulting change in conductance of ZnO [ 88 ]. The gas-sensing m echanis m s suggested in the past include the desorption of adsorbed surface hydrogen and grain boundaries in poly-ZnO [ 93 ],

PAGE 34

34 exchange of charges between adsorbed gas spec ies and the ZnO surface leading to changes in depletion depth [ 94] and changes in surface or grain-boundary conductio n by gas adsorption/desorption [ 95 ] Finally, Figure 2-3 shows an incubation tim e for response of the sensors to hydrogen. This could be due to some of the Pd (or Pt) becom ing covered with native oxide, which is removed by exposure to hydrogen. A potential solution is to use a bi-layer deposition of the Pt/Pd followed by a very thin Au layer to protect the Pd from oxidation. However, this adds to the complexity and cost of the process and, since the Pd is not a continuous film, the optimum coverage of Au wo uld need to be determined. We should also point out that the IV characteristics were the same when measured in vacuum as in air, indicating that the sensors are not sensitive to humidity. The power requirements for the sensors were very low. Figure 2-4 shows the IV characteristics m easured at 25 C in both a pure N2 ambient and after 15 min in a 500 ppm H2 in N2 ambient. Under these conditions, the resistance response is 8 % and is achieved for a power requirement of only 0.4 mW. This compares well with competing nanotechnologies for hydrogen detection such as Pd-loaded carbon nanotubes [ 48, 49]. Moreover, the 8 % response compares very well to the existing SiC-based sensor s, which operate at temperatures >100 C through an on-chip heater in order to enhance the hydrogen dissociation efficiency [64]. Figure 2-5 shows the sensors can detect 100 ppm H2. In conclusion, Pt-coated ZnO nanorods a ppear well suited to detection of ppm concentrations of hydrogen at room temperatur e. The recovery charac teristics are fast upon removal of hydrogen from the ambient. The Zn O nanorods can be placed on cheap transparent substrates such as glass, making them attrac tive for low-cost sensing applications, and can operate at very low power conditions. Of course, th ere are many issues still to be addressed, in

PAGE 35

35 particular regarding the reliabil ity and long-term reproducibility of the sensor response before it can be considered for space-flight applications. In addition, the slow response of the sensors at room temperature is a major issue in some applications Figure 2-1 Scanning electron micrograph of ZnO multiple nanorods

PAGE 36

36 Figure 2-2 Schematic of contact geometry for multiple nanorod gas sensor (left) and a picture of packaged sensor (right). 051015202530 0 2 4 6 8 500ppm H2Air Time(min)| R|/R (%) Pt Pd Au Ag Ti Ni Figure 2-3 Time dependence of relative resist ance response of metal coated multiple ZnO nanorods as the gas ambient is switched from N2 to 500 ppm of H2 in air as time proceeds. There was no response to O2.

PAGE 37

37 -0.4-0.20.00.20.4 -1.0 -0.5 0.0 0.5 1.0 Current(mA)Voltage(V)Pt coated ZnO nanowires 15 min exposed to 500ppm H2 air Figure 2-4 IV characteristic of Pt-coated nanowires in air an d after 15 min in 500 ppm of H2 in air. 051015202530 0 2 4 6 8 10 Air 10~500 ppm H2Pt-ZnO nanowires 500ppm 250ppm 100ppm 10ppmTime(min)| R|/R (%) Figure 2-5 Time dependence of resistance change of Pt-coated multiple ZnO nanorods as the gas ambient is switched from N2 to various concentrations of H2 in air (10 ppm) and then back to N2.

PAGE 38

38 2.3 Hydrogen Sensing Using Pd-Coat ed ZnO Multiple Nanowires The device fabrication details are as described in section 2.2. Figure 2-6 shows the time dependence of resistance of either Pd-coated or uncoated multiple ZnO nanorods as the gas ambient is switched from N2 to various concentrations of H2 in air (10 ppm) as time proceeds. There are several aspects of the data. Fi rst, there is a strong in crease (approximately a factor of 5) in the response of the Pd-coate d nanorods to hydrogen relative to the uncoated devices. The addition of the Pd appears to be effective in catalytic dissociation of the H2 to atomic hydrogen. Second, there was no response of either type of nanorod to the presence of O2 in the ambient at room temperature. Third, th e effective conductivity of the Pd-coated nanorods is higher due to the presence of the metal. Fourth the recovery of the initial resistance is rapid (<20 s) upon removal of the hydrogen from the ambi ent, while the nanorod resistance is still changing at least 15 min after th e introduction of the hydrogen. The reversible chemisorption of reactive gases at the surface of metal oxides su ch as ZnO can produce a large and reversible variation in the conductance of the material [ 93]. Fifth, the relative response of Pd-coated nanorods is a function of H2 concentration in N2. The Pd-coated nanrods detected hydrogen down to <10 ppm, with relative responses of >2.6 % at 10 ppm and >4.2 % at 500 ppm H2 in N2 after a 10 min exposure, as shown in Figure 2-7 By com parison, the uncoated devices showed relative resistance changes of ~0.25 % for 500 ppm H2 in N2 after a 10 min exposure, and the results were not consistent for lower concentrat ions. The gas-sensing mechanisms suggested in the past include the desorption of adsorbed surface hydrogen and grain boundaries in poly-ZnO [ 94], exchange of charges between adsorbed gas species and the ZnO surface leading to changes in depletion depth [ 89] and changes in surface or grain bound ary conduction by gas adsorption/desorption [ 95 ] The detection mechanism is stil l not f irmly established in these

PAGE 39

39 devices and needs further study. It should be remembered th at hydrogen introduces a shallow donor state in ZnO and this change in near-surface conductivity may also play a role. Figure 2-8 s hows the time dependence of relative resistance change of Pd-coated multiple ZnO nanorods as the gas am bient is switched from vacuum to N2, oxygen or various concentrations of H2 in air (10 ppm) and then back to ai r. These data confirm the absence of sensitivity to O2. The resistance change during the e xposure to hydrogen was slower in the beginning and the rate resistance change reached maximum at 1.5 min of the exposure time. This could be due to some of the Pd becoming covered with native oxide, which is removed by exposure to hydrogen. Since the available surfac e Pd for catalytic ch emical absorption of hydrogen increased after the removal of oxide, the rate of resist ance change increased. However, the Pd surface gradually saturated with the hydroge n and the rate of resist ance change decreased. When the gas ambient switched from hydrogen to air, the oxygen reacte d with hydrogen right away, with the resistance of the nanorods changed back to the original value instantly. Moreover, the data were recorded at a power level of ~ 0.4 mW, which is low even in comparison with CNTs [ 48, 49 ]. This is attractive for long-te rm hydrogen sensing applications. Figure 2-9 shows the Arrhenius plot of nanorod resistance ch ange rate. T he rate of resistance change for the nanorods exposed to the 500 ppm H2 in N2 was measured at different temperatures. An activation energy of 12 KJ/mole was calculated from the slope of the Arrhenius plot. This value is larger than that of a t ypical diffusion process. Therefore, the dominant mechanism for this sensing process is more likel y to be the chemisorption of hydrogen on the Pd surface. In conclusion, Pd-coated ZnO nanorods a ppear well suited to detection of ppm concentrations of hydrogen at room temperatur e. The recovery charac teristics are fast upon

PAGE 40

40 removal of hydrogen from the ambient. The Zn O nanorods can be placed on cheap transparent substrates such as glass, making them attractive for low-cost sensing appl ications and operate at very low power conditions. 0306090120150 640 650 660 670 950 960 H2H2H2H2Air Air Air Air500ppm 250ppm 100ppm 10ppmO2N2 Pd-coated ZnO nanorods ZnO nanorods Resistance(ohm)Time(min) Figure 2-6 Time dependence of resistance of either Pd-coated or uncoated multiple ZnO nanorods as the gas ambient is switched from N2 to various concentrations of H2 in air (10 ppm) as time proceeds. There was no response to O2.

PAGE 41

41 0306090120150 -0.04 -0.03 -0.02 -0.01 0.00 0.01 H2H2H2H2 ZnO nanorod with PdAir Air Air Air O2500ppm 250ppm 100ppm 10ppmN2R/R (Sensitivity) Time(min) Figure 2-7 Relative response of Pd-c oated nanorods as a function of H2 concentration in N2. 051015202530 0 1 2 3 4 5 Air 10 ~ 500 ppm H2|R|/R (%) Time(min)Pd-ZnO nanorods 500 ppm 250 ppm 100 ppm 10 ppm Figure 2-8 Time dependence of relative resistance of Pd-coated multiple ZnO nanorods as the gas ambient is switched from N2 to oxygen or various concentrations of H2 in air (10 500 ppm) and then back to N2.

PAGE 42

42 0.00200.00250.00300.0035 e0e1e2e3e4 adsorption curve Arrhenius fitting slope= -1420.00457 activation energy (E)= 11.805 kJ/molroom T 50C 100C 150C 200C1/T (K-1) |dR|/dt Figure 2-9 Arrhenius plot of rate of resi stance change after exposure to 500 ppm H2 in N2.

PAGE 43

43 CHAPTER 3 HYDROGEN SENSOR USING AL GAN/GAN SCHOTTK Y DIODE AND HIGH E LECTRON MOBILITY TRANSISTOR 3.1 Background There is great current interest in detection of hydrogen sensors for use in hydrogen-fueled automobiles and with proton-exchange membrane (PEM) and solid oxide fuel cells for space craft and other long-term sensing applications. These sensors are required to selectively detect hydrogen near room temperature with minimal po wer consumption and weight and with a low rate of false alarms. Due to their low in trinsic carrier concentrations, wide bandgap semiconductor sensors based on GaN or SiC can be operated at lower current levels than conventional Si-based devices and offe r the capability of detection to ~600 C [ 8 96117]. The ability of ele ctronic devices fabricated in thes e materials to function in high temperature, high power and high flux/energy radiation conditions en able performance enhancements in a wide variety of spacecraft, satellite, homeland defense, mining, automobile, nuclear power, and radar applications. AlGaN/GaN high electron mobility transistor s (HEMTs) show promising performance for use in broad-band power amplifiers in base sta tion applications due to the high sheet carrier concentration, electron mobility in the two di mensional electron gas (2DEG) channel and high saturation velocity. The high electron sheet carrie r concentration of nitr ide HEMTs is induced by piezoelectric polarization of the strained AlGaN la yer and spontaneous pola rization is very large in wurtzite III-nitrides. This provides an increased sensitivity relative to simple Schottky diodes fabricated on GaN layers [ 8 99117] An additional attractive attribute of AlGaN/ GaN diodes is the fact that gas sensors base d on this m aterial could be in tegrated with high-temperature electronic devices on the same chip. The advant ages of GaN over SiC for sensing include the presence of the polarization-induced charge, the availability of a heterostructure and the more

PAGE 44

44 rapid pace of device technology development for GaN which borrows from the commercialized light-emitting diode and laser diode businesses. Section 3.2 discusses a comparison of two modes of operation for the detection of hydrogen with AlGaN/GaN HEMTs, namely through monitoring changes in either the drainsource current at different gate biases or in the gate current at zero drain-source bias. These correspond to a comparison of Schottky diode versus field effect transistor (FET) operation. The FET mode of operation provides much higher cu rrent changes but the diode mode shows a higher relative sensitivity over a limited range of forward biases. Section 3.3 reports on the use of a differential pair of AlGaN/GaN HEMT diode s for hydrogen sensing near room temperature. This configuration provides a built-in control diod e to reduce false alarms due to temperature swings or voltage transients We demonstrate fast res ponse of the diodes to 1 % H2 in air. Section 3.3 shows that use of Ti/Al/TiB2/Ti/Au Ohmic contacts on AlGaN/GaN HEMT diodes produces less noise in the gate cu rrent of the sensor at fixed fo rward bias voltage compared to conventional Ti/Al/Ni/Au contacts. This is attr active for reducing false alarms and reducing the ultimate detection threshold of the sensors. 3.2 Comparison of Gate and Drain Current Detection of Hydrogen at Room Temperature with AlGaN/GaN High Electron Mobility Transistor The HEMT layer struct ures were grown on c-plane Al2O3 substrates by Metal Organic Chemical Vapor Deposition (MOCVD). The layer structure included an initial 2m thick undoped GaN buffer followed by a 35 nm thick unintentionally doped Al0.28Ga0.72N layer. The sheet carrier concentration was ~11013 cm-2 with a mobility of 980 cm2/V s at room temperature. Mesa isolation was achieved by usin g an inductively coupled plasma system with Ar/Cl2 based discharges. The Ohmic contacts were formed by lif t-off of e-beam deposited Ti

PAGE 45

45 (200 )/Al (800 )/Pt (400 )/Au (800 ). The contacts were annealed at 850 C for 45 sec under a flowing N2 ambient in a Heatpulse 610T system. A 200 thick circular Pt Schottky contact was deposited for the gate metal. The fi nal step was deposition of e-beam evaporated Ti/Au (200 /2000 ) interconnection contacts. The gate dimension of the device was 1 m2. The devices were bonded to electrical feed-through and exposed to either pure N2 or 500 ppm H2 in N2 ambient in an environmental chamber in which the gases were introduced through electronic mass flow controllers. Figure 3-1 shows the HEMT drainsource current-voltage (IDS-VDS) characteristics (top) and the transfer characteristics (bottom) at 25 C measured in both the pure N2 and the 500 ppm H2 in N2 ambients. The increase in drain current at each applied gate voltage current is consistent with the hydrogen molecules dissociating into at oms through the catalytic action of the Pt gate contact and diffusing to the P t/AlGaN interface wher e it screens some of the piezo-induced channel charge [ 118]. Pr evious measurements have shown an effective decrease in the effective barrier height of Pt on GaN by 30~60 m eV by introduction of hydrogen into a N2 ambient [ 103]. This data represents the FET mode of operati on f or hydrogen gas sensing. The transconductance of the HEMT also increases slightly when meas ured in the hydrogen-containing ambient due to the increase in effective channel ch arge, as shown at the bottom of Figure 3-1 Figure 3-2 s hows the gate I-V characteristics at 0 V IDS measured at 25 C under pure N2 ambient or in a 500 ppm H2 in N2 ambient. Both the forward and reverse currents increase due to the reduction in barrier height. This data represents the Schottky diode mode of operation for hydrogen gas sensing. Figure 3-3 shows the change in drainsource o r gate currents as a function of gate voltage (top) and percentage change in these cu rrents (bottom) for measurement under pure N2 ambient

PAGE 46

46 or in a 500 ppm H2 in N2 ambient. The FET mode of operation (monitoring of change in drainsource current) shows a much larger signal over a broad range of ga te voltage. By sharp contrast, the diode mode of detection shows a large relativ e change in current onl y at high forward gate biases. This shows the advantage of using the 3-terminal device structure, with its attendant current gain. The percentage cha nge in both drain-source and gate currents when the hydrogen is introduced into the measurement ambient are shown at the bottom of Figure 3-3 The relative change can be m u ch larger in the diode mode of operation at small forward bias (~1 V) due to the lower baseline current. At higher forward bias the effects of series resistance dominate the current both in N2 and H2-containing ambients. Figure 3-4 s hows some of the recovery char acteristics of the HEMTs upon m ultiple cycling of the ambient from N2 to 500 ppm H2 in N2 and back again .The sensors show good recyclability and recovery in both modes of operation. Once again the change in drain-source current is much larger in the FET mode. The in itial response to hydrogen in both cases is rapid (<5 sec), while the recovery back to the N2 ambient value takes much longer (100-200 secs) because of the mass transport characteristics of ga s in our test chamber. From the fast initial response of the sensors in both modes, the eff ective diffusivity of the atomic hydrogen through the Pt is >410-13 cm2/V s at 25 C. This is only an estimate, since the response time includes the gas flow dynamics of the gas into the test ch amber, the dissociation of the molecular hydrogen and the diffusion to the Pt/AlGaN interface of the atomic hydrogen. In conclusion, Pt/AlGaN/GaN HE MTs operated in either a diode mode or in an FET mode show the ability to detect 500 ppm H2 in N2 at room temperature. The FET mode provides a larger total current change with introduction of hydrogen into the ambient, but the diode mode shows a higher relative se nsitivity over a limited range of forward biases.

PAGE 47

47 Figure 3-1 IDS-VDS characteristics (top) and transfer charac teristics (bottom) of Pt-gated HEMT measured at 25 C under pure N2 ambient or in a 500 ppm H2 in N2 ambient.

PAGE 48

48 Figure 3-2 Gate I-V characteristics at 0 V IDS measured at 25 C under pure N2 ambient or in a 500 ppm H2 in N2 ambient.

PAGE 49

49 Figure 3-3 Change in drain-source or gate curre nts as a function of ga te voltage (top), and percentage changes in these currents (bottom) for measurement under pure N2 ambient or in a 500 ppm H2 in N2 ambient.

PAGE 50

50 Figure 3-4 Time dependence of drain-source (top) or gate current (bottom) when switching from pure N2 ambient to a 500 ppm H2 in N2 ambient and back again.

PAGE 51

51 3.3 Robust Detection of Hydrogen Using Differential AlGaN/GaN High Electron Mobility Sensing Diode A maskset was designed for fabricating differe ntial diodes to eliminate the temperature effect on the diode characteristics. Schottky cont acts of 100 Pt for the active diode and Ti (200 )/Au (1200 ) for the reference diodes were deposited by e-beam evaporation. Details of device fabrication are as de scribed in Section 3.2. Figure 3-5 s hows an optical microscope image of the com pleted devices. The devices were bonded to an electrical feed-through and exposed to a 1 % H2 ambient in an environmental chamber. Figure 3-6 s hows the absolute and differential forw ard curren t-voltage (I-V) characteristics at 25 C of the HEMT active (t op) and reference (bottom) diodes, both in air and in a 1 % H2 in air atmosphere. For the active diode, the cu rrent increases upon in troduction of the H2, through a lowering of the effective barrier height. The H2 catalytically decomposes on the Pt metallization and diffuses rapidly to the interface where it forms a dipole layer [ 117]. The differential change in forward current upon introduction of the hydroge n into the am bient is ~1-4 mA over the voltage rang e examined and peaks at low bias. This is roughly double the detection sensitivity of comparable GaN Schottky gas sensors tested und er the same conditions, confirming that the HEMT based diode has advantages for applicatio ns requiring the ability to detect hydrogen even at room temperature. As the detection temperature is increased to 50 C, the differential cu rrent response of the HEMT diode pair was almost constant over a wide range of voltages due to more efficient cracking of the hydrogen on the metal contact, as shown in Figure 3-7 The maximum dif ferential current is similar to that at 25 C, but the voltage control to achieve maximum detection sensitivity for hydrogen is not as important at 50 C.

PAGE 52

52 To test the time response of the HEMT diode sensors, the 1 % H2 ambient was switched into the chamber through a mass flow controller for 200 seconds and then switched back to air. Figure 3-8 shows the tim e dependence of forward curr ent for the active and reference diodes at a fixed bias of 2.5 V under these conditions. The re sponse of the sensor is rapid (<1 sec based on a series of switching tests). Upon switching out of the hydrogencontaining ambient, the forward current decays exponentially back to its initial value. This time constant is determined by the volume of the test chamber and the flow rate of the input gases and is not limited by the response of the HEMT diode itself. Note that the use of the differential pair geometry removes false alarms due to changes in ambient temperature or voltage drifts. In conclusion, AlGaN/GaN HEMT differentia l sensing diodes appear well-suited to hydrogen detection applications and suggest that integrated chips i nvolving gas sensors and HEMT-based circuitry for off-chip communicat ion are feasible in the AlGaN/GaN system. Future work will involve design and fabrication of an integrated sensor chip with GaN HEMT amplifier and transmitter. Figure 3-5 Microscopic images of differential se nsing diodes. The opening of the active diode was deposited with 10 nm of Pt, and the re ference diode was deposited with Ti/Au.

PAGE 53

53 Figure 3-6 Absolute (a) and differential (b) current of HEMT diode measured at 25 C.

PAGE 54

54 Figure 3-7 Absolute (a) and differential (b ) current of HEMT diode measured at 50 C.

PAGE 55

55 Figure 3-8 Time dependent test of differential HEMT diodes at 25 and 50 C. 3.4 Stable Hydrogen Sensors from AlGaN /GaN Heterostructure Diodes with TiB2-Based Ohmic Contacts We compared two types of Ohmic contacts formed by sputter deposition and lift-off, i.e. Ti (200 ) /Al (1000 ) /Pt (600 ) /Au ( 800 ) or Ti (200 ) /Al (1000 ) /TiB2 (400 ) /Ti (200 ) /Au (800 ). All of the metals were deposited by Ar plasma-assisted rf s puttering at pressures of 15-40 mTorr and rf (13.56 MHz) powers of 200-250 W. Other deta ils of device fabrication are as described in Section 3.2. Figure 3-9 shows a schem a tic of a completed device. The device was bonded to an electrical feed-through and exposed to a 1 % H2 ambient in an environmental chamber. Figure 3-10 shows the linear (top) and log scale (bottom ) forward current-voltage (I-V) characteristics at 25 C of the HEMT active (ie. Pt-gate) diod e, both in air and in a 1 % H2 in air atmosphere. For these diodes, the current increases upon intr oduction of the H2, through a

PAGE 56

56 lowering of the effective barrier height. The H2 catalytically decomposes on the Pt metallization and diffuses rapidly to the interface where it forms a dipole layer [ 117]. The differential change in forward current upon introduction of the hydrogen into the am bient is ~1 m A over the voltage range examined. By sharp contrast, the passive diodes with Ti/Au gates showed no difference in current when measured in H2-containing ambients. Figure 3-11 shows the tim e dependence of forward current for the active diodes at a fixed bias of 1.5 V as the 1 % H2 ambient was switched into th e chamber through a mass flow controller for 200 seconds and then switched back to air. The response of the sensor is rapid (<1 sec based on a series of switching tests). The deca y time of the forward current back to its initial value is a function of the volume of the test chamber and the flow rate of the input gases and is not limited by the response of the HEMT diode its elf. As we have previously noted, the use of the differential pair geometry removes false alar ms due to changes in ambient temperature or voltage drifts [ 119]. The TiB2-based Ohmic contact devices showed much more stable forward currents at fixed bias than their conve ntional counterparts. Figure 3-12 shows the tim e dependence of forward current at 1.5 V gate bias for devices with both ty pes of Ohmic contacts. These tests were carried out in the field, where temperat ure and humidity were not controlled. There are several features of note. First, the current is much higher in the diodes with TiB2based contacts because of their lower contact resistance (1.6-6 cm2 vs 7.5-6 cm2 for the conventional Ti/Al/Pt/Au). Second, there is much better stab ility of the devices with TiB2based contacts. There is much less temperature dependence to the contact resistance of the boride contacts and this translates to less variation in gate current as the temperature cycles from day to night. We know from previous results that the TiB2 is an effective diffusion barrier and prevents degradation of the

PAGE 57

57 contact morphology [ 120].W e expect that the AlGaN/Ti inte rface is therefore m o re uniform with the TiB2 overlayer and this translates to less noise in the current at fixed voltage. Note that this leads to a lower threshol d for hydrogen detection. In conclusion, Pt-gated AlGaN/GaN HEMT di odes show greatly improved current stability under field conditions with use of Ti/Al/TiB2/Ti/Au contacts replacing the more conventional Ti/Al/Pt/Au. Combined with the superior thermal stability of these boride-based contacts, this metallization system appears attractive for se nsors for long-term monitoring applications. Figure 3-9 Schematic of HEMT diode hydrogen sensor using either conventional or TiB2-based Ohmic contacts.

PAGE 58

58 Figure 3-10 I-V characteristics in li near (a) or log (b) form of Pt-g ated diode measured in air or 1 % hydrogen ambient at 25 C.

PAGE 59

59 Figure 3-11 Time-dependence of current test biased by 1.5 V of Pt-gated di ode as the ambient is switched from air to 1 % hydrogen and back to air. Figure 3-12 Variation in forward current at fixed bias for di odes with boride-based Ohmic contacts (top) or conventional Ohmic contac ts (bottom) as a function of time under field conditions where the temperature in creases during the day and decreases at night.

PAGE 60

60 CHAPTER 4 MERCURY ION SENSOR USING ALGA N/GAN HIGH E LEC TRON MOBILITY TRANSIST OR 4.1 Background The toxicity of heavy metal ions, including mercury(II) (Hg2+) lead(II) (Pb2+), copper(II) (Cu2+), and zinc(II) (Zn2+) has long been recognized as a chronic environmental problem [ 121125] In particular, m e rcury is released into the environment through a variety of courses including the combustion of fossil fuels, mining, volcanic emissions and solid waste incineration. Mercury has attracted a great d eal of attention around the worl d for its impact on wild life ecology and human health. Certain bact eria convert inorganic mercury Hg2+ into neuro-toxic organic-mercury compounds, which bio-accumulate through the plant, animals, and can food chain and affect the entire eco-system [ 126, 127 ]. It is highly desirable to de velop sensitive and selective analytical m e thods for the quantitative detection of Hg2+, which are applicable in a wide range of different sites and environments. Traditionally, there are several methods for heavy metal detection including spectroscopic (atomic absorption spectroscopy (AAS), Auger el ectron spectroscopy (AES), or inductively coupled plasma-Mass Spectrometry (I CP-MS)), or electrochemical (ion selective electrodes (ISE) or polarography), ho wever, these methods are either expensive or not useful for detection on-site, where hand-held portable devices could be inva luable for metal detections at low concentrations [ 128130]. To date, a num ber of selective Hg2+ ion sensors have been devised utilizing redox, chromogenic or fluorogenic changes. Most of these systems display shortcomings in practical use, such as interference from othe r metal ions, delayed response to Hg2+, and/or lack of water solubility [ 131134]. Therefore, development of fast response and inexpensive m ethods for detection of bioavailable heavy metal concen trations is highly desirable.

PAGE 61

61 GaN/AlGaN high electron mobility transistors (HEMTs) have also shown promise for gas and liquid sensor applications due to primarily two reasons: 1) they consist of a high electron sheet carrier concentration channel induced by both piezoelectric polariza tion of the strained AlGaN layer and the difference in spontaneous polarization between AlGaN and GaN. Unlike conventional semiconductor field effect transist ors, there is no inten tional dopant in the AlGaN/GaN HEMT structure. 2) the electrons in the two-di mensional electron gas (2DEG) channel are located at the lowe r interface between the AlGaN laye r and GaN layer. The electron carrier concentration in 2DEG strongly depends on the ambient [ 7 9 135139]. W e have recently exploited these properties to detect a variety of species in gases and liquids using appropriately functionalized AlGaN/GaN HEMTs [ 135139]. For these reasons, nitr ide HEMTs are versatile devices that m a y be used for a variety of sensing applications. Section 4.2 presents the detection of Hg2+ with sensors fabricated with Au-gated and thioglycolic acid functionalized Au-gated GaN/AlGaN HEMTs. We investigated a wide range of concentration from 10 M to 10 nM. The temporal resolution of the device was quantified, along with limit of detection selectivity over sodi um as well as magnesium and precision of measurements. Section 4.3 illustrates the detection of Hg2+ and Cu2+ ions with sensors fabricated with Au-gated and thioglycolic acid functionalized Au-gated Ga N/AlGaN HEMTs. We investigated a wide range of concentration from 10 M to 10 n M. The temporal resolution of the device was quantified, along with limit of detec tion and selectivity over sodium, magnesium and lead ions. The recyclability of the sensors between measurements was also explored.

PAGE 62

62 4.2 Fast Electrical Detection of Hg(II) Ions with AlGaN/GaN High Electron Mobility Transistors The HEMT structures consisted of a 2 m thick undoped GaN buffer and 250 thick undoped Al0.25Ga0.75N cap layer. The epi-layers were gr own by metal-organic chemical vapor deposition on 100 mm (111) Si substrates at Nitronex Corpor ation. Mesa isolation was performed with an Inductively Coupled Plasma (ICP) etching with Cl2/Ar based discharges at 90 V dc self-bias, ICP power of 300 W at 2 MHz and a process pressure of 5 mTorr. 50 m2 Ohmic contacts separated with gaps of 10, 20, and 50 m consisted of e-beam deposited Ti/Al/Pt/Au patterned by lift-off and a nnealed at 850 C, 45 sec under flowing N2 for source and drain metal contacts and 5-nm thin gold film was deposited as gate metal to functionalize a selfassembled monolayer of thioglycolic acid. 500 nm-thick polymethyl methacrylate (PMMA) was used to encapsulate the source/drain regions, with only the gate region open to allow the liquid solutions to cross the surface by e-beam lit hography. The source-drain current-voltage characteristics were measured at 25 C using an Agilent 4156C parameter analyzer with the Augated region exposed to diffe rent concentrations of Hg2+, Mg2+ or Na+ solutions. Ac measurements were performed to prevent side electrochemical reactions with modulated 500-mV bias at 11 Hz. A schematic cross-sectio n of the device with Hg2+ ions bound to thioglycolic acid functionalized on the gold gate region and plan view photomicrograph of a completed device is shown in Figure 4-1 Th e thiog lycolic acid, HSCH2COOH, is an organic compound and contains both a thiol (mercaptan) and a car boxylic acid functional group. A self assembled monolayer of thioglycolic acid molecule was adsorbed onto the gold gate due to strong interaction between gold and the thiol-group. The extra thioglycolic acid molecules were rinsed off with de-ionized (DI) water. An increase in th e hydrophilicity of the treated surface by thioglycolic acid

PAGE 63

63 functionalization was confirmed by contact angle measurement ( Figure 4-2 ) which sho w ed a change in contact angle from 58.4 to 16.2 after the surface treatment. X-ray Photoelectron Spectroscopy (XPS) and electrical measurements confirming a high surface coverage and Au-S bonding formation on the GaN surface and the re sults have been previously published [ 139 ]. Unlike conventional semiconductor field effect tr ansistors, there is no intentional dopant in the AlGaN/GaN HEMT structure. The electrons in the two-di mensional electron gas (2DEG) channel of the AlGaN/GaN HEMT are induced by piezoelectric and s pontaneous polarization effects. This 2DEG is located at the interf ace between the GaN layer and AlGaN layer. There are positive counter charges at the AlGaN su rface layer induced by the 2DEG. Any slight changes in the ambient of the AlGaN/GaN HEMT affect the surface charges of the AlGaN/GaN HEMT. These changes in the surface charge are tran sduced into a change in the concentration of the 2DEG in the AlGaN/GaN HEMTs. Based on this principle, we have de monstrated the use of appropriately functionalized AlGa N/GaN HEMTs as mercury ion (Hg2+) sensors As shown in Figure 4-3(a) the drain current of both sensor s further reduc ed after exp osure to different concentrations of Hg2+ ion solutions. Being exposed to 10-5 M Hg2+, the drain current reduced ~55 % for the thioglycolic acid func tionalized AlGaN/GaN HEMT sensors and bare-Augate sensor had less than ~8 % changes of the dr ain current. The mechanism of the drain current reduction for bare Au gate and thioglycolic acid functionalized AlGaN/GaN HEMT sensors was quite different. For the bare Au-gate devices, Au-mercury amalgam formed on the surface of the bare Au-gates when the Au-gate electrode exposed to Hg2+ ion solution. The formation rate of the Au-mercury amalgam depended on the soluti on temperature and the concentration of the Hg2+ ion solution. Figure 4-3(a) also shows the time dependence of the drain current for the two types of sensors. For the higher Hg2+ ion concentration solution, 10-5 M, the bare Au-gate based

PAGE 64

64 sensor took less than 15 seconds for the drain to reach steady state. However, the drain current required 30-55 seconds to reach steady state, when the sensor was exposed to the less concentrated Hg2+ ion solutions. A less than 5 second response time was obtained for the thioglycolic acid functionalized AlGaN/GaN HEMT sensors, when th e sensor was exposed to the 10-5 M of the Hg2+ ion solution. This is the shortest response time of Hg2+ ion detection ever reported. For the thioglycolic acid functionalized AlGaN/GaN HEMT, the thioglyco lic acid molecules on the Au surface align vertically with carboxylic acid func tional group toward the solution [ 140]. The carbox ylic acid functional g roup of the adjacent thioglycolic ac id molecules form chelates of R-COO-(Hg2+)-OOC-R with Hg2+ ion, when the sensors are exposed to the Hg2+ ion solution. The charges of trapped Hg2+ ion in the R-COO-(Hg2+)-OOC-R chelates changed the polarity of the thioglycolic acid molecules, which were bon ded to the Au-gate through -S-Au bonds. This is why the drain current changes in response to me rcury ions. Similar surface functionalization was used by Chang et. al. and the fluorescence was use for the detections [ 141]. The di fference of drain current for the devi ce exposed to different Hg2+ ion concentration to the DI water is illustrated in Figure 4-3(b). The Hg2+ ion concentration detection limit for the thioglycolic acid functionalized sensor is 10-7 M, which is approximately equivale nt to 27 ppb (parts per billion). The thioglycolic acid functionalized sensor also showed excellent sens ing selectivity (over 100 times higher selectivity) over Na+ and Mg2+ ions, as illustrated in Figure 4-4 Since our sensor chip is very comp act (1 mm 5 mm) and opera tes at extremely low power (8 W based on 0.5 V of drain voltage and 80 A of drain current operated at 11 Hz), it can be integrated with a commercial availabl e hand-held wireless transmitter to realize a portable, fast response and high sensitivity Hg2+ ion detector.

PAGE 65

65 In summary, we have demonstrated Al GaN/GaN HEMT to be an excellent Hg2+ ion sensor through a chemical modification on the Au-gate surface. The thioglycolic acid functionalized Au-gate based sensor showed good sensitivity an d shortest response time ever reported. The sensor also showed excellent detection selectivity over Na+ and Mg2+ ions. Figure 4-1 (a) A schematic of AlGaN/GaN HEMT. The Au-coated gate area was functionalized with thioglycolic acid. (b) Plan view phot omicrograph of a completed device with a 5 nm Au film in the gate region. Figure 4-2 Photographs of contact angle of water drop on the su rface of bare Au (left) and thioglycolic acid functio nalized Au (right).

PAGE 66

66 Figure 4-3 (a) Time dependent response of the drain current for bare Au-gate AlGAN/GaN HEMT sensor and thioglycolic acid functi onalized Au-gate HEMT sensor. (b) Drain current of a thioglycolic aci d functionalized Au-gate HEMT sensor as a function of the Hg2+ ion concentration.

PAGE 67

67 Figure 4-4 Time dependent response of the drain current for detecting Na+, Mg2+ or Hg2+ with a thioglycolic acid functionaliz ed Au-gate HEMT sensor. 4.3 Selective Detection of Hg(I I) Ions from Cu(II) and Pb(II) Using AlGaN/GaN High Electron Mobility Transistors The HEMT structures consisted of a 2 m thick undoped GaN buffer and 250 thick undoped Al0.25Ga0.75N cap layer. The epi-layers were grown by molecular beam epitaxy system on 2 sapphire substrates at SVT Associates. Details of the device fabrication are as described in Section 4.2. 5-nm thin gold film was deposited as gate metal for two set of samples. One was for the bare Au-gate sensor and the other was for functionalizing a self-ass embled monolayer of thioglycolic acid on the Au-gate. An increase in the hydrophilicity of the treated surface by thioglycolic acid functionaliza tion was confirmed by contact angle measurement as well. A schematic cross-section of the device with Hg2+ ions bound to thioglycolic acid functionalized on the gold gate region is as shown in Figure 4-1(a) The source-drain current-voltage characteristics were m easured at 25 C using an Agilent 4156C parameter analyzer with the Au-

PAGE 68

68 gated region exposed to di fferent concentrations of Hg2+, Cu2+, Pb2+, Mg2+ or Na+ solutions. AC measurements were performed to prevent side electrochemical reactions with modulated 500 mV bias at 11 Hz. Figure 4-5 s hows the change in drain current of a bare Au-gated AlGaN/GaN HEMT sensor and a thioglycolic acid functionaliz ed AlGaN/GaN HEMT sensor exposed to 10-5 M Hg2+ ion solutions as compared to exposed to DI wa ter. The drain current of both sensors decreased after exposure to Hg2+ ion solutions. The drain current reduction of the thioglycolic acid functinalized AlGaN/GaN HEMT sensors was almost 80 % more than that of the bare Au-gate sensor. The mechanism of the drain current redu ction for bare Au-gate and thioglycolic acid functionalized AlGaN/GaN HEMT se nsors is probably quite differe nt. For the thioglycolic acid fictionalized AlGaN/GaN HEMT, the thioglycolic acid molecules on the Au surface align vertically with carboxylic acid func tional group toward the solution [ 140]. The carbox ylic acid functional g roup of the adjacent thioglyc olic acid molecules probably forms ch elates (R-COO-(Hg2+)-OOC-R) with the Hg2+ ions. If the chelates are indeed forming, one would expect the charges of trapped Hg2+ ion in the R-COO-(Hg2+)-OOC-R to change the polarity of the thioglycolic acid molecules. Th is is probably why the drain cu rrent changes in response to mercury ions. A similar type of surface functi onalization was used by Chang et. al. and the detection performed with goldnanoparticle-based fluorescence[ 141] but the detection time is longer than the nitride H EMT based sensor. Because Hg2+ ions were used in our experiments, we do not expect an Au-mercury amalgam to form on the bare Au-surface. The detailed mechanism for mercury ion induced reducti on in drain current of the Au-gate device is not clear and currently under further investigation.

PAGE 69

69 Figure 4-6 shows tim e dependence of the drain curre nt for the two types of sensors for detecting Hg2+, Cu2+, and Pb2+ ions. Both type of sensors show ed very short response time (less than 5 seconds), when exposed to Hg2+ ion solution. The limits of detection for Hg2+ ion detection for the bare Au-gate and thioglyc olic acid functionalized sensor were 10-6 and 10-7 M, respectively. Neither sensor could detect Pb2+ ions. For the Cu2+ ions, the detection limit of the thioglycolic acid functiona lized sensor was around 10-7 M. However, the bare Au-gate could not detect the Cu2+ ions as shown in Figure 4-6 Figure 4-7 show s the drain current changes in response to Hg2+and Cu2+ ions as a function of the ion c oncentration for the two different surfaces. The difference in the response between the bare Au-gate and the thioglycolic acid functionalized sensor offers the possi bility for selective detection for Hg2+ and Cu2+ ions presented in a single solution with a sensor chip containing both type of sensors, as shown in Figure 4-8 The dim e nsion of the active area of the AlGaN/GaN HEMT sensor is less than 50 m 50 m, and the sensors can be fabricated as an array of individual sensors. The fabrication of both sensors is identical except for the thiogl ycolic acid functionalized sensor, which has an additional functionalization step. This step can be accomplished with micro-inkjet system to locally functionalize surfaces. The ba re Au-gate and thioglycolic ac id functionalized sensors also showed excellent sensing se lectivity (over 100 times hi gher selectivity) over Na+ and Mg2+ ions. As illustrated in Figure 4-9 there was alm ost no detection of Na+ and Mg2+ ions for both types of sensors with 0.1 M concentrations. Most semiconductor based chemical sensors are not reusable. The bare Au-gate and thioglycolic acid functionaliz ed sensors showed very good recyclability, as shown in Figure 410. After a sim p le rinse with DI water, the sensors can be reused for Hg2+ ion detection repeatedly and the responses to different ioni c solutions remain uncha nged. The stability of

PAGE 70

70 thioglycolic acid functionalized Au surface is aff ected by several factors, like oxygen level, light, initial packing quality, chain leng th, and terminal functional group [ 142, 143]. Our devices has been stored in nitrog en ambient and repeated ly used over a couple of weeks. The long term stability of the thioglycolic acid functi onalized Au surface is under investigation. The current sensor operates at 0.5 V of drain voltage and 2 mA of drain current. However, the operation voltage and device size can be furt her reduced to minimize the power consumption to W range. The sensor can be integrated with a commercial available hand-held wireless transmitter to realize a portable, fast response and high sensitivity Hg2+ and Cu2+ ion detector. In summary, we have demonstrated bared Au-gate and thioglycolic acid functionalized AlGaN/GaN HEMT sensors to heavy ion detections. The bare Au-gate sensor was sensitive to Hg2+ and thioglycolic acid functiona lized sensors could detect both Hg2+ and Cu2+ ions. By fabricating an array of the sensors on a single chip and selectively functionalizing some sensors with thioglycolic acid, a multi-functional specific detector can be fabricated. Such a sensor array can be used to de tect quantitatively Hg2+ ions in Cu2+ ion solution or Cu2+ ions in Hg2+ ion solution. Both bare Au-gate and thioglycolic acid functionalized sensor can be repeatedly used after a simple DI water rinse.

PAGE 71

71 Figure 4-5 Changes in HEMT drain-source cu rrent for bare Au-gate and Au-gate with thioglycolic acid functio nalization exposed to 10-5 M Hg2+ ion solutions. Figure 4-6 (a) Time dependent response of the drain current as a function of Hg2+, Cu2+, Pb2+ ion concentrations for bare Au-gate AlGaN/ GaN HEMT sensor. (b) Time dependent response of the drain curr ent as a function of Hg2+, Cu2+, Pb2+ ion concentrations for thioglycolic acid functionalized Au-gate AlGaN/GaN HEMT sensor.

PAGE 72

72 Figure 4-7 Drain current ch anges in response to Hg2+ and Cu2+ ions as a function of the ion concentration for (a) the bare Au-gate and (b) the thioglycolic acid functionalized Augate AlGaN/GaN HEMT sensor. Figure 4-8 Plan view photograph of a multiple cell AlGaN/GaN HEMT sensors.

PAGE 73

73 Figure 4-9 Time dependent change in the drain current in response to Na+ and Mg2+ with bare Au-gate and thioglycolic acid func tionalized Au-gate HEMT sensor. Figure 4-10 Recyclability for (a) the bare Au-gat e, and (b) the thioglyc olic acid functionalized Au-gate surface.

PAGE 74

74 CHAPTER 5 DISEASE BIOMARKER SENSOR USING AL GAN/GAN HIGH ELE CT RON MOBILITY TRANSISTOR 5.1 Background Disease diagnosis by detecting specific biom arkers (functional or structural abnormal enzyme, low molecular weight proteins, or antigen ) in blood, urine, saliva, or tissue samples has been established into several approaches incl uding enzyme-linked immunsorbent assay (ELISA), particle-based flow cytometric assays, electr ochemical measurements based on impedance and capacitance, electrical measur ement of microcant ilever resonant frequency change, and conductance measurement of semiconductor nano stuctures. ELISA possess a major limitation that only one analyte will be measured at one time [ 144, 145]. Particle-based assay opens a spotligh t for multiple detections by using multip le beads but the whole detection process over than 2 hours is not practical to bedside detecting [ 146]. Electrochem ical devices h ave attracted attention due to their low cost and simplicity, but significant improvements in their sensitivities are still needed for use with clinical samples [ 147 148]. Microcantilever cap able f o r detecting concentration as low as 10 pg/ml, unfortunately, suffers from an undesira ble resonant frequency change due to viscosity of the medium and can tilever damping in the solution environment [ 149, 150]. Nanom a terial devices so far have provided the best option toward fast, label-free, sensitive, selective, and multiple detections for both preclin ical and clinical applications. Examples of electrical measurements of semiconductor devices include carbon nanotubes for lupus erythematosus antigen detection [ 151], com pound semiconducting nanowires, In2O3 nanowires, for prostate-specific antigen detection[ 152], and silicon nanowire array detecting prostatespecific antigen, carcinoem bryonic antigen, and mu cin-1 in serum for diagnosis of prostate cancer [ 153156]. Recently, AlGaN/GaN high electron m obility trans i stors (HEMTs) have

PAGE 75

75 shown promise for such applications due to a hi gh electron sheet carrie r concentration channel induced by both piezoelectric polariza tion and spontaneous polarization [ 7 8, 135139, 157]. Acute Kidney Injury (AKI) or Acute Renal Failu re (ARF) is one of the most comm on and serious m edical complications that is cl osely associated with high mortality [158160] Despite the im provements in dialysis and kidney transp lantation techniques over the past two decades, the high mortality rate has remained. The AKI diagnosis by detecting the urinary biomarker, kidney injury molecule-1 or KIM-1 (a specific sensitive AKI biomarker) [ 161], have been proved with enzym e -linked immuns orbent assay (ELISA) technology [ 145]. Most state-of-art testing m e thods for kidney injury disease biomarkers have limitations du e to the laboratoryoriented nature of the measurements requiring sample transportation, time consuming analysis and high cost of detection. In this chapter, we report the detection of KIM-1 with KIM-1 antibody functionalized Au-gat ed GaN/AlGaN HEMTs (Section 5.2). We quantified the sensitivity of the HEMT sensor and the temporal resolution, along with the limit of detection (LOD) and selectivity. 5.2 Kidney Injury Molecule-1 Detection Us ing AlGaN/GaN High Electron Mobility Transistors The HEMT structures consisted of a 2 m thick undoped GaN buffer and 250 thick undoped Al0.25Ga0.75N cap layer. The epi-layers were gr own by metal-organic chemical vapor deposition on 100 mm (111) Si substrates. Mesa isolation wa s performed with Inductively Coupled Plasma (ICP) etching with Cl2/Ar based discharges at V dc self-bias, ICP power of 300 W at 2 MHz and a process pressure of 5 mTorr. 50 50 m2 Ohmic contacts separated with gaps of 20 m consisted of e-beam deposited Ti/Al/Pt/Au patterned by lift-off and annealed at 850 C for 45 sec under flowing N2. 5 nm thin gold film was deposited as gate metal to

PAGE 76

76 functionalize a self-assembled monolayer of thioglycolic acid. 500-nm-thick polymethyl methacrylate (PMMA) was used to encapsulate the source/drain regions, with only the gate region opened using e-beam lithography. A plan vi ew photomicrograph of a completed device is shown in Figure 5-1(a). Before thioglycolic acid coating, the sam ple was exposed to UV ozone for 5 m ins to clean surface contamination. The thioglycolic acid, HSCH2COOH, is an organic compound and contains functional groups of a thiol (mercaptan) and a carboxylic acid functional group. A selfassembled monolayer of thioglycolic acid molecu le was adsorbed onto the Au-gate due to strong interaction between gold and the thiol-group. The extra thioglyco lic acid molecules were rinsed off with de-ionized water. An increase in the hyd rophilicity of the treated surface by thioglycolic acid functionalization was confirmed by contact angle measurements which showed a change in contact angle from 58.4 to 16.2 after the su rface treatment. The sample was treated with monoclonal anti rat KIM-1 antibody in a soluti on of 10 mM phosphate-buffered saline (PBS) buffer solution containing 4 mM sodium cyanoborohydride, pH 8.8 at room temperature for 2 hours. This antibody immobilization is base d on a strong reaction between carboxyl group on thioglycolic acid and amine gr oup on KIM-1 antibody. Excess KIM1 antibodies were washed off by PBS buffer and the unreacted surface ca rboxyl groups were passivated by a blocking solution of 100 mM ethanolamine in 10 mM phosphate buffer pH 8.8. Figure 5-1(b) sh ows a schem atic device cross section with thioglyco lic acid followed by KIM-1 antibody coating. The source-drain current-voltage char acteristics were measured at 25 C using an Agilent 4156C parameter analyzer with the KIM-1 antibody functi onalized Au-gated region exposed to different concentrations of KIM-1/PBS buffer. AC meas urements were performed to prevent side electrochemical reactions with modulated 500 mV bias at 11 Hz.

PAGE 77

77 The electrical properties of the de vices, source and drain current (IDS) vs. voltage (VDS), were measured in PBS buffer and 100 ng/ml KIM-1 in PBS buffer, as shown in Figure 5-2 There is a clear conductance decrease with KIM-1 exposure and this suggests that through the selective binding of KIM-1 with antibody, there are char ges accum u lated at th e surface and these surface charges are transduced into a change in the carrier concentra tion of AlGaN/GaN 2DEG, leading to the obvious decreas e in the conductance of the device after KIM-1 exposure. Figure 5-3 shows the tim e dependent source-drain current signal with constant bias of 500 mV for KIM-1 detection in PBS bu ffer solution. No current change can be seen with the addition of buffer solution around 50 sec. This stability is important to exclude possible noise from the mechanical change of the buffer solution. By sh arp contrast, the current change showed a rapid response in less than 20 seconds when target 1 ng/ml KIM-1 was switched to the surface at 150 sec. The abrupt current change due to the expos ure of KIM-1 in a buffer solution stabilized after the KIM-1 thoroughly diffused into buffer to re ach a steady state. 10 ng/ml KIM-1 was then applied at 350 sec and it was accompanied with a larger signal correlated to the higher KIM-1 concentration. Further real time tests were carried out to explor e the limit of detection of KIM-1 antibody ( Figure 5-4 ). T h e device was exposed to 10 pg/ml, 100 pg/ml, 1ng/ml, 10ng/ml, and 100ng/ml individually and each concentration was re peated five times to obtain the standard deviation of source-drain current response for each concentration. The limit of detection of this device was 1ng/ml KIM-1 in PBS buffer solutio n and the source-drain current change is nonlinearly proportional to KIM-1 concentration. Between each test, the device was rinsed with a wash buffer of 10 M phosphate buffer solution containing 10 M KCl with pH 6 to strip the antibody from the antigen. These results suggest that our HEMTs are compatible with AKI biomarker, KIM-1, are very sens itive compared to nano-devices [ 151156] and are u s eful for

PAGE 78

78 preclinical and clinical applications. Similar surface modifica tions can be applied for detecting other important disease biomarkers and a compac t disease diagnosis arra y can be realized for multiplex disease analysis. In summary, we have shown that the Au-gat ed region of an AlGaN/GaN HEMT structure can be functionalized with KIM-1 (a kidney injury disease biom arker) antibody for the detection of KIM-1 with a limit of detecti on of 1 ng/ml in PBS buffer. This electronic detection of disease biomarker is a significant step towards a compact sensor chip, which can be integrated with a commercial available hand-held wireless transmitter to realize a portable, fast and high sensitive device for multiple disease diagnosis.

PAGE 79

79 Figure 5-1 (a) Plan view photomicr ograph of a completed device with a 5 nm Au film on the gate region. (b) schematic device cross section. The Au-coated gate area was functionalized with KIM-1 an tibody on thioglycolic acid.

PAGE 80

80 Figure 5-2 IDS-VDS characteristics of HEMT in both PBS buffer and 100 ng/ml KIM-1. Figure 5-3 Time dependent current signal when exposing the HEMT to 1 ng/ml and 10 ng/ml KIM-1 in PBS buffer.

PAGE 81

81 Figure 5-4 Current change in HEMT as a function of KIM-1 concentration.

PAGE 82

82 CHAPTER 6 ZNO BASED LIGHT EMITTING DIODE 6.1 Background ZnO is attracting renewed interest for use in blue/UV light-emitting diodes (LEDs) and photodectors with potential advant ages over the III-nit ride system due to the higher exciton binding energy, availability of high quality bulk substrates and ease of wet etching [ 162166]. The reports of ZnO metal-insula tor-sem i conductor (MIS) electroluminescent diodes go back to the 1970s, with most of the emission being due to defect bands in the blue/green and infra-red (IR) [ 167171]. However, in som e cases, small band-edge emission was observed at low temperatures [ 171] with little understanding of the origin of the holes in these n-type ZnO structures. Generally, no electroluminescence was obs erved in these devices in the reverse bias or without the i-layer. More recently, a numbe r of groups have reported hybrid heterojunction LEDs using n-type ZnO deposited on top of p-type layers of GaN, AlGaN or conducting oxides [ 165, 166, 172 175]. Homojunction ZnO LEDs have been reported by Ts ukazaki [ 176 177] who used temp erature modulation epitaxy for p-type doping of ZnO using N as dopant and fabricated a p-ZnO/i-ZnO/n-ZnO LED on a ScAlMgO4 substrate. Most of the emission consisted of bands at 420 and 500 nm, with a small shoulder at 395 nm assigned to radiative recombination in the pZnO through donor-acceptor pair transitions. Another homojunction ZnO LEDs have also been recently reported by Jae-Hong [ 178, 179] who used rf sputtering technique for P dopant p-type ZnO and fab r icated p-ZnO/n-ZnO LED. The emissi on consisted of a near band edge emission at 380nm and broad deep level emission at appr oximately 640 nm. With p-ZnO/barrier-MgZnO/nZnO/b-MgZnO/n-ZnO structure, carrier combination process is confined in high quality n-type ZnO thin film and the defect related emission at 640 nm is removed. In addition, it has been suggested that semiconducting nanowires may offer additional advantages for light emission due

PAGE 83

83 to the increased junction area, reduced temp erature sensitivity, enhanced polarization dependence of reflectivity and improved carri er confinement in 1-D nanostructures [ 180181]. In this cha p ter, we demonstrate that N+ implantation into bulk single-crystal ZnO substrates can be used to achieve bandedge elec troluminescence (EL) in simple diode structures. The mechanism for bandedge EL is most likely hole creation by impact ionization in the MIS structure. 6.2 Band-Edge Electroluminescence from N+-Implanted Bulk ZnO There have also been recent breakthroughs in the understanding of damage creation and annealing in ion implanted ZnO [182 189] and reports of p-type doping using As im plantation at low tem peratures, followed by multiple step annealing [ 190]. Ion implantation is an attractive process for low-cost, high throughput device m anuf acturing and in this section we show that N+ implantation into bulk single-crystal ZnO subs trates can be used to achieve bandedge electroluminescence (EL) in simple diode structures. The mechanism for bandedge EL is most likely hole creation by impact ionization in the MIS structure. The ZnO samples were (0001) undoped grad e I quality bulk, singlecrystal ZnO crystals from Cermet. They were epiready with one-side-Zn-face-polished by the manufacturer. The room temperature electron concentr ation and mobility established by van der Pauw measurements were 1017 cm and 190 cm2/V s, respectively. Ion implantation was performed at 300K with N+ ions of energy 5 keV (dose of 1.513 cm-2), 20 keV (dose of 513 cm-2) plus 50 keV (dose of 1.314 cm-2) and 130 keV (dose of 3.514 cm-2), followed by rapid thermal annealing (RTA)for 2 mins under a flowing O2 ambient. We also annealed some of the samples in either a conventional t ube furnace or a pulsed laser deposition chamber under O2 ambients for 45 mins, with the same basic trends observed in diode behavior as for the RTA processed

PAGE 84

84 devices. The backside of the substr ates was deposited with full ar ea contacts of e-beam deposited Ti (20 nm)/Au (200 nm) annealed at 400 C [ 191 ]. Circular front-side co ntacts of Ni (20 nm )/Au (80 nm ) with diameter 200 m were deposited by e-beam evaporation and patterned by lithography and lift-off. A schematic of the completed diodes is shown in Figure 6-1 The current-voltage (I-V) characteristics were m easured at 300 K us ing a probe station and Agilent 4145B param eter analyzer. The EL spectrum and output power from the structures were measured using a spectrometer and Si photodi ode, respectively while the photoluminescence (PL) was excited with a He-Cd laser. Figure 6-2 s hows the I-V characteristics from the im planted structures as a function of post-implant RTA temperature under an O2 ambient for 2 mins. The I-V s are characteristic of back-to-back diodes for low anneal temperatures and transition to Schot tky-diode like-behavior at the highest anneal temperat ure. Note that for anneals at 800 C the behavior might be misinterpreted as that from a pn junction because the forward turn -on voltage is that expected from a material with bandgap around 3 eV but this is misleading if not considered in the context of all the data. Thus we do not believe that we create a p-type re gion by activation of the implanted N acceptors. This is consistent with our relatively low dose, the large ionization energy of the N and the residua l n-type background of the s ubstrate. With all of these considerations, it is not likely we have converted the implanted region to p-type conductivity. Figure 6-3 shows room t emperature PL from the bulk ZnO before and after N+ implantation and annealing at 800 C for 2 mi ns (top) and EL from MIS diode at room temperature and 120 K (bottom). The unimplanted ZnO shows strong band-edge (~380 nm) PL, whereas after implantation and anne aling the intensity of this tr ansition is decreased and deep level-related emission peaked at >600 nm is intr oduced. This is expected, since the annealing of

PAGE 85

85 point defects will not be co mplete for 600 C anneals [ 182185] and m ost of the small band-edge peak m ay actually come from the undamaged Zn O underneath the implanted region. In the EL spectrum, we did not observe any band-edge em ission at room temperature, but at lower temperatures (120 K), there is a small peak shifted to higher wavelengths. This is similar to the results previously in ZnO MIS diodes [ 171] and to the EL spectra reported for the ZnO PositiveIntr insic-Negative (PIN) homoj unction diodes. The band-edge emission from our diodes was absent for higher annealing temperatures, although the deep level emission was still present. This is also consistent with our diodes be ing MIS structures and not pn junctions. Figure 6-4 s hows I-V characteristics and forward bias curren t de pendence of integrated EL intensity measured by a Si photodiode from a stru cture annealed at 800 C. The device shows an apparent threshold of about 4.5 V and the forw ard current above this threshold is limited by a series resistance of about 25 much lower than reported fo r the LEDs grown on insulating oxide substrates [ 176, 177]. The EL intensity increases alm ost linearly with drive current above thresho ld. Figure 6-5 ( top) shows an optical microscope image of the light emitted from a single device whereas the bottom of the figure shows a photograph of a device unde r bias in the light and dark. The diodes em it a yellowish light due to the dominance of the deep level emission. We would expect a more uniform emission if we add a transparent conducting layer on the implanted layer to obtain improved current spreading. Given that we do not believe the N+ implanted region is p-type, th en the origin of the holes needed for observation of the band-edge EL needs to be established. Mahan et al. [ 192] in presenting a theory to explain the conduction in ZnO-based metaloxide varistors suggested that holes could be created by impact ionization durin g biasing. Direct eviden ce of the production of holes in forward-biased ZnO varistor s was later reported by Pike et al. [ 193], with the detection

PAGE 86

86 of band-edge EL in addition to the broad subbandgap luminescence peaked near 600 nm. We therefore suggest that the role of the N+ implantation and subsequent anneal in our samples is to create a resistive layer [ 182185] that leads to the re alization of an MIS di ode upon metallization. It is im portant that such e ffects are accounted for in any pn junction ZnO LEDs where the low hole density and propensity for p-layers in ZnO to exhibit unstable conductivity [ 194] m ay lead to m isinterpretation of the device results. In conclusion, band-edge and yellow EL has been obtained from N+-implanted bulk ZnO diodes similar to that observed in MIS diode s. Future work on acceptor implantation should focus on achieving p-type conductivity in the ZnO so that true injection LEDs may be realized. Figure 6-1 Schematic of ZnO MIS diode formed by N+ implantation into a bulk single crystal substrate.

PAGE 87

87 -15-10-5051015-0.04 -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04 N+ implanted ZnO 600C, O2, 2 mins. 800C, O2, 2 mins. 950C, O2, 2 mins. Current(A)Voltage(V) Figure 6-2 I-V characteristics as a function of post-implant ann ealing temperature under an O2 ambient for 2 mins.

PAGE 88

88 0 50000 100000 150000 200000 250000 300000 350000 400000 400450500550600 0 2000 4000 6000 8000 10000 12000 I= 30 mA EL intensity (arb. unit)wavelength (nm) T= 120 K T= 298 KT= 298 KPL intensity (arb. unit) Un-implanted ZnO Implanted ZnO Figure 6-3 Room temperature PL from ZnO before and after N+ implantation and annealing at 800 C for 2 mins (top) and EL from MI S diode at room temperature and 120 K (bottom).

PAGE 89

89 0.000.020.040.060.080.100 2 4 6 8 800C annealed device Voltage Power 0.0 1.0x10-82.0x10-83.0x10-84.0x10-85.0x10-8 Voltage (V)Current (A) Power (arb. uint) Figure 6-4 I-V characteristics and forward bias current dependen ce of integrated EL intensity from an MIS diode annealed at 800 C. The EL intensity was measured by a Si photodiode. Figure 6-5 Optical microscope image of the emissi on from the diode in the dark (top) and photos of the diode under bias from the probe contact taken both in the light and dark (bottom).

PAGE 90

90 CHAPTER 7 INCREASING SCHOTTKY BARRIER HEIGHT W I TH CRYOGENIC METAL DEPOSITION 7.1 Background There is no established gate oxide for II I-V compound semiconductors and therefore all field effect transistors (FETs) in GaAs [ 195] and GaN [196201 ] are based on m e tal Schottky gates. This has some advantages in terms of switching speed because of the low parasitic capacitance of metal-semiconducto r FETs (MESFETs) but is less thermally stable than the metal-oxide-semiconductor FET (MOSFET) approach. Another significan t drawback is the limited range of barrier heights available, especially for metals on GaAs, where surface Fermi level pinning generally limits the barrier height to ~0.72 eV [ 195]. Higher barrier heights would enable larger gate-drain brea kdown voltage, output resistance and power gain and lower gate leakage current and nois e in GaAs MESFETs. Ther e have been a number of reports of enhancing barrier heights on III-V semiconductors by use of cryogenic temperatures during the gate metal deposition Metal films deposit ed at cryogenic temperatures have been shown to enhance Schottky barrier heights on InP, GaAs, InGaAs and some II-VI compounds [ 12, 13, 202204]. The barrier height enh a ncements have been as hi gh as 0.5 eV relative to those deposited at room temperature. The mechanism for the barrier height enhancement is still not firmly established. In the case of Au contacts on InP [ 204], room temperature deposition pr odu ced an ideality factor of 1.02 nearly independent of temp erature and the current transpor t was controlled by thermionic emission (TE). For the case of cryogenic depositi on, the ideality factor was increased and the current transport was controlled by thermionic field emission (TFE) [ 204]. The barrier height enhancem ent and the difference in transport mech anism was attributed to the formation of an amorphous-like structure at the cryogenic diode interface. This amorphous layer was suggested to act as an insulator to create a metalinsulator-semiconductor (MIS)-like structure [ 204].

PAGE 91

91 However, others have disputed this interpreta tion and suggested the results were due to an inhomogeneous Schottky barrier height in the di odes due to a dependence of the local interface dipole on the local interface structure [ 205]. This chapter m a inly examines the effect of cryogenic deposition temperatures on the properties of Au Schottky contacts on n-type Ga As. Section 7.2 presents the role of deposition temperature on the electrical properties of Au/G aAs diodes. We find the barrier height is increased by cryogenic deposition and the in terfacial roughness is decreased. Section 7.3 examines the effect of post-deposition annealin g temperature on the barr ier height and reverse breakdown voltage of Au/n-GaAs diodes deposited at either 77 or 300K. Th e barrier height is increased by cryogenic deposition and remains higher throughout the annealing temperature range up to 300 C. The reverse br eakdown voltage is also increased by the low temperature Au deposition. Finally, section 7.4 reports on X-ra y reflectivity (XRR) studies of the interface between Au deposited on n-GaAs at either 77 K or 300 K, followed by post-deposition annealing at temperatures up to 300 C, for comparing with the results in section 7. 2 and 7.3. The barrier height is increased by ~0.09 eV by cryogenic deposition relative to the room temperature deposition value of 0.73 eV. This is accomp anied by a smoother metal surface, while the metal/GaAs interfacial roughness is similar. As the diodes are annealed to 300 C, the barrier height enhancement disappears; the Au/GaAs inte rfaces continue to show the same degree of roughness while the metal surface becomes rougher. Ot her metals such as Pt, Ti, Pd and Ni were also examined for barrier height enhancement.

PAGE 92

92 7.2 Improved Au Schottky Contacts on GaAs Using Cryogenic Metal Deposition An evaporator system with a load-lock and five pockets for different metals was used in these experiments. The load-lock maintained the background pressure in the metal deposition chamber in the range of 10-10 Torr. The background pressure of a typical commercial evaporator is in the range of low 10-7 to high 10-8 Torr and the theoretical monolay er formation time in this vacuum environment is around 1 minute. With MB E-like background pressures, the rate of gas molecule impingement on a sample surface is significantly reduced and the theoretical monolayer formation time is extended to approximately an hour or two. In comparison to conventional evaporation techni ques, this system results in enhanced integrity of the semiconductor surface before metal deposition. Front-side contacts of 1000 thick Au were deposited at 77 K or 300 K onto n-GaAs (n~1017 cm-3) substrates with full area back Au/Ge/Ni/Au contacts that had b een alloyed at 400 C for 3 mi ns. Prior to insertion in the evaporator, the samples were cleaned in 3:1:50 of HNO3: HF: H2O for 1 min. The Au contacts ranged in diameter from 200-800 m and were patterned by lift-off of photoresist. The currentvoltage (I-V) characteristics of the resulting diodes were measured on an Agilent 4156C parameter analyzer. The barrier height, b, and diode ideality factor, n were extracted from the relation for the thermionic emission over a barrier [ 206], )exp()exp(2*nk T eV k T e TAJb F (7-1) where JF is the forward current density, A* is the Richardsons constant for n-GaAs, T is the absolute temperature, e is the electronic charge, k is Boltz manns constant, and V is the applied voltage. In addition, the surface and interfacial roughness of thin Au films (~100 ) deposited on GaAs substrates was examined by X-Ray Reflectivity (XRR).

PAGE 93

93 Figure 7-1 s hows I-V characteristics of Au/GaAs Schottky diodes deposited at either 7 7 K ( ) or 300 K ( ) with 200 m (left) or 800 m (right) in contact diameter. An expanded view of the forward voltage part of the curves is show n at the bottom of the figure. There is a clear decrease in both reverse and forward bias current for the diodes with Au deposited at 77 K, consistent with an increase in the effective Schott ky barrier height. It has been reported that the crystal structure and grain size of the low temper ature deposited metal are different from metals deposited at room temperature. The changes in electrical behavior were consistent both spatially within a 2 wafer and for different contact diameters. Figure 7-2 shows the forward current dens ities as a function of bias for diodes of different diam eter, deposited at e ither 77 or 300 K. The Sc hottky barrier height enhancement by deposition at cryogenic temperatur es is thought in part to result from the minimized interaction between the metal and semiconductor during the metal deposition. The stability of the metal-semiconductor interface an d the stability of the small grain size metal contact are important for device re liability. Typically after the gate metallization step, the device is encapsulated by a SiNx dielectric for packaging. This di electric is deposited by PECVD and the temperature that the substrate is exposed to is 200-300 C. At these temperatures, it is desirable that interface diffusion and grain growth be minimized. Table 7-1 lis ts th e common metals used for Schottky contacts on GaAs and their melting temperatures. A recognized measure for the onset of grain growth is 0.4 Tmelt. As indicated in the table, contacts of Ti, Pt, and Pd should not recrystallize at PECVD depos ition temperatures and therefore these may be more effective choices for actual GaAs devices. Fo r significant interfacial diffusion to take place, the constituents of either the semiconductor or th e metal contact must become mobile and form a solid solution. The contact metals of Ti, Pt and Pd have little diffusion below 0.4 Tmelt.

PAGE 94

94 The origin of the reverse current can be ex amined by looking at the dependence of current on perimeter/area ratio. Figure 7-3 sh ows reverse curren t at -4 V for diodes deposited at either 77 or 300 K, as a function of either contact diameter (top) or area (bottom).The reverse current was proportional to both the perimeter and area of the rectifying contact, suggesting that both surface and bulk contributions are present in this voltage range. Theref ore, low temperature deposition does not seem to reduce Fermi leve l pinning by surface states in GaAs. Table 7-2 gi ves a summary of the electrical propertie s of the diodes. The m ain differe nces are an effective increase in barrier height of 10-13 % fo r cryogenic deposition, from a m ean value of 0.73 eV for room temperature deposition to a mean value of 0.82 eV for 77 K deposition. However, it should also be noted that the ideality factors were larger for the low te mperature diodes, perhaps suggesting the presence of interfacial contamination gett ered to the cold surface during the pump-down and initial stages of deposition. This could be improved by including gettering sources such as tungsten filaments within the ch amber during evaporation. It has been proposed that these low temperature contacts could be governed by the lack of a thermally diffused interfacial layer, geometry differences due to reduced metal cluste ring, or strain imposed on the interface caused by thermal expansion differences between the meta l and substrate, or some combination of the three. TEM and XPS studies by D. S. Cammack et al support the assertio n of increased barrier heights due to an interfacial layer differences [ 207]. Their TEM study of low tem perature deposited co ntacts showed a more abrupt metalsemiconductor interface, than room temperature contacts. All of the diodes showed excellent stability under forward bias aging. Figure 7-4 sho w s the time dependence of forward bias at a current of 10 mA for diodes deposite d at either 77 or 300 K, with no evidence of drift due to trapping effects. There was also no apparent difference in

PAGE 95

95 contact morphology at low resolution. Figure 7-5 shows optical m i croscope images of Au contacts deposited at 77 K (left) or 300 K (right), with both exhibiting excellent morphology. Cryogenic metal deposition also reduces the resistivity, of very thin films, four or five orders of magnitude compared to contacts deposited at room temperature [ 208]. By lim iting surf ace diffusion, atoms deposited on cryogenically cooled substrates tend to stay close to where they impinge on the substrate surface. Conversely, room temperature deposited atoms are more likely to re-evaporate or diffuse along the surface. Diff usion can lead to clustering, and eventually coalescence leaving voids on the substrate [ 12]. The lack of tem perature related diffusion at 77 K causes the cryogenic m etal film to become continuous at lower film thicknesses than films deposited at room temperature. As metal thickn ess increases, room temperature films experience secondary nucleation, and the voids fill. Void filling makes the room temperature films more continuous, while the larger kinetic energy of room temperature deposited makes these atoms more likely to form regular crystal lattices than low temperature films. The increased continuity of room temperature films coupled with better crystal quality lead to a cross over, where the room temperature films then have lower resistances than comparable thickness cryogenic films. The lack of temperature related lateral surface di ffusion of metal deposited at 77 K implies that cryogenic metal deposition may also be used to enhance the adhesion of refractory metals (such as molybdenum). The adhesion of such metals limits their application at short gate lengths. Successful use of refractory metals at short gate lengths may enhance the reliability of devices at elevated temperatures. Figure 7-6 s hows XRR of thin (~100 ) Au layers of GaAs for the two different deposition tem peratures, and the associated Au surf ace roughness and Au/GaAs interfacial roughness derived from the XRR. The Au surface is clearl y smoother for low temperature deposition when

PAGE 96

96 measured by this higher resolu tion technique. There is also a slight decrease in metal/GaAs interfacial roughness with cryogenic deposition, suggesting less diffusion of the initially deposited Au atoms, as discussed above. Details of XRR results are discussed in section 7.4. In conclusion, the results of this study are summarized as follows: The use of low temperature deposition of Au on n-GaAs produces an increase in Schottky barrier height of 10-13 % relative to conventional room temperature deposition. The improved barrier height is accompanie d by a smoother Au surface and more abrupt interface between the Au and the underlying GaAs. Additional work is needed to determine the or igin of the increased ideality factors in low temperature diodes. Table 7-1 Metals, melting temperature, and recrystallization temperature. Melting Temp. (C) Recrystallization Temp. 0.4Tmelt (C) Aluminum 660 264 Gold 1064 425 Indium 156 62 Nickel 1453 581 Palladium 1552 620 Platinum 1769 707 Titanium 1668 667 Table 7-2 Summary of Au/GaAs diode characteristics fo r deposition of the Au at either 77 K or 300 K slope J(A, intercept) Diameter ( m) R(cm) Js(A/cm2) n barrier height d d % 14.21 7.0x10-11 200 0.01 2.23x10-7 1.170.73 14.24 2.8x10-10 400 0.02 2.27x10-7 1.170.73 300 K 14.23 9.6x10-10 800 0.04 1.91x10-7 1.170.73 11.66 1.6x10-12 200 0.01 5.27x10-9 1.430.83 0.097 13.3 11.12 1.0x10-11 400 0.02 8.18x10-9 1.500.82 0.086 11.8 77 K 11.06 5.1x10-11 800 0.04 1.02x 10-8 1.510.81 0.076 10.3

PAGE 97

97 Figure 7-1 I-V characteristics of Au/GaAs Schottky diodes deposited at either 77 K( ) or 300 K( ) for both 200 m dia.(left) and 800 m dia.(right) contact. An expanded view of the forward voltage part of the curves is shown at bottom. Figure 7-2 Forward current densities as a functio n of bias for diodes of different diameter deposited at either 77 or 300 K.

PAGE 98

98 Figure 7-3 Reverse current at -4 V for diodes deposited at either 77 or 300 K, as a function of either contact diameter (top) or area (bottom). Figure 7-4 Time dependence of forward bias at a current of 10 mA for diodes deposited at either 77 or 300 K.

PAGE 99

99 Figure 7-5 Optical microscope images of Au c ontacts deposited at 77 K (left) or 300 K (right). Figure 7-6 XRR of thin (~90 ) Au layers of Ga As for the two different deposition temperatures and the associated Au surface roughness and Au/GaAs interfacial roughness derived from the XRR.

PAGE 100

100 7.3 Thermal Stability of Au Schottky Diodes on GaAs Deposited at Either 77 K or 300 K The devices used for this study are those repor ted in section 7.2. The I-Vs were obtained as a function of post-deposition annealing temp erature (up to 300 C, 30 minutes anneals under air ambient). The Schottky barrier height, b, and diode ideality factor, n were extracted from the relation for the thermionic emission (eq. 7-1) as well. Figure 7-7 show s I-V characteristics of 400 m diameter diodes deposited at either 300 K (l eft) or 77 K (right), as a function of postdeposition annealing temperature. The diodes deposited at low te mperature have reverse current densities approximately two orders of magn itude lower than those deposited at room temperature. The respective barrier heights extr acted from the forward I-V characteristics were 0.73 eV for the room temperature diodes and 0. 82 eV for the low temperature samples. Both types of diodes show increases in reverse current density after annealing at 200 C or higher, with very significant increases after 300 C anneals. The forw ard I-V characteristics as a function of annealing temperature are shown in more detail in Figure 7-8 Note that while both diode types show a deteriorati on in rectifying behavior, the s am ples with low temperature deposited contacts still re tain lower current densities at al l annealing conditions. The forward turn-on voltage for rectifiers is given by [ 206], FONB F FJRn T A J e nkT V )ln(2** (7-2) where k is Boltzmanns constant, T is the absolute temperature, e is the electronic charge, A** is Richardsons constant and RON the on-state resistance. Thus the observed increase in turn-on voltage is also consistent as resulting from a larger barrier height. The on/off ratio of the diodes was ~104 at 1 V/-4 V for the low temperature deposited devices and ~102 for the room temperature diodes.

PAGE 101

101 Figure 7-9 shows the Schottky barrier height as a function of annealing temperature f o r diodes deposited at either 77 or 300 K. The enhancement in barrier height of ~0.09 eV between the two types of diodes is retain ed over the entire annealing temperature range investigated. After annealing at 200 C, the low temperature diodes exhibit a barrie r height of 0.53 eV, compared to 0.44 eV for the comparable room temperature deposited diode. Note the ideality factor for the low temperature diodes was always larger than those for the room temperature devices. For example, the as-deposited diodes sh owed a value of 1.17 for the room temperature devices and 1.43 for the cryogenic diodes. This would be consistent with the presence of an interfacial layer that produces more of an MIS behavior th an a true Schottky contact. In the case of breakdown being initiated in the bulk, the reverse breakdown voltage of a diode VB can be expressed as [ 209], 22BD BWeN V (7-3) where ND is the doping on the epilayer, WB is the depletion de pth at breakdown and is the dielectric constant of GaAs. Figure 7-10 shows the reverse leakage current (@ -4 V) o n two different scales as a function of annealing temperature for diodes de posited at either 77 or 300 K. We defined the breakdown voltage as the reve rse bias needed to re ach a current of 100 A cm-2. These values are shown as a func tion of anneal temperature in Figure 7-11 There is roughly a 50 % increase in VB for the diodes deposited at 77 K as a result of the increased barrier height. Previous electron microscopy and chemical bonding studies are supportive of interfacial layer differences being the cause of the differences in barrier height, with a more abrupt metalsemiconductor interface for low temperature deposited contacts [ 207]. Au rectifying contacts deposited at 77 K on n-GaAs show an enhancem ent in barrier height of ~12 % over their value for room t emperature deposition (0.73 eV ). This enhancement persists

PAGE 102

102 to annealing temperatures of 200 C, while anneal ing at 300 C produces a severe degradation in rectifying behavior for both type s of diodes. The increase in ba rrier height translates to a decrease in reverse current dens ity of several orders of magnit ude for Au/GaAs diodes deposited at 77 K. This process is a relatively simple one with many potential advantages in the dc performance of GaAs MESFETs. Figure 7-7 I-V characteristics of 400 m diameter diodes deposited at either 300 K (left) or 77 K (right), as a function of post-de position annealing temperature. Figure 7-8 Forward I-V characteristics of 400 m diameter diodes deposited at either 300 K (left) or 77 K (right), as a function of pos t-deposition annealing temperature.

PAGE 103

103 Figure 7-9 Schottky barrier height as a function of annealing temperature for diodes deposited at either 77 or 300 K. Figure 7-10 Reverse leakage current (@ -4 V) on two different scales as a function of annealing temperature for diodes deposited at either 77 or 300 K.

PAGE 104

104 Figure 7-11 Reverse brea kdown voltage (@ -100 A) as a function of annealing temperature for diodes deposited at either 77 or 300 K. 7.4 Interfacial Differences in Enhanced Sc hottky Barrier Height Au/n-GaAs Diodes Deposited at 77 K 100 thick Au films deposited at 77 K or 300 K, followed by post-deposition annealing at temperatures up to 300 C, were used in this X-Ray Reflectivity (XRR) study. We interest in interfacial differences of 77 K and 330 K deposited Au on n-GaAs substrate before and after thermal treatment, which correlate with Au/n-G aAs diode characteristic s and thermal behaviors as discussed in section 7.2 and 7.3. In addition, front-side contacts of 1000 thick Au, Pt, Ni, Pd or Ni were deposited at 77 K or 300 K onto n-GaAs (n ~1017 cm-3) substrates to examine their individual barrier height enhancements. As shown in Figure 7-1 (top), the diodes deposited at low tem p erature have reverse current densities approximately two orders of magn itude lower than those deposited at room temperature. The respective barrier heights ex tracted from the forwar d I-V characteristics ( Figure 7-1 bottom ) were 0.73 eV for the room temperat ure diodes and 0.82 eV fo r the low temperature

PAGE 105

105 samples. Similar measurements were performed for the diodes deposited with Ni, Pt, Pd or Ti at 77 K or 300 K. Table 7-3 shows a summary of the extracted b arrier he ights for all the metals examined. Only Au, Pd and Ni showed any significant change in barrie r height for cryogenic metal deposition. Note also that we observed cr acking and peeling of the metal in the case of Pt, Pd and Ni, leaving Au as the only metal that sh owed an enhancement in barrier height and good adhesion to the GaAs. Optical micrographs of the Ti and Au diodes deposited at different temperatures are shown in Figure 7-12 as an example of a situation where the m e tal (Ti in this case) peels off when deposited at low temperature. Figure 7-13 shows the XRR spectra from 77 K Au /GaAs diodes as a function of postdeposition anneal tem perature. From this data it is possible to d econvolute the interfacial roughness between the Au and GaAs and also the metal roughness at the air interface. To make the differences in the spectra more obvious, Figure 7-14 shows the XRR spectra from Au/GaAs diodes deposited at either 77 K or 300 K before (left) or after (ri ght) annealing at 300 C.Two things a re obvious from this data, firstly, there is a clear difference between the samples deposited at different te mperatures and secondly, the effect of the 300 C annealing is to wash out these differences. Figure 7-15 shows the interfacial Au /GaAs roughness and m e tal/air roughness data derived from the XRR spectra for samples deposited at ei ther 300 K(left) or 77 K( right) as a function of post-deposition annealing temperature. The interfacial roughne ss in the room temperature deposited diodes is basically constant with anne al temperature, whereas that for the cryogenic diodes is initially smoother but roughens with ann ealing, reaching a similar value to that in the room temperature diodes ( Figure 7-16 ). By contrast, the m etal/air roughness im proves above 200

PAGE 106

106 C in the room temperature diodes, but worsens w ith annealing in the 77 K diodes. Note that the metal is initially smoother on the low temperature diodes ( Figure 7-16 ). The correlation of electrical and x -ray data clear ly shows that the enhanced barrier height is associated with a smoother Au/GaAs interf ace and that post-deposition annealing of the cryogenic diodes roughens this inte rface to a value similar to that of diodes deposited at room temperature while at the same time reducing the Schottky barrier height back to the values obtained on the room temperature diodes annealed under the same conditions. Previous electron microscopy and chemical bonding studies are also supportive of interfacial layer differences being the cause of the differences in barrier height, with a more abrupt metal-semiconductor interface for low temperature deposited contacts [ 207].There are still m any issues to b e resolved, for example, why only certain metals exhibit the increased barrier height when deposited at low temperatures. This may be related to the crystal structure and grain size of the particular metal layers on GaAs obtained at diffe rent deposition temperatures. The melting temperature of Au (1064 C) is well below that of the other metals studied here(1453 C for Ni, 1552 C for Pd, 1769 C for Pt, and 1668 C for Ti) and thus the recrystal lization (onset of grain growth) temperature will also be lower, since this is typi cally about 40 % of the melting temperature. It is desirable that that interface di ffusion and grain growth be minimized during encapsulation of the GaAs device by a SiNx dielectric for packaging. This di electric is deposited by PECVD and the temperature that the substrate is exposed to is 200-300 C. For si gnificant interfacial diffusion to take place, the constituents of either the semic onductor or the metal contact must become mobile and form a solid solution. The contact metals st udied here should have little diffusion below 0.4Tmelt.

PAGE 107

107 The results of our study may be summarized as follows: The deposition of Au at low temperatures produc es an increase in barrier height from 0.73 eV to 0.82 eV. Ni shows an even larger enha ncement, but the metal contact in that case shows cracking and peeling. No significant enhancement in barrier height was observed for low temperature deposited Pt and Ti. The improved barrier height in the case of Au is accompanied by a sharper metal/GaAs interface. As the samples are annealed to 300 C, this interface roughens to the same value as in room temperature deposited diodes and th e enhancement in barrier height disappears. Table 7-3 Barrier height enhancement obs erved for different metals on n-GaAs Metal Cracks at 77K (300K)(eV) (77K)(eV) d (%) Au No 0.73 0.82 12 Pt Yes 0.79 0.79 0 Ti No 0.69 0.71 3 Pd Yes 0.76 0.81 7 Ni Yes 0.67 0.79 18

PAGE 108

108 Figure 7-12 Optical micrograph images of Ti de posited at either 77 K (top left) or 300 K (top right) and Au at 77 K (bottom left) or 300 K (bottom right) on GaAs. Figure 7-13 XRR spectra from 77 K Au/GaAs diodes as a function of post-deposition annealing temperature.

PAGE 109

109 Figure 7-14 XRR spectra from Au/G aAs diodes deposited at either 77 K or 300 K before (left) or after (right) annealing at 300 C. Figure 7-15 Interfacial Au/GaA s roughness and metal/air roughness data derived from the XRR spectra for samples deposited at either 300 K (left) or 77 K (right), as a function of annealing temperature.

PAGE 110

110 Figure 7-16 Comparison of meta l roughness (left) and metal-se miconductor interfacial roughness (right) for the two types of diodes as a function

PAGE 111

111 CHAPTER 8 SUMMARY AND FUTURE WORK 8.1 Hydrogen Sensor Using Multiple ZnO Nanorods A variety of different metal cata lyst cluster coatings (Pt, Pd Au, Ag, Ti, and Ni) deposited on multiple ZnO nanorods were compared for their effectiveness in enhancing sensitivity for detecting hydrogen at room temper ature. The metal cluster coated nanorods were biased at 0.5 V and power levels for these diode sensors were ~0 .4 mW. Pt-coated nanor ods showed an increase of conductance up to 8 % in room temperature upon exposure to 500 ppm hydrogen in N2. This is a factor of two larger than that obtained with Pd, and more than an order of magnitude larger than that achieved with the remaining metals. Pt-coated ZnO nanorods easily detected hydrogen down to 100 ppm, with 4 % increase of conductance at this concentration after 10 min exposure. It took a few minutes for the nanorods to return to their original conduc tance after switching hydrogen off and back to air. The slow response at room temperature is a drawback in some applications, but the sensors do offer low power operation and very good detection sensitivity. The sensitivity for detecting hydrogen using mu ltiple ZnO nanorods with cluster coating of Pd on the surface is further investigated with different hydrogen concentrations. The nanorods show changes of conductance upon exposure to hydrogen concentrations of 10-500 ppm balanced with N2 approximately a factor of five larger than without Pd. Pd-coated ZnO nanorods detected hydrogen down to ~10 ppm, with a increa se of conductance >2.6 % at 10 ppm and >4.2 % at 500 ppm H2 in N2 after 10 min exposure. The nanorods had no response to O2 at room temperature. As opposed to the slow recovery for Pt coated nanorods, Pd coated nanorods showed a much quicker recovery time upon switching the ambient from hydrogen to either air or pure O2, for which approximately 95 % of the initial ZnO conductance after exposure to hydrogen was recovered within 20 s. This rapid and easy recoverability make the Pd-coated

PAGE 112

112 nanorods suitable for practical applications in hy drogen selective sensing at ppm levels at room temperature with ~0.4 mW power consumption. In conclusion, both Pt-coated and Pd-coated ZnO nanorods appear well suited to detection of ppm concentrations of hydrogen at room temp erature. Pd coated na norods showed better recovery characteristics. The ZnO nanorods can be placed on cheap transparent substrates such as glass, making them attractive for low-cost sens ing applications, and can also operate at very low power conditions. Of course, there are many i ssues still to be addressed, in particular regarding the reliability and longterm reproducibility of the sensor response before it can be considered for space-flight applicat ions. In addition, the slow respons e of the Pt coated sensors at room temperature is a major issue in some applications. 8.2 Hydrogen Sensor Using AlGaN/GaN Schottky Diode and High Electron Mobility Transistor Pt-gated AlGaN/GaN high electron mobility tran sistors can be used as room temperature hydrogen gas sensors at hydrogen co ncentrations as low as 100 ppm. A comparison of the changes in drain and gate current-voltage (I-V) characteristics with in troduction of 500 ppm H2 into the measurement ambient shows that monitoring the change in dr ain-source current (FET mode) provides a wider gate voltage operation range for maximum detection sensitivity and higher total current change than measuring the ch ange in gate current (Schottky diode mode). However, over a narrow gate voltage range, the relative sensitivity of detection by monitoring gate current changes (Schottky diod e mode) is up to an order of magnitude larger than that of drain-source current changes (FET mode). In both cas es, the changes are fully reversible in ~2-3 mins at 25 C upon removal of the hydrogen from the ambient. These Pt/AlGaN/GaN HEMTs operated in either a diode mode or in an FET mode show the ability to detect 500 ppm H2 in N2

PAGE 113

113 at room temperature. The FET m ode provides a larger total current change with introduction of hydrogen into the ambient, but the diode mode shows a higher relative sensitivity over a limited range of forward biases. The design of AlGaN/GaN differential sensing diodes is shown to provide robust detection of 1 % H2 in air at 25 C. The active device in the differe ntial pair is coated with 10 nm of Pt to enhance catalytic dissociation of molecular hydroge n, while the reference diode is coated with Ti/Au. The active diode in the pair shows an incr ease in forward current of several mA at a bias voltage of 2.5 V when exposed to 1 % H2 in air. The use of the differential pair removes false alarms due to ambient temperature variations These AlGaN/GaN HEMT differential sensing diodes appear well-suited to hydrogen detection applications. The use of TiB2-based Ohmic contacts (Ti/Al/TiB2/Ti/Au) on Pt-Schottky contact AlGaN/GaN heterostructure hydrogen sensing diodes is shown to provide very stable operation for detection of 1 % H2 in air under field conditions where temperature is allowed to vary. By contrast, the use of more conventional Ti/A l/Pt/Au Ohmic contacts led to higher background variations in current that affect the ultimate detection threshold of the sensors. Combined with the superior thermal stability of these boride-based contacts, this metallization system appears attractive for sensors of longterm monitoring applications. In conclusion, combined with a differential pa ir geometry that compares current from an active diode with Pt Schottky co ntact and a passive diode with Ti/Au Schottky contact, the more stable TiB2-based Ohmic contacts reduce false alarms due to ambient temperature changes, and suggest that integrated chips involving gas sensors and HEMT-based circuitry for off-chip communication are feasible in the AlGaN/GaN sy stem. Future work will involve design and fabrication of an integrated sensor chip with GaN HEMT amplifier and transmitter.

PAGE 114

114 8.3 Mercury Ion Sensor Using AlGaN/GaN High Electron Mobility Transistor Bare Au-gated and thioglycolic acid func tionalized Au-gated Al GaN/GaN high electron mobility transistors (HEMTs) were used to detect mercury(II) ions. Fast detection of less than 5 seconds was achieved for thioglycolic acid functiona lized sensors. This is the shortest response time ever reported for mercury detection. Thioglycolic acid functionalized Au-gated AlGaN/GaN HEMT based sensors showed 2.5 time s larger response than bare Au-gated based sensors. The sensors were able to detect mercury (II) ion concentration as low as 10-7 M. The sensors showed an excellent sens ing selectivity of more than 100 for detecting mercury ions over sodium or magnesium ions. Bare Au-gated and thioglycolic acid func tionalized Au-gated Al GaN/GaN HEMTs were further used to detect both mercury(II) and copper(II) ions. The bare Au-gate sensor was only sensitive to Hg2+, and thioglycolic acid functionalized sensors could detect both Hg2+ and Cu2+ ions. Both surfaces had a selec tivity of approximately a hundredfold over other contaminating ions of sodium, magnesium and lead. Both bare Au-gate and thioglycolic acid functionalized sensor can also be repeatedly used after a simple DI water rinse. By fabr icating an array of the sensors on a single chip and selectively functionalizing some sens ors with thioglycolic acid, a multi-functional specific detector can be fabricated. Such a sensor array can be used to detect quantitatively Hg2+ ions in Cu2+ ion solution or Cu2+ ions in Hg2+ ion solution. The dimensions of the active area of the sensor a nd the entire sensor chip are 50 m 50 m and 1 mm 5 mm, respectively. Our results show that portable, selective, and fast Cu2+ and Hg2+ sensors can be realized by combining bare Au-gated and thioglyc olic acid-functionalized surface in one sensor.

PAGE 115

115 8.4 Disease Biomarker Sensor Using AlGaN/ GaN High Electron Mobility Transistor AlGaN/GaN high electron mobility transistors (H EMTs) were used to detect kidney injury molecule-1 (KIM-1), an important biomarker for early kidney injury detection. The gate region consisted of 5 nm gold deposited onto the AlGaN surface. The gold was conjugated to highly specific KIM-1 antibodies through a self-assembled monolayer of thioglycolic acid. The HEMT source-drain current showed a clear dependence on the KIM1 concentration in phosphate buffered saline solution. The limit of de tection was 1 ng/ml using a 20 m 50 m gate sensing area. This electronic detection of disease biom arker is a significant step towards a compact sensor chip, which can be integrated with a co mmercial available hand-held wireless transmitter to realize a portable, fast and high sensitive devi ce for multiple disease diagnosis. Our approach shows potential for both pr eclinical and clinical disease di agnosis with accurate, rapid, noninvasive, and high throughput capabilities. 8.5 ZnO Based Light Emitting Diode N+ ion implantation at moderate doses (1013-1014 cm-2) into nominally undoped (n ~1017 cm-3) bulk single crystal ZnO substrates followed by annealing in the range 600-950 C was used to fabricate diodes that show band-edge elec troluminescence at 120 K (~390 nm) under forward bias conditions. The current-voltage (I-V) behavior s of the diodes are characteristics of metalinsulator-semiconductor (MIS) devices but not p-n junctions, and suggest the implantation creates a more resistive region in the n-ZnO in which holes ar e created by impact ionization during biasing, similar to the cas e of electroluminescence in ZnO varistors. The series resistance is only 25 due to the use of the conducting ZnO subs trate. We demonstrat ed that band-edge and yellow EL could be obtained from N+-implanted bulk ZnO diodes similar to that observed in

PAGE 116

116 MIS diodes. Future work on acceptor implantation should focus on achieving p-type conductivity in the ZnO so that true injection LEDs may be realized. 8.6 Increasing Schottky Barrier Height with Cryogenic Metal Deposition The use of low temperatures (~77 K) duri ng Au Schottky contact deposition onto n-GaAs produces an increase in barrier height from 0. 73 eV for room temperature diodes to 0.82 eV. The increase in barrier height translates to a decrease in reverse current density of several orders of magnitude for Au/GaAs diodes deposited at 77 K. The reverse breakdown voltage of low temperature deposited diodes was ~50 % larger than conventional Au/GaAs diodes. There is no evidence of drift in the forward current in either type of diode, and the low temperature deposited samples show smoother Au layers and more abru pt Au/GaAs interfaces as determined by X-Ray Reflectivity measurements. Both types of diode s show surface and bulk contributions to the reverse bias current. The ideality factor of th e cryogenically processed devices (~1.43) was higher than for room temperature diodes (~1.17) and it may result from contaminants gettered to the cold GaAs surface. Not all Schottky metals show this enhancement; for example Pt and Ti do not show any significant change in barrier height whereas Au, Pd and Ni show increases between 7-18 %. The enhancement of ~0.09 eV (a 12 % incr ease) in Schottky barrier height for Au deposited at cryogenic temperatur es on n-type GaAs relative to conventional deposition at 300 K is shown to persist for anneali ng temperatures up to 200 C. At higher anneal temperatures (300 C), both types of diodes show a severe deterioration in rectifying be havior. We used X-Ray Reflectivity to show that the main differe nce between Au deposited at 77 K and room temperature is a decreased interfacial roughness between the Au and GaAs. As the diodes are annealed to 300 C both the difference in barrier height a nd interfacial roughness is lost. This is a

PAGE 117

117 simple method has many potentials for improving the performance of Ga As metal semiconductor field effect transistors (MESFETs). The results of our study may be summarized as follows: The use of low temperature deposition of Au on n-GaAs produces an increase in Schottky barrier height of 10-13 % relative to conventional room temperature deposition. The improved barrier height is accompanied by a smoother Au surface, and more abrupt interface between the Au and the underlying GaAs. Additional work is needed to determine the or igin of the increased ideality factors in low temperature diodes. Ni shows an even larger enhancement, but th e metal contact in that case shows cracking and peeling. No significant enhancement in barrier height was observed for low temperature deposited Pt and Ti. The improved barrier height in the case of Au is accompanied by a sharper metal/GaAs interface. As the samples are annealed to 300 C, this interface roughens to the same value as in room temperature deposited diodes, and the enhancement in barr ier height disappears.

PAGE 118

118 LIST OF REFERENCES [1] Sensors: A Global Strategic Business Report, released by Global Indu stry Analys ts, Inc. http://www.strategyr.com /MCP-2740.asp [2] S. J. Pearton, D. P. Norton, K. Ip, Y. W. Heo, and T. Steiner, J. Vac. Sci. Technol. B 22, 932 (2004). [3] S. J. Pearton, J. C. Zolper, R. J.Shul, and F. Ren, J. Appl. Phys. 86, 1 (1999). [4] M. H. Huang, Y. Wu, H. Feick, N. Tran, E. Weber, and P. Yang, Adv. Mater. 13, 113 (2001). [5] Z. W. Pan, Z. R. Dai, and Z. L. Wang, Science 291, 1947 (2001). [6] W. Park, G. Yi, M. Kim, and S. Pennycook, Adv. Mater. 15, 526 (2003). [7] Y. Alifragis, A. Volosirakis, N. A. Chanio takis, G. Konstantinidis, E. Iliopoulos, and A. Georgakilas, Phy. Stat. Sol. (a) 204, 2059 (2007). [8] M. Eickhoff, J. Schalwig, G. Steinhoff, O. Weidemann, L. Grgens, R. Neuberger, M. Hermann, B. Baur, G. Mller, O. Ambacher, and M. Stutzmann, Phys. Stat. Sol. (c) 0, 1908 (2003). [9] O. Ambacher, M. Eickhoff, A. Link, M. Hermann, M. Stutzmann, F. Bernardini,V. Fiorentini, Y. Smorchkova, J. Speck, U. Mishra W. Schaff, V. Tilak, and L. F. Eastman, Phys. Stat. Sol. (c) 0, 1878 (2003). [10] B. Claflin, D. C. Look, S. J. Park, and G. Cantwell, J. Cryst. Growth 287, 16 (2006). [11] S. Limpijumnong, S. B. Zhang, S.-H. Wei, and C. H. Park, Phys. Rev. Lett. 92, 155504 (2004). [12] H. J. Lee, W.A. Anderson, H. Hardtdegen, and H. Luth, Appl. Phys. Lett. 63, 1939 (1993). [13] S. A. Clark, S. P. Wilks, A. Kestle, D. I. Westwood, and M. Elliot, Surf. Sci. 850, 352-354, (1996). [14] D. S. Cammack, S. M. McGregor, J. J. McCh esney, S. A. Clark, P. R. Dunstan, S. R. Burgess, S. P. Wilks, F. Peiro, J. C. Ferrer, A. Cornet, J. R.Morante, A. Kestle, D. I. Westwood, and M. Elliott, Appl. Surf. Sci. 501, 123-124 (1998). [15] S. L. Rumyantsev, M. S. Shur, and M. E. Levinshtein, Selected Topics in Electronics and System 33, 1-19 (2004). [16] D. W. Palmer, http://www.semiconductors.co.uk 2002. 06. [17] J.-Y. Duboz, Phy. Stat. Sol. (a) 176, 5 (1999). [18] A. R. Hutson, Phy. Rev. 108, 222 (1957).

PAGE 119

119 [19] D. C. Look, J. W. He msky, and J. R. Sizelove, Phys. Rev. Lett. 82, 2552 (1999). [20] B. J. Jin, S. H. Bae, S. Y. Lee, and S. Im, Mater. Sci. Eng. B 71, 301 (2000). [21] D. M. Hofmann, A. Hofstaetter, F. Leiter, H. Zhou, F. Henecker, B. K. Meyer, S. B. Orlinskii, J. Schmidt, and P. G. Baranov, Phys. Rev. Lett. 88, 045504-1 (2002). [22] C. G. Van de Walle, Phys. Status Solidi B 229, 221 (2002). [23] S. F. J. Cox, E. A. Davis, P. J. C. King, J. M. Gil, H. V. Alberto, R. C. Vilao, J. Piroto Duarte, N. A. De Campos, and R. L. Lichti, J. Phys.: Condens. Matter 13, 9001 (2001). [24] C. G. Van de Walle, Phys. Rev. Lett. 85, 1012 (2000). [25] D. M. Bagnall, Y. F. Chen, M. Y. Shen, Z. Zhu, T. Goto, and T. Yao, J. Cryst. Growth 184, 605 (1998). [26] B. J. Jin, S. Im, and S. Y. Lee, Thin Solid Films 366, 107 (2000). [27] E. G. Bylander, J. Appl. Phys. 49, 1188 (1978). [28] N. Riehl, O. Ortman, Z. Elektrochem. 60, 149 (1952). [29] F. A. Krger and H. J. Vink, J. Chem. Phys. 22, 250 (1954). [30] I. Y. Prosanov and A. A. Politov, Inorg. Mater. 31, 663 (1955). [31] N. Y. Garces, L. Wang, L. Bai, N. C. Giles, L. E. Halliburton, and G. Cantwell, Appl. Phys. Lett. 81, 622 (2002). [32] H.-J. Egelhaaf and D. Oelkrug, J. Cryst. Growth 161, 190 (1996). [33] D. Hahn, R. Nink, and Physik, Cond. Mater. 3, 311 (1965). [34] M. Liu, A. H. Kitai, P. Mascher, and J. Lumin. 54, 35 (1992). [35] H. Xing, S. Keller, Y.-F. Wu, L. McCarthy, I. P. Smorchkova, D. Buttari, R. Coffie, D. S. Green, G. Parish, S. Heikman, L. Shen, N. Zhang, J. J. Xu, B. P. Keller, S. P. DenBaars, and U. K. Mishra, J. Phys.: Condens. Matter 13, 7139 (2001). [36] R. J. Trew, Proceedings of the IEEE 90, 1032 (2002). [37]D. Pavlidis, Technical Digest Internati onal Electron Devices Meeting 795 (2004). [38] E. T. Yu, G. J. Sullivan, P. M. Asbeck, C. D. Wang, D. Qiao, and S. Lau, Appl. Phys. Lett. 71, 2794 (1997).

PAGE 120

120 [39] O. Ambacher, J. Smart, J. R. Shealy, N. G. Weimann, K. Chu, M. Murphy, W. J. Schaff, L. F. Eastman, R. Dimitrov, L. Wittmer, M. St utzmann, W. Rieger, and J. Hilsenbeck, J. Appl. Phys. 85, 3222 (1999). [40] O. Ambacher, B. Foutz, J. Smart, J. R. Shealy, N. G. Weimann, K. Chu, M. Murphy, A. J Sierakowski, W. J. Schaff, L. F. Eastman, R. Dimitrov, A. Mitchell, and M. Stutzmann, J. Appl. Phys. 87, 334 (2000). [41] P. Bergveld, IEEE Trans. Biomed. Eng. 17, 70 (1970). [42] P. Bergveld, IEEE Trans. Biomed. Eng. 19, 342 (1972). [43] G. F. Blackburn, Biosensors: Fundamentals a nd Applications, A. P. F. Turner, G. Karube, G. S. Wilson, Oxford University Press, Oxford, 481 (1987). [44] D. G. Hafeman, J. W. Parce, and H. M. McConnell, Science 240, 1182 (1988). [45] J. Kong, N. R. Franklin, C. Zhou, M. G. Chapline, S. Peng, K. Cho, and H. Dai, Science 287, 622 (2000). [46] P. Qi, O. Vermesh, M. Grecu, A. Javery, Q. Wang, H. Dai, S. Peng, and K. J. Cho, Nano Lett. 3, 347 (2003). [47] A. Star, T. R. Han, V. Joshi, J.-C. P. Gabriel, and G. Grner, Adv. Mater. 16, 2049 (2004). [48] I. Sayago, E. Terrado, E. Lafuente, M. C. Horillo, W. K. Maser, A. M. Benito, R. Navarro, E. P. Urriolabeita, M. T. Martinez, and J. Gutierrez, Syn. Metals 148,15 (2005). [49] Y. Lu, J. Li, H. T. Ng, C. Binder, C. Partridge, and M. Meyyapan, Chem. Phys. Lett. 391, 344 (2004). [50] A. Kolmakov, D. O. Klenov, Y. Lilach, S. Stemmer, and M. Moskovits, Nano Lett. 5, 667 (2005). [51] E. Katz and I. Willner, Chem. Phy. Chem. 5, 1084 (2004). [52] R. J. Chen, S. Bangsaruntip, K. A. Drouva lakis, N. Wong Shi Kam, M. Shim, Y. Li, W. Kim, P. J. Utz, and H. Dai, Proc. Natl. Acad. Sci. USA 100, 4984 (2003). [53] A. Star, J.-C. P. Gabriel, K. Bradley, and G. Grner, Nano Lett. 3, 459-463 (2003). [54] C. Li, M. Curreli, H. Lin, B. Lei, F. N. Is hikawa, R. Datar, R. J. Cote, M. E. Thompson, C. Zhou, J. Am. Chem. Soc. 127, 12484 (2005). [55] B. R. Azamian, J. J. Davis, K. S. Coleman, C. B. Bagshaw, M. L. H. Green, J. Am. Chem. Soc. 124, 12664 (2002). [56] J. J. Davis, K. S. Coleman, B. R. Azamian, C. B. Bagshaw, and M. L. H. Green, Chem. Eur. J. 9, 3732 (2003).

PAGE 121

121 [57] F. Patolsky and C. M. Lieber, Materialstoday April, 20 (2005). [58] Y. Cui, Q. Wei, H. Park, and C. M. Lieber, Science 293, 1289 (2001). [59] J. Hahm and C. M. Lieber, Nano Lett 4, 51 (2004). [60] W. Wang, C. Chen, K. Lin, Y. Fang, and C. M. Lieber, Proc. Natl. Acad. Sci. USA 102, 3208 (2005). [61] F. Patolsky, G. Zheng, O. Hayd en, M. Lakadamyali, and C. M. Lieber, Proc. Natl. Acad. Sci. USA 101, 14017 (2004). [62] M. Curreli, C. Li, Y. Sun, B. Lei, M. A. Gundersen, M. E. Thompson, and C. Zhou, J. Am. Chem. Soc. 127, 6922 (2005). [63] Byoung Sam Kang, University of Florida dissertation, 2005. [64] See, for example, the revi ew by G. W. Hunter, P. G. Neudeck, R. S. Okojie, G. M. Beheim, V. Thomas, L. Chen, D. Lukco, C. C. Liu, B. Ward, and D. Makel, Proc.-Electrochem. Soc. 114, 93 (2002). [65] Q. Wan, Q. H. Li, Y. J. Chen, T. H. Wang, X. L. He, X. G. Gao, and J. P. Li, Appl. Phys. Lett. 84, 3654 (2004). [66] K. Keem, H. Kim, G. T. Kim, J. S. L ee, B. Min, K. Cho, M. Y. Sung, and S. Kim, Appl. Phys. Lett. 84, 4376 (2004). [67] M. H. Huang, S. Mao, H. Feick, H. Yan, Y. Wu, H. Kind, E. Weber, R. Russo, and P. Yang, Science 292, 1897 (2001). [68] Q. H. Li, Q. Wan, Y. X. Liang, and T. H. Wang, Appl. Phys. Lett. 84, 4556 (2004). [69] B. H. Kind, H. Yan, B. Messer, M. Law, and P. Yang, Adv. Mater. (Weinheim, Ger.) 14, 158 (2002). [70] C. H. Liu, W. C. Liu, F. C. K. Au, J. X. Ding, C. S. Lee, and S. T. Lee, Appl. Phys. Lett. 83, 3168 (2003). [71] W. I. Park, G. C. Yi, J. W. Kim, and S. M. Park, Appl. Phys. Lett. 82, 4358 (2003). [72] J. J. Wu and S. C. Liu, Adv. Mater. (Weinheim, Ger.) 14, 215 (2002). [73] H. T. Ng, J. Li, M. K. Smith, P. Nguygen, A. Cassell, J. Han, and M. Meyyappan, Science 300, 1249 (2003). [74] J. Q. Hu and Y. Bando, Appl. Phys. Lett. 82, 14019 (2003). [75] W. I. Park, G. C. Yi, M. Y. Kim, and S. J. Pennycook, Adv. Mater. (Weinheim, Ger.) 15, 526 (2003).

PAGE 122

122 [76] Y. W. Heo, V. Varadarjan, M. Kaufman, K. Kim, D. P. Norton, F. Ren, and P. H. Fleming, Appl. Phys. Lett. 81, 3046 (2002). [77] D. P. Norton, Y. W. Heo, M. P. Ivill, K. Ip S. J. Pearton, M. F. Chisholm, and T. Steiner, Mater. Today 7, 34 (2004). [78] P. J. Poole, J. Lefebvre, and J. Fraser, Appl. Phys. Lett. 83, 2055 (2003). [79] M. He, M. M. E. Fahmi, S. Noor Mohammad, R. N. Jacobs, L. Salamanca-Riba, F. Felt, M. Jah, A. Sharma, and D. Lakins, Appl. Phys. Lett. 82, 3749 (2003). [80] X. C. Wu, W. H. Song, W. D. Huang, M. H. Pu, B. Zhao, Y. P. Sun, and J. J. Du, Chem. Phys. Lett. 328, 5 (2000). [81] M. J. Zheng, L. D. Zhang, G. H. Li, X. Y. Zhang, and X. F. Wang, Appl. Phys. Lett. 79, 839 (2001). [82] S. C. Lyu, Y. Zhang, H. Ruh, H. J. L ee, H. W. Shim, E. K. Suh, and C. J. Lee, Chem. Phys. 327, 104 (1999); B. P. Zhang, N. T. Binh, Y. Segawa, K. Wakatsuki, and N. Usami, Appl. Phys. Lett. 83, 1635 (2003). [83] W. I. Park, Y. H. Jun, S. W. Jung, and G. Yi, Appl. Phys. Lett. 82, 964 (2003). [84] B. D. Yao, Y. F. Chan, and N. Wang, Appl. Phys. Lett. 81, 757 (2002). [85] J. Y. Lao, J. Y. Huang, D. Z. Wang, and Z. F. Ren, Nano Lett. 3, 235 (2003). [86] X. W. Sun, S. F. Yu, C. X. Xu, C. Yuen, B. J. Chen, and S. Li, J. Appl. Phys. 42, L1229 (2003). [87] Z. L. Wang, Mater. Today 7, 26 (2004). [88] A. A. Tomchenko, G. P. Harmer, B. T. Marquis, and J. W. Allen, Sens. Actuators B 93, 126 (2003). [89] H. L. Hartnagel, A. L. Dawar, A. K. Jain, and C. Jagadish, Semiconducting Transparent Thin Films (IOP, Bristol, 1995). [90] Y. W. Heo, D. P. Norton, L. C. Tien, Y. Kwon, B. S. Kang, F. Ren, S. J. Pearton, and J. R. LaRoche, Mater. Sci. Eng. R 47, 1 (2004). [91] See the databases at http://www.rebresearch.com/H2perm2.htm and http://www.rebresearch.com/H2sol2.htm [92] W. Eberhardt, F. Gr eunter, and E.W. Plummer, Phys. Rev. Lett. 46, 1085 (1981). [93] K. D. Mitzner, J. St ernhagen, and D. W. Galipeau, Sens. Actuators B 93, 92 (2003). [94] P. Mitra, A. P. Chatterjee, and H. S. Maiti, Mater. Lett. 35, 33 (1998).

PAGE 123

123 [95] J. F. Chang, H. H. Kuo, I. C. Leu, and M. H. Hon, Sens. Actuators B 84, 258 (1994). [96] G. W. Hunter, P. G. Neudeck, R. S. Okojie G. M. Beheim, V. Thomas, L. Chen, D. Lukco, C. C. Liu, B. Ward, and D. Makel, Proc. ECS Vol 01-02, 212 (2002). [97] A. Baranzahi, A. Lloyd Spetz, and I. Lundstrm, Appl. Phys. Lett. 67, 3203 (1995). [98] A. Lloyd Spetz, L. Unus, H. Svenningstorp P. Tobias, L.-G. Ekedahl, O. Larsson, A. Gras, S. Savage, C. Harris, P. Mrtensson, R. Wigren, P. Salomonsson, B. Hggendahl, P. Ljung, M. Mattsson, and I. Lundstrm, Physica Status Solidi A 185,15 (2001). [99] B. P. Luther, S. D. Wolter, and S. E. Mohney, Sensors Actuators B 56, 164 (1999). [100] J. Schalwig, G. Mller, O. Ambacher, and M. Stutzmann, Phys Status Solidi A 185, 39 (2001). [101] J. Schalwig, G. Mller, M. Eic khoff, O. Ambacher, and M. Stutzmann Sens Actuat B 87, 425 (2002). [102] J. Schalwig, G. Mller, U. Karrer, M. Eick hoff, O. Ambacher, M. Stutzmann, L. Grgens, and G. Dollinger, Appl. Phys. Lett. 80, 1222 (2002). [103] J. Kim, F. Ren, B. Gila, C. R. Abernathy, and S. J. Pearton, Appl. Phys. Lett. 82, 739 (2003). [104] H. T. Wang, B. S. Kang, F. Ren, R. C. Fitc h, J. Gillespie, N. Moser, G. Jessen, R. Dettmer, B.P. Gila, C. R. Abernathy, and S. J. Pearton, Appl. Phys. Lett. 87, 172105 (2005). [105] A. El. Kouche, J. Lin, M.E. Law, S. Ki m, B. S. Kim, F. Ren, and S.J. Pearton, Sensors and Actuators B: Chemical 105, 329 (2005). [106] B. S. Kang, R. Mehandru, S. Kim, F. Ren, R. Fitch, J. Gillespie, N. Moser, G. Jessen, T. Jenkins, R. Dettmer, D. Via, A. Crespo, B.P. Gila, C. R. Abernathy, and S.J. Pearton, Appl. Phys. Lett. 84, 4635 (2004). [107] B. S. Kang, F. Ren, B.P. Gila, C. R. Abernathy, and S. J. Pearton, Appl. Phys. Lett. 84, 1123 (2004). [108] J. Kim, B. Gila, G.Y. Chung, C. R. Abernathy, S. J. Pearton, and F. Ren, Solid-State Electron. 47, 1069 (2003). [109] J.-R. Huang, W.-C. Hsu, Y.-J. Chen, T.-B Wang, K.-W. Lin, H.-I. Chen, and W.-C. Liu, Sensors Actuators B 117, 151 (2006). [110] B. S. Kang, S. Kim, F. Ren, B. P. Gila, C. R. Abernathy, and S. J. Pearton, Sensors Actuators B 104, 232 (2005). [111] K. Matsuo, N. Negoro, J. Kota ni, T. Hashizume, and H. Hasegawa, Appl. Surf. Sci. 244, 273 (2005).

PAGE 124

124 [112] J. Song, W. Lu, J. S. Flynn, and G. R. Brandes, Solid-State Electronics 49, 1330 (2005). [113] M. Ali, V. Cimalla, V. Lebedev, H. Romanu s, V. Tilak, D. Merfeld, P. Sandvik, and O. Ambacher, Sensors Actuators B 113, 797 (2006). [114] O. Weidemann, M. Hermann, G. Steinhoff, H. Wingbrant, A. Lloyd Spetz, M. Stutzmann, and M. Eickhoff, Appl. Phys. Lett. 83, 773 (2003) [115] L. Voss, B. P. Gila, S. J. Pearton, H. T. Wang, and F. Ren J. Vac. Sci. Technol. B 23, 2373 (2005). [116] J. Song, W. Lu, J.S. Flynn, and G. R. Brandes, Appl. Phys. Lett. 87, 133501 (2005). [117] F. Yun, S. Chevtchenko, Y.-T. Moon, H. Morko, T. J. Fawcett, and J. T. Wolan, Appl. Phys. Lett. 87, 073507 (2005) [118] M. Eickhoff, R. Neuberger, G. Steinhoff, O. Ambacher, G. Muller, and M. Stutzmann, Phys. Status Solidi B 228, 519 (2001). [119] H. T. Wang, T. J. Anderson, F. Ren, C. Li, Z. N. Now, J. Lin, B. P. Gila, S. J. Pearton, A. Osinsky, and A. Dabiran, Appl. Phys. Lett. 89, 242111 (2006). [120] R. Khanna, S. J. Pearton, F. Ren, and I. I. Kravchenko, J. Vac. Sci. Technol. B 24, 4 (2006). [121] A. Renzoni, F. Zino, and E. Franchi, Environ. Res. 77, 68 (1998). [122] J. M. Llobet, G. Falco, C. Casas, A. Teixido, and J. L. Domingo, J. Agric. Food Chem. 51, 838 (2003). [123] A. H. Stern, Environ. Res. 98, 133 (2005). [124] F. Zahir, S. J. Rizwi, S. K. Haq, and R. H. Khan, Environ. Toxicol. Pharmacol. 20, 351 (2005). [125] J. Mutter, J. Naumann, R. Schneider, H. Walach, and B. Haley, Neuroendocrinol. Lett. 26, 439 (2005). [126] W. Zheng, M. Aschner, and J.-F. Ghersi-Egea, Toxicol. Appl. Pharmacol. 192, 1 (2003). [127] I. Hoyle and R. D. Handy, Aquat. Toxicol. 72, 147 (2005). [128] A. L. Burlingame, R. K. Boyd, and S. J. Gaskell, Anal. Chem. 68, 599 (1996). [129] K. W. Jackson and G. Chen, Anal. Chem. 68, 231 (1996). [130] J. L. Anderson, E. F. Bowden, and P. G. Pickup, Anal. Chem. 68, 379 (1996).

PAGE 125

125 [131] A. Caballero, R. Martnez, V. Lloveras, I. Ratera, J. VidalGancedo, K. Wurst, A. Trraga, P. Molina, and J. Veciana, J. Am. Chem. Soc. 127, 15666 (2005). [132] E. Coronado, J. R. Galn-Mascars, C. Ma rt-Gastaldo, E. Palomares, J. R. Durrant, R. Vilar, M. Gratzel, and Md. K. Nazeeruddin, J. Am. Chem. Soc. 127, 12351 (2005). [133] Y. K. Yang, K. J. Yook, and J. Tae, J. Am. Chem. Soc. 127, 16760 (2005). [134] M. Matsushita, M. M. Meij ler, P. Wirsching, R. A. Lerner, and K. D. Janda, Org. Lett. 7, 4943 (2005). [135] H. T. Wang, B. S. Kang, F. Ren, R. C. Fitch, J. K. Gillespie, N. Moser, G. Jessen, T. Jenkins, R. Dettmer, D. Via, A. Crespo, B. P. Gila, C. R. Abernathy, and S. J. Pearton, Appl. Phys. Lett. 87, 172105 (2005). [136] B. S. Kang, S. Kim, F. Ren, J. W. Johnson, R. Therrien, P. Rajagopal, J. Roberts, E. Piner, K. J. Linthicum, S. N .G. Chu, K. Baik, B. P. Gila, C. R. Abernathy, and S. J. Pearton, Appl. Phys. Lett. 85, 2962 (2004). [137] B. S. Kang, F. Ren, L. Wang, C. Lofton, Weihong Tan, S. J. Pearton, A. Dabiran, A. Osinsky, and P. P. Chow, Appl. Phys. Lett. 87, 023508 (2005). [138] B. S. Kang, F. Ren, M. C. Kang, C. Lofton Weihong Tan, S. J. Pearton, A. Dabiran, A. Osinsky, and P. P. Chow, Appl. Phys. Lett. 86, 173502 (2005). [139] B. S. Kang, J. J. Chen, F. Ren, S. J. Pearto n, J. W. Johnson, P. Rajagopal, J. C. Roberts, E. L. Piner, and K. J. Linthicum, Appl. Phys. Lett. 89, 122102 (2006). [140] F. Malem and D. Mandler, Anal. Chem. 65, 37 (1993). [141] C. C. Huang and H. T. Chang, Anal. Chem. 78, 8332 (2006). [142] M. H. Schoenfisch and J. E. Pemberton, J. Am. Chem. Soc. 120, 4502 (1998). [143] N. J. Brewer, T. T. Foster, G. J. Leggett, M. R. Alexande r, and E. McAlpine, J. Phys. Chem. B 108, 4723 (2004). [144] R. Lequin, Clin. Chem. 51, 2415 (2005). [145] V. S. Vaidya, V. Ramirez, T. Ichimu ra, N. A. Bobadilla, and J. V. Bonventre, Am. J. Physiol. Renal. Physiol. 290, F517 (2006). [146] Dario A. A. Vignali, Journal of Immunological Methods 243, 243 (2000). [147] J. Wang, Biosens. Bioelectron. 21, 1887 (2006) [148] C. F. Sanchez, C. J. McNe il, K. Rawson, and O. Nilsson, Anal. Chem. 76, 5649 (2004).

PAGE 126

126 [149] K. S. Hwang, J. H. Lee, J. Park, D. S. Yoon, J. H. Park, and T. S. Kim, Lab Chip. 4, 547 (2004). [150] K. W. Wee, G. Y. Kang, J. Park, J. Y. Kang, D. S. Yoon, J. H. Park, and T. S. Kim, Biosens. Bioelectron. 20, 1932 (2005). [151] R. J. Chen, S. Bangsaruntip, K. A. Drouvalak is, N. W. S. Kam, M. Shim, Y. Li, W. Kim, P. J. Utz, and H. Dai, Proc. Natl. Acad. Sci. USA 100, 4984 (2003). [152] C. Li, M. Curreli, H. Lin, B. Lei, F. N. Ishikawa, R. Datar, R. J. Cote, M. E. Thompson, and C. Zhou, J. Am. Chem. Soc. 127, 12484 (2005). [153] G. Zheng, F. Patolsky, Y. Cu i, W. U. Wang, and C. M. Lieber, Nature Biotechnology 23, 1294 (2005). [154] F. Patolsky, G. Zheng, and C. M. Lieber, Nanomedicine 1, 51 (2006). [155] F. Patolsky, G. Zheng, and C. M. Lieber, Nature Protocals 1, 1711 (2006). [156] F. Patolsky, B. P. Timko, G. Zheng, and C. M. Lieber, MRS Bulletin 32, 142 (2007). [157] H.-T. Wang, B. S. Kang, T. F. Chancellor, Jr., T. P. Lele, Y. Tseng, and F. Ren, S. J. Pearton, J. W. Johnson, P. Rajagopal, J. C. R oberts, E. L. Piner, and K. J. Linthicum, Appl. Phys. Lett. 91, 042114 (2007). [158] R. Thadhani, M. Pascua l, and J. V. Bonventre, N. Engl. J. Med. 334, 1448 (1996). [159] G. M. Chertow, E. M. Levy, K. E. Ha mmermeister, F. Grover, and J. Daley, Amer. J. Med. 104, 343 (1998). [160] J. V. Bonventre and J. M. Weinberg, J. Am. Soc. Nephrol. 14, 2199 (2003). [161] T. Ichimura, J. V. Bonventre, V. Bailly, H. Wei, C. A. Hession, R. L. Cate, and M. Sanicola, J. Biol. Chem. 273, 4135 (1998). [162] S. J. Pearton, D. P. Norton, K. Ip, Y. W. Heo, and T. Steiner, Prog. Mater. Sci. 50, 293 (2005). [163] D. C. Look, Mater. Sci. Eng. B 80, 383 (2001). [164] G.-C. Yi, C. Wang, and W. I. Park, Semicond. Sci. Technol. 20, S22 (2005). [165] Ya. I. Alivov, E. V. Kalinina, A. E. Ch erenkov, D. C. Look, B. M. Ataev, A. K. Omaev, M. V. Chukichev, and D. M. Bagnall, Appl. Phys. Lett. 83, 4719 (2003). [166] A. Osinsky, J. W. Dong, M. Z. Kauser, B. Hertog, A. M. Dabiran, P. P. Chow, S. J. Pearton, O. Lopatiuk, and L. Chernyak, Appl. Phys. Lett. 85, 4272 (2004). [167] B. W. Thomas and D. Walsh, Electron. Lett. 9,362 (1973).

PAGE 127

127 [168] T. Minami, M. Tanigawa, M. Yamanishi, and T. Kawamura, Jpn. J. Appl. Phys. 13, 1475 (1974). [169] T. Minami, S. Takata, M. Yamanishi, and T. Kawamura, Jpn. J. Appl. Phys. 18, 1617 (1979). [170] A. Shimizu, M. Kanbara, M. Hada, and M. Kasuga, Jpn. J. Appl. Phys. 17, 1435 (1978). [171] S. Takata, T. Minami, H. Nanto, and T. Kawamura, Phys. Stat. Solidi A 65, K83 (1981). [172] Y. I. Alivov, J. E. Van Nostrand, D. C. Look, M. V. Chukichev, and B. M. Ataev, Appl. Phys. Lett. 83, 2943 (2003). [173] H. Hosono, H. Ohta, K. Ha yashi, M. Orita, and M. Hirano, J. Cryst. Growth 469, 237 (2001). [174] D. K. Hwang, S.H. Kang, J. H. Lim, E. J. Yang, J. Y. Oh, J. H.Yang, and S. J. Park, Appl. Phys. Lett. 86, 222101 (2005). [175] H. S. Yang, S. Y. Han, Y. W. Heo, K. H. Baik, D. P. Nort on, S. J. Pearton, F. Ren, A. Osinsky, J. W. Dong, B. Hertog, A. M. Dabiran, P. P. Chow, L. Chernyak, T. Steiner, C. J. Kao, and G. C. Chi, Jpn. J. Appl. Phys 44, 7296 (2005). [176] A. Tsukazaki, A Ohtomo, T. Onuma, M. Ohta ni, T. Makino, M. Sumiya K. Ohtani, S. F. Chichibu, S. Fuke, Y. Segawa, H. Ohno, H. Koinuma, and M. Kawasaki, Nature Mater. 4, 42 (2005). [177] A. Tsukazaki, M. Kubota, A. Ohtomo, T. Onum a, K. Ohtani, H. Ohno, S. F. Chichibu, and M. Kawasaki, Jpn. J. Appl. Phys. 21, L643 (2005). [178] K. K. Kim, H. S. Kim, D. K. Hwang, J. H. Lim, and S. J. Park, Appl. Phys. Lett. 83, 63 (2003). [179] J. H. Lim, C. K. Kang, K. K. Kim, I. K. Park, D. K. Hwang, and S. J. Park, Adv. Mater. 18, 2720 (2006). [180] R. Konenkamp, R. C. Word, and C. Schlegel, Appl. Phys. Lett. 85, 6004 (2006). [181] Y. Du, S. Han, W. Jin, C. Zhou, and A. F. J. Levi, Appl. Phys. Lett. 83, 996 (2003). [182] S. O. Kucheyev, J. S. Williams, C. Jagadish J. Zou, C. Evans, A. J. Nelson, and A. V. Hamza, Phys. Rev. B 67, 094115 (2003). [183] S. O. Kucheyev, P. N. K. Deenapanray, C. Jagadish, J. S. Williams, Mitsuaki Yano, Kazuto Koike, Shigehiko Sasa, Masa taka Inoue, and Ken-ichi Ogata, Appl. Phys. Lett. 81, 3350 (2002). [184] S. O. Kucheyev, C. Jagadish J. S. Williams, P. N. K. Deen apanray, M. Yano, K. Koike, S. Sasa, M. Inoue, and K.-I. Ogata, J. Appl. Phys. 93, 2972 (2003).

PAGE 128

128 [185] V. A. Coleman, H. H. Tan, C. Jagadish, S. O. Kucheyev, and J. Zou, Appl. Phys. Lett. 87, 231912 (2005) [186] E. Alves, E. Rita, U. Wahl, J. G. Corre ia, T. Monteiro, J. Soares, and C. Boemare, Nucl. Instrum. Meth. Phys. Res. B 206, 1047 (2003). [187] T. S. Jeong, M. S. Han, C. J. Youn, and Y. S. Park, J. Appl. Phys. 96, 175 (2004). [188] F. Reus, C. Kirchner, Th. Gruber, R. Kli ng, S. Maschek, W. Limmer, A. Waag, and P. Ziemann, J. Appl. Phys. 95, 3385 (2004). [189] K. Lorenz, E. Alves, E. Wendler, O. Bilani, W. Wesch, and M.Hayes, Appl. Phys. Lett. 87, 191904 (2005). [190] G. Braunstein, A. Muraviev, H. Saxe na, N. Dhere,V.Richter, and R. Kalish, Appl. Phys. Lett. 87, 192103 (2005). [191] H. S. Yang, D. P. Norton, S. J. Pearton, and F. Ren, Appl. Phys. Lett. 87, 212106 (2005). [192] G. D. Mahan, L. M. Levinson, and H. R. Philipp, J. Appl. Phys. 50,2799 (1979). [193] G. E. Pike, S. R. Kurtz, P. L. G ourley, H. R. Philipp, and L. M. Levinson, J. Appl. Phys. 57, 5512 (1985). [194] T. M. Barnes, K. Olson, and C. A. Wolden, Appl. Phys. Lett. 86, 112112 (2005). [195] See, for exampl e, R. E. Williams, GaAs Processing Techniques, Artech House, Dedham, MA (1984). [196] See, for example, L. F. Eastman and U. K. Mishra, IEEE Spectrum Magazine 39(5), 28 (2002). [197] H. Xing, P. Chavarkar, S. Keller S. P. DenBaars, and U. K. Mishra, IEEE Electron. Dev. Lett. 24, 141 (2003). [198] S. Rajan, H. Xing, D. Jena, S. P. DenBaars, and U. K. Mishra, Appl. Phys. Lett. 84, 1591 (2004). [199] J. Sun, H. Fatima, A. Koudymov, A. Chitnis X. Hu, H.-M. Wang, J. Zhang, G. Simin, J. Yang, and M. A. Khan, IEEE Electron Dev. Lett. 24, 375 (2003). [200] A. P .Zhang, L. B. Rowland, E. B. Kamins ky, J. W. Kretchmer, R. A. Beaupre, J. L. Garrett J. B. Tucker, B. J. Edward, J. Foppes, and A. F. Allen, Solid-State Electron 47, 821 (2003). [201] A. P. Zhang, L. B. Rowla nd, E. B. Kaminsky, V. Tilak, J. C. Grande, J. Teetsov, A. Vertiatchikh, and L. F. Eastman, J. Electron. Mater. 32, 388 (2003). [202] A. Z. Wang and W. A. Anderson, Appl. Phys. Lett. 66, 1963 (1995).

PAGE 129

129 [203] Z. Q. Shi and W. A. Anderson, J. Vac. Sci. Technol. A 11, 985 (1993). [204] Z. Q. Shi and W. A. Anderson, J. Appl. Phys. 72, 3803 (1992). [205] R. T. Tung, J. Appl. Phys. 73, 7993 (1993). [206] D. K. Schroder, Semiconductor Material and Device Characterization Wiley and Sons, New York (1990). [207] D. S. Cammack, S. M. McGregor, J. J. McCh esney, S. A. Clark, P. R. Dunstan, S. R. Burgess, S. P. Wilks, F. Peiro, J. C. Ferrer, A. Cornet, J. R. Morante, A. Kestle, D. I. Westwood, and M. Elliott, Appl. Surface Science 123, 501 (1998). [208] J. E. Siewenie and L. He, J. Vac. Sci. Technol. A 17, 1799 (1999). [209] B. J. Baliga, Power Semiconductor Devices PWS, Boston (1996).

PAGE 130

130 BIOGRAPHICAL SKETCH Hung-Ta Wang was born in Nantou, Taiwan, in 1977. He received bo th the bachelor and the master degrees in the Department of Chemical Engineering of National Cheng Kung University, Tainan, Taiwan, in 1999 and 2001 resp ectively. He attended Taiwan Army for the mandatory training and service, and served as vice company commander with the position of 1st lieutenant officer from 2001 to 2003. After military service, he worked for Taiwan Semiconductor Manufacturing Company (TSMC) as a process integration engineer monitoring front-end process until 2004. In 2004 Fall, he was enrolled in the Ph.D. program of the Department of Chemical Engineering of the University of Florida. He was under the guidance of Professor Fan Ren studying wide bandgap semiconductor chemical/bio sensors, light emitting diodes, as well as high speed devices. He earned his Doctor of Philosophy degree from the Chemical Engineering Department of the University of Florida in May 2008 with 2 filed pa tents, 32 SCI journal publications in highly recognized journals ( Applied Physics Letters Nanotechnology, Journal of Electronic Materials, Applied Surface Science Electrochemical and So lid-State Letters etc.), and 8 international conference oral presentations.